The Role of Myo1c Phosphorylation in

GLUT4 Translocation

by

Ming Fai Freddy Yip

B.Sc (Hons)

Garvan Institute of Medical Research

Sydney, Australia

A thesis submitted to the

School of Biotechnology and Biomolecular Sciences

University of New South Wales

Sydney, Australia

for the degree of

Doctor of Philosophy

March, 2009

I Table of Contents

Abstract……………………………………….……………………………………….i

Originality Statement…………………………..………..…………………………...ii

Acknowledgements…………………………..……………………………..……….iii

Publications……………………………..………………………..………..……....…v

List of Figures and Tables……………………………..…………………………..viii

Abbreviations……………………………..………………………………………....xi

Chapter 1 General Introduction……………………………….………………..1 i. Insulin action……………………………..………………………………….2 ii. GLUT4…………………..…………………………………………………..2 iii. Insulin signaling…………………..…………………………………………3 iv. GLUT4 trafficking…………………..………………………………………6 v. The key insulin-regulated steps of GLUT4 trafficking…………………….12 vi. Identification of Ser/Thr phosphoproteins…………………..……………..13 vii. 14-3-3 proteins…………………..…………………………………………15 viii. 14-3-3 proteins and insulin signaling…………………..…………………..16 ix. Conclusion…………………..……………………………………………...16

Chapter 2 General Materials and Methods…………………..……..………...23 i. Materials…………………..………………………………………………..24 ii. Methods…………………..………………………………………………...26 A. Cell Culture…………………..………………………………………26 B. Purification of 14-3-3 and preparation of 14-3-3-sepharose…...……27 C. Pulldown and immunoprecipitation……………………………….…30 D. PM silica preparation of plasma membrane………………….…...…30 E. SDS-PAGE gel staining and immunoblotting……….…………..……31 F. Immunoflorescence analysis…………………..….………………..…33 G. 2-Deoxyglucose uptake assay…………………..………….…………33 H. GLUT4 translocation assay…………………..……………...………34 I. Statistical analysis…………………..…………………………..……34

Chapter 3 Phosphoproteomic Analysis of Insulin Signaling Pathway in Adipocytes…………………………………...………………………35 i. Abstract…………………..………………………………………………...36 ii. Introduction…………………..…………………………………….………36 iii. Methods…………………..………………………………………...………38

II A. Mass spectrometry analysis…………………..………………………38 iv. Results…………………..……………………………………………….…40 v. Discussion…………………..…………………………………………...…61

Chapter 4 CaMKII-mediated Phosphorylation of the Myosin Motor Myo1c Is Required for Insulin-Stimulated GLUT4 Translocation in Adipocytes……………………………………………………...……69 i. Abstract…………………..…………………………………………...……70 ii. Introduction…………………..……………………………………….……70 iii. Methods…………………..…………………………………………...……74 A. In vitro and in vivo phosphorylation assay…………...…….…..……74 B. In vitro ATPase assay…………………..…………………….………74 iv. Results…………………..…………………………………………….……75 v. Discussion…………………..………………………………………….…109

Chapter 5 Identification of a Novel Myo1c-Interacting Protein, Armcx5, in Adipocytes…………………..……..……………...……………..…113 i. Abstract…………………..……………………………..…………………114 ii. Introduction…………………..………………………...…………………114 iii. Methods…………………..……………………………….………………116 A. Yeast culture…………………..…………………….….……………116 B. Preparation of yeast lysate………………….………………………117 C. Yeast two-hybrid library screening…………………..………...……117 D. Yeast two-hybrid interaction studies…………………..…….………118 E. Subcellular fractionation…………………..………………..………119 iv. Results…………………..…………………………………...……………119 v. Discussion…………………..…………………………………….………135

Chapter 6 General Discussion…………………..…………....……….………142

Chapter 7 References…………………..……………...…………..……..……155

III Abstract

Glucose is a primary and essential energy source for humans. It is broken down from complex carbohydrates in the diet and absorbed across the gut epithelium into the blood stream. Glucose homeostasis is important as hyperglycermia causes damage of pancreatic and peripheral cells. In response to a meal glucose is principally taken up by fat and muscle tissues and this response is activated by insulin release from pancreatic beta cells. Insulin stimulates the translocation of GLUT4 from the intracellular storage vesicles to the plasma membrane in fat and muscle cells.

Although many proteins have been implicated in this process, the key insulin-regulated substrate has not been determined yet. In the present study, the phosphoserine/threonine binding protein 14-3-3 was used as a tool to affinity-purify insulin-stimulated phosphoproteins from 3T3-L1 adipocytes. By using mass spectrometry 38 proteins were identified, reflecting the important role of 14-3-3 in mediating many insulin-regulated processes. Among the potential phosphoproteins was Myosin 1C (Myo1c), an actin-associated molecular motor, which has previously been implicated in insulin-stimulated GLUT4 trafficking in adipocytes. I showed that insulin stimulates the activation of CaMKII which phosphorylates Myo1c at S701 in a

Ca2+/PI3K-dependent manner. Myo1c phosphorylation induced its interaction with

14-3-3-proteins, reduced calmodulin-binding and stimulated its in vitro ATPase activity. Insulin-dependent stimulation of Myo1c phosphorylation and its ATPase activity were both required for GLUT4 translocation. By using yeast two-hybrid techniques, I identified a candidate ligand of the Myo1c tail, Armcx5, and demonstrated the in vivo interaction in 3T3-L1 adipocytes. The siRNA-mediated knockdown of Armcx5 inhibited insulin-stimulated glucose uptake and GLUT4 translocation. These results suggest that the regulation of Myo1c and its ligand

i Armcx5 are essential in insulin-regulated GLUT4 trafficking, possibly playing a key role in vesicle fusion.

ii Originality Statement

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of materials which have been accepted for the awards of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception or in style, presentation and linguistic expression is acknowledge.

Ming Fai Freddy YIP

31st March, 2009

iii Acknowledgements

My special thanks go to Prof. David James for giving me an opportunity to work with such an exciting project.

I would like to thank Dr Georg Ramm for his patience and guidance throughout these three years.

I would also like to thank A/Prof. Michael Guilhaus for giving me a chance to work with those world-class instruments at BMSF.

Many thanks go to Dr Mark Larance, Dr Kyle Hoehn and Dr Jacqueline Stoeckli for helping me so much in MS analysis, CaMKII study and yeast-two hybrid study.

I would like to acknowledge everyone in the James lab at the Garvan Institute of

Medical Research and in the Bioanalytical Mass Spectrometry Facility at UNSW who has given me such an enjoyable time.

To my family and friends, I would not be able to do it without your supports.

Last but not the least, thank God for what I have been given.

iv Publications

Part of the work in this thesis has been published or presented at meetings as follow.

Paper in refereed journal:

Yip, M.F., Ramm, G., Larance, M., Hoehn, K.L., Wagner, M.C., Guilhaus, M., and

James, D.E. (2008). CaMKII-mediated phosphorylation of the myosin motor Myo1c is required for insulin-stimulated GLUT4 translocation in adipocytes. Cell Metab 8,

384-398.

v

The front cover of Cell Metabolism, volume 8, issue 5, 2008 by Yip M.F.. This cover depicts the pathway as an electric circuit, with the Myo1c as a motor pulley that conveys GLUT4 vesicles to the plasma membrane. Insulin increases the Myo1c ATPase activity via PI3K/Ca2+/CaMKII-dependent phosphorylation. The background depicts adipocytes stained for Myo1c (green).

vi Conference abstracts:

The 8th Hunter Cellular Biology meeting, 2008

Poster presentation: CaMKII-mediated phosphorylation of the myosin motor Myo1c is required for insulin-stimulated GLUT4 translocation in adipocytes.

Yip, M.F., Ramm, G., Larance, M., Hoehn, K.L., Wagner, M.C., Guilhaus, M., and

James, D.E.

American Diabetes Association’s 68th Scientific Session, 2008

Oral presentation: CaMKII-mediated phosphorylation of the myosin motor Myo1c is required for insulin-stimulated GLUT4 translocation in adipocytes.

Yip, M.F., Ramm, G., Larance, M., Hoehn, K.L., Wagner, M.C., Guilhaus, M., and

James, D.E.

vii List of Figures and Tables

Figure 1.1 The insulin-PI3K pathway and the downstream effects of Akt in adipocytes. Figure 1.2 The schematic model of GLUT4 trafficking. Figure 1.3 Principle of TIRF-M analysis. Figure 1.4 The 14-3-3 proteins. Figure 3.1 Flow diagram of preparation of 14-3-3-sepharose. Figure 3.2 Purification of 14-3-3 and preparation of 14-3-3-sepharose. Figure 3.3 Confirming the binding activity of immobilized 14-3-3 by Western blot analyses of IRS-1 and pGSK-3. Figure 3.4 14-3-3-pulldown from 3T3-L1 adipocytes. Figure 3.5 Alkaline phosphotase treatment of blot and GST-14-3-3 Far Western blotting. Figure 3.6 Western blot analysis of 14-3-3-associated proteins using phosphor-Akt substrate (pAS) antibody. Figure 3.7 SDS-PAGE analysis and mass spectrometry identification of insulin-regulated 14-3-3-associated proteins. Figure 3.8 Subcellular localization of Myo1c in 3T3-L1 adipocytes. Figure 4.1 Mechanical cycle of myosin Figure 4.2 in vitro association of 14-3-3 and Myo1c in 3T3-L1 adipocytes. Figure 4.3 in vivo association of 14-3-3 and Myo1c in CHO/IR/IRS-1 cells. Figure 4.4 Direct and phosphorylation-dependent association of 14-3-3 and Myo1c. Figure 4.5 in silico analysis of Myo1c 14-3-3-binding and phosphorylation motifs. Figure 4.6 Mapping Myo1c phosphorylation sites. Figure 4.7 Effect of different kinase inhibitors on 14-3-3-binding of Myo1c. Figure 4.8 in vivo 32P labeling of Myo1c in 3T3-L1 adipocytes. Figure 4.9 Effect of Ca2+ chelator and ionophore on 14-3-3-binding of Myo1c. Figure 4.10 Effect of siRNA-mediated CaMKII knockdown on in vitro 14-3-3-binding of Myo1c.

viii Figure 4.11 in vitro CaMKII phosphorylation assay. Figure 4.12 Insulin stimulates CaMKII activation in 3T3-L1 adipocytes. Figure 4.13 Effect of CaMKII inhibitors on insulin-stimulated [3H]-2-deoxyglucose (2DOG) uptake in 3T3-L1 adipocytes. Figure 4.14 Domain structure of Myo1c. Figure 4.15 in vivo association of different Myo1c mutants and calmodulin in CHO/IR/IRS-1 cells. Figure 4.16 in vitro association of different Myo1c mutants and 14-3-3 in CHO/IR/IRS-1 cells. Figure 4.17 Effect of different kinase inhibitors on in vivo calmodulin-binding of Myo1c. Figure 4.18 Effect of in vitro CaMKII phosphorylation and 14-3-3-binding on ATPase activity of Myo1c. Figure 4.19 Effect of insulin-stimulation on ATPase activity of EYFP-Myo1c. Figure 4.20 Immunofluorescence analysis of surface HA in Myo1c knockdown cells. Figure 4.21 Immunofluorescence analysis of surface HA in Myo1c re-expressing cells. Figure 5.1 Principle of the yeast two-hybrid assay. Figure 5.2 Yeast two-hybrid screening of 3T3-L1 adipocytes cDNA library using Myo1c tail domain as the bait. Figure 5.3 Interaction study of Armcx5 and Myo1c or Myo1b. Figure 5.4 Mapping Myo1c-binding domain of Armcx5. Figure 5.5 in vivo association of Armcx5 and Myo1c truncated mutants in CHO/IR/IRS-1 cells. Figure 5.6 Western blot analyses of subcellular fractions of 3T3-L1 adipocytes. Figure 5.7 in vivo association of endogenous Armcx5 and Myo1c in 3T3-L1 adipocytes. Figure 5.8 Effect of shRNA-mediated knockdown of Armcx5 on 3T3-L1 adipocytes. Figure 5.9 in silico analysis of Armcx5. Figure 6.1 Effects of wortmannin on activation of CaMKII and phosphorylation

ix of Myo1c. Figure 6.2 Multiple roles of Myo1c in adipocytes. Table 3.1 The 14-3-3-affinity purified proteins from insulin-treated 3T3-L1 adipocytes. Table 3.2 The 14-3-3-isoforms purified by 14-3-3-affinity purification. Table 5.1 A list of proteins identified in the yeast two-hybrid screen.

x Abbreviations

2DE Two-dimensional electrophoresis ADP Adenosine diphosphate Alk Phos Alkaline phosphatase Arm Armadillo Armcx5 Armadillo repeat containing, X-linked 5 AS160 Akt substrate of 160 kDa ATP Adenosine triphosphate BCA Bicinchoninic acid BSA Bovine serum albumin °C degree Celsius CaMKII Ca2+/Calmodulin-dependent kinase II CHO Chinese hamster ovary CMFS Calcium- and magnesium-free saline kDa kilo-Dalton DMEM Dulbecco’s Modified Eagle solution DTT Dithiothreitol EDTA Ethylenediaminetetraacetate ER Endoplasmic reticulum FCS Foetal calf serum FOXO Forkhead transcription factor family g gravity mg milli-gram μg micro-gram GFP Green fluorescent protein GLUT4 Glucose transporter 4 GSH Glutathione GSK-3 Glycogen synthase kinase 3 GST Glutathione S-Transferase GSV GLUT4 storage vesicle h hour(s) HRP Horseradish peroxidase IgG Immuno-γ-Globulin IP Immunoprecipitation IPTG Isopropylthio-β-D-galactoside IR Insulin receptor IRS-1 Insulin receptor substrate 1 K Lysine

xi KSR Kinase suppressor of Ras L Leucine ml milli-liter μl micro-liter nl nano-liter m minute(s) mM milli-molar μM micro-molar nM nano-molar MEK MAPK/ERK kinase MS Mass spectrometry M.W. Molecular weight MudPIT Multidimensional protein identification technology Myo1c Myosin 1C NCS Newborn calf serum ml milli-litre μl micro-litre LB Luria Broth LC Liquid chromatography OD Optical density P Proline pAkt Phospho-Akt pAS Phospho-Akt substrate PBS Phosphate buffered saline PD Pulldown PDK1 Phosphoinositide-dependent kinase 1 pGSK-3 phospho-GSK-3 PI3K Phosphatidylinositide 3-kinase

PIP2 Phosphatidylinositol (4,5)-bisphosphate

PIP3 Phosphatidylinositol (3,4,5)-trisphosphate PKA Protein kinase A PKC Protein kinase C PM Plasma membrane PMSF Phenylmethanesulfonyl fluoride pSer Phosphoserine pThr Phosphothreonine pTyr Phosphotyrosine PTB Phospho-tyrosine binding

xii PVDF Polyvinylidene fluoride R Arginine rpm resolutions per minute S Serine SB Sample buffer SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis Ser Serine shRNA Small hairpin RNA siRNA Small interfering RNA T Threonine TBS Tris buffered saline Thr Threonine Tyr Tyrosine U Unit(s) UV Ultra-violet v volume V Volt WB Western blot X any amino acids YFP Yellow fluorescent protein

xiii Chapter 1

General Introduction

1 i. Insulin action

The regulation of glucose uptake, utilization and storage by tissues is critical to sustain blood glucose homeostasis. The blood glucose level is maintained even during long term starvation, and elevated blood glucose level as a result of carbohydrate ingestion is rapidly returned to normal within a narrow range of 5 – 6 mM (Lanner et al., 2008). This metabolic control mechanism is important since hypoglycemia may lead to loss of consciousness while hyperglycermia may result in toxicity in peripheral tissues. Insulin is an anabolic hormone that regulates cellular growth, development and whole-body metabolism (Le Roith and Zick, 2001). It is produced by the beta cells of the pancreas in response to elevated blood glucose to maintain euglycemia. It achieves this by promoting glucose uptake by muscle and fat cells and inhibiting glucose release from the liver (Evans et al., 2004; Flatt, 1995). Type II diabetes is associated with defective insulin action, or insulin resistance, in these tissues ultimately leading to defective pancreatic insulin secretion and hence dysregulation in blood glucose homeostasis (Bryant et al., 2002; Taylor, 1999). ii. GLUT4

The insulin-stimulated uptake of glucose into muscle and fat is the major cellular mechanism for disposal of blood glucose, which is then stored as glycogen or lipid and oxidized to produce energy. Glucose uptake is facilitated by a glucose transporter protein, GLUT4, which is one of the 13 sugar transporter proteins in the human genome (Gould and Holman, 1993; Joost and Thorens, 2001). GLUT4 is expressed in insulin-responsive tissues including adipose tissues, skeletal muscle and cardiac muscle (James et al., 1988). Heterozygous GLUT4+/- mice display peripheral insulin resistance while re-expression of GLUT4 in skeletal muscle of GLUT4+/- mice improves insulin sensitiviy and glucose tolerance (Rossetti et al., 1997; Tsao et al.,

2 1999). The glucose transporter GLUT4 is thus a key regulator of the whole-body glucose homeostasis.

GLUT4 is a type IIIa membrane protein with 12 transmembrane domains. It contains a unique N-terminal cytoplasmic domain with a critical phenylalanine residue, as well as dileucine and acidic motifs in the C-terminus which together determine the kinetics of exocytosis and endocytosis in a regulated trafficking system (Al-Hasani et al., 2002;

Araki et al., 1996; Huang and Czech, 2007). The uptake of glucose into fat or muscle cells in response to insulin is facilitated by translocation of GLUT4-containing vesicles from the cytosol to the plasma membrane (PM) (Bryant et al., 2002; Larance et al., 2008). In the basal state, the GLUT4 storage vesicles (GSVs) are preferentially located in the perinuclear region. Upon stimulation, these vesicles move to and fuse with the PM. Once the GLUT4 is inserted into the membrane, glucose can be transferred across the membrane down its concentration gradient (Antonescu et al.,

2005). In the case of skeletal muscle, contraction or exercise also stimulates translocation of GLUT4 to the PM (Rose and Richter, 2005). iii. Insulin signaling

The Insulin receptor

The insulin signaling pathway starts by activation of the insulin receptor (IR) (Figure

1.1). IR is a tyrosine kinase and its activation by insulin leads to the activation of the mitogen-activated protein kinase (MAPK) and the phosphoinositide 3-kinase (PI3K) pathways (Kasuga et al., 1982). The IR is a heterotetrameric transmembrane protein consisting of two extracellular insulin-binding α subunits and two transmembrane tyrosine kinase β subunits (Ebina et al., 1985). Insulin binding to the extracellular α subunits activates the tyrosine kinase activity of the β subunits, resulting in auto-phosphorylation at other tyrosine residues in the juxtamembrane regions and

3 intracellular tail (Saltiel and Pessin, 2003; Watson et al., 2004a). The activated IR recruits and phosphorylates other substrate proteins including Shc, Gab-1 and the family of insulin receptor substrate (IRS) proteins (White, 2003; Zick, 2001) (Figure

1.1).

The PI3K/Akt signaling pathway

In adipocytes, metabolic actions regulated by insulin such as glucose uptake and lipid synthesis are mediated by the Phosphatidylinositide 3-kinase (PI3K) pathway (Zick,

2001). Inhibition of PI3K with inhibitors such as wortmannin completely blocks insulin-stimulated glucose uptake (Okada et al., 1994). This pathway begins with

IRS-1 which is one of the major substrates of the IR (Sun et al., 1991) (Fig 1.1).

IRS-1 is widely expressed in most cells and tissues including adipocytes, and it is responsible for controlling body growth and peripheral insulin action (Rothenberg et al., 1991; Sun et al., 1992). IRS-1 transiently binds to auto-phosphorylated IR via a phosphotyrosine binding (PTB) domain. IRS-1 is phosphorylated by IR at several tyrosine residues, creating binding sites for downstream effectors such as PI3K

(Virkamaki et al., 1999; White, 1998). Recruitment of PI3K by IRS-1 leads to the localization of PI3K in close proximity to its major substrate phosphatidylinositol

(4,5)-bisphosphate in the inner leaflet of the plasma membrane (White, 2003), thereby generating phosphatidylinositol (3,4,5)-trisphosphate (PIP3) (Cantley, 2002). PIP3 serves as a docking site for two Ser/Thr kinases, PDK1 and Akt. Co-localization of

PDK1 and Akt at the plasma membrane allows phosphorylation of Akt at Thr308

(Alessi et al., 1997). In addition, the mTOR/Rictor complex which is also known as

PDK2 phosphorylates Akt at Ser473 (Sarbassov et al., 2005). These phosphorylation events in turn lead to the activation of Akt and permits the phosphorylation of many other downstream effectors involved in diverse processes (White, 2003).

4 Among the three different isoforms of Akt, only the Akt2 isoform has been shown to be required for insulin-stimulated GLUT4 translocation in skeletal muscle via siRNA-mediated knockdown of specific Akt isoforms (Bouzakri et al., 2006).

Moreover, recent study using a rapid drug-induced heterodimerization system showed that activation of Akt2 is sufficient to stimulate GLUT4 translocation in adipocytes

(Ng et al., 2008). These data indicate that the PI3K/Akt pathway is the key signal that mediates insulin-stimulated GLUT4 translocation in muscle and fat cells.

The Akt substrates

Being an important node in the insulin signaling pathway, Akt can regulate many downstream targets via Ser/Thr phosphorylation (Figure 1.1). Analysis of Akt substrate phosphorylation sites in these direct targets has revealed a minimal consensus site consisting of R/KXR/KXXS/T (Scheid and Woodgett, 2001). Akt promotes glycogen synthesis via covalent regulation of glycogen synthase kinase 3β

(GSK3β). In addition, Akt inhibits the GAP activity of TSC1-TSC2, a Rheb-GAP, involved in the regulation of protein synthesis (McManus and Alessi, 2002). Akt also supresses apoptosis and gene transcription via cytosolic localization of BAD and

Forkhead-related transcription factor 1 (FKHR-L1/FOXO1) (Cantley, 2002; Fresno

Vara et al., 2004; McManus and Alessi, 2002; White, 2003; Zeigerer et al., 2004).

One of the direct substrates of Akt is the Akt substrate of 160 kDa (AS160).

Phosphorylation of AS160 by Akt in response to insulin is believed to induce the translocation of GLUT4 from intracellular compartments to the PM (Larance et al.,

2005; Sano et al., 2003; Zeigerer et al., 2004). AS160 is a Rab GTPase-activating protein (GAP) that regulates the GTP/GDP cycle of Rab GTPases (Deneka et al.,

2003). AS160 is thought to associate with a Rab protein involved in GLUT4 translocation. When AS160 is dephosphorylated, it is active and keeps the Rab in an

5 inactive GDP-bound form. Following activation of Akt via insulin action, Akt phosphorylates AS160, presumably inhibiting its GAP activity and resulting in the activation of a Rab protein required for GLUT4 translocation to the plasma membrane

(Sano et al., 2003; Zeigerer et al., 2004). Expression of an AS160 mutant, in which all four insulin responsive phosphorylation sites were mutated to alanine, in adipocytes blocks insulin stimulated GLUT4 translocation (Sano et al., 2003; Zeigerer et al.,

2004). Although the exact role of AS160 is still unclear, it can be viewed as a brake that is removed via phosphorylation by Akt. iv. GLUT4 trafficking

The nature of GSVs

The evidence for the presence of specialized GLUT4 storage compartments was shown by early studies using electron microscopy, which revealed GLUT4 to be primarily localized in membrane vesicles with diameter of about 50 – 100 nM in basal adipocytes and muscle cells (Hahn and Karchner, 1995; Rodnick et al., 1992; Slot et al., 1991a; Slot et al., 1991b). Approximately 50% of the entire GLUT4 pool is translocated to the PM in response to a maxiumum insulin stimulation (Pilch, 2008).

The insulin-responsive translocated pool is composed of GSVs which are specialized compartments that can only be found in adipocytes, skeletal and cardiac muscle cells

(Ishiki and Klip, 2005).

The nature of GSVs has been widely studied by using mass spectrometry to identify proteins that associate with GSVs (Guilherme et al., 2000; Larance et al., 2005). The three major protein components of GSVs other than GLUT4 include insulin-reponsive aminopeptidase (IRAP) (Kandror et al., 1994), vesicle associated membrane proetin 2

(VAMP2) (Cain et al., 1992) and sortilin (Lin et al., 1997). IRAP has been shown to translocate to the PM in response to insulin and its trafficking kinetics is very similar

6 to GLUT4 (Garza and Birnbaum, 2000; Keller et al., 1995). Although the physiological role of IRAP still remains unclear, siRNA knockdown of IRAP results in decreased GLUT4 expression and translocation in adipocytes (Keller et al., 2002;

Yeh et al., 2007). IRAP has also been shown to interact with AS160 (Peck et al., 2006;

Ramm et al., 2006). Another major component of GSVs is VAMP2. It is a v-SNARE that mediates fusion events of GSVs and synaptic vesicles with the PM via the interaction with t-SNAREs (Grusovin and Macaulay, 2003; Pevsner et al., 1994).

Botulinum toxin-mediated cleavage of VAMP2 completely abolishes GSV fusion with the PM (Macaulay et al., 1997; Tamori et al., 1996). Sortilin, similar to GLUT4, is dramatically up-regulated during differentiation of adipocytes (Lin et al., 1997;

Morris et al., 1998). Knockdown of sortilin decreases both the amount of GSVs and insulin-regulated glucose uptake in adipocytes (Shi and Kandror, 2005). The copy number for GLUT4, IRAP and sortilin per GSV is thought to be similar and range between 6 – 12 (Pilch, 2008). Recent studies showed that there are approximately 70 copies of VAMP2 per synaptic vesicle and a similar amount is believed to be present in GSVs (Takamori et al., 2006).

Another important component of GSVs is Rab proteins, which is critical to vesicle trafficking (Grosshans et al., 2006). Rabs are considered to be molecular switches linking signaling cascades to molecular effectors (Zaid et al., 2008). Rabs 2, 4, 8, 10,

11 and 14 have been detected in GLUT4-enriched membrane compartments (Cormont et al., 1993; Larance et al., 2005; Miinea et al., 2005), and Rabs 4, 5, 11 and 31 have been implicated in GLUT4 trafficking (Ishikura et al., 2008; Kaddai et al., 2008;

Watson et al., 2004a). In addition, the GAP domain of AS160 displays in vitro GAP activity towards some of these Rabs including Rabs 2A, 8A, 10 and 14 (Miinea et al.,

2005). In particular, overexpression of Rab8A or Rab14 in muscle cells is able to

7 rescue the inhibition of GLUT4 translocation due to expression of the inhibitory phosphorylation mutant (4P) of AS160 (Ishikura et al., 2007). In another study, knockdown of Rab10 attenuates GLUT4 translocation to the PM in adipocytes and inhibits the increase in basal translocation due to AS160 knockdown, indicating that

Rab10 is a target of AS160 RabGAP (Sano et al., 2007). It will be important in the future to clarify the role of AS160 and its corresponding Rabs in the specific steps of

GSV trafficking.

The trafficking of GSVs can be dissected into multiple steps including budding of

GSVs from the trans-Golgi network (TGN) or endosomes, transport of the GSVs along the cytoskeletal proteins to the cell peripheral region, interaction of the GSVs with the PM (tethering), assembly of the soluble N-ethylmalemide-sensitive factor attachment receptor (SNARE) complex (docking), and fusion of the GSVs with the

PM (Larance et al., 2008) (Figure 1.2).

Biogenesis of GSVs

GSVs are believed to form either from endosomes or TGN or both, and adaptor proteins play an important role in sorting GLUT4 and IRAP into GSVs (Larance et al.,

2008; McNiven and Thompson, 2006). The clathrin adaptor AP1 and ARF GTPases have been shown to associate with GLUT4 vesicles (Gillingham et al., 1999). Yeast two-hybrid analyses revealed that the medium chain adaptins mu1, mu2 and mu3A directly interact with the N-terminal FQQI motif which may play a role in targeting

GLUT4 to clathrin-coated vesicles (Al-Hasani et al., 2002). In addition, GGA (the

Golgi-localized, gamma-ear-containing, Arf-binding protein) is also involved in sorting newly synthesized GLUT4 and IRAP from the TGN into GSVs (Hou et al.,

2006; Watson et al., 2004b). Instead of interacting with GLUT4 directly, GGA has been shown to interact with sortilin, and the cation-dependent and -independent

8 mannose-6-phosphate receptor (Li and Kandror, 2005). Expression of a dominant negative mutant of GGA in adipocytes inhibits formation of GSVs as well as insulin-stimulated glucose uptake (Li and Kandror, 2005). Knockdown of sortilin also results in similar effects while overexpression of sortilin increases the amount of

GSVs and insulin-regulated glucose uptake (Shi and Kandror, 2005). The same study also showed that double transfection of sortilin and GLUT4 into preadipocytes leads to formation of functional GSVs (Shi and Kandror, 2005). More recently, the luminal

Vps10p domain of sortilin has been shown to interact directly with the luminal domains of GLUT4 and IRAP (Shi and Kandror, 2007). These data suggest an essential role of sortilin in biogenesis of GSVs and acquisition of insulin responsiveness.

Cytoskeletal Transport

Microtubules have been implicated in the transport of GSVs to the cell periphery though the evidence is controversial. Serveral microtubule-depolymerizing agents have been shown to abolish insulin-stimulated GLUT4 translocation to the PM and glucose uptake (Fletcher et al., 2000; Huang et al., 2005; Liu et al., 2003). In contrast other studies using lower concentration of nocodazole that is sufficient to disrupt microtubules showed no effects, suggesting GLUT4 translocation does not require microtubule integrity (Ai et al., 2003; Molero et al., 2001). In fact, microtubule-based kinesin motors including KIF3 and KIF5B have been shown to play a role in movement of GLUT4 vesicles (Emoto et al., 2001; Imamura et al., 2003; Semiz et al.,

2003). In addition, expression of the microtubule-binding protein hTau40 which reduces kinesin attachment to microtubules delays the appearance of GLUT4 at the plasma membrane in response to insulin (Emoto et al., 2001). These studies indicated that kinesin motors allow long-range trafficking of vesicles along microtubules to the

9 cell periphery.

Several studies suggested that insulin promotes the formation of cortical actin which enhances GLUT4 translocation to the cell surface in adipocytes and muscle cells

(Kanzaki and Pessin, 2001; Khayat et al., 2000; Patki et al., 2001).

Actin-depolymerization agents Latrunculin A and Latrunculin B have been shown to disrupt cortical actin filaments and inhibit GLUT4 translocation in rat adipocytes and muscle cells (Brozinick et al., 2004; Omata et al., 2000). Expression of a dominant negative N-WASP mutant blocks cortical F-actin rearrangement and insulin action on

GLUT4 translocation (Jiang et al., 2002). More recently, the actin-based motor Myo1c has been implicated in GLUT4 translocation, playing a role in cortical actin remodelling and membrane ruffling (Bose et al., 2002; Bose et al., 2004). Another myosin MyoVa was identified as a substrate of Akt2. Insulin-induced phosphorylation appears to enhance its association with actin filaments and in turn facilitates movement of GSVs along actin the the cell surface (Yoshizaki et al., 2007). However these studies do not differentiate between a role of actin as cables for processive vesicular movement, versus its role as a cortical filamentous mesh serving to position vesicles for docking and fusion. In fact, α-actinin-4 (ACTN4) was identified as a

GLUT4 interacting protein which links GSVs to cortical F-actin (Foster et al., 2006).

Knockdown of ACTN-4 abolished the gain in surface-exposed GLUT4 in response to insulin, indicating a role of ACTN-4 and cortical actin as a tether of GSVs

(Talior-Volodarsky et al., 2008). Taken together, these studies on the cytoskeleton revealed the requirement of microtubule for long-range trafficking to the periphery, potentially handing on vesicles to cortical actin for a late stage short-range movement or vesicle priming at the PM (Zaid et al., 2008).

Tethering

10 Tethering is the step where the vesicle first encounters the target membrane and it is thought to allow a degree of regulation before the vesicle commits to docking and fusion (Larance et al., 2008). As discussed before, ACTN-4 and cortical actin may play a role in tethering (Talior-Volodarsky et al., 2008). The yeast exocyst, a complex of proteins Sec3p, Sec5p, Sec6p, Sec8p, Sec10p, Sec15p, Exo70p and Exo84p, is a tethering mchainery that is conserved in mammalian cells and is involved in GLUT4 trafficking (Ewart et al., 2005; Inoue et al., 2003; TerBush et al., 1996). Exo70 interacts with all the other subunits of the exocyst. Exo70 is translocated to the PM upon insulin stimulation of adipocytes through an interaction with active TC10 (Inoue et al., 2003; Matern et al., 2001; Vega and Hsu, 2001). TC10, a Rho GTPase, is localized to lipid rafts at the PM and is activated by GTP-loading in response to insulin (Inoue et al., 2006; Saltiel and Pessin, 2003). Crystallographic studies indicated that four of the eight subunits of the exocyst contain long and rod-like domains, thereby bringing the GSV at relatively long distance from the PM into close proximity to allow SNARE complex formation (Munson and Novick, 2006).

Docking and fusion

The SNARE complex is thought to mediate both docking and fusion of GSVs at the

PM. It is composed of four parallel α-helices formed from coiled-coil domains of the

SNARE proteins located in two opposite membranes (Sollner et al., 1993). The v-SNARE VAMP2 at the GSV interacts with the t-SNAREs Syntaxin-4 and SNAP-23 at the PM (Foster and Klip, 2000; Kawanishi et al., 2000). Other accessory proteins are also required in this process. NSF (N-ethylmaleimide-sensitive factor), which associates with GSVs, is required to open up the cis-SNARE complex on the vesicle to allow trans-SNARE formation (Jahn and Scheller, 2006; Mastick and Falick, 1997).

Moreover, Munc18c and Synip are implicated in the regulation of SNARE complex

11 formation. Munc18c is required to keep syntaxin-4 in an inactive conformation to prevent VAMP2/Syntaxin-4 interaction in the basal state (Thurmond et al., 1998).

Syntaxin-4 is thought to switch to an alternate conformation with insulin exposing a

VAMP2 binding site, allowing SNARE complex formation (Smithers et al., 2008).

