Regulation of nerve growth factor signaling by protein ...

University of Iowa

Iowa Research Online Theses and Dissertations

2008

Regulation of nerve growth factor signaling by protein phosphatase 2A Michael J. Van Kanegan

Recommended Citation Van Kanegan, Michael J.. "Regulation of nerve growth factor signaling by protein phosphatase 2A." PhD diss., University of Iowa, 2008. http://ir.uiowa.edu/etd/279.

This dissertation is available at Iowa Research Online: http://ir.uiowa.edu/etd/279

REGULATION OF NERVE GROWTH FACTOR SIGNALING BY PROTEIN PHOSPHATASE 2A

by Michael J. Van Kanegan

An Abstract Of a thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Pharmacology in the Graduate College of The University of Iowa May 2008 Thesis Supervisor: Associate Professor Stefan Strack

1

ABSTRACT The goal of this dissertation research is to determine novel regulatory mechanisms of neurotrophin signaling mediated by protein phosphatase 2A (PP2A). PP2A is a ubiquitous Ser/Thr phosphatase that removes phosphates from proteins to switch their activity on or off. The substrate specificity and subcellular localization of PP2A is determined by almost 20 regulatory subunits that associate with a core dimer built of catalytic and scaffold subunits. Since there are more than 48 possible heterotrimers, studying the function of PP2A poses many challenges. Therefore we have devised a strategy, using scaffold subunit knockdown and mutant replacement, to discern the function of specific families of regulatory subunits. With this approach, I have identified specific PP2A holoenzymes that modulate nerve growth factor (NGF) signaling pathways by positively regulating TrkA receptor tyrosine kinase activity. Many studies have shown that NGF is required for the survival and differentiation of sensory and sympathetic neurons. Additionally, NGF is implicated in many neurodegenerative diseases including Alzheimer’s disease, Parkinson’s disease as well as neuropathic pain. NGF elicits its biological effect through sustained activity of the TrkA receptor and stimulated signaling cascades, including the MAP kinase pathway. Although PP2A has been shown to modulate the mitogen-activated protein (MAP) kinase pathway both positively and negatively at multiple levels, work described herein introduces yet another level of regulation. Specifically, I have shown that PP2A/B’ holoenzymes complex with the TrkA neurotrophin receptor to potentiate receptor tyrosine kinase activity, downstream effector kinase activation, neurite outgrowth, and neuronal differentiation. On the other hand, extracellular signal regulated kinase (ERK), a terminal effector in the MAP kinase pathway was shown to phosphorylate a residue in the juxtamembrane region of TrkA and impose feedback inhibition of receptor activity. Collectively, these data suggest a model in which PP2A and ERK oppose each other in the regulation of TrkA

2

receptor activity and downstream signaling cascades that govern neuronal differentiation and maintenance.

Abstract Approved: ____________________________________ Thesis Supervisor ____________________________________ Title and Department ____________________________________ Date

REGULATION OF NERVE GROWTH FACTOR SIGNALING BY PROTEIN PHOSPHATASE 2A

by Michael J. Van Kanegan

A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Pharmacology in the Graduate College of The University of Iowa May 2008 Thesis Supervisor: Associate Professor Stefan Strack

Graduate College The University of Iowa Iowa City, Iowa

CERTIFICATE OF APPROVAL _______________________ PH.D. THESIS _______________ This is to certify that the Ph.D. thesis of Michael J. Van Kanegan has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Pharmacology at the May 2008 graduation. Thesis Committee: ___________________________________ Stefan Strack, Thesis Supervisor ___________________________________ Steven Green ___________________________________ Johannes Hell ___________________________________ John Koland ___________________________________ Fred Quelle

To my loving and supportive family

ii

“Success is going from failure to failure without losing enthusiasm.” Winston Churchill

iii

ACKNOWLEDGMENTS I would first like to thank Dr. Stefan Strack for the opportunities in your lab and providing an exciting project that would challenge and develop my scientific skills. Your scientific knowledge, insight and wisdom has guided me through my dissertation research and inspired me throughout many years of graduate school. I would also like to thank my committee members Drs. Steven Green, Johannes Hell, John Koland and Fred Quelle for excellent guidance and support. I am grateful for the research funding provided by the PhRMA foundation. I want to give a special thanks to all the Strack lab members, both past and present, for making my time in “purgatory” both memorable and enjoyable. Tom for loads of technical assistance and many thought-provoking discussions. My fellow gradstudents: Audrey, Shanna, Beth and Amit for entertaining grad room banter. A special thanks to Ron for many inspiring discussions, frequent technical support and a bottomless cup of java. I have really learned a lot from your wisdom. I also want to thank the members of Hell, both those who escaped and those who remain, for many fun times and commiseration. I especially want to thank my family. My parents Jeff and Gail, and my brother Tim – thank you for your endless support and instilling in me the confidence, independence and determination needed to complete my degree. My in-laws, MIL and FIL – thanks for the abundant support and many weekends of child care. Melissa, thanks for your expert editorial skills and family support during hectic times. I would like to thank my adorable daughters, Indira and Maya for bringing joy to my life every day. Finally, I would like to thank my loving wife, Michelle. You have provided neverending encouragement and support throughout my graduate career. Your forethought, insight and wisdom have guided me through my scientific path. You have been, and always will be an inspiration in every aspect of my life.

iv

ABSTRACT The goal of this dissertation research is to determine novel regulatory mechanisms of neurotrophin signaling mediated by protein phosphatase 2A (PP2A). PP2A is a ubiquitous Ser/Thr phosphatase that removes phosphates from proteins to switch their activity on or off. The substrate specificity and subcellular localization of PP2A is determined by almost 20 regulatory subunits that associate with a core dimer built of catalytic and scaffold subunits. Since there are more than 48 possible heterotrimers, studying the function of PP2A poses many challenges. Therefore we have devised a strategy, using scaffold subunit knockdown and mutant replacement, to discern the function of specific families of regulatory subunits. With this approach, I have identified specific PP2A holoenzymes that modulate nerve growth factor (NGF) signaling pathways by positively regulating TrkA receptor tyrosine kinase activity. Many studies have shown that NGF is required for the survival and differentiation of sensory and sympathetic neurons. Additionally, NGF is implicated in many neurodegenerative diseases including Alzheimer’s disease, Parkinson’s disease as well as neuropathic pain. NGF elicits its biological effect through sustained activity of the TrkA receptor and stimulated signaling cascades, including the MAP kinase pathway. Although PP2A has been shown to modulate the mitogen-activated protein (MAP) kinase pathway both positively and negatively at multiple levels, work described herein introduces yet another level of regulation. Specifically, I have shown that PP2A/B’ holoenzymes complex with the TrkA neurotrophin receptor to potentiate receptor tyrosine kinase activity, downstream effector kinase activation, neurite outgrowth, and neuronal differentiation. On the other hand, extracellular signal regulated kinase (ERK), a terminal effector in the MAP kinase pathway was shown to phosphorylate a residue in the juxtamembrane region of TrkA and impose feedback inhibition of receptor activity. Collectively, these data suggest a model in which PP2A and ERK oppose each other in the regulation of TrkA

v

receptor activity and downstream signaling cascades that govern neuronal differentiation and maintenance.

vi

TABLE OF CONTENTS LIST OF TABLES .........................................................................................................ix LIST OF FIGURES.........................................................................................................x LIST OF ABBREVIATIONS .........................................................................................xi CHAPTER I INTRODUCTION .....................................................................................1 The structural and functional complexity of PP2A.........................................2 PP2A catalytic subunit ...........................................................................3 PP2A scaffold subunit............................................................................4 PP2A regulatory subunits .......................................................................4 Holoenzyme assembly............................................................................7 Neurotrophin signaling..................................................................................8 Receptor tyrosine kinase family..............................................................8 TrkA receptor tyrosine kinase.................................................................9 MAP kinase signaling pathway ..............................................................9 Cell systems ................................................................................................10 Dissertation research focus..........................................................................11 CHAPTER II SPECIFIC PP2A HETEROTRIMERS MEDIATE TRKA RECEPTOR SIGNALING AND NEURONAL DIFFERENTIATION........15 Abstract ......................................................................................................15 Introduction ................................................................................................15 NGF signaling......................................................................................16 PP2A holoenzyme in signal transduction..............................................16 Materials and methods ................................................................................18 Cell culture and plasmids .....................................................................18 Antibodies............................................................................................18 Other reagents ......................................................................................19 Immunoprecipitations...........................................................................19 Animal tissue preparation.....................................................................20 Microcystin pulldown ..........................................................................20 Quantitative immunoblotting with phospho-specific antibodies ............21 MAPK reporter assays..........................................................................21 Neurite outgrowth assay .......................................................................22 DNA content FACS analysis ................................................................22 Surface biotin labeling..........................................................................23 Results ........................................................................................................23 PP2A is required to sustain TrkA activity.............................................23 PP2A/B’ family of regulatory subunits modulate TrkA autophosphorylation.............................................................................25 PP2A/B’
positively regulates TrkA receptor activity ..........................26 Multiple B’ family members associate with the TrkA receptor and regulate its activity ...............................................................................28 B’
and B’ holoenzymes potentiate NGF dependent neurite outgrowth.............................................................................................30 B’
holoenzymes potentiate NGF dependent differentiation in PC12 cells. ...........................................................................................31 vii

Discussion...................................................................................................32 PP2A regulation of NGF signaling .......................................................32 PP2A mediates long term NGF signaling .............................................34 B’ regulatory subunits target specific substrates ...................................35 PP2A regulates cellular fate by sustaining NGF signaling ....................36 CHAPTER III ERK TARGETS THE TRKA JUXTAMEMBRANE DOMAIN AND NEGATIVELY REGULATES RECEPTOR ACTIVITY THROUGH SERINE PHOSPHORYLATION ............................................46 Abstract ......................................................................................................46 Introduction ................................................................................................46 Nerve growth factor signaling ..............................................................46 Transient vs. sustained signaling ..........................................................48 Serine phospho-regulation of RTKs......................................................49 Materials and methods ................................................................................49 Cell culture and plasmids .....................................................................49 Antibodies............................................................................................50 Other reagents ......................................................................................50 ERK activation and inhibition studies...................................................51 Immunoprecipitations...........................................................................51 Quantitative immunoblotting with phospho-specific antibodies ............52 Metabolic labeling................................................................................52 GST fusion protein plasmid generation and protein purification ...........53 In vitro kinase assays............................................................................54 Results ........................................................................................................55 PP2A inhibition causes an increase in TrkA serine phosphorylation....................................................................................55 Screen for specific kinases that modulates TrkA activity ......................56 TrkA activity is modulated by ERK......................................................57 ERK regulates TrkA serine phosphorylation.........................................59 The TrkA juxtamembrane domain is targeted by ERK in vitro. ............60 Functional effects of S471 mutations....................................................63 Discussion...................................................................................................65 Okadaic acid promotes TrkA serine phosphorylation............................65 TrkA serine phosphorylation is a potential form of receptor crosstalk.......................................................................................................66 Discovering a Ser/Thr kinase that targets TrkA ....................................67 TrkA regulation through the juxtamembrane domain............................68 ERK regulates TrkA Tyr phosphorylation ............................................68 ERK targets S471 in the TrkA juxtamembrane domain ........................69 CHAPTER IV CONCLUSIONS AND FUTURE DIRECTIONS..................................80 Implications of sustained TrkA signaling ....................................................80 Involvement of PP2A in sustained signaling................................................80 TrkA ubiquitination.....................................................................................81 TrkA juxtamembrane regulation..................................................................83 TrkA juxtamembrane binding proteins ........................................................84 REFERENCES..............................................................................................................87

viii

LIST OF TABLES Table 3.1 Summary of pharmacological agents used to screen TrkA kinase activity. ........................................................................................................70 Table 3.2 Summary of GST-TrkA mutant intrinsic kinase activity. ...............................71

ix

LIST OF FIGURES Figure 1.1 The composition and structure of the protein phosphatase 2A holoenzyme..................................................................................................12 Figure 1.2 PP2A regulatory subunit sequence homology...............................................13 Figure 1.3 NGF signaling cascades. ..............................................................................14 Figure 2.1 Sustained TrkA signaling requires PP2A activity. ........................................37 Figure 2.2 PP2A does not modulate TrkA receptor activity or downstream signaling after acute NGF stimulation...........................................................38 Figure 2.3 PP2A does not affect TrkA receptor surface expression. ..............................39 Figure 2.4 PP2A B’ family of regulatory subunits is required to sustain TrkA activity. ........................................................................................................40 Figure 2.5 B’
potentiates TrkA signaling upstream of Ras. .........................................41 Figure 2.6 Multiple B’ subunits associate with the TrkA receptor. ................................42 Figure 2.7 Multiple B’ subunits regulate TrkA activity. ................................................43 Figure 2.8 PP2A B’ members potentiate neurite outgrowth in PC12 cells. ....................44 Figure 2.9 PP2A B’
potentiates neuronal differentiation in PC12 cells........................45 Figure 3.1 PP2A inhibition fosters TrkA Ser/Thr hyperphosphorylation while decreasing Tyr phosphorylation after prolonged NGF stimulation. ...............72 Figure 3.2 Enhanced TrkA Ser/Thr phosphorylation corresponds with decreased TrkA activity after PP2A inhibition..............................................................73 Figure 3.3 ERK activity modulates TrkA phosphorylation in PC12 cells.......................74 Figure 3.4 ERK activity modulates TrkA phosphorylation in PC6-3 cells. ....................75 Figure 3.5 ERK inhibition decreases TrkA Ser/Thr phosphorylation after prolonged NGF treatment.............................................................................76 Figure 3.6 ERK targets TrkA at Serine 471 in vitro.......................................................77 Figure 3.7 Ser 471 is targeted for phosphorylation after NGF stimulation. ....................78 Figure 3.8 Ser 471 impacts TrkA activity in HEK cells.................................................79 Figure 4.1 Model of TrkA function regulated by PP2A and ERK..................................86

x

LIST OF ABBREVIATIONS Akt

protein kinase B

ATP

adenosine triphosphate

BDNF

brain-derived growth factor

BSA

Bovine serum albumin

Cdk

Cyclin-dependent kinase

cDNA

Complementary deoxyribonucleic acid

CTRL

control

DMSO

dimethyl sulfoxide

DNA

deoxyribonucleic acid

Dox

Doxycycline

EGF

epidermal growth factor

EGFR

epidermal growth factor receptor

ERK

extracellular regulated kinase

FACS

fluorescence activated cell sorting

FBS

fetal bovine serum

FGF

fibroblast growth factor

FGFR

fibroblast growth factor receptor

Grb2

growth factor receptor-bound protein-2

GST

glutathione S-transferase

HA

hemagglutinin

HEAT

Huntingtin, elongation, A subunit, TOR

HRP

horse radish peroxidase

IB

immunoblot

IgG

immunoglobulin

IP

Immunoprecipitate/immunoprecipitation

xi

JM

juxtamembrane

kDa

kilodalton

MAPK

mitogen activated protein kinase

Mek1

MAPK kinase 1


g

microgram

ng

nanogram

NGF

nerve growth factor

NT-3

neurotrophin-3

NT4/5

neurotrophin-4/5

OA

okadaic acid

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate buffered saline

PC12

Pheochromocytoma clone 12

PC6-3

Pheochromocytoma clone 6-3

PCR

polymerase chain reaction

PDGFR

platelet-derived growth factor receptor

PI

propidium iodide

PI3K

phosphatidylinositol 3-kinase

PKA

cAMP dependent protein kinase A

PLC

phospholipase C

PP1

Protein phosphatase 1

PP2A

Protein phosphatase 2A

PP2B

Protein phosphatase 2B (Calcineurin)

PP2C

Protein phosphatase 2C

Raf1

MAP kinase kinase kinase

Rb

retinoblastoma

RNA

Ribonucleic acid xii

RNAi

RNA interference

RTK

receptor tyrosine kinase

SDS

sodium dodecyl sulfate

Ser

serine

Shc

src homology 2/a-collagen-related protein

shRNA

Short hairpin RNA

SOS

son-of-sevenless

TBS

tris-buffered saline

TfnR

transferrin receptor

Thr

threonine

TrkA

tropomyosin related kinase A

TrkB

tropomyosin related kinase B

TrkC

tropomyosin related kinase C

Tyr

tyrosine

WD

Tryptophan aspartate domain

WT

Wild-type

YOP

Yersinia phosphatase

xiii

1

CHAPTER I INTRODUCTION Maturing neurons first begin to emerge from neural progenitor cells shortly after neural tube closure in human embryos at ~30 days of development (O'Rahilly and Muller, 2007). At this point, cell division accelerates to more than 250,000 neurons per minute (Sterberg, 2003). Remarkably, a single layer of cells will continue to divide for the duration of fetal development and the entire central and peripheral nervous system will materialize. Both genetic determinants as well as extrinsic molecular cues shape the fate of each cell as it migrates to a specific anatomical region of the nervous system, differentiates into a specific neuronal sub-type, and develops synaptic connections that facilitate higher order cognitive functions. Surprisingly, between 45 and 75% of all neurons created during development do not survive due to insufficient growth factors and the inability to maintain stable synapses. The remaining surviving neurons lay the foundation for neural networks that will last a lifetime. The anatomical development of the nervous system is very well established. The mysteries that are slowly being understood are the molecular mechanisms that propel neuronal differentiation, survival and eventually cell death. To fully understand how neuronal physiology operates, we must look at operations that occur at the cellular level. We must ask how communication occurs between cells and how communication occurs within cells to maintain proper cellular function. Since neurons are not likely a rejuvenating commodity, it is important to understand how normal, thriving neurons function, in efforts to prevent and treat pathologies that eventually destroy our brain and nervous system. Cellular communication, termed ‘signal transduction’, that supports the differentiation and survival of central and peripheral nervous tissue is the focus of this doctoral research. Typically, extracellular cues are transmitted through transmembrane

2

receptors that initiate signaling networks that can induce cell motility, promote cell survival or apoptosis, activate gene transcription and regulate neuronal differentiation. Receptors have the ability to amplify a small originating stimulus by activating many kinases, phosphatases and signaling molecules that, in turn, drive broad cellular activities that define each cell. These networks of signaling molecules are vast, and they intertwine and interplay as they adjust and fine-tune the information flowing through the cell. Even though signal transduction is typically represented and studied as canonical in nature and linear from start to finish, this is simply not an accurate view of cellular biochemistry and physiology. Our understanding of how cells function is limited because there are many unknown players within these networks that we don’t see, can’t see, with our current technology and understanding. However, our understanding is continually becoming clearer. As each new protein is identified, new interactions are determined and new functions are assigned, our knowledge of the biological puzzle becomes one step closer to being complete. With complete understanding, diseases can be avoided, treated and cured. Herein, I present data illustrating novel regulation of signaling networks within the nervous system that are vital for neuronal development and survival. These data describe the interaction and regulation of the prototypical neurotrophin receptor, TrkA by the ubiquitously expressed regulator of signaling pathways, protein phosphatase 2A (PP2A). The structural and functional complexity of PP2A Protein phosphatase 2A (PP2A) is an enzyme that removes phosphate modifications from serine (Ser) and threonine (Thr) residues on protein substrates to regulate their activity. PP2A is found in every cell and likely composes almost 1% of cellular protein content (Walter). Being such an abundant protein, PP2A has been shown to regulate numerous signaling pathways, neuronal differentiation, cell cycle progression,

3

DNA replication, transcription and translation in both a physiological and pathological setting (Janssens and Goris, 2001; Schonthal, 2001). PP2A is not just one enzyme, but a collection of holoenzymes composed of three groups of subunits: A-scaffold, C-catalytic and regulatory subunits (Figure 1.1). The A and C subunits form a core dimer that associates with any of four families of regulatory subunits: B, B’, B” and B’” (Janssens and Goris, 2001; Junttila et al., 2007). These regulatory subunits dictate substrate specificity and subcellular localization and are differentially expressed in many cell types. PP2A catalytic subunit In vertebrates, both the scaffold and catalytic subunits are encoded by two genes, A
/ and C
/ respectively. While the exact tissue expression patterns differ slightly between
and  isoforms; the
isoform is generally ten times more abundant than the  isoforms (Usui et al., 1988; Zhou et al., 2003). The C
and  isoforms are globular proteins that are 97% identical with divergent N-termini (Arino et al., 1988; Cho and Xu, 2007; da Cruz e Silva and Cohen, 1987; Stone et al., 1987). Despite being almost identical, there is evidence that each isoform has distinct roles during embryonic development (Gotz et al., 1998; Gotz et al., 2000). The catalytic subunit functions by orienting the substrate and clamping two -sheets on a phosphorylated Ser or Thr (Denu et al., 1996). Histidine and aspartate residues coordinate the phosphate as divalent metal ions (Mn2+, Zn2+, Fe2+ /3+) in each -sheet activate water molecules. These high-energy molecules catalyze a nucleophilic attack that removes the phosphate group, allowing the histidine to protonate the Ser/Thr residue (Barford, 1995; Denu et al., 1996). Although the C-subunit can be purified from bacteria and has catalytic activity, only the A/C dimers and A/B/C heterotrimers have been isolated from tissue (Usui et al., 1988).

