RIO»!
‘ . é .. 5
5.15.1.9!!! n» ”trig: ._
2.14%: , r

 

 

21’“:
3.11:“... I
:3 .L .ndJJ.
4.1.... x:

in. a.

it?!-

J . I
.. :39.»
.53.. vs... #i
a. 2!... a».
, , ‘ .11.... an... 5.
‘ hr ‘,
:3... x.
‘ .211
N... 1
K i 31....
:3
3
. P.

5.9.?
. «
EELr a: w.
61.... .. . v -.
Edwas .- hut“: \ x ha.

.3 5...".
6...... i.
15.3.:

.2 I ,

ax?

, .335.

.
its r
, :53. .
.V .fiy) .a . A . $5.

LU; .m.;.;...,,.a..amw wflfih . .muaau
9. s a 1:..sz Lg
. .b. 2:!

1?. qup' '1. s

I! I'.‘ .m-’ up}! '1‘ b

..:n§4.fi\.5..x\flufin.flnii Ir IA“:

numb”: Gsflvudv . .5. A ..
2L4:

ar:1¢h:~.t.n§ V

an)“.

..
. 15.1. . Ali‘m

. . ‘
Slag" \ :13.

film“)! .. . n
3......
a... zmuflrzdm
I. f.-
51351: 3 It... .(i....z.:.li.2n.mm.lxu u
1 I“! ‘ 5‘... I "av "‘l 2 n l
.« “"‘. s I? .‘£.4.-’ ‘2‘
.E a 2...; _

3...)?!“ r.
i. a de?

:1 t...»
. :4!
DJ“
i1). V
.1: 21:35... I
27:: 3.3.6)...

32.5.3.1

A z

n .53..” .

x £9

3. l1.,..:
£3.

 

 

79.8813

t) r
.7
\f-

This is to certify that the :
dissertation entitled

DYNAMIC REGULATION OF MIXED LINEAGE KINASE 3 BY
c-JUN N-TERMINAL KlNASE-MEDIATED
PHOSPHORYLATION

presented by

KAREN ALEXANDRA SCHACHTER

— -.-.--.-- -

has been accepted towards fulfillment
of the requirements for the

PhD degree in Biochemistry 8. Molecular
Biology

 

 

Major Professor’s Signature
55/5/54:

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

LiBRAR‘
Michigan State
University

—..——

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/CIRC/DateDue.indd-p.1

 

 

 

DYNAMIC REGULATION OF MIXED-LINEAGE KINASE 3 BY c-JUN N-
TERMINAL KINASE-MEDIATED PHOSPHORYLATION

By

Karen Alexandra Schachter

A DISSERTATION
Submitted to
Michigan State University
In partial mlfillment of the requirements
For the degree of
DOCTOR OF PHILOSOPHY

Department of Biochemistry and Molecular Biology

2006

ABSTRACT

DYNAMIC REGULATION OF MIXED-LINEAGE KINASE 3 BY c-JUN N-
TERMINAL KINASE-MEDIATED PHOSPHORYLATION

By

Karen Alexandra Schachter

Mixed-Lineage Kinase 3 (MLK3) is a widely expressed mammalian
serine/threonine protein kinase. MLK3 functions as a Mitogen Activated Protein Kinase
(MAPK) Kinase Kinase (MAPKKK) to activate the c-Jun N-terminal kinase (JN K)
pathway. In addition to its kinase domain, MLK3 contains a Src-homology 3 (SH3)
domain, a leucine zipper domain, a Cdc42/Rae interactive binding (CRIB) motif, and a
C-terminal region which is rich in proline, serine and threonine residues.

The small GTPase Cdc42 in its activated state binds to MLK3 through the CRIB
motif. As a result of this interaction, the catalytic activity of MLK3 increases. Several
lines of evidence suggest that Cdc42 induces dimerization of MLK3 and trans-
autophosphorylation of MLK3 in its activation loop sites, and targets MLK3 to the
plasma membrane. Additionally, Cdc42 alters the in vivo phosphorylation pattern of
MLK3. Mass spectrometry coupled with phosphopeptide mapping was used to identify
11 phosphorylation sites on MLK3. Six of the identified sites correspond to
phosphorylated serines followed by proline residues, thus conforming to the consensus
for phosphorylation by proline-directed kinases. The work described in this dissertation
examines the role of proline-directed phosphorylation of MLK3.

The group of proline-directed kinases comprises the MAPKs, the cyclin-

' dependent kinases (Cdks) and Glycogen Synthase Kinase 3 (GSK-3). Since MLK3 is an

upstream activator of the MAPKs, it is possible that MAPKs feedback-phosphorylate
MLK3 on the proline-directed sites. Pharmacological inhibition of IN K, but not other
MAPK pathways, decreases the in viva phosphorylation of MLK3 on proline-directed
sites, suggesting that JNK phosphorylates MLK3. In fact, an in vitro phosphorylation
assay using recombinant, active JNK demonstrates that MLK3 is a bonafide substrate of
IN K.

Interestingly, the degree of proline-directed phosphorylation of MLK3 correlates
with its levels in cellular lysates. Inactivation of JNK in cells, by using dominant
negative versions of MKK4 and MKK7, the upstream activators of JNK, results in a
reduction on the levels of MLK3 in cellular lysates. On the other hand, overexpression of
a constitutively active form of JNK in cells results in an increase on the levels of MLK3
in cellular lysates. These results were corroborated using a pharmacological inhibitor of
IN K, and extended to demonstrate that inhibition of JNK allows (a) phosphatase(s) to
dephosphorylate MLK3; the dephosphorylated MLK3 is inactive, and redistributes to a
Triton-X100 insoluble fraction. Since long term JNK inhibition does not result in
degradation of MLK3 in this experimental system, and given that this redistribution
process is reversible, a possible mechanism to modulate MLK3 signaling in cells could
rely on sequestering MLK3 in a “stand-by” mode in a dephosphorylated, Triton-X100

resistant compartment.

ACKNOWLEDGEMENTS

“Looking back, I realize that I have been favored extraordinarily by external
circumstances, the proper place at the proper time in terms of my PhD thesis, my first
employment in the USA (...) and in particular having had incredibly brilliant
coworkers. "

Richard Ernst

Working towards a PhD has been more than just running experiments at the bench
on my own; I have had the input and help from various people and have interacted with
many more, and they all deserve a big Thank You.

First of all, I will be forever thankful of my thesis advisor, Dr. Kathy Gallo, who
taught me as much about science as about being a scientist. Not only did I learn about
hypothesis-driven experimentation, interaction and communication with other scientists,
and mentoring students; most importantly, I learnt about being a woman in science, with
all the joys, difficulties and compromises that it takes.

I would also like to thank the members of my committee, Dr. David Amosti, Dr.
David DeWitt, Dr. Lee Kroos and Dr. Katherine Osteryoung for their input, comments
and suggestions that improved my work. I want to thank Dr. David Amosti and Dr.
Susan Conrad for providing recommendation letters on my behalf.

To the Gallo lab: Thank you!! You are a fun bunch to work with, and I appreciate
all the help and support during all these years. Special thanks to Dr. Hua Zhang, Dr. Yan

Du and Geou-Yarh (Stancy) Liou for their good vibes and thoughtful discussions.

iv

I would also like to thank the people that work in the Fourth floor at the BPS
building, for being a continuous source of reagents, equipment and advice. Thanks go to
the staff of the Physiology Department for embracing me as one of their own and helping
me with all sorts of administrative issues.

Finally, I want to thank my family, my mother and father who have and will
always cheer for me wherever I am, my sister that will become an accomplished PhD

herself, and the love of my life, Dr. Soren Ottosen, for all their care and support.

TABLE OF CONTENTS

List of Tables ...................................................................................
List of Figures .................................................................................
Key to abbreviations .........................................................................

I. Literature Review ...........................................................................

1. Overview of protein kinase classification, structure and function ..........
2. Mitogen Activated Protein Kinase pathways ..................................

2.1 The MAPK pathways ...................................................
2.11 The ERK pathway .............................................
2.12 The p38 pathway ..............................................
2.13 The JNK pathway .............................................
2.14 The ERK3 and ERKS pathways .............................
2.2 Specificity on the MAPK pathways ...................................

3. Mixed Lineage Kinases ............................................................

3.1 The MLK family members .............................................
3.11 MLK1-4 .......................................................
3.12 DLK and LZK .................................................
3.13 ZAK ............................................................
3.14 MLK orthologs ................................................
3.2 MLKs signaling activities .............................................
3.21 Activation of the IN K pathway ............................
3.2.2 Activation of the p38 pathway ............................
3.2.3 Activation of the ERK pathway ............................
3.2.4 Activation of NF-KB .........................................
3.2.5 Other MLK substrates .......................................
3.3 Regulation of MLK5 activities .........................................
3.3.1 Regulation of MLK5 by extracellular signals ............
3.3.2 Regulation of MLKs by autoinhibition ....................
3.3.3 Regulation of MLK3 by dimerization .....................
3.3.4 Regulation of MLK5 by small GTPases ..................
3.3.5 Regulation of MLK5 by scaffold proteins ................
3.3.6 Regulation of MLK8 by chaperone proteins ..............
3.3.7 Regulation of MLK3 by phosphorylation .................
3.3.8 Regulation of MLK3 by subcellular localization. . . . .
3.4 The physiological roles of the MLKs ................................
3.4.1 The Drosophila MLK, Slpr .................................

vi

Page
ix

x

xii

1

1

5

6

6

8

9

11
12
14
15
15
17
17
18
18
18
19
20
21
21
23
23
28
29
3O
32
35
36
39
4O
4O

3.4.2 Mammalian MLKs ........................................... 41

4. Objective of the Thesis ................................................... 45
II. Phosphorylation of MLK3 by proline-directed kinases ............................... 47
1. Abstract .............................................................................. 47
2. Introduction ........................................................................ 48
3. Materials and Methods ............................................................ 50
3.1 Reagents and Antibodies ................................................ 50
3.2 Construction of mammalian expression vectors and site-directed 50
mutagenesis ...............................................................
3.3 Cell culture and transfections .......................................... 51
3.4 Cell lysis and immunoprecipitations .................................. 52
3.5 Gel electrophoresis and western blot analysis ....................... 52
3.6 Pro-Q Diamond and Silver Staining .................................. 53
3.7 In vivo labeling and phosphopeptide mapping ....................... 53
3.8 In vitro phosphorylation of MLK3 .................................... 54
3.9 Northern blot analysis ................................................... 55
4. Results ............................................................................... 56
4.1 JNK is required for in viva phosphorylation of MLK3 ............. 56
4.2 Manipulation of IN K activity alters the phosphorylation of 65
MLK3 .....................................................................
4.3 Feedback phosphorylation regulates MLK3 protein levels in 69
cellular lysates ............................................................
4.4 Inhibition of JNK reduces the levels of activated MLK3 in 70
cellular lysates ............................................................
4.5 Inhibition of other proline-directed kinase pathways has no 76
impact on MLK3 levels .................................................
4.6 Dephosphorylation of MLK3 reduces its levels in cellular 76
lysates .....................................................................
4.7 Dephosphorylation of MLK3 upon JN K inhibition does not 78
result in its degradation .................................................
5. Discussion ........................................................................... 85
111. Feedback phosphorylation of MLK3 by JNK regulates its distribution to 89
Triton-insoluble fractions .................................................................
1. Abstract .............................................................................. 89
2. Introduction ........................................................................ 9O
3. Materials and Methods ............................................................ 92
3.1 Reagents and Antibodies ................................................ 92
3.2 Cell culture ............................................................... 93
3.3 Cell lysis and western blotting ......................................... 93
3.4 Preparation of detergent-resistant membrane fractions by 93
sucrose-density gradient ................................................
3.5 Immunofluorescence Confocal Microscopy ......................... 94
4. Results ............................................................................. 96

4.1 JNK inhibition redistributes MLK3 to a triton-insoluble fraction 96

vii

4.2 Distribution of MLK3 between Triton-soluble and triton- 98

insoluble fractions is reversible ........................................
4.3 Triton-insoluble MLK3 is not associated with lipid rafts .......... 102
4.4 Inhibition of JNK results in the localization of MLK3 to 104
punctate, vesicle-like structures ........................................
5. Discussion ........................................................................... 121
IV. Concluding Remarks ...................................................................... 125
V. Appendix A: Determination of the in vitro kinase activity of various MLK3 129
mutants .......................................................................................
Generation of phosphorylation site mutants ....................................... 129
Effect of mutagenesis on MLK3 in vitro kinase activity ........................ 131
1. Mutagenesis of the glycine-rich region ................................ 131
2. Mutagenesis of the Cdc42 inducible sites .............................. 135
VI. References ................................................................................... 139

viii

LIST OF TABLES

Page
III. Feedback phosphorylation of MLK3 by JN K regulates its distribution to triton-
insoluble fractions
Table 1. Distribution of organelle markers in crude Triton X-100 fractionation of
MCF-7/iFlag-MLK3 cells ............................................................ 120
V. Appendix
Table 1. Oligonucleotides used for Site-directed mutagenesis .............................. 130

ix

LIST OF FIGURES
Page
I. Literature Review

Figure 1. Crystal structure of the kinase domain of B-Raf in complex with a small 3

molecule inhibitor, BAY 93-4006 ................................................
Figure 2. Mammalian MAPK signaling cascades .......................................... 7
Figure 3. Regulation of ERK activation by the scaffold KSR ........................... 13
Figure 4. Block diagram structure of human Mixed Lineage Kinases .................. 16
Figure 5. Regulation of MLK3 by extracellular signals ................................... 24

Figure 6. Alignment of the sequences of the activation loops of the MLK subfamily 37
members ..............................................................................

II. Phosphorylation of Mixed Lineage Kinase 3 by proline-directed kinases

Figure 1. Schematic representation of MLK3 .............................................. 57
Figure 2. Phosphorylation levels of MLK3 ................................................. 59
Figure 3. Effect of MAPK pathway inhibitors on in vivo phosphorylation of MLK3. 61

Figure 4. Phosphotryptic peptide maps of MLK3 from cells metabolically labeled 62

in the presence of MAPK inhibitors .............................................
Figure 5. Phosphopeptide mapping of MLK3 6A ......................................... 64
Figure 6. JNK phosphorylation of MLK3 .................................................. 66
Figure 7. Regulation of the levels of MLK3 variants in cellular lysates in response 71

to modulation of JNK signaling ...................................................
Figure 8. Electrophoretic mobility of activated MLK3, and impact of 72

pharmacological inhibition of JNK on MLK3 levels in cellular lysates. .

Figure 9. Effect of SP600125 on inducibly expressed F lag-ER and on Flag-MLK3 75
mRNA levels ........................................................................

Figure 10. Effect of proline-directed kinase inhibitors on MLK3 levels in cellular 77
lysates and activation loop phosphorylation ...................................

Figure 11. Impact of phosphatase inhibition on MLK3 levels in cellular lysates ...... 79
Figure 12. Effect of MGl32 on the protein levels of MLK3 on cellular lysates ....... 81
Figure 13. Ubiquitination of MLK3 on HEK 293 cells .................................... 82

111. Feedback phosphorylation of MLK3 by JNK regulates its distribution to triton-
insoluble fractions

Figure 1. Redistribution of MLK3 to detergent insoluble fractions ..................... 97
Figure 2. Effect of long-term JNK inhibition on the distribution of MLK3 in 100
Triton X-100 ........................................................................
Figure 3. Effect of SP600125 removal on MLK3 distribution ........................... 101
Figure 4. Distribution of MLK3 in a sucrose-density gradient ........................... 103
Figure 5. Subcellular localization of MLK3 in MCF-7/iF1ag-MLK3 cells ............. 105
Figure 6. Costaining of MLK3 and B-tubulin in MCF-7/iFlag-MLK3 cells ............ 107
Figure 7. Costaining of MLK3 and vinculin on MCF-7/iFlag-MLK3 cells ............ 109
Figure 8. Costaining of MLK3 and GM130 on MCF-7/iFlag-MLK3 cells ............. 111
Figure 9. Costaining of MLK3 and mitochondria on MCF-7/iFlag-MLK3 cells ...... 114
Figure 10. Costaining of MLK3 and endocytic vesicles on MCF-7/iFlag-MLK3 115
cells ..................................................................................
Figure 11. Costaining of MLK3 and clathrin on MCF-7/iFlag-MLK3 cells ............ 118
Appendix
Figure 1. In vitro kinase assay of MLK3 mutants in the glycine-rich region ........... 133
Figure 2. In vitro kinase assay of the Cdc42-inducible sites mutants of MLK3 ........ 137

xi

AGC
AP- 1
ASK- 1
ATF-2
Bsk
CAMK
CDK
CMBG
CNK 1/2
CREB
CRIB
CKl
COX4
DLK
DMEM
DRMs
Dpp
EEAl
EGF
ERK

GEF

KEY TO ABBREVIATIONS

containing Protein Kinase A, Protein Kinase G, Protein Kinase C families
Activating Protein 1

Apoptosis Signal-regulating Kinase 1
Activating Transcription Factor 2

Basket

Calcium/calmodulin protein Kinase

Cyclin Dependent protein Kinase

containing CDK, MAPK, GSK3, CLK families
Connector Enhancer of KSR 1/2
c-AMP-Response Element Binding protein
Cdc42/Rac interactive binding

Casein Kinase 1

Cytochrome Oxidase 4

Dual leucine zipper protein kinase

Dulbecco’s Modified Eagle’s Medium
Detergent Resistant Membranes
Decapentaplegic

Early Endosome Antigen 1

Epidermal Growth Factor

Extracellular signal-Regulated Kinase

Guanine nucleotide Exchange Factor

xii

GFP
GSK
GST
HA
HEK
Hep
HPKl
Hsp
IKB
IKK
IL
IMP
JIP
JN K
JNKK
Kay
KIF3
KSR
Lamp 1
LC-MS/MS
LPS
LZK

MALDI-MS

Green Fluorescent Protein

Glycogen Synthase Kinase

Glutathion S Transferase
Hemagglutinin

Human Embryionic Kidney
Hemipterous

Hematopoietic Progenitor Kinase 1
Heat shock protein

inhibitor of NF-KB

IKB kinase

Interleukin

Impedes Mitogenic signal Progression
JNK interacting protein

c-Jun N-terminal Kinase

JNK Kinase

Kayak

Kinesin superfamily motor protein 3
Kinase Suppresor of Ras

Lysosomal associated membrane protein 1
Liquid Chromatography-Mass Spectrometry/Mass Spectrometry
Lipopolysaccharide

Leucine Zipper-bearing Kinase

Matrix-Assisted Laser Desorption/lonization- Mass Spectrometry

xiii

MAPK

MAPKAPKZ

MAPKKK

MEF
MEK
MEKK
MKK
MLK
Mnk
MP 1
Msn
NGF
NIMA
PAGE
PAK
PDGF
PI3K
PIP3
PKA
PKC
PMA
PMSF

POSH

Mitogen Activated Protein Kinase
MAPK activated Protein Kinase 2
MAPK kinase kinase

Mouse Embryonic Fibroblast
Mitogen-activated protein/Erk Kinase
MEK Kinase

MAPK Kinase

Mixed Lineage Kinase
MAPK-interacting kinase

MEK Partner 1

Misshapen

Nerve Growth Factor

Never In Mitosis A

Polyacrylamide Gel Electrophoresis
p21 Activated protein Kinase
Platelet-derived Growth Factor
Phosphoinositide-3 Kinase
Phosphatidylinositol 3,4,5 triphosphate
Protein Kinase A

Protein Kinase C

Phorbol Myristoyl Acetate
phenylmethylsulfonylfluoride

Plenty of SH3s

xiv

PP2A
PSD-95
SCG
RGC
RNAi
Rsk
SAM
SH2
SH3
STE
TAKl
TGF
TK
TKL
TLC

TLE

Tp12

ZAK

Protein Phosphatase 2A
Post-Synaptic Density protein 95
Superior Cervical Ganglion
Receptor Guanylate Cyclase
RNA intereference

Ribosomal S6-Kinase

Sterile Alpha Motif
Src-Homology 2

Src-Homology 3

homologs of yeast Sterile 7, 11, 20 kinases
TGF-B Activated protein Kinase 1
Transforming Growth Factor
Tyrosine Kinase

Tyrosine Kinase Like

Thin Layer Chromatography

Thin Layer Electrophoresis
Tumor Necrosis Factor

Tumor progression locus 2

leucine-Zipper and sterile-Alpha motif Kinase

XV

I. Literature Review

1. Overview of protein kinase classification, structure and function

Protein kinases are enzymes that catalyze the transfer of the ‘y-phosphate of ATP
to a substrate. A comprehensive analysis of human protein kinase genes reveals that 518
protein kinases are encoded in the human genome (Manning et al., 2002). The human
kinome is divided in nine groups: AGC (PKA, PKG and PKC families), CAMK
(calcium/calmodulin dependent protein kinases), CMGC (CDK, MAPK, QSKB and CLK
families), TK (tyrosine kinases), STE (homologs of yeast Sterile 7, 11, 20 kinases), CKl
(casein kinase 1), TKL (tyrosine kinase-like), RGC (receptor guanylate cyclase), and
atypical kinases (Manning et al., 2002). Mutations and deregulation of protein kinases
play underlying roles in human diseases. The availability of the complete list of human
protein kinases will help in the identification of genes involved in human diseases and in
the development of therapies to treat them.

Phosphorylation is an important posttranslational modification that regulates most
of the signal transduction in cells. Modification of substrate activity by phosphorylation
allows the modulation of cellular processes such as transcription, metabolism, cell cycle
progression, cytoskeleton rearrangements, differentiation, and apoptosis. Protein
phosphorylation also regulates key aspects of development, and the general physiological
homeostasis.

Protein kinases phosphorylate their substrates in a sequence-specific manner. In
this “consensus sequence” model, the kinase recognizes residues directly flanking the

phosphoacceptor site (Kemp & Pearson, 1990; Pearson & Kemp, 1991). This has

allowed in vitro studies using synthetic peptides containing the recognition motif for a
protein kinase. However, it seems unlikely that substrate specificity inside the cells
would rely exclusively on recognition motifs; several kinases share a similar consensus
sequence for phosphorylation, and in vitro studies using synthetic peptides do not always
reflect the phosphorylation kinetics observed for the full length protein substrate in viva.
In fact, docking interactions between the kinase and the substrate are usually required for
efficient substrate phosphorylation (reviewed in (Biondi & Nebreda, 2003)).

Most mammalian protein kinases phosphorylate hydroxyl amino acids, and are
therefore classified as either Ser/Thr kinases or Tyr kinases. A small group of kinases
can phosphorylate Ser, Thr and Tyr residues, and are hence called dual-specificity
kinases (Dhanasekaran & Premkumar Reddy, 1998; Lindberg et al., 1992). Based on
sequence similarity and structural information, Ser/Thr and Tyr kinases form a
superfamily that is distinct from other eukaryotic kinases, like inositol and inositol
phosphate kinases. Crystal structures of the kinase domains of several protein kinases
have been solved, and the structure is well conserved within the superfamily. Figure 1
shows the structure of the kinase domain of B-Raf in complex with a small molecule
inhibitor, BAY 43-9006 (Wan et al., 2004). Like other protein kinases, the Raf kinase
fold consists of a NHz-terminal small lobe and a COOH-terminal large lobe (Zoller et al.,
1991). In the interface between the two lobes Mg”, ATP and the protein substrate bind.
BAY 43—9006 is an ATP analogue, and therefore is located where ATP would normally

bind.

P-Ioop -.

     
 

DFG motif

activation x/
segment

Figure 1. Crystal structure of the kinase domain of B-Raf in complex with a small
molecule inhibitor, BAY 43-9006. The NHz-terminal lobe is in magenta, and the
COOH-terminal lobe is in marine. The P-loop is in orange. The DFG and APE motifs are
in yellow, the rest of the activation segment and the N—terminus are in red. Dashed lines
indicate a disordered region in the activation loop (Reprinted from Cell (116) 855—67,
Wan P.T.C. et al. Mechanism afActivatian of the RAF -ERK Signaling Pathway by
Oncogenic Mutations of B-RAF , Copyright 2004 with permission from Elsevier).

Images in this dissertation are presented in color.

The catalytic mechanism of phosphorylation involves direct transfer of the )4
phosphoryl group from ATP to the hydroxyl group in the protein substrate (reviewed in
(Kaiser, 1987)). A number of conserved residues or motifs are found in all protein
kinases and they are required for proper substrate orientation and for the stabilization of
ATP (Figure 1). The glycine loop or P-loop, whose consensus sequence is GXGXXG, is
found at the top of the clefi between lobes. The Gly residues of this loop coordinate the
phosphates of ATP via backbone interactions. The P-loop is very flexible in the absence
of ATP, and this allows small molecule inhibitors to bind. An invariant Lys also
participates in the orientation of the phosphate groups of ATP for catalysis, and a
conserved Glu stabilizes this Lys by ionic interactions. Mutation of this conserved Lys
renders protein kinases inactive. At the bottom of the cleft is located the catalytic loop.
A conserved Asp residue interacts with the nucleophilic hydroxyl group on the substrate,
and an Asn residue orients this Asp. Another Asp residue, located in the conserved motif
DF G, makes crucial contacts with the Mg2+ ions, which are coordinated with the
phosphate groups of ATP (Hanks et al., 1988).

Within physiological systems, most protein kinases are normally maintained in an
inactive state. One of the mechanisms by which protein kinases are “turned on” is
through phosphorylation of the activation loop. The activation loop or activation
segment, provides a stage for peptide binding (Huse & Kuriyan, 2002; Morgan & De
Bondt, 1994). The activation loop is flanked by the motifs DFG and APE, present in
nearly all protein kinases. Upon phosphorylation, the activation loop is stabilized in an
open and extended conformation that allows the binding of the substrate (Huse &

Kuriyan, 2002).

Most protein kinases contain non-catalytic domains that assist in the regulation of
the protein function, by binding activators or inhibitors, by controlling subcellular
localization, by mediating the interaction with other proteins, or by influencing protein
stability. These regulatory mechanisms are required to provide specificity in response to
different extracellular signals. In addition, protein kinases participate in signaling
networks, in which, as a result of the stimulation of a particular signaling pathway,
feedback mechanisms are initiated to regulate the activity of protein kinases in the same

or in different pathways.

2. Mitogen Activated Protein Kinase pathways

The Mitogen Activated Protein Kinases (MAPKs) are a family of protein kinases
that respond to extracellular stimuli, such as grth factors, cytokines, irradiation, high
osmolarity, etc. (reviewed in (Johnson & Lapadat, 2002; Raman & Cobb, 2003)), to
coordinate a variety of cellular processes. MAPKs are conserved fi'om yeast to humans.
The MAPK signaling pathways operate by sequential phosphorylation and activation: an
upstream MAPK kinase kinase (MAPKKK) phosphorylates the activation loop of the
MAPK kinase (MAPKK). MAPKKs are dual-specificity kinases; upon activation, they
phosphorylate the MAPKs on both the Thr and Tyr residues of a conserved TXY motif in
the MAPK activation loop. Although activated MAPKs have mainly been associated
with the phosphorylation and activation of transcription factors in the cell nucleus,
MAPKs also have cytosolic substrates. MAPKs are proline-directed kinases; they
phosphorylate Ser or Thr residues that are followed by a Pro residue in their substrates

(Aitken, 1999).

In multicellular organisms, there are three major MAPK cascades, each of which
includes multiple isoforrns (Figure 2): the extracellular-signal regulated kinase (ERK),
ERK1 and ERK2; four p38 kinases, p38a, p380, p387 and p386; and c-Jun N-tenninal
Kinase (JNK), JNK] , JNK2 and JNK3. A few other MAPKs have been discovered,

namely ERK3 and ERKS.

2.1 The MAPK pathways
2.1.1 The ERK pathway

The ERK pathway is described as the canonical signaling cascade that is activated
by growth factors. At the top of the pathway, the Raf family of MAPKKKs is activated
by their association with the small GTPase Ras (reviewed in (Pearson et al., 2001)).
Activated Raf phosphorylates the dual specificity kinases MEKl and MEK2 in their
activation loops. MEK] and MEK2 in turn, phosphorylate ERK] and ERKZ in the Thr
and Tyr residues of the TEY motif in their activation loop (Zheng & Guan, 1993).
Additional MAPKKKs that activate the ERK pathways have been described, such as
MAPK/extracellular—signal regulated kinase kinase kinase 1 (MEKKl) (Xu et al., 1996),
MEKK2 and MEKK3 (Blank et al., 1996), tumor progression locus-2 (Tp12) (Ceci et al.,
1997) and Mos (Nebreda & Hunt, 1993).

