Characterization of TPX2 Phosphorylation at Threonine 72 in Human Cancer Cells

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2014-04-30 Characterization of TPX2 phosphorylation at Threonine 72 in Human cancer cells

Shim, Su Yeon

Shim, S. Y. (2014). Characterization of TPX2 phosphorylation at Threonine 72 in Human cancer cells (Unpublished master's thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/25295 http://hdl.handle.net/11023/1459 master thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

Characterization of TPX2 phosphorylation at Threonine 72 in Human cancer cells

Su Yeon Shim

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE

DEPARTMENT OF BIOCHEMISTRY AND MOLECULAR BIOLOGY

CALGARY, ALBERTA

April, 2014

The Targeting Protein for Xenopus kinesin-like protein 2 (TPX2) is a microtubule- associated protein required for mitotic spindle assembly. In interphase cells, TPX2 regulates the amplification of the -H2AX signal in the nucleus during the DNA damage response. TPX2 functions are likely to be regulated by the putative phosphorylation sites identified by several mass spectrometry-based screenings. However, most of these 41 sites have not been validated and their roles have not been characterized. In my thesis, I characterized the phosphorylation of

TPX2 at Threonine 72 (Thr72) in vivo. First, I confirmed the specificity of our homemade phospho-Thr72 antibodies. I then found out that phosphorylation at Thr72 is cell cycle- dependent, peaking at the M phase of the cell cycle. I also confirmed the existence of Thr72 phosphorylation in mitotic HeLa cells by mass spectrometry. Using a pharmacological approach,

I discovered that cyclin-dependent kinases can mediate phosphorylation at this site in mitotic

HeLa cells. Finally, using TPX2 siRNAs and phospho-mutants, I examined the role of Thr72 in mitotic spindle assembly and DNA damage response. Understanding the significance of phosphorylation of Thr72 may provide new insights into the roles of TPX2 in healthy and diseases conditions, such as cancers.

ii ACKNOWLEDGEMENTS

I would like to express my thanks and gratitude to a number of people. First, I would like to thank my supervisor, Dr. Minh Dang Nguyen for his continuous support, training and patience during my graduate studies in his laboratory. Next, I would like to thank my committee members,

Dr. Susan Lees-Miller, Dr. Karl Riabowol and Dr. Randal Johnston for their invaluable comments, continuous guidance and advice. Also, my special thanks goes to Dr. Donna Senger for accepting the invitation to be an internal/external examiner and for helpful comments and suggestions for my thesis. I also would like to thank Dr. Julie Deans for her advice on my graduate studies and life, the members of the Nguyen laboratory and my two Korean friends, Dr.

Bo Young Ahn and Dr. Young Hee Ahn for their friendship, encouragement and helpful advice on science and life. I am also grateful to Dr. Jennifer Rahn for giving me valuable comments on my thesis.

I really want to thank my family, particularly my husband, Young ho Danny Lee for his love and support. Thank you to my parents for their love and trust in me. Last but not least, I would like to thank my two lovely daughters, Haeun Esther and Seo-eun Serena for their true love and their understanding. Girls, I really want you to know I love you so much and I promise that I will continue my efforts to be a mom you can be proud of.

iii TABLE OF CONTENTS

ABSTRACT ...... II

ACKNOWLEDGEMENTS ...... III

TABLE OF CONTENTS ...... IV

LIST OF TABLES ...... IX

LIST OF FIGURES AND ILLUSTRATIONS...... X

LIST OF SYMBOLS, ABBREVIATIONS AND NOMENCLATURE ...... XIII

CHAPTER ONE: INTRODUCTION ...... 1

1.1 Overview of the Targeting Protein for Xenopus like-protein 2 (TPX2) ...... 1

1.1.1 The discovery of TPX2 ...... 1

1.1.2 TPX2 protein sequence and function are conserved across several species .....2

1.1.3 Known protein domains of TPX2 ...... 5

1.2 TPX2 during cell cycle ...... 7

1.2.1 Localization and expression of TPX2 are strictly regulated during cell cycle..7

1.2.2 Different localization of TPX2 in the cells during cell cycle is regulated by the

small GTPase, Ras-related nuclear protein (Ran) and the importin / complex .....9

1.2.3 The distinct functions of TPX2 during mitosis and interphase ...... 12

1.2.3.1 The functions of TPX2 during mitosis ...... 14

1.2.3.2 The functions of TPX2 during interphase ...... 23

1.3 TPX2 in post-mitotic cells ...... 26

iv 1.4 TPX2 is a phospho-protein ...... 27

1.4.1 The functions of TPX2 are likely regulated by phosphorylation ...... 27

1.4.2 Some putative kinases important for the functions of TPX2 ...... 28

1.4.3 Mass spectrometry based-phosphoproteome screenings to identify in vivo

phosphorylation sites of mitotic phosphoproteins ...... 33

1.4.4 TPX2 phosphorylation sites already characterized in the literature ...... 35

1.5 TPX2 overexpression in cancers ...... 36

1.6 Rationale and hypothesis ...... 37

1.6.1 Rationale for Studying phosphorylation of Threonine 72 on TPX2 ...... 37

CHAPTER TWO: MATERIALS AND METHODS ...... 43

2.1 Cell culture ...... 43

2.2 Generation of GFP-TPX2 wild-type and mutant constructs ...... 43

2.3 Generation of phospho-specific antibodies ...... 44

2.4 RNA interference (RNAi) ...... 44

2.5 Transfection of cells with plasmid constructs and siRNAs ...... 45

2.6 Cell cycle synchronization ...... 45

2.7 Protein extraction and western blotting ...... 46

2.8 Antibodies ...... 47

2.9 Flow cytometry analysis ...... 49

2.10 Mass spectrometry analysis to identify in vivo phosphorylation sites of TPX2 ....50

2.11 Quantification of the cells with different numbers of mitotic spindles ...... 50

2.12 Treatment with Cdk inhibitors ...... 51

2.13 Immunoprecipitations (IP) ...... 52

v 2.14 Lambda protein phosphatase treatment on Immunoprecipitation samples ...... 52

2.15 Chromatin fractionation for determining the levels of -H2AX protein ...... 53

2.16 Immunofluorescent staining and confocal microscopy analysis ...... 53

2.17 Observation of ionizing radiation-induced foci formation and -PPase treatment

of U2OS cells on coverslips ...... 54

2.18 In vitro Cdk kinase assay ...... 54

2.19 Blocking peptide experiments ...... 55

2.20 Statistical analysis ...... 56

CHPTER THREE: CHARACTERIZATION OF TPX2 PHOSPHORYLATION AT

THREONINE 72 IN HUMAN CANCER CELLS ...... 57

3.1 Rationale ...... 57

3.2 Results ...... 58

3.2.1 Characterization of the phospho-specific antibodies against Thr72 of TPX2

in vivo ...... 58

3.2.2 In vivo TPX2 phosphorylation at Thr72 is cell cycle-dependent...... 66

3.2.3 In vivo mass spectrometry analysis confirmed that TPX2 is phosphorylated

at Thr72 in mitotic HeLa cells ...... 69

3.2.4 Intracellular localization and expression of the phosphorylated Thr72-TPX2 in

HeLa cells ...... 74

3.2.5 TPX2 is phosphorylated in vivo at Thr72 by Cdks ...... 77

3.3 Discussion ...... 84

3.3.1 Summary ...... 84

vi 3.3.2 Critiques and interpretations of results...... 85

CHAPTER 4: DETERMINATION OF THE ROLE OF THREONINE 72

PHOSPHORYLATION ON TPX2 MITOTIC FUNCTION ...... 88

4.1 Rationale ...... 88

4.2 Results ...... 89

4.2.1 Effects of GFP-TPX2 WT and T72A mutant on mitotic spindle assembly ....89

4.3 Discussion ...... 98

4.3.1 Summary ...... 98

4.3.2 Critical interpretations in each experiment ...... 98

4.3.3 Caveats and alternate suggestions for the systems used ...... 100

CHAPTER FIVE: DETERMINATION OF THE ROLES OF TPX2 PHOSPHORYLATION

AT THREONINE 72 IN DNA DAMAGE RESPONSE ...... 102

5.1 Rationale ...... 103

5.2 Results ...... 106

5.2.1 Analyzing the formation of IR-induced foci positive for pThr-72 upon IR

treatment ...... 106

5.2.2 Determination of the role of TPX2 phosphorylation at Thr72 in the amplification

of -H2AX signals upon exposure to ionizing radiation...... 112

5.3 Discussion ...... 116

5.3.1 Summary ...... 116

5.3.2 Alternative interpretations for each experiment...... 116

vii CHAPTER SIX: GENERAL CONCLUSION AND FUTURE DIRECTIONS ...... 119

6.1 General Discussion and Future directions ...... 119

6.1.1 Considering other potential phosphorylation sites of TPX2 ...... 119

6.1.2 TPX2 and cancers ...... 121

6.1.3 Further characterization of Thr72 TPX2 phosphorylation in mitotic spindle

assembly ...... 122

6.1.4 Further characterization of TPX2 phosphorylation at Thr72 in DNA damage

response ...... 126

6.2 Conclusion ...... 128

CHAPTER SEVEN: CRITICAL EVALUATION OF MY WORK ...... 130

REFERENCES ...... 134

APPENDIX ...... 145

viii LIST OF TABLES

Table 1.1 Human Cdks, their associated cyclins, and their functions…….………………….. 32

Table 2.1 List of antibodies…………...………………………………………………………. .52

Appendix Table 1 Matching kinase motif information for the phosphorylation sites of TPX2 identified from my own mass spectrometry data analysis……………………………………..152

Appendix Table 2 P values of ANOVA and Neuman-Keulus tests for the experiments in Figure

4.1……………………………………………………………………………………………….156

Appendix Table 3 P values of ANOVA and Neuman-Keulus tests for the experiments in Figure

4.2……………………………………………………………………………………………….157

ix LIST OF FIGURES AND ILLUSTRATIONS

Figure 1.1 The phylogram tree of TPX2 proteins for various species ...... 4

Figure 1.2 Summary of the known in vivo phosphorylation sites of TPX2 reported in

PhosphoSitePlus and the main domains of TPX2 protein ...... 6

Figure 1.3 TPX2 localization and expression during cell cycle of HeLa cells ...... 8

Figure 1.4 Intracellular compartmentalization of TPX2 during cell cycle is controlled by Ran and the importin complex………………………………………………………….……….……11

Figure 1.5 Summary of TPX2 functions during the cell cycle………………………………….13

Figure 1.6 Models of mitotic spindle assembly……………………………………………..…..15

Figure 1.7 TPX2 mediates branching MT nucleation…………………………………………...18

Figure 1.8 Multiple sequence alignment of TPX2 shows that Thr72 on TPX2 is conserved across different species……………………………………………………………………...…………..39

Figure 1.9 In vitro kinase assay showing that TPX2 is phosphorylated at Thr72 in vitro by Cdk1 and Cdk2 and the in vitro specificity of phospho-TPX2 antibodies against Thr72……………...41

Figure 3.1 The pThr-72 antibodies are specific to endogenous TPX2 in vivo…………….….....60

Figure 3.2 The pThr-72 antibodies are specific in western blots to the immunoprecipitated TPX2 phosphorylated at Thr72 in vivo……………………………..…………………………………..64

Figure 3.3 In vivo TPX2 phosphorylation at Thr72 is cell cycle-dependent………...………….67

Figure 3.4 TPX2 is phosphorylated at Thr72 in mitotic HeLa cells as detected by mass spectrometry analysis………………………………………………………………………...…..72

x Figure 3.5 Localization and expression of TPX2 phosphorylated at Thr72 in HeLa cells and the specificity of pThr-72 antibodies in immunostaining……………………………………………78

Figure 3.6 Phosphorylation of TPX2 at Thr72 is inhibited by the Cdk inhibitor roscovitine...... 79

Figure 3.7 Phosphorylation of TPX2 at Thr72 is inhibited by treatment with the Cdk1-specific inhibitor alsterpaullone…………………………………………………………………………..81

Figure 4.1 Effects of GFP-TPX2 WT and T72A mutant overexpression in HeLa cells on the numbers of mitotic spindles……………………………………………………………………..97

Figure 4.2 Effects of GFP-TPX2 WT and T72A in HeLa cells lacking endogenous TPX2 on the numbers of mitotic spindles………………………………………………………………….…100

Figure 5.1 Overexpression of GFP-TPX2 reverses the increase the levels of -H2AX triggered by the absence of endogenous TPX2 in HeLa cells in response to IR………………...……….105

Figure 5.2 Formation of IRIF positive for pThr-72 in irradiated U2OS cells and specificity of pThr-72 antibodies against pThr-72-positive IRIF signal using blocking peptides…………....108

Figure 5.3 -PPase Treatment on pThr-72-positive IRIF shows that this IRIF formation is phosphorylation-dependent…………………………………………………..…………………110

Figure 5.4 Effect of GFP-TPX2 T72A on the levels of -H2AX in response to

IR………………………………………………………………………………………………..114

Appendix Figure 1 Sequence coverage and phosphorylation sites of TPX2 identified by LC-

MS/MS analysis……………………………………………………… …..………………...….146

xi Appendix Figure 2 Multiple sequences alignment of TPX2 protein from different species and the location of phosphorylation sites of TPX2 from PhosphoSitePlus and my own mass spectrometry analysis data…………………..………………………………………………….150

Appendix Figure 3 The localization of pThr-72 protein during mitosis……………………....151

xii LIST OF SYMBOLS, ABBREVIATIONS AND NOMENCLATURE a.a.: amino acid GSK3: glycogen synthase kinase 3 Ab: antibody GST: glutathione S-transferase ANOVA: analysis of variance GTP: guanosine-5'-triphosphate APC/C: anaphase-promoting complex/ GTPase: guanosine triphosphatase cyclosome IP: immunoprecipitation ATP: adenosine triphosphate H2AX: histone H2AX ATM: ataxia telangiectasia mutated IR: ionizing radiation 53BP1: p53 binding protein 1 IRIF: ionizing radiation-induced foci BRCA1: breast cancer susceptibility gene1 LC-MS: liquid chromatography-mass Cdk: cyclin-dependent kinase spectrometry CK: casein kinase MAP: microtubule-associated protein Chk1: checkpoint kinase-1 MAPK: microtubule-associated protein kinase/ CIN: chromosomal instability mitogen-activated protein kinase DAPI: 4',6-diamidino-2-phenylindole MDC1: mediator of DNA damage checkpoint DNA: deoxyribonucleic acid MT: microtubule DNA-PK: DNA-dependent protein kinase MRE11: meiotic recombination protein 11 DDR: DNA damage response MRN: MRE11-RAD50-NBS1 complex DMEM: dulbecco's modified eagle medium NBS1: Nijmegen breakage syndrome 1 DSB: double strand break NEK6: nima-related kinase 6 ERK: extracellular signal-regulated kinase NLS: nuclear localization signal FBS: fetal bovine serum NS: not significant or non-synchronized GDP: guanosine di-phosphate PBS: phosphate buffered saline GFP: green fluorescent protein Phe: phenylalanine RanGAP: ranGTPase activating protein PI(3)K: phosphatidylinositol-3 kinase RNAi: ribonucleic acid interference PIKK: phosphatidylinositol 3-kinase-related RSK: ribosomal S6 kinase. kinase RCC: regulator of chromosome P100: protein 100 condensation 1 pThr-72: phospho-Threonine 72 RT: room temperature Ran: ras-related nuclear protein

xiii -PPase: lambda protein phosphatase PVDF: polyvinylidene difluoride SDS-PAGE: sodium dodecyl sulfate polyacrylamide gel electrophoresis S.E.M.: standard error of the mean STDEV: standard deviation TPX2: targeting protein for Xklp2 (Xenopus kinesin-like protein 2) Thr: threonine Tyr: tyrosine Val: valine WB: western blot WT: wild-type

xiv

CHAPTER ONE: INTRODUCTION

The main focus of our laboratory is to study the cytoskeleton and its associated proteins.

Microtubules (MTs), one component of the cytoskeleton, are completely reorganized during mitosis to build the mitotic spindle through the activity of many MT-associated proteins (MAPs) and motor proteins [1]. Accurate mitotic spindle assembly is crucial for the correct segregation of sister chromatids during cell division [2].

My project focused on one of these MT-associated proteins, Targeting Protein for

Xenopus kinesin-like protein 2. For the last 15 years, TPX2 has been studied as a critical player for mitotic spindle assembly. Recently, TPX2 has emerged as a new regulator of DNA damage response [3]. Even though TPX2 is a phosphoprotein with over 40 phosphorylation sites identified in various literature sources using a mass spectrometry approach [4], most of these sites have never been validated and their roles have not been characterized. In my thesis, I aimed to characterize TPX2 phosphorylation at a particular threonine residue, Thr72. I hypothesized that phosphorylation of Thr72 regulates the cell cycle-dependent functions of TPX2.