4

PP2A scaffold subunit The A
and  scaffolds have 87% identity (Hemmings et al., 1990) and show slightly different expression patterns and functions (Zhou et al., 2003). The scaffold forms a horseshoe-shaped structure composed of 15 non-identical HEAT (Huntingtin, elongation/A subunit/TOR) repeats (Andrade and Bork, 1995) that mediate proteinprotein interactions with the catalytic and regulatory subunits (Cho and Xu, 2007; Groves et al., 1999; Xu et al., 2006). The catalytic subunit associates with the C-terminal repeats, 11-15 (Kremmer et al., 1997), while the regulatory subunits associate with HEAT repeats 1-10 (Cho and Xu, 2007; Ruediger et al., 1994; Ruediger et al., 1992; Xu et al., 2006). The various subunit families (B, B’, B”, B”’) associate with the scaffolding subunit through slightly different interaction sites unique to each family. Thereby, point mutations have been developed to disrupt regulatory subunit binding and selectivity offering excellent tools for determining PP2A function in culture. For example, mutations at Glu100 and 101 to Arg (EE100RR) disrupt all regulatory subunit binding, while the Asp139/ Trp140 to Arg (DW139RR) mutation only allows B’ family subunits to bind and Asp 177/ Thr178/ Pro179 to Ala (DTP177AAA) mutations only disrupt B’ subunit binding, leaving B and B” subunit association intact (Ruediger et al., 1999). These three mutants are described in further detail in Chapter 2 (Figure 2.3) when they are used to identify specific PP2A holoenzymes that regulate TrkA activity. PP2A regulatory subunits PP2A regulatory subunits have been divided into three well-characterized families (B, B’, B”) and one debated family (B’”) (Moreno et al., 2000). Subunit association with the AC dimer is mutually exclusive making each heterotrimer a unique functional phosphatase. Interestingly, the families are extremely diverse in structure and sequence homology (Figure 1.2 A), but each links with high affinity to partially overlapping regions of the A scaffolding subunit (see above). The regulatory subunits have gone

5

through many confusing generations of nomenclature that are described below. For simplicity, B, B’, B” will be used throughout this thesis. The B-family (B55, PR55 or PPP2R2) comprises four gene products (
, , , ) that are predicted to fold into a seven bladed -propeller formed degenerate WD-repeats. WD-repeats are minimally conserved, 40 amino acid -sheets that generally end in tryptophan-aspartate (WD) dipeptides (Strack et al., 2002). B subunits are 86% identical and tissue expression differs among subunits where
and  are abundantly and ubiquitously expressed in many tissues, while  and  are highly enriched in the brain (Mayer et al., 1991; Strack et al., 1999). There is also differential expression within specific brain regions as well as distinct subcellular localizations within neurons. B
, B and B are mainly cytosolic, while B is associated with the cytoskeleton (Strack et al., 1998). The cellular distribution dictates the numerous proteins targeted by PP2A/B. For example, B
regulates TGF receptor activity at the membrane (Griswold-Prenner et al., 1998) and controls cell cycle progression by dephosphorylating cytosolic CDC proteins (Clarke et al., 1993; Healy et al., 1991; Mayer-Jaekel et al., 1993). B
has also been shown to regulate cell survival by targeting Akt (Nawa et al., 2008). Likewise B
and B have been shown to directly regulate ERK activity (Adams et al., 2005; Van Kanegan et al., 2005), while B2 is targeted to mitochondria in response to cellular stress (Dagda et al., 2003). Importantly, B is shown to be a positive regulator of MAPK signaling by potentially regulating B-Raf (Strack, 2002) The B’-family (B56, PR61 or PPP2R5) contains five isoforms (
, , , , ) that share 71-88% sequence identity (Figure 1.2 B), (McCright et al., 1996a) with major divergence occurring on the extreme N- and C-termini. Although B’ subunits are very similar in sequence they are reported to have diverse functions on numerous substrates. B’
, , and  are typically cytosolic in localization, while B’ shuttles between the nucleus and cytoplasm and B’ is mostly nuclear. These subunits are phospho-proteins that share a structure described as a HEAT-containing subunit, very similar to the

6

scaffold subunit (Figure 1.1), (Cho and Xu, 2007; Xu et al., 2006). The repeating HEAT structure is proposed to be dynamic in nature to allow interactions with various proteins (Mumby, 2007). B’ subunits have been shown to regulate CDC proteins and cell cycle progression (Bennin et al., 2002), paxillin and cell motility (Ito et al., 2000), APC and Wnt signaling (Hsu et al., 1999), ERK and MAPK signaling (Letourneux et al., 2006; Silverstein et al., 2002), Akt and survival signaling (Rocher et al., 2007), MDM2, p53 and tumor suppression (Chen et al., 2004; Janssens et al., 2005; Okamoto et al., 2002) as well as tyrosine hydroxylase phosphorylation and catecholamine synthesis (Saraf et al., 2007). Distinguishing B’ subunit substrate specificity is a major goal of the experiments described in Chapter 2. Interestingly, results shown herein that B’, B’ and B’ have distinct specificity on TrkA function as compared to B’
and B’ despite extremely high sequence homology (Figure 1.2 B). The B”-family (PR72 or PPP2R3) consists of three isoforms (
/PR72/130, /PR59, /PR48), with B”
/PR72 having a 130 kD splice variant (B”
/PR130). B” subunits share a 56-65% sequence homology (Figure 1.2 A). These proteins are both nuclear and cytosolic with B”
/PR72 being highly expressed in the heart and muscle and the 130 kD splice variant, B”
/PR130, additionally being expressed in brain, lung and placenta (Hendrix et al., 1993). B” subunits have EF hands that bind calcium and structurally look similar to calmodulin. B” (subunits regulate Wnt signaling (PR130) {Creyghton, 2006 #173) B” subunits have EF hands that bind calcium and structurally are similar to calmodulin. B” (subunits regulate Wnt signaling (PR130) (Creyghton et al., 2006)) and proteins such as DARRP-32 (PR72) (Ahn et al., 2007b). B” and B” regulate cell cycle though Rb-related protein p107 and Cdc6 respectively (Voorhoeve et al., 1999; Yan et al., 2000). The B”’-family (PR93/PR110) consists of three conserved calmodulin-binding proteins, SG2NA, striatin and mMOB1(Moreno et al., 2000, Moreno, 2001 #215). They structurally resemble B-family subunits by having WD repeats as well as contain a

7

domain conserved with B’ subunits. Each poses distinct subcellular localization; SG2NA is found in the nucleus, while striatin is enriched in the post-synaptic densities and membrane (Moreno et al., 2000). Both proteins associate with calmodulin in a Ca++dependent manner and likely play a role in regulating calcium signaling (Moreno et al., 2000). There is debate within the field whether B’” proteins are indeed regulatory subunits as opposed to A/C interacting proteins. Holoenzyme assembly The complexity of PP2A is indeed grand. With two catalytic and scaffolding isoforms and three families of regulatory subunits each having several members and multiple splice variants, the number of possible heterotrimers is well over 50. Although regulatory subunit binding regions have been identified as discussed above, mechanisms that dictate holoenzyme assembly and subunit specificity still remains largely undefined. Subunit expression can clearly regulate subunit availability. However this cannot solely account for subunit composition given that every cell expresses multiple genes simultaneously. Regulatory subunit phosphorylation certainly plays a key role in B’ subunit association (Letourneux et al., 2006), however there is limited information that defines this mechanism. Additionally, many reports have shown the importance of catalytic subunit methylation and phosphorylation on holoenzyme activity and hypothesized that methylation was required for regulatory subunit association (Bryant et al., 1999; Chung et al., 1999; Ogris et al., 1997; Tolstykh et al., 2000; Yu et al., 2001). However, a recent study has determined that only the B-family requires C-subunit methylation, as B’ and B” can associate with or without methylation (Longin et al., 2007). This report also shows that selective amino acid phosphorylation of the catalytic subunit C-terminus modulates B and B’ subunit association but has no impact on B” association. Only now are the multimeric intricacies of PP2A assembly beginning to be discovered and understood. With the aid of new crystal structures and improved

8

biochemical techniques, holoenzyme assembly, as well as substrate recognition, is gradually being defined. Neurotrophin signaling Neurotrophin signaling supports the differentiation and survival of central and peripheral nervous tissue. The family of neurotrophins includes nerve growth factor (NGF), brain-derived growth factor (BDNF), neurotrophin-3 (NT-3) and neurotrophin4/5 (NT-4/5) (Huang and Reichardt, 2003). While all neurotrophins can bind with low affinity to the p75 receptor, each neurotrophin displays high affinity for specific members of the Trk family of receptor tyrosine kinases (Huang and Reichardt, 2003). NGF, the first-described neurotrophin, preferentially binds to the receptor tyrosine kinase TrkA, BDNF and NT-4/5 to TrkB and NT-3 to TrkC (Segal, 2003). Receptor tyrosine kinase family Receptor tyrosine kinase (RTK) activation is essential for the growth, differentiation, and survival of every cell (Huang and Reichardt, 2001). Members of the RTK family share common mechanisms of activation wherein ligand binding induces receptor dimerization and inter-receptor autophosphorylation of Tyr residues, which serve as docking sites for signaling molecules (Schlessinger, 2000). Principle signaling pathways emanating from RTKs include mitogen-activated protein kinase (MAPK), phosphoinositide 3-kinase (PI3 kinase) and phospholipase C
(PLC
) pathways. Since uncontrolled RTK activation contributes to a variety of malignancies (Zwick et al., 2002), mechanisms of receptor desensitization are of extreme importance and are being vigorously investigated. Tyr phosphorylation is an established read-out of receptor activity, whereby either total receptor pTyr content or specific phosphorylated Tyr residues are measured. Total pTyr detects both phosphorylation of the active kinase domain as well as Tyr residues important for protein recruitment and potentiating RTK signaling pathways. Phosphospecific antibodies can distinguish which pathways are

9

being activated through pTyr recruitment and provide insight to the mechanism of activity. Reversible Ser/Thr phosphorylation has received relatively little attention but is most often linked to desensitization of RTKs including the epidermal growth factor receptor (EGFR) (Bao et al., 2000), c-MET (Hashigasako et al., 2004), c-KIT (Chan et al., 2003) and insulin receptors (Zick, 2003). Ser/Thr phosphorylation of the TrkA receptor is the focus of the research and discussed in detail in Chapter 3. TrkA receptor tyrosine kinase TrkA (tropomyosin-related kinase A), the high affinity NGF receptor, consists of an extracellular ligand binding domain, a transmembrane domain, an intracellular juxtamembrane (JM) domain followed by a tyrosine kinase domain and a C-terminal tail (Wiesmann and de Vos, 2001). NGF stimulation results in receptor dimerization and autophosphorylation of C-terminal tyrosine residues, including Tyr-490, a Shc and FRS2 (FGF receptor substrate 2) docking site critical for both MAP kinase and PI3 kinase activation (Segal, 2003) (Figure 1.3). The TrkA receptor is also heavily Ser phosphorylated upon NGF treatment (MacPhee and Barker, 1997), an event that is largely unresolved in terms of mechanism or consequence. NGF signaling through the TrkA receptor alone is sufficient to drive differentiation of certain classes of neurons as well as PC12 cells (Chao and Hempstead, 1995; Cowley et al., 1994; Qui and Green, 1992; Traverse et al., 1992). MAP kinase signaling pathway The MAPK pathway is activated by many different growth factors including NGF. Activation of the MAPK pathway can lead to various cellular activities based on the cell type and the receptor from which it originates. For example, the EGFR generates a transient MAPK signal that leads to cellular proliferation in most cells. TrkA generates a sustained MAPK signal that leads to growth arrest and neuronal differentiation of PC12 cells (Cowley et al., 1994; Qui and Green, 1992). The difference in the time course of

10

MAPK signaling arising from the two RTKs is due to differences in adaptor protein recruitment, feedback regulation, and trafficking (Huang and Reichardt, 2001; Kao et al., 2001). EGFR initiates the cascade through the recruitment of membrane-bound Ras by the adaptor/guanine exchange factor (GEF) complex Shc:Grb2:SOS, which primarily results in activation of the Raf-1 isoform. The cascade is terminated as the receptor loses proximity to its substrates as it is traffics through the endocytic pathway and the Grb2:SOS complex dissociates due to ERK-mediated feedback phosphorylation (Holt et al., 1996). TrkA also signals through the Ras/MAPK pathway during initial activation at the plasma membrane. However, once internalized, Ras signaling is terminated and through an unknown mechanism, the TrkA MAPK signaling complex shifts to one that sustains signaling. TrkA complexes with the Crk:C3G adaptor/GEF complex that activates the endosomal small-GTPase Rap1 that promotes sustained activation of the BRaf isoform and potentiates extended ERK signaling (Kao et al., 2001; York et al., 1998). How TrkA signaling is sustained is explored in Chapters 2 and 3. These data attribute mechanisms that involve Ser phosphorylation/dephosphorylation of the receptor itself to the control of sustained ERK signaling. Cell systems PC12 cells are derived from a rat pheochromocytoma (Greene and Tischler, 1976). In the presence of NGF, PC12 cells undergo differentiation into a neuronal phenotype: division is halted, and the cells elaborate neurites capable of generating action potentials (Greene, 1978; Toledo-Aral et al., 1995). This presents an elegant system for studying neuronal signal transduction and differentiation pathways. PC6-3 cells are a sub-line of PC12 cells that were selected to display increased sensitivity to NGF withdrawal once differentiated, a characteristic of sympathetic neurons (Pittman et al., 1993). We prefer working with PC6-3 cells, because they can be transfected with higher efficiency and display less tendency to aggregate than parental PC12 cells. Initial studies

11

using PC6-3 cells implicated PP2A as a positive regulator of sustained MAPK activation that leads to neuronal differentiation (Strack, 2002). PC6-3 cells with inducible knockdown of the A
scaffold subunit of PP2A (A
RNAi cells (Strack et al., 2004)) and PP2A regulatory subunit replacement cell lines (A
-exchange cell lines (Strack et al., 2004)) are used to determine PP2A activity as discussed in Chapter 2. Additionally, tissue extracts from mouse brain, spinal cord and dorsal root ganglion (DRG) are used to examine biochemistry in Chapter 2. Dissertation research focus NGF can have profound effects on differentiation, proliferation, survival, cell motility, membrane trafficking, pain, inflammation and synaptic plasticity, depending on the cellular context. Sustained signaling through TrkA is critical for the ability of NGF to differentiate PC12 cells into neurons and is required for the development of sympathetic and dorsal root ganglion, as well as cholinergic projections in the forebrain (Smeyne et al., 1994). Published TrkA biochemical data, structural analysis, and functional studies suggest Ser phosphorylation plays an important role in maintaining TrkA activity (Kahl and Campanelli, 2003; MacPhee and Barker, 1997; Till et al., 2002; Yoon et al., 1997). I present data herein addressing the hypothesis that sustained TrkA signaling is negatively regulated by ERK-mediated phosphorylation within the juxtamembrane domain of the receptor, and that a specific PP2A holoenzyme counteracts kinase activity to maintain TrkA activity. To address this hypothesis I have defined the following specific aims: 1.) Identify the kinase(s) and PP2A holoenzyme(s) that regulate TrkA Tyr phosphorylation through Ser phosphorylation in PC12 cells, 2.) Identify intracellular Ser residues that regulate TrkA receptor activity and 3.) Begin to determine the mechanism by which Ser phosphorylation influences TrkA kinase activity. The results of my research have identified novel regulators of NGF signaling and provided important insight into the underlying mechanisms that control RTK signaling pathways.

12

Figure 1.1 The composition and structure of the protein phosphatase 2A holoenzyme. The PP2A heterotrimeric crystal structure (Protein Data Bank accession code 2NPP) is manipulated using Swiss PDB viewer to depict the subunits of PP2A. A. The PP2A core dimer is represented by the A-scaffolding subunit (yellow) associating with the C-catalytic subunit (grey). The A-scaffolding subunit is horseshoe shaped and composed of
-helical HEAT-repeats. The C-catalytic subunit is globular with a protruding C-terminus that is thought to regulate holoenzyme composition. B. The B’ subunit crystal structure (blue) is composed of
-helical HEAT-repeats and associates at the N-terminal portion of the A/C dimer. The predicted structure of B is a toroid composed of seven WD repeats. The predicted structure of PR48/B” shows only the calcium binding EF hands (25% of protein). The regulatory subunits compete for a binding regions and form mutually exclusive heterotrimers with the A/C dimer.

13

Figure 1.2 PP2A regulatory subunit sequence homology. A. Phylogentic tree of human subunit families is generated by clustalW sequence alignment software. Protein sequences from major PP2A regulatory subunits were compared. No homology is seen between families, but within families, sequences are highly conserved. The length of horizontal line represents the number of evolutionary amino acid replacements. B. Sequence alignment using clustalW showing B’ family of regulatory subunits. Extremely high sequence identity is seen throughout the core domain (71-88%) with divergent N and C-termini. Highlights: Blue - identical, Yellow - conserved, White - non-conserved.