Synip has been shown to interact with Syntaxin-4 using a yeast two-hybrid approach

(Min et al., 1999). Similar to Munc18c, synip is thought to prevent non-specific

VAMP2/Syntaxin-4 interaction in the basal state. Upon insulin stimulation, Akt was shown to phosphorylate Synip at Ser99, causing it to dissociate from Syntaxin-4

(Okada et al., 2007; Yamada et al., 2005). However, the relevance of this phosphorylation event to GLUT4 translocation has been controversial. Other studies have demonstrated that the expression of a Synip mutant S99A has no effects on insulin-stimulated GLUT4 translocation in adipocytes (Sano et al., 2005). v. The key insulin-regulated steps of GLUT4 trafficking

More than 60 proteins have been implicated in GLUT4 trafficking and many more are expected to be discovered. While many of them are passive participants in the whole process, at least some are believed to be important insulin-regulated drivers. What is/are the key insulin-regulated step(s) of the GLUT4 trafficking? Early studies pointed to the early stage of GLUT4 trafficking. Using an in vitro cell-free reconstitution assay evidence was provided to show that insulin stimulates the formation of GSVs (Xu and Kandror, 2002). Insulin was also shown to release the tethers, TUG, from GSVs allowing their movement towards the cell periphery (Bogan et al., 2003). The identification of myosin motors, Myo1c and MyoVa, involved in

GLUT4 translocation suggested that the movement of GSVs may also be regulated by insulin. MyoVa was of particular interest since it was shown to be a substrate of Akt2

(Bose et al., 2002; Yoshizaki et al., 2007). Recently attention has focused on the later

12 stages of GLUT4 trafficking. Studies using an in vitro fusion assay indicated that the rate of GSV fusion is significantly stimulated by insulin and is Akt-dependent

(Koumanov et al., 2005). This observation was supported by other studies that utilized total internal reflection fluorescence microscopy (TIRFM) to visualize the movement of GSVs at the PM (Figure 1.3). Huang et al. showed that insulin decreases the tethering/docking duration and modulates the fusion kinetics, resulting in a 4-fold increase in the fusion frequency (Huang et al., 2007). Studies by Lizunov et al. suggested that GSVs loosely tether to the PM in the absence of insulin and the primary mechanism of insulin action is to stimulate tethering and fusion vesicles to specific fusion sites at the PM (Lizunov et al., 2005). Bai et al. showed that preparation of vesicles for fusion after docking at the PM is the key insulin-regulated step (Bai et al., 2007). They further suggested that Akt and its downstream substrate

AS160 regulate the docking of vesicles, and the fusion of vesicles is regulated by an alternative substrate or pathway. Gonzalez et al. suggested that the fusion step is

PI3K-dependent but not Akt-dependent (Gonzalez and McGraw, 2006). Since the PM is likely the major insulin-regulatory node for GLUT4 translocation, further studies looking at insulin-regulated substrates at the PM will be important for identifying the key factors involved in GLUT4 translocation. vi. Identification of Ser/Thr phosphoproteins

Mass spectrometry has been widely used to identify Ser/Thr phosphoproteins. The most commonly used mass spectrometry-based Ser/Thr phosphoproteomic approaches requires prior purification of the relevant phosphoproteins/phosphopeptides. An effective way to selectively enrich for Ser/Thr phosphoproteins is immunoprecipitation with phospho-specific antibodies. Depending on the affinity and specificity of the antibodies, these antibodies can be powerful tools

13 to enrich for low abundance phosphorylated proteins prior to 1DE or 2DE (Blagoev et al., 2003). Immunoprecipitation with anti-pTyr antibodies combined with either 1DE or 2DE has been shown to be a sensitive tool for identification of Tyr phosphoproteins

(Imam-Sghiouar et al., 2002; Vercoutter-Edouart et al., 2000). However, since the specificity and affinity of the pSer and pThr antibodies are generally regarded as insufficient, immunoprecipitation with these antibodies has not been widely used

(Machida et al., 2003). Recently, Gronborg et al. (2002) demonstrated that an anti-pSer/Thr PKA substrate antibody was capable of enriching pSer/Thr-containing proteins, leading to the identification of a novel PKA regulated signaling molecule using mass spectrometry (Gronborg et al., 2002). In addition, anti-pSer/Thr Akt substrate antibody (anti-pAS) was also previously used for immunoprecipitation and this led to identification of a novel insulin-regulated Ser/Thr phosphoprotein (Kane et al., 2002). Although these kinase-substrate antibodies that specifically detect individual kinase substrates have shown to be promising tools for identification of phosphoproteins, the generation of sufficient phospho-specific antibodies for large-scale immunoprecipitation coupled with mass spectrometry identification involves significant labor and cost.

As an alternative to anti-pSer/Thr antibodies, pSer/Thr binding proteins can also be used to purify phosphoproteins from complex mixtures. This non-antibody-based affinity purification approach takes advantage of the fact that many Ser/Thr phosphoproteins serve as high-affinity binding sites for protein interaction modules, such as 14-3-3, WW, forkhead-associated, WD40 and LRR domains (Yaffe and Elia,

2001). These powerful tools are comparable to antibodies and are more cost-effective for large-scale phosphoprotein purification. As far as insulin action and Akt signaling are concerned, one of the most valuable binding domains is 14-3-3 proteins because

14 they have been shown to bind to a number of Akt substrates such as GSK-3, AS160 and FOXO (Wilker and Yaffe, 2004). vii. 14-3-3 proteins

In 1967, 14-3-3 proteins were first identified as abundant, acidic brain proteins by

Moore and Perez (Moore and Perez, 1967). The name is derived from the fraction number on DEAE-cellulose chromatography and the migration position after subsequent starch gel electrophoresis (Fu et al., 2000). 14-3-3 proteins are highly conserved and are ubiquitously expressed in all eukaryotic cells. There are at least 7 mammalian isoforms (β, γ, ε, σ, ζ, τ and η), 15 plant isoforms and 2 yeast isoforms of

14-3-3 (Fu et al., 2000; Rosenquist et al., 2001; Wang and Shakes, 1996). 14-3-3 proteins exist as homo- and heterodimers with a monomeric mass of 30 kDa. Aitken

(Aitken, 2002) has shown that different 14-3-3 dimers bind to different targets or the same targets with different affinities. By screening peptide libraries, two high-affinity phosphorylation-dependent binding motifs have been defined: RSXpS/TXP and

RXXXpS/TXP where pS and pT represent phosphoserine and phosphothreonine respectively (Figure 1.4) (Rittinger et al., 1999; Yaffe et al., 1997). However, an increasing number of other Ser/Thr phosphorylation-dependent motifs including those in the IGF-1 receptor and p53 have been described (Craparo et al., 1997; Waterman et al., 1998). Moreover, it has been reported that some protein interactions with 14-3-3 are phosphorylation-independent, including the interactions with exoenzyme S and

RhoGEF (Hallberg, 2002; Masters et al., 1999; Zhai et al., 2001).

It is believed that 14-3-3-binding to a particular target functions to inhibit or mediate protein interactions, prevent dephosphorylation, promote protein stability, modulate enzymatic activity and alter localization (Dougherty and Morrison, 2004). Targets for

14-3-3 can be found in all subcellular compartments and they include transcription

15 factors, biosynthetic enzymes, signaling molecules, cytoskeletal proteins, apoptosis factors and tumor suppressors. These targets reflect the important roles of 14-3-3 in the regulation of many crucial biological processes, namely transcription, metabolism, signal transduction, protein trafficking, cell cycle control, apoptosis and malignant transformation (Dougherty and Morrison, 2004). viii. 14-3-3 proteins and insulin signaling

The two high-affinity phosphorylation-dependent 14-3-3-binding motifs, RSXpS/TXP and RXXXpS/TXP, are similar to the basophilic kinase phosphorylation consensus motifs including that of Akt, which is R/KXR/KXXS/T (Figure 1.4B) (Benzinger et al., 2005). Upon insulin stimulation, 14-3-3 binds to GSK-3, FOXO and AS160, which are substrates of Akt (Wilker and Yaffe, 2004). Phosphorylation of mouse

GSK3β at Ser 9 by Akt inhibits the kinase catalytic activity. 14-3-3 binds to this site and prevents its dephosphorylation, thus playing an important role in insulin regulation of glycogen synthesis (Wilker and Yaffe, 2004). 14-3-3-binding also plays an important role in the insulin regulation of gene expression. Members of the FOXO transcription factor family are active in the nucleus. Akt-dependent phosphorylation of

FOXO results in binding of 14-3-3 and the FOXO/14-3-3 complex is excluded from the nucleus, inhibiting a specific set of gene expression (Brunet et al., 1999).

Additionally, 14-3-3 has been implicated in insulin-stimulated GLUT4 translocation.

14-3-3 binds to AS160 at Thr642 as a result of Akt phosphorylation in response to insulin. While expression of the inhibitory phosphorylation mutant (4P) of AS160 abolishes insulin-stimulated GLUT4 translocation, introduction of a constitutive

14-3-3-binding site on AS160 overcomes this blockage (Ramm et al., 2006). These studies clearly demonstrated the important role of 14-3-3 in insulin signaling pathway. ix. Conclusion

16 Type II diabetes, which is characterized by insulin resistance in peripheral tissues and/or abnormal insulin secretion, is clearly one of the main threats to human health in the 21st century (Zimmet et al., 2001). Defining the key insulin dependent steps in fat and muscle cells will not only facilitate a better understanding of the physiological role of insulin, but also help to define new therapeutic approaches for the treatment of this disease. Although the translocation of the glucose transporter GLUT4 to the PM in fat and muscle cells was known to be the primary mechanism for clearance of exogenous glucose load since 1980’s, the underlying molecular mechanism has not been fully resolved yet. The advent of emerging biochemical assays and imaging technology has pinpointed the PM as a major node for insulin regulation of GLUT4 translocation. Hence, identifying the key factors that act principally at the PM becomes one of the important goals in the field of GLUT4 biology. In combination of mass spectrometry, the diverse range of binding partners and the physiological relevance of 14-3-3 proteins make it a useful tool to study the insulin-regulated phosphoproteome.

17

Figure 1.1 The insulin-PI3K pathway and the downstream effects of Akt in adipocytes. When insulin binds to the insulin receptor (IR), IRS-1 transiently binds to auto-phosphorylated IR via a phosphotyrosine binding (PTB) domain. IRS-1 is phosphorylated by IR at several tyrosine residues, creating binding sites for downstream effectors, including PI3K (Virkamaki et al., 1999; White, 1998).

Recruitment of PI3K leads to localization of PI3K in close proximity to its major substrate PIP2 in the

inner leaflet of the plasma membrane (White, 2003), thereby generating PIP3 (Cantley, 2002). PIP3 serves as a docking site for PDK1 and Akt. Co-localization of PDK1 and Akt at the plasma membrane allows phosphorylation of Akt by PDK1 (White, 2003). The phosphorylation of Akt activates the

Ser/Thr kinase activity of Akt so that it can phosphorylate many other downstream effectors involved in diverse processes, including stimulation of GLUT4 vesicle translocation from intracellular pools to the plasma membrane (AS160), glycogen synthesis (GSK-3β), protein synthesis (TSC1-TSC2) and

18 gene transcription (FOXO) (Cantley, 2002; Fresno Vara et al., 2004; McManus and Alessi, 2002;

White, 2003; Zeigerer et al., 2004).

AS160, Akt substrate 160; FOXO, forkhead-related transcription factor; GSK3, glycogen synthase kinase 3; IRS-1, insulin receptor substrate 1; PDK-1, phosphotidylinositide-dependent protein kinase 1;

PI3K, phosphotidylinositide 3-kinase; PIP2, phosphotidylinositol (4,5)-bisphosphate; PIP3, phosphotidylinositol (3,4,5)-trisphosphate.

19

Figure 1.2 The schematic model of GLUT4 trafficking. The trafficking of GSVs can be dissected into multiple steps: formation of GSVs from the endosomes or TGN, transport of the GSVs along the microtubule and/or actin to the cell peripheral region, initial tethering of the GSVs with the PM mediated by the tethering complex, SNARE-mediated docking of the GSVs at the PM, and fusion of the GSVs with the PM

20

Figure 1.3 Principle of TIRF-M analysis. The laser illumination at incident angles is greater than the critical angle, resulting in TIRF that creates an evanescent wave immediately adjacent to the coverglass-specimen interface. This evanescent wave excites single molecules in the thin section (~100 nm) in contact with the coverglass. This allows imaging of vesicles approaching to the PM, docking and fusing with the PM.

21

Figure 1.4 14-3-3 proteins. A. The structure of 14-3-3 bound to the phosphoserine peptide RLYHpSLP

(stick representation shown with carbons colored yellow, nitrogens blue, oxygens red, and phosphate green) (Rittinger et al., 1999). B. The two consensus binding motifs for 14-3-3 proteins and the phosphorylation substrate motifs for Akt and CaMKII are shown. Each motif is shown with X being any amino acid and with amino acids in each column being those most commonly found in each position. The red serine/threonine residue denotes the site of phosphorylation.

22 Chapter 2

General Materials and Methods

23 i. Materials

The pEYFP-Myo1c plasmid was a gift from Michael Czech (University of

Massachusetts Medical School, MA) and was used to generate pEYFP-Myo1c S142A,

T564A, S701A, IQ1A (I706A, Q707A, R711A and G712A), S701AIQ1A and K111A using the QuikChange II XL site-directed mutagenesis kit from Stratagene (La Jolla,

CA). The pBabe-HA-GLUT4, pGEX-14-3-3β plasmid and the mammalian expressing

GST-14-3-3β plasmid was previously described (Ramm et al., 2006). Full length

Myo1b cDNA and Armcx5 cDNA were purchased from Origene (Rockville, MD).

Recombinant 14-3-3β expressed as a glutathione-S-transferase (GST) fusion protein in E. coli was purified by standard procedures using Glutathione-4B sepharose (GE

Healthcare, Uppsala, Sweden). The GST tag of the recombinant GST-14-3-3β protein was removed by Thrombin protease before coupling to CNBr-activated sepharose (GE

Healthcare).

For immunoblotting and immunoprecipitation of Myo1c previously characterized antibody was used (Wagner et al., 1992). Anti-14-3-3β, anti-Actin and anti-CaMKII antibodies were from Santa Cruz (Santa Cruz, CA). Anti-calmodulin antibody was from Upstate (Lake Placid, NY). Anti-pAkt (pS473) was from Cell Signaling

Technology (Beverly, MA). Anti-pCaMKII (pT286) antibody was from Promega

(Madison, WI). Anti-GST antibody was from GE Healthcare. Anti-HA antibody was from Covance Research Products (Richmond, CA). Anti-GFP antibody for immunoprecipitation and Western blot was from Roche (Indianapolis, IN), and anti-GFP antibody for immunoflorescence was from Molecular Probes (Leiden, the

Netherlands). Secondary antibodies for Western blot were purchased from GE

Healthcare. Alexa Fluor 488- and Cy3-conjugated secondary antibodies for immunoflorescence staining were from Molecular Probes and Jackson

24 ImmunoResearch (West Grove, PA).

Akt inhibitor (Akti-1/2) was obtained from Peter Shepherd (University of Auckland,

New Zealand) and previously described (DeFeo-Jones et al., 2005). CaMKII inhibitor peptide tat-CN21 and tat-ctrl were obtained from Ulrich Bayer (University of

Colorado Denver, Aurora, CO) and previously described (Vest et al., 2007).

LY294002 and BAPTA-AM were purchased from Calbiochem (San Diego, CA).

Ampicillin and complete protease inhibitor cocktail tablets were obtained from Roche

(Indianapolis, IN). Isopropylthio-β-D-galactoside (IPTG) was purchased from Progen

(Madison, WI).

GST-14-3-3 fusion protein was dialysed using 12 - 14,000 molecular weight cut-off molecularporous membrane tubing (Spectrum, Rancho Dominguez, CA). All protein concentrations were assayed using Bicinchoninic acid (BCA) Protein Assay Kit

(Pierce, Rockford, IL). Absorbance at 540 nm was measured on a BMG

Labtechnologies FLUOstar Galaxy microplate reader.

For SDS-PAGE analysis, 30% Acrylamide/Bis solution from Bio-Rad (Hercules, CA) was used. SeeBlue Plus2 Pre-Stained Standard (Invitrogen, Carlsbad, CA) was used as protein standard. SYPRO Ruby gel stain was purchased from Molecular Probes. For

Western Blotting, Immobilon-P PVDF membrane was obtained from Millipore

(Billerica, MA), SuperSignal West Pico Chemiluminescent Substrate was obtained from Pierce and Super RX medical X-ray film was purchased from Fujifilm

(Brookvale, Australia).

For LC-MS/MS analysis, the HPLC system was purchased from Waters (Milford,

MA). The C18 trapping column, Famos autosampler, capillary column and Ultimate pump were purchased from LC Packings (Amsterdam, The Netherlands) The Qstar mass spectrometer was purchased from Applied Biosystems (Melbourne, Australia).

25 The 3T3-L1 murine fibroblasts, the Chinese hamster ovary cells expressing insulin receptor and IRS-1 (CHO/IR/IRS-1), PlatE cells were obtained from the American

Type Culture Collection (Rockville, MD). Dulbecco’s Modified Eagle solution

(DMEM), F-12 nutrient mixture and newborn calf serum (NCS) were purchased from

Gibco (Carlsbad, CA). Foetal calf serum (FCS) was from purchased Trace Scientific

(Melbourne, Australia). Insulin was obtained from Calbiochem. Lipofectamine 2000 reagent was purchased from Invitrogen.

All other chemicals were of analytical grade and obtained from Sigma (Sydney,

Australia). ii. Methods

A. Cell Culture

3T3-L1 fibroblasts and adipocytes

The 3T3-L1 fibroblasts were maintained in NCS/DMEM medium (DMEM supplemented with 10% NCS, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate) at 37 °C in an atmosphere of 10% CO2 in air. The cells were allowed to grow to 100% confluence prior to differentiation. To differentiate into addipocytes, the growth medium was replaced with differentiation medium (DMEM supplemented with 10% FCS, 2 μg/ml insulin, 100 ng/ml dexamethasone, 500 μM isobutyl-1-methyl-xanthine, 100 ng/ml biotin, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate). After 3 days of differentation, the differentiation medium was replaced with post-differentiation medium (DMEM supplemented with 10% FCS, 2 μg/ml insulin, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate). After 3 days of post-differentiation, the post-differentiation medium was replaced with FCS/DMEM medium (DMEM

26 supplemented with 10% FCS, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate).

3T3-L1 fibroblasts were infected with retrovirus as previously described (Ramm et al.,

2006). Cells were incubated with relevant virus for 5 h in the presence of 4 μg/ml polybrene. After 24 h, infected cells were selected in relevant antibiotics (1 mg/ml puromycin or 0.4 mg/ml geneticin). Differentiated 3T3-L1 adipocytes were transfected by electroporation as described (Bose et al., 2002). For transfection of siRNA, 20 nmol of siRNA was used per 10 cm dish. For cotransfection of plasmids and siRNA, 100 μg of DNA and 20 nmol of siRNA were used per 10 cm dish.

CHO/IR/IRS-1 cells and PlatE cells

CHO/IR/IRS-1 cells were maintained in FCS/F12 medium (F12 medium supplemented with 10% FCS, 0.8 mg/ml geneticin, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate) in 6-cm dishes at 37 °C in an atmosphere of

10% CO2 in air. PlatE cells were maintained in FCS/DMEM medium (DMEM supplemented with 10% FCS, 103 units/ml penicillin, 103 μg/ml streptomycin and 2 mM L-glutamate) with the addition of 1 μg/ml puromycin and 10 μg/ml blastcidine.

CHO/IR/IRS-1 cells and PlatE cells were transiently transfected with plasmids using

Lipofectamine 2000 according to manufacturer’s protocol.

B. Purification of 14-3-3 and preparation of 14-3-3-sepharose

Expression of GST-14-3-3βfusion protein

GST-14-3-3β fusion protein isoform was prepared for use in pulldown experiments. A stab of bacteria transformed with a plasmid containing the GST-tagged 14-3-3β gene was streaked out onto an LB agar plate containing ampicillin at 100 μg/ml and incubated at 37 °C overnight. A single colony was used to inoculate 20 ml of LB

27 media containing ampicillin at 100 μg/ml. This culture was incubated overnight at

37 °C in a shaking incubator at 180 rpm. The following day, the overnight culture was used to inoculate 1 L of LB media containing ampicillin at 100 μg/ml. The culture was then incubated at 37 °C in a shaking incubator at 180 rpm until the optical density at 600 nm (OD600) reached 0.6. Expression of GST-14-3-3 fusion protein was induced by the addition of IPTG to a final concentration of 1 mM and cells were incubated for a further 3 h. Bacterial pellets were collected by centrifugation at 10,000 g at 4 °C for

10 min and stored at -20 °C overnight.

Immobilization of GST-14-3-3 fusion protein

Each bacterial pellet was resuspended on ice in 20 ml of PBS supplemented with a complete protease inhibitor cocktail. The resuspended pellets were then sonicated five times on ice with 30 s duration each at output level 5. Triton X-100 and PMSF were added to the final concentration of 1% (v/v) and 300 μM, respectively, and the lysates were incubated on ice for 15 min. Lysates were centrifuged at 10,000 g at 4 °C for 30 min to pellet the cellular debris. The supernatant was incubated with 625 μl of glutathione-sepharose, which was pre-washed 3 times with 10 volumes of cold PBS, at 4 °C on a rotating wheel for 2 h. The supernatant was removed by centrifugation.

The sepharose was washed 5 times with 10 volumes of PBS and once with

10 volumes of thrombin cleavage buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl,

0.1% (v/v) β-mercaptoethanol, 2.5 mM CaCl2).

Thrombin cleavage of immobilized GST-14-3-3 fusion protein

The GST tag was removed from the immobilized GST-14-3-3 fusion protein bound on the glutathione-sepharose using thrombin. Thrombin cleavage buffer (2 ml) containing 100 U of thrombin was added to the GST-14-3-3 bound glutathione-sepharose. The bound sepharose and thrombin cleavage buffer were

28 incubated at room temperature for 6 h on a rotating wheel. The supernatant was collected by centrifugation.

The thrombin was removed from the thrombin cleavage reaction to obtain purified

14-3-3. The supernatant of the thrombin cleavage reaction was added to 300 μl of benzamidine-sepharose which was pre-washed 3 times with 10 volumes of thrombin cleavage buffer. The mixture was incubated at room temperature for 1 h on a rotating wheel to immobilize the thrombin on the benzamidine-sepharose. The supernatant was collected by centrifugation.

Coupling of 14-3-3 to sepahrose

The purified 14-3-3 was covalently coupled to the CNBr-activated sepharose via the reaction between the N-terminal primary amine of the protein and the CNBr active group of the sepharose. The purified 14-3-3 elute was dialyzed against 2 L coupling buffer (0.1 M NaHCO3, pH 8.3, 0.5 M NaCl) overnight at 4 °C. CNBr-activated sepharose (500 mg) in freeze-dried powder form was suspended in 100 ml of 1 mM

HCl. The swollen sepharose was washed for 15 min with 1 mM HCl on a sintered glass filter containing a Whatman No. 1 filter paper. Dialyzed 14-3-3 solution (2 ml) was mixed with the sepharose and the mixture was incubated for 4 h at room temperature on a rotating wheel. The supernatant was removed by centrifugation and the sepharose was washed with 10 volumes of coupling buffer. Any remaining active groups on the sepharose were blocked by incubation with 0.1 M Tris-HCl buffer, pH

8.0, for 2 h at room temperature on a rotating wheel. The control sepharose was prepared by directly blocking active groups of CNBr-sepharose with 0.1 M Tris-HCl buffer, pH 8.0 without any prior protein-coupling. The supernatant was removed by centrifugation. The sepharose was washed with 3 cycles of alternating pH, in which each cycle consisted of a wash with 10 volumes of low pH wash buffer (0.1 M

29 CH3COONa, pH 4.0, 0.5 M NaCl) followed by a wash with 10 volumes of high pH wash buffer (0.1 M Tris-HCl, pH 8.0, 0.5 M NaCl). The 14-3-3-sepharose was stored as 50% slurry in PBS supplemented with 0.02% sodium azide at 4 °C. Protein concentrations of coupling buffer before and after coupling were measured by the

BCA assay to determine the coupling efficiency.

C. Pulldown and immunoprecipitation

Transfected CHO/IR/IRS-1 cells or 3T3-L1 adipocytes were incubated in serum-free

Dulbecco's modified Eagle's medium (Invitrogen) for 2 h, stimulated with 100 nM insulin (Calbiochem) for 15 min or preincubated with kinase inhibitors prior to insulin stimulation, and lysed in lysis buffer (1% Nonidet P-40, 137 mM NaCl, 10% glycerol,

25 mM Tris, pH 7.4) supplemented with Complete protease inhibitor mixture (Roche) and phosphatase inhibitors (2 mM sodium orthovanadate, 1 mM pyrophosphate, 10 mM sodium fluoride) at 4 °C. Pulldown or immunoprecipitation was performed overnight at 4 °C with 14-3-3-coupled sepharose, glutathione sepharose or specific antibodies (anti-GFP, anti-GST, anti-Myo1c, anti-CaMKII, anti-FLAG and anti-Armcx5) coupled to Protein G-sepharose (Pierce, Rockford, IL). Beads were washed three times with lysis buffer and two times with PBS and boiled in sample buffer.

D. PM silica preparation of plasma membrane

Basal or insulin stimulated 3T3-L1 adipocytes were washed twice with PBS, and twice with coating buffer (20 mM MES, 150 mM NaCl, 280 mM sorbitol pH 5.5).

Coating buffer containing 1% silica was added to the cells for 2 min on ice and the cells were washed once with coating buffer. Coating buffer containing 1 mg/ml sodium polyacrylate (pH 6.5) was added to the cells and incubated at 4°C for 2 min.

Cells were washed once with coating buffer and once with modified HES buffer

30 (20 mM HEPES pH 7.4, 250 mM sucrose, 1 mM DTT, 1 mM magnesium acetate, 100 mM potassium acetate, 0.5 mM zinc chloride) and lysed in modified HES buffer.

Nycodenz (100%) in modified HES buffer was added to the lysate to a final concentration of 50%. The lysate was layered onto 0.5 ml of 70% Nycodenz in modified HES and centrifuged at 41,000 g for 20 min at 4°C. The pellet was washed three times with modified HES buffer and resuspended in SDS-PAGE sample buffer.

E. SDS-PAGE gel staining and immunoblotting

Silver staining

After electrophoresis, the gel was incubated with a fixative solution (30% (v/v) ethanol, 10% (v/v) acetic acid) for 30 min. The gel was washed 3 times with 30% (v/v) ethanol for 15 min and 3 times with milliQ water for 5 min. The gel was sensitized in

1.4 mM Na2S2O4 for 1 min. After washing twice with milliQ water for 1 min, the gel

-5 was incubated in 8.6 mM Ag(NO3)2 solution in the presence of 2.8 × 10 % (v/v) formaldehyde for 1 h with agitation. The gel was washed with milliQ water for 1 min and the bands were developed in a developer solution (0.6 M Na2CO3, 253 μM

-4 Na2S2O3, 1.9 × 10 % (v/v) formaldehyde) for 20 min. The reaction was stopped by incubation in 3.5% (v/v) acetic acid for 10 min. The gel was washed with milliQ water.

SYPRO Ruby staining

For protein identification using mass spectrometry, the SDS-PAGE gel was stained with SYPRO Ruby gel stain prior to in-gel digestion. After electrophoresis, the gel was incubated in SYPRO Ruby fix solution (50% (v/v) methanol, 7% (v/v) acetic acid) at room temperature for 30 min. The gel was incubated in SYPRO Ruby gel stain at room temperature overnight with agitation. The gel was then washed with a wash

31 solution (10% (v/v) methanol, 7% (v/v) acetic acid) for 30 min followed by milliQ water for 5 min. Proteins stained with SYPRO Ruby gel stain were visualized using a

300 nm UV transilluminator.

Western blotting

Subsequent to SDS-PAGE, proteins were transferred to PVDF membrane for 75 min at 120 V. After transfer, the membrane was stained with Ponceau S solution (0.1%

(w/v) Ponceau S, 1% (v/v) acetic acid) for 1 min and destained with milliQ water. The membrane was incubated in a blocking solution (2% (w/v) BSA in TBS/Tween 20

(0.1%, v/v)) for 1 h. The membrane was incubated with primary antibody in blocking solution overnight 4 °C with agitation. After washing 3 times with TBS/Tween 20

(0.1%, v/v) for 10 min, the membrane was incubated with HRP conjugated secondary antibody in 2% (w/v) skim milk in TBS/Tween 20 (0.1%, v/v) for 1 h at room temperature with agitation. The membrane was washed 3 times with TBS/Tween 20

(0.1%, v/v) for 10 min and 2 times with TBS for 5 min. After incubation in Super

Signal West Pico Chemiluminescence Substrate for 5 min, the membrane was exposed to X-ray film.

Far Western blotting

The membrane was incubated with GST-14-3-3 prior to primary antibody incubation.

After incubating the membrane in blocking solution, the membrane was incubated with GST-14-3-3 (2.15 μg/ml) in blocking solution for 1 h with agitation. Another membrane was incubated with the same molar concentration of GST in blocking solution for 1 h with agitation and used as a control. After washing 3 times with

TBS/Tween 20 (0.1%, v/v) for 10 min, the membrane was subjected to primary and secondary antibody incubations.

32 Phosphatase treatment of blot

For phosphatase treatment, the blot was incubated with alkaline phosphatase before blocking with blocking solution. After transfer, the membrane was washed twice with

TBS/Tween 20 (0.1%, v/v) for 5 min. The blot was incubated in shrimp alkaline phosphatase buffer (200 mM Tris-HCl, pH 8.0, 100 mM MgCl2) containing 100 U/ml of shrimp alkaline phosphatase overnight at 37 °C with agitation. After washing 2 times with TBS for 5 min, the blot was blocked in blocking solution and subjected to immunoblotting.

F. Immunoflorescence analysis

After insulin stimulation, 3T3-L1 adipocytes on glass coverslips were fixed with 3% paraformaldehyde and block with 2% BSA. For surface labeling of with anti-HA antibody, the cells were permeabilized after primary antibody incubation. Otherwise, the cells were blocked and permeabilized with 0.1% saponin/2% BSA prior to primary and secondary antibody incubations. Images were obtained using the Leica

TCS SP2 confocal laser scanning microscope. For quantification of surface anti-HA antibody labeling, a region of interest (R.O.I) was set around each cell and the amount of fluorescence per unit area was quantified using ImageJ software (National

Institutes of Health).

G. 2-Deoxyglucose uptake assay

3T3-L1 adipocytes were preincubated with either DMSO for 45 min, 10 μM KN-62 for 45 min, 5 μM tat-CN21 for 30 min or 5 μM tat-ctrl for 30 min in KRP buffer

(120 mM NaCl, 600 μM Na2HPO4, 400 μM NaH2PO4, 6 mM KCl, 1.2 mM MgSO4,

1 mM CaCl2, 12.5 mM HEPES, pH 7.4) prior to 100 nM insulin stimulation for another 15 min. Uptake of 2DOG and 3H-2DOG (Perkin Elmer Life Science,

33 Melbourne, Australia) was measured over the final 5 min of insulin stimulation and analyzed by scintillation counting.

H. GLUT4 translocation assay

3T3-L1 fibroblasts were grew and differentiated on 96-well plate. After insulin stimulation, the cells were fixed with 3% paraformaldehyde for 45 min and quenched with 50 mM Glycine in PBS. For surface labeling of HA epitope tag, the cells were block with 5% normal swine serum (NSS) and incubated with anti-HA antibody before permeabilized with 0.1% saponin/5%NSS/PBS. Otherwise, the cells were blocked and permeabilized with 0.1% saponin/5% NSS/PBS prior to primary antibody incubations for whole cell labeling of HA epitope tag. Then the cells were incubated with Alexa Fluor 488-conjugated mouse secondary antibody and the fluorescence intensity (emm 485 nm/exc 520 nm) was measured with the bottom-reading mode in a fluorescence microtiter plate reader.

I. Statistical analysis

Results are given as mean ± SEM. Statistical analyses were performed by using a

Student’s t-test and p<0.05 was taken to indicate a significant difference.

34 Chapter 3

Phosphoproteomic Analysis of Insulin Signaling Pathway in Adipocytes

35 i. Abstract

Insulin stimulates a lot of cellular processes via receptor-mediated signaling pathways.

The phosphorylation of proteins is a central regulatory mechanism that alters enzymatic activity and protein-protein interaction, thereby allowing signal transduction from the insulin receptor to the downstream signaling proteins and effectors. Due to the diverse effects of insulin, not all of the insulin-regulated substrates have been identified yet. These include a critical insulin-regulated substrate in GLUT4 translocation. The 14-3-3 proteins are phosphoserine/threonine binding proteins that play a critical role in mediating the effects of protein phosphorylation and have been implicated in many cellular processes. In order to identify novel insulin-regulated phosphoproteins in adipocytes, a proteomic analysis of

14-3-3-interacting proteins in adipocytes was performed. 14-3-3 protein was used as a tool to affinity-purify phosphoproteins from 3T3-L1 adipocyte lysate. After separation with SDS-PAGE, the proteins were extracted and identified by liquid chromatography tandem mass spectrometry (LC-MS/MS). Thirty-eight proteins including several novel 14-3-3-associated proteins were identified and they have been implicated in metabolism, signal transduction, intracellular trafficking and cytoskeletal dynamics.

This study not only represents a novel strategy in studying growth factor-regulated phosphoproteome, but also further explores the 14-3-3-interactome. ii. Introduction

Insulin increases glucose uptake into muscle and fat cells by triggering the translocation of the facilitative glucose transporter GLUT4 from intracellular storage vesicles (GSVs) to the plasma membrane (PM). Recent progress in defining key insulin-regulated steps has been derived from two separate approaches: the use of a membrane fusion assay in which some aspects of the process have been reconstituted

36 in vitro (Koumanov et al., 2005), and the use of total internal reflection fluorescence microscope (TIRFM), a high resolution method for studying events close to the PM in living cells (Bai et al., 2007; Huang et al., 2007). Both of these methods indicate that the PM is a major regulatory node for insulin action. In particular, the latter method dissected the encounter of GSVs with the PM into multiple steps and showed that a step downstream of vesicle docking is the major insulin-regulated step (Bai et al.,

2007). This has provided important information in the search for key molecules that function at the intersection point between signaling and vesicle transport.