The ERK pathway is generally associated with proliferation, and deregulation of
the components of the pathway often results in cellular transformation. ERK is
hyperactivated in 30% of cancers (Allen et al., 2003). B-Raf is also mutated in human
cancers; the most common mutation is a substitution of Val 599 by Glu in the activation

loop of B-Raf, which generates a gain-of-function, activated kinase (Davies et al., 2002).

Stimulus Growth factors Stress, Cytokines

l l / \.

mks, .“ 111175:
MAPKKK Raf, B-Raf + _ _ _ MEKKs, MEKKS,
‘ others ‘ others "
MAPKK MKK 1/2 MKK 3/6 MKK 4/7 ‘
MAPK Erk 1/2 p38 JNK
Transcription Elk1 ATP-2 c-Jun
factor
_ Apoptosis Apoptosis
3'0'09163' Deg‘lggtm'ent Inflammation Inflammation
response Differentiation Growth Growth
Differentiation Differentiation

Figure 2. Mammalian MAPK signaling cascades. The three best-defined mammalian
MAPK pathways are the ERK, p38 and JNK pathways. The core module of the MAPK

pathway is composed of three kinases that are sequentially activated by phosphorylation.

Consequently, inhibitors of the ERK pathway are been tested in clinical trials as potential
anti-cancer agents.

The regulation of gene expression by ERK is accomplished at two levels. ERK
phosphorylates a number of protein kinases, including Ribosomal S6-kinase 1-3 (Rskl -3)
and MAPK-interacting kinase 1 (Mnkl) and Mnk2 (reviewed in (Pearson et al., 2001)).
Rsk proteins activate gene expression through phosphorylation of several transcriptional
regulators, such as c-AMP-response element binding protein (CREB), serum response
factor, and the coactivator CBP (Pearson et al., 2001). In addition, ERK proteins directly
phosphorylate transcription factors, such as the activating protein 1(AP-1) family
members c-Jun, c-Fos and activating transcription factor 2 (ATP-2), and members of the
ternary complex factors (TCFs) like Elk-l (reviewed in (Chen et al., 2001)).

Among the numerous cytosolic substrates for ERK are other signaling molecules,
like phosphodiesterase 4D3 (Hoffrnann et al., 1999) and cytoskeletal elements such as

neurofilament subunits (Veeranna et al., 1998).

2.1.2 The p38 pathway

The human p3 8a isoform was identified as a target for drugs that block tumor
necrosis factor alpha (TNF-oz) and interleukinl (IL-1) release from monocytes stimulated
with lipopolysaccharide (LPS) (Lee et al., 1994). Subsequentially, other isoforrns were
cloned by homology. Both p3 8a and p380 have splicing variants, making a total of six
different p38 MAPK isoforms. The p38 MAPKs are activated by phosphorylation on
their activation loops by MKK3 and MKK6 (Enslen et al., 1998). MAPKKKs that

activate the p38 pathways include transforming growth factor beta (TGF-B)-activated

kinase 1 (TAK1)(Moriguchietal., 1996), apoptosis signal-regulating kinase 1 (ASKl)
(Ichijo et al., 1997), and some Mixed Lineage Kinases (MLKs) (Gallo & Johnson, 2002),
among others.

Isoforrns of p38 differ in their expression patterns, extracellular activators and
cellular substrates; for instance p3 8a, but not p380, is activated by LPS (Nick et al.,
1999). The p38 MAPKs play a major role in immunity, regulating the expression of
several cytokines and activating the immune response (Crawley et al., 1997). Death
receptors, such as F as and TNF-a receptors, trigger the activation of the p38 pathways,
implicating p38 in apoptosis (Beyaert et al., 1996; Juo et al., 1997).

p38 phosphorylates a variety of different cellular substrates that include
phospholipase A2 (Kramer et al., 1996) and the microtubule associated protein Tau
(Reynolds et al., 1997). Analogous to ERK, p38 regulates gene expression directly, by
phosphorylating transcription factors like ATP -1 and -2, Ets-l, Myocyte Enhancer Factor
2A (MEF 2A) and p53 (reviewed in (Kyriakis & Avruch, 2001)), and by phosphorylating
other protein kinases involved in gene regulation like Msk1-2 and MAPK activated

protein kinase 2 (MAPKAPK 2) (reviewed in (Roux & Blenis, 2004)).

2.1.3 The JNK pathway

JNKs, or stress-activated protein kinases (SAPKs), as their name imply, are
activated by environmental stress, such as heat shock, UV and osmotic stress (reviewed
in (Ip & Davis, 1998; Weston & Davis, 2002)). Ten JNK isoforms, generated by

alternative splicing from the JnkI, Jnk2 and Jnk3 genes, exist. JNK] and JNK2 are

ubiquitously expressed, while JNK3 is restricted to brain, heart and testis (Gupta et al.,
1996)

Regulation of the JNK pathway is very complex, and many MAPKKK have been
implicated in its activation, including MEKK1—4 (Blank et al., 1996; Gerwins et al., 1997;
Xu et al., 1996), TAKl (Yamaguchi et al., 1995), ASKl (Ichijo et al., 1997), and the
MLKs (reviewed in (Gallo & Johnson, 2002)). Dual phosphorylation of the activation
loop of JNK is accomplished by the MAPKKs MKK4 and MKK7 (Lin et al., 1995;
Toumier et al., 1997; Toumier et al., 1999).

The best characterized substrate of JNK is the transcription factor c-Jun.
Activated JNK phosphorylates c-Jun on serine 63 and serine 73 of its transactivation
domain, increasing the half life of c-Jun by inhibiting its ubiquitination and degradation
through the 26S proteasome pathway (Fuchs et al., 1996). c-Jun is a component of the
AP-l transcription complex, an important regulator of gene expression. Cellular
responses to environmental stresses, which activate JNK, are controlled through AP-l-
mediated transcriptional regulation.

JNK has been implicated in promoting apoptosis. Mouse embryonic fibroblasts
derived from Jnk1'/‘Jnk2'/' knockout mice are resistant to UV-induced apoptosis
(Toumier et al., 2000). Apoptosis induced by nerve growth factor (NGF) withdrawal in
rat PC-12 cells is suppressed when the JNK pathway is inhibited (Le-Niculescu et al.,
1999). However, in pro-B lymphocytes, JNK promotes cell survival. In response to 1L3,
JNK phosphorylates the pro-apoptotic Bcl-2 family protein BAD, which is then

sequestered in the cytosol by 14-3-3, inhibiting apoptosis (Yu et al., 2004). Thus, it

10

seems likely that JNK activation modulates apoptotic responses in a cell type and

stimulus dependent manner.

2.1.4 The ERK3 and ERKS pathways

ERK3 displays high similarity in its kinase domain to ERKl/2; however, it has a
unique COOH-terminal extension and instead of the TEY motif in its activation loop,
ERK3 possesses a SEG motif, where the Ser residue is constitutively phosphorylated
(Cheng et al., 1996). ERK3 mRNA is widely expressed, and its expression is upregulated
during development. ERK3 is a highly unstable protein with a fast turnover rate
regulated through ubiquitin-mediated degradation (Coulombe et al., 2003). Interestingly,
ERK3 is stabilized and accumulates during the differentiation of PC12 cells into neuronal
cells, and of C2C12 cells into muscle cells (Coulombe et al., 2003), suggesting that
protein abundance, rather than activation loop phosphorylation, might regulate ERK3
activities. There is not much information on physiological substrates of ERK3; it does
not phosphorylate ERK1/2 substrates in vitro or in viva. Recently, MAPKAPKS has
been identified in a yeast-two-hybrid screen as an ERK3 interacting partner that can also
be phosphorylated by ERK3 (Setemes et al., 2004). The interaction between these
proteins seems to stabilize ERK3 and results in nuclear exclusion and accumulation in the
cytoplasm of cells. The biological function of this interaction remains to be elucidated.

ERKS, also called Big MAPK 1 (BMKl), is two times the size of ERK1/2, and
like ERK3, it has a unique COOH-terminal extension. ERKS follows the classical
MAPK activation cascade. Its activation loop segment contains the TEY motif that is

phosphorylated by an upstream MAPKK, MEK5 (Kato et al., 1997; Zhou et al., 1995).

11

At the MAPKKK level, MEKK2 and MEKK3 have been shown to activate ERK5 (Chao
et al., 1999; Sun et al., 2001). The repertoire of ERK5 substrates includes transcription
factors from the MEF 2 family, serum- and glucocorticoid-inducible kinase (SGK),
connexin 43 and the pro-apoptotic protein Bad (reviewed in (Hayashi & Lee, 2004)).
Like ERK1/2, ERK5 is activated by growth factors; the ERK5 knockout is lethal, and the
embryos die of cardiovascular defects. Studies in ERK5 conditional knockout mice
indicate that ERK5 is required for survival of endothelial cells and the maintenance of

vascular integrity (Hayashi et al., 2004).

2.2 Specificity on the MAPK pathways

MAPK cascades respond to extracellular stimuli in an organized manner; yet the
great variety of MAPKKKs and MAPKKs that can be utilized in response to a signal
raises the question of how specificity is achieved. Part of the answer comes from the
identification of scaffold proteins for each MAPK family. Scaffold proteins interact
through different domains with specific subsets of kinases, as well as with other
molecules (reviewed in (Dard & Peter, 2006; Morrison & Davis, 2003; Nicolas Dard,
2006; Raman & Cobb, 2003)). In addition to promoting kinase-substrate interactions,
scaffolds can spatially and temporally regulate the activation of MAPKs and can
modulate the kinetics of activation by recruiting phosphatases (reviewed in (Dard &
Peter, 2006)). For example, kinase suppressor of Ras (KSR) is a scaffold protein for the
ERK pathway (Figure 3) (Roy et al., 2002). KSR is constitutively associated with MEK

and 14-3-3 through phosphorylated Ser residues. In addition, KSR is associated with

12

 

Cytosolic
substrates

 

HQ

1

Nuclear
substrates

 

Plasma membrane

 

 

Figure 3. Regulation of ERK activation by the scaffold KSR. Cytosolic KSR is kept

inactive through its interaction with 14-3-3 proteins and Impedes Mitogenic signal
Propagation (IMP). Upon binding of EGF to its receptor, Ras is activated (Ras-GTP).
Ras—GTP induces dephosphorylation of KSR by PP2A, releasing 14-3—3 and translocating

KSR to the membrane; in addition, Ras-GTP also induces ubiquitination and degradation
of IMP. At the membrane, KSR facilitates the phosphorylation of MEK by Raf, and the
phosphorylation of ERK by MEK. Activated ERK translocates to the nucleus to activate

transcription factors, or phosphorylates cytosolic substrates (adapted from (Kolch,

2005)).

Impedes Mitogenic signal Propagation (IMP) and this interaction keeps KSR in a Triton-
insoluble compartment (Matheny et al., 2004). In response to cell stimulation, IMP is
degraded through ubiquitination-mediated degradation. KSR is dephosphorylated by the
protein phosphatase PP2A, and 14-3-3 is released. KSR translocates to the plasma
membrane, and brings MEK in contact with activated Raf. In addition, KSR provides a
docking stage for ERK, facilitating the sequential phosphorylation events required for
ERK activation (Ory & Mon'ison, 2004). Additional scaffold proteins that regulate the
ERK pathway include MEK partner 1 (MP1), B-arrestins, and connector enhancer of
KSR 1/2 (CNK1/2) (reviewed in (Kolch, 2005; Morrison & Davis, 2003)). Scaffold
proteins for the JNK and p38 pathways have also been described, and will be discussed in

detail later in the chapter.

In summary, the MAPKs are a complex family of protein kinases with many
different subcellular functions. Various upstream activators impinge in the different
pathways to regulate a different subset of biological responses. For the rest of the
chapter, detailed consideration will be given to one of the families of MAPKKKs, the

MLKs; their functions in the MAPK pathways; the regulation of their activities.

3. Mixed Lineage Kinases

Mixed Lineage Kinases are a family of serine/threonine kinases that constitutes
one group of MAPKKKs. The MLKs, which share sequence similarity with both Ser/T hr
kinases and Tyr kinases in their catalytic domains, are found in the TKL branch of the

human kinome. To date Tyr phosphorylation activity has not been demonstrated for the

14

MLKs. The best described function of the MLKs is as upstream activators of the JNK
and p38 pathways. MLKs have been implicated in neuronal apoptosis, and a pan MLK
inhibitor, CBP-1347 (Maroney et al., 1998), was in clinical trials for Parkinson’s disease
(Parkinson Study, 2004). A derivative of CBP-1347, CBP-11004 (Murakata et al., 2002),

is currently available for research studies.

3.1 The MLK family members
The MLK family can be divided into 3 subgroups (Figure 4): MLK1-4; the Dual
Leucine zipper bearing Kinases (DLKs); and Zipper sterile-a motif Kinase (ZAK)

(reviewed in (Gallo & Johnson, 2002)).

3.1.1 MLK1-4

The MLK subfamily includes MLKl (Dorow et al., 1993), MLK2 (also called
MST, (Dorow et al., 1995; Katoh et al., 1995)), MLK3 (also known as SPRK/PTKl,
(Ezoe et al., 1994; Gallo et al., 1994; Ing et al., 1994)) and the splice isoforms MLK4a
and MLK4B (Kvasha et al, from Genebank). MLK1-4 contain a NHz-terminal SH3
domain followed by the catalytic domain, a centrally located leucine zipper region, and a
Cdc42/Rac interactive binding (CRIB) motif. MLK1-4 share over 75% sequence identity
in their kinase domain, and 65% sequence identity from the SH3 through their CRIB
motif. The COOH-terminal regions of the MLKs diverge, but they are all proline-rich.

In spite of their similarities, MLK1-4 have different cellular expression patterns.
MLK] mRNA has been detected in epithelial cell lines of breast, colonic and esophageal

origin (Dorow et al., 1993). MLKl protein is detected in an immature B-cell line, RIN-

15

SH3 Kinase LZ CRIB

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

( MLK1 1055
...... W 9154

MLK MLK3 [GT 1" i 7 U E W

K MLK” LL [ 1:31 l 1036
DLK [DLK [ [ E] E] Elsa

.sz l r ‘11 n .1...

, SAM
ZAK ZAKa l 7 7 7 i solo

.ZAKB I_!J-

 

Figure 4. Block diagram structure of human Mixed Lineage Kinases. The relative

position of the SH3 domain, leucine zipper (LZ), CRIB motif, and sterile a motif (SAM)
is indicated (Adapted from (Gallo & Johnson, 2002))

SAH cells, but is absent in a more mature B-cell line, PIN-A12 cells, suggesting that
MLK1 might be regulated during B-cell development (Dorow et al., 1993). MLK2
mRNA is predominantly expressed in brain, testis and skeletal muscle, and at lower
levels in pancreas (Dorow et al., 1995; Katoh et al., 1995). MLK3 mRNA is widely
expressed in human tissues, with lower expression in brain and heart (Ezoe et al., 1994;
Gallo et al., 1994; Ing et al., 1994). There is currently no published information on the

expression pattern of MLK4.

3.1.2 DLK and LZK

The DLK subfamily includes DLK (also called ZPK/MUK, (Holzman et al.,
1994)) and Leucine-Zipper Kinase (LZK) (Sakuma et al., 1997). While DLKs lack an
SH3 domain and a CRIB motif, they do have two small leucine zipper motifs separated
by 31 amino acids, which follow the catalytic domain. The catalytic domains of DLK
and LZK share 87% sequence identity with each other, but only 35% sequence identity
with MLK1-4. DLK mRNA is highly expressed in the adult mouse brain, primarily in
the hippocampus, cerebral cortex and the Purkinje cells of the cerebellum (Blouin et al.,
1996; Holzman et al., 1994). LZK mRNA is widely expressed, with the highest

expression in the pancreas (Sakuma et al., 1997).

3.1.3 ZAK

The only member of the third MLK subfamily is ZAK (Gotoh et al., 2001;
Gross et al., 2002; Liu et al., 2000a). Two splicing forms of ZAK have been described,

ZAKa (also known as MRK) and ZAKB (also called MLTK/MLK7/ZAK). In addition to

17

the catalytic domain, ZAKa contains a leucine zipper region and a sterile-a motif (SAM).
Due to the differential splicing, ZAKB lacks the SAM. ZAK mRNA is widely expressed
in human tissues, with highest expression in heart and skeletal muscle (Gotoh et al.,

2001; Gross et al., 2002).

3.1.4 MLK orthologs

MLKs are absent from yeast, but MLK orthologs are found in Drasaphila, C.
elegans and Xenapus. The Drasaphila ortholog of the MLK subfamily is called Slipper
(Slpr), and is involved in regulating the process of dorsal closure in the fly embryo
(Stronach & Penimon, 2002). The expression of the Xenapus MLK ortholog, which,
among the human MLKs, is most similar to MLK2, varies during differentiation (Poitras
etal., 2003), and antisense inactivation of Xenapus MLK2 impairs normal tubule
development of the cement gland (Poitras et al., 2003). By analysis of the sequences
from the catalytic domain and zipper regions of DLK against the National Center for
Biotechnology Information (N CBI) protein database, both Drasaphila and C. elegans
have a DLK ortholog (Gallo & Johnson, 2002). C. elegans also has an ortholog of ZAK,
and another distantly related SH3-containing kinase whose catalytic domain is similar to

the MLKs.

3.2 MLKs signaling activities
3.2.1 Activation of the JNK pathway
Of the seven mammalian MLKs, all except MLK1 and MLK4 have been

reported to activate the JNK pathway when overexpressed in cells (Cuenda & Dorow,

18

1998; Fan et al., 1996; Gotoh et al., 2001; Liu et al., 2000a; Rana et al., 1996; Sakuma et
al., 1997). MLK2 and MLK3 phosphorylate the MAPKKs MKK4 and MKK7 in in vitro
kinase assays (Cuenda & Dorow, 1998; Rana et al., 1996), while DLK only
phosphorylates MKK7 in vitro (Merritt et al., 1999), suggesting that DLK may only
utilize MKK7 to transduce signals to JNK in viva. Overexpression of ZAK enhances the
catalytic activity of MKK7, and to a lesser degree, that of MKK4 (Gotoh et al., 2001). In
addition, a catalytically inactive form of MKK7, but not a catalytically inactive MKK4,
can block ZAK-mediated JNK activation, (Yang, 2002), suggesting that the relevant

substrate for ZAK in viva may be MKK7.

3.2.2 Activation of the p38 pathway

Concomitant to the activation of JNK, a modest activation of the p38 pathway,
through phosphorylation of MKK3 and MKK6, was initially observed for overexpressed
MLK3 (Tibbles et al., 1996), and later for DLK (Fan etal., 1996), MLK2 (Cuenda &
Dorow, 1998; Hirai et al., 1997), and ZAKa (Gotoh et al., 2001; Gross et al., 2002). The
scaffold protein JIP2 has been shown to mediate the MLK-induced p38 activation
(Buchsbaum et al., 2002; Schoorlemmer & Goldfarb, 2001). Endogenous MLK3 can
activate the p38 pathway in hepatoma cell lines upon TGF-B treatment (Kim et al., 2004),
although the slow kinetics of activation suggest that p38 activation by MLK3 might occur
indirectly, perhaps as a result of activating some other signaling pathway. Suppression of
MLK3 expression by RNAi reduces TNF-a-induced activation of p38 on CCD-18Co

human colon fibroblasts and Jurkat T lymphocytes (Chadee & Kyriakis, 2004).

19

3.2.3 Activation of the ERK pathway

ERK activation can be detected in cells overexpressing MLK2 (Hirai et al.,
1997; Nagata et al., 1998) and ZAK (Gotoh et al., 2001). The data regarding the
activation of the ERK pathway by ectopically expressed MLK3 are conflicting.
Overexpression of MLK3 can result in ERK activation (Gallo KA, unpublished
observations, and (Kim et al., 2004)). In contrast to these results, some data indicate that
overexpression of MLK3 blocks ERK activation in response to Phorbol Myristoyl
Acetate (PMA) and Platelet-Derived Growth Factor (PDGF) in mesangial cells
(Parameswaran et al., 2002a; Parameswaran et al., 2002b). Although MLK3 can
phosphorylate and activate MEKl in vitro and in viva, a report indicates that
overexpression of MLK3, which results in sustained JNK activation, blocks mitogenic
activation of ERK (Shen et al., 2003). The proposed mechanism suggests that negative
cross-talk between JNK and ERK exists, in which c-Jun mediated gene transcription is
required to somehow uncouple the MLK3-mediated MEKl activation from ERK
activation.

Studies using RNA interference (RNAi) to silence MLK3 expression in human
and murine cell lines had reveal a novel role of MLK3 in the activation of the ERK
pathway. The results indicate that MLK3 is required for the activation of ERK upon
various extracellular stimuli, and for the proliferation of human tumor cell lines (Chadee
& Kyriakis, 2004). MLK3 appears to play a scaffolding role, rather than a catalytic role
in mediating ERK activation. Specifically, MLK3 is proposed to form a complex with B-

Raf and Raf-1 that is required for efficient signaling to ERK.

20

3.2.4 Activation of NF-KB

NF-KB is a transcription factor involved in immune, inflammatory and anti-
apoptotic responses. NF-KB is normally sequestered in the cytosol through the
interaction with IKB. Phosphorylation of IKB by the Ser/Thr kinases IKB kinases (IKKs),
IKKa and IKKB, results in its proteasomal-mediated degradation, allowing NF-KB to
translocate to the nucleus and activate transcription (reviewed in (Israel, 2000)). In
J urkat cells, overexpression of MLK3 induces phosphorylation and activation of IKKa
and IKKB (Hehner et al., 2000). Catalytically inactive MLK3 blocks NF -KB dependent
transcription in response to CD3/CD28 T-cell costimulation. Other MLKs that have been

reported to activate NF -KB are ZAK (Liu et al., 2000a) and LZK (Masaki et al., 2003).

3.2.5 Other MLK substrates

Most MLK substrates are protein kinases whose activation loops are
phosphorylated by the MLKs. Recently, it has been shown that MLK3 can phosphorylate
golgin-l60 in vitro (Cha et al., 2004). Golgin-160 is a peripheral membrane protein
localized to the cytoplasmic face of the Golgi complex (Hicks & Machamer, 2005),
which upon genotoxic stress, is cleaved by caspases, resulting in the disassembly of the
Golgi complex during programmed cell death (Mancini et al., 2000). Coexpression of
MLK3 and golgin-l60 in CHO cells results in a decrease in the mobility of golgin-160 on
SDS-PAGE, suggesting that MLK3 can promote phosphorylation of golgin-160 in viva
(Cha et al., 2004). The function of this phosphorylation is not clear; overexpression of

MLK3 on HeLa cells that stably overexpress golgin-160 results in the caspase-induced

21

cleavage of a very small fraction of golgin-160, but its effect on the induction of
apoptosis was not tested.

Using a yeast-two-hybrid system, the nuclear receptor corepressor Alien was
identified as an interacting partner of MLK2 (Eckey et al., 2003). In the absence of
thyroid hormone, Alien is recruited by thyroid hormone receptor through its silencing
domain located in its COOH-terminus. Thyroid hormone receptor-corepressor complexes
repress target gene expression partly through association with histone deacetylases
(Eckey et al., 2003). MLK2 can phosphorylate Alien in vitro, and reporter-gene assays
indicated that coexpression of MLK2 with Alien results in increased transcriptional
repression when compared to Alien expressed alone (Eckey et al., 2003). The relevance
of the MLK2-Alien interaction as well as the effect of the phosphorylation of Alien by
MLK2 needs further investigation.

Another MLK2-interactive partner discovered in a yeast-two-hybrid screen is
NeuroD. NeuroD is a member of the basic helix-loop-helix family of transcription
factors, and it plays a key role in neuronal development and survival in vertebrates (Lee
et al., 1995). NeuroD and MLK2 can be coprecipitated from transfected mouse
neuroblastoma N2A cells, and coexpression of MLK2 with NeuroD induces a retardation
in the mobility of NeuroD in SDS-PAGE (Marcora et al., 2003). Furthermore, purified
recombinant MLK2 phosphorylates NeuroD isolated from cells in an in vitro assay.
However, purified recombinant NeruroD is a poor substrate for MLK2 in vitro,
suggesting that additional cofactors or previous modification of NeuroD in cells is
required for the phosphorylation by MLK2. Microinj ection of the NeuroD mRNA into

Xenapus embryos results in ectopic neurogenesis, which is used as an indicator of

22

NeuroD activity. Coinjection of NeuroD and MLK2 into Xenapus embyos further
potentiates the ability of NeuroD to ectopically induce neurogenesis (Marcora et al.,

2003).

3.3 Regulation of MLK5 activities

A considerable amount of work has focused on deciphering the biochemical
mechanisms that regulate MLKs signaling, by analyzing the roles that the different
domains of the MLKs play on MLKs activities. Recently, the efforts have been
concentrated in understanding the biological contexts in which MLKs are activated. The

following sections summarize the current knowledge on how MLKs are regulated.

3.3.1 Regulation of MLK5 by extracellular signals

Upon overexpression, the MLKs display relatively high catalytic activity judged
by their ability to autophosphorylate in in vitro kinase assays. However, the extracellular
signals that activate and regulate endogenous MLKs have remained largely unknown.
Most of the work has focused on MLK3, for which there are commercially available
antibodies. Figure 5 summarizes some of the data.

Several studies have been undertaken to look at the effects of extracellular
signals on MLK3, using as a readout changes in electrophoretic mobility on SDS-PAGE
gels, as an indication of phosphorylation. For instance, in Jurkat T lymphocytes, PMA
treatment and co-stimulation with CD3/CD28 antibodies induce retardation in the
electrophoretic mobility of MLK3, implying that activation of lymphocytes results in

MLK3 phosphorylation (Hehner et al., 2000). Incubation of PMA-treated J urkat T

23

 

 

  
 

Sorhrml

membrane

CD3IC028

    

   

 
 

m... GD‘q}, \/

 

 

Figure 5. Regulation of MLK3 by extracellular signals. Several extracellular stimuli
have been shown to activate MLK3. Cross-linking of the T-cell receptor (TCR) with anti
CD3/CD28 antibodies results in MLK3 activation and signaling through the JNK and/or
NF-KB pathways. NGF withdrawal results in MLK3 activation and JNK-mediated
apoptosis. Activation of the TNF-a receptor (TNF R) also initiates MLK3-mediated
signaling to the JNK and p38 pathways that results in program cell death. TGF-B binding
to its receptor (TGFR) triggers MLK3-mediated activation of the p38 pathway. Growth
factors like EGF, upon binding to their receptors, initiate a signaling cascade that requires
the interaction of Raf and MLK3 for proper activation of the ERK pathway. Stressers
like sorbitol can also activate MLK3 signaling. The precise mechanisms that connect
MLK3 to the activation of the different receptors are not clear. MLK3-mediated
activation of the MAPK pathways initiates a wave of transcriptionally-regulated gene

expression that elicits a variety of different biological responses.

24

lymphocytes with inorganic phosphate results in an increase in the incorporation of
radioactive phosphate into endogenous MLK3, confirming that MLK3 is phosphorylated
in viva upon PMA treatment (W. Chen, unpublished results). It is unclear whether the
phosphorylation changes detected in viva correlate with changes in the catalytic activity
of MLK3.

Incubation of J urkat T lymphocytes with sorbitol, which results in hyperosmotic
stress and activation of all the MAPK pathways, dramatically retards the mobility of
MLK3 in SDS-PAGE (unpublished results). Recently, a phospho-specific antibody
directed against the activation loop of MLK3 has become commercially available. Using
this antibody, it has been possible to detect activation of endogenous MLK3 in J urkat T
lymphocytes treated with sorbitol (unpublished results), as well as in A431 human
epidennoid carcinoma cells (Xu et al., 2005). When subjected to hyperosmotic
conditions, cells turn on the expression of several proteins that regulate the intracellular
levels of osmolytes, like inositol transporters, as well as the expression of Hsp70 to
protect proteins from unfolding and aggregating (reviewed in (Sheikh-Hamad & Gustin,
2004)). Although the mechanism by which cells sense hyperosmolarity is not clear, it has
been established that hyperosmotic stress activates several signaling pathways, including
the p38, JNK, PKA and the tyrosine kinase F yn, resulting in the transcription of the
osmolyte regulators (Sheikh-Hamad & Gustin, 2004). It is possible that MLK3 plays a
role in the transduction of the signal from the osmotic sensor to the MAPK pathways. In
addition, other stress conditions, like DNA damage induced with carnptothecin, a
t0poisomerase inhibitor, results in activation of endogenous MLK3 in HeLa cells and in

PC12 cells as judged by reactivity with the phospho-specific MLK3 antibody (Xu et al.,

25

2005). Overall, the data indicate that MLK3 might be playing a role in cellular responses
against stress insults.