14

Figure 1.3 NGF signaling cascades. NGF activates the TrkA receptor by forming homodimers that auto/trans-phosphorylate Tyr residues within the intracellular domain. Tyr490 phosphorylation recruits Shc/Grb2/SOS to activate Ras. Ras can potentiate the MAPK signaling module by activating Raf/MEK/ERK. Ras can also activate PI3 Kinase pathway leading to Akt activation. PLC1 is recruited to Tyr780 to induce PKC activation. PP2A/B family members target the MAPK pathway. PP2A/B’ family members target Akt pathway. PP2A/B’
targets TrkA.

15

CHAPTER II SPECIFIC PP2A HETEROTRIMERS MEDIATE TRKA RECEPTOR SIGNALING AND NEURONAL DIFFERENTIATION. Abstract Nerve growth factor (NGF) signaling is critical for the differentiation and maintenance of certain neuronal populations. Sustained receptor signaling is a hallmark of TrkA function, yet our knowledge of the molecular components necessary for prolonged activity is incomplete. Protein phosphatase 2A (PP2A) is a heterotrimeric Ser/Thr phosphatase composed of a scaffolding, catalytic and regulatory subunit (B, B’, B” gene families). Specific PP2A heterotrimers are known to regulate the Ras-MAP kinase pathway activated by NGF. Here, we employ pharmacological inhibitors and RNA interference in PC12 cells to show that B’ family regulatory subunits foster sustained Tyr autophosphorylation of TrkA in response to NGF treatment. Specifically, B’
potentiates TrkA Tyr phosphorylation and downstream signaling that enhances NGFdependent neuritogenesis and differentiation in PC12 cells. Conversely, NGF signaling is reduced after B’
and B’ subunits are downregulated. PP2A is shown to complex with Trk receptors in PC12 cells as well as in neuronal tissue. Collectively, our data support a model in which PP2A/B’ holoenzymes associate with the TrkA receptor tyrosine kinase to sustain NGF-mediated developmental and survival signaling. Introduction Extracellular signaling through receptor tyrosine kinases is critical for the maintenance and survival of the developing nervous system (Huang and Reichardt, 2001). Neurotrophins target high and low affinity Trk receptors to activate signal transduction cascades that regulate neuronal function, synaptic plasticity and survival in many neuronal populations (Huang and Reichardt, 2003). Nerve growth factor (NGF) is

16

the prototypical neurotrophin that binds to and activates the TrkA receptor tyrosine kinase. TrkA activity is required for the development and preservation of centrally located cholinergic neurons in the brain, as well as sympathetic and sensory neurons in the peripheral nervous system (Chen et al., 1997; Fagan et al., 1997; Kuruvilla et al., 2004). NGF signaling Upon ligand binding, receptors dimerize and autophosphorylate tyrosine residues that serve as docking sites for adaptor proteins that facilitate the activation of the MAP kinase, PI3 kinase and PLC
1 signaling cascades (Huang and Reichardt, 2003). Receptor complexes are then internalized into signaling endosomes that remain active while undergoing retrograde transport down axons towards the perinuclear region where gene regulation ensues (Ginty and Segal, 2002). Prolonged exposure to NGF forces PC12 cells into a differentiated state through a process that is dependent on MAP kinase activity (Cowley et al., 1994; Greene and Tischler, 1976; Qui and Green, 1992). Likewise, persistent NGF stimulation is essential for neuronal survival that is directed through the PI3 kinase pathway (Ashcroft et al., 1999; Greene and Tischler, 1976; Klesse et al., 1999). Despite the well accepted belief that receptor phosphorylation is maintained to allow prolonged signaling of the MAP kinase and PI3 kinase pathways, the molecular machinery that influences these activities is largely undetermined. PP2A holoenzyme in signal transduction Reversible Ser/Thr phosphorylation is a widespread mechanism that allows a balanced control of many protein kinase signaling modules. Protein phosphatase 2A (PP2A) is a ubiquitous serine/threonine phosphatase that targets a large spectrum of substrates that govern the delicate balance of molecular signals emanating from numerous trophic receptors (Gallego and Virshup, 2005; Janssens and Goris, 2001). This multimeric enzyme is comprised of a conserved dimeric core consisting of a scaffolding

17

A subunit and a catalytic C subunit, which associates with divergent regulatory subunits. There are three families of regulatory subunits, B (PR55), B’(B56/PR61) and B” (PR93/PR110) which provide substrate specificity and subcellular localization of the functional holoenzyme (Janssens and Goris, 2001). Each regulatory subunit family contains three to five genes that share 70-90% sequence identity (Hendrix et al., 1993; Mayer et al., 1991; McCright et al., 1996a). The B family of regulatory subunits can impose both positive and negative regulation of the MAP kinase pathway depending on targeted substrate (Junttila et al., 2007). For example, PP2A/B overexpression promotes neuronal differentiation of PC12 cells in the absence of NGF by enhancing Raf1 activity and potentiating the dynamic effects of MAP kinase signaling (Strack, 2002). Conversely, ERK activity is negatively regulated by B
and B holoenzymes through direct dephosphorylation (Van Kanegan et al., 2005). The recent crystal structure of a B’ family member depicts a helical phosphoprotein that is able to recognize basic residues in target proteins (Xu et al., 2006, Cho, 2007 #212). Many biochemical studies have implicated B’ subunits in regulating cell cycle, Wnt/-catenin signaling, survival and synaptic plasticity (Bennin et al., 2002; Fukunaga et al., 2000; Okamoto et al., 2002; Ratcliffe et al., 2000; Rocher et al., 2007; Ruvolo et al., 2002; Seeling et al., 1999). The B” family subunits contain 2 EF hands that require Ca++ for PP2A activity (Ahn et al., 2007a; Janssens et al., 2003). PP2A plays a multifaceted role in regulating many signal transduction pathways arising from numerous membrane receptors (Janssens and Goris, 2001; Junttila et al., 2007). The current study affixes yet another level of PP2A regulation to signaling pathways emanating from the TrkA receptor. Herein, we show PP2A activity is required to sustain TrkA phosphorylation. PP2A complexes with Trk receptors and imposes a positive regulation that enhances downstream kinase activity required for differentiation of PC12 cells.

18

Materials and methods Cell culture and plasmids PC12 (6-24) TrkA overexpression cells (a gift from Phil Barker, (Hempstead et al., 1992)) and parental PC12 (PC6-3) cells (Pittman et al., 1993) were cultured (37°C, 5% CO2) in RPMI 1640 (Gibco) containing 10% horse and 5% fetal bovine serum (both heat-inactivated). A
-RNAi cells (Strack et al., 2004) were cultured in medium supplemented with 2
g/ml blasticidin and 200
g/ml G418 to maintain tetracycline repressor and inducible short hairpin (sh)RNA constructs, respectively. A-exchange cell medium included the addition of 200
g/ml hygromycin for maintenance of the inducible A
EE100RR, A
DTP177AAA and DWF139AAA constructs (Strack et al., 2004). B’ inducible cells (Saraf et al., 2007) were cultured in medium supplemented with 2
g/ml blasticidin and 200
g/ml hygromycin to maintain the inducible overexpression construct. Antibiotics were omitted from cultures seeded for experiments. Plasmids encoding HA-Ras V12 and LCK-mCherry were donated by Phillip Stork (Vollum Institute) and Steven Green (University of Iowa), respectively. pFC-MEK1, a plasmid encoding constitutively active MEK1 (S218, 222D), was supplied as part of the PathDetect Elk1 trans-reporting system (Stratagene, La Jolla, CA). HA-tagged B’ subunits are described elsewhere. PP2A/B’ subunit pSuper shRNA constructs are described elsewhere. Antibodies Polyclonal antibodies against Bpan, B' and B’ were a gift from David Virshup (University of Utah). Transferrin receptor antibody was a gift from David Sheff (University of Iowa). Antibodies commercially available are as follows: pan-Trk C-14 and agarose conjugate, Trk B-3, total ERK, rabbit IgG-agarose, Protein A agarose (Santa Cruz Biotechnology, Santa Cruz, CA), HA epitope and HA-conjugated beads (Sigma, St. Louis, MO); PP2A catalytic subunit (Pharmingen); phospho-TrkA Y490, phospho-Ser-

19

473 Akt, total Akt, phospho-ERK1/2 (Cell Signaling, Beverly, MA); phospho-Tyr (4G10) (Upstate Biotechnology, Lake Placid, NY). Secondary antibodies include goat
Rabbit HRP and goat
Mouse HRP (Santa Cruz Biotechnology, Santa Cruz, CA); goat
Rabbit IR800 and goat
Mouse IR680 (Licor, NE). Other reagents Rat nerve growth factor (NGF, 2.5 S) and recombinant human fibroblast growth factor-2 (FGF2) were purchased from Upstate Biotechnology, Sigma, and Alomone Labs (Jerusalem, Israel), respectively. Growth factors were stored at –20 °C in lyophilized aliquots and dissolved to 100X in 1/10th culture medium prior to use. Okadaic acid and microcystin-LR were purchased from Alexis (Lausanne, Switzerland). Microcystin-LR agarose beads were purchased from Upstate Biotechnology. Bradford reagent was purchased from BioRad (Hercules, CA) Immunoprecipitations PC6-3, A
-RNAi or A
-exchange cells were seeded in 100 mm dishes and treated ±doxycycline (Dox) for 3 days as indicated. HA-tagged B’ subunits were transiently transfected into PC6-3 cells for 2 days using Lipofectamine 2000. Cells were serum-starved for at least 2 h and pretreated in some cases with ±300 nM OA prior to adding growth factors at staggered times. After washing with phosphate-buffered saline, cells were harvested in RIPA lysis buffer (1% TX-100, 0.5% sodium dodecyl sulfate (SDS), 0.5% deoxycholate (DOC), 150 mM Tris pH 8.0, 300 mM NaCl, 1 mM EDTA, 1 mM EGTA supplemented with protease/phosphatase inhibitor cocktail – 1 mM benzamidine, 10
g/ml (20
M) leupeptin, 1mM pepstatin and 250
M PMSF, 1 mM glycerolphosphate, 2.5 mM sodium pyrophosphate and 0.5
M Microcystin-LR). Soluble protein was quantified using Bradford reagent and equal lysates were immunoprecipitated using 1.5
g of TrkA (C-14) antibody and protein-A agarose, TrkAagarose conjugate or HA-agarose beads for 4-6 hrs. Beads were washed 4 times with

20

0.5% TX-100 TBS, extracted in SDS sample buffer, subjected to SDS-PAGE and transferred to PVDF membrane (0.45μM Millipore) for western blotting. Images were gathered using either a Kodak Imager 440 or Odyssey Imaging System (Licor, NE). Animal tissue preparation DRGs, spinal cord and brain were harvested immediately following cervical dislocation of male Taconic C57black/6 mice. The DRGs and spinal cord were pooled separately from two mice. Tissues were homogenized in ice-cold buffer including 320 mM sucrose, 10 mM Tris-HCl, pH 7.4, 10 mM EGTA, and 10 mM EDTA supplemented with protease inhibitors (1 mM benzamidine, 10
g/ml (20
M) leupeptin, 20
g/ml aprotinin, 1mM pepstatin and 250
M PMSF and phosphatase inhibitors – 50 mM NaF, 1 mM
-glycerolphosphate, 2.5 mM sodium pyrophosphate and 1
M Microcystin-LR). Supernatants were recovered from a low-speed spin (2 min, 3500 x g in a microfuge) to remove debris and then subjected to ultracentrifugation (30 min, 250,000 x g in a Beckman Ti-50 rotor) to pellet membranes. Membranes were solubilized and homogenized in 1% Triton X-100, 10 mM Tris-HCl, pH 7.4, 10 mM EDTA, and 10 mM EGTA supplemented with protease and phosphatase inhibitors in the same concentration used above. Non-solubilized material was pelleted for 15 min, 15,000 x g in a microfuge. TrkA was IPed from normalized lysates as described above. Control immunoprecipitations were performed with 5
g of purified rabbit IgG that originated from naive, i.e., non-immunized, animals. Total extract input (5% of volume used for immunoprecipitation) was loaded in parallel. Microcystin pulldown Mouse brain lysates were prepared as described above in the absence of microcystin-LR. Normalized lysates were divided into three. As control, one sample is incubated for 30 min with 1
M free microcystin-LR to block microcystin-LR-agarose binding. Samples are incubated with 15
ls of microcystin-LR-agarose or GST-agarose

21

as control for 2 h. Beads are washed three times in 0.5% TX-100 TBS and extracted in SDS sample buffer, subjected to SDS-PAGE and transferred to PVDF membrane (0.45μM Millipore) for western blotting. Images were gathered using an Odyssey Imaging System (Licor, NE). Total extract input (5% of volume used for immunoprecipitation) was loaded in parallel. Quantitative immunoblotting with phospho-specific antibodies Enhanced chemiluminescence (SuperSignal, Pierce) images were captured using a Kodak Imager 440, and band intensities were quantified with the ImageJ software gel analyzer plug-in. Phospho-specific antibody signals were divided by total protein antibody signals to control for loading differences. All signal intensities were normalized to maximum intensity per experiment. Statistical analysis was performed on data from at least four independent experiments using Prism 4.0 to run two-way ANOVA with Bonferroni post-tests for individual time point comparisons. MAPK reporter assays The PathDetect Elk1 trans-reporting system was modified for the dual-luciferase assay (Promega, Madison, WI) to quantify ERK activation according to the manufacturers' instructions. PC6-3 cells overexpressing the B’
subunit were plated at 100,000–150,000 cells/well in 24-well plates and were transfected in triplicate using Lipofectamine 2000 (BD Biosciences) with 0.5
g/well reporter plasmid mix (by mass: 92.5% pFR-Luc, 5% pFA2-Elk1, 2.5% pRL-SV40) and 0.5–2 ng/well activator plasmid (Ha-Ras V12, pFC-MEK1 S218, 222D) or pcDNA3.1 empty vector. Media was exchanged 4-6 hours later with Doxycycline (1
g/ml) or 0.1% ethanol vehicle. PC6-3 cells subjected to transient RNAi were transfected with 0.25
g/well pSUPER-based plasmids expressing shRNA’s directed against B’ subunits and 0.25
g/well reporter plasmid mix. After 2 days ±Dox treatment or 3 days shRNA expression, cells were

22

preincubated for 2 h in medium with one-tenth original serum concentration followed by 5–6 h of stimulation with growth factors as indicated. Cultures were lysed in passive lysis buffer according to protocol (PathDetect) and subjected to dual-luciferase assays using a Berthold Sirius tube luminometer. Photinus and Renilla luciferase activity ratios were expressed relative to basal conditions without ERK activator plasmids or growth factor stimulation. Neurite outgrowth assay B’
inducible cells were plated on collagen coated 12 well plates and transfected with 1
g/well LCK-mCherry to visualize neurites using Lipofectomine2000 and blindly dosed ±Dox for 2 days to induce subunit expression then stimulated with 2 ng/ml NGF for 18hrs. PC6-3 cells were blindly transiently transfected with B’ subunit shRNA and LCK-mCherry for 3d and stimulated with 2 ng/ml NGF for 26 hours. Cells were washed with PBS before and after 5 min fixation with 4% PFA. 50-100 cell images per condition were taken at 20X using an inverted epifluorescent microscope equipped with a digital camera. Total neurite length from each cell was measured using ImageJ (NIH) and the data was converted to microns. Statistical analysis was performed on data from three independent experiments using Prism 4.0 to run unpaired t-test. DNA content FACS analysis B’
inducible cells were plated on collagen coated 60 mm well plates at 150,000/plate and dosed ±Dox for 2 days to induce subunit expression then treated with ±10 ng/ml NGF for 2 and 4 days to stimulate differentiation. Cells were removed from plates with trypsin and quenched with culture media. Cells were pelleted at 1000 x g and lysed in PI Lysis buffer (0.1% TX-100, 0.1% sodium citrate, 0.05 mg/ml propidium iodide and 2
g/ml Ribonuclease A (Roche)) for 30 min at RT and placed on ice. DNA content was analyzed by flow cytometry with a Becton Dickinson FACScan. For each sample 1.5 x 104 events were recorded. PI histograms were generated and analyzed by

23

FlowJo software (Tree Star) to determine percentages of cells in each phase (G0/G1, S, G2/M). The percentages of cells in a given phase from four independent experiments were compared using an unpaired t-test using Prism 4.0. Surface biotin labeling PC6-3 cells were grown to 80% confluency on 100 mm dishes, serum starved for 2 hrs and pretreated ±300 nM OA for 1 hr prior to 20 ng/ml NGF stimulation for 0-60 min. Cells were chilled on ice, washed two times with cold PBS and incubated with 500nM EZ-link Sulfo-NHS-LC-LC-Biotin (Pierce, Rockford, IL) for 30 minutes on ice. Cells were harvested in RIPA buffer and normalized lysates were IPed with NeutrAvidin beads (Pierce, Rockford, IL) for 2 hours. Beads were washed three time in 0.5% Triton X-100 TBS, extracted in SDS sample buffer and subjected to SDS PAGE and western blotting for total TrkA and Transferrin receptor as a labeling and loading control. Total surface was determined using quantitative western blotting analysis comparing TrkA/TfnR. Results PP2A is required to sustain TrkA activity It is well established that prolonged NGF treatment of PC12 cells directs molecular machinery to halt proliferation and force the cells into a differentiated state (Greene, 1978; Huang and Reichardt, 2003). To address whether the Ser/Thr protein phosphatase PP2A is involved in sustaining TrkA activity, we monitored TrkA receptor activity after treatment with saturating levels of NGF using PC12 (6-24) cells that overexpress TrkA 15-20 fold over native PC12 cells (Hempstead et al., 1992). TrkA receptor autophosphorylation activates several key tyrosine residues involved in receptor activity. Tyrosine 490 phosphorylation facilitates the recruitment of the adaptor protein Shc and subsequently activates the Ras/MAPK cascade, which is required for PC12

24

differentiation (Stephens et al., 1994; Yoon et al., 1997). Figure 2.1A shows that NGF activated the TrkA receptor and lead to Tyr490 phosphorylation that was sustained for more than 60 minutes. Treating PC6-3 cells with 250 nM okadaic acid (OA), a selective Ser/Thr phosphatase inhibitor that targets PP2A as well as PP4, PP5, PP6 at nanomolar concentrations and PP1 at higher concentrations (>1 μM), diminished TrkA Tyr phosphorylation by ~50% following 15 and 60 minutes of NGF treatment (Figure 2.1A, C). Not only was Tyr490 reduced after OA treatment, global pY detection was reduced as shown in Figure 2.1B, C. These results are supported by a similar TrkA activity profile obtained in native PC6-3 cells (Figure 2.1E). Since okadaic acid inhibits PP2A like phosphatases (PP2A, 4, 5, 6) with similar potency (OA inhibition rank order is: PP2A=4=5>6>>>1), definitive receptor regulation by PP2A cannot be concluded. Therefore, we inducibly downregulated the A
scaffolding subunit of PP2A to deplete holoenzyme activity prior to NGF treatment and monitored TrkA Tyr phosphorylation. Treating A
RNAi cells (Strack, 2002) ±Dox for three days and then with NGF for the indicated times showed similar reductions of TrkA activity as compared to OA treatments. These data implicate PP2A as the phosphatase responsible for sustaining TrkA activity (Figure 2.1D, E). Furthermore, comparable pY490 and pY western blot profiles after both pharmacological inhibition and selective RNA interference verified a functional role of PP2A in potentiating TrkA activity. These data implicate a role for PP2A regulation during sustained NGF signaling. To address if both transient and sustained TrkA signaling is regulated by PP2A, comprehensive NGF time course was analyzed ±PP2A activity using A
RNAi cells (Figure 2.2 A). Interestingly, TrkA Tyr phosphorylation was not decreased after 5 minutes of NGF, showing a slight, yet not significant increase in TrkA phosphorylation in the absence of PP2A (Figure 2.2 A, B). This is in contrast to the additional time points analyzed through 180 minutes, where PP2A inactivity caused a decrease in TrkA Tyr phosphorylation. These data suggest that PP2A does not impact the transient phase of