The PI3K/Akt pathway plays a central role in regulating glucose transport in muscle and fat cells and so identifying downstream targets of Akt, particularly those acting at the PM, is an important goal. One such target is the RabGAP TBC1D4/AS160, the phosphorylation of which plays an essential yet incomplete role in GLUT4 trafficking

(Sano et al., 2003). Overexpression of a TBC1D4 mutant, in which each of the Akt phosphorylation sites have been mutated, in adipocytes blocks insulin-stimulated

GLUT4 translocation (Sano et al., 2003). However, using TIRFM it has been shown that the function of TBC1D4 is most closely linked to GSV docking which as described above seems to be a relatively minor insulin-regulated step (Bai et al.,

2007).

Recent studies have shown that in vitro 14-3-3-affinity purification followed by mass spectrometry analysis could serve as an efficient and rapid method for identification of 14-3-3-binding proteins. Meek et al. (2004) have recently performed such a study to identify proteins that are differentially bound to 14-3-3 between interphase and mitotic phase (Meek et al., 2004). In this study, 14-3-3ζ-interacting proteins from extracts of both interphase and mitotic HeLa cells were bound to GST-14-3-3 beads.

The interacting proteins were eluted and separated by SDS-PAGE and subsequently

37 identified by high performance liquid chromatography (HPLC) tandem mass spectrometry (MS/MS) (Meek et al., 2004). Among the 250 proteins that were identified, more than 100 proteins were cell cycle related proteins. Rubio et al. (2004) have identified over 200 14-3-3-binding phosphoproteins in HeLa cells using a similar approach (Rubio et al., 2004). These proteins were involved in metabolism, cell proliferation, cytoskeletal dynamics and trafficking. Jin et al. (2004) performed immunoprecipitation of different 14-3-3 isoforms to identify in vivo 14-3-3-binding proteins in HEK293 cells (Jin et al., 2004). In this study, 70 14-3-3β-binding, 127

14-3-3γ-binding, 26 14-3-3ζ-binding and 19 14-3-3θ-binding proteins were identified.

The proteins identified by Jin et al. only showed a 26% overlap with the in vitro

14-3-3-binding proteins identified by Meek et al. and Rubio et al., which may be due to differences in purification procedures, the 14-3-3 isoforms or the cell types used

(Jin et al., 2004). More recently, Benzinger et al. (2005) combined the tandem affinity purification and the multidimensional protein identification technology to characterize

117 proteins presumably associated with 14-3-3σ in vivo in HEK293 cells (Benzinger et al., 2005). All of these studies indicated that both in vitro and in vivo

14-3-3-purification assays give promising results in HeLa and HEK293 cells.

However, as of yet this kind of approach has not been used in adipocytes to study insulin action. I describe here the method development and the use of this approach for the identification of novel insulin-regulated phosphoproteins in adipocytes.

For the continuity of this chapter, part of the results from this chapter was extracted from my honours thesis as this project started in my honours year. iii. Methods

A. Mass spectrometry analysis

38 In-gel tryptic digestion

The protein bands were excised from the SDS-PAGE gel as gel plugs (each about 1 mm × 5 mm) after staining with SYPRO Ruby gel stain. Each gel plug was destained with 1 ml of 25 mM NH4HCO3, pH 8 for 15 min. The gel plug was washed with 1 ml of 50% CH3CN for 15 min and dehydrated in 1 ml of 100% CH3CN for 10 min. After removing all CH3CN by centrifugation, the gel plug was incubated in 20 μl of Trypsin

(12.5 ng/μl) in NH4HCO3 (100 mM), pH 8.0 overnight at 37 °C. The following day, the gel plug was incubated with 100 μl of 5% formic acid for 1 h at room temperature.

Then, 100 μl of 100% CH3CN was added and incubated for another 1 h at room temperature. The supernatant, containing extracted peptides, was collected from the gel plug. The peptides were dried using a speedivac and resuspended in 5 μl of 5% formic acid.

LC-MS/MS analysis

Peptide samples (5 µl) were loaded onto a trapping column (C18, 0.3 × 1 mm) from a

Famos autosampler at 20 µl/min. After a 4 min wash with buffer A (98% H2O, 2%

CH3CN, 0.1% formic acid), the trapping column was switched into line with the capillary column (0.3 × 1 mm). Peptides were eluted at 200 nl/min using a gradient of

95% buffer A to 60% buffer B (20% H2O, 80% CH3CN, 0.1% formic acid) in 30 min using an Ultimate pump. Information-dependent acquisition (IDA) data were acquired using the QStar mass spectrometer with Analyst QS software.

MASCOT (Martix Sciences) was used as the search engine for the MS/MS analysis.

MASCOT search parameters were the following: SwissProt database with all entries of taxonomy, allowing one missed cleavage of trypsin and variable modification of oxidation (Met), peptide charge of 2 +, 2.0 Da tolerance in MS and 0.8 Da tolerance in MS/MS; a match with p < 0.05 was accepted for identification of peptides. The

39 SCANSITE program (http://scansite.mit.edu/) was used to detect 14-3-3-binding consensus motifs of the identified proteins. iv. Results

Purification of 14-3-3 and preparation of 14-3-3-sepharose

The GST-14-3-3β fusion protein was produced and purified from transformed E. coli.

Glutathione-sepharose was used to purify the GST-14-3-3 fusion protein (Figure 3.1).

GST-14-3-3 fusion protein was visible as the major protein bands in the crude E. coli extract and the sample after the glutathione-sepharose affinity purification (Figure

3.2B, lane 1 and 3). Figure 3.2A shows an immunoblot of crude extract of transformed E. coli and GST-14-3-3 purified by glutathione-sepharose using an anti-14-3-3 antibody. The ratio of specific activity of GST-14-3-3, measured as the intensity of immunoreactive GST-14-3-3 band over the total amount of protein resolved by the SDS-PAGE resolving gel, in the crude extract to that in the purified sample was calculated as 1 to 2.7. This ratio indicated that GST-14-3-3 fusion proteins have been enriched for 2.7-fold in the purified sample compared to the crude extract.

However, some other bacterial proteins were also present in this purified sample

(Figure 3.2B, lane 3). These proteins were thought to be bacterial proteins associating with GST or GSH.

In order to remove GST-binding bacterial proteins, the GST-tag was removed by thrombin cleavage (Figure 3.1). Moreover, the use of 14-3-3-sepharose, instead of

GST-14-3-3-sepharose, in subsequent pulldown experiments avoids isolation of

GST-binding proteins from mammalian cells. Therefore, after binding of GST-14-3-3 to glutathione-sepharose, 14-3-3 was eluted by cleavage with thrombin. This cleavage specifically eluted 14-3-3 from the sepharose, leaving GST, GST-binding and

GSH-binding bacterial proteins bound on the sepharose (Figure 3.2B, lane 4). The

40 thrombin was then removed by binding to benzamidine-sepharose (Figure 3.1). This sample was then used to prepare 14-3-3-sepharose. Although the supernatant from the benzamidine-sepharose step was not completely pure, the GST tags of many

GST-14-3-3 fusion proteins and some of the contaminating bacterial proteins were removed using this approach (Figure 3.2B, lane 5 and 6).

Western blot analysis of known 14-3-3-binding proteins

After purification of 14-3-3 and coupling to sepharose, the binding activity of immobilized 14-3-3 was tested. The binding activity of 14-3-3 was checked by blotting for previously reported 14-3-3-binding proteins. Eluates were immunoblotted with antibodies specific for IRS-1 and phospho-GSK-3 since they have previously been shown to bind to 14-3-3 in an insulin-dependent manner (Kosaki et al., 1998;

Xiang et al., 2002; Yuan et al., 2004). As shown in Figure 3.3, the amount of IRS-1 and phospho-GSK-3 bound to 14-3-3-sepharose was significantly greater using lysate from insulin-treated cells compared to lysate from basal cells. There was no significant IRS-1 or phospho-GSK bound to sepharose beads that had not been cross-linked with 14-3-3 proteins. These data indicate that the 14-3-3 proteins used in these experiments are capable of binding to known 14-3-3-binding proteins in the adipocyte lysate in an insulin-regulated and specific manner.

SDS-PAGE and Western blot analysis of 14-3-3-associated proteins

The goal of the next series of experiments was to show that insulin-stimulated phosphoproteins in adipocytes could be isolated using 14-3-3-affinity purification.

To test this hypothesis, the 14-3-3-associated protein profiles of basal cells and insulin-treated cells were compared via silver staining of SDS-PAGE resolving gel.

The basal and insulin-stimulated lysates of 3T3-L1 adipocytes were incubated with

41 14-3-3-sepharose. During cell lysis, minimum amount of NP40 buffer was used to maintain the cell extract as concentrated as possible. This may improve the purification of low-affinity 14-3-3-interacting proteins. The sepharose was then washed with a buffer containing salt (NaCl) and a mild non-ionic detergent (NP-40) to remove non-specific binding proteins. Proteins that bound to the 14-3-3-sepharose were then eluted by boiling in sample buffer. Figure 3.4 shows the silver stained

SDS-PAGE resolving gel of 14-3-3-associated proteins under basal and insulin-stimulated conditions. In the silver stained gel, certain bands were much more intense in the lane corresponding to the insulin-stimulated lysate compared to that from basal cells. Some of these proteins were also present in the basal condition but the intensities were much weaker. This experiment suggested that this technique is capable of isolating insulin regulated 14-3-3-associated proteins from adipocytes and that some of the 14-3-3-associated proteins are insulin dependent.

Although 14-3-3 generally binds to phosphorylated sites on target proteins, non-phosphorylated 14-3-3 binding sequences have been identified in certain interacting proteins (Fuglsang et al., 2003). Since this project aimed to identify phosphoproteins, the phosphorylation-dependence of 14-3-3-binding of

14-3-3-purified proteins was tested. This was achieved by using a Far Western blotting approach following treatment of the PVDF membrane with alkaline phosphatase (Figure 3.5). For phosphatase treatment, after the transfer of proteins from the SDS-PAGE resolving gels to the PVDF membranes, the membranes were incubated in the buffer with or without alkaline phosphatese which hydrolyzes the phosphate groups of the phosphoproteins.

Phosphatase treatment of blot was first checked by immunoblotting of total cell extract. Figure 3.5A shows the Western blot of the total cell extract after the

42 phosphatase treatment. As indicated, treatment with alkaline phosphatase effectively removed phosphate groups from target proteins on the membrane. This was most readily apparent for Akt where recognition of phospho-Akt with the phospho-specific antibody was completely absent after treatment with alkaline phosphatase. This was not simply a non-specific effect as binding of the total Akt antibody was unaffected by alkaline phosphatase and this antibody does not relay upon phosphorylation of the target protein for its binding.

Figure 3.5B shows the GST-14-3-3 Far Western blot of 14-3-3-associated proteins after phosphatase treatment of blot. After 14-3-3-pulldown, the 14-3-3-associated proteins were transferred to the PVDF membrane which was then subjected to

GST-14-3-3 overlay after phosphatase treatment. The GST-14-3-3 overlay allowed the binding of GST-14-3-3 fusion proteins to the immobilized proteins on the membrane.

Since the fusion protein recognized both GST-binding and 14-3-3-binding proteins,

GST overlay was performed in parallel to distinguish GST-binding proteins. As indicated in Figure 3.5B, the GST-14-3-3 fusion protein, but not GST, labeled a large number of proteins. The labeling of the majority of the proteins was substantially diminished subsequent to phosphatase treatment. This experiment suggested that, for the majority of the 14-3-3-associated proteins purified in this study, phosphorylation was required for the 14-3-3-binding.

As mentioned before, 14-3-3 can bind to many phosphoproteins that are substrates of different kinases. As far as the insulin signaling pathway is concerned, Akt-dependent phosphoproteins are the major focus. Hence, in order to determine if 14-3-3-affinity purification is isolating a range of Akt-dependent phosphoproteins, a commercially available anti-phospho-Akt substrate antibody (α-pAS) was employed. This antibody, which has been raised against a degenerate Akt phosphorylation peptide library,

43 preferentially recognizes the Akt substrates containing a motif of R/KXR/KXXpS/T.

Figure 3.6 shows the immunoblot of the 14-3-3-associated proteins using anti-pAS antibody to determine if Akt phosphorylates 14-3-3-associated proteins. Few bands were labeled with this antibody in the lane corresponding to the 14-3-3-pulldown from basal cells. However, an increased number of Akt substrates were present in the sample from insulin-treated cells. Although the Akt phosphorylation sites of these proteins are not necessarily the 14-3-3-binding sites, this experiment supports the hypothesis that 14-3-3-affinity purification was able to isolate insulin-dependent Akt substrates from adipocytes.

Collectively, these experiments validate the hypothesis that most of the

14-3-3-binding proteins purified using 14-3-3-affinity purification were phosphoproteins and some of them were insulin-regulated and substrates of Akt.

Identification of 14-3-3-assocaited protein by mass spectrometry

In order to identify the 14-3-3-associated proteins from insulin-treated adipocytes, a large scale 14-3-3-affinity purification was performed using lysate from insulin-treated adipocytes. After separation of 14-3-3-associated proteins by

SDS-PAGE, the gel was stained with SYPRO Ruby gel stain, which is a fluorescent gel stain, because it is more compatible with mass spectrometry than silver staining

(Finehout and Lee, 2003). The relevant gel lane was sliced into 16 separate slices, each of which was subjected to tryptic digestion. A gel slice from a non-stained region was used as a negative control. After in-gel tryptic digestion, the peptides were identified by LC-MS/MS which provided amino acid sequence information and allowed the identification of the corresponding proteins in each gel piece.

A number of protein matches were obtained by the Mascot search engine corresponding to each gel slice. These matches had to fulfill certain criteria in order to

44 be considered as potential 14-3-3-associated proteins. The protein match should have a score of at least 43 and a p-value of less than 0.05. The molecular weight of each protein identified in a gel slice should correlate with its relative migration by

SDS-PAGE. Trypsin and keratins were detected in all gel slices and the negative control and hence they were regarded as contaminants.

After subtraction of proteins that did not fulfill the above criteria, 38 protein identifications representing potential 14-3-3β-associated proteins were obtained

(Table 3.1). The 14-3-3-associated proteins were grouped according to the processes in which they had been implicated previously. These proteins reflect the important roles of 14-3-3 in different processes including cytoskeletal dynamics, metabolism, signal transduction, protein processing, oxidative stress, protein trafficking, transcription and translation.

Among these proteins, 17 proteins were previously shown to associate with 14-3-3 isoforms in either detailed case-by-case studies or large-scale proteomic analyses. It is noted that some known 14-3-3-binding proteins in adipocytes, including IRS-1 and

GSK-3, are not included in the list. However, IRS-1 and GSK-3 were detected but their scores were not high enough to be considered as potential 14-3-3-associated proteins. Thus it is likely that further scaling up of the experiment will be required to identify these and other proteins. Moreover, the exact 14-3-3-isoforms that interact with these proteins are not clear. The β isoform may not be the best binding partner for these proteins and hence they were not detected in the 14-3-3β-pulldown.

It is however possible that some of the proteins detected here indirectly associate with

14-3-3β via interaction with other 14-3-3β-associated proteins or represent contaminants. To get a further indication about the specificity of the interaction, in silico detection of 14-3-3-binding motifs was performed. The presence and location of

45 putative 14-3-3 consensus binding motifs (mode 1, RSXpS/TXP) were determined for the 38 identified 14-3-3β-associated proteins using the software SCANSITE

(Obenauer et al., 2003). When a query with the highest stringency was performed, only 6 out of 38 (16%) proteins contained at least one 14-3-3-binding consensus motif

(mode 1, RSXpS/TXP). When medium and low stringency settings were applied, 16

(42%) and 34 (89%) proteins showed 14-3-3-binding sites, respectively. The highest stringency setting allowed more selective scanning while the lowest stringency setting led to a more sensitive scanning of 14-3-3-binding sites. However, the sites identified by the SCANSITE program have to be treated cautiously. The high stringency setting may lead to false negative identification while the low stringency setting may result in false positive identification. Moreover, the identified sites are not necessarily direct binding sites of 14-3-3 since the 14-3-3-binding of a protein may also depend on the flanking regions of the putative 14-3-3-binding site of the protein. Although the specificity of binding to 14-3-3 of every individual protein must be checked, the protein that contains 14-3-3-binding sites identified by the SCANSITE program is more likely to be a direct 14-3-3-binding protein, and hence a phosphoprotein, than the protein that does not contain any 14-3-3-binding sites under any stringency settings. The limitation of the SCANSITE program is that it does not allow scanning of the other classic 14-3-3 consensus binding motif (mode 2, RXXXpS/TXP) and non-classic 14-3-3-binding motifs, including those in the IGF-1 receptor and p53

(Craparo et al., 1997; Waterman et al., 1998).

Besides those 38 proteins listed in Table 3.1, some 14-3-3 isoforms were also identified. Table 3.2 shows that, except for the τ isoform, all 14-3-3 isoforms were purified in 14-3-3β-affinity purification which is in accordance with previous reports showing heterodimerization between different 14-3-3 isoforms (Chaudhri et al., 2003).

46 However, the sequence coverage for other isoforms was relatively low compared to that of the β isoform, confirming previous experimental results that most of the immobilized 14-3-3β formed stable dimers on the sepharose. Purification of other

14-3-3 isoforms may be due to binding of these isoforms to a relatively small amount of immobilized 14-3-3β monomer or displacement of the β isoform from the dimer by one of the endogenous isoforms.

In the next experiment, I aimed to identify proteins that were clearly seen to be up-regulated in response to insulin. The SYPRO Ruby stained gel of

14-3-3-associated proteins from basal and insulin-treated cells is shown in Figure 3.7.

The protein bands that showed at least 1.5-fold increased intensity in insulin-stimulated conditions were excised, digested with trypsin and subjected to

LC-MS/MS as described. This allowed identification of 7 insulin-regulated

14-3-3-associated proteins. All of the 7 proteins were identified before in the previous experiment (Table 3.1). These proteins represent some of the highly abundant insulin-regulated 14-3-3-associated proteins and that the insulin effects were able to be observed in SYPRO Ruby stained gel. Importantly, they do not represent all the insulin-regulated 14-3-3-associated proteins since SYPRO Ruby is not as sensitive as silver staining or Western blotting.

The insulin-dependent and direct 14-3-3-binding of the proteins listed in Table 3.1 must be checked in case-by-case studies in order to determine if they are phosphorylated upon insulin stimulation. However, due to time restriction, only

Myo1c was chosen for further study as it has previously been implicated in GLUT4 trafficking (Bose et al., 2002; Bose et al., 2004). Myo1c has previously been described as a molecular motor which directs the GLUT4 vesicles along actin filaments to the plasma membrane. This process is believed to be insulin-regulated.

47 However, the molecular mechanisms of how Myo1c is regulated by insulin are unknown. Immunoflorescence analysis of endogenous Myo1c and PM silica preparation revealed that Myo1c is localized at the PM (Figure 3.8). The subcellular localization of Myo1c does not change in response to insulin. In another

14-3-3-pulldown experiment performed by other member of our laboratory using PM and cytosolic fraction followed by MS analysis showed that the sequence coverage of

Myo1c purified from PM was 37.2% compared to only 3.6% from the cytosol, suggesting that the PM is the major location where 14-3-3 interacts with Myo1c. The

14-3-3-binding of Myo1c described here indicates that Myo1c may potentially be regulated by Ser/Thr phosphorylation.

48

Figure 3.1 Flow diagram of preparation of 14-3-3-sepharose. A. Expression of GST-14-3-3 fusion

protein. B. Immobilization of GST-14-3-3 fusion protein on GSH-sepharose. C. Thrombin cleavage of immobilized GST-14-3-3 fusion protein. D. Removal of thrombin using benzamidine-sepharose. E.

Coupling of 14-3-3 to CNBr-sepharose. F. Affinity-purification of phosphoproteins using the

14-3-3-sepharose.

49

Figure 3.2 Purification of 14-3-3 and preparation of 14-3-3-sepharose. A. Western blot analysis of

GST-14-3-3. Crude extract (1 μg) and purified GST-14-3-3 sample (1 μg) were loaded onto a 10%

SDS-PAGE resolving gel, transferred to PVDF membrane and immunoblotted with anti-14-3-3β antibody. Specific activity of GST-14-3-3 (Unit/mg) was calculated as the intensity of immunoreactive

GST-14-3-3 band over total amount of proteins resolved by the gel. B. Silver stain of samples at various stages of the preparation. After lysis of bacterial cells, the crude extract (lane 1, 10 μg) was

50 mixed with GSH-sepharose. The supernatant (lane 2, 10 μg) was removed and the immobilized

GST-14-3-3 (lane 3, 5 μg) on GSH-sepharose was eluted to check the purity of bound proteins. The

immobilized GST-14-3-3 was cleaved by thrombin, leaving GST bound on the sepharose. The proteins bound on GSH-sepharose after thrombin cleavage (lane 4, 5 μg) was eluted to evaluate the efficiency of

cleavage. After cleavage of immobilized GST-14-3-3, thrombin was removed from the supernatant by

mixing with benzamidine-sepharose. The supernatant of benzamidine-pulldown (lane 5, 5 μg),

containing 14-3-3, was coupled to CNBr-sepharose. The proteins bound on the resulting

14-3-3-sepharose (lane 6, 2 μl of sepharose) were eluted to check the purity of 14-3-3-sepharose.

51

Figure 3.3 Confirming the binding activity of immobilized 14-3-3 by Western blot analyses of

IRS-1 and pGSK-3. 14-3-3-sepharose (containing 45 μg of total bound 14-3-3) was incubated with lysate (2 mg protein) obtained from either basal or insulin-stimulated adipocytes. The control sepharose was prepared by directly blocking active groups of CNBr-sepharose with 0.1 M Tris-HCl buffer, pH

8.0 without any prior protein-coupling. The control sepharose was mixed with the same amount of basal and insulin-stimulated lysate (2 mg). After overnight incubation, the sepharose was washed and the proteins were eluted, run on 10% SDS-PAGE resolving gel, transferred to PVDF membrane and immunoblotted with the indicated antibodies. The experiment was performed three times and the images are from a representative experiment.

52

Figure 3.4 14-3-3-pulldown from 3T3-L1 adipocytes. 3T3-L1 adipocytes were stimulated with 100 nM insulin for 30 min, and lysates were incubated with 14-3-3-sepharose beads overnight, beads were washed and bound proteins eluted with SDS-PAGE sample buffer followed by separation of the proteins by SDS-PAGE. 14-3-3-sepharose without incubation with lysate was used as control. To detect total protein, the gel was silver-stained.

53

Figure 3.5 Alkaline phosphotase treatment of blot and GST-14-3-3 Far Western blotting. A.

Alkaline Phosphatase treatment of blot. Cell extracts (10 μg) of basal and insulin-treated 3T3-L1 adipocytes were run on 7.5% SDS-PAGE resolving gel, transferred to PVDF membrane. The membrane was incubated with or without alkaline phosphatase in phosphatase buffer overnight at

54 37 °C and immunoblotted with the indicated antibodies. B. Phosphatase treatment and GST-14-3-3 Far

Western blot analysis of 14-3-3-associated proteins. 14-3-3-sepharose (containing 23 μg of total bound

14-3-3) was incubated with lysate (1 mg protein) obtained from either basal or insulin-stimulated adipocytes. After overnight incubation, the sepharose was washed and the proteins were eluted, run on

7.5% SDS-PAGE resolving gel and transferred to PVDF membrane. The membrane was incubated with or without alkaline phosphatase in phosphatase buffer. The membrane was probed for binding to either GST or GST-14-3-3. Anti-GST antibody was used as primary antibody. The experiment was performed two times and the images are from a representative experiment.

55

Figure 3.6 Western blot analysis of 14-3-3-associated proteins using phospho-Akt substrate (pAS) antibody. 14-3-3-sepharose (containing 23 μg of total bound 14-3-3) was incubated with lysate (1 mg protein) obtained either from basal or insulin-stimulated adipocytes. After overnight incubation, the sepharose was washed and the proteins were eluted, run on 7.5% SDS-PAGE resolving gel, transferred to PVDF membrane and immunoblotted with anti-pAS antibody. The experiment was performed two times and the images are from a representative experiment.

56 Accession Protein Sequence Function Symbol Name MW (kDa) SCANSITE number score coverage P05064 ALDOA_MOUSE Fructose-bisphosphate aldolase A 39 53 3% 0, 0, 2 P06151 LDHA_MOUSE L-lactate dehydrogenase A chain # 36 106 7% 0, 0, 0 Glycolysis P16858 G3P_MOUSE Glyceraldehyde-3-phosphate dehydrogenase # 36 182 12% 0, 0, 1 P52480 KPYM_MOUSE Pyruvate kinase, isozyme M2 # 58 54 3% 0, 1, 1

Fatty acid P19096 FAS_MOUSE Fatty acid synthase # 272 445 5% 0, 2, 11 biosynthesis Q91V92 ACLY_MOUSE ATP-citrate synthase 120 104 3% 0, 0, 7 P41216 ACSL1_MOUSE Long chain fatty acid CoA ligase 1 # 78 536 18% 0, 0, 2

Hydroxyacyl-Coenzyme A dehydrogenase/3-ketoacyl-Coenzyme β-oxidation Q64428 ECHA_RAT 83 254 9% 0, 0, 3 A thiolase/enoyl-Coenzyme A hydratase

Q99JY0 ECHB_MOUSE Trifunctional enzyme beta subunit, mitochondrial precursor 51 89 3% 0, 1, 6 P04117 FABPA_MOUSE Fatty acid-binding protein, adipocyte 15 155 33% 0, 0, 1 Lipid Metabolism P54310 LIPS_MOUSE Hormone-sensitive lipase 83 57 4% 1, 2, 7 Q8CGN5 PLIN_MOUSE Perilipin (lipid-droplet associated protein) 56 158 9% 0, 2, 9 P11762 LEG1_MOUSE Galectin-1 (lactose-binding lectin 1) 15 76 17% 0, 0, 0 P40142 TKT_MOUSE Transketolase # 68 156 9% 1, 1, 3 Other Metabolisms P56480 ATPB_MOUSE ATP synthase beta chain, mitochondria precursor * # (75) 56 54 3% 0, 2, 4 Q03265 ATPA_MOUSE ATP synthase alpha chain, mitochodria precursor # 60 86 5% 0, 0, 2 Q61753 SERA_MOUSE D-3-phosphoglcerate dehydrogenase 56 71 8% 0, 0, 2

Serine/Threonine-protein kinase MARK2 (microtubule affinity- Q05512 MARK2_MOUSE 86 88 2% 2, 5, 11 Protein kinase regulating kinase 2) # Q99N57 RAF1_MOUSE RAF proto-oncogene Serine/Threonine-protein kinase * # (53) 73 45 1% 3, 4, 10 P20029 GRP78_MOUSE 78kDa glucose-regulated protein precursor 72 185 10% 0, 0, 0 Protein processing P63017 HSP7C_MOUSE Heat shock 70 protein 8 # 71 440 19% 0, 0, 4 Stress response P35700 PRDX1_MOUSE Peroxiredoxin 1 (Thioredoxin peroxidase 2) # 22 340 39% 0, 0, 1 Nucleic acid binding P62806 H4_MOUSE Histone H4 * # (77) 11 125 29% 0, 0, 1 O70133 DHX9_MOUSE ATP-dependent RNA helicase A # 149 58 1% 0, 1, 3 P16381 PL10_MOUSE ATP-dependent RNA helicase PL10 73 71 3% 0, 2, 8 Transcription Q92499 DDX1_MOUSE DEAD-box protein 1 # 82 115 4% 0, 0, 1 O54724 PTRF_MOUSE Polymerase 1 and transcript release factor 44 154 13% 0, 0, 3 P10126 EF1A1_MOUSE Translation elongation factor 1-alpha # 51 183 8% 0, 0, 1 P29341 PABP1_MOUSE Polyadenylate-binding protein 1 # 71 137 4% 0, 0, 3 P62082 RS7_MOUSE 40S ribosomal protein S7 # 22 59 4% 1, 1, 4 Translation P62264 RS14_MOUSE 40S ribosomal protein S14 16 50 7% 0, 1, 2 P62270 RS18_MOUSE 40S ribosomal protein S18 18 96 13% 0, 1, 1 P62830 RL23_MOUSE 60S ribosomal protein L23 # 15 110 23% 0, 0, 1 Cellular trafficking P61205 ARF3_MOUSE ADP-ribosylation factor 3 20 49 6% 0, 0, 0

Basement membrane-specific heparan sulfate proteoglycan core Cell attachment Q05793 PGBM_MOUSE 398 376 3% 4, 8, 32 protein (HSPG)

P13020 GELS_MOUSE Gelsolin precursor (actin-depolymerizing factor) 86 229 9% 0, 0, 3 Actin dynamics P60710 ACTB_MOUSE Beta-actin * # (78) 42 43 4% 0, 0, 2 Q9WTI7 MYO1C_MOUSE Myosin 1C 118 226 6% 0, 1, 6

Table 3.1 14-3-3-affinity-purified proteins from insulin-treated 3T3-L1 adipocytes.

14-3-3-sepharose was mixed with insulin-stimulated lysate. After overnight incubation, the sepharose was washed and the proteins were eluted and run on a 7.5% SDS-PAGE gel. The protein bands were excised, tryptic digested and the peptides were analyzed by LC-MS/MS. The data obtained from

LC-MS/MS analysis were searched against the SwissProt database using the Mascot search algorithm.

The proteins were grouped by their functions. The molecular weight, protein score and percentage of sequence coverage of each protein were shown. “*”, 14-3-3-binding of the protein was confirmed in a

57 case-by-case study; “#”, the proteins identified in other large scale proteomic analysis of

14-3-3-binding proteins (Benzinger et al., 2005; Jin et al., 2004; Meek et al., 2004; Rubio et al., 2004).

“SCANSITE” The number of putative 14-3-3 consensus binding motif (mode 1, RSXpS/TXP) of each protein was determined using the software SCANSITE under (high, medium, low) stringency settings.

14-3-3 isoforms Protein score Sequence coverage β 682 53% γ 122 12% ε 189 14% σ 92 6% ζ 256 20% τ Not detected η 145 17%

Table 3.2 The 14-3-3-isoforms purified by 14-3-3-affinity purification. The data obtained from

LC-MS/MS analysis were searched against the SwissProt database using Mascot search algorithm. The protein score and percentage of sequence coverage of each isoform were shown

58

Figure 3.7 SDS-PAGE analysis and mass spectrometry identification of insulin-regulated

14-3-3-associated proteins. 14-3-3-sepharose (containing 450 μg of total bound 14-3-3) was mixed with 20 mg of basal and insulin-stimulated lysate. After overnight incubation, the sepharose was washed and the proteins were eluted and run on 7.5% SDS-PAGE resolving gel. The gel was stained with SYPRO Ruby gel stain. The protein bands that showed at least a 1.5-fold increase in intensity in insulin-stimulated conditions were excised, tryptic digested and identified by LC-MS/MS. The data obtained from LC-MS/MS analysis were searched against the SwissProt database using the Mascot search algorithm. The accession number for each protein was shown in blanket.

59

Figure 3.8 Subcellular localization of Myo1c in 3T3-L1 adipocytes. A. Immunofluorescence analysis of Myo1c in 3T3-L1 adipocytes. Basal and insulin-treated cells were fixed, permeabilized and stained for endogenous Myo1c with anti-Myo1c antibody (green). The scale bar represents 30 μm. B.

Western blot analysis of PM silica preparation from basal and insulin-treated 3T3-L1 adipocytes using anti-Myo1c and anti-Syntaxin4 antibodies. The experiment was performed three times and the images are from a representative experiment.

60 v. Discussion

In this study, 14-3-3-affinity purification in combination with mass spectrometry analysis was developed to identify insulin-regulated phosphoproteins in adipocytes.

Previous studies focused on using this approach to study the physiological roles of

14-3-3 in the cells. In the present study, 14-3-3 was used as a tool to identify a number of potential phosphoproteins. This study has proven that the methodology described is suitable for comparing a subset of the phosphoproteome under different physiological conditions. This approach may also be applied to identify the substrates of specific kinases.

This approach started with production and purification of GST-14-3-3. As opposed to phospho-specific antibodies, the GST-14-3-3 fusion proteins can easily be produced in bacteria and bind with high affinity to their Ser/Thr-phosphorylated ligands in vitro. A preliminary experiment was performed in which GST-14-3-3 fusion proteins were used as the affinity ligands to purify phosphoproteins from an adipocyte lysate, and the protein complexes were then captured using glutathione-sepharose. The purified proteins were then identified by mass spectrometry. Among the identified proteins, endogenous GST from the adipocytes was the most abundant (results not shown). It is believed that these endogenous GST associated with the glutathione-sepharose or the

GST-tag of the GST-14-3-3 fusion protein, even though GST may also interact with

14-3-3 as well. To overcome this problem Meek et al. (2004), who used a similar proteomic strategy to study 14-3-3-binding proteins, pre-incubated the cell lysate in a

GST column in order to remove GST-associated proteins prior to incubation with the

GST-14-3-3 column (Meek et al., 2004). However, some 14-3-3-binding proteins may also bind non-specifically to the GST column and pre-incubation may remove such proteins. To overcome this problem, in the present study the GST-tag was removed

61 from the GST-14-3-3 fusion protein by thrombin protease cleavage before coupling to sepharose. This avoided the pre-incubation step using a GST column. Hence, the present study suggested that thrombin cleavage of GST-tag is an alternative strategy to avoid GST-associated proteins. Nevertheless, Figure 3.2 and 3.4 indicate that the thrombin cleavage reaction of GST-14-3-3 fusion proteins may need to be further optimized to completely remove the GST tags. This may be achieved by increasing the incubation time or amount of thrombin protease in the cleavage reaction.

The 14-3-3 proteins were coupled covalently to the CNBr-sepharose. The

14-3-3-sepharose was then used as a tool to purify phosphoproteins from basal and insulin-treated adipocytes. The silver stained gel revealed that a number of proteins were purified from adipocytes using this approach (Figure 3.4). These results are in agreement with previous findings that 14-3-3 can associate with up to 250 proteins

(Meek et al., 2004).

The phosphorylation state of the purified proteins was confirmed by GST-14-3-3 overlay combined with treatment with alkaline phosphatase which removed phosphorylation. Strikingly numerous bands were labeled upon incubation with

GST-14-3-3 and this labeling was almost completely abolished following alkaline phosphatase treatment. This indicated that a significant number of phosphoproteins were purified using the 14-3-3-affinity purification.