Also in J urkat T lymphocytes, TNF-az stimulation results in MLK3 activation,
as judged by a two fold increase in the phosphorylation of recombinant GST-MKK4 by
MLK3 in vitro (Sathyanarayana et al., 2002). Similar results have been obtained in
MCF-7 cells (Zhang et al., 2004). Interestingly, no changes in the electrophoretic
mobility of MLK3 can be detected upon stimulation of cells with TNF-OL In addition,
treatment of Jurkat T cells with the lipid C6-ceramide increases endogenous MLK3
catalytic activity as judged by in vitro kinase assay (Sathyanarayana et al., 2002).
Ceramide is generated in cells challenged with TNF-a through the activation of
sphingomyelinases that degrade sphingomyelin into ceramide and phosphocholine.
Ceramide promotes apoptosis by inducing perrneabilization of the mitochondria
membranes (reviewed in (Ding & Yin, 2004)). The MLK inhibitor CBP-11004
attenuates both TNF-a—induced and cerarnide-induced JNK activation (Sathyanarayana et
al., 2002). These data suggest that MLK3 participates in the activation of apoptotic
pathways in response to TNF-az.

It has been reported that TGF-B treatment of FaO rat hepatoma cells and Hep3B
human hepatoma cell lines induces activation of MLK3 as judged by reactivity with the
phospho—specific antibody (Kim et al., 2004), although the slow kinetics of MLK3
activation suggest that MLK3 does not lie directly under the TGF-B receptor signaling
cascade, but is rather indirectly activated. As with TNF-a, no change in the
electrophoretic mobility of total MLK3 is observed upon TGF-B treatment.

Overexpression of kinase defective MLK1, MLK2 or MLK3 in these hepatoma cell lines

26

diminishes TGF-B induced apoptosis (Kim et al., 2004), suggesting that activation of the
MLKs is a required step in the process of TGF-B-induced cell death.

Recent work has looked at the effect of grth factors on the activation of
MLK3. Incubation of normal human colon cells CCD-18Co with epidermal growth
factor (EGF) activates endogenous MLK3, as judged by in vitro kinase assay of
immunopurified MLK3 using recombinant GST-MEKl as a substrate (Chadee &
Kyriakis, 2004). However, MLK3 has been reported to be in a complex with B—Raf and
Raf-1(Chadee & Kyriakis, 2004; Chadee et al., 2006), which are upstream activators of
MEKl. Therefore, it is possible that the ascribed ability of MLK3 to phosphorylate
MEK] in vitro might be due to the presence of B-Raf and Raf-1 in complex with the
immunoprecipitated MLK3. Nonetheless, RNAi against MLK3 abolishes the EGF-
induced activation of ERK, JNK and p38, indicating that MLK3 is required to transduce
signals from the EGF receptor (Chadee & Kyriakis, 2004).

Treatment of neuronal-glial cultures from rat embryonic telencephalon with
okadaic acid, a serine/threonine phosphatase inhibitor, results in a retarded
electrophoretic mobility of DLK (Mata et al., 1996). In addition, JNK is activated,
presumably by inhibition of phosphatases that negatively regulate DLK and/or JNK.

The mechanisms by which extracellular ligands regulate the activation of
different MLKs are still unsolved. Specifically, how the signals are transmitted from the
plasma membrane to the MLKs, and what determines which MLK is activated, are
puzzling questions still unresolved. It is likely that the expression levels of the different
MLKs as well as cell type specific interacting partners are modulating the specificity of

the activation of the MLKs.

27

3.3.2 Regulation of MLKs by autoinhibition

MLK1-MLK4 have an NHz-terminal SH3 domain. SH3 domains are
independently folded modules of about 60 amino acids. SH3 domains bind proline-rich
sequences, the consensus for binding being PXXP, preceded or followed by basic amino
acids (Yu et al., 1994). One prediction is that the SH3 domain of MLK1 -MLK4 would
mediate the interaction with other proteins and/or the localization to different cellular
compartments, to regulate signaling events. The SH3 domain of MLK3 has been shown
to interact with a proline-rich region in hematopoietic progenitor protein kinase 1
(HPKl), and HPKl can phosphorylate kinase inactive MLK3 in vitro (Kiefer et al.,
1996). However, no effect of the HPKl interaction with MLK3, as well as of the HPKl-
mediated phosphorylation of MLK3, on MLK3 activity has been reported in the literature
or has been observed in our laboratory.

The SH3 domain of the protein kinase Src is involved in an autoinhibitory,
intrarnolecular interaction (Hubbard et al., 1998). Similarly, the SH3 domain of MLK3
has been shown to negatively regulate its kinase activity, by means of an interaction with
a region of MLK3 located between the zipper and the CRIB motif (Zhang & Gallo,
2001). Disruption of the SH3 domain of MLK3 by mutation of a conserved Tyr to Ala
results in increased MLK3 catalytic activity. Interestingly, the SH3 domain binding
region contains only one Pro residue. Mutation of that Pro to Ala also results in
increased MLK3 kinase activity. The critical Pro in the SH3-binding region is conserved
in MLK1, MLK2, MLK4, Slpr and the frog ortholog of MLK2, suggesting that all the
MLK subfamily members may use a common mechanism of SH3-mediated

autoinhibition.

28

3.3.3 Regulation of MLKs by dimerization

All the MLK family members have a leucine zipper. Leucine zippers mediate
protein oligomerization by forming coiled coil structures. These structures are stabilized
mainly by leucine residues spaced seven residues apart, which interact at the interface of
opposing helices (Hu et al., 1990; O'Shea et al., 1991). The data available indicate that
the leucine zipper region of MLK3 mediates homodimerization (Leung & Lassam, 1998).
Although MLK3 and DLK can be coimmunoprecipitated from cells (Tanaka & Hanafusa,
1998), the DLK leucine zipper does not form efficient heterodimers with other MLKs
(Nihalani et al., 2000), suggesting that heterodimerization of MLKs is unlikely to occur
through this region.

The leucine zipper region of MLK3 is necessary for MLK3 oligomerization
(Bock et al., 2000), and deletion of the entire zipper region generates a MLK3 variant that
is catalytically inactive in vitro and that fails to activate the JNK pathway (Leung &
Lassam, 1998). Destabilization of the leucine zipper region by mutating one of the Leu
residues to Pro, L410P, disrupts dimerization and MLK3-mediated JNK activation
(Vacratsis & Gallo, 2000). Strikingly, although the leucine zipper mutant displays nearly
null in vitro autophosphorylation activity, coexpression of the small GTPase Cdc42
greatly stimulates this activity (Vacratsis & Gallo, 2000). Similarly, the leucine zipper
mutant does not display reactivity with the phospho-MLK3 antibody, while coexpression
with Cdc42 potently induces activation loop phosphorylation (Yan Du, unpublished
observations). A possible explanation for this observation is that Cdc42 itself can
dimerize (Zhang & Zheng, 1998), and dimerized Cdc42 can force two molecules of

MLK3 into close proximity, so that transphosphorylation and activation can occur.

29

However, even in the presence of activated Cdc42, the leucine zipper mutant fails to
phosphorylate Thr 258 in the activation loop of MKK4 when used as an in vitro substrate.
These results indicate that MLK3 dimerization is required for proper substrate
phosphorylation and activation of the downstream pathway.

Leucine zipper-mediated DLK dimerization is necessary for DLK activation
and stimulation of the JNK pathway (Nihalani et al., 2000). Induced dimerization of
DLK promotes autophosphorylation and activation of JNK (N ihalani et al., 2001).
Consistent with a model in which dimerization results in transphosphorylation and
activation, overexpression of the DLK leucine zipper inhibits dimerization and activation
of fiill length DLK, by forming inactive complexes (Nihalani et al., 2001). LZK also
forms oligomers, and deletion of the LZK leucine zipper region prevents LZK-induced
JNK activation (Ikeda et al., 2001).

In summary, leucine zipper-mediated dimerization is a common regulatory
mechanism among the members of the MLK family. Dimerization precedes

autophosphorylation, and is required for subsequent JNK activation.

3.3.4 Regulation of MLKs by small GTPases

Small GTPases are molecular switches that cycle between an active, GTP-
bound state, and an inactive, GDP-bound state. The Rho family of small GTPases
includes RhoA-E and RhoG; Racl and Rac2; Cdc42; and TC10 (Mackay & Hall, 1998).
They regulate many cellular processes, including cell growth, cellular transport, cell

motility and cytoskeleton rearrangements (Bar-Sagi & Hall, 2000). Constitutively active

30

forms of Rae and Cdc42 have been shown to activate the JNK and p38 pathways (Coso et
al., 1995; Minden et al., 1995).

MLK1-4 contain a CRIB motif. CRIB motifs are composed of 14-16 amino
acids, eight of which are conserved and mediate binding with activated forms of Rac and
Cdc42 (Burbelo et al., 1995). MLK3, which has six of the eight conserved residues in its
CRIB motif, binds to activated Cdc42 and Rac (Bock et al., 2000; Burbelo et al., 1995;
Teramoto et al., 1996). Ablation of MLK3 expression by RNAi blocks ectopically
expressed Cdc42-induced JN K activation (Du et al., 2005), reaffirming the notion that
Cdc42 is an upstream activator of MLK3, and that MLK3 is a relevant interactor of
Cdc42 in cells. Coexpression of MLK3 with activated Cdc42 promotes MLK3
oligomerization (Leung & Lassam, 1998; Nihalani et al., 2001), increases MLK3
catalytic activity, potentiates MLK3-induced JNK activation (Bock et al., 2000), and
increases the in viva phosphorylation of MLK3, which is accompanied by a retarded
mobility of MLK3 in SDS-PAGE (Bock et al., 2000). Coexpression of MLK3 with
Cdc42 also results in the phosphorylation of the activation loop of MLK3, translocation
of MLK3 to membranes, and further phosphorylation of MLK3 (Du et al., 2005).
Maximal JNK activation by MLK3 is dependent on its localization to membranes,
suggesting that full activation of MLK3 is achieved once targeted by Cdc42 to
membranes (Du et al., 2005).

The binding region for Cdc42 lies in close proximity to the autoinhibitory
binding site. Although the mechanism by which Cdc42 activates MLK3 is not completely
understood, the following model has been pr0posed: binding of Cdc42 to the CRIB motif

on MLK3 disrupts the SH3-mediated intrarnolecular inhibition, resulting in induced

31

dimerization, transphosphorylation on the activation loop, and relocalization to
membranes. Here, MLK3 undergoes full activation and utmost JNK activation is

obtained.

3.3.5 Regulation of MLKs by scaffold proteins

JNK interacting proteins 1-4 (J IPs) are a group of scaffold proteins encoded by
four genes. These proteins can be divided into two sub-groups based on sequence
similarities, but all of the J IPs share common biochemical properties, including potential
functions as cargo molecules for kinesin motor proteins and as scaffold proteins for JNK
and p38 (Morrison & Davis, 2003).

J [P] and J 1P2 are structurally similar. J [PI is a scaffold protein that binds some
members of the JNK signaling cascade. In addition to JNKs and MKK7, several MLKs,
including MLK2, MLK3, DLK and LZK, have been shown to interact with J 1P1
(Whitrnarsh et al., 1998; Yasuda et al., 1999). Furthermore, guanine nucleotide exchange
factors (GEFs) like Tiam, kinases like AKT, and phosphatases like MAPK phosphatase 7
(MKP7) are also found in association with the JIP scaffolds (reviewed in (Morrison &
Davis, 2003)). Coexpression of JIP] with MLK3 potentiates JNK activation, suggesting
that the role of scaffolds is to facilitate signaling by organizing the cascade module.
However, the proposed mechanism for J IP-mediated DLK activation of JNK is more
complex than just providing docking regions for the different components of the
signaling pathway. Under basal conditions, DLK is bound to J [P in a monomeric,
inactive form. Upon stimulation, JNK is recruited to JIP] and phosphorylates it. DLK

now dissociates from J IPl, dimerizes and becomes catalytically competent. Activated

32

DLK phosphorylates JIP1-bound MKK7, which in turn phosphorylates JIP1-bound JNK
and activates it (Nihalani et al., 2001; Nihalani et al., 2003).

Although initial studies suggested that JIP1 and J IPZ bound the same repertoire
of proteins of the JNK pathway, later experiments indicated that the binding of J IPZ to
JNK is weaker than that of J 1P1, and that JIP2 instead prefers p38oz and p3 86 as binding
partners (Buchsbaum et al., 2002; Schoorlemmer & Goldfarb, 2001). J 1P2 associates
with MLK3, MKK3 and p38 isoforms. In addition, J IPZ associates with Tiarnl and Ras-
GRF, which are GEFs for the small GTPase Rac. Expression of these GEFs enhances the
association between J IP2 and MLK3 (Buchsbaum et al., 2002), presumably by
facilitating the activation of MLK3 through Rac and the formation of an activated
complex between MLK3 and JIP2.

JIP3 and J IP4 are similar to each other. J 1P3 (also known as JNK/stress-
activated protein kinase-associated protein 1 (J SAP1)) has been shown to interact with
JNK, MKK7 and members of the MLKs (Kelkar et al., 2000). However, JIP3 can also
associate with the MAPKKK MEK kinase 1 (MEKKl) and MKK4 (Ito etal., 1999), and
with ASKl (Matsuura et al., 2002). Similarly, JIP4 interacts with MEKK3 and MKK4
(Lee et al., 2002).

Because all of the J [P proteins associate with kinesin motor complexes, it is
possible that the JIP proteins play a role in kinesin-mediated transport and could have
roles in regulating the subcellular localization of the JNK signalosome.

In addition to JIPs, Plenty 9f SL133 (POSH) is another scaffold protein that
associates with the small GTPase Rae, and in overexpression studies it promotes

apoptosis by activating the JNK pathway (Tapon et al., 1998). POSH interacts directly

33

with activated Racl, MLKs (MLK1-3 and DLK), and indirectly with MKK4/7 and JNKs
(Xu et al., 2003). POSH-induced apoptosis can be blocked by treatment with the MLK
inhibitor CBP-1347 (Xu et al., 2003). Akt2 has also been reported to bind to the POSH-
MLK-MKK-JNK complex, and the interaction of Akt2 with POSH seems to disassemble
the JNK signalosome and to downregulate JNK signaling (Figueroa et al., 2003).
Recently, it has been shown that POSH and J IPs directly bind to each other to form a
multiprotein complex that includes all of the known kinase components of the pathway
(Kukekov et al., 2006).

Post-synaptic density protein 95 (PSD-95) is a neuronal protein composed of
three PDZ domains, an SH3 domain, and a catalytically inactive guanylate kinase domain
(Garcia et al., 1998). PSD-95 interacts with glutamate receptor 6 (GluR6) through one of
the PDZ domains. The SH3 domain of PSD-95 interacts with MLK2 and MLK3.
Coexpression of kinase defective MLK2 or MLK3 significantly diminishes JNK
activation and neuronal toxicity mediated by GluR6. These data suggest that PSD-95
acts as a scaffold protein to mediate signaling from ion channels to JNK by anchoring
MLK2 or MLK3 to the GluR6 receptor complex (Savinainen et al., 2001).

Connector enhancer of KSR (CNK) was identified in a genetic screen for genes
that could enhance the phenotype observed in the fly when the kinase domain of KSR is
overexpressed (reviewed in (Kolch, 2005)). There are three mammalian homologues of
CNK, CNK1-3. The widely expressed CNK] interacts with Raf-1 and c-Src (Ziogas et
al., 2005) to enhance Raf-l activity. However, another report indicates that CNKl can
also bind members of the JNK pathway, namely MLK2 and MLK3, and MKK7, to

induce c-Jun phosphorylation downstream of the small GTPase Rho (J affe et al., 2005).

34

At present, it is not known whether these activities are performed by different pools of
CNK molecules, or whether overlapping binding partners link the different regulatory

functions of CNKl.

3.3.6 Regulation of MLKs by chaperone proteins

Recently, an affinity purification procedure was used to isolate MLK3 protein
complexes fi'om cells inducibly expressing MLK3. Electrospray/Mass spectrometry
analysis of the mixture identified Hsp90 and Cdc37 as binding partners of MLK3 (Zhang
et al., 2004). Hsp90 is one of the most abundant proteins in eukaryotic cells. It functions
as a molecular chaperone to prevent aggregation of its client proteins. In mammalian
cells, there are two genes encoding cytosolic Hsp90 homologs, and there are also distinct
isofonns localized in the endoplasmic reticulum (Grp94) and in mitochondria (TRAP-l).
Hsp90 consists of three major domains: a highly conserved NHz-tenninal ATPase
domain, a middle domain, and a COOH-terrninal dimerization domain. The clients of
Hsp90 include steroid hormone receptors and protein kinases, such as Src, Raf, and Cdk4
among others. The co-chaperone Cdc37 is required for the interaction of Hsp90 with its
client protein kinases, which occurs through the kinase domain (reviewed in (Pearl, 2005;
Wegele et al., 2004)).

MLK3 also associates with Hsp90/Cdc37 through the kinase domain, and like
other protein kinases, the interaction is not dependent on the activation state of MLK3.
Inhibition of the ATPase activity of Hsp90 with geldanamycin decreases the protein
levels of endogenous MLK3, which suggests that Hsp90 helps MLK3 to fold properly,

and that in the absence of chaperone activity MLK3 is unstable (Zhang et al., 2004).

35

3.3.7 Regulation of MLKs by phosphorylation

Phosphorylation is a versatile mechanism to regulate protein kinase activities.
Activation loop phosphorylation is the predominant event involved in the activation of
most protein kinases. As mentioned before, the activation loop is flanked by two motifs
very conserved among protein kinases, DF G and APE. Analysis of the sequence of the
activation loop of MLK3 indicates that there are four potential phosphorylation sites
within the loop: Thr 277, Thr 278, Ser 281 and Thr 285. Replacement of Thr 277 or Ser
281, but not Thr 278, with Ala generates a form of MLK3 with low kinase activity,
whereas replacement of the Thr 277 and Ser 281 with Asp residues results in a kinase as
active as wild type MLK3, suggesting that Thr 277 and Ser 281 could be activation loop
sites (Leung & Lassam, 2001). A commercially available phospho-specific antibody has
been generated against a peptide containing the activation loop sequence of MLK3
phosphorylated at Thr 277 and Ser 281. This antibody recognizes MLK3 when it has
been coexpressed in cells with activated Cdc42 (Du et al., 2005). Under the same
experimental conditions, a kinase defective form of MLK3 does not react with the
phospho-specific antibody, consistent with the idea that MLK3 autophosphorylates its
activation loop sites. The reactivity with the phospho-specific antibody correlates with
increased catalytic activity of MLK3, supporting the idea that Thr 277 and Ser 281 are
the banafide activation loop sites of MLK3.

An alignment of the sequences of the activation loop of the MLK subfamily is
shown in Figure 6. The sequences of the activation loop of MLK1 and of the activation
loop of MLK3 differ in only 3 out of 26 residues, with the MLK3 activation loop sites

conserved in MLK1. Surprisingly, mutagenesis studies indicate that the activation of

36

hMLK3
mmx3
hMLKl
hMLK2
hMLK4
Slpr

xMLK

Figure 6. Alignment of the sequences of the activation loop of the MLK subfamily
members. The activation loop sequences from the human MLKs, as well as mouse
MLK3 (mMLK3) and the Drasophila (Slpr) and Xenapus (xMLK) orthologs of the
MLKs are displayed. The phosphorylation sites identified in human MLK3 are
highlighted in sky blue box. The phosphorylation site identified in human MLK1 is

267
268
242

289
290

highlighted in yellow box.

 

37

MLK1 requires phosphorylation of Thr 312, which is equivalent to Thr 285 in MLK3
(Durkin et al., 2004). The availability of the crystal structures of the kinase domains of
the MLKs might enlighten the reasons for this difference.

In HepG2 liver cancer cells, insulin treatment results in the activation of
phosphatidylinositol 3-kinase (PI3K), which generates phosphoinositol 3 phosphate
(PIP3) at the cell membrane. The protein kinase Akt is activated by its association with
PIP3, as well as by phosphorylation by other protein kinases at the membrane. Akt is
suggested to have a positive impact on cell survival, blocking the processes that lead to
apoptosis. It has been shown that upon insulin treatment, Akt is activated; under those
conditions, MLK3, MKK7 and JNK catalytic activities are attenuated. The data suggest
that Akt phosphorylates Ser 674 on MLK3, negatively affecting MLK3 activity and
MLK3-mediated JNK activation and apoptosis (Barthwal et al., 2003).

Mass spectrometry coupled with phosphopeptide mapping analysis was used to
identify in viva phosphorylation sites on MLK3 (Vacratsis et al., 2002). One of the sites
is located between amino acids 11-37 of MLK3, but the exact residue could not be
assigned. Eleven serines were positively identified as phosphorylation sites; all of them
are located beyond the CRIB motif on the COOH-terminal region of MLK3. From these
eleven sites, seven correspond to Ser residues immediately followed by a Pro residue,
which suggests that MLK3 might be a substrate for proline-directed kinases. This idea

will be discussed in detail in Chapter II.

38

3.3.8 Regulation of MLKs by subcellular localization

One way to regulate the activity of protein kinases is by modulating their
accessibility to their substrates. Changes in subcellular localization of protein kinases
might bring them in close proximity to a particular target, thus regulating signal
specificity.

Initial immunohistochemical studies in NIH 3T3 cells indicate that endogenous
DLK localizes to a perinuclear region. Treatment of cells with Brefeldin A, a fimgal
toxin that dismembers the stacks of the Golgi apparatus, results in a dispersed cytosolic
localization of DLK, strongly suggesting that DLK is localized to the Golgi (Douziech et
al., 1999). DLK present in a crude membrane fraction preparation could only be
solubilized in the presence of Triton-X1 14, indicating that DLK is a peripheral protein
associated with the cytosolic face of the Golgi membranes (Douziech et al., 1999). The
activation status of the Golgi-associated DLK has not been examined; therefore, it is not
currently known how changes in DLK activity correlate with its localization to the Golgi
apparatus. Similarly, ectopically expressed MLK3 shows some localization to a
perinuclear region in HeLa cells (Du et al., 2005). Since both the wild type form and a
kinase defective mutant form of MLK3 have similar subcellular distributions, at least in
biochemical fractionations (Du et al., 2005), the presence of MLK3 in the
Golgi/perinuclear region is unlikely to be regulated through activation of MLK3.
However, as mentioned in the previous section, maximal activation of MLK3 upon
coexpression with Cdc42 requires translocation to the plasma membrane.

MLK2 microinjected into the nuclei of Swiss 3T3 cells localizes along

microtubules (Nagata et al., 1998). The staining pattern is punctate, and radial.

39

Interestingly, a truncated form of MLK2 which lacks the SH3 domain and the kinase
domain retains the same punctuate cellular distribution. However, only when wild type
MLK2, but not the truncated form of MLK2, is microinjected into cells, does activated
JNK colocalize to microtubules, supporting the idea that they form a functional signaling
complex which might play a role in cellular trafficking. Yeast-two-hybrid studies using
MLK2 as a bait identified members of the kinesin superfamily proteins (KIFs) as
interacting partners of MLK2 (Nagata et al., 1998). KIFs are molecular motor proteins
that transport organelles, mRNAs and protein complexes in vesicles along microtubules
to different locations within the cell, in an ATP-dependent manner (Miki et al., 2001).
MLK2 and MLK3 were shown to interact with the kinesin motor protein KIF 3, and with
the putative cargo recognition molecule KAP3. These data suggest that MLK2 and

MLK3 could be involved in regulating vesicular trafficking.

3.4 The physiological roles of the MLKs
Studies regarding the fimction of the MLKs have been related to their role as
upstream activators of the JNK pathway and their effect in apoptosis, specifically in

neuronal cells. Some of the studies are outlined below.

3.4.]. The Drasaphila MLK, Slpr

In Drasaphila, the process of dorsal closure is regulated by the JNK pathway.
In this process, the dorsal ectoderrn moves from a lateral position to the dorsal midline, to
cover the embryo in a continuous protective epidermis (Noselli, 1998). The components

of the signaling pathway required for dorsal closure are the GTPase dRacl , a

40

MAPKKKK called Misshapen (Msn), Slpr, the MKK7 ortholog Hemipterous (Hep), the
JNK ortholog Basket (Bsk), and the AP-l transcription factors dJun and dFos, which are
encoded by Jun-related antigen (Jra) and kayak (kay), respectively (reviewed in
(Stronach, 2005; Xia & Karin, 2004)).

JNK has two functions during epithelial cell-sheet morphogenesis. One of them
is the expression of the cytokine Decapentaplegic (Dpp), which is orthologous to bone
morphogenic protein 4, at the leading edge of the epithelial sheet. AP-l activation is
required for the expression of Dpp. Slpr mutants fail to express dpp, consistent with the
lack of activation of Bsk (Stronach & Penimon, 2002). The other fimction of JNK is the
control of stretched epithelial cell morphology. Mutants in the Slpr gene cannot maintain
the stretched morphology and the cells round up, indicating that this MLK is involved in
the regulation of cell shape during dorsal closure. Interestingly, whole—mount in situ
hybridization of Xenapus Iaevis embryos shows early expression of the Xenapus ortholog
of the MLKs (xMLK2) in the cement gland, a mucus-secreting epithelium that forms
fi'om the extreme anterior ectoderm (Poitras et al., 2003). In later stages, xMLK2 is
found in the brain and pronephros. In the case of the cement gland and pronephros, the
expression of xMLK2 correlates with the formation of epithelia (Poitras et al., 2003).
Thus, it seems that the MLKs might be playing a general role during differentiation of

epithelial layers.

3.4.2 Mammalian MLKs

Using RNAi, it has been shown that MLK3 is required for the proliferation of

tumor cell lines (Chadee & Kyriakis, 2004). The data indicate that MLK3 and B-Raf

41

exist in a complex in cells, and that MLK3 is required for the activation of B-Raf, which
in turn results in the activation of ERK and in the induction of cell proliferation.
Recently, the generation of the mlk3 knockout mice has been reported (Brancho et al.,
2005). Mice derived from targeted gene disruption of the mlk3 gene are viable and show
no obvious defects. In primary mouse embryonic fibroblasts (MEFs) derived from mlk3
knockout mice, there is a 50% reduction in TNF-oz mediated activation of JNK,
suggesting that MLK3 is necessary for full activation of the JNK pathway upon TNF-az
stimulation, but that other MAPKKKs may also partially contribute or at least
compensate for the lack of MLK3 (Brancho et al., 2005). MEFs from mlk3’/' mice do not
display deficient activation of the MAPKs in response to various growth factors (Brancho
et al., 2005), whereas the RNAi studies indicate that the activation by cytokines and
mitogens of JNK, ERK and p38 is dependent on MLK3 (Chadee & Kyriakis, 2004). One
possible explanation for this discrepancy is that in the knockout mouse, compensatory
mechanisms are prone to occur during development. If other MLKs can substitute for
MLK3 functions, then only non-redundant roles for MLK3 will emerge. However during
transient knock down of MLK3 expression by RNAi, the cellular phenotypes observed
would correspond to more immediate effects of MLK3 ablation.

A cumulative body of work implicates the MLKs in JNK-mediated apoptosis in
neuronal cells. NGF deprivation induces apoptosis, and this event is dependent on the
activation of the JNK pathway (Mielke & Herdegen, 2000). Overexpression of MLK2,
MLK3 or DLK induces apoptosis in neuronal-like PC12 cells (Xu et al., 2001).
Overexpression of MLK3 in superior cervical ganglion (SCG) sympathetic neurons

activates JNK and induces apoptosis, while expression of a dominant negative MLK3

42

blocks apoptosis induced by NGF deprivation (Mota et al., 2001). Likewise, the MLK
inhibitor CBP-1347 protects primary neurons in culture from NGF-deprivation-induced
apoptosis (Maroney et al., 1998). In addition, animal models of chemical-induced
neurotoxicity show that CBP-1347 effectively protects in viva (Saporito et al., 1999).
Altogether the data suggest that the MLKs are activated by NGF withdrawal and this
turns on the JNK pathway and promotes apoptosis.

Huntington’s disease is a neurodegenerative disorder produced by mutations in
the gene of the Huntingtin protein. Polyglutarnine stretches are added at the NH;-
terminus of the protein causing it to aggregate inside neuronal bodies (Morfini et al.,
2005). The SH3 domain of MLK2 has been shown to interact with normal Huntingtin,
but not with the polyglutamine expanded mutant protein. Likewise, MLK3 also interacts
with normal Huntingtin in GST-pulldown assays using the SH3 domain of MLK3 (Geou-
Yiarh Liou, Gallo lab). Expression of catalytically inactive MLK2 can diminish neuronal
apoptosis induced by mutant Huntingtin (Liu et al., 2000b). The proposed model
suggests that normal Huntingtin and MLK2 interact in neuronal cells, and that this
association keeps MLK2 inactive. In Huntington’s disease, polyglutamine expanded
Huntingtin can no longer associate with MLK2. This free MLK2 can now be activated
and can promote apoptosis. However, the inability of MLK2 to interact with
polyglutamine expanded Huntingtin may just be due to the aggregation of the mutated
Huntingtin. Further experiments are needed to clarify the role of the MLKs in
Huntington’s disease.