25

TrkA activity but instead modulates sustained signaling. The activity of two downstream signaling proteins, Akt and ERK, mirrors the decrease in TrkA activity following PP2A inhibition, as measured by western blotting (Figure 2.2 C-F). Repeated experiments showed that Akt and ERK activity were not affected at 5 minutes, while each kinase activity diminished throughout the extended NGF time course. While these studies suggest that PP2A regulates sustained TrkA activity on a signaling endosome, PP2A may be impacting TrkA expression levels or affecting surface delivery of the functional receptor. Next, cell surface biotinylation assays analyzed cell surface expression ±PP2A activity using A
RNAi cells (Figure 2.3 A). These results showed that after NGF stimulation, the absence of PP2A had no significant effect on the number of TrkA receptors that were available at the surface for ligand stimulation. However PP2A inhibition did cause a trend towards increased basal level surface receptors. This may account for the slight increase in TrkA activity at 5 minutes of NGF stimulation. Nevertheless, these data further support the idea that PP2A contributes to the sustained TrkA activity following NGF treatment. PP2A/B’ family of regulatory subunits modulate TrkA autophosphorylation PP2A substrate specificity and subcellular localization is dictated by core dimer association with regulatory subunits from three families of genes (B, B’, B”). With several regulatory subunits being expressed simultaneously in a given cell, determining specific holoenzyme regulation is difficult to delineate. To identify the family of holoenzymes that positively regulate TrkA activity, we utilized A
-exchange cell lines that inducibly knockdown the endogenous A
scaffolding subunit and concomitantly express a mutant that selectively binds specific families of regulatory subunits (Figure 2.4A) (Ruediger et al., 2001; Strack et al., 2004). The absence of an A
subunit

26

association caused a rapid degradation of the catalytic subunit, as well as the B and B’ regulatory subunits. Conversely, the B” subunits were not destabilized (Strack et al., 2004). When the A
mutant (EE100RR) that is unable to associate with any regulatory subunits was exchanged, a ~50% decrease in TrkA activity was detected (Figure 2.4B, D) similar to results following OA treatment and scaffolding subunit knockdown (Figure 1). As expected, B and B’ family regulatory subunits were rapidly degraded without dimer interaction. Replacement with the A
mutant did not cause catalytic subunit degradation indicating that the A
mutant formed a functional dimer that was unable to recruit regulatory subunits (Figure 2.4B). An A
-exchange (DWF139AAA) that only allowed interaction with B family regulatory subunits (no B’ or B” interaction) significantly downregulated TrkA receptor activity is quantified in Figure 2.4E. Using the DTP177AAA mutant that selectively prevents B’ family binding similarly showed an attenuation of pTyr signal (Figure 2.4C, F). Unfortunately, the mutant that only binds B’ family members in vitro (DW139RR) (Ruediger et al., 2001), had unreliable heterotrimer formation as seen by a inconsistent down regulation of both B and B’ regulatory subunits in repeated experiments. Therefore reliable conclusions could not be formulated using these cell lines. Therefore, inducible B
knockdown cells were used as a positive control and had no effect on TrkA Tyr phosphorylation (data not shown). Taken together, these data deductively advocate B’ members as regulators of TrkA activity. PP2A/B’ positively regulates TrkA receptor activity PP2A/B’ is a neuron-enriched subunit that is found in most neuronal populations in the adult rat brain (Fukunaga et al., 2000; McCright et al., 1996b; Saraf et al., 2007). B’ mRNA levels also show a high expression during embryonic and neonatal rat brain development (van Lookeren Campagne et al., 1999), a time when Trk receptors are also highly expressed. Furthermore B’ mRNA is increased 4-fold in neuroblastoma cells when differentiation is induced using retinoic acid (McCright et al., 1996b). Therefore,

27

we hypothesized that the  isoform of the B’ family is responsible for regulating TrkA activity. Inefficient transient transfection of the B’ regulatory subunit created heterogeneous populations of cells in which the effects of B’ on TrkA receptor phosphorylation are masked. To overcome this problem, we utilized a Dox inducible B’ (B’
) cell line to investigate the PP2A subunit in TrkA function. After 2 days ±Dox treatment, we observed a significant potentiation of TrkA Tyr phosphorylation after 15 and 60 minutes of NGF treatment (Figure 2.5A, B). The elevated receptor activity translated into an amplification of downstream signaling pathways that drive survival and differentiation: the Akt and MAPK pathways, respectively. Extending NGF treatment showed enhanced Akt and MAPK phosphorylation that was sustained throughout the time course (Figure 2.5C-F). To pinpoint where in the NGF signaling cascade B’ is regulating activity, an Elk1 reporter assay was employed in epistasis experiments (Figure 2.5G). NGF stimulation of the TrkA receptor initiates activity of the MAP Kinase cascade resulting in the activation of Ras/Rap1, MEK and ERK successively. Active ERK can target cytosolic substrates as well as translocate to the nucleus to phosphorylate transcription factors such as Elk1 (Gille et al., 1995; Gille et al., 1992). The Elk1 reporter assay detects nuclear ERK activity by measuring ERK dependent, Elk1-mediated transactivation of a luciferase reporter construct. ERK can be activated by growth factor stimulation, expression of constitutively active Ras V12 or constitutively active CAMEK1 enzymes that are titrated in PC6-3 cells to achieve similar levels of Elk1 activity as NGF treated cells. After transfection and 2 days ±Dox treatment to overexpress B’ subunits, the cells were treated with NGF and Elk1 luciferase activity was measured. Repeated experiments showed a ~2-3 fold increase in reporter activity in B’ expressing cells with NGF treatment compared to NGF treated control cells. Neither Ras nor MEK activation of the pathway triggered an elevation in ERK activity following B’ overexpression. In fact, stimulating the ERK pathway with Ras and MEK showed a decrease in Elk1 activity, suggesting B’ may negatively regulate the MAPK pathway

28

downstream of TrkA. B’ demonstrated high receptor selectivity, because B’ expression did not enhance FGF stimulation of the ERK pathway. In combination with the data showing B’ expression enhances TrkA autophosphorylation (Figure 2.5), these data indicate that PP2A/B’ regulates TrkA receptor activity at the receptor level (Figure 2.5H). Multiple B’ family members associate with the TrkA receptor and regulate its activity The family of B’ regulatory subunits consists of 5 members (, , , , ) that share 80% identity within a conserved core region (McCright et al., 1996b). These phospho-proteins differ in the extreme N and C-terminus, which are thought to direct subcellular localization and substrate specificity. To determine if the PP2A/ B’ holoenzyme can associate with the TrkA receptor we utilized the PC6-3 cell line that inducibly overexpresses the B’ subunit (B’
). After 2 d ±Dox and stimulation with NGF for 60 minutes, the TrkA receptor was immunoprecipitated. Results in Figure 2.6A showed a dramatically enriched B’ association following Dox treatment. Recruitment of the B’ subunit to TrkA increased the PP2A catalytic subunit association to the receptor complex by 60%, suggesting enhanced PP2A activity is localized to the receptor. To investigate selectivity among B’ family members, HA-tagged B’ subunits were transiently expressed in PC6-3 cells. After NGF stimulation, TrkA receptor immunoprecipitations from normalized lysates were probed with HA antibody to detect associated PP2A subunits. Figure 2.6B shows TrkA receptor complexes that strongly associated with the B’ and B’ subunits and a weak interaction with the B’ subunit as determined in repeated experiments. Overexpressed B’ and B’ showed minimal association with the TrkA receptor when compared to a HA-tagged negative control protein (CTRL). PP2A/B’ is a major subunit found in PC6-3 cells and Figure 2.6C demonstrates that this subunit does indeed associate with the TrkA receptor in a native

29

cells. PP2B/calcineurin does not associate with TrkA as shown in Figure 2.6C. Utilizing mouse dorsal root ganglions (DRGs) (data not shown) and spinal cord, PP2A was shown to associate with TrkA receptors as seen by robust catalytic subunit detection after Trk immunoprecipitation (Figure 2.6C). Additional mouse spinal cord coIPs exhibited B’ association (Figure 2.6C). Brain extracts showed similar PP2A association with Trk receptors and demonstrated B’ subunit association with TrkA receptors (Figure 2.6C). The reverse association was demonstrated by a microcystin-LR pulldown from mouse brain extract that brought down the Trk receptor with PP2A (Figure 2.6D). These precipitation studies clearly exhibit a Trk/PP2A complex in PC12 cells, as well as in mouse neuronal tissue, that implicate Trk receptor regulation being governed by a specific B’ and B’ PP2A holoenzyme. To address the functional significance of specific holoenzyme assembly with Trk receptors, we utilized an Elk1 luciferase assay to measure downstream receptor activity after B’ subunit RNAi. With high degrees of sequence identity, it is plausible that each holoenzyme may demonstrate functional overlap and regulate the TrkA receptor. Therefore, differences in Elk1 activity after RNAi may be explained by differences in the effectiveness of B’ subunit knockdown. Figure 2.7A shows the efficacy of subunit depletion when HA tagged subunits were expressed in the presence of directed shRNAs. These results show shRNAs directed against B’
and B’ have the most effective knockdown of overexpressed subunits, while there is no impact on B-family subunit stability. Endogenous PP2A/B’ subunits downregulation with shRNA showed subunit specific regulation of TrkA activity in the Elk1 luciferase assay. Compared to control hairpin-expressing cells, the loss of B’ and B’ showed a 40 and 50% decrease in Elk1 luciferase activity in three independent experiments (Figure 2.7B). These results are similar in magnitude to receptor phospho-signals measured in the PP2A inhibition studies conducted in Figure 1 and Figure 2. Similar, yet more variable in repeated experiments, the loss of B’ subunits lead to a 30% decrease in Elk1 activity, while knockdown of the

30

B’
and B’ showed less than 17% decrease in reporter activation. These data identify that B’ and B’ are required for full NGF responsiveness in PC12 cells. B’ and B’ holoenzymes potentiate NGF dependent neurite outgrowth Given that neuronal tissue is highly enriched in B’ and B’ expression (Fukunaga et al., 2000; McCright et al., 1996b; Saraf et al., 2007) and both are induced upon differentiation in neuroblastoma cells (McCright et al., 1996b), we asked if specific B’ isoforms are involved in regulating NGF-induced differentiation of PC12 cells. Activation of the MAPK signaling cascade is shown to be both necessary and sufficient for NGF induced neuronal differentiation in PC12 cells (Cowley et al., 1994; Qui and Green, 1992). Based on the TrkA autophosphorylation and the Elk1 luciferase assays studies (Figure 2.5A and 2.5G respectively), B’ overexpression should rapidly increase neuritogenesis in PC12 cells. The B’ overexpression cells were transiently transfected with a membrane targeted Lck-mCherry construct permit accurate quantification of neurite length. After 3 days of transfection and ±Dox treatment to induce B’ expression, the cells were treated with 2 ng/ml NGF for 18 hours to initiate differentiation. The short NGF treatment was used to capture augmented neuritogenesis that may be masked after prolonged NGF treatment and the low dose was optimized using the Elk1 luciferase assay. As seen in Figure 2.8A and B, total neuron length per cell was indeed increased by 25% with B’ expression. To verify that PP2A regulation of NGF induced differentiation was not an artifact of overexpression, we used specific shRNA to silence endogenous B’ subunits. PC6-3 cells were co-transfected with membrane targeted LCK-mCherry as well as shRNAs directed towards B’
, B’ and B’ and a scrambled hairpin as control (CTRL). Allowing 3d for knockdown, the cells were treated with 2 ng/ml NGF for 26 hours. Cells transfected with LCK-mCherry and empty vector had no impact on neurite outgrowth when compared to control hairpin (data not

31

shown). Knockdown of B’ and B’ resulted in a 40 and 55% decrease in total neurite length as seen in Figure 2.8C and D. There was no difference in process length after B’
knockdown when compared to CTRL. These data correspond nicely with the B’ shRNA MAPK reporter assay and again shows PP2A/B’ subunit specificity. Thus, PP2A/B’ and B’ holoenzymes demonstrate selective subunit targeting to the TrkA receptor and positively regulate NGF mediated neuritogenesis. B’ holoenzymes potentiate NGF dependent differentiation in PC12 cells. Experiments described earlier showed that PP2A/B’ is a key regulator of NGFinduced neurite outgrowth. These assays are a good measure of morphological changes induced by NGF, but do not indicate true differentiation, as defined as an exit from the cell cycle. In the presence of NGF, PC12 cells exit the cell cycle, cease proliferation and differentiate into a sympathetic neuron-like cell type with distinct physiological and morphological features (Chao and Hempstead, 1995; Greene and Tischler, 1976). To address whether PP2A/B’ can potentiate differentiation in PC12 cells, B’ was inducibly expressed and cells were treated with 10 ng/ml NGF for 2 and 4 days prior to FACS analysis to analyze DNA content. As shown in Figure 2.9A, B’ expression elevated the G0/G1 peak, indicative of an increase in the number of growth arrested cells. Likewise, proliferative cells in S-phase/G2 were decreased compared to control. B’ had a more profound effect on differentiation at 2 days than 4 days of NGF treatment (Figure 2.9B, C). These data suggest that B’ overexpression causes cells to bypass the initial mitogenic phase of NGF treatment and immediately arrest due to enhanced TrkA signaling. This is substantiated by seeing little difference between differentiated cells ±B’ at 7 days of NGF treatment (data not shown).

32

Discussion In PC12 cells, prolonged NGF exposure promotes neuronal differentiation through the MAPK signaling module (Cowley et al., 1994; Qui and Green, 1992; Vaudry et al., 2002). PP2A regulates many molecules within this pathway at different levels in both a positive and negative manner. The present study implicates, for the first time, PP2A as a master regulator of neurotrophin signal transduction by promoting TrkA receptor activity. Our data suggest that PP2A forms a signaling complex that sustains TrkA function and facilitates ERK and Akt activity required for neuronal differentiation. TrkA receptors are targeted by specific PP2A/B’ and B’ holoenzymes that positively regulate receptor function. PP2A regulation of NGF signaling How does PP2A regulate NGF-mediated signal transduction? Previous studies have shown that PP2A targets many enzymes in the Ras/MAPK pathway including Shc, Raf and ERK (Dougherty et al., 2005; Ugi et al., 2002; Van Kanegan et al., 2005). In this study, we showed that PP2A also regulated NGF signaling at the receptor level. Okadaic acid, a potent inhibitor of PP2A-like phosphatases, caused an attenuation of TrkA activity after prolonged NGF treatment (Figure 2.1). These data suggest that phosphatase activity is required to sustain receptor activity. Paradoxically, we previously observed an increase in basal ERK and Akt activity after OA treatment in PC6-3 cells (Van Kanegan et al., 2005), supporting a divergent regulation of each pathway that can be attributed to the global inhibition of different PP2A holoenzymes that specifically target ERK (PP2A/B
and B) and Akt. However, experiments using the A
knockdown cell line displayed the converse ERK and Akt activity profiles; showing a diminished signal after prolonged NGF stimulation (Van Kanegan et al., 2005). These experiments were repeated while analyzing the TrkA receptor (Figure 2.2 A, B) showing that TrkA activity is also decreased with extended NGF treatment. The PP2A/A
knockdown cells likely

33

induce compensatory regulation of signaling pathways, since EGF and FGF Akt and ERK activity is dampened (Van Kanegan et al., 2005). However, NGF-stimulated downstream kinase activity likely affords a truer reflection of TrkA activity and not compensatory downregulation of kinase activity in the absence of PP2A activity. The current studies suggest PP2A targets receptor activity after TrkA is activated and entered into a signaling endosome. Many reports have described TrkA signaling complexes that are spatially and temporally regulated by differently recruited signaling molecules (Grimes et al., 1996; Huang and Reichardt, 2003; Schlessinger, 2000). Since there was no effect of PP2A downregulation on surface receptor levels or initial activity after NGF treatment, these data suggest that PP2A activity is enhanced after TrkA activation (Figure 2.2) to sustain NGF signaling. PP2A is a complex enzyme that utilizes any of three multi-membered families of regulatory subunits to direct the scaffolding and catalytic dimer to specific substrates (Janssens and Goris, 2001). To eliminate families of regulatory subunits that do not modulate TrkA receptor activity, we used a previously described protein exchange system (Strack et al., 2004). Replacement of the A
scaffold with mutants that selectively prohibit binding of regulatory subunit families allowed us to deduce that the B’ family was responsible for regulating TrkA activity. Since each exchange cell line used shows a decrease in TrkA Tyr phosphorylation and the DW139RR cells are show erratic results, a better positive control is necessary. Future experiments will utilize cells that exchange wild-type A
after endogenous A
knockdown to ensure TrkA Tyr phosphorylation is not affected by the exchange itself. the attenuation of TrkA activity in the absence of B’ regulatory subunits in each exchange cell line, while TrkA phosphorylation was unaltered in cell lines that inducible downregulate B family members only (data not shown).