Western blot of 14-3-3-associated proteins with anti-pAS antibody revealed that some of the purified proteins were Akt substrates. As mentioned in the introduction, the two

14-3-3-binding motifs, RSXpS/TXP and RXXXpS/TXP, are similar to the basophilic kinase phosphorylation consensus motifs. The examples of basophilic kinases include

Akt, PKA and CaMKII. The Akt consensus phosphorylation motif is R/KXR/KXXS/T, the PKA consensus is R/LR/LXS/TX and the CaMKII consensus is RXXS/T

62 (Benzinger et al., 2005). By analyzing the sequences of these motifs, it would be expected that some Akt, PKA and CaMKII substrates may have perfect

14-3-3-binding motifs. It is believed that combining 14-3-3-affinity purification, anti-pSer/Thr kinase substrate immunoblotting with SDS-PAGE represents a direct approach to selectively purify and label a subset of the kinase substrates. These phosphoproteins can be identified by mass spectrometry analysis of gel bands excised from a reference gel which correspond to bands detected by immunoblotting. In the present study, it has been shown that 14-3-3 is a useful tool to purify Akt substrates and potentially other basophilic kinase substrates.

Using 14-3-3-affinity purification and mass spectrometry, 38 proteins were purified and subsequently identified in this study. Recently, identification of potential in vitro

14-3-3-binding proteins in HeLa cells using yeast 14-3-3 proteins (mixture of BMH1 and BMH2) by Rubio et al. (2004) or 14-3-3ζ by Meek et al. (2004) was reported, in which 200 and 250 proteins were identified, respectively (Meek et al., 2004; Rubio et al., 2004). The number of proteins that were identified varies depending on the amount of cell extract. Rubio et al. used 200 mg while Meek et al. used 150 mg of cell extract. In the present study, only 54 mg was used. It is likely that more proteins will be identified if the quantity of cell extract is increased. Furthermore, in the present study, the overlap with the proteins identified by Rubio et al. and Meek et al. was 29% and 37%, respectively. Out of the 38 proteins identified in this study, only 11 and 14 of them were detected by Rubio et al. and Meek et al., respectively. It indicated that only a small portion of proteins identified in the present study matched with those previously reported. The discrepancy may be due to the use of different cell lines and isoform-specific differences in ligand binding. It may also result from the different experimental approaches including the use of binding and wash buffers,

63 gel stain and mass spectrometry. The differences make it difficult to compare these results. In another study, Benzinger et al. (2005) co-immunoprecipitated 117 proteins from 14-3-3σ transfected HEK293 cells (Benzinger et al., 2005). Only 2 of the proteins identified by Benzinger et al. were detected as potential 14-3-3β ligands in the present study. Moreover, Jin et al. (2004) performed co-immunoprecipitation of 4 tagged 14-3-3 isoforms to identify in vivo 14-3-3-binding proteins in HEK293 cells

(Jin et al., 2004). Only 2 of the 19 proteins identified as potential in vivo 14-3-3β ligands by Jin et al. were detected in vitro in the present study. This is presumably due to the differences in 14-3-3-isoforms, cell types and amount of cell extract used. More importantly, this may also be due to the differences between the 14-3-3-ligand associations in intact cells (in vivo) and those on columns (in vitro). Despite the similar methodology used by Benzinger et al. and Jin et al. in HEK293 cells, the in vivo 14-3-3σ-purified proteins only showed 17% overlap with the in vivo

14-3-3ζ-purified proteins. All of these works suggested that isoform-specific effects and heterodimerization complexity of 14-3-3 dimers would further enlarge the 14-3-3 interactome. It is expected that as more proteomic analysis of 14-3-3-binding proteins are conducted, more isoform- and cell line-specific interactors will be identified. In order to understand the biological roles of 14-3-3 proteins, it may be important to determine which method is most accurate from a physiological standpoint (Bridges and Moorhead, 2005). Nevertheless, as a novel approach to purify phosphoproteins, all these studies showed that 14-3-3 is an emerging tool in phosphoproteomic analysis.

The major drawback of this approach is similar to that of immunoprecipitation with phospho-specific antibodies in which not all detected proteins are phosphorylated.

Some non-phosphorylated proteins can bind to or co-precipitate with phosphoproteins.

64 Hence, the direct 14-3-3-binding of every individual protein of interest must be checked to determine if it is phosphorylated.

Although the proteins identified in this study only show limited overlap with those detected in other literature previously reported (Benzinger et al., 2005; Jin et al., 2004;

Meek et al., 2004; Rubio et al., 2004), all of these studies have purified proteins with diverse functions. They are involved in metabolism, signaling transduction, stress response, transcription, translation, intracellular trafficking and cytoskeletal dynamics.

Prominent among the 14-3-3-associated proteins identified in the present study were enzymes involved in carbohydrate and fat metabolisms. It was reported that adipocytes require a basal level of glycolysis in order to provide the glycolytic intermediate dihydroxyacetone phosphate as an intermediate in the synthesis of triacyglycerols. Hence, both glycolysis and lactate production are stimulated by insulin in adipocytes (Vasta et al., 1989). In the present study, the identification of several glycolytic enzymes in 14-3-3-pulldown may indicate the 14-3-3 scaffolding role in the spatial co-localization of a metabolic pathway. The enzymes involved in the fourth (fructose-bisphosphate aldolase), sixth (glyceraldehyde-3-phosphate dehydrogenase), and tenth (pyruvate kinase) step of glycolysis and lactate-dehydrogenase were all affinity-purified in this study. Furthermore, the association of 14-3-3 with both glyceraldehyde-3-phosphate dehydrogenase and lactate-dehydrogenase was increased by insulin (Figure 3.6). Although direct 14-3-3 interaction of this glycolytic complex has not yet been confirmed, spatial co-localization of enzymes may represent an exciting new mechanism by which insulin regulates glycolysis in adipocytes. The 14-3-3 proteins may act as scaffold to anchor these glycolytic proteins within close proximity of each other. Besides glycolytic enzymes, fatty acid synthase and ATP-citrate synthase, which are involved

65 in fatty acid synthesis, were also purified in this study. Figure 3.6 shows that the

14-3-3-associations of these two enzymes were increased by insulin. However, such interactions have not yet been studied in further detail.

14-3-3 is known to be involved in regulation of signal transduction. One of the protein kinases identified in this study, Raf or MAP-kinase-kinase-kinase, is known to interact with 14-3-3 directly (Light et al., 2002; Roy et al., 1998; Thorson et al., 1998).

Microtubule-affinity regulating kinase 2 was identified as a 14-3-3-associated protein in this study. Since it contains two high-affinity 14-3-3-binding motifs, it is likely that it may bind 14-3-3 directly. The mechanism of how 14-3-3 may regulate this protein kinase has not yet been studied.

Besides cytosolic proteins, 14-3-3 also associated with proteins in different subcellular compartments. It associated with nuclear proteins (histone, RNA helicase and polymerase) and proteins localized in the endoplasmic reticulum (78 kDa glucose-regulated protein precursor). Several mitochondrial enzymes involved in

β-oxidation and ATP synthesis were also purified. This suggested that 14-3-3 can be present in all of these subcellular compartments and may function to localize the proteins in these organelles once the proteins were synthesized in the cytoplasm.

This study also purified some other proteins that were previously shown to be

14-3-3-binding proteins in case-by-case studies. The plant ATP synthase β chain was previously identified as a phosphorylation-dependent 14-3-3-binding protein. The

14-3-3-binding of ATP synthase β chain at the mitochondrial inner membrane results in inactivation of the enzyme and prevents it from functioning in reverse to hydrolyze

ATP when ATP levels are high. This may represent a similar regulatory mechanism of mammalian ATP synthase since the 14-3-3-binding region of the β chain is conserved in mammalian cells (Bunney et al., 2001; Moorhead et al., 1999). Histone is another

66 known 14-3-3-binding protein. During chromatin remodeling, it is phosphorylated and the binding of 14-3-3 prevents the dephosphorylation of histone by phosphatase

(Chen and Wagner, 1994).

Insulin-dependent actin remodeling is known to be an important event for transport of

GLUT4 vesicles to the cell surface. Beta-actin was shown to be a direct

14-3-3-binding protein in a previous study and the interaction plays a role in cell division and apoptosis (Chen and Yu, 2002). In the present study, β-actin was purified and the 14-3-3-binding of actin was increased by insulin (Figure 3.6). Moreover, some actin-associated proteins were also purified in this study. These include gelsolin

(actin-depolymerizing factor), myosin 1C (Myo1c) and fructose-bisphosphate aldolase A which is a glycolytic enzyme that has multiple functions. In particular,

Myo1c and fructose-bisphosphate aldolase A have previously been implicated in linking the GLUT4 vesicles to the actin filaments and the transport of GLUT4 vesicles upon insulin stimulation in 3T3-L1 adipocytes (Bose et al., 2002; Bose et al.,

2004; Kao et al., 1999). These results suggested that 14-3-3 may be involved in insulin-dependent actin remodeling and actin-based transport of GLUT4 vesicles. The present study also showed that Myo1c is localized at the PM where it interacts with

14-3-3. It is particularly relevant to insulin-stimulated GLUT4 translocation since the

PM is a major insulin-regulatory node. Hence, Myo1c is possibly an essential insulin-regulated phosphoprotein that involves in GLUT4 translocation.

These data have revealed a broad range of 14-3-3-associated proteins with numerous potential functions. The binding of 14-3-3 to phosphorylated proteins connects signaling pathways for regulations of even greater diversity of target proteins. In particular, this study showed that 14-3-3 proteins play important roles in the regulation of metabolic enzymes and actin dynamics. Further characterization of these

67 interactions and the cellular consequences may improve our current understanding of many cellular processes.

68 Chapter 4

CaMKII-Mediated Phosphorylation of the Myosin Motor Myo1c Is Required for

Insulin-Stimulated GLUT4 Translocation in Adipocytes

69 i. Abstract

The unconventional myosin Myo1c has been implicated in insulin-regulated GLUT4 translocation to the plasma membrane in adipocytes. I show that Myo1c undergoes insulin-dependent phosphorylation at Ser701. Phosphorylation was accompanied by enhanced 14-3-3 binding and reduced calmodulin binding. Recombinant CaMKII phosphorylated Myo1c in vitro and siRNA knockdown of CaMKIIδ abolished insulin-dependent Myo1c phosphorylation in vivo. CaMKII activity was increased upon insulin treatment and the CaMKII inhibitors, CN21 and KN-62 or the Ca2+ chelator BAPTA-AM blocked insulin-dependent Myo1c phosphorylation and insulin-stimulated glucose transport in adipocytes. Myo1c ATPase activity was increased after CaMKII phosphorylation in vitro and after insulin-stimulation of

CHO/IR/IRS-1 cells. Expression of wild type Myo1c, but not S701A or ATPase dead mutant K111A, rescued the inhibition of GLUT4 translocation by siRNA-mediated

Myo1c knockdown. These data suggest that insulin regulates Myo1c function via

CaMKII-dependent phosphorylation and these events play a role in insulin-regulated

GLUT4 trafficking in adipocytes likely involving Myo1c motor activity. ii. Introduction

As described in Chapter 3, the Myosin Motor Myo1c was identified as a novel insulin-regulated phosphoprotein in adipocytes. This is a particularly exciting observation because Myo1c is an excellent candidate to regulate intracellular vesicle transport (Durrbach et al., 2000). Myo1c is a 118 kDa molecular motor protein, which converts chemical energy into motion, and belongs to the myosin-I family. This family of unconventional myosins, unlike the conventional myosins, is single-headed and does not self-associate into bipolar filaments (Pollard and Korn, 1973). Myo1c contains three domains. The first 700 amino acids form a globular head domain which

70 hydrolyzes ATP and binds actin. The alternative cycle of ATP hydrolysis and binding controls the binding to actin (Figure 4.1). When free of ATP, Myo1c binds strongly to actin. ATP-binding reduces the affinity of Myo1c to actin and Myo1c alters its conformation to a state where it can exert force. ATP is then rapidly hydrolyzed by

Myo1c, and the subsequent release of ADP and inorganic phosphate triggers a tight interaction of Myo1c with actin. These repeating cycles result in sliding of Myo1c along the actin filament (Yount et al., 1995; Zhu et al., 1996). The next 91 amino acids of Myo1c comprise a neck domain, containing 4 calmodulin binding domains. The calmodulin-binding of Myo1c is Ca2+-dependent in which the binding of Ca2+ to calmodulin causes its dissociation from Myo1c. It is believed that the Ca2+-binding negatively regulates the ATPase activity and the cargo interaction of Myo1c in general

(Stoffler and Bahler, 1998; Wolenski et al., 1993; Zhu et al., 1996). However, the detail consequences underlying the dissociation of each of the 4 calmodulin molecules are poorly understood. The last 237 amino acids of Myo1c form a tail domain which binds negatively charged phospholipids, such as phosphatidylserine and phosphatidylinositol-4,5-bisphosphate (PIP2), at biological membranes (Barylko et al.,

2000). This domain plays a role in connecting the tail domain of Myo1c to cargo or organelle (Gillespie and Cyr, 2004).

Phosphorylation sites within Myo1c are poorly understood. Myo1c was shown to be phosphorylated by PKCγ in vitro in rat liver cells. Phosphorylation by PKCγ reduces the binding of calmodulin to Myo1c. The exact phosphorylation site is unknown but it is believed to be present in the tail domain (Williams and Coluccio, 1995). However, whether Myo1c is phosphorylated by any kinases in adipocytes is still unknown.

As PM is believed to be a major insulin regulatory node for GLUT4 translocation, it is particularly exciting that Myo1c acts principally at the PM playing an important role

71 in a diverse set of physiological processes. These include the translocation of ion channels within the membranes of stereocilia in vestibular hair cells of the ear (Batters et al., 2004; Phillips et al., 2006; Stauffer et al., 2005), cytoskeletal rearrangement and membrane protrusion in neuronal growth cones (Wang et al., 2003) and antidiuretic hormone-regulated delivery of sodium channels to the brush border of kidney epithelial cells (Wagner et al., 2005). Myo1c is also involved in the insulin-regulated delivery of GLUT4 to the PM in adipocytes (Bose et al., 2002; Bose et al., 2004).

Reduced expression of Myo1c in adipocytes using siRNA or overexpression of the

Myo1c tail domain, as a dominant inhibitor, both impair insulin stimulated translocation of GLUT4 to the PM (Bose et al., 2002; Huang et al., 2005). Ultrafast miscroscopic analysis revealed that Myo1c enhances localized membrane ruffling at the PM and insulin stimulates the mobilization of GSVs to these regions (Bose et al.,

2004). However, the molecular mechanism of insulin-dependent regulation of Myo1c is poorly understood.

In this study I have shown that CaMKII phosphorylates Myo1c at S701 in response to insulin, thereby reducing calmodulin-binding and stimulating ATPase activity of

Myo1c. These events together play an important role in insulin-regulated GLUT4 translocation in adipocytes.

72

Figure 4.1 Mechanical cycle of myosin. A. When ADP and free phosphate bind to the head domain, the myosin binds tightly to the actin filament. B. Release of ADP and free phosphate triggers a conformational change and the myosin head bends. C. Binding of ATP reduces the interaction with the actin filament. D. The myosin head hydrolyses ATP to ADP and free phosphate, and attaches to the actin filament.

73 iii. Methods

A. In vitro and in vivo phosphorylation assay

For in vivo γ32P-labeling, 3T3-L1 adipocytes were incubated in serum-free and phosphate-free DMEM for 1 h followed by serum-free DMEM containing 2.7 mCi

[32P]orthophosphate per 10-cm dish for 3 h prior to insulin stimulation. In vitro

CaMKII phosphorylation assay was performed as described (Konstantopoulos et al.,

2007) with modifications. To activate GST-CaMKIIα fusion protein (Cell Signaling

Technology), the kinase was preincubated in kinase buffer (50 mM Tris, pH 7.5, 10 mM MgCl2, 2 mM DTT, 0.1 mM EDTA, 2 mM CaCl2, 100 μM ATP, 1.2 μM calmodulin) for 10 min at 30 °C. Inactive GST-CaMKIIα, which was preincubated in kinase buffer in the absence of CaCl2 and calmodulin, was used as control.

Immunoprecipitated Myo1c was incubated with kinase buffer only, active

GST-CaMKIIα or inactive GST-CaMKIIα for 5 min at 30 °C. For in vitro 32P-labeling of EYFP-Myo1c, 3 μCi [γ32P]ATP was added to each sample prior to CaMKII phosphorylation. The reaction was stopped by boiling in sample buffer. The phosphorylation of Myo1c and EYFP-Myo1c was analyzed by 14-3-3 overlay assay and autoradiography respectively. For measuring CaMKII kinase activity in 3T3-L1 adipocyte lysates, the assay was carried out directly on the agarose-beads bound with immunoprecipitated CaMKII using SignaTECT CaMK Assay System (Promega) according to the manufacturer’s protocol. The 32P-labeling of the peptide substrate was analyzed by scintillation counting.

B. In vitro ATPase assay

For assay of EYFP-Myo1c immunoprecipitated from CHO/IR/IRS-1 cells, the

Myo1c-bound beads were washed three times with ATPase reaction buffer (30 mM

Tris, pH 7.5, 20 mM KCl, 1 mM MgCl2, 1 μM CaCl2) and the ATPase reactions were

74 carried out directly on the beads in the ATPase reaction buffer containing 100 μM actin, 500 μM ATP and 2 μCi[γ32P]ATP at 37˚C. For ATPase assay after in vitro

CaMKII phosphorylation of Myo1c, the CaMKII was removed and the beads were washed three times with ATPase reaction buffer. The beads were incubated with either

46 ng GST or 100 ng GST-14-3-3 for 30 min. The beads were washed and the ATPase reactions were carried out in the presence of 1 μM tat-CN21. The reactions were stopped by placing on ice. Three microliters of each reaction were spotted onto polyethyleneimine cellulose TLC plate and the inorganic phosphate was separated as described (Bultman et al., 2005). The amounts of [γ32P]Pi and [γ32P]ATP were measured by autoradiography and quantified by densitometry. iv. Results

Insulin regulates Myo1c phosphorylation and 14-3-3-binding

The in vitro interaction of Myo1c with 14-3-3 was confirmed by immunoblotting using a Myo1c specific monoclonal antibody (Figure 4.2). Insulin increased the amount of Myo1c associated with 14-3-3 compared to that observed under basal conditions and this effect was abolished by the PI3K inhibitor Wortmannin. Similar data were obtained when EYFP-tagged Myo1c and GST-tagged 14-3-3β were co-expressed in CHO/IR/IRS-1 cells indicating that the insulin-dependent association of Myo1c and 14-3-3 occurs in vivo (Figure 4.3). One possible explanation for the insulin-dependent interaction between Myo1c and 14-3-3 is that Myo1c may interact with another protein that is phosphorylated in response to insulin. To exclude this possibility I performed a GST-overlay assay using GST-14-3-3 (Figure 4.4).

EYFP-Myo1c was immunoprecipitated from CHO/IR/IRS-1 cells, subjected to

SDS-PAGE and transferred to PVDF membrane and incubated with GST-14-3-3.

There was a direct association between 14-3-3 and Myo1c using this approach and I

75 observed increased binding of 14-3-3 to Myo1c isolated from insulin treated cells compared to non-insulin treated cells. To prove that this interaction was due to Myo1c phosphorylation, the PVDF membrane was incubated with alkaline phosphatase prior to the overlay and this completely abolished the interaction. These findings demonstrate that insulin stimulates Myo1c phosphorylation in adipocytes resulting in its interaction with 14-3-3 proteins.

Myo1c is phosphorylated on S701 in response to insulin

Analysis of the Myo1c primary amino acid sequence by PhosphoSite identified multiple potential phosphorylation and 14-3-3-binding sites (Figure 4.5) (Hornbeck et al., 2004). Site directed mutagenesis was used to map putative phosphorylation sites in Myo1c. Residues S142, T564 and S701 were initially chosen for site directed mutagenesis. S142 has a high stringency mode 2 consensus 14-3-3-binding motif

(RXXXpS/TXP), T564 and S701 were identified as potential mode 1 consensus

14-3-3-binding motifs (RSXpS/TXP) and T564 is also a potential Akt phosphorylation motif. Each of these sites was mutated to Ala and mutants were expressed in CHO/IR/IRS-1 cells. As shown in Figure 4.6, Myo1c WT and the S142A and T564A mutants interacted to a similar extent with 14-3-3 in an insulin-regulated manner. However, mutation of S701 completely inhibited the 14-3-3/Myo1c interaction. Based on these data I conclude that S701 is the major insulin-regulated phosphorylation site in Myo1c that regulates 14-3-3-binding.

Calcium/Calmodulin-dependent protein Kinase II (CaMKII) phosphorylates Myo1c.

I next sought to identify the Myo1c kinase. In silico analysis revealed S701 is a potential phosphorylation site for PKA, PKC and CaMKII. As shown in Figure 4.7A, the PKCζ pseudo-substrate and the Akt inhibitor had no significant effect on insulin-stimulated 14-3-3-binding to Myo1c. In contrast, Wortmannin and KN-62

76 caused a substantial reduction in the amount of 14-3-3 co-immunoprecipitated with

Myo1c from insulin-treated 3T3-L1 adipocytes (Figure 4.7B). To confirm the role of

CaMKII in Myo1c phosphorylation I utilized a highly specific cell-penetrating

CaMKII inhibitor tat-CN21 (Vest et al., 2007). The insulin-stimulated interaction between 14-3-3 and Myo1c was significantly inhibited by tat-CN21 whereas the control peptide, tat-ctrl, was without effect (Figure 4.7C). To confirm the status of endogenous Myo1c phosphorylation, 3T3-L1 adipocytes were labeled to equilibrium with [32P]orthophosphate, and Myo1c was immunoprecipitated following incubation of cells with or without kinase inhibitors and insulin. Insulin increased the amount of phosphorylated Myo1c in adipocytes in a Wortmannin and tat-CN21-sensitive manner

(Figure 4.8), consistent with the in vitro 14-3-3 pulldown data.

CaMKII is activated in vivo by increased cytosolic calcium, which triggers calmodulin binding to CaMKII. KN-62 prevents the activation of CaMKII by inhibiting calmodulin binding (Tokumitsu et al., 1990). Intriguingly, chelation of cytosolic calcium with BAPTA-AM inhibits insulin-stimulated glucose transport and

GLUT4 translocation in adipocytes (Whitehead et al., 2001). Hence, I next set out to establish if BAPTA-AM inhibits insulin-stimulated Myo1c phosphorylation and

14-3-3-binding. BAPTA-AM almost completely inhibited insulin stimulated Myo1c phosphorylation as indicated by impaired binding to 14-3-3 (Figure 4.9A). Conversely, the Ca2+ ionophore A23187 increased 14-3-3 binding to Myo1c to a similar extent as insulin (Figure 4.9B). To further investigate the role of CaMKII in phosphorylation of

Myo1c I utilized an siRNA construct targeting CaMKIIδ, a major isoform expressed in 3T3-L1 adipocytes (Figure 4.10). Reduction of CaMKIIδ protein expression by

~90% percent resulted in a significant reduction in insulin-stimulated Myo1c binding to 14-3-3 whereas in cells transfected with a control siRNA I observed an ~2.5 fold

77 increase in Myo1c binding to 14-3-3 with insulin. Based on these results, I propose that CaMKII mediates the insulin-stimulated phosphorylation of Myo1c at S701.

Next, I performed an in vitro phosphorylation assay using recombinant CaMKII and endogenous Myo1c immunoprecipitated from 3T3-L1 adipocytes. An insulin-dependent and KN-62-sensitive Myo1c/14-3-3 interaction was observed in the absence of recombinant CaMKII (Figure 4.11A). Incubation with active CaMKII, but not inactive CaMKII, enhanced Myo1c/14-3-3-binding under all conditions. Moreover,

CaMKII phosphorylation of Myo1c in vitro was sensitive to the CaMKII inhibitor tat-CN21, but not to the control peptide tat-ctrl (Figure 4.11B).

To confirm that CaMKII phosphorylates Myo1c directly on S701, EYFP-Myo1c WT and the S701A mutant were immunoprecipitated from CHO/IR/IRS-1 cells and incubated with recombinant CaMKII in the presence of [γ32P]ATP. Phosphorylation of

Myo1c WT but not the Myo1c S701A mutant was observed with CaMKII (Figure

4.11C). Hence, I conclude that CaMKII phosphorylates Myo1c both in vitro and in vivo at S701 and that the upstream regulation of CaMKII relies upon

Ca2+/CaM-dependent pathways.

Insulin induces CaMKII phosphorylation and activation in adipocytes

CaMKII is activated following autophosphorylation at T286 (Colbran et al., 1988). I next sought to determine if insulin activates CaMKII in adipocytes using an anti-pT286 CaMKII antibody. Phosphorylated CaMKII was detected in unstimulated cells and insulin increased CaMKII phosphorylation in a dose- and time-dependent manner (Figure 4.12A and B). I also observed increased CaMKII kinase activity following incubation of adipocytes with insulin (Figure 4.12C). These data indicate that CaMKII is activated by insulin in 3T3-L1 adipocytes. Moreover, KN-62 and tat-CN21 inhibited insulin-stimulated glucose transport by 63% and 46%, respectively

78 (Figure 4.13), consistent with an important role for this kinase in insulin action.

The effect of phosphorylation on calmodulin binding of Myo1c

It is of interest that S701 is located at the start of the Myo1c neck domain proximal to the four calmodulin-binding IQ motifs (Figure 4.14). Increased calcium displaces calmodulin from one or more of these IQ domains (Gillespie and Cyr, 2002) and so I reasoned that phosphorylation may play a role in calmodulin binding.

Phosphorylation at S701 may either affect calmodulin binding to IQ1 or calmodulin binding may regulate Myo1c phosphorylation at S701. To test these possibilities, I generated an IQ1A mutant in which four of the five residues of the IQ consensus sequence of IQ1 were mutated to Ala (Figure 4.14). This mutation was previously shown to prevent calmodulin binding to the IQ domain (Phillips et al., 2006). A significant interaction was observed between calmodulin and Myo1c WT in non-stimulated cells whereas this interaction was reduced by 45% following insulin stimulation (p<0.05 versus basal, Figure 4.15). Mutation of the IQ1 domain resulted in a ~31% reduction in calmodulin binding to Myo1c (p<0.05 versus WT/basal) compared to WT Myo1c. The fact that calmodulin binding was not completely ablated in the IQ1 mutant suggests that this domain is not required for calmodulin binding to the adjacent IQ domains. Whereas calmodulin binding to WT Myo1c was reduced with insulin this was not observed for the S701A and IQ1A mutants (Figure 4.15).

This suggests that Myo1c S701 phosphorylation and calmodulin binding to IQ1 are mutually exclusive events raising the question of whether phosphorylation regulates calmodulin binding or vice versa. If calmodulin binding to IQ1 is required for Myo1c phosphorylation then the IQ1A mutant should display reduced insulin-stimulated phosphorylation and 14-3-3-binding. However, dissociation of calmodulin from IQ1 neither inhibited nor stimulated phosphorylation at S701 (Figure 4.16). In contrast,

79 insulin did not stimulate 14-3-3-binding to the double mutant S701AIQA, which contains both the phosphorylation site mutation and the IQ1 mutation. These results suggest that calmodulin binding to IQ1 is not required for phosphorylation at S701.

Although insulin may also regulate calmodulin binding to other IQ domains, these data suggest that insulin inhibits calmodulin binding to the first IQ domain in Myo1c via phosphorylation at S701.

To determine if CaMKII activity may regulate the insulin-dependent dissociation of calmodulin and Myo1c, I pretreated adipocytes with KN62 or tat-CN21. Both compounds, as well as Wortmannin, inhibited the insulin-dependent reduction of calmodulin binding to Myo1c (Figure 4.17).

The effect of phosphorylation on the ATPase activity of Myo1c

As the myosin motor family is responsible for actin-based motility, I next examined the role of phosphorylation on Myo1c ATPase activity. Endogenous Myo1c was immunoprecipitated from basal adipocytes and incubated with recombinant CaMKII.

CaMKII was removed and Myo1c-bound sepharose was extensively washed with reaction buffer. [γ32P]ATP hydrolysis was assayed in the presence of tat-CN21 to abolish the activity of any remaining CaMKII. An equivalent amount of Myo1c was present in all conditions (Figure 4.18A). CaMKII, but not inactive CaMKII, stimulated Myo1c ATPase activity by 3-4 fold (Figure 4.18). Inclusion of GST-14-3-3 in this assay had no further effect on Myo1c ATPase activity. I next explored if insulin stimulation affects Myo1c ATPase activity in vivo by measuring the ATPase activity of EYFP-Myo1c WT immunoprecipitated from transiently transfected CHO/IR/IRS-1 cells with or without insulin stimulation. Insulin increased the ATPase activity of

Myo1c WT by 2 fold (Figure 4.19) whereas a Myo1c ATPase mutant (EYFP-Myo1c

K111A), only showed background hydrolysis of ATP though it was phosphorylated in

80 response to insulin. These data suggest that insulin-stimulated phosphorylation of

Myo1c regulates its motor activity and 14-3-3 binding does not appear to play a significant role in this effect at least in vitro.

The role of phosphorylation of Myo1c on GLUT4 translocation

I next set out to establish if Myo1c phosphorylation plays a role in GLUT4 translocation using a knock-in approach. An siRNA construct was designed to target the untranslated region of Myo1c mRNA. Expression of this siRNA in adipocytes resulted in a substantial reduction in endogenous Myo1c levels (~85%) (Figure 4.20).

In agreement with previous studies (Bose et al., 2002), siRNA-mediated reduction in

Myo1c in adipocytes significantly reduced insulin-stimulated GLUT4 translocation in

3T3-L1 adipocytes (Figure 4.20). Re-expression of WT Myo1c in cells expressing the

Myo1c siRNA rescued the block in GLUT4 translocation (Figure 4.21). Strikingly, re-expression of the Myo1c S701A mutant was unable to rescue the defect in insulin-stimulated GLUT4 translocation observed in cells expressing the Myo1c siRNA. These data indicate that Myo1c phosphorylation at S701 plays an essential role in insulin-stimulated GLUT4 translocation in vivo.

As described above, insulin-dependent Myo1c phosphorylation/14-3-3-binding triggers release of calmodulin from the first and possibly additional IQ domains in

Myo1c. Hence, this raises the question of whether phosphorylation of Myo1c per se plays an essential role in insulin action or if the major effect is mediated via calmodulin release. To explore this relationship in more detail I first examined the consequences of reintroducing the Myo1c calmodulin binding mutant into adipocytes lacking endogenous Myo1c (Figure 4.21). Intriguingly, upon insulin stimulation of cells expressing the IQ1A mutant insulin-stimulated GLUT4 translocation was partially restored. However, since calmodulin is not displaced from the Myo1c S701A

81 mutant in response to insulin this is consistent with a possible role for both phosphorylation/14-3-3-binding and calmodulin release in the function of Myo1c in insulin-mediated GLUT4 translocation. To further elucidate the importance of both steps I used the S701AIQ1A double mutant. If phosphorylation/14-3-3-binding simply regulates Myo1c function via displacement of calmodulin, then the S701AIQ1A mutant should phenocopy the IQ1A mutant. Intriguingly, that was not the case pointing toward a role of phosphorylation/14-3-3 independent of calmodulin displacement as the important insulin-regulated step. Notably, expression of the IQ1A mutant in adipocytes resulted in a significant increase in non-insulin dependent

GLUT4 translocation (Figure 4.21). As shown in Figure 4.16 this mutant does not display increased phosphorylation under basal conditions suggesting that regulated calmodulin binding to Myo1c per se may play an additional role in this process.

Hence, these data suggest that both calmodulin binding in the basal state and phosphorylation after insulin-stimulation may regulate the function of Myo1c in

GLUT4 translocation.

To explore the relationship between Myo1c ATPase activity and GLUT4 translocation in adipocytes, I reintroduced the ATPase catalytic domain mutant in Myo1c knockdown adipocytes. Re-expression of Myo1c K111A was unable to rescue the defect in insulin-stimulated GLUT4 translocation mediated by knockdown of Myo1c

(Figure 4.21). This suggests that Myo1c ATPase activity is required for insulin-stimulated GLUT4 translocation in adipocytes.

82

Figure 4.2 in vitro association of 14-3-3 and Myo1c in 3T3-L1 adipocytes. Cell lysate from either basal, insulin-treated or Wortmannin/insulin-treated (100 nM Wortmannin for 10 min prior to insulin stimulation) 3T3-L1 adipocytes was subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-Myo1c and anti-14-3-3 antibodies. The experiment was performed three times and the images are from a representative experiment.

83

Figure 4.3 in vivo association of 14-3-3 and Myo1c in CHO/IR/IRS-1 cells. CHO/IR/IRS-1 cells were transiently cotransfected with EYFP-Myo1c and GST-14-3-3β. Cell lysate from either basal, insulin-treated or Wortmannin/insulin-treated (100 nM Wortmannin for 10 min prior to insulin stimulation) CHO/IR/IRS-1 cells were subjected to pulldown with glutathione-sepharose or immunoprecipitation with anti-GFP antibody. Input cell extracts, pulldowns and immunoprecipitates were analyzed by Western blot with the use of anti-GFP and anti-GST antibodies to detect

EYFP-Myo1c and GST-14-3-3β respectively. The experiment was performed three times and the images are from a representative experiment.

84

Figure 4.4 Direct and phosphorylation-dependent association of 14-3-3 and Myo1c.

CHO/IR/IRS-1 cells were transiently transfected with EYFP-Myo1c. Cell lysate from either basal or

insulin-treated cells was subjected to immunoprecipitation with anti-GFP and IgG antibodies. Input cell

extracts and immunoprecipitates were analyzed by Western blotting with the use of anti-GFP antibody

(upper panel). The PVDF membrane containing immunoprecipitated Myo1c was incubated with or without alkaline phosphatase in phosphatase buffer overnight at 37 °C. The membrane was probed for binding to either GST or GST-14-3-3 with the use of anti-GST antibody (lower panel). The experiment was performed three times and the images are from a representative experiment.

85

Figure 4.5 In silico analysis of Myo1c 14-3-3-binding and phosphorylation motifs. T564 and S701 were detected as potential 14-3-3-binding motifs (mode 1, RSXpS/TXP); S148 was identified manually as a potential perfect 14-3-3-binding motif (mode 2, RXXXpS/TXP). These sites were then scanned for basophilic kinase phosphorylation motifs. T564 were identified as potential Akt phosphorylation motifs

(R/KXR/KXXS/T); S701 was identified as potential PKA phosphorylation motifs (R/LR/LXS/T), PKC phosphorylation motif (S/TXR/K) and calmodulin-dependent kinase phosphorylation motif

(R/KXXS/T). The amino acid sequences of potential 14-3-3-binding motifs are underlined and the potential phosphorylation sites are in red. Scanning was done under medium stringency settings in

PhosphoSite program.