1.1 Overview of the Targeting Protein for Xenopus like-protein 2 (TPX2)

1.1.1. The discovery of TPX2

Human TPX2 was originally called P100 based on its molecular mass of ~ 100kDa. This protein was first identified by the Heidebrecht group in 1997 from the nuclear lysates of L428 cells (human cell lines derived from patients with Hodgkin's disease) [5]. They first raised a

novel mouse monoclonal antibody, named Ki-S2 by immunizing mice with nuclear lysates of

L428 cells. The 100 kDa protein recognized by Ki-S2 turned out to be a novel proliferation- associated nuclear protein. Through examination by immunohistochemistry of various types of human tissues, both healthy and cancerous, they found that this protein is preferentially enriched in the nuclei of proliferating cells during S and G2 phases. In contrast, during mitosis, it is directly associated with the mitotic spindle and spindle poles. Interestingly, they observed that expression levels of TPX2 were considerably higher in cells going through the G1/S transition to the end of cytokinesis, than in G1 phase cells; immediately after cytokinesis, TPX2 levels decrease rapidly and become almost imperceptible in G1 [10]. Based on its tightly controlled expression during cell cycle, TPX2 was advanced as a proliferation marker used for the diagnosis and prognosis of cancers.

In 1998, independently from the Heidebrecht’s group, the Wittmann group identified and characterized the functions of Xenopus TPX2. They found that TPX2 is a novel MT-associated protein (MAP) required for localizing Xenopus kinesin-like protein 2 (Xklp2) to spindle poles and to the minus ends of MTs during mitosis [6-8]. Xklp2 is a plus end-directed MT motor essential for centrosome separation and spindle bipolarity maintenance [9]. Wittmann and his colleagues named this protein TPX2, Targeting Protein for Xklp2, based on their finding of the first function of the protein [6].

1.1.2 TPX2 protein sequence and function are conserved across several species

According to NCBI (The National Center for Biotechnology Information), the sequences of the tpx2 gene and protein are well conserved in human, mouse, rat, Xenopus, lizard, chimpanzee, Rhesus monkey, dog, cow, chicken, and zebrafish [10]. The homologous protein of

TPX2 cannot be found in yeast. Even though TPX2 protein is present, certain functional domains are missing in Drosophila and Caenorhabditis elegans. Generally speaking, the functions of

TPX2 protein appear to be conserved in various organisms such as vertebrates, C. elegans,

Drosophila, ascidians and plants [11-14]. Figure 1.1 shows the phylogram tree of TPX2 proteins.

Each percentage represents the proportion of the amino acid sequences in each species that is identical to the amino acid sequences of human TPX2. Of note, sequence comparison between human and other vertebrate species shows that the human TPX2 protein sequence is relatively well conserved in mouse, rat and frog (78, 78 and 53 %, respectively). The details are shown in

Appendix Figure 2.

Figure 1.1 The phylogram tree of TPX2 proteins for various species [10]

Each percentage shown beside the name of the species represents the percentage of identical amino acid sequences in each species when aligned with human TPX2 sequence.

1.1.3 Known protein domains of TPX2

Human TPX2 contains a total of 747 amino acids (a.a.). Human TPX2 has three main functional domains that are conserved across vertebrate species (Figure 1.2). First, the N- terminus of TPX2 contains a region called the Aurora A binding domain (a.a 1-66). This domain binds two different sites of Aurora A. This binding results in autophosphorylation and activation of Aurora A kinase at Threonine 288 (Threonine 295 in Xenopus Aurora A) as well as protection of this kinase from being dephosphorylated by Protein phosphatase 6 (PP6) [15]. The N-terminal residues (a.a. 1-43) of TPX2 have shown to be sufficient for Aurora A binding and activation

[16]. Additionally, the importin binding domain in the middle of human TPX2 (a.a. 236-352) includes a short nuclear localisation signal (NLS) comprising residues 315-318 that corresponds to a.a. 284-287 in Xenopus TPX2 [17-19]. This domain binds to the nuclear transport receptor importin , which regulates the activity of TPX2 through the RanGTP gradient (see below) [19].

Finally, the Eg5 interaction domain at the C-terminus of TPX2 (a.a. 680-747) is required for targeting the motor protein Eg5 to the mitotic spindle, for regulating its activity and for the formation of a normal bipolar spindle [10, 20].

Figure 1.2 Summary of the known in vivo phosphorylation sites of TPX2 reported in

PhosphoSitePlus and the main domains of TPX2 protein [4, 10, 17-19, 21]

Human TPX2 contains a total of 747 amino acids. TPX2 has more than 40 phosphorylation sites identified from in vivo mass spectrometry-based screening [T: Threonine, S: Serine, Y: Tyrosine]. PhosphositePlus analysis was done in May 30, 2013. The blue box shows the Aurora A binding domain (amino acids 1-66), which activates Aurora A. The red box indicates the importin binding domain (a.a. 236-352) with an NLS (indicated as purple, a.a. 315-318), which confers binding to the nuclear transport receptor protein importin , allowing TPX2 transport into nucleus. The yellow box indicates the motor protein Eg5 interaction domain (a.a. 680-747), which mediates the binding of the motor protein Eg5 to mitotic spindles and regulates its activity.

1.2 TPX2 during cell cycle

1.2.1 Localization and expression of TPX2 are strictly regulated during cell cycle

The expression levels and localization of the TPX2 protein are tightly regulated during cell cycle [5, 22]. In interphase, TPX2 is located in the nucleus and continuously resides there throughout S phase until G2 phase. During mitosis, TPX2 is associated with the mitotic spindle as a MAP (Figure 1.3B). As shown in Figure 1.3A, western blot analysis from Gruss et al. shows that expression of TPX2 increases from S phase to mitosis and peaks at mitosis. At mitotic exit, the levels of TPX2 sharply decline and TPX2 is degraded by the anaphase-promoting complex/cyclosome (APC/C), an E3 ubiquitin ligase responsible for controlling mitotic progression [23]. Of note, the expression levels of TPX2 in cells at G1 phase are low, but detectable (Figure 1.3A). After 24 hours, when cells complete an entire cycle, TPX2 re- accumulates in the nucleus.

Figure 1.3 TPX2 localization and expression during cell cycle of HeLa cells (modified from Figure 3 in [22]). HeLa cells were synchronized at the boundary of G1/S phase by a double thymidine block. A) After release (0 hour as indicated time, G1/S), cells were harvested every 2 hours and TPX2 expression was analyzed by western blot, fixed at the indicated times. Western blots were probed with human TPX2 (hTPX2) and -tubulin antibodies. B) After release (at 0 hour time), cells on cover slips were fixed with 3% paraformaldehyde and stained with hTPX2 (Cy3-red) and - tubulin (FITC-Green) antibodies. DNA was stained with 4', 6-diamidino-2-phenylindole (DAPI).

1.2.2 Different localization of TPX2 in the cells during cell cycle is regulated by the small

GTPase, Ras-related nuclear protein (Ran) and the importin / complex

During interphase, TPX2 is located in the nucleus. Intracellular distribution of TPX2 is controlled by the nuclear transport receptor protein importin / complex and the small guanosine tri-phosphatase (GTPase) Ran gradient system [24]. Ran, one member of the Ras superfamily, is a GTP-binding protein with essential roles in mediating transport of macromolecules such as RNA and proteins, between the nucleus and cytoplasm throughout the cell cycle [25]. Ran regulates the ability of importins to bind and release their cargo proteins including TPX2 [25]. Cargo proteins usually contain an NLS that allows binding to the nuclear transport proteins, facilitating entry to the nucleus [26]. In cells, Ran exists in two different nucleotide bound states, RanGDP (guanosine diphosphate) and RanGTP (guanosine triphosphate) modulated by several proteins. The GTP hydrolysis (RanGTP conversion to

RanGDP) is stimulated by RanGTPase-activating protein (RanGAP) and two RanGTP binding proteins, RanBP1 and RanBP2. Conversely, efficient RanGDP exchange for RanGTP requires the nucleotide exchange factor RCC1 (Regulator of chromosome condensation 1). Due to different localization of RCC1 and RanGAP in the cells (RCC1 is bound to chromatin in the nucleus whereas RanGAP and RanBP1/2 are in the cytoplasm), the concentrations of RanGTP and RanGDP are also different between the cytoplasm and nucleus. This difference in the concentrations creates concentration gradients of RanGTP and RanGDP [27, 28] (RanGTP is in the nucleus, while RanGDP is in the cytoplasm). For nuclear import, RanGTP binding to importin α and β leads the importins to release their cargo proteins into the nucleus.

To enter the nucleus, cytoplasmic TPX2 binds to the importin / complex via its NLS

(around a.a. 315 in human and a.a. 284 in Xenopus) and gets transported into the nucleus from the cytoplasm [19]. In the nucleus, there is the high concentration of RanGTP mediated by RCC1.

RanGTP binds to importin β dissociating it from importin α. Then importin α has reduced affinity for NLS sequences in TPX2, thereby releasing TPX2 (Figure 1. 4). During mitosis,

TPX2 is localized on the mitotic spindle. High concentrations of RCC1 are found around chromosomes, which generate a RanGTP gradient in the vicinity of chromosome [29]. RanGTP associates with the importin α /β complex, releasing TPX2 in the same way as the way as during interphase. The released TPX2 is free to mediate MT nucleation and chromosome-induced spindle assembly (Figure 1.4). Therefore, Ran and the importin complex regulate intracellular compartmentalization of TPX2 during the cell cycle.

Figure 1.4 Intracellular compartmentalization of TPX2 during cell cycle is controlled by Ran and the importin complex (Modified and redrawn from [24]). During interphase, TPX2 is associated with the nuclear transport receptor proteins importin α/β complex in the cytoplasm via its NLS and it gets transported into the nucleus by importin α and β. RanGTP produced by RCC1 in the nucleus binds to importin β dissociating it from importin α. Then, importin α has reduced affinity for NLS sequences in TPX2, and releases TPX2. TPX2 is free for mediating MT nucleation and activating the Aurora A protein kinase through direct binding.

1.2.3 The distinct functions of TPX2 during mitosis and interphase

Due to the different localizations of TPX2 during cell cycle, the functions of TPX2 are also cell-cycle dependent. During mitosis, TPX2 is localized on the mitotic spindles where it acts as a critical factor in chromosome-induced spindle assembly. Conversely, during interphase, which lasts up to 23 hours (as in the case of HeLa cells), TPX2 gets transported into the nucleus and resides in the nucleus throughout interphase. The function of TPX2 in the nucleus had remained obscure for quite a long time. The recent publication from our laboratory shows that in the nucleus, during interphase, nuclear TPX2 regulates the amplification of the -H2AX signal in response to DNA damage. This is the first demonstrated function of nuclear TPX2 [3] (See

Figure 1.5 for the summary of TPX2 functions during cell cycle).

Figure 1.5 Summary of TPX2 functions during the cell cycle [3, 19, 20, 30]

TPX2 plays distinct roles during mitosis and interphase. During mitosis, TPX2 is associated with mitotic spindles and spindle poles and plays diverse roles during mitotic spindle assembly as a spindle assembly factor. During interphase, and particularly in G1 phase, TPX2 regulates the levels of -H2AX in response to ionizing radiation.

1.2.3.1 The functions of TPX2 during mitosis

During mitosis, TPX2 is associated with the mitotic spindle and spindle poles. Mitotic

TPX2 plays multiple roles in MT assembly and spindle formation (Figure 1.5). TPX2 is required for 1) MT nucleation in the vicinity of chromosomes and MT bundling, 2) branching MT nucleation, 3) activation of the mitotic kinase Aurora A, 4) regulation of the localization and activity of a mitotic motor protein, Eg5 [18-20, 22, 31-33].

1.2.3.1.1 TPX2 is important for chromosome-induced spindle assembly

Currently, there are two models of mitotic spindle assembly that are not mutually exclusive. The first model is centrosome-mediated spindle assembly, in which centrosomes act as microtubule organizing centers (MTOC) and nucleate spindle MTs (Figure 1.6A) [34]. The other model is chromosome-mediated spindle assembly, also called the self-assembly model, in which MTs are nucleated around chromosomes, then, are organized by motor proteins into a bipolar array (Figure 1.6B) [35, 36]. The current view is that a combination of the two models results in spindle formation in cells containing centrosomes (Figure 1.6C) [34] and only chromosome-induced spindle assembly forms mitotic spindles in cells that lack centrosomes [37,

38]. Within these models, TPX2 plays several essential roles in chromosome-mediated spindle assembly [22]:

1) TPX2 functions in RanGTP-induced MT nucleation and MT bundling

MT nucleation is the generation of new MTs around chromosomes, which is an initial step in spindle assembly [29]. After nucleation, MT bundling is required to form mitotic spindles

[19]. In the Xenopus system, when TPX2 is depleted from egg extracts, MT nucleation and the formation of spindles around chromatin beads are prevented [29].

Figure 1.6 Models of mitotic spindle assembly (redrawn from [34]) A) Centrosome-mediated spindle assembly. Also called the “Search and capture” model. Centrosomes play roles as microtubule organizing centers (MTOC): They nucleate spindle MTs and search for chromosomes and kinetochores. They contact chromosomes and kinetochores “by chance”, then form mitotic spindles

B) Chromosome-mediated spindle assembly. Also, known as the “self-organization” model. MTs are nucleated around chromosomes in the absence of centrosomes and are organized into bipolar spindles by motor proteins.

C) Combined model. Both MTs nucleating from the centrosome and around chromosomes are incorporated into a bipolar array to generate mitotic spindles

Conversely, when recombinant TPX2 is added to the TPX2-depleted extracts, the ability of chromatin beads to nucleate MTs is restored [22]. Also, when human TPX2 recombinant protein is added into a pure tubulin solution, in vitro MT polymerization (de novo MT formation) is initiated and MT bundling is induced [19]. Interestingly, this nucleation activity of TPX2 is regulated by importin  and . However, MT bundling activity of TPX2 is not inhibited by importins. Thus, TPX2-induced MT bundling is independent of importins [19]. Consistent with these results, Tulu et al. found that TPX2 knockdown in HeLa cells by siRNA treatment prevents

MT nucleation around chromosomes [39]. Gruss et al. revealed that TPX2 depletion in HeLa cells significantly decreases chromatin-mediated MTs without affecting centrosome-mediated

MT nucleation [22]. Consequently, TPX2 has been shown to be critical for spindle formation both in cultured cells and Xenopus egg extracts [22, 40]. Furthermore, primary cell cultures from a conditional TPX2 knockout mouse have defective MT nucleation around chromosomes, leading to aberrant spindle formation and chromosome segregation defects [41]. In sum, TPX2 plays essential roles in MT nucleation around chromosomes and MT bundling.

2) TPX2 mediates branching MT nucleation

In addition to the roles of TPX2 in MT nucleation and bundling, Petry et al. found that

TPX2 is also required for MT nucleation from existing MTs. This process is called, branching

MT nucleation [33]. Due to the very high density of MTs, it has been quite hard to observe this branching MT nucleation. However, Petry et al. visualized the growth and formation of MT in

Xenopus egg extracts on a glass surface using internal reflection (TIRF) microscopy in vitro [33].

In this in vitro assay, they observed that MT nucleation also occurs from the sides of existing MTs. Daughter MTs newly nucleated from mother MTs grow at low angle and therefore

grow out in the same direction as mother MTs (from a minus end to a positive end) (Figure 1.7).

Augmin, a protein complex required for microtubule nucleation, has been implicated to be important for this process [33, 42, 43]. Petry et al. confirmed that when Augmin is depleted, branching MT nucleation does not occur in Xenopus egg extracts. RanGTP and TPX2, also promote branching MT nucleation. When TPX2 is immunodepleted from the egg extracts, individual MT growth still occurs, but, branching MT nucleation is inhibited. Branching MT nucleation is the main mechanism for amplification of MT numbers and preservation of MT polarity. In sum, in concert with Augmin, TPX2 plays an essential role in branching MT nucleation, thereby allowing MT mass amplification.

Figure 1.7 TPX2 mediates branching MT nucleation (Redrawn and modified from [33, 44])

TPX2 together with Augmin and -tubulin promotes branching MT nucleation. Newly nucleated

“daughter MTs” grow from the existing “mother MTs” at a low angel and still keep the same polarity (from a minus end to a plus end).

3) TPX2 is required for activation of Aurora A

Another important role of TPX2 during mitosis is to activate the mitotic kinase Aurora A.

Aurora A is one of the three members of the serine/threonine Aurora kinase family essential for mitotic spindle assembly and cell cycle progression. Aurora A is important for separation and maturation of centrosomes [45]. It is also required to ensure the proper formation of bipolar spindles for correct chromosome segregation [45]. Aurora A is overexpressed in several human cancers and deregulation of this kinase results in mitotic defects and aneuploidy (abnormal number of chromosomes) [46-50]. Aurora A contains an N-terminal regulatory domain and a C- terminal catalytic domain. Phosphorylation of Threonine 287/288 in the kinase activation loop is essential and leads to significant increase in its kinase activity [51].