Finally, the MLKs have also been involved in the regulation of the cell cycle.

Cell cycle studies in the filamentous fungus Aspergillus nidulans have demonstrated that

43

proper control of the G2/M transition requires the activity of a serine/threonine protein
kinase called never in mitosis A (N IMA). Loss of NIMA function causes cell cycle arrest
in G2 phase (Osmani et al., 1991). In addition to its kinase catalytic domain, NIMA has a
long COOH-terminal region that is necessary and sufficient to induce G2/M arrest when
overexpressed in human cells (Lu & Hunter, 1995). The COOH-tenninal region of
NIMA, and the region between amino acids 365-847 of MLK3 were found to share 16%
identity and 22% similarity (Swenson et al., 2003). In essence, the sequence similarities
between NIMA and MLK3 are due to the presence of a kinase domain, central domains
predicted to be coiled-coils, and a COOH-tenninal region rich in Ser/Thr—Pro motifs.

The catalytic activity of NIMA is enhanced during G2/M; during early mitosis, the
electrophoretic mobility of MLK3 is retarded due to phosphorylation, which correlates
with increased catalytic activity of MLK3. However, this MLK3 activation does not
translate into JNK activation, suggesting a JNK-independent role for MLK3 during the
cell cycle. Analysis of the subcellular distribution of ectopically expressed NIMA or
MLK3 revealed that they both localize to the centrosome, as judged by costaining with )9
tubulin. Interestingly, the subcellular distribution of endogenous MLK3 varies in the
different stages of the cell cycle; during interphase and prophase, MLK3 appears as a
very bright spot located near the nucleus that corresponds to the centrosomes. During
metaphase, MLK3 staining becomes totally cytosolic, and it only relocates to the
centrosome after cytokinesis is complete (Swenson et al., 2003). Other reports indicate
that MLK3 localizes to a Golgi/perinuclear region (Cha et al., 2004; Du et al., 2005), but
the bright spot of centrosomal localization has not been reproduced in our hands. Loss of

NIMA leads to cell cycle arrest; however, ablation of MLK3 expression by RNAi does

not result in a cell cycle arrest in SAOS human osteosarcoma cells (Swenson et al.,
2003). In contrast, the MLK inhibitor CBP-11004 delays mitotic progression and
promotes the formation of aberrant mitotic spindles in HeLa cells (Cha et al., 2006). The
results obtained using CBP-11004 should be carefully considered, since its ability to
inhibit other protein kinases has not been well characterized and therefore the effects seen
may very well correspond to off target events. Interestingly enough, the levels of the
MLK ZAK increase during G2/M in MDCK cells stably expressing ZAK (Gross et al.,
2002; Liu et al., 2000a). Catalytically inactive ZAK attenuates 'y-irradiation-induced cell
cycle arrest. At present, the exact role of the MLKs in the regulation of cell cycle

progression is not clear.

4. Objective of the Thesis

Protein kinases are key regulators of a variety of cellular processes. Regulation
of their activities is at the center of cellular homeostasis. My thesis work has been
directed towards understanding the molecular mechanisms that regulate the various
activities of MLK3.

As described in the previous sections, MLK3 is a widely expressed
serine/threonine kinase whose major function has been ascribed as an upstream activator
of the JNK pathway. MLK3 contains several domains that regulate its activity, including
an SH3 domain, a leucine zipper region and a CRIB motif. Our current working model
for MLK3 activation proposes that disruption of the SH3-mediated autoinhibitory
intramolecular interaction allows MLK3 to dimerize and transphosphorylate in its

activation loop. Activated MLK3 is competent to modulate JNK activation through

45

phosphorylation of MKK4 and MKK7. The interaction of MLK3 with the small GTPase
Cdc42 might trigger the activation process, and also results in relocalization of MLK3 to
the plasma membrane.

MLK3 is a multiphosphorylated protein. Activation of MLK3 relies on
phosphorylation of its activation loop. In addition, several other phosphorylation sites
have been identified in MLK3, most of which correspond to phosphorylated Ser residues
followed by Pro residues (SP sites) that could potentially be phosphorylated by proline-
directed kinases. The research work presented in this dissertation investigates whether
MLK3 is regulated by proline-directed kinases.

Chapter II investigates which proline-directed kinase phosphorylates MLK3 on
SP sites. This study also evaluates the conditions in which phosphorylation at the SP
sites occurs and the dynamics of the process. Chapter III examines the changes in the
subcellular distribution of MLK3 as a result of phosphorylation by proline-directed
kinases. Chapter IV summarizes the results presented in this work and presents the most
current working model for the regulation of MLK3, as well as final remarks and future
directions. Finally, an Appendix compiles the information obtained from site-directed
mutagenesis studies of some of the phosphorylation sites on MLK3 and their effect on

MLK3 catalytic activity.

46

II. Phosphorylation of Mixed Lineage Kinase 3 by Praline-directed

Kinases

1. Abstract

Mixed-lineage kinase 3 (MLK3) is a widely expressed, mammalian
serine/threonine protein kinase that activates multiple MAPK pathways. MLK3 induces
JNK activation through phosphorylation of MKK4 and MKK7. MLK3, and other MLKs,
have been implicated in JNK-mediated neuronal cell death in response to trophic factor
withdrawal. However, MLK3 also appears to play key roles in cellular proliferation,
especially in the context of cancer cells. Previously our lab used in viva labeling/mass
spectrometry to identify phosphorylation sites of activated MLK3. Seven of eleven
identified sites correspond to the consensus motif for phosphorylation by proline-directed
kinases. Based on these results, I hypothesized that JNK, or another proline-directed
kinase, phosphorylates MLK3 as part of a feedback loop. Using a combination of
pharmacological inhibitors and modulators of JNK activity, I show that MLK3 can be
phosphorylated by JNK in vitro and in viva. Blockade of JNK results in
dephosphorylation of MLK3. My data indicate that COOH-terminal JNK
phosphorylation is critical for the regulation of MLK3 levels in cellular lysates. This
work describes a novel mode of regulation of MLK3 and its signaling to JNK. My
findings have important implications for processes involving sustained JNK activation,

including apoptosis.

47

2. Introduction

Protein kinases and their signaling cascades are dynamically regulated through
protein-protein interactions, subcellular targeting, and phosphorylation. By differential
integration of these regulatory mechanisms, a protein kinase and its signaling pathway
can lead to distinct cellular responses. In addition, many protein kinases are subjected to
feedback regulation, and biological signaling pathways are modulated by positive and
negative regulatory feedback loops.

Our lab has focused on understanding the mechanisms that regulate MLK3
activity and signaling. In addition to its catalytic domain, MLK3 contains several other
regions important for its regulation, including an SH3 domain and a centrally located
zipper followed by a Cdc42/Rae Interactive Binding (CRIB) motif. Notably, the COOH-
terminal region of MLK3 is rich in Ser, Thr, and Pro residues, with the final 220 amino
acids of MLK3 being comprised of 12%, 13%, and 24%, respectively, of these amino
acids. Activated forms of the small GTPases, Cdc42 and Rac, can bind to MLK3, retard
MLK3 electrophoretic mobility, increase MLK3 catalytic activity and potentiate
signaling to JNK (Bock et al., 2000; Burbelo et al., 1995; Teramoto et al., 1996; Tibbles
et al., 1996). Our recent work demonstrates that binding of activated Cdc42 promotes
phosphorylation of Thr 277 and Ser 281 within the activation segment of MLK3 leading
to increased MLK3 activity (Du et al., 2005). Additionally Cdc42, in a prenylation—
dependent fashion, induces translocation of MLK3 to cellular membranes (Du et al.,
2005)

MLK3 is multiphosphorylated in viva. In addition, data from our lab indicate that

coexpression of activated Cdc42 changes the in viva phosphorylation pattern of MLK3

48

(Bock et al., 2000). Several in viva phosphorylation sites of MLK3 have been identified
using metabolic labeling coupled with mass spectrometry (Figure 1) (Vacratsis et al.,
2002). Strikingly, seven of the identified sites, serines 524, 705, 727, 740, 758, 770 and
793, are followed immediately by a proline, thus conforming to the consensus for
phosphorylation by proline-directed kinases. There are three well-characterized groups
of proline directed kinases: the MAPKs, the Cyclin-dependent kinases (Cdks), and
Glycogen Synthase Kinase 3 (GSK3). GSK3 is special in that it phosphorylates Ser or
Thr residues in a “Ser-Pro” motif in proteins that have been already primed, by
phosphorylation of another Ser or Thr (reviewed in (Frame & Cohen, 2001)). For the
MAPKs and Cdks, some effort has been directed towards identifying the precise
sequence requirements for substrate phosphorylation, but it seems that in most cases the
exact sequence surrounding the phosphorylation site is not as relevant as the presence of
a Ser/Thr-Pro motif.

The identification of sites of proline-directed phosphorylation of MLK3, together
with the ability of MLK3 to activate multiple MAPK pathways, led me to hypothesize
that MLK3 might be regulated through feedback phosphorylation by MAPKs. In this
chapter, data is presented supporting JNK—mediated phosphorylation of MLK3 in cells.
JNK also phosphorylates MLK3 in vitro. Manipulation of JNK activity in cells
dramatically impacts net phosphorylation and the levels of MLK3 in cellular lysates.
Taken altogether the data provides evidence for a novel positive feedback loop in which
the MAPKKK, MLK3, is phosphorylated by its downstream MAPK, JNK, leading to

stabilization of MLK3 and enhanced signaling through JNK.

49

3. Materials and Methods
3.1 Reagents and Antibodies

The following reagents were purchased from Calbiochem: SP600125, 88203580,
U0126, Roscovitine, and Kenpaullone. Calyculin A was obtained from BioMol. The
MLK inhibitor CBP-11004 was generously provided by Cephalon, Inc. Mouse
monoclonal antibodies against phospho-c-Jun (KM-1) and GF P, were purchased from
Santa Cruz Biotechnology, Inc. The phospho-MLK3 antibody and phospho-JNK
antibody were from Cell Signaling Technology. Other antibodies used were the MLK3
rabbit polyclonal antibody (Bock et al., 2000), actin and Flag mouse monoclonal antibody
(Sigma), hemagglutinin (HA) mouse monoclonal antibody (BAbCO), and horseradish

peroxidase-conjugated secondary antibodies (Bio-Rad).

3.2 Construction of mammalian expression vectors and site-directed mutagenesis

Construction of the expression vectors pRKS-MLK3, pRKS-MLK3 K144R,
pRKS-Flag-MLK3, pRK5—Flag-MLK3 1-598, pCGN-HA-MLK3 and pRKS N-F lag
Cdc42V12 has been described elsewhere (Bock et al., 2000; Zhang & Gallo, 2001).
MLK3 K144M was generated by the Quick Change Site-directed Mutagenesis method
(Stratagene) using the following oligonucleotide:
5’CTGGTGGCTGTGATGGCAGCTCGCC 3’, and its reverse complement. The MLK3
variant containing six point mutations was constructed by introducing successive point
mutations into pCGN-HA-MLK3. The primers used for mutagenesis were:

5’-GACCTCGGCCCGCGCCCCTTCGCAG-3’ (S770A);

50

5 ’-ACCCCACGTGCACCACCCCTG-3 ’ (S75 8A);
5 ’-GGTCAGCCAAGGCCCCCCGACGTGA-3 ’ (S727A);
5’-TCTCCCCTGCCAGCACCACAGCCTGCACCC-3’ (S793A);
5’-CTCAAGACGCCCGCAGCCCCGCCCAC-3’ (S705A) and
5’-GGAGGCACTGTCGCACCCCCACCGG-3’ (S740A), and their reverse
complements.

F lag-MKK7 K165A and GST-MKK4 K129R were generated using pCDNA3-
F lag-MKK7 and pEBG-GST-MKK4 as templates, and the following oligonucleotides
and their reverse complements: 5’- CATTGCTGTTGCGCAAATGCG-3’ for MKK7,
and 5’- CAGATAATGGCAGTTAGAATTCGGTCAACT-3’ for MKK4. Other
mammalian expression vectors include pSRa3HA-JNKK2-JNK1 (kindly provided by Dr.
Anning Lin, U. of Chicago), pCDNA3-Flag-ubiquitin (kindly provided by Dr. Kun-Liang
Guan, U. of Michigan) and pCMS-EGFP (kindly provided by Dr. Lloyd Greene,

Columbia University).

3.3 Cell culture and transfections

Human embryonic kidney (HEK) 293 cells were maintained in Ham’s F 12:low
glucose Dulbecco’s modified Eagle’s media (DMEM) (1:1) supplemented with 8% fetal
bovine serum, 2 mM glutarnine and penicillin/streptomycin. HeLa cells were cultured in
DMEM supplemented with 10% fetal bovine serum, 2 mM glutarnine and
penicillin/streptomycin. Transfections were performed using the calcium phosphate
technique or using LipofectAMINE 2000 (Invitrogen), following manufacturer’s

instructions. MCF-7/iFLAG-MLK3 cells inducibly expressing F lag-MLK3 have been

51

described elsewhere (Zhang et al., 2004), and were maintained in DMEM supplemented
with 5% fetal bovine serum, 40 ug/mL neomycin and 10 pg/mL hygromycin. MCF7
cells inducibly expressing Flag-Estrogen Receptor were kindly provided by Dr. Susan
Conrad (Michigan State University). F lag-MLK3 expression as well as Flag-Estrogen

Receptor expression was induced by the addition of 50 nM AP21967 for 20 h.

3.4 Cell lysis and immunoprecipitation

Cells were washed with ice cold PBS and lysed by addition of lysis buffer (50
mM HEPES (pH 7.5), 150 mM NaCl, 1.5 mM MgC12, 2 mM EGTA, 1% Triton X-100,
10% glycerol, 10 mM NaF, 1 mM Na4PPi, 100 1.1M B-glycerophosphate, 1 mM Na3VO4,
2 mM PMSF) complemented with a cocktail of protease inhibitors (Sigma). Lysates
were clarified by centrifugation for 15 min at 14,000 rpm at 4°C. For
immunoprecipitations, antibodies against the proteins of interest were prebound to
Protein A-agarose beads at room temperature for 30 min and incubated with clarified cell
lysates for 90 min at 4°C. The immunoprecipitates were washed with HNTG buffer (20

mM HEPES (pH 7.5), 150 mM NaCl, 0.1% Triton X-100, 10% glycerol).

3.5 Gel electrophoresis and western blot analysis

Cellular lysates and immunoprecipitates were separated by SDS polyacrylamide
gel electrophoresis. Proteins were transferred to nitrocellulose membrane and
immunoblotted using appropriate antibodies. Western blots were developed by the

chemiluminescence method as described previously (Bock et al., 2000).

52

3.6 Pro-Q Diamond and Silver Staining

HEK 293 cells were transfected with vectors encoding MLK3 or MLK3 K144M
as described. MLK3 was immunoprecipitated as described above and resolved by SDS-
PAGE. The gel was stained with Pro-Q Diamond stain (Invitrogen) following the
manufacturer’s instructions. The image was acquired using a Molecular Irnager FX Pro
Plus scanner and PDQuest v7.4 (Bio-Rad). The gel was then silver stained using
SilverQuest (Invitrogen) following manufacturer’s instructions. Quantitations were done

by densitometry (ImageJ, http://rsb.info.nih.gov/ii/).

 

3.7 In viva labeling and phosphopeptide mapping

Twelve hours after transfection, HEK 293 cells were washed twice with
phosphate-free medium (Dulbecco's modified Eagle's medium supplemented with 10%
dialyzed fetal bovine serum (Summit Biotechnology», and incubated at 37 °C for 1 h.
The cells were then pre-treated with the different MAPK pathway inhibitors (40 pM
SP600125; 6 11M SBZO3580; 10 uM U0126) or vehicle DMSO for 1 h before addition of
l mCi/ml [32P]-orthophosphate (PerkinElmer Life Sciences, Inc), and firrther incubated
for4 h at 37°C. Cells were washed twice with ice-cold PBS and then lysed as described
above. MLK3 was immunoprecipitated as described above and resolved by SDS-PAGE.
Incorporation of radiolabel into MLK3 was quantified by phosphorimaging, and
normalized to the MLK3 expression levels, determined by densitometry.

Radiolabeled bands were excised from the gel, transferred to siliconized

eppendorf tubes and rehydrated with 4 p1 of water. The gel pieces were washed twice

with 500 pl of 0.1 M NH4HCO3, 50% acetonitrile at 30°C for 50 min. The gel pieces

53

were completely dried using a SpeedVac, partially rehydrated with 4 pl of 0.1 M
NH4HC03, 0.02% Tween 20, and incubated with 40 pl of 50 mM NH4HC03 containing 2
pg of trypsin (Roche) overnight at 30°C. The supernatant was then collected, and the gel
pieces were sonicated in 100 ul of a 60% acetonitrile, 0.1% trifluoroacetic acid solution,
for 30 min. The supernatant was collected and the sonication repeated in 50 pl of the
same solution. The supematants were combined and concentrated using a speedvac. The
peptides were separated on cellulose thin layer chromatography (TLC) plates by thin
layer electrophoresis (TLE) in the first dimension in a pH 1.9 buffer (formic acid/glacial
acetic acid/water, 25:78:897, v/v/v) at 0°C and 1000 V for 30 min, and separated by TLC
in n-butanol/pyridine/glacial acetic acid/water, 15:10:3z12, v/v/v/v. The radiolabeled

phosphopeptides were visualized and quantitated using a Phosphorlrnager.

3.8 In vitro phosphorylation of MLK3

HEK 293 cells were transfected with a vector encoding MLK3 K144M. Twenty
hours after transfection the cells were washed and lysed in Lysis buffer (50 mM Tris pH
7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100). MLK3 K144M was
immunopurified from clarified cellular lysates. Kinase reactions were carried out in 50
pl Kinase buffer (50 mM Tric-Cl pH 7.5, 0.1 mM EGTA, 10 mM MgC12, 0.1% B-
mercaptoethanol) containing 10 pM ATP, 10 pCi [y-32P]-ATP (3000 Ci/mmol,
PerkinElmer Life Sciences, Inc) and 0.5 pg (57 mU) of recombinant JNKlocl (Upstate),
at 37°C for 30 min. Proteins were fractionated by SDS-PAGE, and the incorporation of

radioactivity into MLK3 was measured by phosphorimaging.

54

3.9 Northern blot analysis

F lag-MLK3 expression was induced by incubation of MCF-7/iFLAG-MLK3 cells
with 50 nM AP21967 for 20 h. Cells were then incubated with 15 pM SP600125 for
different periods of time. Total RNA was extracted using TRIzol (Invitrogen) following
manufacturer’s instructions. Forty micrograms of RNA were fractionated on a 1%
fonnaldehyde/agarose gel and transferred to a nitrocellulose membrane. pRK-MLK3
was digested with BstEII to generate a probe encompassing nucleotides 198-1836 of the
MLK3 gene. The random primer DNA labeling system (Invitrogen) was used together
with 50 pCi of [or-32P]dCTP to radiolabel the probe, following manufacturer’s
instructions. The membrane was pre-hybridized in 15 mL of pre-hybridization solution
(50% forrnamide, 5X SSPE (20X SSPE: 3M NaCl, 0.23M NaHzPO4, 0.02M EDTA, pH
7.7), 2X Denhardt’s, 0.1% SDS, 100 pg/mL salmon sperm DNA) 1 h at 42°C. The
radiolabeled probe was then added to the solution and the membrane was incubated
overnight at 42°C. The membrane was then washed once at room temperature in 1X SSC
(20X SSC: 3M NaCl, 0.3M sodium citrate, pH 7.0)/0.l % SDS, and twice at 42°C in
0.2X SSC/0.1% SDS. The radiolabeled F lag-MLK3 mRNA was visualized by

phosphorimaging.

55

4. Results
4.1 JNK Is Required for in viva Phosphorylation of MLK3

Using mass spectrometry coupled with comparative phosphotryptic peptide
mapping, we previously identified eleven in viva phosphorylation sites on MLK3 (Figure
1) (Vacratsis et al., 2002). In our strategy, HEK 293 cells expressing MLK3 with its
activator Cdc42Vlz for 16 h were subsequently labeled with [32P]-orthophosphate for an
additional 4 h. After affinity purification and trypsin digestion of MLK3, the resultant
radiolabeled phosphotryptic peptides were resolved by chromatography and their
sequences were determined using a combination of MALDI-MS and LC-MS/MS. Thus
only those sites which are relatively labile (undergo dephosphorylation) and incorporate
substantial radiolabel (undergo rephosphorylation) during the 4 h labeling period could
potentially be identified.

MLK3 contains 27 potential proline-directed kinase sites, most of which are
found in the COOH-terminal region, which is rich in proline, serine, and threonine
(Figure 1) (Gallo et al., 1994). Examination of the identified phosphorylation sites
revealed that seven of the eleven sites conform to the consensus sequence for
phosphorylation by proline-directed kinases; and six of the seven sites reside in the
COOH-terminal region. This suggests that MLK3 is phosphorylated by proline-directed
kinases. Additionally we have found that, whereas wild type MLK3 incorporates
substantial radiolabel under our experimental conditions, a kinase-defective MLK3
variant incorporates very little radiolabel (P.O. Vacratsis and K.A. Gallo, unpublished

data).

56

S705

 

 

 

 

 

 

 

 

 

 

8724
$555 727
S524 8558 $654 8740
5&3?
197%:
l SH3 Kinase Domain Zippeii E SerrfhrIPro rich
1 4’3 103115 3134400 462492506 832 847

 

RGSSSGTPKLIQRALLRGTALLASLGLGRDLQPPGGPGRERGESPTTPPTPTPAP
CPTEPPPSPLICFSLKTPDSPPTPAPLLLDLGIPVGQRSAKSPRREEEPRGGTVSP
PPGTSRSAPG TPGTPRSPPLGLISRPRPSPLRSRIDPWSFVSAGPRPSPLPSPQP
APRRAPWTLFPDSDPFWDSPPANPFQGGPQDCRAQTKDMGAQAPVtNPEAGP

 

 

 

Figure 1. Schematic representation of MLK3. The Ser residues of the identified
phosphorylation sites on MLK3 are indicated above. Numbers below the diagram
correspond to amino acid numbers. The amino acid sequence of the
proline/serine/threonine-rich region is shown below the diagram. TP and SP motifs are

highlighted in bold italic. Identified phosphorylation sites are highlighted in red.

57

Total phosphate incorporation into MLK3 was determined using the Pro-Q
diamond dye. This fluorescent stain detects phosphate groups attached to serine,
threonine or tyrosine residues in a protein (Schulenberg et al., 2004; Steinberg et al.,
2003). MLK3 or a kinase inactive version MLK3 K144M was expressed in HEK 293
cells. Irnmunopurified MLK3 variants were resolved by SDS-PAGE. After staining the
gel with Pro-Q Diamond, densitometry was used to quantitate the relative incorporation
of Pro-Q Diamond stain into the MLK3 variants (Figure 2). Afier correcting for
expression levels, the extent of Pro-Q diamond staining was determined to be 40% less
for kinase-defective MLK3 than for wild type MLK3. The increased phosphorylation of
wild type MLK3 over that of inactive MLK3 is consistent with MLK3
autophosphorylation and/or feedback phosphorylation emanating from active MLK3.
Given that MLK3 is implicated in the activation of multiple MAPK pathways, it is
reasonable to hypothesize that, upon MLK3-induced activation, MAPKs might
phosphorylate MLK3.

To decipher which MAPK(s) might contribute to proline-directed phosphorylation
of MLK3, the impact of MAPK pathway inhibitors on the incorporation of radiolabeled
phosphate into MLK3 was determined using the same conditions that were used in the
original mass spectrometry studies. Concentrations of SP600125, U0126, and SB203580
required for inhibition of JNK, ERK (through inhibition of its upstream activator MEK),
and p3 8, respectively, were experimentally determined in HEK 293 cells coexpressing
MLK3 and activated Cdc42. Cells were pretreated with MAPK pathway inhibitors prior
to and during in viva labeling and, afier affinity purification and SDS-PAGE, the

incorporation of radiolabel into MLK3 was quantitated by phosphorimaging. Based on

58

Pro-Q Stain Silver Stain
MLK3 wt K144M wt K144M
. . m

 

Relative Phosphorylatlon
s a e a g

0

wt K144M

Figure 2. Phosphorylation levels of MLK3. Ectopically expressed MLK3 and MLK3
K144M were immunopurified from HEK 293 cells and resolved by SDS-PAGE. Pro-Q
Diamond staining was used to detect total phosphorylation levels (top panel, left). Total
amounts of proteins were detected by silver staining of the same gel (top panel, right).
Asterisk denotes the position of the MLK3 band. Data from three independent
experiments were normalized to total protein levels and quantitated as described under

“Materials and Methods” (bottom panel).

59

three independent experiments, incorporation of radiolabel into MLK3 was reduced by
70% in the presence of SP600125 (Figure 3). In contrast, inhibition of the p38 and ERK
pathways does not reduce in viva phosphorylation of MLK3 (Figure 3). In fact, treatment
with 88203580 enhances the phosphorylation of MLK3 by about 2-fold. This may
reflect the established antagonism between p38 and JNK signaling (Hall & Davis, 2002).
Phosphotryptic peptide maps of MLK3 isolated from cells treated with each of the
MAPK pathway inhibitors separately, as well as all three together, were generated. The
radiolabeled MLK3 was excised from the gel and trypsin-digested, and identical cellular
equivalents of the phosphotryptic peptides were resolved by TLE/TLC, and radioactive
phosphopeptides were detected by phosphorimaging. Phosphotryptic peptide maps from
the different samples were processed and imaged in parallel so that the intensities of spots
between maps could be directly compared. Treatment of cells with SP600125, but not
with other MAPK pathway inhibitors, yielded phosphopeptide maps that showed a
decrease in intensity of essentially all of the spots corresponding to identified
phosphorylation sites (Figure 4). These findings are not altogether unexpected;
phosphopeptide mapping analysis of a mutant version of MLK3, in which six of the
serines phosphorylated on Ser-Pro motifs previously identified by mass spectrometry
were mutated to alanine (MLK3 6A), indicated that proline-directed phosphorylation
accounts for most of the phosphorylation detected on MLK3 when coexpressed with
activated Cdc42 (Figure 5). Thus, the activity of JNK, but not of other MAPKs, is
apparently responsible for the dynamic COOH terminal proline-directed phosphorylation

of MLK3.

60

SP600125
I ' 88203580

U0126

All three

I"

[Mm-MLK3

3 ‘3 omso

- . . g - MLK3 IPIMLK3 we

.- .. - . . FIag-Cdc42v‘2wa

OJ

N
or

I

_l

    

Relative incorporation of [’2P]
into MLK3
o ..L
'01 an N

 

 
 

 

 

 

H u, " E

DMSO SP600125 88203580 U0126 All three

 

Figure 3. Effect of MAPK pathway inhibitors on in viva phosphorylation of MLK3.
HEK 293 cells transiently coexpressing MLK3 and Flag-Cdc42V12 were pre-incubated in
phosphate-free medium containing either SP600125, SB203580, U0126, the combination
of all three, or carrier (DMSO) as described in “Materials and Methods”. After in viva
labeling with [32P]-inorganic phosphate for 4 h in the presence of the MAPK pathway
inhibitors, MLK3 was immunoprecipitated from cleared lysates and resolved by SDS-
PAGE. T op panel, autoradiogram of immunoprecipitated MLK3. Middle panel, western
blot of a fraction of the immunoprecipitated MLK3 using an MLK3 antibody. Bottom

V” in cellular lysates using a Flag antibody.

panel, western blot of Flag-Cdc42
Incorporation of radioactivity into MLK3 was quantified by phosphorimaging and
normalized to MLK3 expression levels as described in “Materials and Methods”. The

means :1: SE. of three independent experiments are shown.

61

Figure 4. Phosphotryptic peptide maps of MLK3 from cells metabolically labeled in
the presence of MAPK inhibitors. Radiolabeled MLK3 from HEK 293 cells incubated
with the different MAPK inhibitors was fractionated by SDS-PAGE and the bands
corresponding to MLK3 were excised from the gel and subjected to trypsin digestion.
Portions of the recovered peptides corresponding to equal cellular equivalents were
analyzed by thin-layer electrophoresis (TLE) in the first dimension and thin-layer
chromatography (TLC) in the second dimension. Identical exposure times of the
phosphopetide maps were analyzed by phosphorimaging. + and — indicate positive and

negative poles, respectively.