86

Figure 4.6 Mapping Myo1c phosphorylation sites. CHO/IR/IRS-1 cells were transiently transfected with either EYFP (empty vector), EYFP-Myo1c WT, S142A, T564A or S701A. Cell lysates from either basal or insulin-treated cells were subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-GFP antibody, which detects EYFP-Myo1c, and anti-14-3-3 antibody. The experiment was performed three times and the images are from a representative experiment.

87

Figure 4.7 Effect of different kinase inhibitors on 14-3-3-binding of Myo1c. A. Effect of different kinase inhibitors on in vitro 14-3-3-binding of Myo1c. 3T3-L1 adipocytes were pretreated with either

100 nM Wortmannin (W) for 10 min, 50 μM LY294002 (L) for 30 min, 5 μM PKCζ pseudosubstrate (P) for 1 h, 10 μM KN-62 (K) for 45 min or 5 μM Akt inhibitor (A) for 30 min prior to insulin stimulation.

88 Cell lysate from either basal, insulin-treated or kinase inhibitors/insulin-treated 3T3-L1 adipocytes was subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-Myo1c, anti-14-3-3 and anti-pAkt (pS473) antibodies. The experiment was performed three times and the images are from a representative experiment. B. Effect of Wortmannin and KN-62 on in vivo 14-3-3-binding of Myo1c. 3T3-L1 adipocytes were pretreated with either 100 nM

Wortmannin for 10 min or 10 μM KN-62 for 45 min for 30 min prior to insulin stimulation. Cell lysate from either basal, insulin-treated or kinase inhibitors/insulin-treated 3T3-L1 adipocytes was subjected to immunoprecipiration using anti-Myo1c antibody and irrelevant antibodies (ctrl-IP). Input cell extracts and immunoprecipitates were analyzed by Western blot with the use of anti-Myo1c and anti-14-3-3 antibodies. The experiment was performed three times and the images are from a representative experiment. C. Effect of CaM-KIIN-derived peptides, tat-CN21, on in vitro

14-3-3-binding of Myo1c. 3T3-L1 adipocytes were pretreated with either 5 μM tat-CN21 or control

peptide (tat-ctrl) for 30 min prior to insulin stimulation. Cell lysate from either basal, insulin-treated or

inhibitor/insulin-treated 3T3-L1 adipocytes was subjected to 14-3-3-pulldown. Input cell extracts and

pulldowns were analyzed by Western blot with the use of anti-Myo1c and anti-14-3-3 antibodies. The

experiment was performed three times and the images are from a representative experiment.

89

Figure 4.8 in vivo 32P labeling of Myo1c in 3T3-L1 adipocytes. 3T3-L1 adipocytes were labeled with

[32P]orthophosphate. Cells were incubated with insulin or kinase inhibitors as indicated. Cell lysates

were subjected to immunoprecipitation with anti-Myo1c and irrelevant antibodies (ctrl-IP).

Immunoprecipitates were analyzed by autoradiography and immunoblotting with the use of anti-Myo1c

antibody. The experiment was performed two times and the images are from a representative

experiment.

90

Figure 4.9 Effect of Ca2+ chelator and ionophore on 14-3-3-binding of Myo1c. A. Effect of Ca2+

chelator, BAPTA-AM, on in vitro 14-3-3-binding of Myo1c. 3T3-L1 adipocytes were pretreated with

50 μM BAPTA-AM for 30 min prior to insulin stimulation. Cell lysate from either basal, insulin-treated

or BAPTA-AM/insulin-treated 3T3-L1 adipocytes was subjected to 14-3-3-pulldown. Input cell

extracts and pulldowns were analyzed by Western blot with the use of anti-Myo1c and anti-14-3-3

antibodies. The experiment was performed three times and the images are from a representative

experiment. B. Effect of Ca2+ ionophore, A23187, on in vitro 14-3-3-binding of Myo1c under basal condition. 3T3-L1 adipocytes were treated with 5 μM A23187 for 15 min. Cell lysate from either basal, insulin-treated or A23187-treated 3T3-L1 adipocytes was subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-Myo1c and anti-14-3-3 antibodies. The experiment was performed three times and the images are from a representative experiment.

91

Figure 4.10 Effect of siRNA-mediated CaMKII knockdown on in vitro 14-3-3-binding of Myo1c.

3T3-L1 adipocytes were electroporated with either scrambled or CaMKIIδ-targeted siRNA. After 48 h, cell lysate from either basal or insulin-treated cells was subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-CaMKII, anti-Myo1c and

anti-14-3-3 antibodies. The experiment was performed three times and the images are from a

representative experiment.

92

Figure 4.11 In vitro CaMKII phosphorylation assay. A. in vitro CaMKII phosphorylation assay of

Myo1c. 3T3-L1 adipocytes were either untreated, treated with KN-62 (10 μM for 1 h) only, insulin only or pretreated with KN-62 (10 μM for 45 min) prior to insulin stimulation. Cell lysates were subjected to immunoprecipitation with anti-Myo1c antibody. Input cell extracts were analyzed by

Western blot with the use of anti-Myo1c and anti-Actin antibodies (upper panel). The immunocomplex-bound Protein G-separose was incubated with kinase assay buffer containing either

93 no GST-CaMKII, inactive GST-CaMKII (I) or active GST-CaMKII (A). The kinase reactions were stopped by boiling in sample buffer, resolved by SDS-PAGE and transferred to PVDF membrane. The amounts of GST-CaMKII were detected by Western blot with anti-GST antibody. To detect the amounts of Myo1c immunoprecipitated in each condition, the PVDF membrane containing immunoprecipitated

Myo1c was stained with SYPRO Ruby protein blot stain. The membrane was then probed for binding to GST-14-3-3 with the use of anti-GST antibody to detect the in vitro 14-3-3-binding of immunoprecipitated Myo1c (lower panel). The experiment was performed three times and the images are from a representative experiment. B. Effect of tat-CN21 on in vitro CaMKII phosphorylation assay of Myo1c. Cell lysate from basal 3T3-L1 adipocytes were subjected to immunoprecipitation with anti-Myo1c antibody and were subjected to in vitro CaMKII phosphorylation in the presence of

[γ32P]ATP and either 1 μM tat-CN21 or 1 μM tat-ctrl. Immunoprecipitates were analyzed by autoradiography and Western blot with the use of anti-Myo1c antibody. The experiment was performed two times and the images are from a representative experiment. C. in vitro CaMKII phosphorylation assay of Myo1c WT and S701A. CHO/IR/IRS-1 cells were transiently transfected with either EYFP

(empty vector), EYFP-Myo1c S701A or EYFP-Myo1c WT. Cell lysate from basal CHO/IR/IRS-1 cells were subjected to immunoprecipitation with anti-GFP antibody and were subjected to in vitro CaMKII phosphorylation in the presence of [γ32P]ATP. Immunoprecipitates were analyzed by autoradiography

and Western blot with the use of anti-GFP antibody. The experiment was performed two times and the

images are from a representative experiment.

94

Figure 4.12 Insulin stimulates CaMKII activation in 3T3-L1 adipocytes. A. Dose response of

insulin signaling. 3T3-L1 adipocytes were serum-starved and stimulated with insulin at the indicated

concentrations for 20 min. The cell extracts were analyzed by Western blot with the use of

anti-pCaMKII (T286), anti-pAkt (pS473) and anti-14-3-3 antibodies. The experiment was performed

three times and the images are from a representative experiment. B. Time course of insulin signaling.

3T3-L1 adipocytes were serum-starved and stimulated with 100 nM insulin for 1 to 20 min as indicated.

The cell extracts were analyzed by Western blot with the use of anti-pCaMKII (T286), anti-pAkt

95 (pS473) and anti-14-3-3 antibodies. The experiment was performed three times and the images are from a representative experiment. C. in vitro CaMKII kinase activity assay. 3T3-L1 adipocytes were either untreated, treated with insulin, or with 5 μM A23187 for 20 min. Cell lysates were subjected to immunoprecipitation with anti-CaMKII antibody and the kinase activity assay was preformed directly on the agarose beads by measuring the 32P-labeling of the peptide substrate. The ratio of autonomous

activity to maximal activity was calculated for each condition and the ratio corresponding to the

unstimulated cells was normalized to 1 arbitrary unit. The value of each condition represents the mean

± SEM of 3 independent experiments. *p<0.05 versus basal condition.

96

Figure 4.13 Effect of CaMKII inhibitors on insulin-stimulated [3H]-2-deoxyglucose (2DOG) uptake in 3T3-L1 adipocytes. 3T3-L1 adipocytes were either treated with DMSO (basal) only, insulin

(100 nM for 15 min) only, KN-62 (10 μM for 1 h) only or pretreated with KN-62 (10 μM for 45 min) prior to insulin stimulation (left panel). In another experiment, 3T3-L1 adipocytes were either treated with tat-ctrl (5 μM for 45 min) only, tat-CN21 (5 μM for 45 min) only or pretreated with either of the

peptides (5 μM for 30 min) prior to insulin stimulation (right panel). The average 2DOG uptake of each

condition is shown as percentage of maximum response of insulin-treated cells and represents the mean

± SEM of 3 independent experiments, in which each experiment was assayed in triplicate.

97

Figure 4.14 Domain structure of Myo1c. Myo1c consists of 3 domains including head, neck and tail domain. S701A, IQ1A and S701AIQ1A double mutants were generated in this study. S701A contains a single amino acid change at S701 while IQ1A contains four amino acid mutations (I706A, Q707A,

R711A and G712A).

98

Figure 4.15 in vivo association of different Myo1c mutants and calmodulin in CHO/IR/IRS-1 cells.

CHO/IR/IRS-1 cells were transiently transfected with either EYFP-Myo1c WT, S701A, IQ1A or EYFP

(empty vector). Cell lysates from either basal or insulin-treated cells were subjected to immunoprecipitation with anti-GFP antibody. Input cell extracts and pulldowns were analyzed by

Western blot with the use of anti-GFP antibody, which detects EYFP-Myo1c, and anti-CaM antibody.

The experiment was performed three times and the images are from a representative experiment.

99

Figure 4.16 in vitro association of different Myo1c mutants and 14-3-3 in CHO/IR/IRS-1 cells.

CHO/IR/IRS-1 cells were transiently transfected with either EYFP (empty vector), EYFP-Myo1c WT,

S701A, IQ1A or S701AIQ1A double mutant. Cell lysates from either basal or insulin-treated cells was subjected to 14-3-3-pulldown. Input cell extracts and pulldowns were analyzed by Western blot with the use of anti-GFP antibody, which detects EYFP-Myo1c, and anti-14-3-3 antibody. The experiment was performed three times and the images are from a representative experiment.

100

Figure 4.17 Effect of different kinase inhibitors on in vivo calmodulin-binding of Myo1c. 3T3-L1 adipocytes were pretreated with either 100 nM Wortmannin (W) for 10 min, 10 μM KN-62 (K) for 45 min or 5 μM tat-CN21 (C) for 30 min prior to insulin stimulation. Cell lysates were subjected to immunoprecipitation using anti-Myo1c antibody. Input cell extracts and immunoprecipitates were analyzed by Western blot with the use of anti-Myo1c and anti-calmodulin antibodies. The experiment was performed three times and the images are from a representative experiment.

101

Figure 4.18 Effect of in vitro CaMKII phosphorylation and 14-3-3-binding on ATPase activity of

Myo1c. A. Basal 3T3-L1 adipocyte lysates were subjected to immunoprecipitation with anti-Myo1c antibody and irrelevant antibody (ctrl). The immunocomplex-bound Protein G-separose was incubated with either inactive GST-CaMKII (I) or active GST-CaMKII (A), followed by either 46 ng GST or 100 ng GST-14-3-3 for 30 min at room temperature. GST-CaMKII and any unbound GST-14-3-3 were

102 removed by washing of the beads with ATPase reaction buffer. The precipitates were incubated with

[γ32P]ATP and reactions were subjected to thin-layer chromatography. Hydrolyzed [γ32P]Pi and unhydrolyzed [γ32P]ATP are indicated on the right. Immunoprecipitates were analyzed by Western blot with the use of anti-Myo1c antibody. The experiment was performed three times and the images are from a representative experiment. B. Quantification of ATPase assays. The values of [γ32P]Pi and

[γ32P]ATP were measured by densitometry. The ratio of [γ32P]Pi to [γ32P]ATP was calculated and the control signal (ctrl-IP) was subtracted. The ratio corresponding to the precipitate incubated with inactive GST-CaMKII and GST was normalized to 1 arbitrary unit. The value of each condition represents the mean ± SEM of 3 independent experiments.

103

Figure 4.19 Effect of insulin-stimulation on ATPase activity of EYFP-Myo1c. A. CHO/IR/IRS-1 cells were transiently transfected with either EYFP (empty vector), EYFP-Myo1c WT or K111A. Cell lysates from either basal or insulin-treated cells were subjected to immunoprecipitation with anti-GFP antibody. The ATPase activities were measured as described in Fig. 5A. Immunoprecipitates were

104 analyzed by Western blot with the use of anti-GFP antibody. The experiment was performed three times and the images are from a representative experiment. B. Quantification of ATPase assays. The values of

[γ32P]Pi and [γ32P]ATP were measured by densitometry. The ratio of [γ32P]Pi to [γ32P]ATP was

calculated and the background signal corresponding to Myo1c K111A was subtracted. The ratio

corresponding to Myo1c WT immunoprecipitated from basal cells was normalized to 1 arbitrary unit.

The value of each condition represents the mean ± SEM of 3 independent experiments. *p<0.05 versus

unstimulated cells.

105

Figure 4.20 Immunofluorescence analysis of surface HA in Myo1c knockdown cells. A. Amounts of Myo1c protein in HA-GLUT4-expressing 3T3-L1 adipocytes after expression of scrambled or

Myo1c-targeted siRNA (si-Myo1c) for 72 h were revealed by Western blot with the use of anti-Myo1c and anti-Actin antibodies. B. Immunofluorescence analysis of surface HA in scrambled and Myo1c knockdown HA-GLUT4-expressing 3T3-L1 adipocytes. Basal or insulin-treated cells were fixed and stained for surface HA with anti-HA antibody (red). The scale bar represents 15 μm.

106

107

Figure 4.21 Immunofluorescence analysis of surface HA in Myo1c re-expressing cells. A. Amounts of different EYFP-Myo1c proteins and endogenous Myo1c protein in HA-GLUT4-expressing 3T3-L1 adipocytes after co-expression of Myo1c-targeted siRNA (si-Myo1c) and EYFP-Myo1c for 72 h were revealed by Western blot with the use of anti-Myo1c and anti-Actin antibodies. B. Immunofluorescence analysis of surface HA in Myo1c-re-expressing cells. HA-GLUT4-expressing 3T3-L1 adipocytes were co-electroporated with si-Myo1c and EYFP-Myo1c WT, S701A, IQ1A, S701AIQ1A double mutant or

K111A. After 72 h, basal and insulin-treated cells were fixed and stained for surface HA with anti-HA antibody (red). After permeabilization with 0.1% saponin, EYFP-Myo1c was stained with anti-GFP antibody (green). Two cells are shown in each condition. The scale bar represents 15 μm. C. The average surface HA fluorescence level (Cy3) of each condition is shown as percentage of maximum response of insulin-treated cells expressing scrambled siRNA and represents the mean ± SEM of 3 independent experiments, within which 15 cells were assayed in each experiment. #p<0.05 versus unstimulated cells expressing si-Myo1c/Myo1c WT; *p<0.05, **p<0.005 versus unstimulated cells.

108 v. Discussion

In the present study I show that the myosin motor Myo1c undergoes increased phosphorylation in response to insulin. In view of the important role of Myo1c in

GLUT4 trafficking (Bose et al., 2002), this represents a key advance in this area.

Insulin stimulates Myo1c phosphorylation at a site that flanks one of its major calmodulin binding domains. Insulin-dependent Myo1c phosphorylation is regulated by CaMKII in a PI3K-dependent manner resulting in reduced calmodulin binding to

Myo1c and concomitant association with 14-3-3 proteins. The observation that the

Myo1c IQ1A mutant undergoes normal insulin-dependent phosphorylation suggests that calmodulin binding to the first IQ domain is not required for Myo1c phosphorylation excluding the possibility that this domain is required for CaMKII binding. Further studies are required to determine if the additional IQ domains are required for CaMKII-dependent Myo1c phosphorylation. Notwithstanding our data suggest a role for Myo1c phosphorylation in calmodulin release. Phosphorylation has been shown to play an important role in calmodulin binding to other IQ domain proteins. Serine phosphorylation of neuromodulin at a variable residue within its consensus IQ domain (IQxpSxRGxxxR) regulates calmodulin binding (Chapman et al., 1991). A crystal structure of the IQ domains from MyoV and apo-Calmodulin reveals a role for acidic residues in calmodulin in this interaction (Houdusse et al.,

2006). Hence, the introduction of an additional negative charge in this vicinity as in phosphorylated Myo1c would likely impair this interaction (pSxxxxIQxxxRGxxxR).

Alternatively, binding of 14-3-3 upon phosphorylation could lead to steric displacement of calmodulin.

How could Myo1c phosphorylation and/or regulated calmodulin binding play a role in the function of this molecule? It has been reported that calmodulin binding to Myo1c

109 inhibits its actin stimulated ATPase activity and Myo1c ATPase activity is stimulated by calcium (Barylko et al., 1992; Chacko et al., 1994; Zhu et al., 1998; Zhu et al.,

1996). In the present study I show that CaMKII-mediated phosphorylation stimulates

Myo1c ATPase activity. Hence, this leads to a model where insulin may regulate the mechanical properties of Myo1c via Ca2+/CaMKII-dependent phosphorylation. To formally test this hypothesis it will be necessary to precisely identify the consequences of Ca2+ loading and phosphorylation on Myo1c function in vivo. It is also conceivable that insulin-dependent Myo1c phosphorylation and 14-3-3 binding may regulate the association of other proteins, such as cargo, since 14-3-3 possesses at least two phosphopeptide binding sites.

Another possibility is that phosphorylation may regulate the association of Myo1c with specific regions of the PM. Myo1c and Myo1c orthologs bind to many different anionic phospholipids (Barylko et al., 2005) via a lipid-binding site located in the

Myo1c tail domain (Hokanson et al., 2006; Reizes et al., 1994). Consistent with a potential role for calmodulin binding in the membrane association of Myo1c, Tang and colleagues (Tang et al., 2002) showed that Ca2+ regulates the binding affinity of

Myo1c to lipids. Since Myo1c was shown to be localized at the PM under both basal and insulin-stimulated-conditions, this poses a potential model where Myo1c phosphorylation at S701 may stabilize membrane association of Myo1c at certain subdomains on the PM. Hence, further studies examining whether Myo1c binds to lipid on the PM directly in adipocytes are important. Intriguingly, another study by our laboratory revealed that a significant reduction in insulin-stimulated CaMKII phosphorylation was not observed in adipocytes at T286 using the PI3K inhibitor

Wortmannin suggesting that activation of CaMKII by insulin is PI3K-independent.

However, insulin-stimulated phosphorylation of Myo1c was PI3K-dependent. This

110 raises the possibility that PI3K may regulate the accessibility of Myo1c to CaMKII in vivo either by regulating subdomain localization of Myo1c or CaMKII itself on the

PM. Noteably, CaMKII phosphorylation is regulated via dynamic subcellular localization of both substrate and kinase (Tsui et al., 2005). Further studies are required to resolve these aspects of Myo1c regulation by Ca2+, CaMKII and PI3K.

Another intriguing element of the present study is the role of CaMKII and Ca2+ in insulin regulated glucose transport. Although controversial, inhibition of this process using Ca2+ chelators (Whitehead et al., 2001) and CaMKII inhibitors has been described (Brozinick et al., 1999; Konstantopoulos et al., 2007) consistent with our data. The use of inhibitors of ill-defined specificity, such as KN-62, the inability to document insulin-dependent changes in cytosolic Ca2+ and CaMKII activity and the lack of insulin regulated CaMKII substrates in adipocytes have provided major obstacles in substantiating the role of Ca2+ in insulin action. In the present study I have overcome several of these obstacles by showing that: a highly specific CaMKII inhibitor tat-CN21 inhibits insulin stimulated glucose transport in adipocytes (Figure

4.13); insulin activates CaMKII in adipocytes consistent with a recent study in rat muscle (Wright et al., 2004); and Myo1c as an important insulin-regulated CaMKII substrate in the fat cell. Notably, I have been unable to observe Myo1c phosphorylation in muscle raising the possibility that an alternate myosin isoform may substitute for Myo1c in this cell type. While I have not directly examined the effects of insulin on intracellular Ca2+ in the adipocyte a previous study has shown that insulin increases the Ca2+ concentration just beneath the PM in isolated skeletal muscle (Bruton et al., 1999). This raises the possibility that the insulin effect on Ca2+ homeostasis may be highly localized in these cells. This may be relevant to the effect of insulin on CaMKII activity, Myo1c phosphorylation and the insulin-dependent

111 release of calmodulin from Myo1c all of which likely occur just beneath the PM.

Intriguingly, CaMKII is activated at specific subcellular locations in other cell types

(Davies et al., 2007; Inagaki et al., 2000), which may explain why there is only a modest increase in CaMKII kinase activity induced by insulin when analyzed in vitro.

This also provides challenges in dissecting some of these steps in vivo. This is particularly evident for the effects of insulin on the Myo1c/calmodulin interaction which may require the development of highly sensitive FRET based methods to ensure that the changes in calmodulin-binding that I have observed are not due to changes after cell lysis. Regardless, the effect of insulin on Myo1c/calmodulin interaction as a minimum likely reflects reduced affinity between Myo1c and calmodulin in vivo thus underpinning a role for S701 phosphorylation in causing a conformational change in Myo1c at or around IQ1. Furthermore, the observation that the IQ1A mutant caused a slight increase in cell surface GLUT4 levels when expressed in Myo1c-knockdown adipocytes is consistent with a role for regulated calmodulin binding at this site.

In conclusion, Myo1c represents a novel target of insulin action in adipocytes. In the basal state, calmodulin binds to the IQ1 domain of Myo1c. Upon insulin stimulation,

Ca2+ influx increases the concentration of Ca2+ just beneath the PM, activating

CaMKII, Myo1c phosphorylation at S701 and 14-3-3 binding to Myo1c.

Phosphorylation and/or 14-3-3 binding leads to displacement of calmodulin from the

Myo1c IQ1 domain. A key functional outcome of this pathway is increased Myo1c

ATPase activity. This likely plays a key role in GLUT4 vesicle docking and/or fusion at the plasma membrane.

112 Chapter 5

Identification of a Novel Myo1c-Interacting Protein, Armcx5, in Adipocytes

113 i. Abstract

To identify novel Myo1c-interacting proteins, a yeast two-hybrid 3T3-L1 adipocytes cDNA library was screened with the carboxy tail domain of Myo1c as bait. The identification and biochemical characterization of a novel Myo1c-interacting protein,

Armcx5, is described. Armcx5 consists of three Armadillo repeat domains (Arm), which are known to form a scaffold for protein-protein interaction. Armcx5 interacted specifically with the Myo1c tail, but not with the Myo1b tail, a highly conserved homolog. This was observed using in yeast two-hybrid system and the

Armcx5/Myo1c tail interaction was confirmed in transiently transfected

CHO/IR/IRS-1 cells. Endogenous Armcx5 was localized at the plasma membrane where it interacted with full length Myo1c in 3T3-L1 adipocytes. Additionally studies indicated that truncation of even minimal regions at the N or C termini of the protein impaired its ability to interact with Myo1c suggesting that this interaction likely involves multiple domains. Finally, a reduction in the level of Armcx5 protein by siRNA knockdown resulted in reduced basal and insulin-stimulated glucose uptake, and GLUT4 translocation. Overall, these results suggest that Armcx5 plays a role in

GLUT4 trafficking in adipocytes, possibly functioning as a plasma membrane scaffold for binding of Myo1c. ii. Introduction

The tail domains of unconventional myosins are highly divergent and play an important role in determining the target cargo. For examples, Melanophilin is the melanosome-specific receptor of the MyoVa tail, creating a link between Rab27a and

MyoVa and functioning in actin-based movement of melanosomes (Wu et al., 2002);

Rab11-FIP2 was identified as an adaptor between Rab11a and the MyoVb tail, and this complex is important in the regulation of plasma membrane composition by

114 membrane recycling (Hales et al., 2002); Dab2 binds to the MyoVI tail and in turn recognizes the NPXY internalizational signal in the cytoplasmic tail of the LDL receptor family members, mediating their endocytosis (Buss et al., 2004); MyoVIIa was known to be anchored by the transmembrane protein vezatin to the cadherin-catenins complex, creating a tension force between adherens and actin filaments (Kussel-Andermann et al., 2000); BIG1, the Arf1 guanine nucleotide exchange factor (GEF), was shown to interact with the tail of MyoIXb which possesses a GTPase activating protein (GAP) activity for Rho (Saeki et al., 2005); the

MyoX tail interacts directly with integrin and regulates its localization and function in cell adhesion (Zhang et al., 2004). Hence, the identification of molecules that bind to myosin tails has provided essential information about the specific function of individual members of this family.

Recent studies have highlighted the crucial role of Myo1c in adipocytes. Bose et al. reported that Myo1c, which is present in the GLUT4-enriched membrane fraction purified by differential centrifugation, facilitates GLUT4 recycling in response to insulin by showing that siRNA knockdown of Myo1c reduced glucose uptake (Bose et al., 2002). Using ultrafast microscopy, Bose et al. further showed that Myo1c is able to drive localized membrane remodeling facilitating the fusion of GSV’s with the plasma membrane (Bose et al., 2004). Recently, RalA was identified as a GSV cargo receptor for Myo1c. RalA was shown to bind to the IQ domain of Myo1c and this interaction is mediated by calmodulin (Chen et al., 2007). In addition, recent studies also showed that Rictor interacts with the head domain of Myo1c and this complex is believed to participate in cortical actin remodeling and membrane ruffling (Hagan et al., 2008). Interestingly, Myo1c was shown to associate with the nuclear factor κB essential modulator (NEMO)/IKK-γ subunit, promoting TNF-α-induced Ser307

115 phosphorylation of IRS-1 which in turn attenuates insulin signaling (Nakamori et al.,

2006).

None of the proteins that were shown so far to associate with Myo1c bind to the tail domain. However, proteins interacting with the tail domain are likely to be very important. Expression of the tail domain of Myo1c abolished insulin-mediated

GLUT4 translocation to the plasma membrane whereas expression of the Myo1b tail showed no effect (Bose et al., 2002). These results suggest that the tail domain of

Myo1c plays an important specific role in GLUT4 translocation likely involving some sort of molecular interaction. It is notable that the Myo1c tail domain has been shown to bind to phosphoinositides through its putative pleckstrin homology (PH) domain and so this interaction could involve lipid. However, it is unlikely that this is the only component of the interaction since the lipid binding domain is also found in the

Myo1b tail (Hokanson et al., 2006). Thus, I hypothesized that Myo1c must bind to a protein via its C terminal tail and that this interaction must be highly specific and central to insulin action.

Therefore, the aim of this study was to identify a protein which binds to the Myo1c tail and that plays a role in Myo1c function at the plasma membrane. Using a yeast two-hybrid screen, I identified Armcx5 as a novel Myo1c interacting protein and showed that it plays an important role in GLUT4 translocation and glucose uptake in adipocytes. iii. Materials and Methods

A. Yeast culture

The Saccaromyces cerevisiae yeast strain AH109 (MATa, trp1-901, leu2-3, 112, ura3-52, his3-200, gal4Δ, gal80Δ, LYS2::GAL1UAS-GAL1TATA-HIS3,

116 GAL2UAS-GAL2TATA-ADE2, URA3::MEL1UAS-MEL1TATA-lacZ) was utilized during these studies. Unless otherwise stated the yeast strain was propagated according to the manufacturer’s protocol (Clontech).

B. Preparation of yeast lysate

Yeast cultures were grown in selective media at 30 ˚C and 22 rpm for 2 d. The yeast pellet was collected by centrifugation at 4,000g for 3 min, and resuspended with a mixture of 135 μl Thorner buffer (50 mM Tris-HCl pH 8.8, 8M urea, 5% (w/v) SDS) and 15 μl β-mercaptoethanol. The activated 425-600 μm glass beads (~200 μl) were added to the suspension and the cells were lysed by vortexing for 1 min. The 2 × SDS reducing sample buffer (150 μl) was added and the cells were further lysed by vortexing for 1 min. The mixture was boiled for 2 min and ~10 ul was loaded onto

SDS-PAGE resolving gel for Western blot analysis.

C. Yeast two-hybrid library screening

The pGBKT7 and pGADT7 plasmids were purchased from Clontech. The pGBKT7-Rab15 and pACT2-15.24 plasmids, which were used as a positive control interaction partner in yeast two-hybrid screening, were gifts from Jagath Junutula

(Genentech, CA). The 3T3-L1 adipocyte cDNA library (cDNA fragments were ligated unidirectionally into EcoRI-XhoI-digested pGAD-GH GAL4 activation domain (AD) plasmid) was a gift from Dr Alan Saltiel (University of Michigan School of Medicine,

MI). The GAL4 DNA binding domain (BD) fusion containing the Myo1c tail domain was constructed by cloning the 711 bp fragment encoding residue 792 to 1028 of

Myo1c into the NcoI-BamHI sites of pGBKT7 (pGBKT7-Myo1cT). The yeast two-hybrid library screening was carried out according to the manufacturer’s protocol

(Clontech). Strain AH109 was transformed with pGBKT7-Myo1cT and Trp+ prototrophs were selected. The cells were made competent in 100 mM lithium acetate

117 and were transformed with 428 μg of 3T3-L1 adipocyte library DNA. Transformants were selected by plating cells on 50 × 15 cm petri dishes of synthetic medium lacking histidine, leucine and tryptophan (SD-His-Leu-Trp/–HLT, medium stringency). After

4 d of incubation at 30 °C, colonies that appeared on the –HLT plates were patched onto SD-Leu-Trp plates (–LT, low stringency), –HLT plates (medium stringency),

SD-Ade-His-Leu-Trp plates (–AHLT, high stringency) and SD-Ade-His-Leu-Trp plate containing 40 μg/ml of 5-bromo-4-chloro-3-indolyl-α-D-galactopyranoside

(–AHLT/+X-gal) for determination of the α-Gal activity. The plates were incubated at

30 °C for 4 days.

The plasmids were rescued from the Ade+ prototrophic colonies as previously described (Robzyk and Kassir, 1992). The cDNA insert of each clone was amplified by PCR using MATCHMAKER 5’ and 3’ AD LD-Insert Screening Amplimers

(Clontech) according to the following conditions: 35 cycles of denaturing at 94 °C for

2 min, annealing at 56 °C for 1 min and extension at 72 °C for 3 min. The PCR product of each cDNA clone was sequenced from both 5’ and 3’ end using

MATCHMAKER 5’ and 3’ AD LD-Insert Screening Amplimers. The resulting sequences were used to identify the corresponding gene in the GenBank database

(NCBI).

D. Yeast two-hybrid interaction studies

Interactions between known proteins were examined using the yeast two-hybrid system. The GAL4 DNA binding domain fusion of the Myo1b tail domain was constructed by cloning the 735 bp fragment encoding residues 893 to 1137 of Myo1b into the NcoI-BamHI sites of pGBKT7 (pGBKT7-Myo1bT). The following GAL4 activation domain fusions of Armcx5 were constructed by cloning the following fragments into EcoRI-XhoI sites of pGADT7 (pGADT7-Armcx5): full length (FL),

118 residues 1 to 341 (N1), residues 1 to 390 (N2), residues 1 to 509 (N3), residues 511 to

1137 (N1), residues 391 to 1137 (N2), residues 342 to 1137 (N3), residues 341 to 606

(C1), residues 391 to 606 (C2) and residues 509 to 606 (C3). Transformants were selected by plating cells on –LT plates, –HLT plates, –AHLT plates and –AHLT/+X-gal plates, and incubated at 30 °C for 4 d.

E. Subcellular fractionation

Subcellular fractionation of 3T3-L1 adipocytes using differential centrifugation was carried out as previously described (Shewan et al., 2003). The cells were incubated in serum-free DME medium for 2 h, stimulated with 100 nM insulin for 15 min, and lysed in HES buffer (20 mM HEPES, 10 mM EDTA, 250 mM sucrose pH 7.4) supplemented with Complete protease inhibitor mixture (Roche) and phosphatase inhibitors (2 mM sodium orthovanadate, 1 mM pyrophosphate, 10 mM sodium fluoride) at 4 °C. The lysate was centrifuged at 500 g for 10 min, followed by further centrifugation at 10,080 g for 12 min, 15,750 g for 17 min, and 175,000 g for 75 min at 4 °C to generate plasma membrane (PM), mitochondria and nuclei, high density microsomes (HDM) and low density microsomes (LDM). HDM contains the endoplasmic reticulum and large endosomes; LDM contains Golgi markers, recycling endosomes and the majority of the insulin responsive GLUT4 storage vesicles. iv. Results

Yeast two-hybrid library screening of 3T3-L1 adipocytes

To identify novel proteins interacting with the cargo binding domain, or tail domain, of Myo1c in adipocytes, I screened a 3T3-L1 adipocyte cDNA library using the yeast two-hybrid system (Figure 5.1). The tail domain of Myo1c was cloned into the bait vector pGBKT7, which generated a GAL4 DNA binding domain fusion of Myo1c tail,

119 and sequentially transformed into strain AH109 with a 3T3-L1 adipocyte cDNA library fused to the GAL4 activation domain in the prey vector pGAD-GH (Printen et al., 1997). A small scale transformation (428 μg of DNA into 1 L overnight culture of yeast containing ~6 × 106 yeast cells) generated ~3.8 × 106 independent clones on –LT plates. The library-scale transformations were plated on 50 × 150 mm plates of –HLT medium to screen for expression of HIS3. Of ~3.8 × 106 total transformants, 193 colonies His+ prototrophic colonies were recovered in the first screen of medium stringency. Subsequently, these His+ colonies were subjected to a second screen of medium stringency by plating on –LT plates and –HLT plates. A positive control

(yeast transformed with pGBKT7-Rab15 and pACT2-15.24) and a negative control

(yeast transformed with pGBKT7-Myo1cT and pGADT7) were included. Out of the

193 colonies, 147 colonies were His+ and grew faster than the negative control in the second screen (Figure 5.2A). These 147 colonies were replicated onto –AHLT plates to screen for ADE2 expression which virtually eliminated false positive interactions.