TPX2 was identified as a major interacting protein of Aurora A in mitotic HeLa cells [16].

Furthermore, TPX2 binding to Aurora A is shown to increase the autophosphorylation activity of

Aurora A, leading to strong activation of this kinase [16]. Pull-down experiments in HeLa cell extracts using the first 1-43 amino acids from the N-terminus fused to GST protein show that amino acids 1-43 can interact with the C-terminus of Aurora A, thereby activating and protecting

Aurora A from dephosphorylation [15, 16, 31]. More precisely, once TPX2 is released by

RanGTP from importin / complex, TPX2 is free to bind Aurora A kinase. TPX2 binding results in the conformational change of Aurora A. This in turn activates the kinase by promoting autophosphorylation at Thr288 (for human) or Thr295 (for Xenopus). Binding of TPX2 to

Aurora A also sterically inhibits the threonine residue from dephosphorylation by Protein phosphatase 6 (PP6) [15]. Additionally, TPX2 is responsible for targeting Aurora A to spindle

MT in human and Caenorhabditis elegans [12, 31]. In mitotic cells, TPX2 co-localizes with

Aurora A along the spindles. However, when TPX2 is depleted from cells using TPX2 siRNA,

Aurora A is no longer found along the mitotic spindle [31]. This TPX2-Aurora A interaction is required for MT nucleation near chromosomes and for spindle length establishment [52]. In C. elegans, when the interaction of TPX2 with Aurora A is abolished by changing a few amino acids in the N-terminus of TPX2 (Phe15/18 and Val23/Tyr26), short bipolar spindles are formed

[12]. Consistently, mutants with point-mutation blocking TPX2/Aurora A interaction in human cells (U2OS cells), form very short, bipolar spindles [52]. These cells are still able to segregate chromosomes. Short spindles are due to failed separation of spindle poles after nuclear envelope breakdown. Interestingly, in these mutants, MT nucleation from chromosomes is abolished during spindle assembly [52]. Taken together, this result suggests that TPX2 mediated-MT nucleation around chromosomes is important for the length of spindles as they form [52].

Interestingly, human TPX2 and Xenopus TPX2 are substrates of Aurora A [31, 32]. In vitro kinase assays using recombinant Aurora A and human TPX2 show that TPX2 is phosphorylated by wild-type Aurora A, but not the kinase-dead mutant form of Aurora A [31].

Eyers et al. found that the N-terminus of Xenopus TPX2 is phosphorylated by Aurora A and they identified three target sites (Ser48, Ser90, or Ser94). However, mutation of these sites does not block Aurora A activation [32]. Instead, when three other putative phosphorylation sites in the

N-terminus of Xenopus TPX2 (Tyr8, Tyr10 and Asp11) that closely interact with the C-terminal catalytic domain of Aurora A in structural studies were mutated to Alanine, they found that theses mutations abolish Aurora A activation. This result suggests that TPX2-mediated activation of Aurora A is regulated by phosphorylation of TPX2 by Aurora A. In summary, the N-terminus of TPX2 binds to Aurora A kinase, activates the kinase and targets it to the mitotic spindle. And

TPX2 is phosphorylated by Aurora A. This reciprocal regulation between TPX2 and Aurora A is

important for chromosome-associated MT nucleation and for establishing the correct length of spindles during spindle assembly.

4) TPX2 regulates the localization and activity of the mitotic motor protein Eg5

Eg5 is a plus-end directed motor protein that belongs to the Kinesin-5 subclass. This protein affects mitotic spindle organization and spindle assembly by MT cross-linking, sliding along MTs and generating outward forces for spindle pole separation at mitotic entry [20, 21].

Eg5 is located at spindle poles and on spindles during mitosis. Eg5 directly binds to and interacts with the last 35 amino acids of the C-terminus of TPX2 (CT-35) [21]. Furthermore, Ma et al. demonstrated that the interaction between TPX2 and Eg5 is essential for the organization and stability of spindle MTs and also is required for Eg5 targeting to mitotic spindles [20]. To examine the exact role of the TPX2-Eg5 interaction, they generated a stable cell line (LLC-Pk1 pig cells) expressing TPX2 lacking the TPX2-Eg5 interacting domain (CT-35). In these cells, morphology of the spindles was abnormal. Cells lacking the TPX2-Eg5 interaction domain have highly disorganized spindles and spindle pole fragmentation (multipolar spindles). These results indicate that the TPX2-Eg5 interaction is important for the organization of normal spindle MTs, as evidenced by the greatly decreased localization of TPX2 on spindles in cells overexpressing

CT-35 [20].

Additionally, TPX2 regulates the motor activity of Eg5. In Xenopus cells, overexpression of the C-terminus of TPX2 (containing the Eg5 interaction domain) blocks centrosome separation, whereas, addition of excess Eg5 protein re-establishes the wild-type phenotype. These results indicate that the C-terminus of TPX2 inhibits the activity of Eg5 [21].

Furthermore, microinjection of C-35 protein into LLC-Pk1 cells causes spindle elongation. This

indicates that the TPX2-Eg5 interaction regulates Eg5 motor activity [20]. Finally, in vitro measurement assay with purified Eg5 and TPX2 protein showed that TPX2 inhibits or abolishes

Eg5 dependent-MT gliding and sliding activity. Thus, through interaction with Eg5, TPX2 reduces the motor activity of Eg5. Consequently, TPX2 is important for balancing the forces regulating spindle length. In summary, TPX2 is required for Eg5 localization to mitotic spindles and for regulation of the motor activity of Eg5 critical for spindle length determination [20].

1.2.3.1.2 TPX2 knockout mice and haploinsufficiency

The recent study of TPX2 knockout mice analysis performed by Aguirre-Portolés al. showed that TPX2 gene knockout leads to defects in early embryos [48]. TPX2-/- homozygous mutants are embryonic-lethal and cannot be detected between E (embryonic day) 8.5 and E17.5.

When cultured at E1.5, TPX2 null embryos are arrested at the morula stage with larger nuclei, indicative of cell division failure. TPX2 null cells displayed spindle aberrancies such as collapsed or monopolar spindles. In contrast, wild-type and heterozygous embryos generated normal blastocysts that are able to form normal biopolar spindles. In addition, some of the TPX2 null cells are tetraploid in their early morula stage. This implies that they are defective in chromosome segregation. Taken together, these results suggest that TPX2 is essential for normal bipolar spindle formation and chromosome segregation during the early developmental stages of mouse embryos [48].

In agreement with the embryonic lethality of TPX2 knockout mice, mouse embryonic fibroblasts (MEFs) with conditional TPX2 knockout display abnormal chromosome segregation and aberrant ploidy caused by failure of MT nucleation from DNA. They also exhibit abnormal nuclear morphology such as micronuclei or multiple nuclei.

Furthermore, Aguirre-Portolés et al. showed that the TPX2 +/- genotype mice are haploinsufficient, and this genotype leads to accumulation of cells with aneuploidy in vivo. The tumor-free lifespan of TPX2+/- mice is shorter than that of TPX2 +/+ mice. Also, TPX2 +/- mice are more susceptible to spontaneous lymphomas and lung tumors. They observed that these

TPX2+/- tumors were highly aneuploid compared to TPX2+/- healthy tissues or TPX2+/+ control tissues. Taken together, these results suggest that TPX2 is important for the maintenance of genomic stability, and deregulation of TPX2 expression possibly favors the development of tumors in vivo [41].

1.2.3.2 The functions of TPX2 during interphase

At the beginning of interphase, TPX2 is localized to the nucleus and resides there throughout interphase until the time of nuclear envelope breakdown. However, the exact role of nuclear TPX2 had remained vague until recently. General assumption was that TPX2 is localized to the nucleus during interphase to prevent MTs from being nucleated or re-organized in the cytoplasm. Our laboratory has recently discovered that during interphase nuclear TPX2 plays an important role in the DNA damage response (Figure 1.5) [3].

1) DNA damage response (DDR)

Eukaryotic cells encounter DNA damage triggered by various endogenous and environmental stresses every day. To overcome this, cells have developed sophisticated molecular signaling mechanisms collectively called the DNA damage response (DDR) to detect and repair DNA damage. Among DNA lesions caused by various DNA damaging agents, DNA double-strand breaks (DSBs) are considered to be the most toxic lesions in cells because a single

DSB has the potential to cause the loss of an entire chromosome arm and induce cell death [53].

Defects in the DDR pathway can result in genomic instability that can lead to cancer predisposition or neurodegenerative diseases [54, 55].

2) H2AX phosphorylation: a key event of DDR to DSBs and its role in DDR

One of the important DDR events in response to DSBs is the phosphorylation of the histone variant H2AX on Serine 139 (Ser139), which produces -H2AX. H2AX phosphorylation is required for assembling DNA damage signaling/repair proteins at the damaged sites of DNA.

It is also required for activating cell cycle checkpoints, mediating DNA repair, or getting rid of damaged cells [56]. Originally, ATM (ataxia telangiectasia mutated) has been considered a major kinase for H2AX phosphorylation upon DSB. ATM is a member of phosphatidylinositol 3- kinase-related kinases (PIKKs) that show sequence homology to phosphatidylinositol 3-kinases

(PI3Ks). However, several research papers reveal that all PIKKs such as ATM, ATR (ataxia telangiectasia and RAD3-related), DNA-PKcs (DNA-dependent protein kinase catalytic subunit) mediate the phosphorylation of H2AX in response to ionizing radiation-induced DNA damage

[57-59].

3) Ionizing radiation-induced foci (IRIF)

When DSBs are generated by ionizing radiation (IR), many DDR proteins such as

Mre11-Rad50-Nbs1 (MRN) complex, p53-binding protein 1(53BP1), Mediator of DNA damage checkpoint protein 1 (MDC1) and Breast cancer 1(BRCA1) are relocated to the break sites to form nuclear foci, referred to as ionizing radiation-induced foci (IRIF). IRIF of these proteins co- localize with -H2AX-positive IRIF and interact with -H2AX [60-62]. -H2AX-positive IRIF recruit repair proteins, causing higher concentrations of repair proteins at the sites of DSBs [63].

H2AX phosphorylation is known to play an important role in the formation of IRIF by accumulating DNA damage signaling and repair proteins to DSB sites. H2AX-knockout mice

demonstrate the phenotypes of growth retardation, radiation sensitivity and immune deficiency.

When the distribution of several factors that co-localize with γ-H2AX at IRIF is examined in

H2AX knockout (-/-) fibroblasts, the IRIF formation of Nbs1, Brca1 and 53bp1 is found to be impaired in those cells [64]. The observed phenotypes in H2AX-knockout mice are due to the failure of recruitment of these proteins to IRIF. This indicates a critical role of -H2AX in accumulating certain DNA repair complexes at sites of damaged DNA.

4) The function of TPX2 in DDR

Our laboratory recently found that nuclear TPX2 plays a role in DDR by the regulation of

H2AX phosphorylation through the gain or loss of TPX2 function experiments [3]. Depletion of

TPX2 in HeLa cells with siRNA treatment or stable expression of doxycycline-inducible TPX2 miRNA leads to significant and transient increases in levels of IR-induced γ-H2AX signals in the

G0 and G1 phases of the cell cycle. In contrast, cells overexpressing TPX2 show decreased levels of the γ-H2AX signal after IR. Interestingly, TPX2 localizes to a subset of IRIF in HeLa,

U2OS, LAN1 (human neuroblastoma cells) and post-mitotic neurons positive for γ-H2AX.

Consistent with the western blot results, upon TPX2 depletion, the percentage of cells with high intensity γ-H2AX-positive IRIF is significantly higher when compared to control cells. This phenotype was also observed in post-mitotic neurons. This result indicated that TPX2-mediated regulation of the DDR is independent from its functions in mitosis [3]. In agreement with our finding, a large-scale proteomic study identified TPX2 as a potential in vivo substrate in response to IR induced-DSB [65]. Among 900 phosphorylation sites from over 700 proteins in their database, TPX2 was found to be phosphorylated at one serine site, Ser634 by ATM or ATR in response to DNA damage. They specifically used the antibodies against the ATM and ATR consensus motif (phospho-SQ/TQ) for immunoprecipitation and SILAC (stable isotope labeling

with amino acids in cell culture) studies in human embryonic kidney 293T (HEK) cells for their screening [65]. Even though this phosphorylation has not been validated and its exact role in the

DDR remains to be determined, the results suggest that the function of TPX2 in DDR might be regulated by phosphorylation.

1.3 TPX2 in post-mitotic cells

TPX2 also plays several important roles in post-mitotic neurons. TPX2 is expressed in dorsal root ganglion (DRG) and mouse cortical neurons [3, 66]. Aurora A is also expressed in these cells, especially at neurite hillock (a part of the neuronal cell body that connects to the neurite). Interestingly, this kinase is phosphorylated and activated during neurite extension [66].

However, unlike in cycling cells, in neurons, Aurora A kinase is phosphorylated at Thr287 by atypical protein kinase C (aPKC). More precisely, TPX2 binds to and activates Aurora A as well as co-localizes with phosphorylated Aurora A (Thr287) at neurite hillocks [66]. Interestingly, phosphorylation of Aurora A at Thr287 by aPKC enhances Aurora A binding with TPX2. This

TPX2-Aurora A binding leads to the enhanced Aurora A activation. In turn, Aurora A activation results in phosphorylation of another MAP, called NDEL1, an important organizer of MTs in neurons [66]. When TPX2, Aurora A or NDEL1 are depleted in DRG neurons by their corresponding siRNAs, neurite extension is severely impaired. This result shows that TPX2 pathway plays an important role in remodelling of MTs during neurite elongation [66].

In addition, the previous finding from our laboratory shows that in neurons TPX2 also plays a role in the DNA damage response by regulating the levels of -H2AX upon ionizing irradiation [3]. In primary cultured cortical neurons, depletion of TPX2 by siRNA treatment results in increased levels of -H2AX in response to IR, whereas overexpression of TPX2 leads

to the opposite phenotype (i.e. decrease of γ-H2AX levels). Whether this phenotype is related to

Aurora A or not remains unknown [3].

1.4 TPX2 is a phospho-protein

1.4.1 The functions of TPX2 are likely regulated by phosphorylation

Protein phosphorylation, the addition of a phosphate group to a protein at an amino acid residue of serine, threonine or tyrosine, is mediated by a protein kinase. Protein phosphorylation is one of the key regulatory mechanisms in many cellular processes, such as mitotic spindle assembly, DNA damage response and cell cycle control. Deregulation of these processes via hyper or hypo- phosphorylation can lead to tumorigenesis [67, 68].

Human TPX2 contains 747 amino acids, for a predicted mass of 86kDa. However, the observed molecular weight of TPX2 on an SDS-PAGE gel is about 100kDa. This suggests post- translational modifications of TPX2 [69]. In support of this idea, when TPX2 was first discovered by the Heidebrecht group, they found that human TPX2 is phosphorylated during M phase [5]. Also, using Xenopus egg extracts, Wittmann et al. showed that TPX2 is phosphorylated specifically during mitosis and this phosphorylation is stimulated specifically by

MT polymerization [8]. Using the MAP fractions purified from mitotic Xenopus egg extracts, they already observed a higher molecular weight band for TPX2 by SDS-PAGE [14]. Using mass spectrometry, they detected several putative cdc2 sites and MAP kinase sites on TPX2 from these extracts. In fact, PhosphoSitePlus, an online database system providing information for protein posttranslational modification research shows that TPX2 has over 40 in vivo putative

phosphorylation sites (Figure 1.2) [4]. These phosphorylation sites are mostly derived from reports in the literature using high-throughput mass spectrometry screening. Most of these phosphorylation sites are not validated and their functions remain to be investigated. Considering the functions of TPX2 during cell cycle and the fact that the levels of TPX2 expression are tightly regulated by cell cycle, the functions of TPX2 are likely to be regulated by phosphorylation.

1.4.2 Some putative kinases important for the functions of TPX2

1) Cyclin-dependent kinases (Cdks)

During cell cycle, the progression into each phase is governed by the sequential assembly, activation and inactivation of cell cycle regulatory molecules. The main players of cell cycle control and progression are Cdks and their regulatory proteins, the cyclins [70-72]. Cdks are a family of serine/threonine kinase proteins that are activated by association with their appropriate binding partner, a cyclin. Each activated Cdk triggers major events of the cell cycle by phosphorylating target proteins to mediate cell cycle progression [72-75]. A Cdk without its cyclin partner has no kinase activity. There are 9 known Cdks in human cells so far (Table 1.1).