34

PP2A mediates long term NGF signaling PP2A phosphatase activity appears to be a tightly regulated cellular event as PP2A inhibition be relevant to the progression of Alzheimer’s disease by increasing tau and amyloid precursor protein phosphorylation which can enhance neurofibrillary tangles and amyloid plaque formation respectively in animal models (Sontag et al., 2007). Conversely, elevated PP2A activity can suppress Akt survival activity in dopaminergic neurons (Chung et al., 2007), perhaps contributing to the progression of Parkinson’s disease. Likewise, our data imply PP2A regulates sustained TrkA activity, which can be altered by phosphatase manipulation. Overexpressing the B’
regulatory subunit enhanced NGF signaling starting at the receptor and preceding though the Akt and MAPK pathways (Figure 2.3). In contrast, downregulating B’
and B’ had profound negative effects on neuritogenesis (Figure 2.8). Sustained activation of the ERK pathway with phorbal esters induces the expression of the early response gene IEX-1, which has been shown to positively regulate ERK activity. IEX-1 induction prevents PP2A/B’
holoenzymes from negatively regulating ERK activity through direct dephosphorylation (Letourneux et al., 2006). After stimulating the ERK pathway with FGF, oncogenic Ras or CA-MEK, we observed a decrease in MAPK reporter activity in the presence of B’
overexpression. These data support a direct downregulation of ERK in the presence of elevated PP2A/B’
holoenzymes. However, stimulation with NGF showed a strong potentiation of ERK activity with B’
overexpression that completely dominated any negative regulation imposed by B’
downstream of the TrkA receptor. The fact that B’
overexpression had no positive effect after FGF treatment emphasizes the selectivity this PP2A holoenzyme had for the TrkA receptor (Figure 2.6). Given that PP2A acts upstream of Ras, it is possible that PP2A targets the receptor directly, and this hypothesis is tested in the next chapter. PP2A has been shown to positively regulate another receptor tyrosine kinase, cMET, through reversible Ser

35

phosphorylation (Hashigasako et al., 2004) supporting this possibility. Although, there are numerous signaling proteins upstream of Ras shown to regulate receptor activity and trafficking that may be targeted by PP2A, data herein provide evidence against trafficking effects (Figure 2.3) and strongly argue for direct receptor targeting rather than associated protein targets. B’ regulatory subunits target specific substrates In pursuit of specific B’ family isoforms that regulate TrkA receptor activity, we asked if multiple isoforms could associate with the TrkA receptor. B’ family members have a very high degree of sequence identity within a central core domain and it is possible that several, if not all these subunits can impart similar regulation. Supporting this notion, all B’ subunits can regulate the Wnt/
-catenin pathway (Seeling et al., 1999), and every B’ subunit associates with the shugoshin protein (Kitajima et al., 2006). On the other hand, several reports demonstrate B’ subunit specificity. B’ has been shown to preferential association with Cdc25 (Margolis et al., 2006), while B’
and B’ confer specificity towards the IEX-1 early gene product (Letourneux et al., 2006). Our data demonstrated that only overexpressed B’
and B’ robustly associated with the TrkA receptor (Figure 2.6). We observed a basal level of association, but upon NGF treatment we saw a two-fold increase in complex formation, suggesting that PP2A is stabilized to facilitate prolonged activity (data not shown). In native PC6-3 cells and nervous tissue TrkA IPs, only B’ could be reliably detected to complex with TrkA. However, other subunit association with Trk receptors cannot be excluded, since antibody detection of native B’ subunits is a limitation in these tissues. Evidence that does distinguish PP2A subunit composition is accomplished through functional MAPK reporter assays. Downregulation of specific subunits in the B’ family exhibited isoform specificity that regulated TrkA receptor function (Figure 2.7). Supporting the TrkA/PP2A complex formation data, B’
and B’ depletion created the most profound decrease in ERK

36

activity. Again, these data implicate specific PP2A/B’ subunits as positive regulators of TrkA activity. PP2A regulates cellular fate by sustaining NGF signaling Sustained neurotrophin signaling is a hallmark of Trk receptor function. The signaling endosome hypothesis states that Trk receptors remain active as they are endocytosed and delivered to cell body to elicit cellular responses (Campenot and MacInnis, 2004; Grimes et al., 1996). Data within this report suggests that PP2A not only maintains Trk receptor activity that is necessary to execute gene regulation, it can potentiate this operation when the phosphatase is upregulated. This is illustrated given that a significant potentiation of early neuritogenesis and differentiation was seen with B’ overexpression (Figure 2.8, 2.9). Analyzing the effects of major neuronal B’ isoforms through RNAi showed a striking specificity in holoenzyme function. Since  and  isoforms are highly enriched in neuronal tissue, downregulation of these subunits effectively blunted neurite outgrowth, while B’
downregulation did not. These data were consistent with the ELK1 reporter assays and punctuate the importance of PP2A regulation of TrkA receptor function in a cellular context. In conclusion, we have shown that PP2A is a positive regulator of TrkA receptor activity. I have shown that PP2A/B’ and PP2A/B’ holoenzymes specifically associate with Trk receptors. In the next chapter, I will test a model where PP2A contests Ser/Thr phosphorylation imposed by negative feedback kinases to sustain Trk activity required for neuronal differentiation and survival.

37

Figure 2.1 Sustained TrkA signaling requires PP2A activity. (A) PC12 (6-24) cells overexpressing TrkA (TrkA) were treated ± 250nM OA for 90 min and stimulated with 20ng/ml NGF for the indicated times. Normalized lysates were probed with phosphospecific antibody directed towards Y490 and pan-Trk antibodies. Arrows indicate mature TrkA receptor. (B) TrkA immunoprecipitates from normalized PC12 (6-24) lysates were probed with phospho-tyrosine (4G10) and pan-TrkA antibodies. (C) Bar graphs represent densitometric analysis of TrkA activity as measured by phospho/total signal of four independent experiments ± SEM (*p value< 0.05, **p-value < 0.001). (D) Native PC6-3 cells were treated for the indicated times. TrkA immunoprecipitates from normalized lysates were probed with phospho-tyrosine (4G10) and pan-TrkA antibodies. (E) Bar graphs represent densitometric analysis of TrkA activity of native PC6-3 cells treated ± 300nM Okadaic acid (OA) for 90 min or PP2A-A
cells treated 3d±1mM Dox and stimulated with 20ng/ml NGF dox. Values illustrate phospho/total signal of four independent experiments ± SEM (*p value< 0.05, **p-value < 0.001).

38

Figure 2.2 PP2A does not modulate TrkA receptor activity or downstream signaling after acute NGF stimulation. (A) PP2A/A
cells are treated 3d±Dox and stimulated with 20ng/ml NGF for indicated times. Lysates solubilized with 1% TX-100 were probed with pY490 and pan TrkA, (C) pAkt and Akt, (E) pERK and ERK antibodies. (B, D, F). Line graphs show densitometric analysis of TrkA, Akt and ERK activity as measured by phospho/total signal of three independent experiments. PP2A only modulates prolonged NGF signaling.

39

Figure 2.3 PP2A does not affect TrkA receptor surface expression. (A) PC6-3 cells are treated with ±300 nM OA for 1hr and stimulated with 20 ng/ml NGF for indicated times. Cells were chilled and surface proteins were labeled with biotin. Cells were lysed and streptavidin IPs were probed for pan-TrkA and transferrin receptor (TfnR) as a labeling and loading control. (B) Graphs showed densitometric analysis of surface TrkA after NGF stimulation as measured by TrkA/TfnR signal of three independent experiments.

40

Figure 2.4 PP2A B’ family of regulatory subunits is required to sustain TrkA activity. (A) Schematic of three A
-exchange cell lines that inducibly knockdown the A
scaffolding subunit with directed shRNA and concomitantly express a mutant that selectively lacks regulatory subunit binding (B, B’, B” families of subunits). Each cell line was treated 3d ± Dox to selectively inhibit regulatory subunit binding. Cells were stimulated with 20ng/ml NGF for the indicated times and TrkA immunoprecipitates were probed with pY and pan TrkA antibodies (B) A
EE100RR (---) cells that lack all subunit association showed a decrease in TrkA activity after NGF stimulation. Lysates were probed for PP2A/B, B’ and B’ regulatory subunits to show subunit depletion. PP2A/C probe showed no depletion. (C) A
DTP177AA (+-+) cells that prohibit B’ family association identified the B’ family as facilitating sustained TrkA activity. Lysates were probed with B’ to show subunit depletion. PP2A/B probe showed no depletion. (D-F). A
EE100RR (---), DWF139AAA (+--), A
DTP177AA (+-+) graphical representation of TrkA activity. Bar graphs show densitometric analysis of TrkA activity as measured by pY/total signal of three independent experiments ± SEM (**p-value < 0.001).

41

Figure 2.5 B’
potentiates TrkA signaling upstream of Ras. (A) PC6-3 cells that inducibly express B’
were treated 3d ± Dox and stimulated with 20ng/ml NGF for indicated times. TrkA immunoprecipitates were probed for pY and pan-TrkA and lysates were probed for B’
to show induced expression. (B) Bar graphs show densitometric analysis of TrkA activity as measured by pY/total signal of four independent experiments ± SEM (*p value< 0.05). (C-D) Extended NGF treatment showed enhanced activation of downstream kinases. B’
expression was induced 3d ± Dox and stimulated with 20ng/ml NGF for indicated times. Normalized lysates were probed for total and phospho-Akt and ERK. (E-F) Graphs showed densitometric analysis of Akt and ERK activity as measured by phospho/total signal of three independent experiments ± SEM. (G) B’
expression PC6-3 cells were transiently transfected with empty vector, RasV12 or CA-MEK1 and treated 2d ± Dox. MAPK activity was measured by a dual-luciferase reporter assay. Representative graph from four independent experiments express data as relative to basal MAPK activity in empty vector transfected cells. SEM ± of triplicate samples. (H) Diagram designates site of pathway activation and hypothesized B’
site of action.

42

Figure 2.6 Multiple B’ subunits associate with the TrkA receptor. (A) PC6-3 cells that inducibly express B’
were treated 3d ± Dox and stimulated for 60 min with 20ng/ml NGF. TrkA and IgG immunoprecipitates were probed for PP2A/B’
and PP2A/C subunits to show association. (B) HA-tagged B’ regulatory subunits were transiently overexpressed in PC6-3 cells for 3d and stimulated for 60 min with 20ng/ml NGF. TrkA immunoprecipitates were probed for pan TrkA and HA to show subunit association. Lysates were probed with HA to show expression levels. HA-tagged CTRL was used as negative control. (C) PC6-3 cells were stimulated for 60 min with 20ng/ml NGF. TrkA and IgG immunoprecipitates were probed for PP2A/B’ and PP2A/C subunits to show endogenous subunit association. PP2B/Calcineurin shows no association. Trk and IgG immunoprecipitates from mouse spinal cord and brain extracts were probed for Trk, PP2A/B’ and PP2A/c subunits. (D) Microcystin-LR pull down from brain extracts. Control sample was incubated with free Microcystin-LR for 30 min to block PP2A/C binding and Microcystin-LR agarose was added to both samples for 60 min. Samples were blotted for Trk to show association to PP2A.

43

Figure 2.7 Multiple B’ subunits regulate TrkA activity. (A) HA-tagged regulatory subunits and vector or shRNAs were coexpressed in PC6-3 cells. Lysates were probed with HA to show relative knockdown of each subunit and PP2A/B pan as loading control and to show regulatory subunit downregulation specificity. (B) B’ regulatory subunit shRNAs were transiently expressed for 3d and MAPK activity was measured by a dualluciferase reporter assay. Bar graph expresses data as relative to control hairpin (CTRL). SEM ± of three independent experiments.

44

Figure 2.8 PP2A/B’ members potentiate neurite outgrowth in PC12 cells. (A) B’
expression cells transiently transfected with membrane-targeted Lck-mCherry and B’
expression was induced 3d ± Dox. Cells were stimulated with 2 ng/ml NGF for 18 h and fixed. Representative images show enhanced neurogenesis with B’
expression. Scale bar depicts 10 microns. (B) Total neurite length per cell was measured using ImageJ and shown in bar graph (microns). SEM ± of three independent experiments (***p value< 0.001). (C) Membrane-targeted Lck-mCherry and B’ regulatory subunit shRNAs, control shRNA (CTRL) or empty vector (pSUPER) were transiently expressed for 3 d and stimulated with 2 ng/ml NGF for 26 h and fixed. Representative images show decreased neurogenesis with B’
and B’ expression. Scale bar depicts 10 microns. (D) Total neurite length per cell was measured using ImageJ, data was converted to microns and shown in bar graph. SEM ± of three independent experiments (***p value< 0.001).

45

Figure 2.9 PP2A/B’
potentiates neuronal differentiation in PC12 cells. B’
expressing cells were induced 2d ±Dox and treated with 10 ng/ml NGF for two or four days to induce differentiation. DNA content was measured from each sample by FACS analysis. (A) Representative histograms of DNA content showing a shift from cycling cells to arrested cells in a differentiated state after B’
-induced expression and 2d NGF. Percentage of cells in growth phase (S/G2) and arrested phase (G0/G1) are depicted in bar graphs. (B) No NGF ± B’
expression, (C) 2d NGF ± B’
expression, (D) 4d NGF ± B’
expression. SEM ± of four independent experiments (*p value< 0.05).

46

CHAPTER III ERK TARGETS THE TRKA JUXTAMEMBRANE DOMAIN AND NEGATIVELY REGULATES RECEPTOR ACTIVITY THROUGH SERINE PHOSPHORYLATION Abstract NGF can have profound effects on differentiation, proliferation, survival, cell motility, membrane trafficking and synaptic plasticity depending on the cellular context. Sustained signaling through TrkA is critical for the ability of NGF to differentiate PC12 cells into neurons. In the previous chapter, I have shown that protein phosphatase 2A (PP2A) plays an important role of in maintaining TrkA activity. PP2A is a heterotrimeric enzyme composed of scaffolding, catalytic and >20 regulatory subunits that control substrate specificity and subcellular localization. Herein, I present data addressing the overlying hypothesize that sustained TrkA signaling is negatively regulated by ERKmediated phosphorylation within the juxtamembrane domain of the receptor. Consistent with this hypothesis, TrkA Ser phosphorylation was shown to be upregulated in the absence of PP2A, while ERK inhibition decreased TrkA Ser phosphorylation. ERK activity caused TrkA activity to decrease while ERK inhibition increased sustained TrkA signaling. Furthermore, ERK was shown to target the juxtamembrane of TrkA in vitro. Mutating S471 decrease TrkA Ser phosphorylation and modulated TrkA activity. Taken together, ERK was established as a negative regulator of TrkA function. Introduction Nerve growth factor signaling Neurotrophins (NTs) are extracellular peptide hormones that mediate cellular fate in terms of survival, differentiation and development as well as higher order functions

47

such as nociception and long-term potentiation (LTP). NTs, such as NGF, BDNF, NT3 and NT4, activate cognate neurotrophin receptors (NTRs) from the Trk receptor tyrosine kinase family (Huang and Reichardt, 2003). Additionally, each NT has the ability to activate the p75 NTR through a low affinity binding site (Barker, 2007; Reichardt, 2006). NTRs are differentially expressed in defined cell types where receptor activity is responsible for differentiation, cell survival, LTP, membrane stability and neuronal maintenance. Nerve growth factor (NGF), the prototypical neurotrophin, selectively binds to and activates the TrkA receptor. TrkA is predominately expressed in peripheral sensory neurons, as well as in cholinergic neurons in the central nervous system (Liebl et al., 2000; Ma et al., 2000; Smeyne et al., 1994). The importance of this NTR receptor is exemplified by TrkA deficient mice that have compromised sensory and sympathetic development and die shortly after birth (Smeyne et al., 1994). TrkA activity is also implicated in peripheral sensitization to pain in sensory neurons (Nicol and Vasko, 2007), and a loss of TrkA activity is associated with the progression of Alzheimer’s disease in several animal models (Schindowski et al., 2008) and congenital insensitivity to pain with anhidrosis (CIPA) (Indo, 2001). TrkA belongs to the family of receptor tyrosine kinases (RTK’s) and is structurally characterized by an extracellular ligand binding domain, a transmembrane domain, an intracellular juxtamembrane (JM) domain and a tyrosine kinase domain that is followed by a C-terminal tail (Huang and Reichardt, 2003). NGF binding to the extracellular domain brings two TrkA receptors together and orients a homodimer that facilitates trans/auto phosphorylation of three tyrosine residues within the kinase core to achieve a fully active conformation (Cunningham et al., 1997; Inagaki et al., 1995). Once activated, the kinase domain targets conserved Tyr residues on TrkA, that when phosphorylated, serve as docking sites for adaptor proteins that signal down the MAP kinase, PI3 kinase and PLC
cascades to initiate cellular responses (Reichardt, 2006). Similarly, a number of other extracellular growth factors, such as NGF, EGF, FGF and

48

IGF, activate these exact pathways by RTK stimulation (Hubbard and Miller, 2007). So, how do different ligands induce different responses? The answer lies in the subtle structural differences distinguishing each receptor and coordinate differential association with intracellular signaling proteins. Transient vs. sustained signaling Unlike other RTKs such as the EGFR that undergoes transient activation, a hallmark of TrkA function is sustained activity, which is required for neuronal differentiation and survival (Marshall, 1995; Qui and Green, 1992). RTK activity is defined by receptor phosphorylation on specific Tyr residues that facilitate the initiation of signaling cascades. The MAP kinase and PI3 kinase pathways are activated via Tyr490 phosphorylation, and phospholipase C
1 is recruited to phospho-Tyr785 in the C-terminal tail to activate PKC-mediated signaling (Stephens et al., 1994). Tyr490 resides in a conserved NPxY motif within the JM domain proximal to the kinase domain. This site is phosphorylated after receptor activation and serves as a docking site for the Shc and FRS-2 adaptor proteins (Dikic et al., 1995; Meakin et al., 1999). Shc recruitment to TrkA initiates the Ras/Raf-1/MEK/ERK signaling cascade resulting in transient activity that potentiates cell division. TrkA association with FRS-2 stabilizes Crk/C3G/Rap1/B-Raf activity resulting in sustained ERK activity (Kao et al., 2001; Meakin et al., 1999). However, recent data by Arevalo et al. (2006) argue that the sustained ERK activity through the Crk/C3G/Rap1/B-Raf module is independent of FRS2 and is mediated by an unusual ankrin-rich transmembrane protein, ARMS. The mechanisms that dictate transient versus extended NGF signaling have puzzled scientists for years, and although many models describe different assemblages of receptor complexes responsible for sustaining ERK signaling, all of the paradigms rely on receptor activity, i.e. Tyr phosphorylation. Ultimately, the receptor controls the fate of

49

the cellular response; therefore, defining mechanisms that regulate receptor Tyr phosphorylation is necessary to understand the physiological responses imposed by NGF. Serine phospho-regulation of RTKs Feedback inhibition by Ser/Thr phosphorylation has been shown to play a significant role in regulating the activity of several receptor tyrosine kinases. For example, PKC targets the EGFR and c-Met receptors to negatively regulate tyrosine phosphorylation (Gandino et al., 1990; Hunter et al., 1984). More recent data have shown the c-Met receptor to be bi-directionally regulated by PKC and PP2A at a specific Ser residue in response to different environmental stimuli (Hashigasako et al., 2004). Since RTKs share a high degree of structural and functional homology, it is conceivable that TrkA may be similarly regulated by Ser/Thr phosphorylation. The TrkA receptor is heavily Ser phosphorylated upon NGF stimulation and receptor crosstalk through p75 and C2-ceramide signaling results in TrkA Ser phosphorylation that correlates with decreased TrkA tyrosine phosphorylation (MacPhee and Barker, 1997). Since experiments in Chapter 2 showed that PP2A inhibition caused a decrease in TrkA activity, it can be inferred that Ser/Thr phosphorylation must, either directly or indirectly, regulate receptor activity. Therefore, it was of interest to test the hypothesis that TrkA is negatively regulated by Ser phosphorylation. This chapter defines a kinase that targets the TrkA receptor and is responsible for a dynamic regulation of prolonged activity. Materials and methods Cell culture and plasmids PC12 TrkA overexpression (6-24) cells (a gift from Phil Barker (Hempstead et al., 1992)) and parental PC12 (PC6-3) cells (Pittman et al., 1993) and were cultured (37 °C, 5% CO2) in RPMI 1640 (Gibco) containing 10% horse and 5% fetal bovine serum (both heat-inactivated). Plasmids encoding pFC-MEK1, a plasmid encoding constitutively