Under this high stringency screen, 43 Ade+ prototrophic colonies were recovered

(Figure 5.2B). The interacting prey fragments of the positive clones were

PCR-amplified and sequenced at the 5’ and 3’ junctions, confirming that all the 43 colonies were in-frame. A total of 17 unique proteins were identified (Table 5.1).

One of the cDNA clones which encodes an entire protein missing only the first 15 amino acids was Armadillo repeat containing X-linked 5 (Armcx5). Armcx5 is widely expressed in most tissues although its function and the expression profile in adipocytes are unknown (Shmueli et al., 2003). It contains 3 Armadillo (Arm) repeat domains at the C-terminus. Arm repeat proteins contain tandem imperfect repeats of a sequence motif of about 42 amino acids. A single Arm repeat is composed of three α helices. Multiple Arm repeats interact with one another to form a right-handed

120 superhelix of helices, creating a scaffold for protein-protein interactions (Huber et al.,

1997). Arm repeat proteins are involved in a wide range of cellular functions. An example is yeast Vac8p which contains 11 Arm domains and is required for vacuole fusion. It interacts with the fusion machinery through its Arm repeats and facilitates bilayer mixing (Wang et al., 2001).

Interaction studies of Armcx5 and Myo1c

To further investigate the interaction between Armcx5 and the Myo1c tail domain, I compared the binding of the Myo1c tail with the Myo1b tail. The full length Armcx5 was obtained from a commercial source and cloned into pGADT7 vector. Yeast strain

AH109 was cotransformed with plasmids encoding the GAL4 AD fusion of full length Armcx5 and the GAL4 BD fusion of either the Myo1c tail or the Myo1b tail.

The strain that coexpressed Armcx5 and Myo1c tail grew as efficiently as the positive control, pGBKT7-Rab15 and pACT2-15.24 cotransformed strain, which confirm the result of the yeast two-hybrid library screen (Figure 5.3). The strain expressing

Armcx5 and the Myo1b tail grew slowly under medium stringency conditions, and did not grow at all under high stringency conditions. Since the Myo1b tail is highly homologous to the Myo1c tail and expression of the Myo1c tail but not the Myo1b tail in adipocytes inhibits insulin-regulated GLUT4 translocation, this suggests that

Armcx5 interacts specifically with the Myo1c tail and that this interaction may play an important role in GLUT4 translocation.

In order to map the Myo1c-binding domain on Armcx5, I generated a series of truncation mutants and assayed for their binding activities using the yeast two-hybrid system. As there are no defined domains or structures located at the N-terminus of the first Arm domain, these mutants were designed based on the three Arm domains of

Armcx5 and they consist of different numbers of Arm domains (Figure 5.4A). Figure

121 5.4B shows that all mutants were expressed at the predicted molecular weights. As shown in figure 5.4C, in contrast to the full length Armcx5 which interacts strongly with the Myo1c tail, none of the N-terminal or the C-terminal truncation mutants showed an interaction with Myo1c. These data suggest that the full length Armcx5 protein is required for the interaction with Myo1c.

To confirm the interaction of Armcx5 with the Myo1c tail in mammalian cells,

CHO/IR/IRS-1 cells were cotransfected with FLAG epitope-tagged Armcx5 and either GFP-tagged Myo1c tail (GFP-MT) or GFP-tagged Myo1c neck/tail (residue

696 to 1028, GFP-MNT) (Figure 5.5). Cell extracts from basal or insulin-treated cells were incubated with an anti-FLAG antibody. Armcx5 co-immunoprecipitated with both the Myo1c tail and Myo1c neck/tail and these interactions were neither stimulated nor inhibited by insulin stimulation. Co-immunoprecipitation was not observed when the GFP tag was used alone.

To investigate the role of Armcx5 in GLUT4 translocation, I first determined the subcellular localization of Armcx5 in adipocytes. Strikingly, Armcx5 displayed a similar subcellular distribution to Myo1c in adipocytes with significant enrichment in the PM fraction. Notably, insulin had no significant effect on the distribution of

Myo1c or Armcx5. To characterize the specificity of the interaction between endogenous Armcx5 and full length Myo1c in adipocytes, I subjected the lysates from

3T3-L1 adipocytes to co-immunoprecipitation using the anti-Myo1c antibody.

SDS-PAGE followed by immunoblotting using an anti-Armcx5 antibody revealed an in vivo interaction between Myo1c and Armcx5 (Figure 5.7). A reciprocal co-immunoprecipitation using anti-Armcx5 antibody confirmed this observation. No band was found in the IgG control samples demonstrating the specificity of the endogenous interaction of Myo1c with Armcx5.

122 The role of Armcx5 in GLUT4 translocation

I next set out to determine if Armcx5 plays an important role in glucose transport in adipocytes using a knockdown strategy. I generated stable shRNA-mediated Armcx5 knockdown cell lines via retroviral infection of HA-GLUT4-expressing 3T3-L1 adipocytes (Figure 5.8A). The shRNA construct 388 resulted in a significant reduction of the Armcx5 protein by ~50% compared to the scrambled shRNA control. Analysis of GLUT4 translocation in these cells revealed that shRNA-mediated knockdown of

Armcx5 GLUT4 translocationwas reduced by 23% under basal conditions, by 10% -

18% at a sub-maximal insulin level and by 15% under maximal insulin stimulation

(Figure 5.8B). Similar changes in glucose transport were also observed (Figure 5.8C).

Thus, despite the fact that there was only a modest reduction in insulin action in these cells in view of the fact that I only observed a 50% reduction in Armcx5 protein levels in these cells this is indicative of a role for this protein in GLUT4 trafficking.

123

Figure 5.1 Principle of the yeast two-hybrid assay. A bait gene is expressed as a fusion to the GAL4

DNA-binding domain (DNA-BD) and the cDNA library is expressed as a fusion to the GAL4 activation domain (AD). When the bait and the library fusion proteins interact, the DNA-BD and AC are brought into close proximity, activating transcription of the reporter genes. Strain AH109 is a derivative of strain PJ69-2A and includes the ADE2 and HIS3 markers, allowing strong and stringent nutritional selection. MEL1 is an endogenous GAL4-responsive gene which encodes α-galactosidase for blue/white screening on X-gal indicator plates. All 3 reporter genes are under the control of 3 different GAL4-responsive UAS and promoter elements.

124

125

Figure 5.2 Yeast two-hybrid screening of 3T3-L1 adipocytes cDNA library using Myo1c tail domain as the bait. A. The 147 His+ colonies which grew on the –HLT plates in the first screen were replicated onto –LT plate (upper panel), –HLT plate (middle panel) and –AHLT plate (bottom panel).

The –AHLT plate is for screening of Ade+ prototrophic colonies. #S10 is the negative control

(pGBKT7-Myo1cT/pGADT7 transformants) and #T10 is the positive control

(pGBKT7-Rab15/pACT2-15.24 transformants). B. The 43 Ade+ prototrophic colonies were replicated again onto –LT plate, –HLT plate, –AHLT plate and –AHLT/X-gal indicator plate for assay of

α-galactosidase activity.

126 Accession Description NM_001009575.4 Mus musculus armadillo repeat containing, X-linked 5 (Armcx5), mRNA NM_008753.4 Mus musculus ornithine decarboxylase antizyme 1 (Oaz1), mRNA Mus musculus suppressor of var1, 3-like 1 (S. cerevisiae) (Supv3l1), mRNA >gb|BC049796.1| Mus musculus suppressor of var1, 3-like 1 (S. NM_181423.2 cerevisiae), mRNA (cDNA clone MGC:59435 IMAGE:6332552), complete cds NM_130864.3 Mus musculus acetyl-Coenzyme A acyltransferase 1A (Acaa1a), mRNA NM_033562.3 Mus musculus Der1-like domain family, member 2 (Derl2), mRNA NM_009984.3 Mus musculus cathepsin L (Ctsl), mRNA NM_029011.2 Mus musculus RIKEN cDNA 4833409A17 gene (4833409A17Rik), mRNA Mus musculus 2 days neonate thymus thymic cells cDNA, RIKEN AK088788.1 full-length enriched library, clone:E430026A19 product:enolase 1, alpha non-neuron, full insert sequence

False Positive

Accession Description NM_007743.2 Mus musculus collagen, type I, alpha 2 (Col1a2), mRNA Mus musculus collagen, type III, alpha 1 (Col3a1), mRNA NM_009930.1 >gb|BC052398.1| Mus musculus collagen, type III, alpha 1, mRNA (cDNA clone MGC:63402 IMAGE:5720872), complete cds NM_009933.3 Mus musculus collagen, type VI, alpha 1 (Col6a1), mRNA Mus musculus ribosomal protein SA, mRNA (cDNA clone MGC:102123 BC081461.1 IMAGE:6820702), complete cds NM_023133.1 Mus musculus ribosomal protein S19 (Rps19), mRNA PREDICTED: Mus musculus similar to ribosomal protein L13 XM_001473362.1 (LOC100039656), mRNA Mus musculus proteasome (prosome, macropain) 26S subunit, ATPase 2 NM_011188.3 (Psmc2), mRNA Mus musculus ATP synthase, H+ transporting mitochondrial F1 complex, NM_016774.3 beta subunit (Atp5b), nuclear gene encoding mitochondrial protein, mRNA Mus musculus ATP synthase, H+ transporting, mitochondrial F1 complex, NM_025313.2 delta subunit (Atp5d), nuclear gene encoding mitochondrial protein, mRNA

Table 5.1 A list of proteins identified in the yeast two-hybrid screen. Out of the 17 proteins, 9 of

them were believed to be false positives (Bartel et al., 1993).

127

Figure 5.3 Interaction study of Armcx5 and Myo1c or Myo1b. Yeast strain AH109 was transformed

with the indicated constructs and all transformants were plated on –LT plate, –HLT plate, –AHLT plate and –AHLT/X-gal indicator plate according to the four dilution-serious. It is noted that the yeast transformed with pGBKT7/pGADT7-Armcx5 (the 3rd row) is negative control and it is not expected to show interaction under –HLT condition (medium stringency). This background interaction disappears under –AHLT condition (high stringency).

128

129

Figure 5.4 Mapping Myo1c-binding domain of Armcx5. A. Different Armcx5 truncation mutants were cloned into pGADT7 vector and expressed as GAL4-AD-tagged and HA-tagged fusion proteins.

B. Western blot analysis of cell lysate from indicated yeast transformants using anti-HA antibody revealed that all fusion proteins were expressed at their predicted molecular weights. C. Yeast strain

AH109 was transformed with the indicated constructs and all transformants were plated on –LT plate and –AHLT plate according to the four dilution-serious.

130

Figure 5.5 in vivo association of Armcx5 and Myo1c truncation mutants in CHO/IR/IRS-1 cells.

CHO/IR/IRS-1 cells were transiently cotransfected with FLAG-Armcx5 and either GFP- Myo1c neck/tail (residue 696 to 1028, GFP-MNT), GFP-Myo1c tail (GFP-MT) or GFP empty vector (negative control). Cell lysate from either basal or insulin-treated cells were subjected to immunoprecipitation with anti-FLAG antibody. Input cell lysates and immunprecipitates were analyzed by western blot with the use of anti-FLAG and anti-GFP antibodies. The experiment was performed two times and the images are from a representative experiment.

131

Figure 5.6 Western blot analyses of subcellular fractions of 3T3-L1 adipocytes. A. Western blot analysis of HDM, LDM, PM and cytosolic fractions from basal and insulin-treated 3T3-L1 adipocytes using anti-Myo1c, anti-Armcx5, anti-GLUT4 and anti-Syntaxin4 antibodies. The experiment was performed three times and the images are from a representative experiment.

Figure 5.7 in vivo association of endogenous Armcx5 and Myo1c in 3T3-L1 adipocytes. Cell lysate

from either basal or insulin-treated 3T3-L1 adipocytes was subjected to immunoprecipiration using

anti-Myo1c, anti-Armcx5 and irrelevant antibodies (ctrl-IP). Input cell extracts and immunoprecipitates

were analyzed by Western blot with the use of anti-Myo1c and anti-Armcx5 antibodies. The

experiment was performed three times and the images are from a representative experiment.

132

Figure 5.8 Effect of shRNA-mediated knockdown of Armcx5 on 3T3-L1 adipocytes. A. Amounts of Armcx5 protein in HA-GLUT4-expressing 3T3-L1 adipocytes which express either scrambled or three different Armcx5-targeted shRNA were revealed by Western blot with the use of anti-Armcx5 and anti-Actin antibodies. B. HA-GLUT4 translocation assay of scrambled and Armcx5 knockdown (#388)

133 HA-GLUT4-expressing 3T3-L1 adipocytes. 3T3-L1 adipocytes were either treated with DMSO (basal) only or insulin at indicated concentrations for 15 min. The average translocation of each condition is shown as fold-change relative to the basal cells expressing srambled shRNA and represents the mean ±

SEM of 3 independent experiments, in which each experiment was assayed in triplicate. C.

[3H]-2-deoxyglucose (2DOG) uptake assay of scrambled and Armcx5 knockdown (#388)

HA-GLUT4-expressing 3T3-L1 adipocytes. 3T3-L1 adipocytes were either treated with DMSO (basal) only or insulin at indicated concentrations for 15 min. The average 2DOG uptake of each condition is shown as fold-change relative to the basal cells expressing srambled shRNA and represents the mean ±

SEM of 3 independent experiments, in which each experiment was assayed in triplicate. *p<0.05 versus cells expressing scrambled siRNA under the same condition;

134 v. Discussion

As the role of Myo1c in GLUT4 translocation in adipocytes is still unclear, the aim of this study was to identify proteins which interact with Myo1c thus encoding new layers of Myo1c regulation of insulin-stimulated GLUT4 translocation. Initially I attempted to identify Myo1c interacting proteins using a proteomic approach in which endogenous Myo1c was immunoprecipitated using anti-Myo1c antibody.

Co-precipitated proteins were digested with trypsin and analysed by mass spectrometry. However, the proteins that were identified were mostly actin-interacting proteins as the head domain of Myo1c interacts with actin filament. Moreover, the anti-Myo1c antibody was raised against the tail domain of Myo1c.

Immunoprecipitation using this antibody may result in competitive binding with

Myo1c interacting proteins. To circumvent these problems I next expressed the

FLAG-tagged Myo1c tail in adipocytes using retroviral infection. The Myo1c tail was then immunoprecipitated using anti-FLAG antibody followed by mass spectrometry analysis. However, again I was unable to detect any significant Myo1c-interacting proteins. One potential reason for this is that the Myo1c tail alone when expressed in adipocytes was mislocalized in vivo and so this may render it inaccessible to potential binding partners.. Because of these limitations of using proteomic approaches to identify Myo1c-interacting proteins, I attempted to utilize a yeast two-hybrid library screening strategy. Etournay et al. used the Myo1c tail domain to screen the mouse inner ear library using yeast two-hybrid library and identified PHR1 as a novel protein ligand of the Myo1c tail (Etournay et al., 2005). Thus, I used a similar approach to screen the Myo1c tail against a 3T3-L1 adipocyte library thus eliminating potential problems with actin and calmodulin binding. In addition, since both BD fusion of

Myo1c and AD fusion of library protein are targeted to the nucleus, mislocalization of

135 the Myo1c tail is not a problem in yeast two-hybrid system.

I detected a total of 17 proteins that interacted with the Myo1c tail under high stringency screening conditions. However, further analysis revelaed that, some of these proteins are rountinely detected as false positives in Y2H screens (Table 5.1).

These include mitochondrial proteins, ribosomal proteins, ATP synthase, proteasome subunits and collagen (Bartel et al., 1993). Excluding these false positive clones, I was left with a series of metabolic enzymes (Oaz1, Acaa1a and enolase 1), proteases

(Derl2 and Ctsl), RNA helicase (Supv3l1) and a protein known as Armcx5 as putative

Myo1c interacting proteins. The interaction of Myo1c with Armcx5 was considered as a strong interaction compared to the positive control. While the positive control was known to interact strongly, the corresponding Armcx5 colonies grew to a similar extent as the positive control under high stringency.

To confirm the specificity of the Myo1c/Armcx5 interaction I next a designed a strategy based on a differential screening approach and showed that Myo1b a highly related homolog of Myo1c does not interact with Armcx5 (Figure 5.3). This is an intriguing result because Myo1b is known to be involved in transferrin endocytosis.

And expression of the tail domain of Myo1b in COS-1 cells abolished internalization of transferrin whereas the Myo1c tail did not (Raposo et al., 1999). Conversely, Bose et al. showed that expression of the tail domain of Myo1c in adipocytes inhibited insulin-stimulated translocation of GLUT4 to the plasma membrane whereas Myo1b tail had no effect (Bose et al., 2002). Therefore, it is an exciting observation that

Armcx5 only interacts with the tail domain of Myo1c but not Myo1b as it suggests that Armcx5 binds specifically to Myo1c and is one of the important factors in insulin-stimulated GLUT4 translocation.

Using a truncation approach in Armcx5 I observed that deletion of only a limited

136 amount of sequence from either end of the protein completely disrupted binding to

Myo1c. Therefore, no specific binding domains were mapped on Armcx5. One possibility is that the truncation mutants may not have folded correctly a common criticism of truncation type experiments. Alternatively it is possible that the interaction between these two proteins involves interactions at multiple sites in

Armcx5 or a 3D structure of Armcx5 that comprises interactions between multiple domains in the protein. In fact, Arm repeat proteins contain tandem copies of a degenerate protein sequence motif which forms a conserved three-dimensional structure (Coates, 2003). A single Arm motif consists of 3 α-helices and tandem Arm repeats fold together and interact with one another to form a right-handed superhelix of helices, creating a surface for protein-protein interaction (Daniels et al., 2001;

Huber et al., 1997). Examples are Importin-α, β-catenin and SPAG6. These proteins form a very similar helical bundle of α helices and are flanked by N- and C-terminal domains, which are thought to stabilize the superhelical structure and to bind to other proteins or membranes (Coates, 2003). Deletion analysis of PF16, the alga homologue of SPAG6 which is required for stability of micotubules, suggested that all eight Arm repeats are required for its function and appear to form a single functional unit (Smith and Lefebvre, 2000). Therefore, it is likely that the Myo1c tail also interacts with a similar, but shorter, helical structure formed by all three Arm repeat domains of

Armcx5.

What is the role of the Armcx5/Myo1c complex in GLUT4 translocation at the molecular level? Many Arm domain proteins, such as Importin-α, β-catenin, yeast

Vac8p and protozoan Nd9p, are adapters that mediate the docking of various cargos at membranes. Importin-α regulates the transport of proteins containing nuclear localization signals, which are recognized by the Arm repeats of importin-α, into the

137 nucleus (Conti and Izaurralde, 2001). The N-terminal non-Arm region of importin-α interacts with importin-β allowing translocation through the nuclear pore (Malik et al.,

1997); β-catenin together with its binding partners anchor the actin cytoskeleton to the plasma membrane and are required for cell-cell adhesion. The Arm repeats of

β-catenin bind directly to the intracellular domain of cadherin and its N-terminal non-Arm region interacts with α-catenin which in turn interacts with actin (Aberle et al., 1996; Rimm et al., 1995); Vac8p plays multiple roles in yeast. It links vacuolar membranes to the actin cytoskeleton through myristolyation and palmitoylation of residues located outside of the Arm domain and this is essential in targeting new vacuolar membrane to the bud site during cell division (Fleckenstein et al., 1998; Pan and Goldfarb, 1998; Wang et al., 1998). Vac8p is also a component of the nucleus-vacuole junction where it binds to the Nyv1p, an integral membrane v-SNARE of the nuclear envelope (Pan et al., 2000). In addition, Vac8p is also required in vacuole-vacuole fusion in vivo and in vitro. It binds to SNARE complexes through its Arm domain and inserts into the vacuole membrane through a palmitoylated residue (Wang et al., 1998). In Protozoan Paramecium, Nd9p is required for an exocytotic membrane fusion step in the regulated secretory pathway.

Similar to Vac8p, the non-Arm region was shown to interact with membranes whereas the Arm domain is thought to be involved in protein-protein interactions (Froissard et al., 2001). These studies suggest that the primary function of Arm domain proteins is to dock binding partners at their target membranes.

The present study shows that knockdown of Armcx5 by ~50% reduced both GLUT4 translocation and glucose uptake. Although the insulin response was not affected significantly, GLUT4 translocation and glucose uptake clearly decreased under both basal and insulin-stimulated conditions. It suggested that Armcx5 proteins are

138 required to position GLUT4 at the PM under both conditions. The present study shows that the Armcx5/Myo1c complex is localized at the PM (Figure 5.6), and previous studies suggested that Myo1c is likely to be involved in the fusion step of

GLUT4 trafficking at the plasma membrane (Bose et al., 2004). Hence, the

Armcx5/Myo1c complex which forms under both basal and insulin-stimulated conditions is likely the limiting factor that facilitates the fusion of GSV. Further experiments using siRNA targeting Armcx5 which may result in more efficient knockdown will likely be useful.

By analogy to other Arm domain proteins, I propose that Armcx5 docks Myo1c at the plasma membrane to facilitate fusion of GSV via interaction between its Arm domains and the tail domain of Myo1c. Armcx5 may in turn bind to the SNARE complex via the regions flanking the Arm repeats. Alternatively, these regions may interact directly with the plasma membrane. Analysis of the Armcx5 primary amino acid sequence by

SCANSITE using a high stringency scan identified a potential phosphatidylinositol

(3,4,5)-triphosphate (PIP3)-binding site outside the Arm repeats at the C-terminal domain (Obenauer et al., 2003) (Figure 5.9). Although insulin stimulation increased the level of PIP3 at the PM, it does not increase the amount of Armcx5 at the PM

(Figure 5.6). Hence, a small pool of PIP3 present under basal conditions maybe responsible for membrane localization of Armcx5. In addition, Armcx5 may potentially be palmitoylated at the residue between the first and the second Arm domains. Although no strict palmitoylation consensus site has been identified, in vitro studies on palmitoylation of synthetic peptides suggested that the presence of basic and hydrophobic residues close to a cysteine facilitates its palmitoylation (Belanger et al., 2001). In silico analysis using the software CSS-Palm 2.0 under high stringency scan revealed two potential palmitoylation sites which are flanked by mostly

139 hydrophobic residues (Ren et al., 2008) (Figure 5.9). Hence, future experiments on identification of other binding partners and lipid modifications of Armcx5 will help clarify the role of the Armcx5/Myo1c complex. In addition, in silico analysis also revealed 2 potential Akt phosphorylation sites located near the N-terminus of the first

Arm domain. Further experiments are required to determine whether Armcx5 is phosphorylated in response to insulin and whether phosphorylation is involved in

GLUT4 translocation.

140

Figure 5.9 In silico analysis of Armcx5. T353 and S355 were identified as potential Akt phosphorylation motifs (R/KXR/KXXS/T); F553 were identified as a potential phosphotidylinositol

(3,4,5)-trisphosphate-binding motifs. Scanning was performed under high stringency settings in

SCANSITE program. C450 and C451 are detected as potential palmitoylation residues using CSS-Palm

2.0 under high stringency scan.

141 Chapter 6

General Discussion

142 Resolving the molecular regulation of the insulin signaling pathway and of GLUT4 trafficking, and more importantly the intersection point of these two events has been a major challenge for at least three to four decades. Recent advances in technology have allowed a more detailed and mechanistic investigation of these cellular processes. For example, TIRFM allows detailed analysis of discrete steps of GLUT4 trafficking, and this has highlighted a major role for a regulatory step at the PM. These TIRFM-based studies provide a solid foundation for further proteomic or genomic analysis which aim to identify important factors that link insulin signaling with GLUT4 trafficking.

Also recent advances in mass spectrometry allow for greater resoluaiton of low abundance proteins making this a practical tool for molecular discrover. The strategy I have employed here was to identify insulin regulated phosphoproteins in adipocytes by using 14-3-3 proteins as an affinity matrix to enrich for potential substrates. Then these 14-3-3-binding proteins were resolved by SDS-PAGE and were identified by mass spectrometry. Using this approach I identified 38 proteins many of which were metabolic enzymes. The appearance of Myo1c on this list was particularly exciting because Myo1c is localized at the plasma membrane (Figure 3.8) and has been implicated in insulin-stimulated GLUT4 translocation in adipocytes (Bose et al., 2002;

Bose et al., 2004).

The present studies represent a significant advance in the field of insulin action for several reasons. Firstly, Myo1c was identified as a downstream target of the

Ca2+-dependent pathway in adipocytes. Calcium has been implicated in the insulin-dependent recruitment of GLUT4 to the PM (Li et al., 2006; Perrini et al.,

2004; Whitehead et al., 2001). It plays a role in maintenance of Akt phosphorylation, the translocation of GSV and the fusion of vesicles at the PM in adipocytes

(Whitehead et al., 2001; Worrall and Olefsky, 2002). However the molecular target of

143 Ca2+ has been unclear. The present study showed that Ca2+ is required for phosphorylation of Myo1c and, more interestingly, incubation of adipocytes with a

Ca2+ ionophore is sufficient to stimulate Myo1c phosphorylation per se. This suggests that Myo1c is likely one of the main end points of the Ca2+-dependent regulation.

Previous study suggested that insulin increases the Ca2+ concentration just beneath the

PM by stimulating Ca2+ entry into muscle cells via L-type Ca2+ channels. (Bruton et al., 1999). A preliminary study by our laboratory revealed that a family of Ca2+ channel blockers inhibits insulin-stimulated GLUT4 translocation to the PM in

3T3-L1 adipocytes, supporting a similar model of insulin-induced PM localized Ca2+ influx in adipocytes thereby activating CaMKII-mediated phosphorylation of Myo1c at the PM.

Secondly, the present study also established a link between insulin, CaMKII and

GLUT4 translocation. Recent studies have implicated a role for CaMKII in glucose transport in adipocytes (Konstantopoulos et al., 2007). Over expression of wild type

CaMKII resulted in a significant increase in insulin-stimulated GLUT4 translocation

(Konstantopoulos et al., 2007). Identification of Myo1c as a downstream target of

CaMKII potentially completes the link between CaMKII and GLUT4 translocation.

The present study also showed for the first time that insulin activates CaMKII in adipocytes. Although the exact mechanism for this effect is still unclear, insulin is likey to activate the CaMKII near the PM by increasing the local Ca2+ concentration.

Thirdly, the present study showed that the phosphorylation of Myo1c is

PI3K-dependent. This is intriguing since another study by our laboratory suggested that the activation of CaMKII is not PI3K-dependent as wortmannin pre-treament did not inhibit insulin-stimulated activation of CaMKII (Figure 6.1). Then how does PI3K fit into the picture? In addition to an increase in the local Ca2+ concentration, the

144 subcellular targeting of CaMKII is known to be another point of regulation to discriminate among the many substrates of CaMKII (Tsui et al., 2005). In neuronal cells, the CaMKII-β isoform which contains a cytoskeletal targeting signal binds to

F-actin and functions as a targeting module that localizes a much larger number of

CaMKII-α subunits to cytoskeletal sites of action (Shen et al., 1998). Although the

α-isoform is not expressed in adipocytes, the β-isoform may dock the δ-isoform, the predominant isoform of CaMKII in adipocytes (Konstantopoulos et al., 2007), to actin for phosphorylation of Myo1c. This cytoskeletal targeting of CaMKII may be

PI3K-dependent. Alternatively, the PI3K-dependent association of CaMKII and

Myo1c may also result in activation of a specific pool of CaMKII at the PM. In the case of Myosin-V in neuronal cells, it interacts with CaMKII, allowing the myosin-bound calmodulin and local Ca2+ to activate the kinase which phosphorylates

Myosin-V at a site that reduces the interaction with its cargo (Costa et al., 1999).

Hence, the association of CaMKII with Myo1c may represent an additional regulatory mechanism besides the insulin-induced influx of Ca2+, and this step is likely to be

PI3K-dependent. Taken together, the PI3K activity may be required for juxtaposing

Myo1c and CaMKII to allow phosphorylation of Myo1c thereby generating a specialized hot zone for Myo1c- and SNARE-mediated fusion.

Furthermore, the present study also led to the identification of Armcx5 as a

Myo1c-interacting protein and an important player in GLUT4 translocation. Since both Myo1c and Armcx5 can potentially bind to phosphoinositol (Figure 5.9), the

Armcx5/Myo1c complex may move horizontally across the PM to a PIP3-enriched region to allow its association with CaMKII. Overexpression of Myo1c may be able to bypass the requirement of PI3K for its localized activation by CaMKII, which explains the observation seen by Bose et al. that high expression of Myo1c overcomes

145 the inhibitory effect of wortmannin on the fusion of GSV at the PM (Bose et al., 2002;

Bose et al., 2004). A similar effect is observed by ionophore treatment in the present study. Figure 4.9B indicates that Myo1c is phosphorylated in response to ionophore treatment in the absence of insulin stimulation. Although I cannot exclude the possibility that the ionophore may lead to activation of PI3K, activation of a large amount of CaMKII by ionophore addition may also bypass the requirement for PI3K in juxtaposing Myo1c and CaMKII.

The elucidation of the phosphorylation site S701 is extremely exciting since this site lies between the head, the functional domain of Myo1c, and the neck, the calmodulin-mediated regulatory domain. I showed that phosphorylation at this site mediates 14-3-3-binding and reduces calmodulin-binding though these events may be independent of each other. Binding of 14-3-3 to Myo1c may mediate protein-protein interaction or prevent dephosphorylation. Calmodulin was thought to regulate protein-protein interaction and/or lipid binding, and the binding of each calmodulin to

Myo1c was known to be regulated at different concentrations of Ca2+ (Gillespie and

Cyr, 2002). The present study suggested that phosphorylation may in fact play a secondary regulatory mechanism of calmodulin-binding, at least at the first IQ domain although the biochemical consequence has yet to be explored. The G protein RalA has been shown to interact with Myo1c through its neck domain (Chen et al., 2007). It is likely that RalA only interacts with Myo1c at the PM in response to insulin since

Myo1c is localized at the PM while RalA is present at the GSV. In this case, the

14-3-3-binding and the stoichiometry of calmodulin bound to Myo1c may be important in mediating the RalA/Myo1c interaction.

Lastly, the present study also indicated that both the phosphorylation and the insulin-activated ATPase activity of Myo1c are required in GLUT4 translocation.

146 Since the ATPase activity is primarily required for the mechanical cycle of Myo1c

(Figure 4.1), it supports the idea that the mechanical properties of the Myo1c motor and actin filaments near the PM are essential at the later stages of GLUT4 trafficking.

This appears to be the case for ADH-regulated translocation of sodium channels to the

PM in epithelial cells (Wagner et al., 2005). In fact, Myo1c is classified as a low-duty-ratio motor because it spends a small fraction of its catalytic cycle strongly bound to actin (El Mezgueldi et al., 2002). Therefore, Myo1c cannot support long distance movement of cargo (De La Cruz and Ostap, 2004). In contrast, MyoVa, a dimeric myosin motor that is present in the GSVs, is a processive motor with a high duty ratio which can travel for a long distance without dissociating from actin filaments (Sato et al., 2007; Yoshizaki et al., 2007). Taken together, MyoVa may be responsible for movement of the GSVs along actin filaments and the GSVs only interact with Myo1c at the PM. The actin-dependent fusion step requires the motor activity of Myo1c, possibly moving the vesicles over a short distance and releasing the vesicles from the actin filament.

Does Myo1c have multiple functions in adipocytes? Myo1c was first reported to be associated with GSV and to transport it to the PM along actin filament in response to insulin (Bose et al., 2002). The present study demonstrated that Myo1c is primarily localized at the PM. TIFRM-based study by our laboratory using the same

EYFP-Myo1c constract as Bose et al. showed that insulin induces GLUT4-EGFP fusion proteins entering to the TIRF zone while EYFP-Myo1c proteins remain static at the TIRF zone. Several Myo1c interacting proteins have been identified including

RalA (Chen et al., 2007), Rictor (Hagan et al., 2008), NEMO/IKK-γ complex

(Nakamori et al., 2006) and Armcx5 in the present study (Figure 6.2). These reflect the involvements of Myo1c in the transport of GSVs along actin (Chen et al., 2007),

147 insulin-mediated actin remodeling (Hagan et al., 2008), the regulation of IRS-1

(Nakamori et al., 2006) and the docking/fusion of GSV at the PM. These events are thought to be regulated by distinct pools of Myo1c proteins that have different subcellular localization and binding partners. In fact, these reported

Myo1c-interacting proteins bind to Myo1c at different domains. Additionally each pool of Myo1c may require its specific and multiple-stage regulatory mechanism. The present study demonstrated that Armcx5-binding, PI3K activity, Ca2+ signaling,

CaMKII-mediated phosphorylation at S701, 14-3-3-binding and calmodulin-binding are likely the major mechanisms of regulating the Myo1c pool at the PM. I have found that Armcx5 and Myo1c are constitutively localized to the PM and that they bind to each other. Neither of these proteins possess a transmembrane domain and so they likely interact with the PM either by binding to lipids or via some other modification. Notably Armcx5 has a potential palmitoylation site and PIP3-binding site (Figure 5.9). The other interesting feature is that Myo1c is found in the lipid rafts

(Gupta et al., 2006). This is intriguing as this may represent a subdomain on the PM for vesicle docking/fusion. Notably t-SNAREs are also found in lipid rafts of adipocytes (Chamberlain and Gould, 2002) and because multiple SNARE complexes are required for fusion of a vesicle (Rickman et al., 2005), I suggest that the

Armcx5/Myo1c complex may recruit the GSV to a very discrete region on the PM where SNAREs are concentrated.

I propose the following model as a framework for future studies (Figure 6.2). Myo1c interacts with its PM receptor, Armcx5, which binds to the PM through palmitoylation and/or its PIP3-binding site. Myo1c may also bind to phosphoinositides at the plasma membrane directly. In the basal state, only a small amount of Myo1c is phosphorylated and results in basal fusion of GSV. Upon insulin stimulation, the Ca2+

148 concentration near the PM increases due to Ca2+ influx by a regulated PM Ca2+ channel, leading to CaMKII activation. PI3K is also activated by insulin and converts

PIP2 to PIP3. Armcx5/Myo1c complexes are recruited to the PIP3-enriched zone into close proximity of CaMKII. CaMKII phosphorylates Myo1c at S701, reducing calmodulin binding at the first IQ domain and inducing 14-3-3-binding which may prevent dephosphorylation of Myo1c and/or mediate the interation with RalA. Myo1c phosphorylation also triggers an increase in the ATPase activity of Myo1c via a conformational change. Activated Armcx5/Myo1c complexes are then recruited to vesicle docking sites by sliding along an actin filament. Then they likely bring the vesicles across to the site for fusion along the actin filament which is fined to a

SNARE hot spot at its plus-end, and stabilize the association of the vesicles with other proteins such as SNARE at the PM.