Cdk1, Cdk2, Cdk3, Cdk4 and Cdk6 function in cell cycle progression. Whereas, Cdk5 is a unique Cdk that does not play any role in cell cycle progression, but it participates in post- mitotic events such as neuronal migration and neurite outgrowth [76, 77]. Even though Cdk5 expression can be found widely in most mammalian tissues and cultured cell lines, high levels of

Cdk5 kinase activity can be detected principally in the nervous system [77, 78]. In addition,

Cdk5 is activated by its three different neuron-specific activators, p25, p35 and p39. Therefore, the function of Cdk5 is most likely specific to neuronal functions [78-80].

Table 1.1 Human Cdks, their associated cyclins (or activators), and their functions [68, 81-85]

CDK Cyclin partner/ activator Function

Cdk1 Cyclin B M phase

Cdk2 Cyclin E G1/S transition S phase Cdk2 Cyclin A G2 phase Cdk3 Cyclin C G1 phase

Cdk4 Cyclin D G1 phase neuronal maturation neuronal migration neurite outgrowth synaptic plasticity Cdk5 p35 learning and memory neurotransmitter release axon transport neurodevelopment actin cytoskeleton dynamics neurite outgrowth Cdk5 p39 neurodevelopment

Cdk6 Cyclin D G1 phase

Cdk-activating kinase Cdk7 Cyclin H transcription

Cdk8 Cyclin C Transcription

Cdk9 Cyclin T Transcription

Cdk1 and Cdk2 play important roles in promoting cell cycle progression. Cdk1/cyclin A or cyclin E permits the transition from G1 to S phase. The Cdk1/cyclin B complex allows mitotic entry from G2 phase [75]. Conversely, Cdk2/cyclin E promotes the transition from G1 to S phase, while the Cdk2/cyclin A complex is required for progression through S phase (see Table

1.1) [74, 75]. Cdk1 and Cdk2 share the common phosphorylation motif, [S/T*]PX[K/R] (S/T: phosphorylated serine or threonine, P: proline, X: any amino acid, K: lysine, and R: arginine).

Because the functions of TPX2 are regulated during the cell cycle, Cdks might be strong candidates to phosphorylate TPX2 and regulate its functions during the cell cycle.

Besides the roles of Cdks in cell cycle control, there is increasing evidence that Cdks are involved in several aspects of the DNA damage response pathway by activating DNA damage checkpoint and initiating DNA repair [86]. For example, it has been reported that Cdk1 phosphorylates BRCA1 at Ser1189, Ser1191 and Ser1497 after DNA damage, and such phosphorylation is important for S phase checkpoint control [87]. Treating cells with RNAi or

Cdk1 inhibitors resulted in impaired activation of S phase checkpoint. Consistent with this, in budding yeast, Cdk1 (cdc28) is essential for activating DNA end resection and DNA damage checkpoint mediated by Mec1 (the yeast homolog of ATR) upon DNA damage [88].

Furthermore, BRCA1 phosphorylation by Cdk1 is important for recruiting Rad51 for repair by homologous recombination [88]. Therefore, these Cdks might also be responsible for regulating

TPX2 function in DDR during interphase.

2) Aurora kinase and Polo-like kinase

Aurora kinase and Polo-like kinase (Plk) are also important for cell cycle regulation in addition to Cdks.

In humans, the Aurora kinase family consists of three members. Aurora A is a major mitotic kinase, known to play critical roles in cell cycle control that regulates accurate mitotic entry and controls G2 checkpoint [30]. Aurora B is also important during mitosis. It is part of the chromosomal passenger complex (CPC) consisting of inner centromere protein (INCENP),

Survivin and Borealin [89, 90]. Aurora B is required for proper chromosome segregation since it mediates mitotic spindle attachment to the centromere and chromosome condensation. Aurora C is the least known kinase and its expression is primarily found in testis [91, 92]. Aurora C can compensate for Aurora B [92]. Recently, TPX2 was found to be a novel activator of Aurora B kinase in the CPC complex, cooperating with other complex proteins to increase Aurora B activity [93].

Polo-like kinases (Plks) are a conserved serine/threonine kinase family consisting of 5 different members (Plk1-5). They mainly regulate cell cycle progression from G2 to mitosis.

They also play roles in mitotic spindle formation and Cdk/cyclin complex activation at the onset of mitosis [94]. Among the Plk family members, Plk1 is the most studied. It promotes maturation of centrosome and mitotic spindle assembly. In addition, Plk1 activates the anaphase-promoting complex/cyclosome (APC/C) for mitotic exit and promotes cytokinesis by recruiting proteins to the spindle and midbody [95]. Finally, Plk1 has a role in DNA damage checkpoints that arrest cells at several points in the G2 and M phases of the cell cycle [96].

According to Fu et al., Aurora kinase, Plks and Cdks collaborate together to regulate the cell cycle [97]. Unlike Cdk-mediated phosphorylation that drives cell cycle progression, Aurora kinases and Plks are important for preventing errors during the cell cycle. They ensure proper cell division through spatial and temporal regulation of phosphorylation of mediators [97].

3) ATM, ATR and DNA-PK

ATM (ataxia telangiectasia mutated), ATR (ATM and Rad3-related) and DNA-PK are three important kinases that mediate the DNA damage response (DDR).

Upon DNA damage, ATM kinase is recruited to DSBs by the MRN (Mre11-Rad50-

NBS1) complex. ATR kinase, together with its regulator ATRIP (ATR-interacting protein), recognizes single-stranded DNA (ssDNA) at stalled replication forks or generated by DSB processing. Both ATM and ATR phosphorylate several proteins to induce activation of checkpoint signaling including several common substrates such as Checkpoint kinase 1 and 2

(Chk1 and Chk2) and p53. This activation of checkpoint signaling can lead to cell cycle arrest,

DNA repair or apoptosis.

DNA-PK is composed of a catalytic subunit (DNA-PKcs) and a heterodimeric DNA binding subunit (Ku). The binding of Ku to DNA is required for DNA-PK activation [98].

This kinase is activated by IR or UV-induced DNA damage. It also plays a critical role in the non-homologous end-joining repair (NHEJ) pathway, a repair pathway that rejoins double-strand breaks [99]. In addition, together with ATR and ATM, Ku is found to be involved in S phase

DNA damage checkpoint [100, 101]. Upon DNA-PK activation, it phosphorylates several downstream substrates to implement DDR leading to p53-mediated apoptosis [102]. Interestingly, these kinases are also involved in mitotic spindle assembly. Brown et al. showed that ATM and

ATR mediated-CEP63 phosphorylation leads to the delocalization of CEP-63 from the centrosomes and prevents centrosome-dependent spindle assembly [103]. Also, DNA-PK seems to have a role in the stabilization of spindle formation and catastrophe prevention during mitosis upon DNA damage [104]. Knockdown of DNA-PKcs in HeLa cells by siRNA treatment leads to

disruption of mitotic spindles, causes cells to have multiple nuclei and results in mitotic cell death in response to ionizing irradiation [104]. Therefore, ATM, ATR and DNA-PK might be good candidate kinases to control the functions of TPX2 in DDR and mitotic spindle assembly.

1.4.3 Mass spectrometry based-phosphoproteome screenings to identify in vivo phosphorylation sites of mitotic phospho-proteins

Protein phosphorylation is considered as one of the key mechanisms regulating mitotic spindle assembly, DNA damage response and cell cycle control. To better understand the molecular mechanisms behind these cellular processes, an inventory of phosphorylation sites of proteins and substrates of the main kinases related to these cellular processes has been established using many different approaches by many research groups.

There are several conventional methods for identifying in vivo protein phosphorylation sites, such as radioactive labeling, thin-layer chromatography-based phosphopeptide mapping,

Edman sequencing or site-directed mutagenesis [105-108]. Due to its sensitivity and speed, mass spectrometry technology has emerged as a popular and powerful technology to map in vivo phosphorylation sites. Numerous papers reported mass spectrometry analysis based on phosphoproteomic analysis of mitotic, DDR or cell cycle control phosphorylation.

In 2006, Nousiainen et al. performed IMAC (immobilized metal ion chromatography)- based phosphopeptide enrichment and large-scale mass spectrometry analysis to identify in vivo phosphorylation sites of proteins purified from mitotic spindles of HeLa cells. They identified

736 phosphorylation sites on 260 proteins. Among those sites, 312 sites were from spindle proteins that had been previously described. Most of the identified phosphorylation sites were mainly consensus sites for Cdk1, Plk1 and Aurora A kinases [109].

In 2007, Matsuoka et al. performed a mass spectrometry-based proteomic screening using antibodies against the consensus phosphorylation sites of ATM and ATR upon DNA damage in human embryonic kidney (HEK) 293 cells and mouse NIH3T3 cells. They found over 700 proteins (human and mouse) and 900 phosphorylated sites in response to DNA damage [65].

From their bioinformatic analysis to group these proteins as modules, they identified a large number of functional modules that are not only previously implicated in DDR pathway, but also never been linked, such as AKT-insulin signaling and RNA metabolism pathways.

In 2008, Chi et al. performed a proteome-wide scale analysis to identify Cdk2 substrates in HEK 293 cell lysates [110]. Using a kinase engineering technology followed by mass spectrometry analysis, 180 putative substrates of Cdk2/cyclin A and more than 200 phosphorylation sites were identified. More precisely, they used a mutant form of Cdk2/cyclin A, and an ATP -S analogue to label proteins in cell lysates with thiophosphates in an in vitro kinase assay to improve the specificity of substrate phosphorylation. Then, the thiophosphorylated peptides were isolated using chemical enrichment and the peptides were analyzed by LC-MS/MS mass spectrometry. The putative substrates were proteins involved in various cellular processes such as cell cycle control, DNA/RNA metabolism and translation.

Some of the identified phosphorylation sites contained both consensus and non-consensus motifs for Cdks. 15% of the identified proteins were already known as Cdk substrates, whereas the remaining proteins have never been reported as Cdk substrates [110]. This paper suggested that

Cdks control cell division through massive networks of cellular substrates and the method used in this paper can be used to study the substrates of other important kinases.

In 2010, Olsen et al. performed a high throughput mass spectrometry analysis using stable isotope labeling of amino acids in cell culture (SILAC) with HeLa cells collected at 6 time

points of the cell cycle (G1, G1/S, early S, late S, S/G2, and M phase) to monitor the dynamics of the proteome and phosphoproteome throughout the cell cycle [111]. They identified 20,443 phosphorylation sites of 6027 proteins. They grouped these proteins and their phosphorylation sites based on cell cycle kinetics and compared their results to the published microarray data of the HeLa transcriptome [112]. They found that most of the phosphorylation sites and over 20% of proteins detected in their analysis were involved in mitotic regulation. In particular, most of the sites showing increased phosphorylation were known to be phosphorylated during mitosis by

Cdk1 or Cdk2.

From the proteomic screenings described above, several phosphorylation sites of TPX2 were consistently represented, thereby strengthening the idea that TPX2 is regulated by phosphorylation during mitosis and DDR.

1.4.4 TPX2 phosphorylation sites already characterized in the literature

In the current literature, the functions of only three phosphorylation sites in TPX2 have been characterized. Using the Xenopus system, Echkerdt et al. identified Serine 204 (Ser204) in

Xenopus TPX2 [113] as a Plx1 (Xenopus Plk1) phosphorylation target. Plx1 is a kinase that promotes mitotic progression in the Xenopus system. When TPX2 is phosphorylated at Ser 204 by Plx1, TPX2 activates Aurora A. Sequence comparison (Appendix Figure 2) shows that this site is not conserved between other species and no research has been done to identify phosphorylation sites for this particular function of TPX2 in other species, including Human.

The two other sites that have been characterized in Xenopus TPX2 are Tyrosine 8 (Tyr8) and Tyrosine 10 (Tyr10) [32].

These two sites are conserved between Xenopus and Human TPX2. These sites are in the region of amino acids of TPX2 that closely bind Aurora A [16]. As mentioned in the introduction, binding of TPX2 to Aurora A kinase results in activation of this kinase. At the same time, it leads to phosphorylation of TPX2. Mutations of both Tyr8 and Tyr10 to Alanine prevent binding of TPX2 to Aurora A and abolished Aurora A activation. Interestingly, these mutations also prevent TPX2 phosphorylation by Aurora A [32]. Thus, Tyr8 and Tyr10 are required to mediate the binding of TPX2 to Aurora A and activate Aurora A. However, these phosphorylation sites observed in vitro have never been confirmed in vivo.

In summary, TPX2 is an important mitotic regulator whose functions are likely controlled by phosphorylation. However, very little research has been done to characterize and investigate the functions of TPX2 phosphorylation sites.

1.5 TPX2 overexpression in cancers

TPX2 is overexpressed in several human cancers such as lung, cervix, liver, prostate, pancreas, and ovary [114-118]. In addition, the location of the human tpx2 gene is on chromosome arm 20q11[119], which is often amplified in various cancers [120]. Overexpression of TPX2 is associated with increased DNA copy number at chromosome region (20q11) in some cancers [114, 117, 118]. Furthermore, expression of TPX2 positively correlates with the progression of several cancers such as malignant astrocytoma and squamous cell lung carcinoma, and its overexpression is associated with decreased patient survival [121-124] . Therefore, it has been suggested that TPX2 can be used as a prognostic indicator for some cancers [122, 124, 125].

Overexpression of TPX2 has been correlated with chromosomal instability (CIN) in several human cancers [126]. CIN is a phenotype of many cancer cell lines having an abnormally

high rate of chromosome loss and increased cell division, which results in aneuploidy [126, 127].

Using in silico methods, Carter et al. reported 70 genes deregulated in tumors with unstable chromosomes. Those genes included TPX2, Cdk1, PCNA (Proliferating cell nuclear antigen) and

Aurora A, all regulators of mitosis. TPX2 shows the strongest correlation with chromosomal instability. Overexpression of TPX2 and aneuploid phenotypes were frequently shown in cancers such as breast carcinoma [126]. Essential mitotic proteins such as Aurora A and Mad2 (Mitotic arrest deficient 2, an essential spindle checkpoint protein) are overexpressed in many cancers and promote tumorigenesis when overexpressed in mouse models, or are correlated with development of tumors in human malignancies [128, 129]. Of note, TPX2 heterozygous mice accumulate cells with aneuploidy with age, This indicates genomic instability, and susceptibility to tumor development. The ratio of tumor development in TPX2 heterozygous mice was much higher than that in TPX2 wild-type mice [41]. All these results suggest that deregulating the expression and function of TPX2 induces aneuploidy, thereby, leading to tumor development

[41].

1.6 Rationale and hypothesis

1.6.1 Rationale for Studying phosphorylation of Threonine 72 on TPX2

Several papers using in vivo mass spectrometry-based screenings revealed that TPX2 is a mitotic phospho-protein with over 40 putative phosphorylation sites [4, 111, 130-136]. These putative phosphorylation sites (Figure 1.2) have never been validated and their characterization has never been performed.

By virtue of our interest for neuroscience research, Dr. James Wang, a former member of the Nguyen laboratory, searched for sites on TPX2 susceptible to be phosphorylated in neurons.

Using phosphorylation site prediction software, Dr. Wang identified Thr72 as a major site for

Cdk5, an atypical Cdk. Importantly, Thr72 residue is conserved among human, mouse, rat and frog (Figure 1.8). For these reasons, Thr72 was selected for further study. Interestingly, Cdk5 recognizes the same phosphorylation motif as Cdk1 and Cdk2. Thus, TPX2 may also be phosphorylated by Cdk1 and Cdk2 at Thr72.

Figure 1.8 Multiple sequence alignment of TPX2 shows that Thr72 on TPX2 is conserved across different species

Part of human TPX2 sequence (a.a. 41 to 80) was aligned with the corresponding sequences from other species (mouse, rat and frog) using DNAMAN software (Lynnon Corperation). The sequence alignment shows that Thr72 in human TPX2 is conserved in other species. Note: the full sequence alignment across these species is shown in Appendix Figure 2.

To prove that Thr72 is one of the Cdk5 phosphorylation sites, Dr. Wang first generated the antibodies specific to pThr-72 (See 2.3 for the method of generating pThr72-TPX2 antibodies). He then went on to perform the in vitro phosphorylation kinase assays using active

Cdk1/cyclin B, Cdk2/cyclin A and Cdk5/p25 complexes and GST-TPX2 fusion proteins with or without mutation at Thr72 (GST-TPX2, GST-TPX2 T72A and GST-TPX2 T72E). The purpose of his experiments was 1) to determine which Cdks mediate the phosphorylation of TPX2 at

Thr72, and 2) to test the specificity of the phospho-antibodies in vitro. Interestingly, he found that TPX2 is phosphorylated at Thr72 in vitro by Cdk1/cyclin B and Cdk2/cyclin A, but not by

Cdk5/p25 (Figure 1.9). It is possible that a different Cdk5/activator complex (such as Cdk5/p35 and Cdk5/p39) may phosphorylate TPX2, but not Cdk5/p25. In addition, the results of the in vitro kinase assay showed that the phospho-TPX2 antibodies recognized only the purified TPX2 wild-type protein, which had been phosphorylated at Thr72 by Cdk1/Cdk2 in vitro, but, not the mutated forms of TPX2. From this result, we found out that TPX2 is phosphorylated at Thr72 by

Cdk1 and Cdk2, not by Cdk5/p25 in vitro. Also, the pThr-72 antibodies we generated are specific to Thr72 in vitro. In support of this result, several studies of mitotic phosphorylation proteomics indicated that Thr72 might be important for TPX2 phosphorylation during mitosis

[109, 130, 132, 134-137].