62

+ DMSO

 

 

 

 

 

 

   

 

 

 

+SP600125
o N o
r—‘ .. _ ,-_I ...
W
" TLE ‘ + TLE
+53203580 +U0126
1.;
0 o ht
'- I— ‘.,-
“F
+ TLE ' + TLE
All three
0
_l
p.
+ TLE ‘

 

63

MLK3 wt 6A

 

 

 

 

.... , . [”P1- MLK3
MLK3+Cdc42m MLK3 6A+Cdc42m
U U
..l _.l
l- i-
* TLE ' ” TLE '

Figure 5. Phosphopeptide mapping of MLK3 6A. HEK 293 cells were transfected
with expression vectors for MLK3 or MLK3 6A, together with activated Cdc42. Cells
were in viva labeled as described for Figure 3, except that the inhibitors were omitted.
MLK3 was immunopurified from clarified lysates and fractionated by SDS-PAGE. T ap
panel, autoradiogram of immunoprecipitated MLK3. Bottom panels, phosphopeptide

maps were developed and analyzed as described for Figure 4.

64

4.2 Manipulation of JNK Activity Alters the Phosphorylation of MLK3

Coexpression with activated Cdc42 retards the electrophoretic mobility of MLK3
due to phosphorylation at sites other than in the activation segment (Du et al., 2005). The
reduced in viva phosphorylation of Cdc42-activated MLK3 observed upon treatment of
cells coexpressing activated Cdc42 and MLK3 with SP600125 is correlated with a more
rapidly migrating, hypophosphorylated form of MLK3 (Figure 3). Because SP600125
may target other protein kinases in addition to JNK (Bain et al., 2003), I sought
alternative means to manipulate cellular JNK activity so as to provide further support for
the idea that JNK phosphorylates MLK3 in viva.

MLK3 activates the JNK pathway through phosphorylation of MKK4 (Rana et
al., 1996) and MKK7/JNKK2 (Toumier et al., 1997). Catalytically inactive versions of
MKK4 (MKK4 K129R) and MKK7/JNKK2 (MKK7 K165A) can function as dominant-
negative proteins to block JNK activity in cells. The blockade of JNK activation is most
efficient when the inactive versions of MKK4 and MKK7 are coexpressed (Liu et al.,
2000a). Therefore increasing amounts of expression vectors for MKK4 K129R and
MKK7 K165A were simultaneously transfected into HEK 293 cells, together with
vectors encoding MLK3 and activated Cdc42. Upon introduction of increasing amounts
of dominant negative MKK4 and MKK7, efficient inhibition of the Cdc42-MLK3-
induced JNK activity in cells was achieved, as judged by the decrease in phospho-c-Jun
levels (Figure 6A, second panel). Under these conditions of efficient JNK inhibition, the
electrophoretic mobility of MLK3 was restored to that observed for MLK3 in the absence
of Cdc42 (Figure 6A, first panel). These results are similar to the effects of SP600125 on

MLK3, suggesting that the effects of SP600125 are, indeed, mediated through inhibition

65

Figure 6. JNK phosphorylation of MLK3. A, HEK 293 cells were cotransfected with

V
212

expression vectors for MLK3 (0.5 pg), Flag-Cdc4 (1 pg), and increasing amounts of

Flag-MKK7/JNKK2 K165A and GST-MKK4 K129R (lanes 4, 5 and 6; 10,15 and 20 pg
each, respectively). Cellular lysates were analyzed by western blotting using the
indicated antibodies. B, HeLa cells were cotransfected with expression vectors for HA-
MLK3, HA-JNKKZ-JNKI and GF P. Twenty four hours post-transfection the cells were
lysed, and clarified lysates were analyzed by western blotting using the indicated
antibodies. C, in vitro JNK phosphorylation of MLK3. Irnmunopurified MLK3 K144M
was incubated with [32P]-ATP in the presence (lane 1) or absence (lane 2) of
recombinant, activated JNKlal. Data shown is representative of three independent

experiments.

66

A

GST-MKK4 K129R A
FIag-JNKK2 K165A
Flag-Cdc42"12
MLK3

I
I
4.
+
+
...

u
+
+
+
+
...

.‘ifi it! MLK3WB

- - .. .. @c-Jun we

.. - .. .. .-- Flag-Cdc42 WB

...... M" W

HA-JNKK2-JNK1- + ' +
HA-MLK3 - ' '1' +

.— a! MLK3 WB
.- III HA-JNKKZ-JNK1WB

---- GFP WB

MLK3 K144M
JNK10t1 - +

 

ea. _, [azpl-MLK3

MLK3 IPI
“" m MLK3 WB

67

of JNK, and that the retarded mobility of MLK3 observed upon coexpression with Cdc42
is mediated, at least in part, through JNK phosphorylation.

In a complementary approach, I sought to determine whether selective activation
of the JNK pathway would promote phosphorylation of MLK3. However, the numerous
extracellular stimuli that activate the JNK pathway invariably activate other MAPK
pathways as well. The only MKK that has been shown to activate exclusively the JNK
pathway is MKK7/JNKK2 (Toumier et al., 1999). Mutant forms of the appropriate
upstream MKKs (MEKl , and MKK3 and MKK6), in which the activation loop sites have
been replaced with acidic residues to mimic phosphorylation, can selectively activate
ERK and p38, respectively (Cowley et al., 1994; Huang et al., 1997) whereas the
corresponding mutant of MKK7/JNKK2 is not highly active (Zheng et al., 1999).
However fusion of MKK7/JNKK2 to JNKl results in a constitutively active JNK when
expressed in cells (Zheng et al., 1999). To determine whether selective activation of JNK
promotes MLK3 phosphorylation, HeLa cells were transfected with expression vectors
for MLK3 and the fusion protein JNKKZ-JNKI. As shown in Figure 6B, the introduction
of activated JNK in cells results in the appearance of slower migrating forms of MLK3,
indicative of phosphorylation of MLK3.

Since the majority of the identified phosphorylation sites of MLK3 conform to the
consensus for proline-directed kinases, the simplest explanation for my data is that JNK
directly phosphorylates MLK3 in cells. To determine whether recombinant, activated
JNK can directly phosphorylate MLK3 in vitro, it was necessary to eliminate the
background of MLK3 autophosphorylation, and therefore a kinase-defective version of

MLK3 (K144M) was expressed, immunopurified from HEK 293 cells, and used as a

68

substrate for recombinant, active JNK in an in vitro kinase assay. As shown in Figure
6C, JNK phosphorylates MLK3 K144M in vitro, and this phosphorylation results in a
decreased electrophoretic mobility of MLK3 (Figure 6C, bottom panel). These data

support the hypothesis that MLK3 is a direct substrate of JNK.

4.3 Feedback Phosphorylation Regulates MLK3 Protein Levels in Cellular Lysates

One intriguing observation in the experiments described above was an apparent
correlation between the cellular levels of activated JNK and the levels of MLK3 protein
present in cellular lysates. As shown in Figure 6A, when JNK activity is diminished, less
MLK3 is detected. This effect is specific to MLK3 since the levels of Flag-Cdc42 in
cellular lysates, which is expressed using an identical vector, are unchanged in response
to JNK inhibition. Likewise the amount of MLK3 in cellular lysates (Figure 6B)
increases when the corresponding cells express the constitutively active JNK fiision
protein, whereas the levels of GP P used as a transfection control remain constant. These
observations suggest the existence of a positive feedback loop involving MLK3-mediated
JNK activation and JNK-mediated phosphorylation of MLK3, leading to increased
hyperphosphorylated, active MLK3 in cellular lysates and sustained activation of the
JNK pathway.

The JNK-phosphorylated MLK3 is activated, as judged by activation loop
phosphorylation. Activation loop phosphorylation due at least in part to
autophosphorylation (Du et al., 2005), COOH-terminal phosphorylation by JNK, or both,
may be required for the increased levels of MLK3 in cellular lysates. As shown

previously (Du et al., 2005), Cdc42 promotes activation loop phosphorylation of wild

69

type MLK3, increases MLK3 protein levels in cellular lysates, and potentiates JNK
activation (Figure 7A). However, catalytically inactive MLK3 K144R is refractory to
regulation by Cdc42 (Figure 7A), indicating that the enhanced levels of MLK3 in cellular
lysates caused by coexpression with the activated GTPase require activation loop
phosphorylation and/or JNK activation. To dissociate activation loop phosphorylation,
which requires MLK3 kinase activity, from JNK-mediated phosphorylation of MLK3, the
constitutively active JNKK2-JNK1 fusion protein was employed. As shown in Figure
7B, expression of the constitutively active JNKKZ-JNKI fusion protein increases the
protein levels in cellular lysates and retards the mobility, of both wild type and kinase-
defective MLK3 indicating that feedback phosphorylation by JNK, rather than activation
segment phosphorylation per se, is the critical determinant in dictating JNK-mediated

enhanced levels of MLK3 in cellular lysates.

4.4 Inhibition of JNK Reduces the Levels of Activated MLK3 in Cellular Lysates

To study the regulation of MLK3 by JNK phosphorylation in a more
reproducible, homogeneous system, a stable clone of a human breast cancer cell line,
termed MCF7/iFlag-MLK3, was used (Zhang et al., 2004). In this cell line, transcription
of Flag-MLK3 is controlled by the small molecule dimerizer AP21967. The induced
Flag-MLK3 potently activates the JNK pathway (Zhang et al., 2004). Consistent with the
high activity of MLK3 in this system, the electrophoretic mobility of MLK3 is retarded to
a degree comparable to that observed when activated Cdc42 is transiently coexpressed

with MLK3 (Figure 8A).

70

A
MLK3 wt K1448

 

Flag-Cdc42” - + - +

....— ._....... MLK3 we
5; 2 281

~- ... #8 ®IllrLK3 e
—‘

...... ~ @JNK we

a- .... FIag-Cdc42m we

- — — - Actin WB

B
MLK3 wt K1448

HA-JNKKZ-JNKl - + - +

 

f“ - - MLK3 we
- .— HA-JNKKZ-JNK1 we

— .— @JNKwe

--- II- Actin WB

Figure 7. Regulation of the levels of MLK3 variants in cellular lysates in response to
modulation of JNK signaling. HeLa cells transiently coexpressing the indicated MLK3
variants and JNK signaling modulators were harvested and cellular lysates were prepared.
The status of MLK3 activation loop phosphorylation and of JNK activation was assessed
by western blotting of cellular lysates using appropriate antibodies. A, HeLa cells were
transfected with vectors for MLK3 or MLK3 K144R with or without Flag-Cdc42v'2. B,
HeLa cells were cotransfected with expression vectors for MLK3 or MLK3 K144R with
or without HA-JNKKZ-JNKI. Levels of expressed proteins, and of endogenous actin as
a loading control, were detected by western blotting using the indicated antibodies. Panel

A is provided by Yan Du, Gallo lab.

71

Figure 8. Electrophoretic mobility of activated MLK3, and impact of
pharmacological inhibition of JN K on MLK3 levels in cellular lysates. A,
Electrophoretic mobility of Flag-MLK3. The electrophoretic mobility of Flag-MLK3
expressed alone or with Flag-Cdc42VIZ in HEK 293 cells and F lag-MLK3 inducibly
expressed in engineered MCF-7 cells (MCF7/iFlag-MLK3) was assessed by western
blotting of cellular lysates after SDS-PAGE. B, F lag-MLK3 expression was induced
using AP21967 in MCF7/iFlag-MLK3 cells for 20h. Cells were then treated with various
concentrations of SP600125 for 2h. Clarified lysates were subjected to western blotting
analysis using the indicated antibodies. C, MCF7/iFlag-MLK3 cells inducibly expressing
Flag-MLK3 as described in B were treated with 15 pM SP600125 or 400 nM CBP-11004
for the indicated times. Clarified lysates were subjected to western blotting analysis

using the indicated antibodies.

72

HEK293
MCF7/Flag
MLK3

 

 

Flag-Cdc42“: - +

H u M Flag-MLK3 we

 

B
+AP21967
SP600125 - - 5 1° 15 20
(W)
E al- ., a... . Flag-MLK3 we
2 281
I @JLKS e
“---—I— ActanB
C
+AP21967

 

SP600125 CBP-11004
- - 3O 60 120 30 60 120 min

- - .- .. .— .. .. Flag-MLK3 we
2 281
— ®JLK3 e
*w-ga-_ @C-JUDWB

------‘- ActinWB

73

To test the effects of JNK inhibition on MLK3, MCF7/iFlag-MLK3 cells were
induced for 20 h with AP21967, washed to remove the inducer, and then treated with
different concentrations of SP600125 for 2 h. Figure 8B shows that inducibly expressed
Flag-MLK3 is active as judged by western blotting with an antibody specific for
activation loop phosphorylation of MLK3. The levels of Flag-MLK3 in cellular lysates
are dramatically decreased upon treatment with the JNK inhibitor, in a concentration
dependent manner. Interestingly, the activation loop phosphorylated form seems to
disappear more quickly than the net Flag-MLK3, which may indicate that the active form
of MLK3 is most sensitive to regulation by JNK.

Recently a pharmacological inhibitor of the MLK family, CBP-11004, has been
shown in animal models to attenuate JNK-induced apoptotic death of dopaminergic
neurons of the substantia nigra (Ganguly et al., 2004). Therefore, CBP-11004 was used
in our inducible expression system to corroborate the effects on MLK3 protein levels
observed using the JNK inhibitor SP600125. MCF-7 cells inducibly expressing F lag-
MLK3 were incubated with 400 nM CEP-11004 and, as shown in Figure 8C, the levels of
phosphorylated c-Jun are decreased dramatically, indicating that JNK activity is potently
inhibited. In addition, CBP-11004 reduces the levels of activated and total MLK3 protein
in cellular lysates, in a similar manner to that of the JNK inhibitor. The effect of
SP600125 appears to be specific for MLK3, since it had no impact on the levels of a
F lag-tagged estrogen receptor inducibly expressed in MCF-7 cells using the identical
expression system and vectors (Figure 9A). Northern blots using mRNA derived from

the MCF-7 cells inducibly expressing Flag-MLK3 showed that SP600125 did not alter

74

A

AP21967 AP21967
SP600125 (15 pM) - - 1 - - 1 (n)
-- MLK3

.._._ H.- ....a Flag WB

—-‘ ER

~—‘ “-w-~~ ActinWB

B
AP21967
SP600125(15pM) - - 30 60 120 (min)
.... ~— .. ..__, e. . MLK3

a swat-00.0%. 23$
. il~ 18$

55

 

Figure 9. Effect of SP600125 on inducibly expressed Flag-ER and on Flag-MLK3

mRNA levels. A, F lag-MLK3 expression and Flag-estrogen receptor expression were

induced using AP21967 for 20 h. Cells were then treated with 15 pM SP600125 for l h.

Clarified lysates were subjected to western blotting analysis using the indicated
antibodies. B, Northern blot analysis of Flag-MLK3 in MCF7/iFlag-MLK3. F lag-MLK3
expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. Cells were
then treated with 15 pM SP600125 for the indicated times. Total RNA was then

extracted and analyzed by northern blotting. The ethidium bromide-stained gel,

displaying the 28S, 18S and 5S ribosomal RNA bands, is shown as a loading control.

75

the mRNA level of F lag-MLK3 (Figure 9B). These data argue for posttranslational

modulation of activated MLK3 by JNK.

4.5 Inhibition of Other Proline-Directed Kinase Pathways Has No Impact on MLK3
Levels

In addition to the MAPKs, Cdks and GSK-3 are proline-directed kinases. To test
whether other MAPKs might directly or indirectly regulate protein levels of MLK3, the
p38 and ERK pathways were blocked using pharmacological inhibitors. As shown in
Figure 10A, neither the p38 inhibitor, SBZO3580, nor the MEK inhibitor, U0126, altered
the levels of Flag-MLK3 protein, in agreement with the results from the in viva labeling
experiment. In addition, reactivity with the MLK3 activation loop phosphorylation-
specific antibody was maintained, indicating that active MLK3 was unaffected.
Likewise, inhibition of the Cdks with roscovitine (Meijer et al., 1997) or of GSK-3 with
kenpaullone (Leclerc et al., 2001) failed to impact levels of total MLK3 or active MLK3
found in cellular lysates (Figure 10B). Together these data support the idea that
inhibition of JNK, but not of other proline-directed kinases, reduces the level of active

MLK3.

4.6 Dephosphorylation of MLK3 Reduces Its Levels in Cellular Lysates
My data provide evidence for positive feedback phosphorylation of MLK3 by
JNK. I surmise that the JNK phosphorylation sites are relatively labile, given that they

incorporate substantial radioactivity during a 4 h labeling period. This raises the idea that

76

+AP21967

 

SP SB U0126
- 30 120 30 120 30 120 min

 

 

 

- a: a. i ii - Flag-MLK3 we

-? -i.. @1'2 281
MLK3 e

-----_‘- ActinWB

+AP21 967

 

SP Rose Kenp
- 3012030120 30120 min

‘ 6 fl - ... Flag-MLK3 we

I“...

-------._ Actln WB

 

 

 

 

2 281
MLK3 B

 

Figure 10. Effect of proline-directed kinase inhibitors on MLK3 levels in cellular
lysates and activation loop phosphorylation. A, F lag-MLK3 expression was induced in
MCF7/iFlag-MLK3 cells using AP21967 for 20h. Cells were then treated with 15 pM
SP600125 (SP), 6 pM SB203580 (SB) or 10 pM U0126 for the indicated times.

Clarified lysates were analyzed by western blotting using appropriate antibodies. B,
Same as in A, but cells were treated with 15 pM SP, 10 pM Roscovitine (Rosc) or 10 pM

Kenpaullone (Kenp). and analyzed as described above.

77

JNK and an opposing phosphatase target the same COOH-terminal phosphorylation sites
and that MLK3 levels are regulated by the balance between the two. JNK inhibition
would be expected to shift the balance towards dephosphorylation and reduction of
MLK3 in cellular lysates. If COOH-terminal dephosphorylation signals to downregulate
active MLK3 levels, inhibition of the responsible phosphatase should protect MLK3
levels from JNK inhibition. To test this hypothesis, the Flag-MLK3-expressing cells
were treated with calyculin A, a broad-spectrum Ser/T hr phosphatase inhibitor (Dawson
& Holmes, 1999; Millward et al., 1999), together with the JNK inhibitor. As predicted,
calyculin A blocked the effect of SP600125 on MLK3 protein levels and MLK3 activity
(Figure l 1), consistent with the model that dephosphorylation of JNK phosphorylation

sites somehow downregulates the levels of MLK3 found in cellular lysates.

4.7 Dephosphorylation of MLK3 upon JNK inhibition does not result in its
degradation

Transcription factors, such as c-Jun and ATP 2, have been reported to be stabilized
by JNK phosphorylation (Fuchs et al., 1998; Fuchs et al., 1997; Musti et al., 1997).
Recent evidence indicates that proteasome-mediated degradation of c—Jun by specific E3
ligases is dependent on JNK-mediated phosphorylation (Laine & Ronai, 2005). For
instance, in PC12 cells, the E3 ligase SCFFbW7 interacts with c-Jun once c-Jun has been
phosphorylated by JNK, promoting its degradation (Nateri et al., 2004). This mechanism
is thought to allow the cells to tolerate the neurotoxic effects of activated JNK. Another
E3 ligase that promotes c-Jun degradation is Itch (Gao et al., 2004). In this case,

activation of JNK in T cells results in phosphorylation of Itch and degradation of c-Jun

78

+AP21967
CalA(50nM) - . . ...

SP(15pM) - - + +

h ...... H Flag-MLK3 WB

.... @T2 281

‘ MLK3 e

"" @c-Jun we

... _ — .- Actin WB

Figure 11. Impact of phosphatase inhibition on MLK3 levels on cellular lysates.

F lag-MLK3 expression was induced in MCF7/iFlag-MLK3 cells using AP21967 for 20
h. Cells were then treated with 15 pM SP600125 for 1 h, in the presence or absence of
50 nM calyculin A. MLK3 protein levels and electrophoretic mobility, as well as
activation loop phosphorylation, were assessed by western blotting using the indicated

antibodies. JNK activation was assessed by western blotting for phospho-c-Jun levels.

79

and JunB, although it is not clear whether Itch recognizes phosphorylated c-Jun or not. It
is conceivable that the changes in MLK3 levels in cellular lysates observed upon JNK
inhibition are due to increased degradation of the dephosphorylated form of MLK3.
Ubiquitin-mediated degradation is commonly used to regulate protein turnover.
Therefore, I set out to investigate whether JNK inhibition results in ubiquitin-mediated
degradation of MLK3. Flag-MLKB-expressing cells were incubated with MG132, a
proteasome inhibitor, prior to treatment with SP600125. The prediction is that
ubiquitination of MLK3 induced upon JNK inhibition would result in the accumulation of
MLK3 in the cells; conversely, phosphatase inhibition, which preserves the
phosphorylated form of MLK3 even when JNK has been inhibited, would override the
effect of MG132. As shown in Figure 12, I did not observe a substantial increase in the
amount of MLK3 upon MG132 treatment. To assess whether MLK3 can be regulated
through ubiquitination, MLK3 and F lag-ubiquitin were ectopically expressed in HEK 293
cells. To (potentially) increase the accumulation of ubiquitinated MLK3, JNK was
pharmacologically inhibited with SP600125, either in the presence or absence of
lactacystin, another type of proteasome inhibitor. The ubiquitinated proteins were
immunoprecipitated from clarified lysates with a Flag antibody, and the presence of
ubiquitin-modified MLK3 was determined by western blot. In accordance with previous
results, SP600125 dramatically reduces the levels of MLK3 in cellular lysates, and
proteasomal inhibition does not protect MLK3 (Figure 13). Interestingly, a single band
corresponding to MLK3 was detected in the Flag pull down, suggesting that MLK3 is
monoubiquitinated. Recently, it has been recognized that while polyubiquitination

usually targets modified substrates for proteasome mediated degradation,

80

AP21967

MG132
CalyculinA - - - + - - +

SP600125 - - + + - + +
-..-....- FIag-MLK3 WB

 

- I... III ®T277@5281MLK3 we
a , , h.“i @c-Jun we

\
~ ~ ~ ~ ...... .- v ACtin WB

Figure 12. Effect of MG132 on the protein levels of MLK3 on cellular lysates. F lag-
MLK3 expression was induced in MCF7/iFlag-MLK3 cells using AP21967 for 20h.
Cells were incubated with 25 pM MG132 for 2h, and then treated with 15 pM SP600125
for 1h, in the presence or absence of 50 nM calyculin A. MLK3 protein levels and
electrophoretic mobility, as well as activation loop phosphorylation, were assessed by
western blotting using the indicated antibodies. JNK activation was assessed by western

blotting for phospho-c-Jun levels.

81

Figure 13. Ubiquitination of MLK3 in HEK 293 cells. MLK3 was transiently
expressed with or without F lag-ubiquitin in HEK 293 cells. After 24 h, cells were
harvested and ubiquitin-modified proteins immunoprecipitated from cellular
lysates were subjected to western blotting with MLK3 and Flag antibodies to
detect ubiquitinated MLK3 and total ubiquitination. An aliquot from the cellular
lysates was analyzed for the expression levels of MLK3, and the activation of the
JNK pathway was indirectly measured by western blotting of phospho-c-Jun

levels.

82

Lactacystin - - . -
SP600125 - - - +
Flag-ubiquitin + - + + +
m U ~ MLK3 WB
Lysates
N - ... @c-Junwe

- d MLK3 we

Flag-ubiquitin IP

Fiag-ubiquitin WB

 

 

83

monoubiquitination regulates location and activity of modified proteins, the best example
being the downregulation and internalization of monoubiquitinated plasma membrane
receptors into endocytic vesicles (Haglund et al., 2003; Hicke, 2001; Hicke & Dunn,
2003). These data altogether suggest that the reduction in MLK3 levels observed upon

JNK inhibition is not due to degradation through the 26S proteasome pathway.

84

5. Discussion

In this work I provide evidence that MLK3 is regulated by JNK-mediated COOH-
tenninal phosphorylation. These studies were prompted by earlier identification of
multiple, proline-directed in viva phosphorylation sites on Cdc42-activated MLK3
(Vacratsis et al., 2002). Since MLK3 acts as a MAPKKK to induce multiple MAPK
pathways, the possibility of feedback phosphorylation was addressed. Inhibition of JNK,
but not other MAPKs, dramatically reduced the dynamic, in viva phosphorylation of
MLK3. Activated JNK promotes MLK3 phosphorylation in viva and phosphorylates
MLK3 in vitro. A kinase defective version of MLK3 is less phosphorylated in cells than
wild type MLK3; furthermore, JNK-mediated phosphorylation of MLK3 requires MLK3
activity, consistent with the existence of a feedback loop.

Feedback phosphorylation of MAPKKKs by their downstream MAPKs is not
without precedence. For instance, in yeast, pheromone induces the three component
MAPK pathway of Stell (MAPKKK)-Ste7(MAPKK)-Fus3/Kssl(MAPK). Fus3/Kssl-
mediated feedback phosphorylation of active Stell targets this MAPKKK for
proteasome-mediated degradation (Esch & Errede, 2002). In the Raf-MEK-ERK
pathway, both B-Raf and Raf-l are negatively regulated by ERK phosphorylation
(Brummer et al., 2003; Wartmann & Davis, 1994; Wartmann et al., 1997). Recently,
ERK-mediated feedback phosphorylation was shown to inactivate and desensitize Raf-1,
rather than target Raf-1 for degradation. Hyperphosphorylated Raf is recycled through
dephosphorylation by PP2A after Pinl -dependent isomerization (Dougherty et al., 2005).

In contrast to these examples of negative feedback phosphorylation of MAPKKKs

by their downstream MAPKs, in the MLK3-MKK4/7-JNK pathway, our evidence

85

indicates that JNK-mediated phosphorylation does not downregulate or drive degradation
of active MLK3, but rather enhances its levels in cellular lysates. In Xenopus oocyte
maturation, progesterone treatment leads to the accumulation of the MAPKKK Mos
(Gotoh et al., 1995; Matten et al., 1996; Roy et al., 1996). This positive feedback of the
Mos-MEK-ERK pathway is manifested primarily through stabilization of mas mRNA
and accompanying increased translation of Mos. JNK has been reported to phosphorylate
other MAPKKKs. For instance, JNK binds and phosphorylates the MAPKKK, MEKKl,
in the MEKKl-MKK4-JNK (Xu & Cobb, 1997) pathway, but this apparently does not
regulate MEKKl stability. Interestingly, MEKKI firnctions not only as a MAPKKK and
scaffold, but also as an E3 ubiquitin ligase and undergoes ubiquitination. However,
ubiquitinated MEKKl is not rapidly degraded, but instead is rendered catalytically
inactive (Witowsky & Johnson, 2003). MLK2 can also be phosphorylated by JNK, but
precise phosphorylation sites and whether they impact MLK2 stability was not reported
(Phelan et al., 2001).

The incorporation of substantial radiolabel into multiple JNK phosphorylation
sites of Cdc42-activated MLK3 indicates that these sites are relatively labile. Inhibition
of phosphatase activity with calyculin A resulted in a marked reduction in the
electrophoretic mobility of MLK3, indicating that in fact, MLK3 is subjected to dynamic
cycles of phosphorylation-dephosphorylation. Interestingly, the activation loop
phosphorylation sites of MLK3 did not incorporate significant radiolabel under our
experimental conditions, as judged by phosphotryptic peptide mapping (Vacratsis et al.,
2002). This unexpected finding may indicate that the activation loop phosphorylation

sites are quite stable, and do not undergo rapid phosphatase-mediated exchange resulting

86

in meager incorporation of radiolabel. In response to JNK inhibition, the disappearance
of activation loop phosphorylated MLK3 precedes the loss of net MLK3. This is
consistent with the notion that activation loop phosphorylated MLK3 is inherently
unstable in the absence of additional COOH-terminal phosphorylation by JNK. I cannot,
however, rule out the possibility that an activation loop phosphatase for MLK3 exists, as
it has been described for ASKl (Morita et al., 2001 ), but I would hypothesize that it
would be inhibited by JNK.

A MLK3 variant in which the phosphorylated serines from the proline-directed
kinase sites located in the COOH-terminal region of MLK3 were substituted with alanine
residues has an incomplete phenotype: it not longer incorporates radioactivity during an
in viva labeling experiment, but yet can still be phosphorylated in vitro by JNK, although
to a much lesser degree than wild type MLK3 (data not shown). Likewise, it can be
phosphorylated in viva by JNK when using the fusion protein between MKK7 and JNK],
but again, to a lesser degree than wild type MLK3 (data not shown). In addition,
activated Cdc42 can still promote a slight decrease in its electrophoretic mobility (data
not shown). All these data suggest that there might be other feedback phosphorylation
sites on MLK3 that have not been identified.