Future directions

It will be of interest to explore the effects of insulin on intracellular Ca2+ levels and on the activity of CaMKII just beneath the PM in the adipocyte. Fluorescence resonance energy transfer (FRET) based methods have been shown to be useful to study the temporal and spatial regulation of CaMKII activity (Kwok et al., 2008). This is done by fluorescence imaging of CaMKII activity by fusing CaMKII with donor and acceptor fluorescent proteins at its amino- and carboxyl-termini. When CaMKII is activated by elevated Ca2+, the fusion protein undergoes a conformational change which brings the donor and acceptor fluorescent proteins into close proximity (Kwok et al., 2008). By targeting this construct to the PM of adipocytes, this method can be used to study the localized CaMKII activity in response to insulin stimulation. In addition, understanding the relationship between the Ca2+/CaMKII-pathway and the

PI3K/Akt-pathway, which is the major insulin-activated pathway, is also important. In

149 fact, the requirement of Akt in the fusion step has been controversial. Activation of

Akt in adipocytes using a dimerizable construct that allows for Akt activation independent of PI3K activity results in full activation of GLUT4 translocation and glucose uptake (Ng et al., 2008). In contrast, TIRFM-based studies by Gonzalez et al. demonstrated that insulin-stimulated fusion of GSVs can occur independently of Akt, but requires PI3K activity (Gonzalez and McGraw, 2006). This supports the present study since insulin/PI3K-dependent Myo1c phosphorylation is not inhibited with an

Akt inhibitor. However, I cannot exclude the possibility that besides the

Akt-independent pathway, phosphorylation of Myo1c by CaMKII may be additionally facilitated by an Akt-dependent mechanism. For example, insulin may result in Ca2+ influx throught a PM channel in an Akt-independent and an Akt-dependent manner.

Therefore, activation of Akt alone may also lead to activation of CaMKII and hence phosphorylation of Myo1c. Further experiments examining the 14-3-3-binding of

Myo1c in response to Akt activation using the dimerizable construct would help clarify the role of Akt in Myo1c phosphorylation.

It will also be important to fully understand the functional role of Myo1c, particularly at the last few steps of GLUT4 trafficking. Future experiments will focus on using

TIRFM to analyse the effects of Myo1c knockdown or Armcx5 knockdown on tethering, docking and fusion which occur at the PM. In the case of other myosin motors an intimate relationship between myosin function and SNARE assembly at the

PM has been shown. For example, MyoVa directly associates with the t-SNARE

Syntaxin1a and facilitates its association with the v-SNARE (Watanabe et al., 2005).

It will be of interest to determine how the function of Armcx5/Myo1c complex may intersect with SNARE assembly in the adipocyte and as to whether such functions are mediated via phosphorylation, calmodulin release or 14-3-3 binding. Resolving the

150 crystal structure of the Armcx5/Myo1c complex under the basal or the insulin-stimulated condition will also help determine the mechanism by which this complex is localized and functions at the PM. Moreover, since the 14-3-3-binding of

Myo1c was not observed in the muscle cells, it will be of interest in the future to determine if other myosin motors may be regulated and act similarly at the PM in muscle cells.

Finally, the 14-3-3-pulldown followed by mass spectrometric analysis has been shown to be a useful approach to study the insulin-regulated phosphoproteome. This technique can be potentially applied in the discovery of protein markers of insulin resistance and type II diabetes. This, however, will require a more quantitative analysis. In this study, only the proteins purified from insulin-treated cells were identified and the levels of most of these proteins have not yet been compared with those from basal cells. In order to quantitatively compare the protein levels,

14-3-3-affinity purification can be potentially combined with a quantitative mass spectrometry-based approach such as stable isotope labeling by amino acids in cell culture (SILAC). SILAC is a simple, inexpensive and accurate technique that can be used as a quantitative proteomic approach in any cell culture system. This technique allows in vivo incorporation of specific amino acids into all mammalian proteins (Ong et al., 2002; Ong et al., 2003). Basal 3T3-L1 adipocytes could be grown in normal media while the insulin-treated adipocytes will be grown in media lacking a standard essential amino acid, either lysine or arginine, but supplemented with a non-radioactive, isotopically labeled form of that amino acid. Protein populations from both samples could be mixed directly after harvesting prior to 14-3-3-affinity purification, SDS-PAGE gel separation and mass spectrometry identification. Peptides with the same sequences from two different samples will appear at different m/z ratios

151 in the mass spectrum. This will allow quantitative comparison between the peptides with the same sequences from basal and insulin-treated cells (Blagoev et al., 2003;

Ibarrola et al., 2003; Ibarrola et al., 2004). Hence, insulin-regulated phosphoproteins can be identified directly without the needs for Western blot analysis. This technique can also be applied to tissues isolated from animals, a more physiologically relevant model of insulin resistance and type II diabetes. Isotope-labeled 3T3-L1 adipocytes or

L6 myotubes could be mixed with tissues isolated from normal or diabetic mice to serve as internal standards. Then the samples would be subjected to phosphoprotein purification and identification independently. The changes in protein level in the two tissue samples could be measured quantitatively (Ishihama et al., 2005). This allows proteome-wide analyses of insulin resistance in peripheral tissues, serving as a powerful tool in the discovery of anti-diabetic drugs.

Conclusion

In conclusion, understanding the full trafficking itinerary of GLUT4 is undoubtedly one of the major focuses in the field of diabetes research. The ultimate goal is to pinpoint the step that is defective in the insulin resistance state and to develop a targeted therapeutic strategy. The present study leads to the identification of a key insulin regulated substrate, Myo1c, and demonstrates its critical role at the PM in

GLUT4 translocation in adipocytes. Futher studies will focus on not only characterizing its role in docking/fusion, but also determining the effects of insulin resistance on its upstream signaling pathway and its functions. Additionally it is essential to identify the myosin isoform that plays a similar role in muscle. These important motor proteins may become useful targets for drug development for prevention or improvement of insulin resistance and type II diabetes.

152

Figure 6.1 Effects of wortmannin on activation of CaMKII and phosphorylation of Myo1c. The present study showed that wortamannin does not inhibit insulin-incuded activation of CaMKII but it abolishes phosphorylation of Myo1c.

153

Figure 6.2 Multiple roles of Myo1c in adipocytes. The present study proposed that Myo1c interacts

with its PM receptor, Armcx5, which binds to the PM. Insulin activates CaMKII by stimulating the

influx of Ca2+ via a Ca2+ channel. CaMKII-mediated phosphorylation of Myo1c is PI3K-dependent. It reduces calmodulin binding, induces 14-3-3-binding and increases the ATPase activity of Myo1c.

These activated Armcx5/Myo1c complexes serve as docking sites for vesicles, possibly via binding of

RalA to the IQ domains of Myo1c (Chen et al., 2007), and thereby bringing the vesicles across to the site for fusion and/or stabilize the association of the vesicles and SNARE. Myo1c has been reported to associate with Rictor to regulate actin remodeling (Hagan et al., 2008) and with the NEMO/IKK-γ complex to mediate TNF-α stimulated phosphorylation of IRS-1 (Nakamori et al., 2006). These different pools of Myo1c are likely regulated by discrete pathways.

154 Chapter 7

References

155 Aberle, H., Schwartz, H., Hoschuetzky, H., and Kemler, R. (1996). Single amino acid substitutions in proteins of the armadillo gene family abolish their binding to alpha-catenin. J Biol Chem 271, 1520-1526.

Ai, H., Ralston, E., Lauritzen, H.P., Galbo, H., and Ploug, T. (2003). Disruption of microtubules in rat skeletal muscle does not inhibit insulin- or contraction-stimulated glucose transport. Am J Physiol Endocrinol Metab 285, E836-844.

Aitken, A. (2002). Functional specificity in 14-3-3 isoform interactions through dimer formation and phosphorylation. Chromosome location of mammalian isoforms and variants. Plant Mol Biol 50, 993-1010.

Al-Hasani, H., Kunamneni, R.K., Dawson, K., Hinck, C.S., Muller-Wieland, D., and Cushman, S.W. (2002). Roles of the N- and C-termini of GLUT4 in endocytosis. J Cell Sci 115, 131-140.

Alessi, D.R., James, S.R., Downes, C.P., Holmes, A.B., Gaffney, P.R., Reese, C.B., and Cohen, P. (1997). Characterization of a 3-phosphoinositide-dependent protein kinase which phosphorylates and activates protein kinase Balpha. Curr Biol 7, 261-269.

Antonescu, C.N., Huang, C., Niu, W., Liu, Z., Eyers, P.A., Heidenreich, K.A., Bilan, P.J., and Klip, A. (2005). Reduction of insulin-stimulated glucose uptake in L6 myotubes by the protein kinase inhibitor SB203580 is independent of p38MAPK activity. Endocrinology 146, 3773-3781.

Araki, S., Yang, J., Hashiramoto, M., Tamori, Y., Kasuga, M., and Holman, G.D. (1996). Subcellular trafficking kinetics of GLU4 mutated at the N- and C-terminal. Biochem J 315 ( Pt 1), 153-159.

Bai, L., Wang, Y., Fan, J., Chen, Y., Ji, W., Qu, A., Xu, P., James, D.E., and Xu, T. (2007). Dissecting multiple steps of GLUT4 trafficking and identifying the sites of insulin action. Cell Metab 5, 47-57.

Bartel, P., Chien, C.T., Sternglanz, R., and Fields, S. (1993). Elimination of false positives that arise in using the two-hybrid system. BioTechniques 14, 920-924.

Barylko, B., Binns, D.D., and Albanesi, J.P. (2000). Regulation of the enzymatic and motor activities of myosin I. Biochim Biophys Acta 1496, 23-35.

Barylko, B., Jung, G., and Albanesi, J.P. (2005). Structure, function, and regulation of myosin 1C. Acta Biochim Pol 52, 373-380.

Barylko, B., Wagner, M.C., Reizes, O., and Albanesi, J.P. (1992). Purification and

156 characterization of a mammalian myosin I. Proc Natl Acad Sci U S A 89, 490-494.

Batters, C., Wallace, M.I., Coluccio, L.M., and Molloy, J.E. (2004). A model of stereocilia adaptation based on single molecule mechanical studies of myosin I. Philos Trans R Soc Lond B Biol Sci 359, 1895-1905.

Belanger, C., Ansanay, H., Qanbar, R., and Bouvier, M. (2001). Primary sequence requirements for S-acylation of beta(2)-adrenergic receptor peptides. FEBS Lett 499, 59-64.

Benzinger, A., Muster, N., Koch, H.B., Yates, J.R., 3rd, and Hermeking, H. (2005). Targeted proteomic analysis of 14-3-3 sigma, a p53 effector commonly silenced in cancer. Mol Cell Proteomics 4, 785-795.

Blagoev, B., Kratchmarova, I., Ong, S.E., Nielsen, M., Foster, L.J., and Mann, M. (2003). A proteomics strategy to elucidate functional protein-protein interactions applied to EGF signaling. Nat Biotechnol 21, 315-318.

Bogan, J.S., Hendon, N., McKee, A.E., Tsao, T.S., and Lodish, H.F. (2003). Functional cloning of TUG as a regulator of GLUT4 glucose transporter trafficking. Nature 425, 727-733.

Bose, A., Guilherme, A., Robida, S.I., Nicoloro, S.M., Zhou, Q.L., Jiang, Z.Y., Pomerleau, D.P., and Czech, M.P. (2002). Glucose transporter recycling in response to insulin is facilitated by myosin Myo1c. Nature 420, 821-824.

Bose, A., Robida, S., Furcinitti, P.S., Chawla, A., Fogarty, K., Corvera, S., and Czech, M.P. (2004). Unconventional myosin Myo1c promotes membrane fusion in a regulated exocytic pathway. Mol Cell Biol 24, 5447-5458.

Bouzakri, K., Zachrisson, A., Al-Khalili, L., Zhang, B.B., Koistinen, H.A., Krook, A., and Zierath, J.R. (2006). siRNA-based gene silencing reveals specialized roles of IRS-1/Akt2 and IRS-2/Akt1 in glucose and lipid metabolism in human skeletal muscle. Cell Metab 4, 89-96.

Bridges, D., and Moorhead, G.B. (2005). 14-3-3 proteins: a number of functions for a numbered protein. Sci STKE 2005, re10.

Brozinick, J.T., Jr., Hawkins, E.D., Strawbridge, A.B., and Elmendorf, J.S. (2004). Disruption of cortical actin in skeletal muscle demonstrates an essential role of the cytoskeleton in glucose transporter 4 translocation in insulin-sensitive tissues. J Biol Chem 279, 40699-40706.

Brozinick, J.T., Jr., Reynolds, T.H., Dean, D., Cartee, G., and Cushman, S.W. (1999).

157 1-[N, O-bis-(5-isoquinolinesulphonyl)-N-methyl-L-tyrosyl]-4- phenylpiperazine (KN-62), an inhibitor of calcium-dependent camodulin protein kinase II, inhibits both insulin- and hypoxia-stimulated glucose transport in skeletal muscle. Biochem J 339 ( Pt 3), 533-540.

Brunet, A., Bonni, A., Zigmond, M.J., Lin, M.Z., Juo, P., Hu, L.S., Anderson, M.J., Arden, K.C., Blenis, J., and Greenberg, M.E. (1999). Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell 96, 857-868.

Bruton, J.D., Katz, A., and Westerblad, H. (1999). Insulin increases near-membrane but not global Ca2+ in isolated skeletal muscle. Proc Natl Acad Sci U S A 96, 3281-3286.

Bryant, N.J., Govers, R., and James, D.E. (2002). Regulated transport of the glucose transporter GLUT4. Nat Rev Mol Cell Biol 3, 267-277.

Bultman, S.J., Gebuhr, T.C., and Magnuson, T. (2005). A Brg1 mutation that uncouples ATPase activity from chromatin remodeling reveals an essential role for SWI/SNF-related complexes in beta-globin expression and erythroid development. Genes Dev 19, 2849-2861.

Bunney, T.D., van Walraven, H.S., and de Boer, A.H. (2001). 14-3-3 protein is a regulator of the mitochondrial and chloroplast ATP synthase. Proc Natl Acad Sci U S A 98, 4249-4254.

Buss, F., Spudich, G., and Kendrick-Jones, J. (2004). Myosin VI: cellular functions and motor properties. Annual review of cell and developmental biology 20, 649-676.

Cain, C.C., Trimble, W.S., and Lienhard, G.E. (1992). Members of the VAMP family of synaptic vesicle proteins are components of glucose transporter-containing vesicles from rat adipocytes. J Biol Chem 267, 11681-11684.

Cantley, L.C. (2002). The phosphoinositide 3-kinase pathway. Science 296, 1655-1657.

Chacko, S., Jacob, S.S., and Horiuchi, K.Y. (1994). Myosin I from mammalian smooth muscle is regulated by caldesmon-calmodulin. J Biol Chem 269, 15803-15807.

Chamberlain, L.H., and Gould, G.W. (2002). The vesicle- and target-SNARE proteins that mediate Glut4 vesicle fusion are localized in detergent-insoluble lipid rafts present on distinct intracellular membranes. J Biol Chem 277, 49750-49754.

Chapman, E.R., Au, D., Alexander, K.A., Nicolson, T.A., and Storm, D.R. (1991).

158 Characterization of the calmodulin binding domain of neuromodulin. Functional significance of serine 41 and phenylalanine 42. J Biol Chem 266, 207-213.

Chaudhri, M., Scarabel, M., and Aitken, A. (2003). Mammalian and yeast 14-3-3 isoforms form distinct patterns of dimers in vivo. Biochem Biophys Res Commun 300, 679-685.

Chen, F., and Wagner, P.D. (1994). 14-3-3 proteins bind to histone and affect both histone phosphorylation and dephosphorylation. FEBS Lett 347, 128-132.

Chen, X.Q., and Yu, A.C. (2002). The association of 14-3-3gamma and actin plays a role in cell division and apoptosis in astrocytes. Biochem Biophys Res Commun 296, 657-663.

Chen, X.W., Leto, D., Chiang, S.H., Wang, Q., and Saltiel, A.R. (2007). Activation of RalA Is Required for Insulin-Stimulated Glut4 Trafficking to the Plasma Membrane via the Exocyst and the Motor Protein Myo1c. Dev Cell 13, 391-404.

Coates, J.C. (2003). Armadillo repeat proteins: beyond the animal kingdom. Trends in cell biology 13, 463-471.

Colbran, R.J., Fong, Y.L., Schworer, C.M., and Soderling, T.R. (1988). Regulatory interactions of the calmodulin-binding, inhibitory, and autophosphorylation domains of Ca2+/calmodulin-dependent protein kinase II. J Biol Chem 263, 18145-18151.

Conti, E., and Izaurralde, E. (2001). Nucleocytoplasmic transport enters the atomic age. Curr Opin Cell Biol 13, 310-319.

Cormont, M., Tanti, J.F., Zahraoui, A., Van Obberghen, E., Tavitian, A., and Le Marchand-Brustel, Y. (1993). Insulin and okadaic acid induce Rab4 redistribution in adipocytes. J Biol Chem 268, 19491-19497.

Costa, M.C., Mani, F., Santoro, W., Jr., Espreafico, E.M., and Larson, R.E. (1999). Brain myosin-V, a calmodulin-carrying myosin, binds to calmodulin-dependent protein kinase II and activates its kinase activity. J Biol Chem 274, 15811-15819.

Craparo, A., Freund, R., and Gustafson, T.A. (1997). 14-3-3 (epsilon) interacts with the insulin-like growth factor I receptor and insulin receptor substrate I in a phosphoserine-dependent manner. J Biol Chem 272, 11663-11669.

Daniels, D.L., Eklof Spink, K., and Weis, W.I. (2001). beta-catenin: molecular plasticity and drug design. Trends Biochem Sci 26, 672-678.

Davies, K.D., Alvestad, R.M., Coultrap, S.J., and Browning, M.D. (2007).

159 alphaCaMKII autophosphorylation levels differ depending on subcellular localization. Brain Res 1158, 39-49.

De La Cruz, E.M., and Ostap, E.M. (2004). Relating biochemistry and function in the myosin superfamily. Curr Opin Cell Biol 16, 61-67.

DeFeo-Jones, D., Barnett, S.F., Fu, S., Hancock, P.J., Haskell, K.M., Leander, K.R., McAvoy, E., Robinson, R.G., Duggan, M.E., Lindsley, C.W., Zhao, Z., Huber, H.E., and Jones, R.E. (2005). Tumor cell sensitization to apoptotic stimuli by selective inhibition of specific Akt/PKB family members. Mol Cancer Ther 4, 271-279.

Deneka, M., Neeft, M., and van der Sluijs, P. (2003). Regulation of membrane transport by rab GTPases. Crit Rev Biochem Mol Biol 38, 121-142.

Dougherty, M.K., and Morrison, D.K. (2004). Unlocking the code of 14-3-3. J Cell Sci 117, 1875-1884.

Durrbach, A., Raposo, G., Tenza, D., Louvard, D., and Coudrier, E. (2000). Truncated brush border myosin I affects membrane traffic in polarized epithelial cells. Traffic 1, 411-424.

Ebina, Y., Ellis, L., Jarnagin, K., Edery, M., Graf, L., Clauser, E., Ou, J.H., Masiarz, F., Kan, Y.W., Goldfine, I.D., and et al. (1985). The human insulin receptor cDNA: the structural basis for hormone-activated transmembrane signalling. Cell 40, 747-758.

El Mezgueldi, M., Tang, N., Rosenfeld, S.S., and Ostap, E.M. (2002). The kinetic mechanism of Myo1e (human myosin-IC). J Biol Chem 277, 21514-21521.

Emoto, M., Langille, S.E., and Czech, M.P. (2001). A role for kinesin in insulin-stimulated GLUT4 glucose transporter translocation in 3T3-L1 adipocytes. J Biol Chem 276, 10677-10682.

Etournay, R., El-Amraoui, A., Bahloul, A., Blanchard, S., Roux, I., Pezeron, G., Michalski, N., Daviet, L., Hardelin, J.P., Legrain, P., and Petit, C. (2005). PHR1, an integral membrane protein of the inner ear sensory cells, directly interacts with myosin 1c and myosin VIIa. J Cell Sci 118, 2891-2899.

Evans, R.M., Barish, G.D., and Wang, Y.X. (2004). PPARs and the complex journey to obesity. Nat Med 10, 355-361.

Ewart, M.A., Clarke, M., Kane, S., Chamberlain, L.H., and Gould, G.W. (2005). Evidence for a role of the exocyst in insulin-stimulated Glut4 trafficking in 3T3-L1 adipocytes. J Biol Chem 280, 3812-3816.

160 Finehout, E.J., and Lee, K.H. (2003). Comparison of automated in-gel digest methods for femtomole level samples. Electrophoresis 24, 3508-3516.

Flatt, J.P. (1995). Use and storage of carbohydrate and fat. The American journal of clinical nutrition 61, 952S-959S.

Fleckenstein, D., Rohde, M., Klionsky, D.J., and Rudiger, M. (1998). Yel013p (Vac8p), an armadillo repeat protein related to plakoglobin and importin alpha is associated with the yeast vacuole membrane. Journal of cell science 111 ( Pt 20), 3109-3118.

Fletcher, L.M., Welsh, G.I., Oatey, P.B., and Tavare, J.M. (2000). Role for the microtubule cytoskeleton in GLUT4 vesicle trafficking and in the regulation of insulin-stimulated glucose uptake. Biochem J 352 Pt 2, 267-276.

Foster, L.J., and Klip, A. (2000). Mechanism and regulation of GLUT-4 vesicle fusion in muscle and fat cells. Am J Physiol Cell Physiol 279, C877-890.

Foster, L.J., Rudich, A., Talior, I., Patel, N., Huang, X., Furtado, L.M., Bilan, P.J., Mann, M., and Klip, A. (2006). Insulin-dependent interactions of proteins with GLUT4 revealed through stable isotope labeling by amino acids in cell culture (SILAC). J Proteome Res 5, 64-75.

Fresno Vara, J.A., Casado, E., de Castro, J., Cejas, P., Belda-Iniesta, C., and Gonzalez-Baron, M. (2004). PI3K/Akt signalling pathway and cancer. Cancer Treat Rev 30, 193-204.

Froissard, M., Keller, A.M., and Cohen, J. (2001). ND9P, a novel protein with armadillo-like repeats involved in exocytosis: physiological studies using allelic mutants in paramecium. Genetics 157, 611-620.

Fu, H., Subramanian, R.R., and Masters, S.C. (2000). 14-3-3 proteins: structure, function, and regulation. Annu Rev Pharmacol Toxicol 40, 617-647.

Fuglsang, A.T., Borch, J., Bych, K., Jahn, T.P., Roepstorff, P., and Palmgren, M.G. (2003). The binding site for regulatory 14-3-3 protein in plant plasma membrane H+-ATPase: involvement of a region promoting phosphorylation-independent interaction in addition to the phosphorylation-dependent C-terminal end. J Biol Chem 278, 42266-42272.

Garza, L.A., and Birnbaum, M.J. (2000). Insulin-responsive aminopeptidase trafficking in 3T3-L1 adipocytes. J Biol Chem 275, 2560-2567.

Gillespie, P.G., and Cyr, J.L. (2002). Calmodulin binding to recombinant myosin-1c

161 and myosin-1c IQ peptides. BMC Biochem 3, 31.

Gillespie, P.G., and Cyr, J.L. (2004). Myosin-1c, the hair cell's adaptation motor. Annu Rev Physiol 66, 521-545.

Gillingham, A.K., Koumanov, F., Pryor, P.R., Reaves, B.J., and Holman, G.D. (1999). Association of AP1 adaptor complexes with GLUT4 vesicles. J Cell Sci 112 ( Pt 24), 4793-4800.

Gonzalez, E., and McGraw, T.E. (2006). Insulin signaling diverges into Akt-dependent and -independent signals to regulate the recruitment/docking and the fusion of GLUT4 vesicles to the plasma membrane. Mol Biol Cell 17, 4484-4493.

Gould, G.W., and Holman, G.D. (1993). The glucose transporter family: structure, function and tissue-specific expression. Biochem J 295 ( Pt 2), 329-341.

Gronborg, M., Kristiansen, T.Z., Stensballe, A., Andersen, J.S., Ohara, O., Mann, M., Jensen, O.N., and Pandey, A. (2002). A mass spectrometry-based proteomic approach for identification of serine/threonine-phosphorylated proteins by enrichment with phospho-specific antibodies: identification of a novel protein, Frigg, as a protein kinase A substrate. Mol Cell Proteomics 1, 517-527.

Grosshans, B.L., Ortiz, D., and Novick, P. (2006). Rabs and their effectors: achieving specificity in membrane traffic. Proc Natl Acad Sci U S A 103, 11821-11827.

Grusovin, J., and Macaulay, S.L. (2003). Snares for GLUT4--mechanisms directing vesicular trafficking of GLUT4. Front Biosci 8, d620-641.

Guilherme, A., Emoto, M., Buxton, J.M., Bose, S., Sabini, R., Theurkauf, W.E., Leszyk, J., and Czech, M.P. (2000). Perinuclear localization and insulin responsiveness of GLUT4 requires cytoskeletal integrity in 3T3-L1 adipocytes. J Biol Chem 275, 38151-38159.

Gupta, N., Wollscheid, B., Watts, J.D., Scheer, B., Aebersold, R., and DeFranco, A.L. (2006). Quantitative proteomic analysis of B cell lipid rafts reveals that ezrin regulates antigen receptor-mediated lipid raft dynamics. Nature immunology 7, 625-633.

Hagan, G.N., Lin, Y., Magnuson, M.A., Avruch, J., and Czech, M.P. (2008). A Rictor-Myo1c complex participates in dynamic cortical actin events in 3T3-L1 adipocytes. Mol Cell Biol 28, 4215-4226.

Hahn, M.E., and Karchner, S.I. (1995). Evolutionary conservation of the vertebrate Ah (dioxin) receptor: amplification and sequencing of the PAS domain of a teleost Ah receptor cDNA. Biochem J 310 ( Pt 2), 383-387.

162 Hales, C.M., Vaerman, J.P., and Goldenring, J.R. (2002). Rab11 family interacting protein 2 associates with Myosin Vb and regulates plasma membrane recycling. J Biol Chem 277, 50415-50421.

Hallberg, B. (2002). Exoenzyme S binds its cofactor 14-3-3 through a non-phosphorylated motif. Biochem Soc Trans 30, 401-405.

Hokanson, D.E., Laakso, J.M., Lin, T., Sept, D., and Ostap, E.M. (2006). Myo1c binds phosphoinositides through a putative pleckstrin homology domain. Mol Biol Cell 17, 4856-4865.

Hornbeck, P.V., Chabra, I., Kornhauser, J.M., Skrzypek, E., and Zhang, B. (2004). PhosphoSite: A bioinformatics resource dedicated to physiological protein phosphorylation. Proteomics 4, 1551-1561.

Hou, J.C., Suzuki, N., Pessin, J.E., and Watson, R.T. (2006). A specific dileucine motif is required for the GGA-dependent entry of newly synthesized insulin-responsive aminopeptidase into the insulin-responsive compartment. J Biol Chem 281, 33457-33466.

Houdusse, A., Gaucher, J.F., Krementsova, E., Mui, S., Trybus, K.M., and Cohen, C. (2006). Crystal structure of apo-calmodulin bound to the first two IQ motifs of myosin V reveals essential recognition features. Proc Natl Acad Sci U S A 103, 19326-19331.

Huang, J., Imamura, T., Babendure, J.L., Lu, J.C., and Olefsky, J.M. (2005). Disruption of microtubules ablates the specificity of insulin signaling to GLUT4 translocation in 3T3-L1 adipocytes. J Biol Chem 280, 42300-42306.

Huang, S., and Czech, M.P. (2007). The GLUT4 glucose transporter. Cell Metab 5, 237-252.

Huang, S., Lifshitz, L.M., Jones, C., Bellve, K.D., Standley, C., Fonseca, S., Corvera, S., Fogarty, K.E., and Czech, M.P. (2007). Insulin stimulates membrane fusion and GLUT4 accumulation in clathrin coats on adipocyte plasma membranes. Mol Cell Biol 27, 3456-3469.

Huber, A.H., Nelson, W.J., and Weis, W.I. (1997). Three-dimensional structure of the armadillo repeat region of beta-catenin. Cell 90, 871-882.

Ibarrola, N., Kalume, D.E., Gronborg, M., Iwahori, A., and Pandey, A. (2003). A proteomic approach for quantitation of phosphorylation using stable isotope labeling in cell culture. Anal Chem 75, 6043-6049.

163 Ibarrola, N., Molina, H., Iwahori, A., and Pandey, A. (2004). A novel proteomic approach for specific identification of tyrosine kinase substrates using [13C]tyrosine. J Biol Chem 279, 15805-15813.

Imam-Sghiouar, N., Laude-Lemaire, I., Labas, V., Pflieger, D., Le Caer, J.P., Caron, M., Nabias, D.K., and Joubert-Caron, R. (2002). Subproteomics analysis of phosphorylated proteins: application to the study of B-lymphoblasts from a patient with Scott syndrome. Proteomics 2, 828-838.

Imamura, T., Huang, J., Usui, I., Satoh, H., Bever, J., and Olefsky, J.M. (2003). Insulin-induced GLUT4 translocation involves protein kinase C-lambda-mediated functional coupling between Rab4 and the motor protein kinesin. Mol Cell Biol 23, 4892-4900.

Inagaki, N., Nishizawa, M., Arimura, N., Yamamoto, H., Takeuchi, Y., Miyamoto, E., Kaibuchi, K., and Inagaki, M. (2000). Activation of Ca2+/calmodulin-dependent protein kinase II within post-synaptic dendritic spines of cultured hippocampal neurons. J Biol Chem 275, 27165-27171.

Inoue, M., Chang, L., Hwang, J., Chiang, S.H., and Saltiel, A.R. (2003). The exocyst complex is required for targeting of Glut4 to the plasma membrane by insulin. Nature 422, 629-633.

Inoue, M., Chiang, S.H., Chang, L., Chen, X.W., and Saltiel, A.R. (2006). Compartmentalization of the exocyst complex in lipid rafts controls Glut4 vesicle tethering. Mol Biol Cell 17, 2303-2311.

Ishihama, Y., Sato, T., Tabata, T., Miyamoto, N., Sagane, K., Nagasu, T., and Oda, Y. (2005). Quantitative mouse brain proteomics using culture-derived isotope tags as internal standards. Nat Biotechnol 23, 617-621.

Ishiki, M., and Klip, A. (2005). Minireview: recent developments in the regulation of glucose transporter-4 traffic: new signals, locations, and partners. Endocrinology 146, 5071-5078.

Ishikura, S., Bilan, P.J., and Klip, A. (2007). Rabs 8A and 14 are targets of the insulin-regulated Rab-GAP AS160 regulating GLUT4 traffic in muscle cells. Biochem Biophys Res Commun 353, 1074-1079.

Ishikura, S., Koshkina, A., and Klip, A. (2008). Small G proteins in insulin action: Rab and Rho families at the crossroads of signal transduction and GLUT4 vesicle traffic. Acta physiologica (Oxford, England) 192, 61-74.

164 Jahn, R., and Scheller, R.H. (2006). SNAREs--engines for membrane fusion. Nat Rev Mol Cell Biol 7, 631-643.

James, D.E., Brown, R., Navarro, J., and Pilch, P.F. (1988). Insulin-regulatable tissues express a unique insulin-sensitive glucose transport protein. Nature 333, 183-185.

Jiang, Z.Y., Chawla, A., Bose, A., Way, M., and Czech, M.P. (2002). A phosphatidylinositol 3-kinase-independent insulin signaling pathway to N-WASP/Arp2/3/F-actin required for GLUT4 glucose transporter recycling. J Biol Chem 277, 509-515.

Jin, J., Smith, F.D., Stark, C., Wells, C.D., Fawcett, J.P., Kulkarni, S., Metalnikov, P., O'Donnell, P., Taylor, P., Taylor, L., Zougman, A., Woodgett, J.R., Langeberg, L.K., Scott, J.D., and Pawson, T. (2004). Proteomic, functional, and domain-based analysis of in vivo 14-3-3 binding proteins involved in cytoskeletal regulation and cellular organization. Curr Biol 14, 1436-1450.

Joost, H.G., and Thorens, B. (2001). The extended GLUT-family of sugar/polyol transport facilitators: nomenclature, sequence characteristics, and potential function of its novel members (review). Molecular membrane biology 18, 247-256.

Kaddai, V., Le Marchand-Brustel, Y., and Cormont, M. (2008). Rab proteins in endocytosis and Glut4 trafficking. Acta physiologica (Oxford, England) 192, 75-88.

Kandror, K.V., Yu, L., and Pilch, P.F. (1994). The major protein of GLUT4-containing vesicles, gp160, has aminopeptidase activity. J Biol Chem 269, 30777-30780.

Kane, S., Sano, H., Liu, S.C., Asara, J.M., Lane, W.S., Garner, C.C., and Lienhard, G.E. (2002). A method to identify serine kinase substrates. Akt phosphorylates a novel adipocyte protein with a Rab GTPase-activating protein (GAP) domain. J Biol Chem 277, 22115-22118.

Kanzaki, M., and Pessin, J.E. (2001). Insulin-stimulated GLUT4 translocation in adipocytes is dependent upon cortical actin remodeling. J Biol Chem 276, 42436-42444.

Kao, A.W., Noda, Y., Johnson, J.H., Pessin, J.E., and Saltiel, A.R. (1999). Aldolase mediates the association of F-actin with the insulin-responsive glucose transporter GLUT4. J Biol Chem 274, 17742-17747.

Kasuga, M., Zick, Y., Blithe, D.L., Crettaz, M., and Kahn, C.R. (1982). Insulin stimulates tyrosine phosphorylation of the insulin receptor in a cell-free system. Nature 298, 667-669.

165 Kawanishi, M., Tamori, Y., Okazawa, H., Araki, S., Shinoda, H., and Kasuga, M. (2000). Role of SNAP23 in insulin-induced translocation of GLUT4 in 3T3-L1 adipocytes. Mediation of complex formation between syntaxin4 and VAMP2. J Biol Chem 275, 8240-8247.

Keller, S.R., Davis, A.C., and Clairmont, K.B. (2002). Mice deficient in the insulin-regulated membrane aminopeptidase show substantial decreases in glucose transporter GLUT4 levels but maintain normal glucose homeostasis. J Biol Chem 277, 17677-17686.