Taken together, there was enough accumulating evidence to suggest that phosphorylation of TPX2 at Thr72 may regulate the functions of TPX2.

Figure 1.9 In vitro kinase assay showing that TPX2 is phosphorylated at Thr72 in vitro by Cdk1 and Cdk2 and the in vitro specificity of phospho-TPX2 antibodies against Thr72 (Data from Dr. Wang)

Purified GST fusion protein TPX2 wild-type, GST-TPX2-T72A (Threonine at Thr72 mutated to Alanine to abolish the phosphorylation) or GST-TPX2-T72E (Threonine at Thr72 mutated to Glutamic acid to mimic the phosphorylation) was incubated with each active Cdk/cyclin complex in the presence of 1 mM cold ATP. All kinase reactions were stopped by adding 2X SDS sample buffer and the samples were run on SDS-Polyacrylamide gel electrophoresis followed by western blot detection using each indicated antibody; TPX2, pThr-72, Cdk1, Cdk2 and Cdk5 antibodies.

1.6.2. Main Hypothesis

Based on Dr. Wang’s preliminary data and published data of phosphorylation screenings,

I hypothesize that TPX2 is phosphorylated at Thr72 in vivo, and this phosphorylation is important for the mitotic spindle assembly and DDR functions of TPX2.

CHAPTER TWO: MATERIALS AND METHODS

2.1 Cell culture

HeLa cells [Catalogue number CCL-2; American Type Culture Collection (ATCC)] were the main cells used in the thesis. They were maintained in high glucose Dulbecco’s Modified

Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin- streptomycin (all from Gibco). U20S cells (Catalogue # 40345; ATCC) were used for DNA damage experiments were maintained in the same conditions as HeLa cells. Both cell lines were maintained in a humidified incubator at 37 C and 5% carbon dioxide.

2.2 Generation of GFP-TPX2 wild-type and mutant constructs

To generate GFP-TPX2 wild-type vector, human TPX2 cDNA (bp 3-2241) was cloned into pEGFP-C1 (Clontech). Site-directed mutagenesis was carried out on this plasmid to generate a phospho-dead mutant and a phospho-mimetic mutant. The phospho-dead mutant (T72A) was generated by PCR using the mutation primer set (5’ phosphorylated T72A forward primer:

CAAGCTATTGTCGCACCTTTGAAACCAG and 5’ phosphorylated T72A reverse primer:

CTGAAGATTAGCCTTTCTCAAAGGAG) and Phusion Hot Start DNA Polymerase

(Finnzymes) to mutate Threonine (ACA) to Alanine (GCA). Mutated PCR products were circularized by ligation using a Rapid DNA ligation kit (Thermo Scientific). The phospho- mimetic mutant (T72E) was generated similarly, using a different forward primer (5’ phosphorylated T72E forward primer: CAAGCATTGTCGAACCTTTGAAACCAG) to mutate

Threonine 72 (ACA) to Glutamic acid (GAA). Both mutations were confirmed by DNA sequencing (University Core DNA Services, University of Calgary).

2.3 Generation of phospho-specific antibodies

The phospho-specific antibodies recognizing phospho-TPX2 at Threonine 72 were raised in rabbits against the phospho-peptide, (K)LQQAI V(pT*72)PLKPVD, which corresponds to amino acids 66 to 78 of human TPX2. The first lysine (K) was added to the N-terminus of the peptides to promote its solubility. The peptide was conjugated to keyhole limpet haemocyanin

(carrier for the phospho-peptide hapten) and was used for immunization of rabbits. The phospho- specific antibodies were then affinity purified from the antiserum using laboratory standard protocol. Briefly, non-phosphorylated peptide or phosphorylated peptide was coupled to two different agarose columns of SulfoLink coupling gel using SulfoLink Immobilization kits

(Pierce). Crude serum was first applied to the column coupled with non-phosphorylated peptide, then, the flow-through was applied to the column coupled with the phosphorylated peptide. The phospho-specific antibodies were eluted from the phosphorylated peptide-column.

2.4 RNA interference (RNAi)

Two unique TPX2-specific siRNA oligos were used to knock down endogenous TPX2.

The first TPX2 siRNA oligo (5'-GAAUGGAACUGGAGGGCUU-3'; called TPX2 cds siRNA in this thesis) has been described previously [22]. This TPX2 cds siRNA targets a coding region of human TPX2 (hTPX2) 160-179 bp from the start codon. The other siRNA oligo (5'-

AAGGCTAATAATGAGATCTAA-3'; called TPX2 UTR siRNA in this thesis) targets a 3’ untranslated region (3’UTR) of hTPX2 mRNA. This TPX2 UTR siRNA was purchased from

Qiagen. “AllStars negative control siRNA” (Qiagen) was used as a negative control.

2.5 Transfection of cells with plasmid constructs and siRNAs

For transient transfection of cells with single siRNAs, cells were seeded on 60 mm-cell culture dishes shortly before transfection. siRNA transfection was performed using HiPerFect transfection reagent (Qiagen) according to the manufacturer’s protocol using 20 nM siRNA and

20 μl of HiPerFect reagent for each dish. Transient transfection of cells with plasmid DNAs only or co-transfection of siRNA with plasmids was performed using Lipofectamine 2000

(Invitrogen). Cells were seeded on 60 mm-cell culture dishes one day before transfection. Cells were transfected according to the manufacturer’s recommended protocol using 3 g of plasmids with 10 μl of Lipofectamine 2000 transfection reagent or 20 nM siRNA oligos and 3 g of plasmids with 10 μl of Lipofectamine 2000 transfection reagent for each dish.

2.6 Cell cycle synchronization

HeLa cells were synchronized at S phase by double thymidine block. Cells were treated with 2 mM thymidine for 18 hours, followed by 8 hours of release in fresh DMEM, and then, re- treated with thymidine for 18 hours.

Cells were synchronized at M phase using nocodazole treatment. For un-transfected cells,

3-5 hours after seeding, cells were treated with 100 ng/ml of nocodazole (Sigma) for 16-17 hours, followed by PBS washes twice and incubation in fresh DMEM for 30 minutes. For synchronization of plasmid-transfected cells at M phase, 24-29 hours after transfection, cells were changed into fresh DMEM containing 100 ng/ml of nocodazole (Sigma), and were incubated for 16-17 hours. Then, cells were washed twice in PBS and incubated in fresh DMEM

for 30 minutes. Cells at each cell cycle stage (S, M phase and non-synchronized cells) were harvested for western blot analysis or were confirmed by flow cytometry analysis.

2.7 Protein extraction and western blotting

After PBS washing, cells were lysed in an appropriate volume of mild and non- denaturing lysis buffer [20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2EDTA, 1 mM

EGTA, 1% (v/v) Triton, 2.5 mM sodium pyrophosphate, 1 mM beta-glycerophosphate, 1 mM

Na3VO4, 1 mM phenylmethylsulfonyl fluoride (PMSF), a protease inhibitor cocktail tablet

(Roche Applied Science) and 1 M microcystin-LR (Cayman chemical)]. Lysates were incubated on ice for 10 minutes and sonicated for 5 seconds twice using a Sonic Dismembrator

Model 100 at level 4. Cells were centrifuged for 10 minutes at 14,000 g to get clear supernatants.

Supernatants were collected and protein concentrations were quantified using the Bradford method (Bio-Rad protein assay dye reagent, Bio-Rad laboratories) and bovine serum albumin

(BSA, EMD millipore) as a standard. Proteins in SDS gel-loading buffer [4% (v/v) sodium dodecyl sulfate (SDS), 0.2% (v/v) bromophenol blue, 20% (v/v) glycerol, 20 mM β- mercaptoethanol] were run on SDS-PAGE and transferred onto PVDF (polyvinylidene fluoride) membrane for western blot analysis. After blocking in 5% skim milk solution in PBS-T (1X PBS containing 0.2% Tween 20), each membrane was incubated with specific primary antibodies overnight at 4 C. After washing using PBS-T for 5 minutes 3 times, the membranes were incubated in 5% skim milk in PBS-T containing secondary antibodies at 1:5000 dilutions for 1 hour at room temperature. After PBS-T washing 3 times each for 5 minutes, Western Lighting 

Enhanced Chemiluminescence solution (Perkin Elmer) was added onto membranes for 2 minutes, and X-ray films (Hyblot CLTM autoradiograph film, Denville Scientific, Inc.) were exposed to the membranes for various times depending on the intensity of the signal. The resulting films 46

were developed using a Kodak X-OMAT 2000A processor. To re-probe membranes with other antibodies, the membranes were incubated in a striping buffer [2% SDS, 62.5 mM Tris (pH6.8),

100 mM -mercaptoethanol] for 15 minutes at 50 C, washed with PBS-T for 3 times, blocked in

5% skim milk in PBS-T, and standard western blot procedures as described above were performed. Results on X-ray films were scanned using a DuoScan T1200 scanner (Agfa).

Quantification of each band signal was performed using Quantity One analysis software (Bio-rad laboratories). Signals were normalized to the levels of loading controls (-actin or non- phosphorylated form of the proteins of interest).

2.8 Antibodies

Antibodies used for western blots (WB), immunoprecipitations (IP) and immunofluorescence (IF) in this thesis are listed in Table 2.1. The information about their specific applications, source, and the dilutions used is included in the table for each antibody.

Table 2.1 List of antibodies

Name of Antibody Application Dilution Source

Primary antibodies -actin mouse monoclonal WB 1:10,000 Chemicon antibody (mAb) cyclin B mouse mAb WB, IP 1:1000 Abcam phospho-TPX2 Thr72 WB 1:2000 Nguyen laboratory rabbit polyclonal antibody (Home-made) (pAb) Described in the section 2.3 phospho -MAPK/Cdk WB 1:1000 Cell signaling substrates (PXS*P or S*PXR/K) (34B2) rabbit mAb phospho-histone H2A.X WB, IF 1:2000, 1:1000 EMD Millipore (Ser139) clone JBW301 mouse mAb H2AX rabbit pAb WB 1:5000 Abcam -Tubulin mouse mAb WB, IF 1:5000, 1:1000 Sigma Cy3 conjugated--Tubulin IF 1:1000 Sigma mouse mAb Cdk1 mouse mAb WB 1:1000 Santa Cruz Biotechnology Cdk2 mouse mAb WB 1:1000 Santa Cruz Biotechnology TPX2 183 rabbit pAb IP 1:100 Novus Biologicals TPX2 184 rabbit pAb WB, IF, IP 1:1000, 1:1000, Novus Biologicals 1:100 TPX2 mouse mAb IF 1:1000 Abcam [18D5-1] Nucleolin mouse mAb IF 1:2000 Abcam [4E2] (a kind gift from Dr. Donna Senger, University of Calgary) Secondary antibodies Donkey anti-rabbit IgG WB 1:5000 GE Healthcare

HRP Sheep anti-mouse IgG WB 1:5000 GE Healthcare HRP Cy3-conjugated anti- IF 1:5000 Jackson Immuno mouse IgG Research Cy3-conjugated anti- IF 1:5000 Jackson Immuno rabbit IgG Research FITC-conjugated anti- IF 1:5000 Jackson Immuno mouse IgG Research FITC-conjugated anti- IF 1:5000 Jackson Immuno rabbit IgG Research Note: Three different TPX2 antibodies were used (TPX2 183, TPX2 184 and TPX2 18D5-1). The immunogen of TPX2 183 is N-terminus of human TPX2 from 150 to 200 a.a. residues. The immunogen of TPX2 184 is C-terminus of human TPX2 from 700 to 749 a.a. residues. The immungen of TPX2 18D5-1 is the recombinant human TPX2 protein (a.a. residues were not specified).

2.9 Flow cytometry analysis

After transfection, with or without synchronization, cells were released by Trypsin-

EDTA (0.25%, Invitrogen), washed with PBS twice, re-suspended in 500 l of PBS and then, fixed with 500 l ethanol for at least 24 hours. After centrifugation at 2,095 g for 10 minutes, the cell pellets were re-suspended in 500 l PBS and 500 l propidium iodide solution

(Molecular Probes) containing Ribonuclease A (RNase A, Sigma) and Triton X-100 (Fisher scientific). The prepared samples were sent to the Flow cytometry Core Facility (University of

Calgary) for cell cycle profiling (% of cells in G1, S and G2/M phase of cell cycle) using a

Becton-Dickinson FACScan flow cytometer.

2.10 Mass spectrometry analysis to identify in vivo phosphorylation sites of TPX2

HeLa cells were synchronized using 100 ng/ml of nocodazole treatment for 16 hours.

After three times of PBS washing, cells were released into fresh DMEM without nocodazole for

30 minutes. Cells were harvested, washed with PBS twice before addition of lysis buffer to get protein lysates. Protein concentrations of each sample were measured using the Bradford protein assay (Bio-rad laboratories). Endogenous TPX2 was immunoprecipitated from 10 mg of total protein using TPX2 antibodies (184) and Protein A/G Sepharose 4 Fast Flow beads. The beads were washed five times with 500 l of lysis buffer containing protease inhibitors. The IP samples were run on SDS-PAGE, and Coomassie blue-stained bands around the expected size of

100 kDa were excised from the gel and sent for mass spectrometry analysis at the University of

Victoria Proteomic Center. Another set of IP samples were analysed by western blotting using pThr-72 and total TPX2 antibodies to confirm the size of the TPX2 IP band. Sample preparation for LC-MS/MS analysis was done in the University of Victoria Proteomic Center. For mass spectrometry data analysis, raw files analyzed with Proteome Discoverer software (Thermo

Scientific) were submitted to Mascot 2.2 against Homo sapiens of Uniprot-Swissprot, Uniprot

Trembl and IPI human database.

2.11 Quantification of the cells with different numbers of mitotic spindles

HeLa cells were seeded on glass coverslips at a density of 5 x 104 cells per well in 24- well cell culture plates. The next day, cells were transfected with 1 g of each plasmid construct

(an empty GFP vector, GFP-TPX2 WT or GFP-TPX2 T72A). 24 hours after transfection, cells were synchronized with nocodazole (100 ng/ml) for 16 hours. Cells were released from the nocodazole block by washing in PBS 3 times, then, incubated in fresh media for 30 minutes.

Cells in mitosis were fixed with 4% PFA and stained with Cy3-conjugated -tubulin antibody for mitotic spindle visualization. Stained cells were observed under a fluorescence microscope and

GFP-positive (transfected) cells were identified. Cells in prometaphase or metaphase were categorized into three different classes based on the number of mitotic spindles; monopolar, bipolar and multipolar spindle. Over 100 prometaphase or metaphase cells were scored per slide and five separate slides were examined for each plasmid construct. The percentages of cells in each category for the indicated group (GFP, GFP-TPX2 WT and GFP-TPX2 T72A) were calculated. The differences in the mean values for each group were analyzed by ANOVA.

In a separate experiment, HeLa cells were seeded on glass coverslips in 24-well cell culture plates (5x104 cells/well). The next day, 20 pmol of TPX2 UTR siRNA duplexes (to deplete endogenous TPX2) were co-transfected with 500 ng of each plasmid construct (an empty

GFP vector, GFP-TPX2 WT or GFP-TPX2 T72A) into cells using 2 ul of Lipofectamine 2000

(Invitrogen) according to the manufacturer’s protocol. 24 hours after transfection, cells were synchronized with nocodazole (100 ng/ml) for 16 hours. For the release, cells were washed in

PBS 3 times and then, released into fresh media for 30 minutes. The rest of the steps for fixation and quantification were exactly the same as already described.

2.12 Treatment with CDK inhibitors

Roscovitine (Sigma) and alsterpaullone (Santa Cruz Biotechnology) were dissolved in

DMSO to get a 20 mM stock solution. HeLa cells were seeded on 100 mm-cell culture dishes at a density of 2x106 cells per dish. 3 hours after seeding, cells were treated with 100 ng/ml of nocodazole for 16 hours. These mitotic cells were incubated with DMSO (as a vehicle control),

20 M or 40 M of drugs (roscovitine or alsterpaullone) for 30 minutes in the presence of

nocodazole in the media. After treatment, cells were trypsinized, and washed in PBS twice. After centrifugation, cell pellets were dissolved in lysis buffer containing protease inhibitor cocktail tablets (Roche) and 1 M microcystin-LR (Cayman chemical). The prepared cell lysates were used for IP experiments. CDK inhibitor treatments were performed in triplicate.