My finding that JNK phosphorylation regulates the cellular levels of a MAPKKK
is novel. Interestingly, it had been previously reported that NGF-withdrawal or UV
treatment of neuronally differentiated PC12 cells results in an increase in the protein
levels of DLK, and in the appearance of slowing migrating forms that correspond to
phosphorylated DLK (Xu et al., 2001). Those studies have been extended, and it has

been recently shown that cellular stresses like camptothecin, a t0poisomerase inhibitor,

87

and sorbitol, an osmotic stresser, can induce activation and stabilization of endogenous
MLK3. Ablation of JNK expression by siRNA or with dominant negative JNKl blocked
camptothecin-induced increase in MLK3 protein levels (Xu et al., 2005). Changes in the
cellular levels of other MLK family members, as well as changes in the levels of the
scaffold proteins JIP1 and POSH were also observed upon modulation of JNK activation
status. Although changes in protein stability were not directly measured, the authors
observed changes in the levels of exogenous MLK3 driven by a constitutive promoter,
but not for a kinase defective form of MLK3 expressed under the same conditions,
concluding that the regulation of MLK3 levels is transcription-independent. In my
experimental system, inhibition of JNK reduces the levels of MLK3 in cellular lysates,
and northern blot analysis indicates that this is not regulated at the mRNA level. To be in
consonance with a model for the regulation of MLK3 stability through JNK-mediated
phosphorylation, blockade of JNK would be expected to increase MLK3 degradation.
This could be accomplished through ubiquitination and proteasome-mediated degradation
of dephosphorylated MLK3. Although MLK3 can be ubiquitinated when ectopically
expressed with ubiquitin in HEK 293 cells, I have not detected increased ubiquitination
of MLK3 upon JNK inhibition, even in the presence of proteasome inhibitors. These
observations indicate that instead of being degraded, MLK3 might have become
“unavailable”, perhaps sequestered in a different subcellular compartment that is resistant
to the cell lysis conditions.

In the next Chapter, I will address this question by looking at the subcellular

distribution of the dephosphorylated MLK3.

88

[11. Feedback Phosphorylation of MLK3 by JNK Regulates its

Distribution to Triton-insoluble Fractions

1. Abstract

Mixed-lineage kinase 3 (MLK3) is a widely expressed, mammalian
serine/threonine protein kinase that activates multiple MAPK pathways. Previously our
lab used in viva labeling/mass spectrometry to identify phosphorylation sites on activated
MLK3. Seven of eleven identified sites correspond to the consensus motif for
phosphorylation by proline-directed kinases. JNK phosphorylates MLK3 in vitro, and it
phosphorylates MLK3 on the COOH-terminal sites in viva. When JNK is inhibited,
MLK3 is rapidly dephosphorylated, consistent with the idea that phosphorylation of the
COOH-terminal sites occurs through a dynamic, JNK-mediated, feedback mechanism.
Herein I present evidence that the hypophosphorylated form of MLK3 is inactive, and
redistributes to a Triton-insoluble fraction. Recovery from JNK inhibition restores
MLK3 solubility and activity, indicating that the redistribution process is reversible.
Sucrose density gradient experiments demonstrate that the Triton-insoluble MLK3 is not
largely associated with lipid rafts. Immunofluoresence microscopy experiments indicate
that the dephosphorylated MLK3 is present in a vesicle-like pattern throughout the cells,
and does not colocalize with major organelle markers. My work describes a novel mode
of regulation of MLK3, by which JNK-mediated feedback phosphorylation of MLK3
regulates its activation and deactivation states by cycling between Triton-soluble and

Triton-insoluble forms.

89

2. Introduction

Our interest in the role of phosphorylation in regulating MLK3 activities led us to
undertake a mass spectrometry/phosphopeptide mapping approach to identify in viva
phosphorylation sites on MLK3. Cells expressing MLK3 and Cdc42 were radiolabeled
with inorganic phosphate, and the residues that incorporated radioactivity were identified
by mass spectrometry. Eleven phosphorylation sites were found, most of them clustered
at the COOH-terminal region of MLK3. Seven of these sites correspond to serine
residues that are followed by proline residues, a consensus sequence that is recognized by
the MAPKs. The work described in Chapter II shows that JNK phosphorylates MLK3 in
viva and in vitro, and that inhibition of JNK-mediated phosphorylation of MLK3 reduces
MLK3 protein levels in cellular lysates. I hypothesized that the decrease in MLK3 levels
could result from ubiquitination and proteasome-mediated degradation of the
hypophosphorylated MLK3. The data presented in Chapter 11 indicate that even in the
presence of a proteasome inhibitor, the levels of MLK3 in cellular lysates decrease when
JNK is blocked in my experimental setting. This raises the possibility that MLK3 had
redistributed to the Triton-insoluble fraction, which had not been previously analyzed for
the presence of MLK3.

Intracellular translocation of signaling proteins is one way to regulate their actions
and fate. Redistribution to different subcellular compartments results in their exposure to
different subsets of molecules, and thus adds specificity to cellular responses. For
instance, upon ligand stimulation of a plasma membrane receptor, signaling molecules
like Src family tyrosine kinases, scaffolds and adaptor molecules like Grb2 and SOS are

recruited to the cytoplasmic leaflet of the plasma membrane (reviewed in (Hunter,

90

2000)). The ability of the recruited proteins to generate productive signaling complexes
by increasing the local concentration of proteins at the membrane has been termed
induced protein proximity. The recruitment of selective signaling molecules to the
plasma membrane generates new micro-environments, which are also characterized by
the presence of specific lipids (Simons & Toomre, 2000). These specialized regions of
the plasma membrane are referred to as ordered lipid microdomains, and include rafts and
caveolae. Due to their lipid content, rafts and caveolae are insoluble in non-ionic
detergents, and are therefore also called detergent-resistant membranes (DRMs) (Morris
etal,2004)

In this chapter, the finding that dephosphorylated MLK3 redistributes to a Triton
X100-insoluble fraction and that it displays a punctate localization pattern in cells is
described. The nature of the Triton X100-insoluble fractions is analyzed and the

importance of this localization for MLK3 regulation is discussed.

91

3. Materials and Methods
3.1 Reagents and Antibodies

SP600125 was purchased from Calbiochem. Calyculin A was obtained from
BioMol. AP21967 was generously provided by Ariad Pharmaceuticals. Mouse
monoclonal antibodies against phospho—c-Jun (KM-1), as well as antibodies against
Hsp90 and Cdc3 7, were purchased from Santa Cruz Biotechnology, Inc. The phospho-
MLK3 antibody and the phospho-JNK antibody were from Cell Signaling Technology.
The monoclonal antibody against cytochrome oxidase subunit IV (COX IV) was from
Invitrogen. The antibody against ERp72 was purchased from Stressgen. The antibody
against flotillin-l was from BD-Transduction Labs. Antibodies against EEAI were
obtained from Affinity BioReagents (for immunoblotting) and from BD-Transduction
Labs (for immunofluorescence). Antibodies against the Golgi marker GM130, Integrin
02 and Nucleoporin p62 were from BD-transduction Labs. The antibody against clathrin
was from Affinity BioReagents. Antibodies against B-tubulin, y-tubulin and vinculin
were from Sigma. The antibody against Histone 3 was from Abcarn. The antibody
against p27 is from Santa Cruz Biotechnology. Other antibodies used were the MLK3
rabbit polyclonal antibody (Bock et al., 2000), actin and Flag mouse monoclonal antibody
(Sigma), and horseradish peroxidase-conjugated secondary antibodies (Bio-Rad). Alexa-
488 and Alexa-543 and Alexa-555 conjugated secondary antibodies, TOPRO iodide and

MitoTracker Red were from Molecular Probes.

92

3.2 Cell culture

MCF-7/iFLAG-MLK3 cells inducibly expressing F lag-MLK3 have been
described in Chapter II (Zhang et al., 2004), and were maintained in DMEM
supplemented with 5% fetal bovine serum, 40 pg/mL neomycin and 10 pg/mL
hygromycin. Flag-MLK3 expression was induced by the addition of 50 nM AP21967 for

20 h.

3.3 Cell lysis and western blotting

Cells were lysed in 1% Triton lysis buffer, containing 50 mM HEPES (pH 7.5),
150 mM NaCl, 1.5 mM MgC12, 2 mM EGTA, 1% Triton X-100, 10% glycerol, 10 mM
NaF, 1 mM NaaPPi, 100 pM B—glycerophosphate, 1 mM Na3VO4 and 2 mM
phenylmethylsulfonylfluoride (PMSF), and complemented with a cocktail of protease
inhibitors (Sigma). Following sonication for 40 s in a sonication bath to enhance
disruption of organelles, lysates were clarified by centrifugation at 14,000 rpm at 4°C for
15 min. For the Triton X-100 biochemical fractionation, the cells were lysed as described
above, and the Triton-insoluble pellets were washed in lysis buffer and then resuspended

in 2X SDS loading buffer and boiled prior to SDS-PAGE.

3.4 Preparation of detergent-resistant membrane fractions by sucrose-density

gradient

Flag-MLK3 expression was induced in MCF7/iFlag-MLK3 for 20 h. Cells were
then treated with carrier or 15 pM SP600125 for 1 h, washed twice in ice-cold PBS, and

lysed in cold TNE buffer (25 mM Tris pH 7.5, 150 mM NaCl, 0.5 mM EDTA) containing

93

1% Triton X-100. Cells were incubated on ice for 20 min and then homogenized using
12 strokes in a Dounce homogenizer. Homogenates were mixed with an equal volume of
an 80% sucrose solution prepared in TNE, to yield a final concentration of 40% sucrose,
and placed at the bottom of an ultracentrifuge tube. A 5%-30% discontinuous sucrose
gradient was layered on top of the homogenate by adding 2 volumes of a 30% sucrose
solution prepared in TNE, and 1 volume of a 5% sucrose solution prepared in TNE.
Samples were spun at 200,000 x g (45,000 rpm in a Beckman TH66O rotor) for 18h at
4°C. Ten fractions (0.4 mL) were collected from the top of the gradient. The insoluble
material was resuspended in 400 pL of 2X SDS loading buffer. Equal volumes from

each fraction were separated by SDS-PAGE and subjected to western blotting analysis.

3.5 Immunofluorescence Confocal Microscopy

MCF-7/iFlag—MLK3 cells were plated into 6-well plates containing cover slips
coated with poly-D-lysine (Sigma). F lag-MLK3 expression was induced for 20 h, and
the cells were treated with 15 pM SP600125 or carrier for 1 h. When costaining for
mitochondria, MitoTracker was added to the cells 30 min prior to fixation at a final
concentration of 50 nM. The cells were rinsed twice with PBS, fixed in 2%
formaldehyde (Polyscience, Inc.) for 30 min at room temperature, and washed three times
with PBS. The cells were perrneabilized with PBS containing 0.2% Triton X-100 or 1%
Triton X-100 for 10 min. Where indicated, the cells were first washed in PBS containing
1% Triton X-100 for 5 min and then fixed in 2% formaldehyde for 30 min. Blocking was
performed in PBS containing 10% fetal calf serum, 2% bovine serum albumin, and

RNase A (100 pg/ml) for l h at room temperature. Primary antibodies were prepared in

94

PBS containing 5% bovine serum albumin (MLK3 rabbit polyclonal antibody 1:1500;
Flag M2 monoclonal antibody 1:1500; B-tubulin mouse monoclonal 1:200; vinculin
mouse monoclonal 1:400; GM130 mouse monoclonal 1:100; EEAl mouse monoclonal
antibody 1:250; clathrin X22 mouse monoclonal antibody 1:1000) and were incubated for
1 h at room temperature. The cells were then washed three times with PBS containing
2% fetal calf serum and incubated for 1 h with the appropriate Alexa F luor 488-, Alexa

F luor 546-, Alexa Fluor SSS-conjugated secondary antibody. After five washes with PBS
containing 2% fetal calf serum, the nuclei were stained for 30 min with TOPRO-3 iodide
(2 pM in PBS) and mounted onto glass slides. The fluorescently labeled cells were
examined with aZeiss LSM Pascal confocal laser scanning microscope. Three separate
tracks were used for capturing fluorescent images; excitation was done using 488-, 543-,
and 633-nm laser lines with BP 500-530 nm, BP 560-615 nm and LP 650 nm,

respectively, as emission filters.

95

4. Results
4.1 JNK inhibition redistributes MLK3 to a Triton-insoluble fraction

The data presented in Chapter II demonstrate that JNK inhibition results in a
decrease in the levels of Flag-MLK3 in cellular lysates. Although MLK3 can be
ubiquitinated when coexpressed in cells with ubiquitin, increased ubiquitination of
MLK3 in the presence of the JNK inhibitor was not observed. This suggests that
ubiquitin-mediated degradation of Flag-MLK3 could not account for the large reduction
in Flag-MLK3 protein levels observed upon JNK inhibition. To explain the reduction in
Flag-MLK3 protein levels in cellular lysates, I speculated that JNK-mediated
phosphorylation of MLK3 might impact the distribution, rather than the protein stability,
of MLK3.

To test this idea, the Triton X-lOO-insoluble pellets obtained after cellular lysis
were examined for both the presence as well as the activation state of MLK3. Briefly,
cells inducibly expressing Flag-MLK3 were incubated with SP600125 alone or with
SP600125 in the presence of calyculin A. Cells were disrupted in the standard 1% Triton
X-lOO-containing buffer, the cellular lysates were removed after centrifugation and, after
rinsing in lysis buffer, the Triton-insoluble pellets were solubilized in SDS-containing
buffer. Equal cellular equivalents from the Triton X-lOO-soluble fractions and the SDS-
resolubilized Triton X-100 pellets were examined by western blotting using identical
exposure times and conditions. As shown in Figure 1, JNK inhibition results in a
dramatic shift of Flag-MLK3 from a Triton-soluble to a Triton-insoluble fraction. This

redistribution is prevented by the addition of the phosphatase inhibitor calyculin A.

96

Triton-soluble

 

Triton-insoluble

 

 

+AP21967 +AP21967
Calyculin A - - - ... - - - +
SP600125 - - + + - - + +
.u- i " ' - Flag-MLK3 we
.9 ®T277@3281
- 7" MLK3 we
@c-Jun we
' " "‘ Hsp90 we
.7" «I... - ~ Cdc37 WB
h!!! 1'" "‘- cox IV we
~- --~ ERp72 WB
7‘ " "" "" Actin we

Figure l. Redistribution of MLK3 to detergent-insoluble fractions. Flag-MLK3
expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. Cells were
then treated with 15 pM SP600125 for l h, in the presence or absence of 50 nM calyculin

A. The cells were lysed as described in “Materials and Methods” and identical cellular

equivalents from the Triton-soluble and Triton-insoluble fractions were analyzed by

western blotting, using appropriate antibodies and identical exposure times.

97

Thus, while the total cellular amount of Flag-MLK3 is largely unaffected by the
manipulation of the phosphorylation state of Flag-MLK3, the distribution of F lag-MLK3
is profoundly impacted by its phosphorylation status. The F lag-MLK3 that redistributes
to the Triton-insoluble fraction upon JNK inhibition is largely inactive, as judged by its
trace reactivity with the phospho-MLK3 antibody. This is consistent with a JNK-
mediated positive feedback loop.

To analyze the solubility in Triton X-100 of different organelles under the cellular
lysis procedure, the distribution of various subcellular markers in the crude fi'actionation
was analyzed by western blotting. The Triton-soluble fractions contain phospho-c-Jun,
cytochrome oxidase IV (COX IV), and ERp72, but these proteins are largely absent from
the Triton-insoluble fractions, indicating that the nucleus, mitochondria, and endoplasmic
reticulum membranes are efficiently solubilized.

In a recent study from our lab, Hsp90 and its co-chaperone Cdc37 were identified
as cochaperones of MLK3 (Zhang et al., 2004). As shown in Figure 1, the Triton-
insoluble fractions are largely devoid of Hsp90 and Cdc37, which may suggest that the
dephosphorylation-dependent redistribution of Flag-MLK3 to detergent-insoluble

domains requires disruption of the MLK3-Hsp90/Cdc37 complex.

4.2 Distribution of MLK3 between Triton-soluble and Triton-insoluble fractions is

reversible
Treatment of cells with geldanamycin, an Hsp90 inhibitor, reduces the protein
levels of MLK3, suggesting that the interaction of Hsp90/Cdc37 with MLK3 stabilizes

MLK3 (Zhang et al., 2004). My findings that the Triton-insoluble domains contain

98

inactive F lag-MLK3 and largely lack the MLK3 chaperones Hsp90/Cdc37 might indicate
that the redistributed, insoluble, hypophosphorylated MLK3 is irreversibly aggregated or
inactivated. It is possible that aggregated MLK3 could be destroyed after a long term
inhibition of JNK. However, as shown in Figure 2, hypophosphorylated, inactive Flag-
MLK3 is still present in a Triton-insoluble compartment, with no evidence of enhanced
degradation even after 12 h of continuous treatment with 15 pM SP600125. This raises
the possibility that the sequestration of Flag-MLK3 is reversible. To test whether
hypophosphorylated, inactive Flag-MLK3 could be reactivated by altering its
phosphorylation status, Flag-MLK3 expression was induced and the cells were treated
with SP600125 for l h. The JNK inhibitor was removed by placing the cells in fresh
media either with or without calyculin A for different periods of time. The cells were
then lysed and the distribution of F lag-MLK3 was analyzed by western blotting. As
shown in Figure 3, upon removal of the JNK inhibitor 3 significant portion of F lag-
MLK3 redistributed to the soluble fraction and regained its activation status, as judged by
activation loop phosphorylation. The portion of F lag-MLK3 that remained insoluble also
recovered its activation status. Removal of the JNK inhibitor, followed by phosphatase
inhibition with calyculin A resulted in quantitative reactivation of Flag-MLK3 and
redistribution to the soluble fraction. These data indicate that the
phosphorylation/dephosphorylation cycle of F lag-MLK3 modulates the distribution of

Flag-MLK3 in a reversible manner.

99

Triton-soluble Triton-insoluble
SP600125(h) 0 1 4 812 0 1 4 8 12

U- .. .. .. nee-au- Flag-MLK3 we

@r277@s281
- "' ' MLK3 we

- e- ..- . @c-Jun WB

Figure 2. Effect of long-term J NK inhibition on the distribution of MLK3 in Triton
X-100. F lag-MLK3 expression was induced using 50 nM AP21967 in MCF7/iFlag-
MLK3 cells for 20 h. Cells were then treated with 15 pM SP600125 for different periods
of time. The cells were lysed as described in “Materials and Methods” and identical
cellular equivalents from the Triton-soluble and Triton-insoluble fractions were analyzed

by western blotting, using appropriate antibodies.

100

Triton-soluble

Triton-insoluble

 

+AP21 967

 

Cal A

212
+++

Recovery (h) - - . 1
SP600125 - - + +

~ .... M

”Iii

....-“—

+A P21 967

 

Cal A

-1212
+++++

 

......» Flag-MLK3 we
2 281
'7 "' '7 ®JLK3 e
@JNK we
Aetin WB

Figure 3. Effect of SP600125 removal on MLK3 distribution. F lag-MLK3 expression
was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. Cells were then
treated with 15 pM SP600125 for 1 h, washed and incubated in fresh media with or

without 50 nM calyculin A for the indicated times. Triton-soluble and Triton-insoluble

fractions were prepared and analyzed as described for Figure 1.

101

4.3 Triton-insoluble MLK3 is not associated with lipid rafts

It is possible that the Triton-insoluble, F lag-MLK3-containing fi'action represents
either cytoskeletal elements or lipid rafts. MLK3 has been shown in
cotransfection/coimmunoprecipitation experiments to interact with the actin cytoskeleton
linker protein, Merlin, which is the product of the neurafibramatasis-Z gene (Chadee et
al., 2006). In addition to its association with cytoskeleton (Bretscher et al., 2002), Merlin
has been recently reported to associate with lipid rafts (Stickney et al., 2004). Detergent-
resistant membranes have been described as microdomains within membranes, highly
enriched in cholesterol and with a distinct protein composition. In particular, lipid rafts
are thought to transiently organize a signaling microdomain, clustering receptor tyrosine
kinases, adaptor molecules and endocytic proteins, among others. Because of their high
cholesterol levels, lipid rafts can be easily isolated by density gradient centrifugation.

To determine whether the hypophosphorylated, inactive form of MLK3 generated
upon JNK inhibition localizes to lipid rafts, a sucrose-density centrifugation assay was
implemented. F lag-MLK3 expression was induced for 20 h, and the cells were then
treated with or without 15 pM SP600125 for 1 h. Homogenates were fractionated by
sucrose-density centrifugation; fractions were collected from top (lighter) to bottom
(heavier) and analyzed by western blotting. Flotillin 1, a component of lipid rafts, was
used as a marker. As shown in Figure 4, F lotillin l, which is present along the entire
gradient, is enriched in the lighter fractions (fractions 3 and 4). Since equal volumes, and
not equal amounts of proteins, were resolved in the gel, the presence of Flotillin l in the
heavier fractions is probably due to the proportionally higher protein content of the

fractions at the bottom of the gradient. Overall, the data indicate that the gradient was

102

 

5%

2345678910?

~-- ~-

40% 5%

SP6001 25

 

40%
23456 78 91°F

-- a- u- . g Flag-MLK3 W8

A I

g u: Flotillin1WB

W ---- EEA1WB
2:3; c222! .. £sz W8
'Hliir‘ii HHHH Lamp1WB

...-0 cu.-- COXIVWB

Figure 4. Distribution of MLK3 in a sucrose-density gradient. F lag-MLK3
expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. Cells were
then treated with 15 pM SP600125 for 1 h and homogenates were loaded onto a 5%-40%
sucrose gradient. After centrifugation at 200,000 x g for 18 h at 4°C, fractions of equal
volume were collected from top to bottom of the gradient and analyzed by western
blotting using the indicated antibodies. Numbers on top indicate fraction number from

top to bottom of the gradient. P, pellet.

103

successfully generated. F lag-MLK3 was found predominantly at the bottom of the
gradient (fractions 7-10), where it cofractionated with disrupted endocytic vesicles,
endoplasmic reticulum, lysosomes and mitochondria, as determined by the presence of
early endosome antigen 1 (EEAl), ERp72, Lysosomal-associated membrane protein 1
(Lampl) and COX IV, respectively. However, only a very small portion of Flag-MLK3
(approximately 6% of the total F lag-MLK3 as determined by densitometry) was
redistributed to a lighter fraction (fraction 3) upon JNK inhibition, while a substantial
amount of MLK3 (approximately 37% of the total F lag-MLK3) remained in the insoluble
pellet. Therefore, relocalization of F lag-MLK3 to lipid rafts cannot explain the dramatic

change in solubility of F lag-MLK3 in Triton X-100 upon inhibition of JNK.

4.4 Inhibition of JNK results in the localization of MLK3 to punctate, vesicle-like
structures

To gain fiirther insight into the redistribution of the hypophosphorylated MLK3 in
cells, immunofluorescence microscopy was used. MCF-7 cells inducibly expressing
F lag-MLK3 were incubated either in the presence or absence of 15 pM SP600125 for 1
h. Cells were then fixed, penneabilized with 1% Triton X-100 and stained for MLK3.
Inducibly expressed Flag-MLK3 is cytosolic, and evenly distributed in the cells (Figure
5). No nuclear localization of F lag-MLK3 was observed. Upon JNK inhibition, Flag-
MLK3 displays a more punctate pattern, and seems to be absent from the plasma
membrane and cellular projections like filopodia. In a parallel experiment, cells were
first incubated in 1% Triton X-100 prior to fixation and staining. Under those conditions,

all the cellular compartments that are solubilized in Triton X-100 are washed away before

104

1% TX-100 after fixation 1% TX-100 before fixation

SP600125

 

Figure 5. Subcellular localization of MLK3 in MCF-7/iFlag-MLK3 cells. Flag-MLK3
expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The cells
were then treated with or without 15 pM SP600125 for 1 h. The cells were incubated in
1% Triton X-100 before or after fixation, and MLK3 was stained using a rabbit MLK3
antibody and a secondary antibody conjugated with Alexa F luor 543. The nuclei were
stained with TOPRO-3 iodide. Representative images are shown with MLK3 and the

nuclei represented in red and blue, respectively. Bar, 5 pm.

105

fixation, and the MLK3 staining is practically lost. However, in the presence of the JNK
inhibitor, the MLK3 staining is preserved, and is undistinguishable from the pattern
observed when the cells are permeabilized afier fixation, suggesting that the punctate,
dot-like structures correspond to the MLK3 present in the Triton X-100 insoluble fraction
from the cellular lysates.

The colocalization of MLK3 with other subcellular markers was analyzed. Like
MLK3, cytoskeletal elements are, under certain conditions, Triton X-100 insoluble. To
determine whether F lag-MLK3 could localize along cytoskeletal elements, a B-tubulin
staining was performed in cells inducibly expressing Flag-MLK3. As shown in Figure 6,
no obvious colocalization of Flag-MLK3 and B-tubulin was observed. In the presence of
the JNK inhibitor, the MLK3 pattern changed to the punctate structures seen before;
however, these structures did not colocalize with B-tubulin either.

I also analyzed the possibility that MLK3 could colocalize with focal adhesions,
Golgi vesicles and mitochondria. Vinculin is a cytoskeletal protein associated with the
cytoplasmic face of cell junctions, where it anchors the actin lattice to the plasma
membrane (reviewed in (Bailly, 2003). As shown in Figure 7, no colocalization of
MLK3 and vinculin was observed, either in the presence or absence of SP600125.

Golgi Matrix protein 130 (GM130) is peripherally associated with the cis-
compartment of the Golgi and was initially isolated as a Triton X-100 insoluble
component of the Golgi matrix (Slusarewicz et al., 1994). When the localization of
GM130 was analyzed in the MCF-7/iFlag-MLK3 cells, it was found that MLK3, even

after JNK inhibition, does not colocalize with GM130 (Figure 8). Finally, I used

106

Figure 6. Costaining of MLK3 and B-tubulin in MCF—7/iFlag-MLK3 cells. Flag-
MLK3 expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The
cells were then treated with or without 15 pM SP600125 for l h and fixed. MLK3 was
stained using a rabbit MLK3 antibody and a secondary antibody conjugated with Alexa
F luor 488. B-tubulin was stained with a mouse B-tubulin antibody and a secondary
antibody conjugated with Alexa F luor 555. The nuclei were stained with TOPRO-3
iodide. Representative images are shown with MLK3, B-tubulin and nuclei represented

in green, red and blue, respectively. Bar, 5 pm.

107

690$ Can—O... :__:n:u-n «v.45—

SZLOOSdS

Figure 7. Costaining of MLK3 and vinculin in MCF-7/iFlag-MLK3 cells. F lag-
MLK3 expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The
cells were then treated with or without 15 pM SP600125 for 1 h and fixed. MLK3 was
stained using a rabbit MLK3 antibody and a secondary antibody conjugated with Alexa
Fluor 488. Vinculin was stained with a mouse vinculin antibody and a secondary
antibody conjugated with Alexa Fluor 555. Representative images are shown with

MLK3 and vinculin represented in green and red, respectively. Bar, 5 pm.

109

696.2

:__:o:_> 93.2

 

SZ LOOSdS

110

Figure 8. Costaining of MLK3 and GM130 in MCF-7/iFlag-MLK3 cells. Flag-MLK3
expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The cells
were then treated with or without 15 pM SP600125 for 1 h and fixed. MLK3 was stained
using a rabbit MLK3 antibody and a secondary antibody conjugated with Alexa F luor
488. GM130 was stained with a mouse GM130 antibody and a secondary antibody
conjugated with Alexa Fluor 555. Representative images are shown with MLK3 and

GM130 represented in green and red, respectively. Bar, 5 pm.

111

 

cars—0

3.1—s.

S
.d
9
0
0
Ir
7v
9

 

112

Mitotracker Red, a dye that metabolically labels mitochondria, to analyze their
distribution in the inducible cell line. No colocalization of Flag-MLK3 and mitotracker
Red was observed, irrespective of JNK inhibition, even when different focal planes were
analyzed (Figure 9).

Because of the distinctly punctate cytoplasmic distribution of MLK3, I turned my
attention to endocytic vesicles. Endosomes are cellular compartments that internalize
extracellular molecules into the cytosol. Early endosomes receive the endocytosed
material, and sort it to late endosomes and lysosomes, or send it for recycling to the
plasma membrane. EEAl is an evolutionarly conserved protein associated with the
membrane of the early endocytic vesicles (Mu et al., 1995). To determine whether the
punctate pattern corresponds to endosomes, cells were costained for MLK3 and EEAl.
As shown in Figure 10, there is a small amount of colocalization of signaling-competent
MLK3 with EEAl. Upon JNK inhibition, the subcellular distribution of the
hypophosphorylated MLK3 looks, at first glance, very similar to the distribution of
EEAl. Some structures seem to harbor both MLK3 and. EEAl staining. However, the
majority of the vesicles are either exclusively MLK3-positive or EEAl-positive. Triton-
soluble and Triton-insoluble fractions from cellular lysates prepared from MCF7/iFlag-
MLK3 cells were analyzed for the presence of early endosomes; EEAl was found
exclusively in the Triton-soluble fractions (see Table 1). These data indicate that JNK
inhibition does not result in quantitative targeting of the dephosphorylated, inactive

MLK3 to endosomes.