Keller, S.R., Scott, H.M., Mastick, C.C., Aebersold, R., and Lienhard, G.E. (1995). Cloning and characterization of a novel insulin-regulated membrane aminopeptidase from Glut4 vesicles. J Biol Chem 270, 23612-23618.

Khayat, Z.A., Tong, P., Yaworsky, K., Bloch, R.J., and Klip, A. (2000). Insulin-induced actin filament remodeling colocalizes actin with phosphatidylinositol 3-kinase and GLUT4 in L6 myotubes. J Cell Sci 113 Pt 2, 279-290.

Konstantopoulos, N., Marcuccio, S., Kyi, S., Stoichevska, V., Castelli, L.A., Ward, C.W., and Macaulay, S.L. (2007). A purine analog kinase inhibitor, calcium/calmodulin-dependent protein kinase II inhibitor 59, reveals a role for calcium/calmodulin-dependent protein kinase II in insulin-stimulated glucose transport. Endocrinology 148, 374-385.

Kosaki, A., Yamada, K., Suga, J., Otaka, A., and Kuzuya, H. (1998). 14-3-3beta protein associates with insulin receptor substrate 1 and decreases insulin-stimulated phosphatidylinositol 3'-kinase activity in 3T3L1 adipocytes. J Biol Chem 273, 940-944.

Koumanov, F., Jin, B., Yang, J., and Holman, G.D. (2005). Insulin signaling meets vesicle traffic of GLUT4 at a plasma-membrane-activated fusion step. Cell Metab 2, 179-189.

Kussel-Andermann, P., El-Amraoui, A., Safieddine, S., Nouaille, S., Perfettini, I., Lecuit, M., Cossart, P., Wolfrum, U., and Petit, C. (2000). Vezatin, a novel transmembrane protein, bridges myosin VIIA to the cadherin-catenins complex. Embo J 19, 6020-6029.

Kwok, S., Lee, C., Sanchez, S.A., Hazlett, T.L., Gratton, E., and Hayashi, Y. (2008). Genetically encoded probe for fluorescence lifetime imaging of CaMKII activity. Biochem Biophys Res Commun 369, 519-525.

Lanner, J.T., Bruton, J.D., Katz, A., and Westerblad, H. (2008). Ca(2+) and

166 insulin-mediated glucose uptake. Current opinion in pharmacology 8, 339-345.

Larance, M., Ramm, G., and James, D.E. (2008). The GLUT4 code. Mol Endocrinol 22, 226-233.

Larance, M., Ramm, G., Stockli, J., van Dam, E.M., Winata, S., Wasinger, V., Simpson, F., Graham, M., Junutula, J.R., Guilhaus, M., and James, D.E. (2005). Characterization of the role of the Rab GTPase-activating protein AS160 in insulin-regulated GLUT4 trafficking. J Biol Chem 280, 37803-37813.

Le Roith, D., and Zick, Y. (2001). Recent advances in our understanding of insulin action and insulin resistance. Diabetes Care 24, 588-597.

Li, L.V., and Kandror, K.V. (2005). Golgi-localized, gamma-ear-containing, Arf-binding protein adaptors mediate insulin-responsive trafficking of glucose transporter 4 in 3T3-L1 adipocytes. Mol Endocrinol 19, 2145-2153.

Li, Y., Wang, P., Xu, J., and Desir, G.V. (2006). Voltage-gated potassium channel Kv1.3 regulates GLUT4 trafficking to the plasma membrane via a Ca2+-dependent mechanism. Am J Physiol Cell Physiol 290, C345-351.

Light, Y., Paterson, H., and Marais, R. (2002). 14-3-3 antagonizes Ras-mediated Raf-1 recruitment to the plasma membrane to maintain signaling fidelity. Mol Cell Biol 22, 4984-4996.

Lin, B.Z., Pilch, P.F., and Kandror, K.V. (1997). Sortilin is a major protein component of Glut4-containing vesicles. J Biol Chem 272, 24145-24147.

Liu, L.B., Omata, W., Kojima, I., and Shibata, H. (2003). Insulin recruits GLUT4 from distinct compartments via distinct traffic pathways with differential microtubule dependence in rat adipocytes. J Biol Chem 278, 30157-30169.

Lizunov, V.A., Matsumoto, H., Zimmerberg, J., Cushman, S.W., and Frolov, V.A. (2005). Insulin stimulates the halting, tethering, and fusion of mobile GLUT4 vesicles in rat adipose cells. J Cell Biol 169, 481-489.

Macaulay, S.L., Hewish, D.R., Gough, K.H., Stoichevska, V., MacPherson, S.F., Jagadish, M., and Ward, C.W. (1997). Functional studies in 3T3L1 cells support a role for SNARE proteins in insulin stimulation of GLUT4 translocation. Biochem J 324 ( Pt 1), 217-224.

Machida, K., Mayer, B.J., and Nollau, P. (2003). Profiling the global tyrosine phosphorylation state. Mol Cell Proteomics 2, 215-233.

167 Malik, H.S., Eickbush, T.H., and Goldfarb, D.S. (1997). Evolutionary specialization of the nuclear targeting apparatus. Proc Natl Acad Sci U S A 94, 13738-13742.

Masters, S.C., Pederson, K.J., Zhang, L., Barbieri, J.T., and Fu, H. (1999). Interaction of 14-3-3 with a nonphosphorylated protein ligand, exoenzyme S of Pseudomonas aeruginosa. Biochemistry 38, 5216-5221.

Mastick, C.C., and Falick, A.L. (1997). Association of N-ethylmaleimide sensitive fusion (NSF) protein and soluble NSF attachment proteins-alpha and -gamma with glucose transporter-4-containing vesicles in primary rat adipocytes. Endocrinology 138, 2391-2397.

Matern, H.T., Yeaman, C., Nelson, W.J., and Scheller, R.H. (2001). The Sec6/8 complex in mammalian cells: characterization of mammalian Sec3, subunit interactions, and expression of subunits in polarized cells. Proc Natl Acad Sci U S A 98, 9648-9653.

McManus, E.J., and Alessi, D.R. (2002). TSC1-TSC2: a complex tale of PKB-mediated S6K regulation. Nat Cell Biol 4, E214-216.

McNiven, M.A., and Thompson, H.M. (2006). Vesicle formation at the plasma membrane and trans-Golgi network: the same but different. Science 313, 1591-1594.

Meek, S.E., Lane, W.S., and Piwnica-Worms, H. (2004). Comprehensive proteomic analysis of interphase and mitotic 14-3-3-binding proteins. J Biol Chem 279, 32046-32054.

Miinea, C.P., Sano, H., Kane, S., Sano, E., Fukuda, M., Peranen, J., Lane, W.S., and Lienhard, G.E. (2005). AS160, the Akt substrate regulating GLUT4 translocation, has a functional Rab GTPase-activating protein domain. Biochem J 391, 87-93.

Min, J., Okada, S., Kanzaki, M., Elmendorf, J.S., Coker, K.J., Ceresa, B.P., Syu, L.J., Noda, Y., Saltiel, A.R., and Pessin, J.E. (1999). Synip: a novel insulin-regulated syntaxin 4-binding protein mediating GLUT4 translocation in adipocytes. Mol Cell 3, 751-760.

Molero, J.C., Whitehead, J.P., Meerloo, T., and James, D.E. (2001). Nocodazole inhibits insulin-stimulated glucose transport in 3T3-L1 adipocytes via a microtubule-independent mechanism. J Biol Chem 276, 43829-43835.

Moore, B.W., and Perez, V.J. (1967). Specific acidic proteins of the nervous system. In Physiological and Biochemical Aspects of Nervous Integration. F.D. Carlson, ed. (Englewood Cliffs: Prentice-Hall), pp. 343-359.

168 Moorhead, G., Douglas, P., Cotelle, V., Harthill, J., Morrice, N., Meek, S., Deiting, U., Stitt, M., Scarabel, M., Aitken, A., and MacKintosh, C. (1999). Phosphorylation-dependent interactions between enzymes of plant metabolism and 14-3-3 proteins. Plant J 18, 1-12.

Morris, N.J., Ross, S.A., Lane, W.S., Moestrup, S.K., Petersen, C.M., Keller, S.R., and Lienhard, G.E. (1998). Sortilin is the major 110-kDa protein in GLUT4 vesicles from adipocytes. J Biol Chem 273, 3582-3587.

Munson, M., and Novick, P. (2006). The exocyst defrocked, a framework of rods revealed. Nature structural & molecular biology 13, 577-581.

Nakamori, Y., Emoto, M., Fukuda, N., Taguchi, A., Okuya, S., Tajiri, M., Miyagishi, M., Taira, K., Wada, Y., and Tanizawa, Y. (2006). Myosin motor Myo1c and its receptor NEMO/IKK-gamma promote TNF-alpha-induced serine307 phosphorylation of IRS-1. J Cell Biol 173, 665-671.

Ng, Y., Ramm, G., Lopez, J.A., and James, D.E. (2008). Rapid activation of Akt2 is sufficient to stimulate GLUT4 translocation in 3T3-L1 adipocytes. Cell Metab 7, 348-356.

Obenauer, J.C., Cantley, L.C., and Yaffe, M.B. (2003). Scansite 2.0: Proteome-wide prediction of cell signaling interactions using short sequence motifs. Nucleic Acids Res 31, 3635-3641.

Okada, S., Ohshima, K., Uehara, Y., Shimizu, H., Hashimoto, K., Yamada, M., and Mori, M. (2007). Synip phosphorylation is required for insulin-stimulated Glut4 translocation. Biochem Biophys Res Commun 356, 102-106.

Okada, T., Kawano, Y., Sakakibara, T., Hazeki, O., and Ui, M. (1994). Essential role of phosphatidylinositol 3-kinase in insulin-induced glucose transport and antilipolysis in rat adipocytes. Studies with a selective inhibitor wortmannin. J Biol Chem 269, 3568-3573.

Omata, W., Shibata, H., Li, L., Takata, K., and Kojima, I. (2000). Actin filaments play a critical role in insulin-induced exocytotic recruitment but not in endocytosis of GLUT4 in isolated rat adipocytes. Biochem J 346 Pt 2, 321-328.

Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, H., Pandey, A., and Mann, M. (2002). Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics 1, 376-386.

169 Ong, S.E., Foster, L.J., and Mann, M. (2003). Mass spectrometric-based approaches in quantitative proteomics. Methods 29, 124-130.

Pan, X., and Goldfarb, D.S. (1998). YEB3/VAC8 encodes a myristylated armadillo protein of the Saccharomyces cerevisiae vacuolar membrane that functions in vacuole fusion and inheritance. Journal of cell science 111 ( Pt 15), 2137-2147.

Pan, X., Roberts, P., Chen, Y., Kvam, E., Shulga, N., Huang, K., Lemmon, S., and Goldfarb, D.S. (2000). Nucleus-vacuole junctions in Saccharomyces cerevisiae are formed through the direct interaction of Vac8p with Nvj1p. Mol Biol Cell 11, 2445-2457.

Patki, V., Buxton, J., Chawla, A., Lifshitz, L., Fogarty, K., Carrington, W., Tuft, R., and Corvera, S. (2001). Insulin action on GLUT4 traffic visualized in single 3T3-l1 adipocytes by using ultra-fast microscopy. Mol Biol Cell 12, 129-141.

Peck, G.R., Ye, S., Pham, V., Fernando, R.N., Macaulay, S.L., Chai, S.Y., and Albiston, A.L. (2006). Interaction of the Akt substrate, AS160, with the glucose transporter 4 vesicle marker protein, insulin-regulated aminopeptidase. Mol Endocrinol 20, 2576-2583.

Perrini, S., Natalicchio, A., Laviola, L., Belsanti, G., Montrone, C., Cignarelli, A., Minielli, V., Grano, M., De Pergola, G., Giorgino, R., and Giorgino, F. (2004). Dehydroepiandrosterone stimulates glucose uptake in human and murine adipocytes by inducing GLUT1 and GLUT4 translocation to the plasma membrane. Diabetes 53, 41-52.

Pevsner, J., Hsu, S.C., Braun, J.E., Calakos, N., Ting, A.E., Bennett, M.K., and Scheller, R.H. (1994). Specificity and regulation of a synaptic vesicle docking complex. Neuron 13, 353-361.

Phillips, K.R., Tong, S., Goodyear, R., Richardson, G.P., and Cyr, J.L. (2006). Stereociliary myosin-1c receptors are sensitive to calcium chelation and absent from cadherin 23 mutant mice. J Neurosci 26, 10777-10788.

Pilch, P.F. (2008). The mass action hypothesis: formation of Glut4 storage vesicles, a tissue-specific, regulated exocytic compartment. Acta physiologica (Oxford, England) 192, 89-101.

Pollard, T.D., and Korn, E.D. (1973). Acanthamoeba myosin. I. Isolation from Acanthamoeba castellanii of an enzyme similar to muscle myosin. J Biol Chem 248, 4682-4690.

170 Printen, J.A., Brady, M.J., and Saltiel, A.R. (1997). PTG, a protein phosphatase 1-binding protein with a role in glycogen metabolism. Science 275, 1475-1478.

Ramm, G., Larance, M., Guilhaus, M., and James, D.E. (2006). A role for 14-3-3 in insulin-stimulated GLUT4 translocation through its interaction with the RabGAP AS160. J Biol Chem 281, 29174-29180.

Raposo, G., Cordonnier, M.N., Tenza, D., Menichi, B., Durrbach, A., Louvard, D., and Coudrier, E. (1999). Association of myosin I alpha with endosomes and lysosomes in mammalian cells. Mol Biol Cell 10, 1477-1494.

Reizes, O., Barylko, B., Li, C., Sudhof, T.C., and Albanesi, J.P. (1994). Domain structure of a mammalian myosin I beta. Proc Natl Acad Sci U S A 91, 6349-6353.

Ren, J., Wen, L., Gao, X., Jin, C., Xue, Y., and Yao, X. (2008). CSS-Palm 2.0: an updated software for palmitoylation sites prediction. Protein Eng Des Sel 21, 639-644.

Rickman, C., Hu, K., Carroll, J., and Davletov, B. (2005). Self-assembly of SNARE fusion proteins into star-shaped oligomers. Biochem J 388, 75-79.

Rimm, D.L., Koslov, E.R., Kebriaei, P., Cianci, C.D., and Morrow, J.S. (1995). Alpha 1(E)-catenin is an actin-binding and -bundling protein mediating the attachment of F-actin to the membrane adhesion complex. Proc Natl Acad Sci U S A 92, 8813-8817.

Rittinger, K., Budman, J., Xu, J., Volinia, S., Cantley, L.C., Smerdon, S.J., Gamblin, S.J., and Yaffe, M.B. (1999). Structural analysis of 14-3-3 phosphopeptide complexes identifies a dual role for the nuclear export signal of 14-3-3 in ligand binding. Mol Cell 4, 153-166.

Robzyk, K., and Kassir, Y. (1992). A simple and highly efficient procedure for rescuing autonomous plasmids from yeast. Nucleic Acids Res 20, 3790.

Rodnick, K.J., Slot, J.W., Studelska, D.R., Hanpeter, D.E., Robinson, L.J., Geuze, H.J., and James, D.E. (1992). Immunocytochemical and biochemical studies of GLUT4 in rat skeletal muscle. J Biol Chem 267, 6278-6285.

Rose, A.J., and Richter, E.A. (2005). Skeletal muscle glucose uptake during exercise: how is it regulated? Physiology (Bethesda, Md 20, 260-270.

Rosenquist, M., Alsterfjord, M., Larsson, C., and Sommarin, M. (2001). Data mining the Arabidopsis genome reveals fifteen 14-3-3 genes. Expression is demonstrated for two out of five novel genes. Plant Physiol 127, 142-149.

Rossetti, L., Stenbit, A.E., Chen, W., Hu, M., Barzilai, N., Katz, E.B., and Charron,

171 M.J. (1997). Peripheral but not hepatic insulin resistance in mice with one disrupted allele of the glucose transporter type 4 (GLUT4) gene. J Clin Invest 100, 1831-1839.

Rothenberg, P.L., Lane, W.S., Karasik, A., Backer, J., White, M., and Kahn, C.R. (1991). Purification and partial sequence analysis of pp185, the major cellular substrate of the insulin receptor tyrosine kinase. J Biol Chem 266, 8302-8311.

Roy, S., McPherson, R.A., Apolloni, A., Yan, J., Lane, A., Clyde-Smith, J., and Hancock, J.F. (1998). 14-3-3 facilitates Ras-dependent Raf-1 activation in vitro and in vivo. Mol Cell Biol 18, 3947-3955.

Rubio, M.P., Geraghty, K.M., Wong, B.H., Wood, N.T., Campbell, D.G., Morrice, N., and Mackintosh, C. (2004). 14-3-3-affinity purification of over 200 human phosphoproteins reveals new links to regulation of cellular metabolism, proliferation and trafficking. Biochem J 379, 395-408.

Saeki, N., Tokuo, H., and Ikebe, M. (2005). BIG1 is a binding partner of myosin IXb and regulates its Rho-GTPase activating protein activity. J Biol Chem 280, 10128-10134.

Saltiel, A.R., and Pessin, J.E. (2003). Insulin signaling in microdomains of the plasma membrane. Traffic 4, 711-716.

Sano, H., Eguez, L., Teruel, M.N., Fukuda, M., Chuang, T.D., Chavez, J.A., Lienhard, G.E., and McGraw, T.E. (2007). Rab10, a target of the AS160 Rab GAP, is required for insulin-stimulated translocation of GLUT4 to the adipocyte plasma membrane. Cell Metab 5, 293-303.

Sano, H., Kane, S., Sano, E., and Lienhard, G.E. (2005). Synip phosphorylation does not regulate insulin-stimulated GLUT4 translocation. Biochem Biophys Res Commun 332, 880-884.

Sano, H., Kane, S., Sano, E., Miinea, C.P., Asara, J.M., Lane, W.S., Garner, C.W., and Lienhard, G.E. (2003). Insulin-stimulated phosphorylation of a Rab GTPase-activating protein regulates GLUT4 translocation. J Biol Chem 278, 14599-14602.

Sarbassov, D.D., Guertin, D.A., Ali, S.M., and Sabatini, D.M. (2005). Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098-1101.

Sato, O., Li, X.D., and Ikebe, M. (2007). Myosin Va becomes a low duty ratio motor in the inhibited form. J Biol Chem 282, 13228-13239.

Scheid, M.P., and Woodgett, J.R. (2001). PKB/AKT: functional insights from genetic

172 models. Nat Rev Mol Cell Biol 2, 760-768.

Semiz, S., Park, J.G., Nicoloro, S.M., Furcinitti, P., Zhang, C., Chawla, A., Leszyk, J., and Czech, M.P. (2003). Conventional kinesin KIF5B mediates insulin-stimulated GLUT4 movements on microtubules. Embo J 22, 2387-2399.

Shen, K., Teruel, M.N., Subramanian, K., and Meyer, T. (1998). CaMKIIbeta functions as an F-actin targeting module that localizes CaMKIIalpha/beta heterooligomers to dendritic spines. Neuron 21, 593-606.

Shewan, A.M., van Dam, E.M., Martin, S., Luen, T.B., Hong, W., Bryant, N.J., and James, D.E. (2003). GLUT4 recycles via a trans-Golgi network (TGN) subdomain enriched in Syntaxins 6 and 16 but not TGN38: involvement of an acidic targeting motif. Mol Biol Cell 14, 973-986.

Shi, J., and Kandror, K.V. (2005). Sortilin is essential and sufficient for the formation of Glut4 storage vesicles in 3T3-L1 adipocytes. Dev Cell 9, 99-108.

Shi, J., and Kandror, K.V. (2007). The luminal Vps10p domain of sortilin plays the predominant role in targeting to insulin-responsive Glut4-containing vesicles. The Journal of biological chemistry 282, 9008-9016.

Shmueli, O., Horn-Saban, S., Chalifa-Caspi, V., Shmoish, M., Ophir, R., Benjamin-Rodrig, H., Safran, M., Domany, E., and Lancet, D. (2003). GeneNote: whole genome expression profiles in normal human tissues. Comptes rendus biologies 326, 1067-1072.

Slot, J.W., Geuze, H.J., Gigengack, S., James, D.E., and Lienhard, G.E. (1991a). Translocation of the glucose transporter GLUT4 in cardiac myocytes of the rat. Proc Natl Acad Sci U S A 88, 7815-7819.

Slot, J.W., Geuze, H.J., Gigengack, S., Lienhard, G.E., and James, D.E. (1991b). Immuno-localization of the insulin regulatable glucose transporter in brown adipose tissue of the rat. J Cell Biol 113, 123-135.

Smith, E.F., and Lefebvre, P.A. (2000). Defining functional domains within PF16: a central apparatus component required for flagellar motility. Cell motility and the cytoskeleton 46, 157-165.

Smithers, N.P., Hodgkinson, C.P., Cuttle, M., and Sale, G.J. (2008). Insulin-triggered repositioning of munc18c on syntaxin-4 in GLUT4 signalling. Biochem J 410, 255-260.

Sollner, T., Whiteheart, S.W., Brunner, M., Erdjument-Bromage, H., Geromanos, S.,

173 Tempst, P., and Rothman, J.E. (1993). SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318-324.

Stauffer, E.A., Scarborough, J.D., Hirono, M., Miller, E.D., Shah, K., Mercer, J.A., Holt, J.R., and Gillespie, P.G. (2005). Fast adaptation in vestibular hair cells requires Myosin-1c activity. Neuron 47, 541-553.

Stoffler, H.E., and Bahler, M. (1998). The ATPase activity of Myr3, a rat myosin I, is allosterically inhibited by its own tail domain and by Ca2+ binding to its light chain calmodulin. J Biol Chem 273, 14605-14611.

Sun, X.J., Miralpeix, M., Myers, M.G., Jr., Glasheen, E.M., Backer, J.M., Kahn, C.R., and White, M.F. (1992). Expression and function of IRS-1 in insulin signal transmission. J Biol Chem 267, 22662-22672.

Sun, X.J., Rothenberg, P., Kahn, C.R., Backer, J.M., Araki, E., Wilden, P.A., Cahill, D.A., Goldstein, B.J., and White, M.F. (1991). Structure of the insulin receptor substrate IRS-1 defines a unique signal transduction protein. Nature 352, 73-77.

Takamori, S., Holt, M., Stenius, K., Lemke, E.A., Gronborg, M., Riedel, D., Urlaub, H., Schenck, S., Brugger, B., Ringler, P., Muller, S.A., Rammner, B., Grater, F., Hub, J.S., De Groot, B.L., Mieskes, G., Moriyama, Y., Klingauf, J., Grubmuller, H., Heuser, J., Wieland, F., and Jahn, R. (2006). Molecular anatomy of a trafficking organelle. Cell 127, 831-846.

Talior-Volodarsky, I., Randhawa, V.K., Zaid, H., and Klip, A. (2008). Alpha-actinin-4 is selectively required for insulin-induced GLUT4 translocation. J Biol Chem 283, 25115-25123.

Tamori, Y., Hashiramoto, M., Araki, S., Kamata, Y., Takahashi, M., Kozaki, S., and Kasuga, M. (1996). Cleavage of vesicle-associated membrane protein (VAMP)-2 and cellubrevin on GLUT4-containing vesicles inhibits the translocation of GLUT4 in 3T3-L1 adipocytes. Biochem Biophys Res Commun 220, 740-745.

Tang, N., Lin, T., and Ostap, E.M. (2002). Dynamics of myo1c (myosin-ibeta ) lipid binding and dissociation. J Biol Chem 277, 42763-42768.

Taylor, S.I. (1999). Deconstructing type 2 diabetes. Cell 97, 9-12.

TerBush, D.R., Maurice, T., Roth, D., and Novick, P. (1996). The Exocyst is a multiprotein complex required for exocytosis in Saccharomyces cerevisiae. Embo J 15, 6483-6494.

Thorson, J.A., Yu, L.W., Hsu, A.L., Shih, N.Y., Graves, P.R., Tanner, J.W., Allen, P.M.,

174 Piwnica-Worms, H., and Shaw, A.S. (1998). 14-3-3 proteins are required for maintenance of Raf-1 phosphorylation and kinase activity. Mol Cell Biol 18, 5229-5238.

Thurmond, D.C., Ceresa, B.P., Okada, S., Elmendorf, J.S., Coker, K., and Pessin, J.E. (1998). Regulation of insulin-stimulated GLUT4 translocation by Munc18c in 3T3L1 adipocytes. J Biol Chem 273, 33876-33883.

Tokumitsu, H., Chijiwa, T., Hagiwara, M., Mizutani, A., Terasawa, M., and Hidaka, H. (1990). KN-62, 1-[N,O-bis(5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazi ne, a specific inhibitor of Ca2+/calmodulin-dependent protein kinase II. J Biol Chem 265, 4315-4320.

Tsao, T.S., Stenbit, A.E., Factor, S.M., Chen, W., Rossetti, L., and Charron, M.J. (1999). Prevention of insulin resistance and diabetes in mice heterozygous for GLUT4 ablation by transgenic complementation of GLUT4 in skeletal muscle. Diabetes 48, 775-782.

Tsui, J., Inagaki, M., and Schulman, H. (2005). Calcium/calmodulin-dependent protein kinase II (CaMKII) localization acts in concert with substrate targeting to create spatial restriction for phosphorylation. J Biol Chem 280, 9210-9216.

Vasta, V., Bruni, P., Vannini, F., and Farnararo, M. (1989). Fructose 2,6-bisphosphate and insulin stimulation of glycolysis in 3T3-L1 adipocytes. Int J Biochem 21, 1359-1363.

Vega, I.E., and Hsu, S.C. (2001). The exocyst complex associates with microtubules to mediate vesicle targeting and neurite outgrowth. J Neurosci 21, 3839-3848.

Vercoutter-Edouart, A., Lemoine, J., Smart, C.E., Nurcombe, V., Boilly, B., Peyrat, J., and Hondermarck, H. (2000). The mitogenic signaling pathway for fibroblast growth factor-2 involves the tyrosine phosphorylation of cyclin D2 in MCF-7 human breast cancer cells. FEBS Lett 478, 209-215.

Vest, R.S., Davies, K.D., O'Leary, H., Port, J.D., and Bayer, K.U. (2007). Dual mechanism of a natural CaMKII inhibitor. Mol Biol Cell 18, 5024-5033.

Virkamaki, A., Ueki, K., and Kahn, C.R. (1999). Protein-protein interaction in insulin signaling and the molecular mechanisms of insulin resistance. J Clin Invest 103, 931-943.

Wagner, M.C., Barylko, B., and Albanesi, J.P. (1992). Tissue distribution and

175 subcellular localization of mammalian myosin I. J Cell Biol 119, 163-170.

Wagner, M.C., Blazer-Yost, B.L., Boyd-White, J., Srirangam, A., Pennington, J., and Bennett, S. (2005). Expression of the unconventional myosin Myo1c alters sodium transport in M1 collecting duct cells. Am J Physiol Cell Physiol 289, C120-129.

Wang, F.S., Liu, C.W., Diefenbach, T.J., and Jay, D.G. (2003). Modeling the role of myosin 1c in neuronal growth cone turning. Biophys J 85, 3319-3328.

Wang, W., and Shakes, D.C. (1996). Molecular evolution of the 14-3-3 protein family. J Mol Evol 43, 384-398.

Wang, Y.X., Catlett, N.L., and Weisman, L.S. (1998). Vac8p, a vacuolar protein with armadillo repeats, functions in both vacuole inheritance and protein targeting from the cytoplasm to vacuole. J Cell Biol 140, 1063-1074.

Wang, Y.X., Kauffman, E.J., Duex, J.E., and Weisman, L.S. (2001). Fusion of docked membranes requires the armadillo repeat protein Vac8p. J Biol Chem 276, 35133-35140.

Watanabe, M., Nomura, K., Ohyama, A., Ishikawa, R., Komiya, Y., Hosaka, K., Yamauchi, E., Taniguchi, H., Sasakawa, N., Kumakura, K., Ushiki, T., Sato, O., Ikebe, M., and Igarashi, M. (2005). Myosin-Va Regulates Exocytosis through the Submicromolar Ca2+-dependent Binding of Syntaxin-1A. Mol Biol Cell 16, 4519-4530.

Waterman, M.J., Stavridi, E.S., Waterman, J.L., and Halazonetis, T.D. (1998). ATM-dependent activation of p53 involves dephosphorylation and association with 14-3-3 proteins. Nat Genet 19, 175-178.

Watson, R.T., Kanzaki, M., and Pessin, J.E. (2004a). Regulated membrane trafficking of the insulin-responsive glucose transporter 4 in adipocytes. Endocr Rev 25, 177-204.

Watson, R.T., Khan, A.H., Furukawa, M., Hou, J.C., Li, L., Kanzaki, M., Okada, S., Kandror, K.V., and Pessin, J.E. (2004b). Entry of newly synthesized GLUT4 into the insulin-responsive storage compartment is GGA dependent. Embo J 23, 2059-2070.

White, M.F. (1998). The IRS-signalling system: a network of docking proteins that mediate insulin action. Mol Cell Biochem 182, 3-11.

White, M.F. (2003). Insulin signaling in health and disease. Science 302, 1710-1711.

Whitehead, J.P., Molero, J.C., Clark, S., Martin, S., Meneilly, G., and James, D.E.

176 (2001). The role of Ca2+ in insulin-stimulated glucose transport in 3T3-L1 cells. J Biol Chem 276, 27816-27824.

Wilker, E., and Yaffe, M.B. (2004). 14-3-3 Proteins--a focus on cancer and human disease. J Mol Cell Cardiol 37, 633-642.

Williams, R., and Coluccio, L.M. (1995). Phosphorylation of myosin-I from rat liver by protein kinase C reduces calmodulin binding. Biochem Biophys Res Commun 216, 90-102.

Wolenski, J.S., Hayden, S.M., Forscher, P., and Mooseker, M.S. (1993). Calcium-calmodulin and regulation of brush border myosin-I MgATPase and mechanochemistry. J Cell Biol 122, 613-621.

Worrall, D.S., and Olefsky, J.M. (2002). The effects of intracellular calcium depletion on insulin signaling in 3T3-L1 adipocytes. Mol Endocrinol 16, 378-389.

Wright, D.C., Fick, C.A., Olesen, J.B., Lim, K., Barnes, B.R., and Craig, B.W. (2004). A role for calcium/calmodulin kinase in insulin stimulated glucose transport. Life sciences 74, 815-825.

Wu, X.S., Rao, K., Zhang, H., Wang, F., Sellers, J.R., Matesic, L.E., Copeland, N.G., Jenkins, N.A., and Hammer, J.A., 3rd (2002). Identification of an organelle receptor for myosin-Va. Nat Cell Biol 4, 271-278.

Xiang, X., Yuan, M., Song, Y., Ruderman, N., Wen, R., and Luo, Z. (2002). 14-3-3 facilitates insulin-stimulated intracellular trafficking of insulin receptor substrate 1. Mol Endocrinol 16, 552-562.

Xu, Z., and Kandror, K.V. (2002). Translocation of small preformed vesicles is responsible for the insulin activation of glucose transport in adipose cells. Evidence from the in vitro reconstitution assay. J Biol Chem 277, 47972-47975.

Yaffe, M.B., and Elia, A.E. (2001). Phosphoserine/threonine-binding domains. Curr Opin Cell Biol 13, 131-138.

Yaffe, M.B., Rittinger, K., Volinia, S., Caron, P.R., Aitken, A., Leffers, H., Gamblin, S.J., Smerdon, S.J., and Cantley, L.C. (1997). The structural basis for 14-3-3:phosphopeptide binding specificity. Cell 91, 961-971.

Yamada, E., Okada, S., Saito, T., Ohshima, K., Sato, M., Tsuchiya, T., Uehara, Y., Shimizu, H., and Mori, M. (2005). Akt2 phosphorylates Synip to regulate docking and fusion of GLUT4-containing vesicles. J Cell Biol 168, 921-928.

177 Yeh, T.Y., Sbodio, J.I., Tsun, Z.Y., Luo, B., and Chi, N.W. (2007). Insulin-stimulated exocytosis of GLUT4 is enhanced by IRAP and its partner tankyrase. Biochem J 402, 279-290.

Yoshizaki, T., Imamura, T., Babendure, J.L., Lu, J.C., Sonoda, N., and Olefsky, J.M. (2007). Myosin 5a is an insulin-stimulated Akt2 (protein kinase Bbeta) substrate modulating GLUT4 vesicle translocation. Mol Cell Biol 27, 5172-5183.

Yount, R.G., Lawson, D., and Rayment, I. (1995). Is myosin a "back door" enzyme? Biophys J 68, 44S-47S; discussion 47S-49S.

Yuan, Z., Agarwal-Mawal, A., and Paudel, H.K. (2004). 14-3-3 binds to and mediates phosphorylation of microtubule-associated tau protein by Ser9-phosphorylated glycogen synthase kinase 3beta in the brain. J Biol Chem 279, 26105-26114.

Zaid, H., Antonescu, C.N., Randhawa, V.K., and Klip, A. (2008). Insulin action on glucose transporters through molecular switches, tracks and tethers. Biochem J 413, 201-215.

Zeigerer, A., McBrayer, M.K., and McGraw, T.E. (2004). Insulin stimulation of GLUT4 exocytosis, but not its inhibition of endocytosis, is dependent on RabGAP AS160. Mol Biol Cell 15, 4406-4415.

Zhai, J., Lin, H., Shamim, M., Schlaepfer, W.W., and Canete-Soler, R. (2001). Identification of a novel interaction of 14-3-3 with p190RhoGEF. J Biol Chem 276, 41318-41324.

Zhang, H., Berg, J.S., Li, Z., Wang, Y., Lang, P., Sousa, A.D., Bhaskar, A., Cheney, R.E., and Stromblad, S. (2004). Myosin-X provides a motor-based link between integrins and the cytoskeleton. Nat Cell Biol 6, 523-531.

Zhu, T., Beckingham, K., and Ikebe, M. (1998). High affinity Ca2+ binding sites of calmodulin are critical for the regulation of myosin Ibeta motor function. J Biol Chem 273, 20481-20486.

Zhu, T., Sata, M., and Ikebe, M. (1996). Functional expression of mammalian myosin I beta: analysis of its motor activity. Biochemistry 35, 513-522.

Zick, Y. (2001). Insulin resistance: a phosphorylation-based uncoupling of insulin signaling. Trends Cell Biol 11, 437-441.

Zimmet, P., Alberti, K.G., and Shaw, J. (2001). Global and societal implications of the diabetes epidemic. Nature 414, 782-787.

178