50

active MEK1 (S218, 222D), was supplied as part of the PathDetect Elk1 trans-reporting system (Stratagene, La Jolla, CA). pCMX HA-TrkA was a gift from Susan Meakin. The HA-TrkA sequence was swapped into pcDNA3.1 using flanking BamHI/EcoRI sites. The cDNA encoding the intracellular domain of TrkA (aa 441-790) was cloned by PCR the from HA-TrkA construct using forward primer 5’-CGC GGA TCC GGA GGT TCC AGC GGC GGC AAC AAA TGT GGA CAG AGG AGC-3’ and reverse primer 5’-CGC GAA TTC TAG CCC AGA ACG TCC AGG TAA C-3’. The fragment was ligated into BamHI/EcoRI site of pGEX2T expression vector and termed GST-TrkAic. pcDNA3.1 HA-TrkA or GST-TrkAic was used as a template for site directed mutagenesis using the QuickChange (Statagene) whole plasmid PCR protocol. TrkA The ERK, PKA and PKC predicted site mutants were made by PCR amplification of the TrkA coding sequence as two fragments linked by introduction of a silent site near the amino acid to be changed. Antibodies Antibodies commercially available are as follows: pan-Trk C-14 and agarose conjugate, Trk B-3, total ERK, (Santa Cruz Biotechnology, Santa Cruz, CA), HA epitope (Sigma, St. Louis, MO); phospho-TrkA Y490, phospho-Ser-473 Akt, total Akt, phosphoERK1/2 (Cell Signaling, Beverly, MA); phospho-Tyr (4G10) (Upstate Biotechnology, Lake Placid, NY). Secondary antibodies include goat
Rabbit HRP and goat
Mouse HRP (Santa Cruz Biotechnology, Santa Cruz, CA); goat
Rabbit IR800 and goat
Mouse IR680 (Licor, Lincoln, NE). Other reagents Rat nerve growth factor (NGF, 2.5 S) Yersinia phosphatase (YOP) and Lambda phosphatase were purchased from Upstate Biotechnology. Growth factors are stored at –20 °C in lyophilized aliquots and dissolved to 100X in 1/10th culture medium prior to use. Okadaic acid and microcystin-LR were purchased from Alexis (Lausanne,

51

Switzerland). U0126 was purchased from Promega (Madison, WI). Bradford reagent was purchased from BioRad (Hercules, CA). ERK activation and inhibition studies PC12 6-24 cells were seeded at 1 x 105/24 well collagen coated plate for 2d; or transiently transfected with pFC-MEK1 using Lipofectamine 2000, serum starved for 2hr and pretreated ±50
M U0126 for 15 min, then stimulated with 50 ng/ml NGF for 0-180 min. Cells were washed with cold PBS and lysed with 1X SDS page sample buffer, sonicated and normalized for total protein content. Samples were subjected to SDSPAGE and transferred to PVDF membrane (0.45 μM Millipore) for western blotting. Images were gathered using either a Kodak Imager 440 or Odyssey Imaging System (Licor, NE) Immunoprecipitations PC6-3 cells were seeded in 100 mm dishes and were transiently transfected with pFC-CA-MEK for 2d in some experiments. Cells were serum-starved for at least 2 h and pretreated in some cases with ±50
M U0126 for 15 min prior to adding growth factors at staggered times. After washing with phosphate-buffered saline, cells were harvested in RIPA lysis buffer (1% TX-100, 0.5% sodium dodecyl sulfate (SDS), 0.5% deoxycholate (DOC), 150 mM Tris pH 8.0, 300 mM NaCl, 1 mM EDTA, 1 mM EGTA supplemented with protease/phosphatase inhibitor cocktail – 1 mM benzamidine, 10
g/ml (20
M) leupeptin, 1mM pepstatin and 250
M PMSF, 1 mM
-glycerolphosphate, 2.5 mM sodium pyrophosphate and 0.5
M Microcystin-LR). Soluble protein was quantified using Bradford reagent and equal lysates were immunoprecipitated using 1.5
g of TrkA (C-14) antibody and protein A-agarose or agarose conjugate for 4-6 h. Beads were washed 4 times with 0.5% TX-100 TBS, extracted in SDS sample buffer, subjected to SDS-PAGE and transferred to PVDF membrane (0.45 μM Millipore) for western

52

blotting. Images were gathered using either a Kodak Imager 440 or Odyssey Imaging System (Licor, NE). Quantitative immunoblotting with phospho-specific antibodies Enhanced chemiluminescence (SuperSignal, Pierce) images were captured using a Kodak Imager 440, and band intensities were quantified with the ImageJ software gel analyzer plug-in. Phospho-specific antibody signals were divided by total protein antibody signals to control for loading differences. All signal intensities were normalized to the maximum intensity within each experiment. Statistical analysis was performed on data from at least four independent experiments using Prism 4.0 to run two-way ANOVA with Bonferroni post-tests for individual time point comparisons. Metabolic labeling PC12 (PC6-3) cells were seeded at 1.2 x106/ 6 well dish and transiently transfected with HA-TrkA WT or S471A for 2d prior to metabolic labeling. PC12 6-24 cells seeded at 8x105/ 6 well dish are used in other experiments. Cells were washed 2X with RPMI without PO4 (Invitrogen) and starved for 30 min with media containing 0.5% dialyzed FBS in RPMI –PO4 with 1X glutamax. Cells were labeled with 500
Ci/ml 32P orthophosphate (Perkin Elmer, Waltham, MA) prior to inhibitor treatment: 30 min before addition of 250
M OA for 1 hr; 60 min before addition of 50
M U0126 for 30 min; or 2 hr in transfected cells. Cells were then stimulated with 50 ng/ml NGF for 0-60 minutes, harvested in RIPA buffer and TrkA was IPed from normalized lysates using TrkA beads or HA-beads. Samples were washed 4X in 0.5% TTBS containing protease/phosphatase inhibitor cocktail and tubes were changed once to minimize background. Samples were split and resuspended in phosphatase buffer (YOP: 50 mM Tris pH 7, 100 mM NaCl, 5 mM DTT, 2 mM EDTA, 0.05% TX-100 or Lambda: 50 mM Tris pH 7.5, 100 mM NaCl, 2 mM DTT, 0.1 mM EGTA, 2 mM MnCl2 0.05% TX-100 ) and supplemented with

53

100U of YOP or 200U of Lambda phosphatase for 15 min at 30ºC with gentle mixing. Samples were quenched with 5X SDS sample buffer + 100 mM ETDA. Samples were subjected to SDS-PAGE and transferred to PVDF membrane (0.45 μm Millipore). Membranes were exposed to phosphoimager and subsequent western blotting. Images were gathered using either a Kodak Imager 440 or Odyssey Imaging System (Licor, Lincoln, NE) GST fusion protein plasmid generation and protein purification The intracellular domain of TrkA (aa 441-790) was cloned using pCMX HATrkA as a template using primers containing flanking XhoI and NdeI sites for insertion into pGEX2T vector to create N-terminal GST-TrkA fusion constructs. Subsequent GST-TrkA mutants were generated using point mutant primers in full plasmid PCR (QuickChange, Stratagene) with pGEX2T GST-TrkA WT template. GST-TrkA was purified from E. coli using the following detailed protocol. Starter culture was grown for 8 hrs shaking 250 rpm at 37ºC. 100 ml SOB +Amp culture is inoculated overnight with 200
ls of starter culture. 25 mls of overnight culture is added to fresh 100 ml SOB +Amp and grown to OD 1-1.2 in 250 ml flask. 50 mls fresh SOB +Amp was added and 95% EtOH was added to make a final concentration of 4%. Cultures were induced with 0.1mM IPTG for 20 hrs shaking 170 rpm at 18ºC. Cells were pelleted at 1000 x g and stored at -70ºC until further processing. Cells are thawed and lysed for 2 hours using GST lysis buffer (1% Triton X-100, 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, protease inhibitor cocktail - 1 mM benzamadine, 1 mM PMSF, 1
g/ml Pepstatin A, 10
g/ml Leupeptin, 0.2
g/ml Aprotinin, or protease inhibitor tablets were used at 1 per 10 mls of buffer (Complete mini tablets, Roche)) supplemented with 500
g/ml lysozyme. Lysates were centrifuged at 40,000 x g for 60 min. Supernatants were collected and incubated with 500
ls of Glutathione-agarose beads (Sigma) for 2-12

54

hours. Beads were washed using 200 mls of GST wash buffer (1% Triton X-100, 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM benzamadine plus fresh 1 mM PMSF). Beads were then washed with 50 mls of elution buffer (0.1% Triton X-100, 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM benzamadine plus fresh 1 mM PMSF). GST-TrkA beads were transferred to a 2 ml microfuge tube and resuspended in an equal volume of GST-storage buffer (0.05% Triton X-100, 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM benzamadine and 50% glycerol) and stored at -20ºC. Alternatively, GST-TrkA was eluted from the glutathione beads using an equal volume of 20 mM free glutathione in elution buffer. Samples were incubated on rotator for 3 hours and supernatant was removed. Elution was repeated to recover more protein from beads. Eluates were pooled and dialyzed in GST dialysis buffer (0.05% Triton X-100, 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM benzamadine with fresh 0.5 mM PMSF) to remove free glutathione. Samples are collected and stored at -20ºC in 50% glycerol. GST-TrkA and bovine serum albumin (BSA) standards were run on SDS-PAGE and concentrations were determined using quantitative coomassie staining and ImageJ analysis. In vitro kinase assays GST-TrkA proteins (200 ng) were suspended in 1X kinase buffer (0.01% Triton X-100, 50 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 2 mM DTT, 1 mM EDTA,) on ice. The reaction was started by the addition of ATP buffer (final concentration: 1X kinase buffer, 200
M ATP, 0.05
Ci 32P
-ATP/
l and 0.05
M microcystin LR) and placed at 30ºC shaking for 5 and 60 minutes. In some experiments 100 ng of GST-Shc was included. Reactions were quenched with 5X SDS sample buffer +100 mM EDTA. Samples were subjected to SDS-PAGE and Coomassie-blue stained for total protein, or transferred to PVDF and subjected to pY490 and total TrkA western blotting. Gels or membranes were exposed to phosphoimager to detect 32P incorporation.

55

Results PP2A inhibition causes an increase in TrkA serine phosphorylation. Several studies using the EGFR, c-MET and KIT receptor systems confirm that RTK activity can be regulated by Ser phosphorylation (Edling et al., 2007; Gandino et al., 1990; Hunter et al., 1984). To date, no reports have examined the role of PP2A in TrkA regulation. However, TrkA receptors are indeed Ser phosphorylated in response to p75NTR activation and ceramide production, and in these states TrkA is less responsive to NGF (MacPhee and Barker, 1999). To determine if PP2A is involved in modulating Ser phosphorylation of TrkA, metabolic radiolabeling was used to monitor TrkA phosphorylation in response to NGF with or without PP2A inhibition. PC12 cells in which the cellular ATP pool was radiolabeled by incubation with 32P orthophosphate were pretreated ± okadaic acid (OA) and stimulated with NGF for various times. Initial experiments measured Tyr phosphorylation by western blot and total Tyr/Ser/Thr phosphorylation through 32P incorporation on the TrkA receptor after 0, 5 and 60 minutes of NGF stimulation. Results showed a detectable increase in total phosphorylation at the 0 and 60-minute time point, while total 32P phosphorylation detected was equal between ±PP2A inhibition at 5 minutes of NGF stimulation (Figure 3.1). In contrast, PP2A inhibition decreases pTyr labeling at 0 and 60 minutes of NGF. However, as demonstrated in previous experiments (Figure 2.1), after 5 minutes of NGF pTyr was indistinguishable between the two treatments. These inverse phosphorylation patterns indicate that Ser/Thr phosphorylation accumulates over time in the absence of PP2A activity. Therefore, in subsequent experiments 0, 15 and 60 minutes of NGF stimulation were used to discriminate TrkA Tyr/Ser/Thr phosphorylation after PP2A inhibition. To better determine whether Ser/Thr phosphorylation is the predominating event in the absence of PP2A, Yersinia tyrosine phosphatase (YOP) (Zhang, 2003) was used to

56

selectively remove phosphates from pTyr residues while leaving Ser/Thr phospho-sites intact to be analyzed. The TrkA receptor was IPed from PC12 6-24 cells pre-equilibrated with 32P orthophosphate and treated with ±OA/NGF. The IPs were treated ±YOP or lambda phosphatase, which selectively removed all phosphate modifications. Figure 3.2 shows the extent of Ser/Thr phosphorylation that occurred on the TrkA receptor after NGF treatment. The first panel of Figure 3.2 again demonstrates an increased total phosphate content after OA and NGF treatment while total pTyr was reduced. The second panel exhibits how YOP selectively removed Tyr phosphates; demonstrating an elevated Ser/Thr phosphate content in the +OA/±NGF treated samples. The third panel confirms that detected 32P signal was indeed phosphoryl modifications removable by phosphatases. This was demonstrated by the lack of pTyr and 32P signal in Lambda phosphatase treated lanes. Repeated experiments consistently showed a significant decrease in pTyr detection after OA/NGF treatments. Conversely, basal levels of total Tyr/Ser/Thr phosphorylation were amplified by OA treatment, as well 60 minutes of NGF stimulation. YOP treatment, by selectively removing Tyr phosphorylation, exposed the effect of PP2A on TrkA Ser/Thr incorporation. The bar graph in Figure 3.2D shows significant increases in phosphorylation after OA treatment at all time points ±NGF. These data demonstrate that PP2A inhibition promotes TrkA Ser/Thr phosphorylation while, concomitantly, diminishing Tyr phosphorylation and TrkA activity. Based on these results, we hypothesized that PP2A positively regulates TrkA activity by modulating Ser phosphorylation, either directly or indirectly through the regulation of other proteins. Screen for specific kinases that modulates TrkA activity TrkA receptor phosphorylation is a dynamic process that must be tightly regulated to elicit proper cellular responses. This can be illustrated by observing Tyr phosphorylation kinetics after exposure to NGF (Chapter 2 Figure 2.1). TrkA was

57

maximally Tyr phosphorylated by 5 minutes of NGF treatment. The signal decreased by 50% after only 15 minutes; thereafter, a steady phospho-state is maintained for greater than 3 hours. Given that OA treatment resulted in elevated TrkA Ser phosphorylation and a decrease in TrkA activity, PP2A is likely to contest an inhibitory kinase that is regulating the receptor through Ser modification. Based on this hypothesis, one would expect kinase inhibition to: 1) decrease receptor Ser phosphorylation and 2) increase receptor activity. To initially identify a kinase that is involved in regulating TrkA activity, specific pharmacological kinase inhibitors and specific kinases activators were employed and receptor activity was tested. Table 1 displays the compounds used and their effects on TrkA activity. All assays were initially done in the TrkA overexpression cell line, PC12 6-24. Cells were pretreated with each compound at the given concentration and then stimulated with 20 ng/ml NGF for 0, 15 and 60 minutes. Lysates were analyzed by western blot using pY490 and pan TrkA antibodies to determine if there was an effect on receptor activity. Of the six kinases tested, this experiments identified only one possible target kinase, ERK. U0126, a selective MEK inhibitor that efficiently blocks NGF activated ERK phosphorylation, produced a modest increase in TrkA phosphorylation as measured using the pY490 antibody, while overexpressing constitutively active MEK (CA-MEK) produced a modest decrease in TrkA activity. No other compound, except the JNK inhibitor (SP600125) caused any significant change in TrkA phosphorylation. The list of compounds tested is by no means exhaustive, and many other kinases may target TrkA to regulate its activity. In addition, nonspecific inhibition of the TrkA kinase (e.g. by SP600125) cannot be ruled out. However, for reasons discussed previously and later in the text, ERK was evaluated in further detail. TrkA activity is modulated by ERK. ERK is a Ser/Thr kinase that is directly activated through phosphorylation by the dual specificity kinase MEK in the MAP kinase signaling module (Yoon and Seger,

58

2006). U0126 is a potent inhibitor of MEK (Ahn et al., 2001) that I have used to effectively inhibit ERK in assays examining receptor activity in the PC12 6-24 TrkA overexpressing cell line. After pretreatment with 50
M U0126 and stimulation with 50 ng/ml NGF, lysates were subjected to western blot analysis measuring total TrkA and pTrkA 490 to analyze the effects of ERK inhibition on TrkA activity. Prolonged TrkA phosphorylation was significantly increased on residue pY490 after 60 minutes of NGF treatment (Figure 3.3 A). The increase in phosphorylation persisted for more than 3 hours in two independent experiments suggesting that ERK inhibits TrkA activity (Figure 3.3 B). Native PC6-3 cells express low levels of TrkA receptors, which have been calculated to be approximately 1000/cell (Hempstead et al., 1992; Landreth and Shooter, 1980). Accordingly, low receptor levels are essential as TrkA overexpression causes membrane crowding and leads to ligand independent activation and downstream signaling that may be detrimental to cellular function. Giving that PC12 6-24 cells express 10 times more receptor than native PC6-3 cells, compensatory mechanisms may play a key role in regulating cellular homeostasis and consequently confound results when examining receptor signaling (Hempstead et al., 1992). To overcome this caveat, PC6-3 cells were utilized to provide a more physiological cell model. Since receptor numbers are low in these cells, TrkA IP’s were analyzed after U0126 pretreatment and NGF stimulation. Verifying the effects seen in the overexpression system, TrkA activity in native cells was also augmented in the absence of ERK activity. This effect was seen after 15 and 60 minutes of NGF stimulation (Figure 3.4 A, B). These experiments suggest that ERK is a negative regulator of TrkA function. If ERK does target the receptor, then increasing ERK activity should have the converse effect and decrease TrkA Tyr phosphorylation. To test this hypothesis, constitutively active MEK1 (CA-MEK1) was introduced to the system to manipulate ERK activity. Typically, Raf1 is activated in response to stimuli and phosphorylates

59

MEK1/2 on Ser218 and Ser222 to fully activate the kinase, which then activates ERK1/2 through phosphorylation on residues Thr202 and Tyr204 (Yoon and Seger, 2006). Introducing point mutations in MEK1 (S218, 222D) creates a constitutively active enzyme that can activate ERK in the absence of growth factor stimulation (Mansour et al., 1994). PC12 6-24 cells were transiently transfected and stimulated with 50 ng/ml NGF prior to western blot analysis of TrkA activity. Figure 3.3 (C, D) shows ERK overactivity decreased TrkA Tyr phosphorylation after 60 min; an effect that persisted through 180 minutes of NGF treatment. To validate these results in another system, native PC6-3 cells were also transiently transfected with CA-MEK and vector control and observed for changes in TrkA activity. In agreement with data obtained form the overexpressing cell line, at the 60 min with NGF, CA-MEK expression in native cells resulted in a significant reduction in TrkA phosphorylation (Figure 3.4C, D). These data strongly implicate ERK as a negative regulator of TrkA activity as kinase upregulation creates a decrease in TrkA activity, while ERK inhibition potentiates TrkA activity. However, these experiments do not address whether TrkA is directly phosphorylated by ERK or if TrkA receptor activity is regulated through some other protein. ERK regulates TrkA serine phosphorylation Precise Trk receptor regulation is critical for downstream signaling that governs the development and maintenance of the nervous system. Earlier experiments measuring pTyr content demonstrated that ERK inhibition could potentiate TrkA activity, while conversely, ERK activation decreased TrkA function. To assess whether ERK affects Ser phosphorylation of the receptor, metabolic labeling studies were used to measure TrkA Ser/Thr phosphorylation after ERK manipulation. As previously described, metabolically labeled TrkA receptors from PC12 6-24 cells pretreated with 50
M U0126 and NGF were stripped of pTyr modifications using YOP. The remaining phospho signal was analyzed to compare the effects of ERK inhibition (Figure 3.5 A). In repeated