113

MLK3 Mitotracker Merge

Figure 9. Costaining of MLK3 and mitochondria in MCF-7/iFlag-MLK3 cells. Flag-
MLK3 expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The

 

SP600125

Different Focal planes

SP600125

 

 

cells were then treated with or without 15 pM SP600125 for 1 h and incubated with
Mitotracker Red 30 min before fixation. MLK3 was stained using a Flag antibody and a
secondary antibody conjugated with Alexa F luor 488. Representative images are shown
with Flag-MLK3 and mitochondria represented in green and red, respectively. Two

different focal planes from the same cell treated with SP600125 are shown. Bar, 5 pm.

114

Figure 10. Costaining of MLK3 and endocytic vesicles in MCF—7/iFlag-MLK3 cells.
Flag-MLK3 expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20
h. The cells were then treated with or without 15 pM SP600125 for 1 h and fixed. MLK3
was stained using a rabbit MLK3 antibody and a secondary antibody conjugated with
Alexa F luor 546. Endocytic vesicles were stained with a mouse EEAl antibody and a
secondary antibody conj gated with Alexa Fluor 488. Representative images are shown

with MLK3 and EEAl represented in red and green, respectively. Bar, 5 pm.

115

 

696: 2mm 9::

92 lOOQdS

116

I then investigated whether MLK3 could localize to other type of vesicles, such as
clathrin-coated vesicles. Figure 11 shows cells inducibly expressing MLK3, which
display a substantial amount of colocalization between MLK3 and clathrin, except at the
cell periphery, where only MLK3 could be detected. When the cells are incubated with
SP600125, a dramatic change in the subcellular distribution of MLK3 is observed, with
loss of staining at the cell edge, as described before. However, there is no substantial
change in the amount of colocalization between MLK3 and clathrin under conditions of
JNK inhibition, when compared with untreated cells. These data indicate that MLK3 and
clathrin can colocalize in cells, but that JNK does not cause a dramatic redistribution of
MLK3 to clathrin-coated vesicles.

Table 1 summarizes the distribution of different subcellular markers in the crude
Triron X-100 fractionation. Most of the markers analyzed are soluble in Triton X-100.
The data presented in this Chapter indicate that upon JNK inhibition, MLK3 becomes
hypophosphorylted and redistributes to a Triton X-100 insoluble fraction that appears as
small dot-like vesicles throughout the cell. The nature of those vesicles could not be

precisely determined, but it does not seem to correspond to any major organelle.

117

Figure 11. Costaining of MLK3 and clathrin in MCF-7/iFlag-MLK3 cells. Flag-
MLK3 expression was induced using AP21967 in MCF7/iFlag-MLK3 cells for 20 h. The
cells were then treated with or without 15 pM SP600125 for 1 h and fixed. MLK3 was
stained using a rabbit MLK3 antibody and a secondary antibody conjugated with Alexa
F luor 546. Clathrin was stained with a mouse clathrin heavy chain antibody and a
secondary antibody conjugated with Alexa F luor 488. Representative images are shown

with MLK3 and clathrin represented in red and green, respectively. Bar, 5 pm.

118

 

09o: 5.55.0 93$.

9Z1009d$

Table 1. Distribution of organelle markers in a crude Triton X-100 fractionation.
Abbreviations: COX IV, cytochrome oxidase IV; ERp72, endoplasmic reticulum protein

72; Lampl , lysosome associated membrane protein 1; EEAl , early endosome antigen 1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Organelle Marker :riiuifl’g £3113:
Mitochondria COX IV ++ -
En lasmi

Rizgulumc ERp72 ++ '-

Golgi GM130 ++ -
Lysosome Lampl ++ -
Plasma membrane Integrin ca ++ -
Endocytic vesicles EEAl ++ -
Lipid Rafis Flotillin 1 ++ +/-
y-tubulin ++ -
Cytoskeleton Actin ++ +/-
Vinculin ++ -
phospho-c-Jun -i-+ -
Histone 3 +/- ++
Nucleus p27 ++ ..
Nucleoporin ++ ++
p62
. Hsp90 ++ +/-
Cytosolic C dc 3 7 ++ __

 

 

 

 

120

5. Discussion

Many signal transduction pathways are modulated by feedback loops. In the case
of MLK3, activation of the JNK pathway results in positive feedback phosphorylation of
MLK3 by JNK. In this chapter I show that when JNK is inhibited, MLK3 is rapidly
inactivated and is redistributed to a Triton-insoluble fraction. This JNK phosphorylation-
dependent, reversible redistribution of MLK3 represents a novel mode by which MLK3
is regulated.

In my studies I made use of MCF-7 cells engineered to inducibly express wild
type Flag-MLK3 which is highly active, and phosphorylated both in its activation loop
(most likely through autophosphorylation) and in its COOH-terminal region by JNK.
This hyperphosphorylated F lag-MLK3 is present in Triton-soluble fractions and in turn
potently activates JN K. Inhibition of JNK results in a rapid redistribution of the majority
of F lag-MLK3 to a Triton-insoluble fraction. Under conditions of JNK blockade, all of
the Flag-MLK3 in the Triton-insoluble fraction, as well as the small portion of Flag-
MLK3 that remains in the Triton-soluble fraction, is inactive, as judged by activation
loop phosphorylation status. The complete loss of reactivity with the activation loop
specific phospho-MLK3 antibody upon JNK inhibition may suggest the existence of an
activation loop phosphatase that is either repressed by JNK or restricted to the cellular
compartment corresponding to the MLK3-containing Triton-insoluble fraction.
Although, activation 100p phosphatases that target MAPKKKs do not seem to be
commonplace, an activation loop phosphatase for ASKl has been described (Morita et

aL,2001)

121

MLK3 distribution depends upon its phosphorylation status. The serine/threonine
phosphatase inhibitor, calyculin A, renders F lag-MLK3 highly phosphorylated and
completely soluble. JNK inhibition results in inactivation and redistribution of Flag-
MLK3 to a Triton-insoluble fraction. Removal of the JNK inhibitor results in active,
solubilized F lag-MLK3, providing evidence for reversibility of this process, at least in
short term treatments.

The finding that JNK phosphorylation regulates the solubility of a MAPKKK is
novel. It has been reported that cellular stressers like camptothecin and sorbitol can
induce MLK3 activation and stabilization. Ablation of JNK expression by siRNA or with
dominant negative JNK] blocked camptothecin-induced increase in MLK3 protein levels,
supporting a model for a positive feedback regulation of MLK3 stability by JNK (Xu et
al., 2005). However, the subcellular distribution of the stabilized MLK3 was not
analyzed. In my experimental setting, JNK inhibition does not reduce the total amount of
MLK3, but rather changes its subcellular distribution. It is unclear whether the
camptothecin-dependent increase in soluble MLK3 is a result of increased protein
stability or increased protein solubility.

Our lab recently published that Hsp90/Cdc37 act as cochaperones for MLK3
(Zhang et al., 2004). The Triton-insoluble fraction containing inactive F lag-MLK3 is
largely devoid of Hsp90 and lacks detectable Cdc37, suggesting that Hsp90/Cdc37
dissociation from F lag-MLK3 might be involved in its redistribution to Triton-insoluble
fractions. Dissociation of Hsp90 from the IKK complex was recently shown to result in
transient, reversible Triton-insolubility of the IKKs (Pittet et al., 2005), whereas long

term stress signals result in degradation of IKK. A role for IKK phosphorylation in this

122

process has not been established. In my case, long term (12 h) inhibition of JNK did not
result in increased degradation of F lag-MLK3. Furthermore, the fact that removal of the
JNK inhibitor results in MLK3 reactivation and solubilization indicates that the MLK3
present in the Triton-insoluble fraction is not irreversibly aggregated, raising the
possibility that there may be other chaperones interacting with F lag-MLK3 that keep it
folded when present in Triton-insoluble compartments. For instance, small Heat shock
proteins (sHsps) help in resolubilizing proteins that become unfolded upon stress,
protecting them fipm irreversible aggregation (revieWed in (Sun & MacRae, 2005)). In
yeast, upon heat stress sHsps like Hsp26 or Hsp42 become insoluble and bind to unfolded
proteins, keeping the whole complex in a state that allows the Hsp104/I-lsp40/Hsp70
machinery to reactivate the proteins (Cashikar et al., 2005). Additionally, translation of
new MLK3 molecules must have occurred during the extended treatment with the JNK
inhibitor; these newly synthesized MLK3 molecules were also targeted to the Triton-
insoluble fraction, suggesting that under non-inhibiting conditions JNK phosphorylation
occurs rapidly, and reinforcing the idea that JNK phosphorylation of MLK3 is required to
keep MLK3 soluble.

Since MLK3 is not degraded, to where has it been redistributed? I analyzed the
Triton-soluble and Triton-insoluble fractions for the presence of markers of different
organelles. The data, summarized in Table 1, indicate that only flotillin-l, which is
present in cholesterol-enriched membranes like lipid rafts and nuclear components
fractionate with MLK3. However, density gradient centrifirgation experiments indicate
that MLK3 does not partition to lipid rafts upon JNK inhibition. Confocal microscopy

studies demonstrate that treatment with the pharmacological inhibitor of JNK changes the

123

MLK3 localization pattern from diffuse cytosolic to vesicular. MLK3 does not localize
to the nucleus, and incubation with the JNK inhibitor does not result in the total
translocation of MLK3 into the nucleus, although some MLK3-containing structures
seem to be perinuclear.

Inducibly expressed MLK3 colocalizes with vesicle markers to some degree. Yet
I did not detect increased colocalization to endocytic vesicles or clathrin-coated vesicles
upon JNK inhibition. In addition, under my cellular lysis conditions, the vesicular
markers fractionate to the Triton-soluble supernatant. The ERK pathway scaffold protein
KSR has been found to fractionate to Triton-insoluble cellular compartments. c-Tak-l-
mediated phosphorylation of KSR, and/or its interaction with the Ras effector protein
IMP, results in hyperphosphorylation and inactivation of KSR, and redistribution of KSR
to a Triton-insoluble fraction (Matheny et al., 2004; Stewart et al., 1999). Interestingly,
coexpression of KSR with IMP results in colocalization of both proteins to Triton-
resistant punctate structures, similar to the ones observed for hyposphopshorylated
MLK3 (Matheny et al., 2004). The authors speculate that KSR could be sequestered into
microdomains where it becomes inaccessible to activators. Like in the case of MLK3,
this distribution is reversible and can be overcome by expression of oncogenic Ras.
However, the nature of the punctate structures was not analyzed.

In summary, this work describes a novel way of regulating MLK3, by targeting its
inactive, dephosphorylated form to a Triton-resistant fraction. The determination of the
exact nature of the Triton-insoluble MLK3, as well as the presence of interactive partners

of MLK3 in the Triton-insoluble fractions requires further investigation.

124

IV. Concluding Remarks

The purpose of the work described in this dissertation was to gain additional
insight into the mechanism by which phosphorylation regulates MLK3 activities. Of
particular interest were the roles of a subset of phosphorylation sites identified on MLK3,
which correspond to proline—directed kinase sites.

The work presented here is the natural continuation of previous studies done in
the laboratory. The identification of in viva phosphorylation sites on MLK3 was thought
as a first step towards understanding how phosphorylation regulates MLK3. The
experiment was designed as follows: MLK3 and its upstream activator Cdc42 were
transfected into cells. Sixteen hours after transfection, the cells were labeled with
inorganic phosphate for four hours. Subsequently, the phosphorylation sites on MLK3
that incorporated radioactivity were determined by mass spectrometry. As a result,
eleven sites were identified. Seven of them contained a proline residue immediately
following the phosphorylated serine residue, suggesting that MLK3 is regulated by
proline-directed kinases.

The work in Chapter II reveals that JNK is responsible for phosphorylating those
sites on MLK3. Since phosphorylation of the proline-directed sites requires activation of
the JNK pathway by MLK3, the data is consistent with positive feedback
phosphorylation of MLK3 as the regulatory mechanism. The rapid exchange of
phosphate into those sites isialso indicative of dynamic phosphorylation. Feedback
phosphorylation of MLK3 generates a slower migrating form of MLK3 in SDS-PAGE,

which is also abolished when JNK is inhibited.

125

In retrospect, the experimental design that led to the discovery of phosphorylation
sites resulted in the identification of sites that occurred as a consequence of MLK3
activation, rather than phosphorylation events that caused MLK3 activation. This helps
explain, for instance, why the activation loop phosphorylation sites were not detected
even though, by means of a phospho-specific antibody, they are real phosphorylation
sites on MLK3: phosphorylation of the activation segment is thought to be one of the first
required steps in the activation of a protein kinase. Therefore, it would have occurred
long before the radiolabeled orthophosphate was added to the cells. This hypothesis
implicitly states that the activation loop sites are quite stable, with a very slow turnover,
resulting in very little incorporation of radioactivity during the labeling period.

Likewise, I believe that the decreased electrophoretic mobility of MLK3 observed upon
coexpression with activated Cdc42 correspond to feedback phosphorylation by JNK,
rather than being a direct effect of Cdc42 on MLK3.

The work in Chapters 11 and III demonstrate that feedback phosphorylation of
MLK3 is a dynamic process. Inhibition of JNK activity not only impaired feedback
phosphorylation of MLK3, but it also dramatically reduces the levels of MLK3 in cellular
lysates. This is not likely due to degradation of MLK3. As shown in Chapter III, the
dephosphorylated MLK3 was rendered Triton-insoluble. However, if the inhibition of
JNK is released, MLK3 moves back to a Triton-soluble fraction, indicating that the
process is reversible and dependent on the phosphorylation state of MLK3. By confocal
microscopy it was shown that the Triton-insoluble MLK3 localizes to dot-type structures
in the cytosol of the cells, although the exact nature of those punctate structures remains

to be determined.

126

An important question that remains to be answered is whether the regulation of
MLK3 by phosphorylation observed in the inducible expression system can be applied to
endogenous MLK3. Under basal conditions, endogenous MLK3 is phosphorylated,
based on phosphatase treatment of cellular lysates, which increases the electrophoretic
mobility of MLK3 (K. Viswanathan). However, activation loop phosphorylation of
MLK3 has not been detected in resting cells. TNF-a treatment, which transiently
activates JNK, does not result in altered mobility of MLK3, suggesting that feedback
phosphorylation has not occurred. However, under conditions of sustained JNK
activation, like a 4 h treatment with camptothecin, feedback phosphorylation of MLK3
has been proposed to occur. In this regard, phospho-specific antibodies generated against
the proline-directed kinase phosphorylation sites would be predicted to recognize MLK3
phosphorylated under sustained, but not transient, JNK activation. Development of such
antibodies becomes necessary to fully clarify this aspect of the molecular regulation of
MLK3. However, for practical reason, it might be necessary to make only one of the
phospho-specific antibody. In that case, selecting a peptide containing phosphorylated
Ser 740 may be the best option; the phosphopeptide maps indicate that Ser 740 is highly
phosphorylated when in viva labeled, which would likely make it easier to detect changes
in the phosphorylation status of the endogenous protein. Whether MLK3 is
phosphorylated on other yet unidentified sites during transient activation also need to be
determined. Mass spectrometry could be applied on immunopurified MLK3 from cells
that have been treated with TNF-a, which has been documented to activate endogenous

MLK3, to identify phosphorylation sites that regulate transient MLK3 activation.

127

The redistribution of inactive, hypophosphorylated MLK3 to a Triton-insoluble
fraction was somewhat surprising. Even more, the fact that MLK3 is kept insoluble,
rather than degraded, is unexpected. It is conceivable that MLK3 is associated with other
molecule(s) that keep MLK3 insoluble, but stable. In fact, during a long term (12 h)
treatment with the JNK inhibitor, the steady state level of MLK3 does not change and
MLK3 does not accumulate in the insoluble fraction, suggesting that MLK3 can be
recognized and turned over normally. One possibility is that the redistribution of MLK3
to Triton-insoluble vesicles is a normal process used by the cell to inactivate MLK3. It
would be important to analyze the distribution of endogenous MLK3 in Triton X-100

upon JNK activation and after JNK activation has ceased.

128

Appendix A: Determination of the in vitro kinase activity of various

MLK3 mutants.

This appendix describes the generation of phosphorylation site mutants of MLK3
and the analysis of their impact on the kinase catalytic activity of MLK3.

Two regions of MLK3 were analyzed: the glycine-rich region, where serines 11,
15 and 35 were independently mutated to either alanine or the phosphomimetic glutamic
acid, and the Cdc42-inducible sites, where serines 555 and 556 were both mutated to

either alanine or glutamic acid.

Generation of phosphorylation site mutants.

Site directed mutagenesis was performed by polymerase chain reaction (PCR)
using the Quick Change method (Stratagene). The oligonucleotides used for PCR, as
well as the DNA template are described in Table 1. The double mutants of the Cdc42
inducible sites were constructed as followed: oligonucleotides containing the 8555A or
the SSSSE point mutation were used in a PCR reaction using pRKS-MLK3 as a template.
The presence of the mutation was confirmed by DNA sequencing (MSU DNA
Sequencing Facility). A second PCR reaction over this mutated DNA was done using
oligonucleotides containing the S555A,S556A or the SSSSE,S556E mutations and the

presence of the mutations was confirmed by DNA sequencing.

129

Table 1. Oligonucleotides used for Site-directed mutagenesis.

 

 

 

 

 

 

 

 

 

 

 

Mutant Oligonucleotide Template
31 1A 5 ’-AGCCTCTTCCTCAAGGCACCTCTAGGGTCATGG-3’ 11:33:55
31 113 5 ’-AGCCTCTICCTCAAGGAGCCTCTAGGGTCATGG-3’ 112/1115133
815A 5’-AAGAGCCCTCTAGGGGCATGGAATGGCAGTGGC-3’ REE;
SlSE 5’-CCTCTAGGGGAATGGAATGGC-3’ 124%.;
S35A 5’-CCTGAGGGGGCTCCAAAGGCA-3’ 112/ES;
S35E 5’-CGGCCTGGAGGGGGAACCAAAGGCAGCG-3’ {412%
S555A 5’-CTGGAGGACGCAAGCAATGGAGAG-3’ Egg
pRKS-
S555A,SSS6A 5’-CTGGAGGACGCAGCCAATGGAGAG-3’ MLK3
8555A
$55513 5’-CTGGAGGACGAAAGCAATGGAGAG-3’ {31%
pRKS-
ssssesssse 5’-CTGGAGGACGAAGAAAATGGAGAG-3’ MLK3
sssse

 

 

 

130

Effect of mutagenesis on MLK3 in vitro kinase activity
1. Mutagenesis of the glycine-rich region

Using mass spectrometry coupled with phosphopeptide mapping analysis, in viva
phosphorylation sites of MLK3 were previously identified, under conditions of ectopic
expression of MLK3 with its upstream activator, Cdc42. Eleven sites were identified; all
are serine residues located towards the COOH-terminal region of MLK3. An additional
phosphorylation site was found in a peptide encompassing Ser 11-Lys 37 of MLK3,
whose sequence is "SPLGSWNGSGSGGGGGGGGGRPEGSPK” (Vacratsis et al.,
2002). Five serines are present in this peptide, and although the mass spectrometry data
indicates that only one phosphate group is present, it was not possible to specifically
assign the phosphorylation site, presumably due to the poor fragmentation of the peptide
on the ion source of the MALDI mass spectrometer (Vacratsis et al., 2002).

In an attempt to implicate one of the Ser residues in that peptide as a
phosphorylation site, I individually mutated Ser 11, Ser 15 and Ser 35 to Ala or to Glu
residues. Ala mutants are used to mimic non-phosphorylated form because the side chain
of alanine cannot be modified with a phosphate moiety. To mimic the effect of having a
serine residue constitutively phosphorylated, Glu or Asp are used to substitute the
phosphorylation site, the idea being that the carboxylate can mimic both the structure and
the charge of a phosphate group. However, substitution of a carboxylate for a phosphate
group is not always successful in reproducing the phenotype of the phosphorylated state
of a protein.

To determine if phosphorylation on any of those serines on the glycine-rich region

could affect MLK3 kinase activity, an in vitro kinase assay was performed. HEK 293

131

cells were transfected using the calcium phosphate technique with expression vectors for
the different MLK3 variants. Twenty hours post-transfection the cells were lysed, and
MLK3 was immunoprecipitated from clarified lysates with a MLK3 antibody. The
immunoprecipitates were washed three times with HNTG buffer (20 mM HEPES (pH
7.5), 150 mM NaCl, 0.1% Triton X-100, 10% glycerol) containing 1 M LiCl, three times
with HNTG buffer, and twice with kinase assay buffer (50 mM Tris-HCl (pH7.5), 100
mM NaCl, 1 mM MnClz, 10 mM MgC12, and 0.1 mM Na3VO4). MLK3 kinase assays
were performed in 20 pl kinase assay buffer using 50 pM ATP and 5 pCi [y-3ZP] ATP
(3000 Ci/mmol) (NEN Life Science Products). The immunoprecipitates were incubated

with 10 pg of mixed histones (Roche Molecular Biochemicals) for 15 min at room

temperature. The kinase reaction was terminated by adding 40 pl of 1.5x SDS loading
buffer (100 mM Tris (pH 6.8), 4% SDS, 20% glycerol, 0.2% bromophenol blue, 100 mM
dithiothreitol, and 1% B-mercaptothanol, 50 mM EDTA (pH 8.0)). The proteins were
resolved by SDS-PAGE. The gels were rinsed in PBS, dried, and the incorporation of
radioactivity into kinase or substrates was determined using Phosphorlmaging (Molecular
Dynamics). The data was normalized to the levels of MLK3 in the immunoprecipitates.
The results shown in Figure 1 indicate that substitution of serine 11 or serine 35
of MLK3 with alanine or glutamic acid residues did not affect MLK3 catalytic activities
in vitro. On the other hand, the S15A substitution on MLK3 reduced its ability to
autophosphorylate as well as its ability to phosphorylate an exogenous substrate by 40%.

However, the correspondent glutamic acid mutation did not increase MLK3 catalytic

132

Figure 1. In vitro kinase assay of MLK3 mutants in the glycine-rich region. HEK 293
cells were transfected with expression vectors for different variants of MLK3 as

indicated. MLK3 was immunoprecipitated from clarified lysates and subjected to in vitro
kinase assay using [y-32P] ATP and histones as substrates. (A) MLK3 and histones were
fractionated by SDS-PAGE. T op panel, autoradiogram showing MLK3
autophosphorylation and histone phosphorylation. Bottom panel, a fraction of the
immunoprecipitated MLK3 was subjected to western blotting using a MLK3 antibody.

(B) The incorporation of radioactivity into MLK3 and into histones was quantified by
phosphorimaging, and normalized to the levels of immunoprecipitated MLK3. Data from

two or three independent experiments are shown as mean t SE.

133

® MLK3

mmnm 9:2 .

(mam ”xi—s.
mmrw “UV—...:
(mrm 83.2
mrrw 93.2
(2% 9:5.

8::

@ Histories

«...-....- MLK3 iPI MLK3 we

S11A S11E $15A S15E $3M $35E

 

.... s m. s L s
2 al 0

noun—Eccamocnouza
c. 62365 Eon.

 

wt

 

 

 

 

S11A S11E S15A S15E'S35A SSSE'

 

i
I
_

azSLSn.
2

 

 

 

 

 

 

 

:ozaicosamofi
2.20.2 E 62335 Eon.

wt

134

activities, the autophosphorylation activity being identical to that of the wild type kinase,
and the substrate phosphorylation activity being even lower than wild type MLK3. It is
possible that the change to glutamic acid is not adequately mimicking the effect of a
phosphate group on serine 15. Nonetheless, the data suggest that Ser 15 is a potential
phosphorylation site on MLK3 that might participate in the regulation of MLK3 catalytic

activity.

2. Mutagenesis of the Cdc42-inducible sites

Cdc42 is a small GTPase that binds, in its activated state, to the CRIB motif of
MLK3. Coexpression of a constitutively active mutant form of Cdc42, Cdc42V'2, with
MLK3, increases the in vitro catalytic activity of MLK3 as judged by
autophosphorylation and histones phosphorylation (Bock et al., 2000). In addition,
activated Cdc42 changes the in viva phosphorylation pattern of MLK3. Two of the
identified phosphorylation sites of MLK3, serine 555 and serine 556, reside in
phosphopeptides that appear at high levels in phosphopeptides maps of MLK3 when
MLK3 is co-expressed with Cdc42V'2, and for that reason they were called “Cdc42
inducible sites”.

To test whether the Cdc42-inducible phosphorylation sites are responsible for the
potentiation of MLK3 catalytic activity by the small GTPase, those sites were mutated to
alanine and to glutamic acid using site-directed mutagenesis. Vectors encoding MLK3,
MLK3 8555A,8556A (double mutation to alanine) and MLK3 SSSSE,SSS6E (double
mutation to glutamic acid) were transfected into HEK 293 cells and the different variants

of MLK3 were immunoprecipitated from cellular lysates. An in vitro kinase assay using

135

histories as a substrate was used to test the activity of the variants. If phosphorylation at
serine 555 and serine 556 were responsible for the increase in MLK3 catalytic activity,
then the double alanine mutant should fail to potentiate MLK3 activity when co-
expressed with Cdc42V12. Similarly, the double glutamic acid mutant would be expected
to exhibit higher catalytic activity than wild type MLK3 when expressed alone.
Furthermore, the catalytic activity of the double glutamate mutant should not be further
increased by coexpression with activated Cdc42. The result shown in Figure 2 did not
match any of these expectations. MLK3 SSSSA,8556A had a basal autophosphorylation
level higher than with wild type MLK3, and its catalytic activity was potentiated upon
co-expression with Cdc42vn at a similar level as wild type MLK3 (Figure 2A, compare
lanes 2 and 4, 3 and 5). Similarly, MLK3 SSSSE,8556E showed increased basal
autophosphorylation activity compared to wild type MLK3, and could still be potentiated
by coexpression with activated Cdc42 (Figure 5A, compare lanes 2 and 6, 3 and 7).
These results suggest that the phosphorylation of the Cdc42 inducible sites is not required
for the increase in MLK3 catalytic activity observed upon coexpression with activated

556

Cdc42. It is possible that phosphorylation of Ser555 and Ser regulates some other

function of MLK3 such as protein-protein interaction or subcellular localization.

555 556

Alternatively the phosphorylation of Ser and Ser may result from, rather than cause,
the increased catalytic activity of MLK3 with Cdc42. Finally, there may be other sites of
Cdc42-induced phosphorylation, as yet unindentified, that are responsible for the Cdc42-

induced effect in MLK3 catalytic activity.

136

Figure 2. In vitro kinase activity of the Cdc42-inducible sites mutants of MLK3.
HEK 293 cells were transfected with expression vectors for different variants of MLK3,
with or without activated Cdc42 as indicated. MLK3 was immunoprecipitated from
clarified lysates and subjected to in vitro kinase assay using [y-32P] ATP and histones as
substrates. (A) MLK3 and histones were fractionated by SDS-PAGE. T ap panel,
autoradiogram showing MLK3 autophosphorylation and histone phosphorylation.
Bottom panel, a fraction of the immunoprecipitated M LK3 was subjected to western
blotting using a MLK3 antibody. (B) The incorporation of radioactivity into MLK3 and
into histones was quantified by phosphorimaging, and normalized to the levels of
immunoprecipitated MLK3. Data from three independent experiments are shown as

mean i SE.

137

Fold increase in
autophosphorylation

Fold increase in histone

$555A $5556
MLK3 Wt S556A S5565

05542V12 - - + -+ - +

. fl @Histones

an -- I .1. ‘ MLK3 lP/MLK3 W8

14~

, _I_

 

 

O
4

41

 

21

V V V ' '

 

 

 

 

 

 

 

 

 

 

 

coca - + - + - +
MLK3 wt 8555.556A 3555,5565

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C
Co ‘
g 6‘
Z‘ 51
3 41
3' 31
2 21 -
a 1 ‘1 [—1
0 I j v . .
Cdc42 ' + - + . +
MLK3 Wt $555,555A 5555,5565

138

VI. References

Aitken A. (1999). Mal Biotechnol, 12, 241-53.

Allen LF, Sebolt-Leopold J and Meyer MB. (2003). Semin Oncol, 30, 105-16.
Bailly M. (2003). Trends in Cell Biology, 13, 163.