2.13 Immunoprecipitations (IP)

Cells were harvested, lysed in lysis buffer (as described in the section 2.7) and incubated on ice for 10 minutes. Cell extracts were sonicated for 5 seconds twice using a Sonic

Dismembrator Model 100 at level 4. After lysates were cleared by centrifugation at 16,500 g at 4

ºC for 10 minutes, supernatants were collected for protein quantification. 1 mg protein of each cell lysate was used for IP. Lysates were precleared by incubating with protein A/G sepharose for 45 minutes, and then supernatants were incubated with the primary antibody against the target protein for 3 hours under gentle rotation at 4 C. Next, the samples were incubated with protein A/G sepharose for an additional hour, continuing rotation at 4 ºC . Immunoprecipitates were washed in lysis buffer for 5 minutes four times, and then the beads were re-suspended in

SDS gel-loading buffer, and analyzed by western blotting.

2.14 Lambda protein phosphatase treatment on Immunoprecipitation samples

Immunoprecipitated proteins on beads were washed 4 times with lysis buffer and equally divided into two aliquots. The beads in each tube were washed with 1X lambda protein phosphatase (-PPase) buffer [50 mM Tris-HCl, 0.1mM Na2EDTA, 5 mM DTT, 0.01% (v/v)

Brij 35 (pH 7.5)] two times. Then, the beads in each tube were treated with or without 400 U of

-PPase (NEB) in 1X -PPase buffer at 30 °C for 30 minutes in the presence of 2 mM MnCl2.

Reactions were stopped by adding 5X SDS gel-loading buffer.

2.15 Chromatin fractionation for determining the levels of -H2AX protein.

HeLa cells were treated with 10 Gy of ionizing-radiation (IR) using a GammaCell 1000

Tissue Irradiator (MDS Nordion) with 137Cs as a -radiation source. After 1 hour of recovery at

37 C, cells were washed with PBS twice, harvested and lysed in 400 l of ice cold NETN buffer

[150mM NaCl, 1mM EDTA, 50 mM Tris-Cl (pH 7.4), 1% (v/v) NP40, protease inhibitor cocktail tablets (Roche), 1 M microcystin-LR (Cayman chemical)]. Cells in NETN buffer were left on ice for 30 minutes and extracts were centrifuged at 4 C for 10 minutes to get the insoluble pellets (The chromatin fraction). The chromatin fraction was dissolved in 200 l of 1%

SDS in PBS and sonicated for 5 seconds at level 4. Protein concentrations of each sample in

NETN soluble fraction were determined using the Bio-Rad Bradford protein assay kit and the quantification of proteins in the chromatin fraction was done using the Bio-Rad DC protein assay kit. For both soluble and chromatin fraction, BSA was used as a standard for protein quantification. The equal amount of protein for each sample was loaded on to 12 % SDS-PAGE.

2.16 Immunofluorescent staining and confocal microscopy analysis

Cells on coverslips were fixed with 4% PFA [4% paraformaldehyde in 1X PBS (pH 7.4)] for 10 minutes at 37 C. Fixed cells were washed with PBS twice and blocked in blocking buffer

(3 % BSA, 0.2% Triton-X 100 in 1X PBS) for 1 hour at room temperature. Next, coverslips were incubated with primary antibody in blocking buffer at 4 C overnight. After washing coverslips

in PBS, coverslips were incubated with secondary antibodies conjugated with Cy3 or FITC

(Jackson ImmunoResearch Laboratories) and DAPI for DNA staining. After washing with 1X

PBS three times, cells on coverslips were mounted on microslides (VWR international) using a drop of Aqua-Mount (Thermo scientific). Images were acquired by Nikon D-Eclipse C1 confocal microscope (Nikon) using EZ-C1 software (Nikon).

2.17 Observation of ionizing radiation-induced foci formation and -PPase treatment of U2OS cells on coverslips

U20S cells were grown on coverslips and treated with 5 Gy of IR, followed by a 1 hour- recovery at 37 C. Cells were fixed with 4% PFA for 10 minutes at 37 C, followed by permeabilization with 0.1% Tween 20 for 10 minutes. For phosphatase treatment, fixed and permeabilized cells on coverslips were incubated with -PPase (NEB) (100U/coverslip) in 1X -

PPase buffer [50 mM Tris-HCl, 0.1 mM Na2EDTA, 5 mM DTT, 0.01% (v/v) Brij 35, (pH 7.5) at 30 C for 30 minutes in the presence of 2 mM MnCl2. Control cells on coverslip were incubated in the buffer containing phosphatase inhibitors (10 mM NAF, 10 mM - glycerophosphate) without -PPase under the same conditions as the -PPase-treated cells. Next, cells were incubated in blocking buffer, and prepared with the standard immunostaining protocol for confocal analysis.

2.18 In vitro Cdk kinase assay

The in vitro kinase assay was performed using active Cdk/cyclin complexes and GST fusion proteins. 10 g of GST-TPX2 WT, T72A and T72E fusion proteins (home-made proteins

purified by Dr. James Wang) were incubated with 1 g of each active Cdk1/cyclin B,

Cdk2/cyclin A, or Cdk5/p25 (Millipore) in the presence of 1 mM ATP (Sigma) at 30 °C for 30 minutes in kinase buffers [25 mM Tris-HCl (pH 7.5), 5 mM -glycerophosphate, 2 mM dithiothreitol (DTT), 0.1 mM Na3VO4, 10 mM MgCl2, 8 mM MOPS (pH 7.0), 0.2 mM EDTA].

10 g of GST-TPX2 protein only was incubated with 1 mM ATP in the same condition as above without any active Cdk/cyclin complex. Kinase reactions were stopped by adding 5X SDS-

PAGE loading buffer into each sample. Western blot analysis was performed with reaction products using TPX2 (184), pThr-72, Cdk1, Cdk2 and Cdk5 antibodies.

2.19 Blocking peptide experiments

The sequence of the blocking peptide is LQQAIVTPLKPVD (from 66 to 78 amino acids of human TPX2), which is the same sequence used to generate the phospho-specific TPX2 antibodies at Thr72. The threonine amino acid underlined has been phosphorylated. The generation and purification of the blocking peptide was performed at the University of Calgary peptide service facility. The peptide was synthesized on preparative HPLC (a Waters HPLC system) using a Vydac C18 column. Phospho-threonine peptide modification was carried out using standard Fmoc chemistry (N-α-Fmoc-O-benzyl-L-phospho-threonine derivative reagent was used to prepare this phospho- threonine peptide). The quality of the peptide was evaluated by analytical HPLC and mass spectrometry. Purity assessment by analytical HPLC showed a single peak representing almost 100 % purity of this peptide (the assessment was performed by the University of Calgary peptide service facility). For the blocking experiments, the antibodies were combined with a five, ten, twenty and fifty-fold excess of blocking peptide and preincubated at 4 C overnight. Next, standard immunofluorescent staining was performed with the 55

antibody-peptide mixture or only the antibodies, without the blocking peptide, on PFA-fixed un- synchronized HeLa cells, and un-irradiated/irradiated U2OS cells.

2.20 Statistical analysis

Data analysis to determine p values was performed using an unpaired Student’s t-test when comparing two groups, or ANOVA for comparing multiple groups. P values of 0.05 were considered statistically significant in both unpaired Student’s t-test and ANOVA. The Newman-

Keuls test was performed when difference between multiple groups was revealed by ANOVA

(significant p values).

CHAPTER THREE:

CHARACTERIZATION OF TPX2 PHOSPHORYLATION AT THREONINE 72 IN

HUMAN CANCER CELLS

3.1 Rationale

The functions of TPX2 are likely to be regulated by phosphorylation based on the fact that TPX2 is a mitotic phospho-protein with over 40 putative phosphorylation sites identified from various large-scale proteomic screenings. However, most of these putative phosphorylation sites have never been validated and characterized. In this study, we aimed to characterize TPX2 phosphorylation at one particular phosphorylation site, Threonine 72 (Thr72), as its importance is indicated by its being conserved between human, mice, rat and frog (Figure 1.8). Most importantly, previous data from our laboratory showed that TPX2 is phosphorylated at Thr72 by

Cdk1/cyclin B and Cdk2/cyclin A in vitro (Figure 1.9). In addition, Thr72 is one of the TPX2 phosphorylation sites which commonly appears in large-scale databases of several human mitotic phosphoproteome studies [109-111, 131, 132, 135].

TPX2 has been studied the most in HeLa cells and Xenopus egg extracts. HeLa, an immortal human cancer cell line, is frequently used in cell cycle research, particularly to study dynamic events of mitosis and post-translational modification due to the cell morphology and ease of cell cycle synchronization [38]. Therefore, we also decided to use HeLa cells as our model system.

The main goal in this section was to characterize TPX2 phosphorylation at Thr72 in

HeLa cells. First, I demonstrated the specificity of our in-house generated pThr-72 antibodies in vivo. To show this, I used siRNAs against TPX2, or performed IP experiments with the cell

lysates from GFP-TPX2 WT and T72A mutant transfectants. Next, I showed that Thr72 phosphorylation is cell cycle-dependent, peaking at the M phase of the cell cycle. Finally, by mass spectrometry analysis with mitotic HeLa cells, I confirmed that TPX2 is phosphorylated at

Thr72 in vivo. In doing so, I also identified other phosphorylation sites from my mass spectrometry data. Furthermore, using Cdk inhibitors, I discovered that Cdks can mediate TPX2 phosphorylation at Thr72 in mitotic HeLa cells.

3.2. Results

3.2.1 Characterization of the phospho-specific antibodies against Thr72 of TPX2 in vivo

The phospho-specific antibodies against Thr72 of TPX2 (pThr-72 antibodies) have been previously generated in our laboratory. The specificity of these antibodies was confirmed in vitro using purified GST-TPX2 WT and mutant proteins (Figure 1.9). In this section, I aimed to characterize pThr-72 antibodies in HeLa cells, therefore I started by confirming the specificity of the pThr-72 antibodies in vivo.

In order to characterize the specificity of the pThr-72 antibodies in the cellular context, I performed two different experiments. First, using two unique siRNAs against TPX2, I confirmed these antibodies are specific to TPX2, as the relevant bands on a western blot significantly decreased after TPX2 knockdown. Second, I transfected GFP-TPX2 WT and GFP-TPX2 T72A mutant into HeLa cells and showed the antibodies are specific to the immunoprecipitated phosphorylated TPX2-WT, but not the T72A mutant. The phosphorylation status of the TPX2-

WT protein was further confirmed by lambda protein phosphatase (-PPase) treatment.

3.2.1.1 siRNA treatment using two unique TPX2-specific siRNAs showed the phospho-TPX2 antibodies against Thr72 are specific to TPX2.

To test the specificity of pThr-72 antibodies to TPX2 in vivo, HeLa cells were transiently transfected with control siRNA or two different TPX2 siRNAs for 24 hours and then, cells were synchronized at M phase with 100 ng/ml of nocodazole treatment for 16 hours. The levels of

TPX2 expression were significantly reduced by 67 % and 70 % respectively in the samples treated with TPX2 siRNA #1 (the siRNA targeting 3’ untranslated region of TPX2; TPX2 UTR siRNA) and #2 (the siRNA targeting coding sequences of humanTPX2; TPX2 cds siRNA), compared to those of the control siRNA-treated samples. Consistent with this, the expression levels of TPX2 phosphorylated at Thr72 was also reduced by 69% and 72% respectively in the samples treated with TPX2 UTR siRNA and TPX2 cds siRNA. In order to make sure the result of western blots was not affected by stripping, the western blots were probed with the phospho-

TPX2 antibodies against Thr72 (pThr-72) first, and after stripping, the blots were re-probed with the total TPX2 (184) antibodies. Figure 3.1 shows that pThr-72 antibodies are specific to TPX2 upon depletion of TPX2 by TPX2 siRNA treatment using two different TPX2-specific siRNAs.

Figure 3.1 The pThr-72 antibodies are specific to endogenous TPX2 in vivo

A) HeLa cells were transfected with control siRNA or one of two different TPX2 siRNAs for 24 hours and synchronized at M phase with nocodazole treatment (100 ng/ml). Cells were harvested and lysed with lysis buffer. Samples were run on SDS-PAGE, followed by western blotting, first probed with the pThr-72 antibodies, then stripped and re-probed with total TPX2 (184) antibodies. Levels of actin were used as loading controls. B) Bar graphs of quantification of relative expression levels of TPX2/pThr-72. Data from four independent western blot experiments are shown. The level of each sample was compared to the sample treated with control siRNA. Mean +/- Standard error of the mean (S.E.M.). For the relative expression levels of pThr-72, control siRNA: 1 +/- 0, TPX2 siRNA #1 (UTR): 0.318 +/- 0.085, TPX2 siRNA #2 (Cds): 0.289 +/- 0.115. For the relative expression levels of TPX2, control siRNA: 1 +/- 0, TPX2 siRNA #1: 0.335 +/- 0.0074, TPX2 siRNA#2: 0.304 +/- 0.091. n=4 (from 4 independent experiments). Unpaired Student’s t-test indicated all the results are significant. *** indicates p<0.001.

3.2.1.2 Immunoprecipitation (IP) experiments with GFP-TPX2 WT and T72A transfection in

HeLa cells and -PPase treatment show that the pThr-72 antibodies are specific in western blot to the immunoprecipitated TPX2 phosphorylated at Thr72 in vivo.

In order to verify the specificity of the pThr-72 antibodies in vivo, IP experiments on

HeLa cells transfected with GFP-TPX2 WT or T72A mutant were performed to show the antibodies only recognize the immunoprecipitated phosphorylated TPX2 WT sample, not the

T72A mutant sample. In detail, HeLa cells were transiently transfected with an empty GFP vector alone, GFP-TPX2 WT and GFP-TPX2 T72A. 24 hours after transfection, cells were synchronized at M phase by nocodazole treatment (100 ng/ml) for 16 hours, and then cells were harvested and lysed with lysis buffer. Both endogenous TPX2 and exogenous TPX2 (GFP-

TPX2) were immunoprecipitated with TPX2 antibodies (NB500-183, Novus biologicals,

Immunogen: 150-200 a.a. region of human TPX2). Each IP sample was divided into two tubes.

IP beads from one set of tubes were treated with -PPase to dephosphorylate TPX2 before SDS-

PAGE. The TPX2 band should then no longer being recognized by the phospho-antibodies. IP beads from the other set of tubes were not treated with -PPase, and thus the phospho-TPX2 band should still appear on the blot. The blot was first probed with pThr-72 antibodies. After stripping, the same blot was re-probed with TPX2 (183) antibodies to detect both the endogenous and exogenous/transfected TPX2. The result shows that pThr-72 antibodies only recognize the immunoprecipitated TPX2 phosphorylated on Thr72 in GFP-TPX2 WT-transfected cells, but not in GFP-TPX2 T72A mutant-transfected cells. However, some apparently non-specific bands appear in the total lysate of pThr72-TPX2 blot. These bands do not appear after IP. Also, in the

IP blot probed with pThr-72 and TPX2 antibodies, the IP levels of endogenous TPX2 in un- transfected and empty GFP vector-transfected HeLa samples are higher than those in GFP-TPX2 62

WT or GFP-TPX2 T72A-transfected HeLa cells. It is possible that IP levels of endogenous

TPX2 in WT and T72A are reduced in the presence of the exogenous TPX2.

Figure 3.2 The pThr-72 antibodies are specific in western blot to the immunoprecipitated TPX2 phosphorylated at Thr72 in vivo.

HeLa cells were transfected with an empty GFP vector (GFP, lanes 4 and 5), GFP-TPX2 WT (lanes 6 and 7) or GFP-TPX2 T72A mutant plasmids (lanes 8 and 9). Also, non-transfected HeLa cells were used to see the expression of endogenous TPX2 (lanes 2 and 3). 24 hours after transfection, cells were synchronized with nocodazole for 16 hours, harvested and lysed. TPX2 immunoprecipitation was performed in each sample with TPX2 antibodies (NB500-183, Novus

biological). Where indicated (lanes 3, 5, 7 and 9), IP beads were treated with -PPase before SDS-PAGE. The blot was first probed with the pThr-72 antibodies (pThr-72). After stripping, the same blot was re-probed with TPX2 (183) antibodies. Lanes 10-13 show the total lysate input used for each IP sample. The bands with the red asterisk (*) sign show non-specific bands on western blots. Positions of molecular weight markers are shown in the middle of IP and total lysate blots (in kDa)

3.2.2 In vivo TPX2 phosphorylation at Thr72 is cell cycle-dependent.

To further characterize the phosphorylation of TPX2 at Thr72 in vivo, we next asked whether Thr72 phosphorylation is cell cycle-dependent. Since the expression of TPX2 is cell cycle-dependent, peaking at the M phase of the cell cycle [23], we predicted that the phosphorylation of TPX2 at Thr72 would follow a similar pattern. HeLa cells were synchronized at S phase by the double thymidine block and at M phase by nocodazole block (See section 2.6 for the detailed protocol). TPX2 was immunoprecipitated with total TPX2 (184) antibodies from

2 mg of total protein lysates derived from cells synchronized at S and M phase, as well as non- synchronized cells. After SDS-PAGE, followed by western blots with pThr-72 and TPX2 (184) antibodies, the relative expression levels of TPX2 phosphorylation at Thr72 in each cell cycle were quantified using the Quantity One program. As shown in Figure 3.3, in vivo phosphorylation of TPX2 at Thr72 is indeed cell cycle-dependent. The levels of phosphorylation are much higher in M phase cells compared to S phase or non-synchronized cells. When the relative levels of pThr72-TPX2/TPX2 expression in non-synchronized cells were set as 1, the relative levels of pThr72-TPX2/TPX2 expression in M phase cells were ~4 fold higher than those in non-synchronized cells and 3.1 fold higher than those in S phase cells (Figure 3.3C). In total, these results suggest that TPX2 is phosphorylated at Thr72 in vivo in a cell cycle-dependent manner and this phosphorylation peaks at the M phase of the cell cycle.