60

experiments, normalized 32P/total TrkA receptor showed a modest trend towards a decrease in phospho-Ser/Thr content after ERK inhibition at 0 and 15 minutes of NGF treatment and a significant, albeit minor difference at 60 minutes when compared to control treated cells (Figure 3.5 B). Interestingly, comparing Ser/Thr phosphorylation to Tyr phosphorylation profiles after ERK inhibition, an inverse pattern of incorporation suggests that Ser/Thr modification of TrkA negatively regulates receptor activity as measured by Tyr phosphorylation. These data nicely complement the OA metabolic labeling experiments that correlate increased Ser/Thr phosphorylation to decreased in TrkA activity. Taken together, these data support a model where ERK negatively regulates TrkA activity and this negative regulation is opposed by PP2A to sustain TrkA activity. The TrkA juxtamembrane domain is targeted by ERK in vitro. Mitogen activated protein kinases, including ERK, are essential signaling molecules that assert numerous biological functions such as cell proliferation and differentiation. ERKs have been shown to target over 150 substrates that range in function from transcription factors and cytoskeleton proteins to other signaling molecules and membrane receptors (Yoon and Seger, 2006). Activated ERK recognizes the consensus phosphorylation site Pro-X-Ser/Thr-Pro, where the -2 Pro is less conserved, leaving Ser/Thr-Pro as the minimal recognition sequence for the kinase (Gonzalez et al., 1991). To determine if ERK or any other kinase is predicted to phosphorylate TrkA, analysis of the intracellular domain (aa 441-790) using NetPhos (NetPhosK) revealed several predicted kinase consensus sites. Also, threading the protein sequence against the crystal structure of a closely related RTK, MuSK, using the 3DPSSM server (3DPSSM) provided insights to consensus sites that are predicted to be exposed for kinase modification. Together, the data suggested that ERK, PKA, CAMKII and PKC might

61

target the receptor. Based on the pharmacological inhibition studies described in Table 1, ERK, PKA and PKC were attractive candidates to pursue in further studies. The prediction software identified two consensus PKA sites in TrkA. One located at S699, just downstream of the kinase domain and the other at S469 in the juxtamembrane domain. One ERK consensus site was identified only two amino acids downstream of the PKA site at S471 in the juxtamembrane domain. A single PKC site was identified in the C-terminal tail at residue S726. To examine in vitro kinase phosphorylation of TrkA, glutathione S-transferase (GST) was fused to the intracellular domain of TrkA and alanine or aspartic acid point substitutions were introduced within the predicted kinase targeting sequences. Single point mutations were introduced to disrupt the PKA and PKC site at S699 and S726 respectively, while a triple mutant spanning S468, S469, S471 (3S>A/D) in the juxtamembrane domain was generated to disrupt the other PKA site as well as the ERK consensus site. Initial characterization of wild type and mutant fusion proteins purified from bacteria showed considerable intrinsic kinase activity as seen by western blot detection of pTyr and 32P incorporation assays (Data not shown). Using bacterially expressed GSTShc as a Tyr substrate for GST-TrkA, in vitro phosphorylation assays demonstrate that the bacterially expressed TrkA tyrosine kinase not only autophosphorylates, but also phosphorylates a known substrate as detected by 32P incorporation and pY western blotting (Figure 3.6B). Each GST-TrkA mutant exhibited unique functional kinase activity and differing kinetics compared to wild type. Table 3.2 summarizes pY content of proteins after bacterial purification as well as in vitro 32P incorporation after incubation with 32P
-ATP. Western blot analysis for pY indicated if the kinase was functional after purification. Most mutants were purified with qualitatively similar levels of Tyr phosphorylation. However, both S>A and S>D mutations at S699 disrupted kinase activity, while only S726A had no intrinsic kinase activity after purification. Mutations at these sites may structurally alter the active site to prohibit kinase activity. Initial in

62

vitro phosphorylation assays measured autophosphorylation as well as Shc phosphorylation. Subsequent experiments pretreated the GST-TrkA proteins with YOP to remove all Tyr phosphates so all enzymes would start in an inactive state. In all instances, Shc phosphorylation corresponded with TrkA autophosphorylation levels. Comparing mutant kinase activity showed differences between mutants. 3S>D had higher activity while 3S>A and S726A had lower activity than wild-type (less 32P incorporation during the incubation time). These studies have not been pursued in enough detail to draw any conclusions. Generating functionally active GST-TrkA was an invaluable tool and has only been reported once in the literature (Llovera et al., 2004). However, initial in vitro phosphorylation experiments using PKA and ERK were flawed by the fact that intrinsic tyrosine kinase activity of TrkA phosphorylated each protein to equal levels, masking any effects of other kinases on TrkA mutants. Therefore, kinase dead fusion proteins were generated by changing a critical lysine (K538A) in the kinase domain to allow accurate analysis of kinase targeting. These constructs were named kinase dead (KD) mutants (e.g. GST-TrkA(KD) WT). Surprisingly, in vitro phosphorylation experiments revealed that WT-TrkA was not phosphorylated by the catalytic subunit of PKA (data not shown). Since PKA does not target the predicted site and mutations in this site render the GSTTrkA protein inactive (Table 3.2), it is likely that S699 may be buried to structurally stabilize the kinase domain, while S469 was never a true PKA site. On the other hand GST-TrkA(KD) was phosphorylated by ERK and the 3S>A mutation blocked ERK phosphorylation (Figure 3.6 C). To further delineate site specificity, a single point substitution at S471 was engineered into the kinase dead GST-TrkA WT protein. Mutating only the S471A site abolished ERK phosphorylation of TrkA equally well as 3S>A mutant and identified the precise site of ERK directed phosphorylation (Figure 3.6 D).

63

Functional effects of S471 mutations. The juxtamembrane domain of the TrkA receptor is predicted to be a very dynamic unstructured coil (prediction) similar to what is predicted in other RTKs. The disordered nature of this region allows induced modifications (i.e. phosphorylation, ubiquitination) to stabilize defined structure that may impose negative or positive regulation on TrkA activity. For example, a single point mutation in the juxtamembrane region of the c-KIT receptor causes enhanced activity resulting in gastrointestinal stromal tumors (Hirota et al., 1998). On the other hand, Tyr mutations in the juxtamembrane region of Eph family receptors, as well as MuSK receptors, greatly impair autophosphorylation and signaling (Binns et al., 2000; Kahl and Campanelli, 2003). To this end, my data supports the hypothesis that TrkA is, in part, regulated by juxtamembrane posttranslational modifications, specifically Ser phosphorylation. Having shown that ERK modulates TrkA activity and is able to target TrkA in vitro, the next step is to determine the impact of S471 mutations on TrkA activity in culture. The TrkA receptor has 20 Ser and 9 Thr residues than can potentially be phosphorylated. Since metabolic labeling studies have shown that TrkA is heavily Ser/Thr phosphorylated after NGF treatment, it is possible that many of these sites are indeed modified. To address if S471 is phosphorylated after NGF stimulation, metabolic labeling experiments were utilized once again. HA-TrkA WT and HA-TrkA S471A were transiently transfected into PC6-3 cells for two days and the previously described 32P labeling protocol was followed. TrkA receptors were IPed using HA beads to enrich for transfected receptor and subjected to ±YOP treatment to remove Tyr phosphates. As demonstrated for many RTKs and shown in Figure 3.7, overexpressing TrkA receptors crowds the membrane forced ligand independent dimerization and activation that results in a high level of basal phosphorylation. However, both WT and S471A are inducibly Tyr phosphorylated with NGF, about 3 fold over basal level (Figure 3.7). YOP effectively eliminated Tyr-phosphates and unmasked the impact of the S471A mutation.

64

Comparing 32P incorporation to WT, S471A had less phospho-signal in both basal and NGF stimulated samples (Figure 3.7). Similar results have been repeated, once in PC6-3 cells and once in PC12 6-24 cells. These results clearly show that S471 is a major residue targeted for phosphorylation during TrkA activation. These experiments will need to be repeated to accurately quantify the S471 contribution to total TrkA Ser/Thr phosphorylation. Earlier experiments demonstrated that ERK inhibition increased TrkA activity while ERK activation decreased TrkA activity. If ERK does indeed target TrkA at S471 in cells, similar effects should be seen with phospho-mimetic or phosphorylationblocking residue substitution on the receptor. Since TrkA homodimerizes when activated by NGF, mutant expression in PC6-3 cells creates a heterogeneous system that can make functional effects difficult to interpret. Therefore human embryonic kidney 293 (HEK 293) cells, which do not natively express TrkA, are utilized to create a system that can express homogeneous populations of TrkA receptors. WT, S471A and S471D HA-TrkA constructs were initially titrated to express receptor levels that have minimal ligand independent autophosphorylation, and display detectible levels of NGF dependent activity. Subsequent experiments compared S471A and S471D phospho-pY490 signal was analyzed comparing the three HA-TrkA constructs (WT, S471A, S471D). The S471A mutation augmented ligand-independent receptor activity (Figure 3.8). This increased receptor activity translated through out 15 minutes of NGF, but was not different from WT at 60 minutes. Autophosphorylation of the S471D mutation in TrkA was not distinguishable from WT in the absence of NGF and was slightly elevated at 5 minutes. Then, after extended NGF treatment, phospho-levels began to decrease compared to WT through 60 minutes of NGF. Results from two and four independent experiments provide supporting evidence that S471 is, in reality, a site that regulates TrkA activity through phosphorylation.

65

Subsequent experiments addressed S471 mutant function in culture utilizing the MAPK reporter assay (described in Chapter 2). In these experiments, HA-TrkA mutants were transfected into PC6-3 cells for 2 days and stimulated with 2-10 ng/ml NGF for 4 hours. In, two independent experiments comparing fold Elk1 activity over basal activity of each transfected construct, one showed no significant difference between each construct in one experiment, while in the other S471D showed a significant increase in activity (data not shown). However in both experiments, HA-TrkA S471A showed higher and S471D showed less basal ERK activity, similar to what was seen at the receptor level using pY490 western blotting. Since this assay has caveats that are difficult to control, such as equal receptor expression and heterodimer contribution, further functional characterization of these mutants is required to draw any conclusions. Discussion Trk receptors belong to a large family of structurally similar receptor tyrosine kinases that activate many of the same signaling pathways, but are regulated by different substrates and signaling molecules that are temporally and spatially limiting. Herein, ERK has been identified as one of many possible kinases that negatively regulate TrkA activity by impacting receptor Ser phosphorylation at a specific residue within the juxtamembrane (JM) domain. These data add TrkA to a growing list of RTKs, including EGFR, c-MET, c-KIT and MuSK receptors, that are regulated by phosphorylation within the JM domain (Chan et al., 2003; Hashigasako et al., 2004; Hubbard and Miller, 2007; Kahl and Campanelli, 2003). Okadaic acid promotes TrkA serine phosphorylation Data in Chapter 2 established that PP2A regulates TrkA activity and suggests that phosphatase activity opposes Ser/Thr kinase activity to sustain receptor signaling. In an attempt to assign mechanism to this phenomenon, the most straightforward hypothesis TrkA activity is mediated through direct receptor Ser/Thr phosphorylation - was initially

66

tested using metabolic labeling in the presence of PP2A inhibition. Treating samples with the selective Tyr phosphatase YOP allowed quantification of receptor Ser/Thr phosphorylation. Without NGF stimulation, PP2A inhibition caused an apparent increase in TrkA Ser/Thr phosphorylation that was seen ± YOP treatment. This effect remained robust after 15 minutes of NGF treatment. Although TrkA pSer/Thr remained significantly elevated in OA treated cells at 60 minutes, the margin of difference greatly decreased as control treated cells accumulated a substantial amount of Ser/Thr phosphorylation over this period (Figure 3.2). This suggests that TrkA Ser/Thr phosphorylation is typically part of a negative feedback loop that responds to saturating doses of NGF by to dampening and controlling a flood of incoming signal. These data suggest that PP2A antagonizes a negative feedback loop to facilitate prolonged signaling as demonstrated by PP2A’s ability to inversely regulate TrkA Ser/Thr and Tyr phosphorylation. TrkA serine phosphorylation is a potential form of receptor cross-talk Negative regulation imposed by Ser/Thr phosphorylation may also serve as a method of receptor cross-talk. MacPhee et al. (1997) demonstrated that activation of p75 resulted in a reduction of TrkA responsiveness to NGF, likely due to TrkA Ser phosphorylation via a C2-ceramide activated kinase. Interestingly, ceramide has many effects on signal transduction pathways, one being activation of the MAP kinase pathway (Pfeilschifter and Huwiler, 1998; Yao et al., 1995). However, in a later report (MacPhee and Barker, 1999), MacPhee et al. (1999) showed prolonged C2-ceramide incubations not only enhanced NGF-induced receptor Tyr phosphorylation, but also forced elevated levels of ligand-independent Tyr phosphorylation. Although not intuitive, these results are consistent with the proposed model of PP2A mediated regulation of TrkA activity. Many studies have also shown that PP2A is activated by C2-ceramide (Chalfant et al.,

67

1999; Dobrowsky et al., 1993; Wolff et al., 1994). Specifically, PP2A/B’ has been shown to be potently activated by C2-ceramide in a dose-dependent manner (Dobrowsky et al., 1993). Taken together, these data suggest that a short C2-cerimide incubation, leading to low local concentrations of ceramide, may activate specific kinases or relieve phosphatase inhibition of kinases that target the receptor (i.e. ERK), while longer incubations would foster greater PP2A/B’ activity and result in enhanced TrkA activity. Discovering a Ser/Thr kinase that targets TrkA In the pursuit of a kinase that may oppose PP2A regulation of TrkA, a panel of pharmacological agents directed towards major kinases was used to screen TrkA activity (Table 3.1). Initially, PKA was a promising candidate. Using PC12 6-24 cells, PKA activation with forskolin caused a dramatic inhibition of TrkA activity. However, repeated experiments in other cell lines showed an opposite effect, which is consistent with most of the data in the literature (Kalman et al., 1990; Kamei and Tsang, 2003; Nagase et al., 2005; Vossler et al., 1997). As compelling as the result was, the data could not be explained and therefore were discarded as artifact of the overexpression system. This assay also uncovered another potential candidate kinase, JNK. Once again PC12 624 cells showed a strong attenuation of TrkA activity, this time after treatment with the SP600125 JNK inhibitor. This result was repeated in native PC6-3 cells (data not shown); however, using a membrane permeable JNK inhibitor peptide (JIP), there was no effect in PC12 6-24 cells. Since there is no data in the literature that would support a JNK positive regulation of TrkA and the selective peptide had no effect, the impressive inhibition imposed by the kinase inhibitor was attributed to off target effects (i.e. TrkA kinase inhibition) and has not been pursued further. Consequently, ERK was the most promising kinase in the panel.

68

TrkA regulation through the juxtamembrane domain Although Trk receptors (TrkA, TrkB and TrkC) have a very high sequence homology, each responds in a unique way to ligand stimulation (Carter et al., 1995; Chen et al., 2005; Sommerfeld et al., 2000). The kinase domain in the Trk receptors is the most highly conserved domain and leaves little room for differential modes of regulation between receptors. However, sequence variations in the juxtamembrane domain provide distinction between TrkA and TrkB/TrkC that could allow exclusive regulation of each receptor (Figure 3.6A). Several studies have swapped the JM regions of TrkA and TrkB and have shown that the TrkA JM domain promotes a longer half-life, while the TrkB JM domain pushes the receptor down the lysosomal degradation pathway (Chen et al., 2005; Sommerfeld et al., 2000). Sequence analysis of Trk JM domains revealed a single putative ERK consensus site in TrkA that is not aligned in TrkB or TrkC. Herein, this site was confirmed to be targeted by ERK in vitro (Figure 3.6D). There are two prolinedirected Ser/Thr sites within the TrkB and TrkC JM region. The first site, at S478 has recently been determined to be a CDK5 targeted site that when phosphorylated, is essential for BDNF induced dendritic growth (Cheung et al., 2007). The second, located at T489 has yet to analyzed for kinase targeting. Assigning unique ERK targeting to TrkA further validates the accumulated data implicating ERK as a negative regulator specific to TrkA that allows for the differential regulation of the TrkA receptors. ERK regulates TrkA Tyr phosphorylation Several experiments, in multiple cell lines, have been performed to determine if ERK is indeed a negative regulator of TrkA activity. In summary, ERK activation by CA-MEK attenuates TrkA activity. Conversely, pharmacological inhibition of ERK facilitates TrkA Tyr phosphorylation (Figures 3.3, 3.4) while decreasing TrkA Ser phosphorylation (Figure 3.5). These effects were modest, but this may be a limitation of poor inhibitor selectivity causing unknown signaling components to either directly or

69

indirectly muffle the expected response. This is illustrated by the difference in magnitude of TrkA Ser phosphorylation detected after U0126 ERK inhibition (Figure 3.6) and after S471A mutation (Figure 3.8). U0126 shows slight, yet significant differences in Ser phosphorylation and the disruption of the ERK consensus site shows a robust change in Ser phosphorylation. These data suggest that a fraction of ERK is not completely inhibited by U0126 and/or the model is more complex than can be tested with pharmacological agents. ERK targets S471 in the TrkA juxtamembrane domain Subsequent analysis of TrkA function is preliminary (Figure 3.8). However, initial results support the idea that S471 is a major site for TrkA regulation. Experiments using a S471A mutant showed a robust increase in TrkA activity after acute NGF treatment when compared to WT activity. Conversely, S471D mutants showed a trend towards decreased activity after prolonged NGF stimulation. Given that the S471A mutant displays a gain of function phenotype, S471 is obviously not required for receptor autophosphorylation, but is a credible site for receptor regulation Several studies and recent crystal structures of RTKs have provided insight into the positioning of the JM region and how it may play a role in regulating receptor kinase activity (Hubbard, 2004; Hubbard and Miller, 2007; Till et al., 2002). This highly dynamic region in TrkA contains many modifiable amino acids that presumably offer many modes of regulation. In conclusion, data presented begin to elucidate a model were ERK imposes one form of negative regulation through phosphorylation of TrkA at S471 in the juxtamembrane domain.