Bain J, McLauchlan H, Elliott M and Cohen P. (2003). Biochem J, 371, 199-204.
Bar-Sagi D and Hall A. (2000). Cell, 103, 227-38.

Barthwal MK, Sathyanarayana P, Kundu CN, Rana B, Pradeep A, Sharma C, Woodgett
JR and Rana A. (2003). J Biol Chem, 278, 3897-902.

Beyaert R, Cuenda A, Vanden Berghe W, Plaisance S, Lee J C, Haegeman G, Cohen P
and Fiers W. (1996). Emba J, 15, 1914-23.

Biondi RM and Nebreda AR. (2003). Biochem J, 372, 1-13.

Blank J L, Gerwins P, Elliott EM, Sather S and Johnson GL. (1996). J Biol Chem, 271,
5361-8.

Blouin R, Beaudoin J, Bergeron P, Nadeau A and Grondin G. (1996). DNA Cell Biol, 15,
631-42.

Bock BC, Vacratsis PO, Qamirani E and Gallo KA. (2000). J Biol Chem, 275, 14231-41.

Brancho D, Ventura JJ, Jaeschke A, Doran B, F lavell RA and Davis RJ. (2005). Mal Cell
Biol, 25, 3670-81.

Bretscher A, Edwards K and F ehon RG. (2002). Nature Reviews Molecular Cell Biology,
3, 586.

Brummer T, Naegele H, Reth M and Misawa Y. (2003). Oncogene, 22, 8823-34.

139

Buchsbaum RJ, Connolly BA and Feig LA. (2002). Ma] Cell Biol, 22, 4073-85.
Burbelo PD, Drechsel D and Hall A. (1995). J. Biol. Chem, 270, 29071-29074.

Cashikar AG, Duennwald M and Lindquist SL. (2005). J. Biol. Chem, 280, 23869-
23875.

Ceci JD, Patriotis CP, Tsatsanis C, Makris AM, Kovatch R, Swing DA, Jenkins NA,
Tsichlis PN and Copeland NG. (1997). Genes Dev, 11, 688-700.

Cha H, Dangi S, Machamer CE and Shapiro P. (2006). Cell Signal, 18, 93-104.
Cha H, Smith BL, Gallo K, Machamer CE and Shapiro P. (2004). J Cell Sci, 117, 751-60.
Chadee DN and Kyriakis JM. (2004). Nat Cell Biol, 6, 770-6.

Chadee DN, Xu D, Hung G, Andalibi A, Lim DJ, Luo Z, Gutrnann DH and Kyriakis J M.
(2006). PNAS, 103, 4463-4468.

Chao TH, Hayashi M, Tapping RI, Kato Y and Lee JD. (1999). J Biol Chem, 274, 36035-
8.

Chen Z, Gibson TB, Robinson F, Silvestro L, Pearson G, Xu B, Wright A, Vanderbilt C
and Cobb MH. (2001). Chem Rev, 101, 2449-76.

Cheng M, Boulton TG and Cobb MH. (1996). J Biol Chem, 271, 8951-8.

Coso OA, Chiariello M, Yu JC, Teramoto H, Crespo P, Xu N, Miki T and Gutkind J S.
(1995). Cell, 81, 1137-46.

Coulombe P, Rodier G, Pelletier S, Pellerin J and Meloche S. (2003). M01 Cell Biol, 23,
4542-58.

Cowley S, Paterson H, Kemp P and Marshall CJ. (1994). Cell, 77, 841-52.

140

Crawley JB, Rawlinson L, Lali FV, Page TH, Saklatvala J and Foxwell BM. (1997). J
Biol Chem, 272, 15023-7.

Cuenda A and Dorow DS. (1998). Biochem J, 333 (Pt 1), 11-5.
Dard N and Peter M. (2006). Biaessays, 28, 146-56.

Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S, Teague J, Woffendin H,
Gamett MJ, Bottomley W, Davis N, Dicks E, Ewing R, Floyd Y, Gray K, Hall S,
Hawes R, Hughes J, Kosmidou V, Menzies A, Mould C, Parker A, Stevens C,
Watt S, Hooper S, Wilson R, Jayatilake H, Gusterson BA, Cooper C, Shipley J,
Hargrave D, Pritchard-Jones K, Maitland N, Chenevix-Trench G, Riggins GJ,
Bigner DD, Palmieri G, Cossu A, Flanagan A, Nicholson A, Ho J W, Leung SY,
Yuen ST, Weber BL, Seigler HF, Darrow TL, Paterson H, Marais R, Marshall CJ,
Wooster R, Stratton MR and F utreal PA. (2002). Nature, 417, 949-54.

Dawson J F and Holmes CF. (1999). Front Biasci, 4, D646-58.
Dhanasekaran N and Premkumar Reddy E. (1998). Oncogene, 17, 1447-55.
Ding WX and Yin XM. (2004). J Cell Mol Med, 8, 445-54.

Dorow DS, Devereux L, Dietzsch E and De Kretser T. (1993). EurJ Biochem, 213, 701-
10.

Dorow DS, Devereux L, Tu GF, Price G, Nichol] J K, Sutherland GR and Simpson RJ.
(1995). Eur J Biochem, 234, 492-500.

Dougherty MK, Muller J, Ritt DA, Zhou M, Zhou XZ, Copeland TD, Conrads TP,
Veenstra TD, Lu KP and Morrison DK. (2005). M01 Cell, 17, 215-24.

Douziech M, Laberge G, Grondin G, Daigle N and Blouin R. (1999). J. Histachem.
Cytochem., 47, 1287-1296.

Du Y, Bock BC, Schachter KA, Chao M and Gallo KA. (2005). J Biol Chem, 280,
42984-93.

141

Durkin J T, Holskin BP, Kopec KK, Reed MS, Spais CM, Steffy BM, Gessner G, Angeles
TS, Pohl J, Ator MA and Meyer SL. (2004). Biochemistry, 43, 16348-16355.

Eckey M, Tenbaum SP, Munoz A and Baniahmad A. (2003). Molecular and Cellular
Endocrinology, 213, 71.

Enslen H, Raingeaud J and Davis RI. (1998). J Biol Chem, 273, 1741-8.
Esch RK and Errede B. (2002). Prac Natl Acad Sci U S A, 99, 9160-5.
Ezoe K, Lee ST, Strunk KM and Spritz RA. (1994). Oncogene, 9, 935-8.

Fan G, Merritt SE, Kortenjann M, Shaw PE and Holzman LB. (1996). J Biol Chem, 271,
24788-93.

Figueroa C, Tarras S, TaylorJ and Vojtek AB. (2003). J Biol Chem, 278, 47922-7.
Frame S and Cohen P. (2001). Biochem J, 359, 1-16.

Fuchs SY, Adler V, Pincus MR and Ronai Z. (1998). Prac Natl Acad Sci U S A, 95,
10541-6.

Fuchs SY, Dolan L, Davis RJ and Ronai Z. (1996). Oncogene, 13, 1531-5.

Fuchs SY, Xie B, Adler V, Fried VA, Davis RJ and Ronai Z. (1997). J Biol Chem, 272,
32163-8.

Gallo KA and Johnson GL. (2002). Nat Rev Mal Cell Biol, 3, 663-72.

Gallo KA, Mark MR, Scadden DT, Wang Z, Gu Q and Godowski PJ. (1994). J Biol
Chem, 269, 15092-100.

Ganguly A, 00 TP, thetskaya M, Pratt R, Yarygina O, Momoi T, Kholodilov N and
Burke RE. (2004). J Neurachem, 88, 469-80.

142

Gao M, Labuda T, Xia Y, Gallagher E, Fang D, Liu YC and Karin M. (2004). Science,
306, 271-5.

Garcia EP, Mehta S, Blair LA, Wells DG, Shang J, Fukushima T, Fallon JR, Garner CC
and Marshall J. (1998). Neuron, 21, 727-39.

Gerwins P, Blank J L and Johnson GL. (1997). J Biol Chem, 272, 8288-95.
Gotoh I, Adachi M and Nishida E. (2001). J Biol Chem, 276, 4276-86.

Gotoh Y, Masuyama N, Dell K, Shirakabe K and Nishida E. (1995). J Biol Chem, 270,
25898-904.

Gross EA, Callow MG, Waldbaum L, Thomas S and Ruggieri R. (2002). J Biol Chem,
277, 13873-82.

Gupta S, Barrett T, Whitmarsh AJ, Cavanagh J, Sluss HK, Derijard B and Davis RJ.
(1996). Emba J, 15, 2760-70.

Haglund K, Di F iore PP and Dikic l. (2003). Trends in Biochemical Sciences, 28, 598.
Hall JP and Davis R]. (2002). JCell Biochem, 86, 1-11.
Hanks SK, Quinn AM and Hunter T. (1988). Science, 241, 42-52.

Hayashi M, Kim SW, Irnanaka-Yoshida K, Yoshida T, Abel ED, Eliceiri B, Yang Y,
Ulevitch RJ and Lee JD. (2004). J Clin Invest, 113, 1138-48.

Hayashi M and Lee JD. (2004). J Mal Med, 82, 800-8.

Hehner SP, Hofrnann TG, Ushmorov A, Dienz O, Wing-Lan Leung I, Lassarn N,
Scheidereit C, Droge W and Schmitz ML. (2000). M01 Cell Biol, 20, 2556-68.

Hicke L. (2001). Nat Rev Mal Cell Biol, 2, 195-201.

143

Hicke L and Dunn R. (2003). Annual Review of Cell and Developmental Biology, 19,
141-172.

Hicks SW and Machamer CE. (2005). .1. Biol. Chem, 280, 28944-28951.

Hirai S, Katoh M, Terada M, Kyriakis JM, Zon LI, Rana A, Avruch J and Ohno S.
(1997). J Biol Chem, 272, 15167-73.

Hoffinann R, Baillie GS, MacKenzie SJ, Yarwood SJ and Houslay MD. (1999). Emba J,
18, 893-903.

Holzman LB, Merritt SE and Fan G. (1994). J Biol Chem, 269, 30808—17.
Hu J C, O'Shea EK, Kim PS and Sauer RT. (1990). Science, 250, 1400-3.

Huang S, Jiang Y, Li Z, Nishida E, Mathias P, Lin S, Ulevitch RJ, Nemerow GR and Han
J. (1997). Immunity, 6, 739-49.

Hubbard SR, Mohammadi M and Schlessinger J. (1998). J Biol Chem, 273, 11987-90.
Hunter T. (2000). Cell, 100, 113.
Huse M and Kuriyan J. (2002). Cell, 109, 275-82.

Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, Moriguchi T, Takagi M, Matsumoto
K, Miyazono K and Gotoh Y. (1997). Science, 275, 90-4.

Ikeda A, Masaki M, Kozutsumi Y, Oka S and Kawasaki T. (2001). FEBS Lett, 488, 190-
5.

Ing YL, Leung IW, Heng HH, Tsui LC and Lassam NJ. (1994). Oncogene, 9, 1745-50.
Ip YT and Davis RJ. (1998). Curr Opin Cell Biol, 10, 205-19.

Israel A. (2000). Trends Cell Biol, 10, 129-33.

144

Ito M, Yoshioka K, Akechi M, Yarnashita S, Takamatsu N, Sugiyama K, Hibi M,
Nakabeppu Y, Shiba T and Yamamoto K-1. (1999). M01. Cell. Biol, 19, 7539-
7548.

Jaffe AB, Hall A and Schmidt A. (2005). Current Biology, 15, 405.
Johnson GL and Lapadat R. (2002). Science, 298, 1911-2.

Juo P, Kuo CJ, Reynolds SE, Konz RF, Raingeaud J, Davis R], Biemann HP and Blenis
J. (1997). Mol Cell Biol, 17, 24-35.

Kaiser ET. (1987). Biochem Soc Trans, 15, 1187-8.

Kato Y, Kravchenko VV, Tapping RI, Han J, Ulevitch RJ and Lee JD. (1997). Emba J,
16, 7054-66.

Katoh M, Hirai M, Sugimura T and Terada M. (1995). Oncogene, 10, 1447-51.
Kelkar N, Gupta S, Dickens M and Davis RJ. (2000). M01. Cell. Biol, 20, 1030-1043.
Kemp BE and Pearson RB. (1990). Trends Biochem Sci, 15, 342-6.

Kiefer F, Tibbles LA, Anafi M, J anssen A, Zanke BW, Lassam N, Pawson T, Woodgett
JR and Iscove NN. (1996). Emba J, 15, 7013-25.

Kim K-Y, Kim B-C, Xu Z and Kim S-J. (2004). J. Biol. Chem, 279, 29478-29484.
Kolch W. (2005). Nat Rev Mal Cell Biol, 6, 827-37.

Kramer RM, Roberts EF, Um SL, Borsch-Haubold AG, Watson SP, Fisher MJ and
Jakubowski JA. (1996). J Biol Chem, 271, 27723-9.

Kukekov NV, Xu Z and Greene LA. (2006). J Biol Chem.

Kyriakis JM and Avruch J. (2001). Physiol Rev, 81, 807-69.

145

Laine A and Ronai Z. (2005). Sci S T KE, 2005, re5.

Le-Niculescu H, Bonfoco E, Kasuya Y, Claret F X, Green DR and Karin M. (1999). Mol
Cell Biol, 19, 751-63.

Leclerc S, Gamier M, Hoessel R, Marko D, Bibb JA, Snyder GL, Greengard P, Biemat J,
Wu YZ, Mandelkow EM, Eisenbrand G and Meijer L. (2001). J Biol Chem, 276,
251-60.

Lee CM, Onesime D, Reddy CD, Dhanasekaran N and Reddy EP. (2002). PNAS, 99,
14189-14194.

Lee J C, Laydon JT, McDonnell PC, Gallagher TF, Kumar S, Green D, McNulty D,
Blumenthal MJ, Heys JR, Landvatter SW and et a1. (1994). Nature, 372, 739-46.

Lee J E, Hollenberg SM, Snider L, Turner DL, Lipnick N and Weintraub H. (1995).
Science, 268, 836-844.

Leung IW and Lassam N. (1998). J Biol Chem, 273, 32408-15.
Leung IW and Lassam N. (2001). J Biol Chem, 276, 1961-7.

Lin A, Minden A, Martinetto H, Claret F X, Lange-Carter C, Mercurio F, Johnson GL and
Karin M. (1995). Science, 268, 286-90.

Lindberg RA, Quinn AM and Hunter T. (1992). Trends Biochem Sci, 17, 114-9.

Liu TC, Huang CJ, Chu YC, Wei CC, Chou CC, Chou MY, Chou CK and Yang JJ.
(2000a). Biochem Biophys Res Commun, 274, 811-6.

Liu YF, Dorow D and Marshall J. (2000b). J Biol Chem, 275, 19035-40.
Lu KP and Hunter T. (1995). Cell, 81, 413-24.

Mackay DJ and Hall A. (1998). JBial Chem, 273, 20685-8.

146

Mancini M, Machamer CE, Roy S, Nicholson DW, Thomberry NA, Casciola-Rosen LA
and Rosen A. (2000). J. Cell Biol, 149, 603-612.

Manning G, Whyte DB, Martinez R, Hunter T and Sudarsanam S. (2002). Science, 298,
1912-34.

Marcora E, Gowan K and Lee J E. (2003). PNAS, 100, 9578-9583.

Maroney AC, Glicksman MA, Basma AN, Walton KM, Knight E, Jr., Murphy CA,
Bartlett BA, Finn JP, Angeles T, Matsuda Y, Neff NT and Dionne CA. (1998). J.
Neurosci, 18, 104-111.

Masaki M, Ikeda A, Shiraki E, Oka S and Kawasaki T. (2003). EurJ Biochem, 270, 76-
83.

Mata M, Merritt SE, Fan G, Yu GG and Holzman LB. (1996). J Biol Chem, 271, 16888-
96.

Matheny SA, Chen C, Kortum RL, Razidlo GL, Lewis RE and White MA. (2004).
Nature, 427, 256-60.

Matsuura H, Nishitoh H, Takeda K, Matsuzawa A, Amagasa T, Ito M, Yoshioka K and
Ichijo H. (2002). J. Biol. Chem, 277, 40703-40709.

Matten WT, Copeland TD, Ahn NG and Vande Woude GF. (1996). Dev Biol, 179, 485-
92.

Meijer L, Borgne A, Mulner O, Chong JP, Blow JJ, Inagaki N, Inagaki M, Delcros JG
and Moulinoux JP. (1997). EurJ Biochem, 243, 527-36.

Merritt SE, Mata M, Nihalani D, Zhu C, Hu X and Holzman LB. (1999). J Biol Chem,
274, 10195-202.

Mielke K and Herdegen T. (2000). Prag Neurabial, 61, 45-60.

Miki H, Setou M, Kaneshiro K and Hirokawa N. (2001). PNAS, 98, 7004-7011.

147

Millward TA, Zolnierowicz S and Hemmings BA. (1999). Trends Biochem Sci, 24, 186-
91.

Minden A, Lin A, Claret FX, Abo A and Karin M. (1995). Cell, 81, 1147-57.
Morfini G, Pigino G and Brady ST. (2005). Trends in Molecular Medicine, 11, 64.
Morgan DO and De Bondt HL. (1994). Curr Opin Cell Biol, 6, 239-46.

Moriguchi T, Kuroyanagi N, Yamaguchi K, Gotoh Y, Irie K, Kano T, Shirakabe K, Muro
Y, Shibuya H, Matsumoto K, Nishida E and Hagiwara M. (1996). J Biol Chem,
271, 13675-9.

Morita K, Saitoh M, Tobiume K, Matsuura H, Enomoto S, Nishitoh H and Ichijo H.
(2001). Emba J, 20, 6028-36.

Morris R, Cox H, Mombelli E and Quinn PJ. (2004). Subcell Biochem, 37, 35-118.
Morrison DK and Davis RJ. (2003). Annu Rev Cell Dev Biol, 19, 91-118.
Mota M, Reeder M, Chemoff J and Bazenet CE. (2001 ). J Neurosci, 21, 4949-57.

Mu F-T, Callaghan JM, Steele-Mortimer O, Stenmark H, Parton RG, Campbell PL,
McCluskey J, Yeo J-P, Tock EPC and Toh B-H. (1995). J. Biol. Chem, 270,
13503-13511.

Murakata C, Kaneko M, Gessner G, Angeles TS, Ator MA, O'Kane TM, McKenna
BAW, Thomas BA, Mathiasen JR and Saporito MS. (2002). Biaarganic &
Medicinal Chemistry Letters, 12, 147.

Musti AM, Treier M and Bohmann D. (1997). Science, 275, 400-2.

Nagata K, Puls A, Futter C, Aspenstrom P, Schaefer E, Nakata T, Hirokawa N and Hall
A. (1998). Emba J, 17, 149-58.

Nateri AS, Riera-Sans L, Da Costa C and Behrens A. (2004). Science, 303, 1374-8.

148

Nebreda AR and Hunt T. (1993). Emba J, 12, 1979-86.

Nick JA, Avdi NJ, Young SK, Lehman LA, McDonald PP, Frasch SC, Billstrom MA,
Henson PM, Johnson GL and Worthen GS. (1999). J Clin Invest, 103, 851-8.

Nicolas Dard MP. (2006). BiaEssays, 28, 146-156.

Nihalani D, Merritt S and Holzman LB. (2000). J Biol Chem, 275, 7273-9.

Nihalani D, Meyer D, Pajni S and Holzman LB. (2001). Emba J, 20, 3447-58.

Nihalani D, Wong HN and Holzman LB. (2003). J. Biol. Chem, 278, 28694-28702.

Noselli S. (1998). Trends Genet, 14, 33-8.

O'Shea EK, Klemm JD, Kim PS and Alber T. (1991). Science, 254, 539-44.

Ory S and Morrison DK. (2004). Curr Biol, 14, R277-8.

Osmani AH, McGuire SL and Osmani SA. (1991). Cell, 67, 283-91.

Parameswaran N, Hall CS, Bock BC, Sparks HV, Gallo KA and Spielman WS. (2002a).
Mol Cell Biochem, 241, 37-43.

Parameswaran N, Hall CS, Bock BC, Sparks HV, Gallo KA and Spielman WS. (2002b).
Cell Physiol Biochem, 12, 325-34.

Parkinson Study G. (2004). Neurology, 62, 330-332.

Pearl LH. (2005). Curr Opin Genet Dev, 15, 55-61.

Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berrnan K and Cobb
MH. (2001). Endocr Rev, 22, 153-83.

Pearson RB and Kemp BE. (1991). Methods Enzymal, 200, 62-81.

149

Phelan DR, Price G, Liu YF and Dorow DS. (2001). J Biol Chem, 276, 10801-10.

Pittet J-F, Lee H, Pespeni M, O'Mahony A, Roux J and Welch WJ. (2005). J Immunol,
174, 384-394.

Poitras L, Bisson N, Islam N and Moss T. (2003). Dev Biol, 254, 200-14.
Raman M and Cobb MH. (2003). Curr Biol, 13, R886-8.

Rana A, Gallo K, Godowski P, Hirai S, Ohno S, Zon L, Kyriakis JM and Avruch J.
(1996). J Biol Chem, 271, 19025-8.

Reynolds CH, Nebreda AR, Gibb GM, Utton MA and Anderton BH. (1997). J
Neurachem, 69, 191-8.

Roux PP and Blenis J. (2004). Microbial Mal Biol Rev, 68, 320-44.

Roy F, Laberge G, Douziech M, F erland-McCollough D and Therrien M. (2002). Genes
Dev, 16, 427-38.

Roy LM, Haccard O, Izumi T, Lattes BG, Lewellyn AL and Maller J L. (1996).
Oncogene, 12, 2203-11.

Sakuma H, Ikeda A, Oka S, Kozutsumi Y, Zanetta JP and Kawasaki T. (1997). J Biol
Chem, 272, 28622-9.

Saporito MS, Brown EM, Miller MS and Carswell S. (1999). J Pharmacol Exp T her,
288, 421-7.

Sathyanarayana P, Barthwal MK, Kundu CN, Lane ME, Bergrnann A, Tzivion G and
Rana A. (2002). M01 Cell, 10, 1527-33.

Savinainen A, Garcia EP, Dorow D, Marshall J and Liu YF. (2001). J Biol Chem, 276,
11382-6.

SchoorlemmerJ and Goldfarb M. (2001). Curr Biol, 11, 793-7.

150

Schulenberg B, Goodman TN, Aggeler R, Capaldi RA and Patton WF. (2004).
Electrophoresis, 25, 2526-32.

Setemes OM, Mikalsen T, Johansen B, Michaelsen E, Armstrong CG, Monice NA,
Turgeon B, Meloche S, Moens U and Keyse SM. (2004). Emba J, 23, 4780-91.

Sheikh-Hamad D and Gustin MC. (2004). Am J Physiol Renal Physiol, 287, F1102-10.

Shen YH, Godlewski J, Zhu J, Sathyanarayana P, Leaner V, Birrer MJ, Rana A and
Tzivion G. (2003). J Biol Chem, 278, 26715-21.

Simons K and Toomre D. (2000). Nat Rev Mol Cell Biol, 1, 31-9.

Slusarewicz P, Nilsson T, Hui N, Watson R and Warren G. (1994). J. Cell Biol, 124,
405-413.

Steinberg TH, Agnew BJ, Gee KR, Leung WY, Goodman T, Schulenberg B,
Hendrickson J, Beechem J M, Haugland RP and Patton WF. (2003). Prateamics,
3, 1128-44.

Stewart 8, Sundararn M, Zhang Y, Lee J, Han M and Guan KL. (1999). Mal Cell Biol,
19, 5523-34.

Stickney JT, Bacon WC, Rojas M, Ratner N and Ip W. (2004). Cancer Res, 64, 2717-24.
Stronach B. (2005). Dev Dyn, 232, 575-84.
Stronach B and Penimon N. (2002). Genes Dev, 16, 377-87.

Sun W, Kesavan K, Schaefer BC, Garrington TP, Ware M, Johnson NL, Gelfand EW and
Johnson GL. (2001). J Biol Chem, 276, 5093-100.

Sun Y and MacRae TH. (2005). Cell Mal Life Sci, 62, 2460-76.

Swenson KI, Winkler KB and Means AR. (2003). Mal Biol Cell, 14, 156-72.

151

Tanaka S and Hanafirsa H. (1998). J Biol Chem, 273, 1281-4.
Tapon N, Nagata K, Larnarche N and Hall A. (1998). Emba J, 17, 1395-404.

Teramoto H, Coso OA, Miyata H, Igishi T, Miki T and Gutkind J S. (1996). J Biol Chem,
271, 27225-8.

Tibbles LA, Ing YL, Kiefer F, Chan J, Iscove N, Woodgett JR and Lassam NJ. (1996).
Emba J, 15, 7026-35.

Tournier C, Hess P, Yang DD, Xu J, Turner TK, Nimnual A, Bar-Sagi D, Jones SN,
Flavell RA and Davis RJ. (2000). Science, 288, 870-4.

Toumier C, Whitmarsh AJ, Cavanagh J, Barrett T and Davis RJ. (1997). Prac Natl Acad
Sci USA, 94, 7337-42.

Toumier C, Whitmarsh AJ, Cavanagh J, Barrett T and Davis RJ. (1999). M01 Cell Biol,
19, 1569-81.

Vacratsis PO and Gallo KA. (2000). J Biol Chem, 275, 27893-900.
Vacratsis PO, Phinney BS, Gage DA and Gallo KA. (2002). Biochemistry, 41, 5613-24.

Veeranna, Amin ND, Ahn NG, Jaffe H, Winters CA, Grant P and Pant HC. (1998). J
Neurosci, 18, 4008-21.

Wan PTC, Gamett MJ, Roe SM, Lee S, Niculescu-Duvaz D, Good VM, Project CG,
Jones CM, Marshall CJ and Springer CJ. (2004). Cell, 116, 855.

Wartmann M and Davis RJ. (1994). J Biol Chem, 269, 6695-701.

Wartmann M, Hofer P, Turowski P, Saltiel AR and Hynes NE. (1997). J Biol Chem, 272,
3915-23.

Wegele H, Muller L and Buchner J. (2004). Rev Physiol Biochem Pharmacol, 151, 1-44.

152

Weston CR and Davis RJ. (2002). Curr Opin Genet Dev, 12, 14-21.

Whitmarsh AJ, Cavanagh J, Toumier C, Yasuda J and Davis RJ. (1998). Science, 281,
1671-4.

Witowsky J A and Johnson GL. (2003). J Biol Chem, 278, 1403-6.
Xia Y and Karin M. (2004). Trends in Cell Biology, 14, 94.
Xu S and Cobb MH. (1997). J Biol Chem, 272, 32056-60.

Xu S, Robbins DJ, Christerson LB, English JM, Vanderbilt CA and Cobb MH. (1996).
Prac Natl Acad Sci U S A, 93, 5291-5.

Xu Z, Kukekov NV and Greene LA. (2003). Emba J, 22, 252-61.
Xu Z, Kukekov NV and Greene LA. (2005). M01 Cell Biol, 25, 9949-59.

Xu Z, Maroney AC, Dobrzanski P, Kukekov NV and Greene LA. (2001). Mol Cell Biol,
21, 4713-24.

Yamaguchi K, Shirakabe K, Shibuya H, Irie K, Oishi I, Ueno N, Taniguchi T, Nishida E
and Matsumoto K. (1995). Science, 270, 2008-11.

Yang JJ. (2002). Biochem Biophys Res Commun, 297, 105-10.

Yasuda J, Whitmarsh AJ, Cavanagh J, Sharma M and Davis RJ. (1999). Mol Cell Biol,
19, 7245-54.

Yu C, Minemoto Y, Zhang J, Liu J, Tang F, Bui TN, XiangJ and Lin A. (2004). M01
Cell, 13, 329-40.

Yu H, Chen JK, Feng S, Dalgamo DC, Brauer AW and Schreiber SL. (1994). Cell, 76,
933-45.

Zhang B and Zheng Y. (1998). J Biol Chem, 273, 25728-33.

153

Zhang H and Gallo KA. (2001). J Biol Chem, 276, 45598-603.

Zhang H, Wu W, Du Y, Santos SJ, Conrad SE, Watson JT, Grammatikakis N and Gallo
KA. (2004). J Biol Chem, 279, 19457-63.

Zheng C, Xiang J, Hunter T and Lin A. (1999). J Biol Chem, 274, 28966-71.
Zheng CF and Guan KL. (1993). J Biol Chem, 268, 11435-9.

Zhou G, Bao ZQ and Dixon J E. (1995). J Biol Chem, 270, 12665-9.

Ziogas A, Moelling K and Radziwill G. (2005). J. Biol. Chem, 280, 24205-24211.

Zoller MJ, Yonemoto W, Taylor SS and Johnson KB. (1991). Gene, 99, 171-9.

154