Figure 3.3 In vivo TPX2 phosphorylation at Thr72 is cell cycle-dependent. HeLa cells were synchronized at the S and M phases of the cell cycle. Cells were lysed and TPX2 immunoprecipitation was performed using cell lysates of non-synchronized, S or M phase- samples with 2 mg of total protein. A) Cell cycle profiles to confirm cell synchronization at each phase by flow cytometry analysis. B) After IPs, western blots were probed first with the pThr-72 antibodies and then, after stripping, re-probed with the TPX2 (184) antibodies. The levels of - actin were used as loading controls. The levels of cyclin B1 were used as positive controls to show that synchronization at M phase worked well, as indicated by the high level of cyclin B1 in M phase compared to that of S phase or non-synchronized cells. The western blot figures are representative of 3 independent experiments. C) The bar graphs for quantification of relative

levels of Thr72 phosphorylation are shown. Non-syn: Non-synchronized cells (1+/- 0), M: M phase cells (4.16 +/- 0.36), S: S phase cells (1.31 +/- 0.4); Group (Mean of the relative levels of pThr72-TPX2/TPX2 expression +/- standard deviation (STDEV)). n=3, the graph indicates the mean value of 3 independent experiments. The error bars indicate +/- STDEV. Statistical analysis from unpaired Student’s t-test, ***: Non-syn vs. M phase, P<0.001, **: M vs. S phase, P<0.01, Not significant: Non-syn vs. S phase.

3.2.3 In vivo mass spectrometry analysis confirmed that TPX2 is phosphorylated at Thr72 in mitotic HeLa cells

Previous data from the Nguyen laboratory shows that TPX2 is phosphorylated at Thr72 by Cdk1/cyclin B and Cdk2/cyclin A in vitro (Figure 1.9). Furthermore, I found this phosphorylation is cell cycle-dependent, peaking at mitosis. Also, Thr72 of TPX2 is frequently identified as a phosphorylation site in the large-scale databases from several proteomic screenings using mitotic HeLa cell lysate or purified mitotic spindle (See introduction; section

1.4.3 for details). Based on these, I hypothesized that this phosphorylation site is one of the important sites for the functions of TPX2 during mitosis. Therefore, using our phospho-specific

TPX2 antibodies against Thr72, I decided to validate and characterize the phosphorylation of

TPX2 at Thr72 in vivo using mitotic HeLa cells.Validation of this phosphorylation site using phospho-specific TPX2 antibodies has never been performed by any other groups.

In order to confirm that TPX2 is phosphorylated at Thr72 in vivo, I performed mass spectrometry analysis on the endogenous TPX2 immunoprecipitated sample from nocodazole synchronized-mitotic HeLa cells using TPX2 antibodies. Figure 3.4A shows a schematic diagram of sample preparation procedures for mass spectrometry analysis. In detail, exponentially growing HeLa cells were synchronized at M phase by nocodazole block and were harvested. After cell lysis, endogenous TPX2 was immunoprecipitated using TPX2 (184) antibodies and phosphopeptides were analyzed to identify phosphorylation sites of TPX2 by mass spectrometry as described in the section 2.10. As shown in Figure 3.4, Thr72 was identified as one of the in vivo phosphorylation sites of TPX2.

The results from mass spectrometry analysis clearly show that TPX2 is phosphorylated at

Thr72 in mitotic HeLa cells (Figure 3.4B, C and Appendix Figure 2). This result was consistent

with the data from several proteomic screenings showing that Thr72 is one of the phosphorylation sites of TPX2 phosphorylated in mitotic HeLa cells [109, 111, 130, 137]. More importantly, in my mass spectrometry analysis data, I confirmed that not only is TPX2 phosphorylated at Thr72, phosphorylation also occurs at several other sites of the TPX2 protein from mitotic HeLa cells; total 16 phosphorylation sites including Thr72 were identified (Figure

3.4C). Among these 16 phosphorylation sites, 15 sites have been already identified from other large-scale proteomic screens, but one site, Serine 102 is a novel phosphorylation site, which never been reported in other literature or proteomic screenings. Sequence coverage of mass spectrometry analysis was 44 % (Appendix Figure 1). Appendix Figure 2 shows the multiple sequence alignment of TPX2 protein orthologs from different species (human, mouse, rat and frog) based on sequence identity and similarity. Appendix Figure 2 also indicates the location of total 41 in vivo phosphorylation sites of TPX2 identified from the Phosphosite database and 16 phosphorylation sites identified from my own mass spectrometry analysis. Appendix Table 1 shows the summary of in vivo TPX2 phosphorylation sites along with some related information such as matching kinase motifs and references for each site. In summary, our result shows that

TPX2 is phosphorylated at Thr72 in mitotic HeLa cells. In addition, we identified several other in vivo phosphorylation sites of TPX2 including one novel site, Ser102. This result confirms all the previous studies that TPX2 is phosphorylated at Thr72 in vivo.

Figure 3.4 TPX2 is phosphorylated at Thr72 in mitotic HeLa cells as detected by mass spectrometry analysis

(A) Schematic diagram of the experimental protocol for mass spectrometry analysis (LC- MS/MS) using mitotic HeLa cells. HeLa cells were synchronized at M phase by nocodazole treatment (100 ng/ml) for 16 hours and released for 30 minutes after nocodazole washout. Endogenous TPX2 was immunoprecipitated from 10 mg of total protein using TPX2 (184) antibodies. IP sample was run on SDS-PAGE and after Coomassie blue staining. The band that size matched to TPX2 (confirmed by western blotting) was cut out and sent for LC-MS/MS analysis. (B) The spectra of the phosphopeptide containing Thr72. The red asterisk signs on the spectra indicate the identified matched fragment ions on mass spectrometry. (C) The in vivo phosphorylation sites of human TPX2 identified from my own mass spectrometry analysis. T: Threonine, S: Serine. The site written in red bold (S102) shows the phosphorylation site that has never been previously reported. Blue, red and yellow boxes represent three main domains of TPX2 already described in Figure 1.2, indicating the location of each phosphorylation site in the context of the functional domains of TPX2.

3.2.4 Intracellular localization and expression of the phosphorylated Thr72-TPX2 in HeLa cells

Protein phosphorylation at threonine, serine or tyrosine residues can possibly rearrange the protein structure, which can lead to changes in subcellular localization of the protein or complex formation with interacting partners [138]. Therefore, we sought to determine whether

TPX2 phosphorylated at Thr72 might change its intracellular localization in HeLa cells and whether the localization and expression of TPX2 phosphorylation at Thr72 is cell cycle- dependent in immunofluorescent analysis, consistent with western blot results. Before we studied the localization of pThr72-TPX2 by immunostaining of HeLa cells, the specificity of the pThr-72 antibodies was validated using antigen preabsorption with the corresponding blocking peptide

(LQQAI VpT72PLKPVD, 66 to 78 a.a. regions of TPX2 used for antigen raising the antibodies).

As shown in Figure 3.5, the pThr-72 signal of immunostaining both in mitosis and interphase of HeLa cells was significantly decreased or absent after the antibodies were preincubated in pThr72-TPX2 blocking peptides. Furthermore, preincubation with pThr-72 and

TPX2 antibodies with blocking peptides only blocked the signal of pThr-72. From these data, we validated the pThr-72 antibodies as specific to TPX2 phosphorylated at Thr72 in immunostaining. Immunostaining with the pThr-72 antibodies in HeLa cells revealed that in interphase, TPX2 phosphorylated at Thr72 is localized in the nucleus (Figure 3.5B), which is the same localization of TPX2 in interphase. The expression levels of pThr-72 in interphase were much lower than those in mitosis. Conversely, during mitosis, the TPX2 phosphorylated at Thr72 is not strictly associated with the mitotic spindle, and is localized on a layer around mitotic spindle or cytoplasm (Figure 3.5A and C). The expression levels of pThr-72 were strong and much higher than those in interphase, in agreement with the western blot results.

Figure 3.5 Intracellular localization and expression of TPX2 phosphorylated at Thr72 in HeLa cells and the specificity of pThr-72 antibodies in immunostaining A) and B) HeLa cells were stained with the antibodies alone [pThr-72 (FITC, green) and tubulin (Cy3, red) antibodies] or with the antibodies preincubated with the corresponding blocking peptides. A) In mitosis, TPX2 phosphorylated at Thr72 is localized around mitotic spindle and cytoplasm, not strictly associated with the mitotic spindle. The expression levels of pThr-72 in mitosis are much higher than those in interphase. The specificity of pThr-72 antibodies in immunostaining was verified by antigen preabsorption with the pThr-72 blocking peptides in several different ratios (from 1:5 to 1:50). Only cells from the 1:20 and 1:50 ratios are shown. In each ratio of blocking peptide-antibody preincubation, the pThr-72 signal was decreased or absent by antigen preabsorption with the phosphorylated blocking peptides. B) In interphase, pThr-72 is localized in the nucleus. The expression levels of pThr-72 in interphase are lower than those in mitosis. The pThr-72 signal was significantly reduced when the antibodies were preincubated with the different ratios of the blocking peptides. C) HeLa cells were stained with the antibodies alone [pThr-72 antibodies (FITC, green) and TPX2 (18D5-1) antibodies (Cy3, red)] or with the antibodies preincubated with the corresponding blocking peptides. When antibodies were preincubated with the pThr-72 blocking peptides, only the pThr- 72 signal was blocked, not the TPX2 signal.

3.2.5. TPX2 is phosphorylated in vivo at Thr72 by Cyclin-dependent kinases (Cdks)

Our previous results showed that TPX2 is phosphorylated at Thr72 by Cdk1 and Cdk2 in vitro (Figure 1.9) and Thr72 phosphorylation site of TPX2 was identified as a Cdk2 substrate from one of mass spectrometry screenings using HEK cell lysates [110] (See 1.4.3). Therefore, we sought to determine whether TPX2 is phosphorylated at Thr72 also in vivo by Cdk1 and

Cdk2. To test this, we used a pharmacological approach; we treated HeLa cells with two Cdk inhibitors, roscovitine and alsterpaullone. Roscovitine is a well-known inhibitor of the kinase activity of Cdk1/cyclin B, Cdk2/cyclin A and Cdk5/p25, as a purine analog that competes for the

ATP binding sites in the catalytic domain of Cdks, thereby preventing the phosphate transfer to

Cdk substrates [139-142]. Alsterpaullone is a specific inhibitor for the kinase activity of

Cdk1/cyclin B and is also an ATP competitive inhibitor [143]. Since Thr72 phosphorylation is the highest in M phase of cell cycle, we performed the roscovitine treatment on mitotic HeLa cells. HeLa cells were synchronized with nocodazole (100 ng/ml) at M phase, so that we would be able to see any changes in levels of phosphorylation of TPX2 at Thr72 during M phase by inhibitor treatment. Nocodazole-arrested mitotic cells were then treated with two different concentrations of roscovitine (20 M and 40 M) for 30 minutes in the presence of nocodazole to avoid premature exit from mitosis [144]. There are several reports that Cdk1 inhibition in cells by some Cdk inhibitors, including roscovitine, can cause cells to rapidly exit from mitosis, which might lead to unspecific results [144]. As shown in Figure 3.6, when mitotic HeLa cells were treated with DMSO as a vehicle control, or 20 M and 40 M of roscovitine, the levels of Thr72 phosphorylation in mitotic cells treated with roscovitine were significantly reduced compared to those in mitotic cells without treatment. We used a phospho-MAPK/Cdk substrate antibody (p-

MAPK/Cdk) as a positive control for Cdk inhibitor treatments. Mitogen-activated protein kinase

(MAPK) and Cyclin-dependent kinase (Cdk) are known to phosphorylate a large family of substrates in M phase [145]. This p-MAPK/Cdk antibody recognizes proteins containing

PX(S*/T*) P or (S*/T*) PXR/K motif (S*: a phospho-serine T*: a phospho-threonine), thereby, the levels of p-MAPK/Cdk substrates should be decreased upon Cdk inhibitor treatments as a positive control. There were several bands on the western blot that appeared from mitotic cells treated with only DMSO control and the intensities of those bands were decreased in the same total cell lysates with 20, 40 M of roscovitine treatments. These results indicate the presence of several substrates of p-MAPK/Cdk and show that Cdk inhibitor treatments were effective.

Furthermore, there were no significant changes in the levels of cyclin B1, indicating that most of the cells treated with roscovitine remained in mitosis. As mentioned, roscovitine is also an inhibitor of Cdk5/p25. However, there is no active Cdk5 in HeLa cells and the result from the in vitro kinase assay performed by Dr. Wang (Figure 1.9) clearly shows that pThr72-TPX2 is not a substrate of Cdk5/p25. Therefore, we were not concerned about the inhibition effects of roscovitine on Cdk5 in our experiments. In summary, our results demonstrate that Cdk1 or Cdk2 phosphorylates TPX2 at Thr72 in vivo during mitosis.

Figure 3.6 Phosphorylation of TPX2 at Thr72 is inhibited by the Cdk inhibitor roscovitine. The levels of Thr72 phosphorylation were reduced by the roscovitine treatment. HeLa cells were treated with 100 ng/ml of nocodazole for 16 hours and mitotic HeLa cells were treated with DMSO (as a control), or 20 M and 40 M of roscovitine for 30 minutes. Cells were harvested and lysed in lysis buffer. TPX2 was immunoprecipitated with the total TPX2 antibodies from cell lysates of each sample with 1 mg of total proteins. SDS-PAGE was performed and followed by western blotting with pThr-72 and total TPX2 antibodies. The levels of actin were used as loading controls. The levels of cyclin B1 were also used as controls to check that most of cells treated with roscovitine remained in mitosis. The levels of p-MAPK/Cdk substrates [PX(S*/T*)P or (S*/T*)PXR/K motif, S*;phospho-serine, T*: phospho-threonine] were also used as positive controls to confirm that Cdk inhibitor treatment in this experiment worked properly. The bar

graph shows the quantification of the relative levels of pThr-72/TPX2 IP samples. The relative levels of pThr-72/TPX2 expression in DMSO-treated control cells were set as 1.

Figure 3.7 Phosphorylation of TPX2 at Thr72 is inhibited by treatment with the Cdk1- specific inhibitor alsterpaullone. The levels of Thr72 phosphorylation was reduced by the Cdk1-specific inhibitor, alsterpaullone treatment. HeLa cells were treated with 100 ng/ml of nocodazole for 16 hours and these nocodazole-arrested mitotic HeLa cells were treated with DMSO (as a vehicle control), or 20 M and 40 M of alsterpaullone for 30 minutes. Cells were harvested and lysed in lysis buffer. TPX2 was immunoprecipitated with TPX2 antibodies from cell lysates of each sample with 1 mg of total proteins. SDS-PAGE was performed and followed by western blotting with pThr-72 and total TPX2 (184) antibodies. The bar graph shows the quantification of the relative levels of pThr-72/TPX2 IP samples. The relative levels of pThr-72/TPX2 expression in DMSO-treated

control cells were set as 1. The levels of actin in the total lysates were used as loading controls. The levels of cyclinB1 were used as comparators to see whether Cdk1 inhibitor treatment caused any change in the cell cycle status of mitotic cells. The levels of p-MAPK/Cdk substrates in the same total cell lysates were also used as positive controls to confirm that alsterpaullone treatment in this experiment worked properly.

As shown in Figure 3.7, we also treated mitotic HeLa cells with alsterpaullone, a Cdk1- specific inhibitor. Consistent with the roscovitine treatment, nocodazole-arrested HeLa cells were treated with two different concentrations of alsterpaullone, 20 M and 40 M, for 30 minutes, which significantly reduced the levels of Thr72 phosphorylation as well as the levels of

MAPK/Cdk substrate phosphorylation (which were used as positive controls for Cdk inhibition).

Unlike the roscovitine treatment, the levels of cyclin B1 were decreased after alsterpaullone treatment, which may imply Cdk inhibition by alsterpaullone caused cells to exit from mitosis.