70

Table 3.1 Summary of pharmacological agents used to screen TrkA kinase activity. Compound

Concentration

Target/action

DMSO

1:1000

Forskolin

TrkA Activity PC12 6-24

PC 6-3

Control vehicle

+++

+++

25
M

PKA activator

+

++++

PKI

*

PKA inhibitor

++++

ND

H89

50
M

PKA inhibitor

+++

ND

Chelerythrine chloride

1
M

PKC inhibitor

+++

ND

Calphostin C

0.1
M

PKC inhibitor

+++

ND

Bisindolylmaleim ide I

1
M

PKC inhibitor

+++

ND

KN-93

50
M

CamKII inhibitor

+++

ND

Trifluoperizine

20
M

Calmodulin inhibitor

+++

ND

W-7

20
M

Calmodulin inhibitor

+++

ND

LY29002

10
M

PI3 kinase inhibitor

+++

ND

U0126

50
M

MEK inhibitor

++++

++++

CA-MEK

*

ERK activator

++

++

SP600125

20
M

JNK inhibitor

+

+

JNK inhibitor peptide

50
M

JNK inhibitor

+++

ND

TrkA activity is measured by pY/pY490 quantitative western blotting and compared to control treated TrkA using two cell types in some experiments. + = relative TrkA Tyr phosphorylation compared to unstimulated control cells * = Transfected protein ND = No Data

71

Table 3.2 Summary of GST-TrkA mutant intrinsic kinase activity. TrkA mutant

Location

WT

-

3S>A

Auto-phosphorylation pY

32

Shc 32

P

pY

P

+++

+++

+++

+++

Juxtamembrane

+++

+

+

+

3S>D

Juxtamembrane

+++

++++

++++

++++

S699A

Kinase domain

-

-

-

-

S699D

Kinase domain

-

-

-

-

S726A

C-terminal tail

-

-

-

-

S726D

C-terminal tail

+++

+

+

+

S471A

Juxtamembrane

+++

ND

ND

ND

S471D

Juxtamembrane

+++

ND

ND

ND

Activity is qualitatively assessed using two measures of TrkA autophosphorylation and Shc phosphorylation. pY represents relative phosphorylation of TrkA or Shc compared to GSTTrkA WT after bacterial purification using pY western blotting. 32P represents relative 32P incorporation of TrkA or Shc compared to GST-TrkA WT after 60 minute incubation with 32 P-
ATP and detected by phosphoscreen imaging. - = no phosphorylation + = subjective phosphorylation ND = No Data

72

Figure 3.1 PP2A inhibition fosters TrkA Ser/Thr hyperphosphorylation while decreasing Tyr phosphorylation after prolonged NGF stimulation. PC12 6-24 cells were metabolically labeled prior to the addition of ± 250nM OA and then stimulated with 50ng/ml NGF for 5 and 60 minutes. TrkA IP’s were subjected to SDS page and transferred to PVDF. Total incorporated 32P (pTyr/Ser/Thr) was detected by phosphoimaging prior to western blot analysis to detect pTyr (pY) and total TrkA. In two experiments, PP2A inhibition with OA had no effect on pY or pTyr/Ser/Thr with acute, 5 min NGF treatment. However, pY levels decreased after prolonged NGF stimulation while pTyr/Ser/Thr levels increased over time implicating PP2A’s role in sustaining TrkA activity.

73

Figure 3.2 Enhanced TrkA Ser/Thr phosphorylation corresponds with decreased TrkA activity after PP2A inhibition. PC12 6-24 cells were metabolically with 32P orthophosphate prior to PP2A inhibition with ± 250nM OA for 2 hours and then stimulation with 50ng/ml NGF for 15 and 60 minutes. TrkA IP’s were treated with blank (-P’tase), Tyr-specific Yersinia phosphatase (YOP) or non-specific Lambda phosphatase for 15 min. Samples were subjected to SDS page and transferred to PVDF. (A) Representative images show pTyr, total incorporated 32P and total TrkA after 60 min of NGF treatment. Each panel shows samples after treatment with indicated phosphatase which represents residues that remain phosphorylated: pTyr/Ser/Thr; Ser/Thr; None. (B) TrkA pTyr phosphorylation was decreased after PP2A inhibition as shown by densitometric analysis of TrkA pY/total signal after 15 and 60 minutes of NGF. (C) TrkA total phosphorylation was increased after PP2A inhibition as shown by densitometric analysis of 32P/total signal of total pTyr/Ser/Thr levels from blank treated samples after 15 and 60 minutes of NGF. (D) TrkA pSer/Thr levels were increased after OA treatment as shown by densitometric analysis of 32P/total signal of total pSer/Thr levels from YOP treated samples after 15 and 60 minutes of NGF. (B-D) summary graphs of six independent experiments ± SEM (p-value < 0.05*,0.01**, p12. Genomics 36(1):168-170. McCright B, Rivers AM, Audlin S and Virshup DM (1996b) The B56 family of protein phosphatase 2A (PP2A) regulatory subunits encodes differentiation-induced phosphoproteins that target PP2A to both nucleus and cytoplasm. Journal of Biological Chemistry 271(36):22081-22089. Meakin SO and MacDonald JI (1998) A novel juxtamembrane deletion in rat TrkA blocks differentiative but not mitogenic cell signaling in response to nerve growth factor. Journal of neurochemistry 71(5):1875-1888. Meakin SO, MacDonald JI, Gryz EA, Kubu CJ and Verdi JM (1999) The signaling adapter FRS-2 competes with Shc for binding to the nerve growth factor receptor TrkA. A model for discriminating proliferation and differentiation. The Journal of biological chemistry 274(14):9861-9870. Moreno CS, Park S, Nelson K, Ashby D, Hubalek F, Lane WS and Pallas DC (2000) WD40 repeat proteins striatin and S/G(2) nuclear autoantigen are members of a novel family of calmodulin-binding proteins that associate with protein phosphatase 2A. The Journal of biological chemistry 275(8):5257-5263. Mumby M (2007) PP2A: unveiling a reluctant tumor suppressor. Cell 130(1):21-24. Nagase H, Yamakuni T, Matsuzaki K, Maruyama Y, Kasahara J, Hinohara Y, Kondo S, Mimaki Y, Sashida Y, Tank AW, Fukunaga K and Ohizumi Y (2005) Mechanism of neurotrophic action of nobiletin in PC12D cells. Biochemistry 44(42):1368313691. Nakagawara A (2001) Trk receptor tyrosine kinases: a bridge between cancer and neural development. Cancer letters 169(2):107-114. Nawa M, Kanekura K, Hashimoto Y, Aiso S and Matsuoka M (2008) A novel Akt/PKBinteracting protein promotes cell adhesion and inhibits familial amyotrophic lateral sclerosis-linked mutant SOD1-induced neuronal death via inhibition of PP2A-mediated dephosphorylation of Akt/PKB. Cellular signalling 20(3):493505. NetPhosK www .cbs.dtu.dk/services/NetPhosK.

96 Nicol GD and Vasko MR (2007) Unraveling the story of NGF-mediated sensitization of nociceptive sensory neurons: ON or OFF the Trks? Molecular interventions 7(1):26-41. O'Rahilly R and Muller F (2007) The development of the neural crest in the human. Journal of anatomy 211(3):335-351. Ogris E, Gibson DM and Pallas DC (1997) Protein phosphatase 2A subunit assembly: the catalytic subunit carboxy terminus is important for binding cellular B subunit but not polyomavirus middle tumor antigen. Oncogene 15(8):911-917. Okamoto K, Li H, Jensen MR, Zhang T, Taya Y, Thorgeirsson SS and Prives C (2002) Cyclin G recruits PP2A to dephosphorylate Mdm2. Molecular cell 9(4):761-771. Peng X, Greene LA, Kaplan DR and Stephens RM (1995) Deletion of a conserved juxtamembrane sequence in Trk abolishes NGF-promoted neuritogenesis. Neuron 15(2):395-406. Pfeilschifter J and Huwiler A (1998) Identification of ceramide targets in interleukin-1 and tumor necrosis factor-alpha signaling in mesangial cells. Kidney international 67:S34-39. Pittman RN, Wang S, DiBenedetto AJ and Mills JC (1993) A system for characterizing cellular and molecular events in programmed neuronal cell death. Journal of Neuroscience 13(9):3669-3680. predictionJ www .sbg.bio.ic.ac.uk/phyre/qphyre output/d00b2821c0b45b29/ summary.html. Qui MS and Green SH (1992) PC12 cell neuronal differentiation is associated with prolonged p21ras activity and consequent prolonged ERK activity. Neuron 9(4):705-717. Ratcliffe MJ, Itoh K and Sokol SY (2000) A positive role for the PP2A catalytic subunit in Wnt signal transduction. Journal of Biological Chemistry 275(46):3568035683. Rechsteiner M and Rogers SW (1996) PEST sequences and regulation by proteolysis. Trends in biochemical sciences 21(7):267-271. Reichardt LF (2006) Neurotrophin-regulated signalling pathways. Philosophical transactions of the Royal Society of London 361(1473):1545-1564. Rocher G, Letourneux C, Lenormand P and Porteu F (2007) Inhibition of B56-containing protein phosphatase 2As by the early response gene IEX-1 leads to control of Akt activity. Journal of Biological Chemistry 282(8):5468-5477. Ruediger R, Fields K and Walter G (1999) Binding specificity of protein phosphatase 2A core enzyme for regulatory B subunits and T antigens. Journal of virology 73(1):839-842. Ruediger R, Hentz M, Fait J, Mumby M and Walter G (1994) Molecular model of the A subunit of protein phosphatase 2A: interaction with other subunits and tumor antigens. Journal of virology 68(1):123-129.

97 Ruediger R, Pham HT and Walter G (2001) Disruption of protein phosphatase 2A subunit interaction in human cancers with mutations in the A alpha subunit gene. Oncogene 20(1):10-15. Ruediger R, Roeckel D, Fait J, Bergqvist A, Magnusson G and Walter G (1992) Identification of binding sites on the regulatory A subunit of protein phosphatase 2A for the catalytic C subunit and for tumor antigens of simian virus 40 and polyomavirus. Molecular and cellular biology 12(11):4872-4882. Ruvolo PP, Clark W, Mumby M, Gao F and May WS (2002) A functional role for the B56 alpha-subunit of protein phosphatase 2A in ceramide-mediated regulation of Bcl2 phosphorylation status and function. Journal of Biological Chemistry 277(25):22847-22852. Saraf A, Virshup DM and Strack S (2007) Differential expression of the B'beta regulatory subunit of protein phosphatase 2A modulates tyrosine hydroxylase phosphorylation and catecholamine synthesis. Journal of Biological Chemistry 282(1):573-580. Schindowski K, Belarbi K and Buee L (2008) Neurotrophic factors in Alzheimer's disease: role of axonal transport. Genes, brain, and behavior 7 Suppl 1:43-56. Schlessinger J (2000) Cell signaling by receptor tyrosine kinases. Cell 103(2):211-225. Schonthal AH (2001) Role of serine/threonine protein phosphatase 2A in cancer. Cancer letters 170(1):1-13. Seeling JM, Miller JR, Gil R, Moon RT, White R and Virshup DM (1999) Regulation of beta-catenin signaling by the B56 subunit of protein phosphatase 2A. Science (New York, NY 283(5410):2089-2091. Segal RA (2003) Selectivity in neurotrophin signaling: theme and variations. Annual review of neuroscience 26:299-330. Silverstein AM, Barrow CA, Davis AJ and Mumby MC (2002) Actions of PP2A on the MAP kinase pathway and apoptosis are mediated by distinct regulatory subunits. Proceedings of the National Academy of Sciences of the United States of America 99(7):4221-4226. Smeyne RJ, Klein R, Schnapp A, Long LK, Bryant S, Lewin A, Lira SA and Barbacid M (1994) Severe sensory and sympathetic neuropathies in mice carrying a disrupted Trk/NGF receptor gene. Nature 368(6468):246-249. Sommerfeld MT, Schweigreiter R, Barde YA and Hoppe E (2000) Down-regulation of the neurotrophin receptor TrkB following ligand binding. Evidence for an involvement of the proteasome and differential regulation of TrkA and TrkB. The Journal of biological chemistry 275(12):8982-8990. Sontag E, Nunbhakdi-Craig V, Sontag JM, Diaz-Arrastia R, Ogris E, Dayal S, Lentz SR, Arning E and Bottiglieri T (2007) Protein phosphatase 2A methyltransferase links homocysteine metabolism with tau and amyloid precursor protein regulation. Journal of Neuroscience 27(11):2751-2759.

98 Stephens RM, Loeb DM, Copeland TD, Pawson T, Greene LA and Kaplan DR (1994) Trk receptors use redundant signal transduction pathways involving SHC and PLC-gamma 1 to mediate NGF responses. Neuron 12(3):691-705. Sterberg RJ (2003) Psychology. Thomson Wadsworth. Stone SR, Hofsteenge J and Hemmings BA (1987) Molecular cloning of cDNAs encoding two isoforms of the catalytic subunit of protein phosphatase 2A. Biochemistry 26(23):7215-7220. Strack S (2002) Overexpression of the protein phosphatase 2A regulatory subunit Bgamma promotes neuronal differentiation by activating the MAP kinase (MAPK) cascade. Journal of Biological Chemistry 277(44):41525-41532. Strack S, Chang D, Zaucha JA, Colbran RJ and Wadzinski BE (1999) Cloning and characterization of B delta, a novel regulatory subunit of protein phosphatase 2A. FEBS letters 460(3):462-466. Strack S, Cribbs JT and Gomez L (2004) Critical role for protein phosphatase 2A heterotrimers in mammalian cell survival. Journal of Biological Chemistry 279(46):47732-47739. Strack S, Ruediger R, Walter G, Dagda RK, Barwacz CA and Cribbs JT (2002) Protein phosphatase 2A holoenzyme assembly: identification of contacts between Bfamily regulatory and scaffolding A subunits. Journal of Biological Chemistry 277(23):20750-20755. Strack S, Zaucha JA, Ebner FF, Colbran RJ and Wadzinski BE (1998) Brain protein phosphatase 2A: developmental regulation and distinct cellular and subcellular localization by B subunits. Journal of Comparative Neurology 392(4):515-527. Till JH, Becerra M, Watty A, Lu Y, Ma Y, Neubert TA, Burden SJ and Hubbard SR (2002) Crystal structure of the MuSK tyrosine kinase: insights into receptor autoregulation. Structure 10(9):1187-1196. Toledo-Aral JJ, Brehm P, Halegoua S and Mandel G (1995) A single pulse of nerve growth factor triggers long-term neuronal excitability through sodium channel gene induction. Neuron 14(3):607-611. Tolstykh T, Lee J, Vafai S and Stock JB (2000) Carboxyl methylation regulates phosphoprotein phosphatase 2A by controlling the association of regulatory B subunits. The EMBO journal 19(21):5682-5691. Traverse S, Gomez N, Paterson H, Marshall C and Cohen P (1992) Sustained activation of the mitogen-activated protein (MAP) kinase cascade may be required for differentiation of PC12 cells. Comparison of the effects of nerve growth factor and epidermal growth factor. The Biochemical journal 288 ( Pt 2):351-355. Ugi S, Imamura T, Ricketts W and Olefsky JM (2002) Protein phosphatase 2A forms a molecular complex with Shc and regulates Shc tyrosine phosphorylation and downstream mitogenic signaling. Molecular & Cellular Biology 22(7):2375-2387.

99 Usui H, Imazu M, Maeta K, Tsukamoto H, Azuma K and Takeda M (1988) Three distinct forms of type 2A protein phosphatase in human erythrocyte cytosol. The Journal of biological chemistry 263(8):3752-3761. Van Kanegan MJ, Adams DG, Wadzinski BE and Strack S (2005) Distinct protein phosphatase 2A heterotrimers modulate growth factor signaling to extracellular signal-regulated kinases and Akt. Journal of Biological Chemistry 280(43):36029-36036. van Lookeren Campagne M, Okamoto K, Prives C and Gill R (1999) Developmental expression and co-localization of cyclin G1 and the B' subunits of protein phosphatase 2a in neurons. Brain Research Molecular Brain Research 64(1):1-10. Varsano T, Dong MQ, Niesman I, Gacula H, Lou X, Ma T, Testa JR, Yates JR, 3rd and Farquhar MG (2006) GIPC is recruited by APPL to peripheral TrkA endosomes and regulates TrkA trafficking and signaling. Molecular and cellular biology 26(23):8942-8952. Vaudry D, Stork PJ, Lazarovici P and Eiden LE (2002) Signaling pathways for PC12 cell differentiation: making the right connections. Science (New York, NY 296(5573):1648-1649. Voorhoeve PM, Hijmans EM and Bernards R (1999) Functional interaction between a novel protein phosphatase 2A regulatory subunit, PR59, and the retinoblastomarelated p107 protein. Oncogene 18(2):515-524. Vossler MR, Yao H, York RD, Pan MG, Rim CS and Stork PJ (1997) cAMP activates MAP kinase and Elk-1 through a B-Raf- and Rap1-dependent pathway. Cell 89(1):73-82. Wiesmann C and de Vos AM (2001) Nerve growth factor: structure and function. Cell Mol Life Sci 58(5-6):748-759. Wolff RA, Dobrowsky RT, Bielawska A, Obeid LM and Hannun YA (1994) Role of ceramide-activated protein phosphatase in ceramide-mediated signal transduction. The Journal of biological chemistry 269(30):19605-19609. Xu Y, Xing Y, Chen Y, Chao Y, Lin Z, Fan E, Yu JW, Strack S, Jeffrey PD and Shi Y (2006) Structure of the protein phosphatase 2A holoenzyme. Cell 127(6):12391251. Yan Z, Fedorov SA, Mumby MC and Williams RS (2000) PR48, a novel regulatory subunit of protein phosphatase 2A, interacts with Cdc6 and modulates DNA replication in human cells. Molecular and cellular biology 20(3):1021-1029. Yao B, Zhang Y, Delikat S, Mathias S, Basu S and Kolesnick R (1995) Phosphorylation of Raf by ceramide-activated protein kinase. Nature 378(6554):307-310. Yoon S and Seger R (2006) The extracellular signal-regulated kinase: multiple substrates regulate diverse cellular functions. Growth factors (Chur, Switzerland) 24(1):2144.

100 Yoon SO, Soltoff SP and Chao MV (1997) A dominant role of the juxtamembrane region of the TrkA nerve growth factor receptor during neuronal cell differentiation. Journal of Biological Chemistry 272(37):23231-23238. York RD, Yao H, Dillon T, Ellig CL, Eckert SP, McCleskey EW and Stork PJ (1998) Rap1 mediates sustained MAP kinase activation induced by nerve growth factor. Nature 392(6676):622-626. Yu XX, Du X, Moreno CS, Green RE, Ogris E, Feng Q, Chou L, McQuoid MJ and Pallas DC (2001) Methylation of the protein phosphatase 2A catalytic subunit is essential for association of Balpha regulatory subunit but not SG2NA, striatin, or polyomavirus middle tumor antigen. Molecular biology of the cell 12(1):185-199. Zhang ZY (2003) Mechanistic studies on protein tyrosine phosphatases. Progress in nucleic acid research and molecular biology 73:171-220. Zhou J, Pham HT, Ruediger R and Walter G (2003) Characterization of the Aalpha and Abeta subunit isoforms of protein phosphatase 2A: differences in expression, subunit interaction, and evolution. The Biochemical journal 369(Pt 2):387-398. Zick Y (2003) Role of Ser/Thr kinases in the uncoupling of insulin signaling. Int J Obes Relat Metab Disord 27 Suppl 3:S56-60. Zwick E, Bange J and Ullrich A (2002) Receptor tyrosine kinases as targets for anticancer drugs. Trends in molecular medicine 8(1):17-23.