This conflicts with the results from another group showing alsterpaullone treatment up to 100

M usually does not cause decreased levels of cyclin B1 and does not lead to premature exit from mitosis [140]. However, even though the levels of cyclin B1 in 20 M and 40 M-treated cells are almost similar, the levels of pThr-72 in 40 M decreased to around 40%, when compared to the levels of pThr-72 in 20 M-treated cells in the IP samples. Furthermore, as already stated in the introduction, the levels of TPX2 sharply decrease right after mitotic exit and entry of G1 [23]. But, as shown in Figure 3.7, the levels of TPX2 in cells treated with 20 M and

40 M of alsterpaullone are similar to the levels of TPX2 in cells without treatment. Therefore, we speculate that a decrease in the levels of pThr-72 in alsterpaullone treated cells is not a non- specific result from cells rapidly exiting from mitosis. It will be necessary to do this experiment one more time with alsterpaullone treatment for a shorter period of time, such as 15 minutes to see the exact effect of Cdk inhibition on the levels of pThr-72.

Taken together, our results suggest that Cdk1/cyclin B phosphorylates TPX2 at Thr72 in vivo during mitosis. Considering that Cdk2/cyclin A is a kinase responsible for S phase phosphorylation, and we have not performed our experiments using cells in S phase or used a

Cdk2-specific inhibitor, we cannot conclude whether TPX2 is phosphorylated by Cdk2/cyclin A at Thr72 in vivo. The kinase responsible for S phase-TPX2 phosphorylation remains to be determined.

3.3 Discussion

3.3.1 Summary

The aim of the present study was to characterize TPX2 phosphorylation at one particular site, Threonine 72 (Thr72) in HeLa cells. In this section, we verified that TPX2 is phosphorylated at Thr72 in mitotic HeLa cells. We previously generated the antibodies against pThr72-TPX2 (pThr-72 antibodies) which are specific in vitro, and here we confirmed the specificity of these antibodies to pThr-72 in vivo. First, blocking peptide experiments were performed in western blotting of IP samples using GFP-TPX2 WT or the T72A mutant, and also in immunostaining of HeLa cells (Figure 3.2 and 3.4). With these phospho-specific antibodies, we found that Thr72 phosphorylation is cell cycle-dependent, peaking at the M phase of the cell cycle (Figure 3.3). Based on this information, we performed mass spectrometry analysis on endogenous TPX2 immunoprecipitated from nocodazole-synchronized mitotic HeLa cells. The mass spectrometry analysis shows that TPX2 is indeed phosphorylated at Thr72 (Figure 3.4). In agreement with our previous data showing TPX2 is phosphorylated by Cdk1 and Cdk2 in vitro, two different Cdk inhibitors, roscovitine and alsterpaullone significantly inhibited phosphorylation of TPX2 at Thr72, suggesting Cdks mediate this phosphorylation in vivo.

3.3.2. Critiques and interpretations of results.

In Figure 3.3, we showed that TPX2 phosphorylation at Thr72 is cell cycle-dependent. In the introduction of my thesis, it was mentioned that the levels of TPX2 gradually increase in S phase through G2 phase and peak at mitosis; the levels of TPX2 in M phase are approximately two fold higher than those in S phase [23]. However, when we performed IP experiments with the TPX2 antibodies, using cell lysates from different cell cycle stages, M, S phase and non- synchronized cells, the levels of TPX2 in M phase and S phase were similar (Figure 3.3). This may be due to the saturated amount of proteins immunoprecipitated by the TPX2 antibodies. If the levels of total TPX2 in the IP sample of M phase were two fold higher than those in S phase, as previously reported, the relative levels of pThr-72 in M phase would also be much higher, as compared to those in S phase or non-synchronized cells.

In Figure 3.4, we showed that TPX2 is phosphorylated at Thr72 in mitotic HeLa cells, by using mass spectrometry analysis. In Figure 3.4A, I performed an IP, ran it on SDS-PAGE and after coomassie blue staining, many bands were observed. Selected bands were cut out based on size, after confirmation by western blotting with TPX2 and pThr72 antibodies. In this experiment, an IP with a control IgG antibody should have been used as a comparator for the bands in the TPX2 IP, to show specificity.

In Figure 3.5, we showed that when TPX2 is phosphorylated at Thr72 in mitosis, the

TPX2 protein surprisingly changes its subcellular localization from the mitotic spindle to a layer around the mitotic spindle or cytoplasm. At first, we suspected the specificity of the pThr-72 antibodies. However, using blocking peptide, we proved these antibodies are specific to pThr-72; the blocking peptide could abrogate the strong signal of pThr-72 in mitotic cells. When HeLa cells were co-stained for pThr-72 and a chromosome periphery protein, nucleolin, we could see

co-localization of pThr72 with nucleolin only in mitotic cells (Appendix Figure 3). Nucleolin is the major protein localized in the nucleoli during interphase, and at the chromosome periphery during mitosis [146]. The chromosome periphery is a layer covering the chromosomes, which is distinct from the chromosome and the cytoplasm. Whether pThr-72 is indeed located in the chromosome periphery or not requires further confirmation. However, it is still quite interesting that pThr72 is not strictly associated with mitotic spindles. Clarification of the exact localization of TPX2 phosphorylated at Thr72 remains to be determined.

In Figure 3.6 and 3.7, we showed that treatment of mitotic HeLa cells with two Cdk inhibitors, roscovitine and alsterpaullone, clearly showed that TPX2 phosphorylation at Thr72 is dependent on Cdk in vivo. However, with alsterpaullone treatment, we could see the levels of pThr72 were significantly decreased at two different concentrations of the drug, however, the levels of cyclin B1 in the same samples were also decreased. Inhibition of Cdk1 by several inhibitors such as roscovitine, flavopiridol and BMI-1026 are known to cause premature mitotic exit in mitotic cells [147-149]. It has been shown that Cdk1 inhibitors induced cytokinesis, therefore resulting in cyclin B1 degradation by APC/C-mediated ubiquitination [144]. When cells were treated with a proteasome inhibitor, MG132, cyclin B1 degradation was inhibited

[144]. However, there is no study clearly demonstrating that alsterpaullone treatment causes this effect in mitotic cells. Alsterpaullone, as a purine analog, has the same mechanism for inhibition of the activity of Cdk1 as the other Cdk1 inhibitors mentioned above (roscovitine, flavopiridol and BMI-1026); it competitively inhibits ATP binding [150]. Therefore, it should induce a similar effect. Hong et al. used alsterpaullone to test the phosphorylation of a Cdk1 substrate in their experiment [151]. To prevent premature mitotic exit caused by alsterpaullone treatment, they only treated cells with alsterpaullone for 15 minutes in the presence of nocodazole in

mitotic cells after nocodazole arrest. Their results showed that the levels of Cdk1/cyclin B1- dependent phosphorylation were decreased, but the levels of cyclin B1 remained the same as the levels in non-treated cells [151]. Therefore, it would be worth repeating our experiment with a shorter time for alsterpaullone treatment in order to see the change in the levels of cyclin B1.

Furthermore, we can try to perform a time course experiments to see how the levels of cyclin B1 change depending on different time of alsterpaullone treatment.

CHAPTER FOUR:

DETERMINATION OF THE ROLE OF THREONINE 72 PHOSPHORYLATION ON

TPX2 MITOTIC FUNCTION

4.1 Rationale

In the previous section of results, we suggested that TPX2 phosphorylation at Thr72 is dependent on Cdk1/cyclin B during mitosis (Figure 3.6 and 3.7). It is known that the activity of the Cdk1/cyclin B complex leads cells into mitosis by phosphorylating many substrates involved in several processes during mitosis such as chromosome condensation, mitotic spindle assembly and nuclear envelope breakdown [152]. In addition, many MT regulators have been identified as

Cdk1/cyclin B substrates [153]. Also, it has been suggested that Cdk1/cyclin B mediated- phosphorylation of particular substrates regulates and maintains mitotic spindle assembly [151].

As mentioned in the introduction, TPX2 is an essential factor for mitotic spindle assembly and

TPX2 phosphorylation at Thr72 occurs mainly in mitosis, which suggests that phosphorylation at

Thr72 might be an important phosphorylation site for the functions of TPX2 in mitotic spindle assembly during mitosis. TPX2 plays multiple roles in mitotic spindle assembly by regulating

MT nucleation, spindle pole organization, Aurora A activation and the activity of the mitotic motor protein, Eg5 [8, 18, 22]. Gain or loss of TPX2 proteins can cause defects in mitotic spindle assembly such as abnormal spindles and failure of spindle formation. When human TPX2 is depleted from HeLa cells, it arrests cells in mitosis with multipolar spindles [40]. Also, when

Gruss et al. depleted hTPX2 using TPX2 siRNA, it causes chromosome condensation, two MT asters to be separated from each other and spindle formation failure, which leads to mitotic block

[22]. Conversely, overexpression of TPX2 using GFP-hTPX2 causes cells to be in a

prometaphase state with abnormal spindles and results in failure of MT organization [22]. Even though the function of TPX2 in mitotic spindle assembly is well characterized, there is little information about which putative phosphorylation sites of TPX2 are actually phosphorylated, except for a few sites related to Aurora A kinase activation [113, 154]. Therefore, we hypothesized that the function of TPX2 in spindle assembly is likely to be regulated by Thr72 phosphorylation, as it occurs specifically during mitosis.

4.2 Results

4.2.1 Effects of GFP-TPX2 WT and T72A mutant on mitotic spindle assembly

To investigate the functional significance of Thr72 phosphorylation on mitotic spindle assembly, we examined the effects of GFP-TPX2 WT or the phospho-dead mutant of TPX2

(T72A) on the morphology of the mitotic spindle in HeLa cells in two different settings; 1) When

GFP-TPX2 or GFP-TPX2 T72A mutant is overexpressed in the presence of endogenous TPX2,

2) GFP-TPX2 or GFP-TPX2 T72A mutant is overexpressed in the absence of endogenous TPX2 knocked down by TPX2 UTR siRNA.

4.2.1.1 Overexpression of GFP-TPX2 T72A mutant increased the percentage of cells with multipolar spindles.

Some of the previous studies showed that GFP-TPX2 overexpression can block some portions of cells with monopolar spindles, while at the same time, some cells (36.5%) expressing

GFP-TPX2 undergo the normal cell cycle without any problem [22]. In addition, the localization pattern of the transfected GFP-TPX2 in cells is very similar as that of endogenous TPX2 in mitosis (in the mitotic spindle). Therefore, we tested whether GFP-TPX2 T72A transfection caused any change in comparison to GFP-TPX2 WT transfection, in terms of morphology of the

mitotic spindle, which could indicate the functions of Thr72 phosphorylation in mitotic spindle assembly. In order to see the effect of GFP-TPX2 WT and T72A mutant overexpression, HeLa cells were transiently transfected with an empty GFP vector only, GFP-TPX2 WT or GFP-TPX2

T72A mutant without endogenous TPX2 depletion (Figure 4.1B). Cells were synchronized at the

G2/M phase with nocodazole treatment for 16 hours, and after nocodazole washout with PBS wash, cells were released in fresh DMEM for 30 minutes to allow the accumulation of mitotic cells. Cells were fixed with 4% PFA and stained with Cy3-conjugated-tubulin for mitotic spindle visualization. Cells in prometaphase or metaphase with mitotic spindles in each sample were categorized into three different classes based on their morphological differences; monopolar, bipolar and multipolar spindle (Figure.4.1A). Then, the percentage of cells in each category on the indicated group (GFP, GFP-TPX2 WT and GFP-TPX2 T72A) were calculated and compared. Interestingly, we found that transfection of the T72A mutant into HeLa cells significantly increased the frequency of mitotic cells with multipolar spindles, compared to cells transfected with GFP-TPX2 WT (Figure 4.1C). We made sure that the levels of GFP-TPX2 WT and T72A mutant expression were similar so that the results would be comparable (Figure 4.1B).

ANOVA comparing the percentage of cells with multipolar spindles in three groups, GFP vector only, GFP-TPX2 WT and T72A mutant showed the results were significant (P value was

P<0.001). To compare the percentage of cells with multipolar spindles in each group, we ran the

Neuman-Keuls test. GFP-TPX2 WT vs. T72A: P<0.001, GFP vs. T72A: P<0.001. GFP vs. WT:

NS (Not significant). The percentage of cells with multipolar spindles in the T72A mutant was

9%, higher than the percentage of cells in GFP-TPX2 WT, which was 12 %, both higher than the percentage of cells in GFP vector only. The other P values from ANOVA, comparing each group with different numbers of spindle poles, are shown in Appendix Table 2. To summarize,

overexpression of the T72A mutant in the presence of endogenous TPX2 significantly increased the frequency of mitotic cells with multipolar spindles compared to that of cells with GFP-TPX2

WT overexpression.

Figure 4.1 Effects of GFP-TPX2 WT and T72A mutant overexpression in HeLa cells on the numbers of mitotic spindles. To examine the effect of GFP-TPX2 WT and the mutant on spindle assembly, HeLa cells were transiently transfected with GFP vector only, GFP-TPX2 WT or T72A mutant without endogenous TPX2 depletion and synchronized at M phase to enrich for cells with mitotic spindles. Cells in one set were harvested for western blotting and cells in the other set were fixed and stained with Cy3-conjugated tubulin for MT/spindle pole visualization. Cells with mitotic spindles in prometaphase and metaphase were observed and categorized based on the morphology of the mitotic spindles. A) Representative figures of mitotic HeLa cells with different numbers of mitotic spindle poles after GFP-TPX2/T72A transfection. B) Western blots showing the levels of GFP-TPX2 WT and GFP-TPX2 T72A. Also, the levels of both endogenous TPX2 and overexpressed GFP-TPX2/T72A in this setting are shown. C) Bar graphs showing the number of cells with different morphology of mitotic spindles in each group. n= at least 100 cells for each set of experiments; these are from 5 independent experiments for WT and T72A and from 2 independent experiments for GFP transfection. A phospho-dead mutant, T72A transfection results in significant increase in the percentage of cells with multipolar spindles. ANOVA comparing three groups in multipolar spindle shows P <0.001. Neuman-Keuls test for each group: *** indicates P<0.001 (multipolar, GFP vs. T72A, TPX2 WT vs. T72A). NS indicates not significant (GFP vs. TPX2). Red (Cy3) indicates tubulin, Green (FITC) indicates GFP signal of TPX2 or TPX2 mutant and Blue (DAP) indicates DNA staining.

4.2.1.2 Expression of GFP-TPX2 T72A mutant with endogenous TPX2 knockdown also increases the percentage of cells with multipolar spindles

Next, in order to examine more directly the effect of GFP-TPX2 T72A, a phospho-dead mutant, on the formation of mitotic spindles, we used TPX2 siRNA to knock down endogenous

TPX2 in HeLa cells. The sequences of GFP-TPX2 contain the coding sequences targeted by the

TPX2 cds siRNA, but not the TPX2 UTR siRNAs, therefore, knocking down TPX2 by TPX2

UTR siRNA did not have an effect on expression of GFP-TPX2 or GFP-TPX2 T72A (See 2.4 for detailed information on TPX2 siRNA). We already have confirmed that TPX2 UTR siRNA does not have any off-target effects. In detail, HeLa cells were transiently co- transfected with

TPX2 UTR siRNA and one of the GFP constructs (GFP vector only, GFP-TPX2 wild-type or

T72A mutant). For transfection of GFP-TPX2 and T72A, the amounts of DNA were calculated to get the same levels of GFP-TPX2 or T72A as expected for the endogenous TPX2 levels in control HeLa cells without TPX2 knockdown. By doing that, we minimized the unwanted effects of overexpression of the GFP constructs. 24 hours after transfection, cells were synchronized with nocodazole block for 16 hours, then released in fresh DMEM without nocodazole, to allow accumulation of cells with mitotic spindles. Mitotic cells were fixed and stained with Cy3- conjugated tubulin. Cells in prometaphase or metaphase with mitotic spindles were categorized into three different classes based on their morphological differences; monopolar, bipolar or multipolar spindles (as shown in Figure 4.1A).

Figure 4.2 Effects of GFP-TPX2 WT and T72A transfection in HeLa cells lacking endogenous TPX2 on the numbers of mitotic spindles HeLa cells were transiently co-transfected with an empty GFP vector, GFP-TPX2 WT and GFP- TPX2 T72A mutant together with TPX2 siRNA targeting the 3’UTR of TPX2 mRNA. As a control, HeLa cells were transfected with an empty GFP vector and control siRNA. Cells were fixed and stained with Cy3-conjugated tubulin for MT visualization. A) Western blot shows the levels of GFP-TPX2 WT/T72A expression were almost similar as endogenous TPX2 in HeLa cells transfected with control siRNA and the level of GFP-TPX2 expression was almost similar as the level of GFP-T72A expression. B) Bar graphs showing the number of cells with different morphology of mitotic spindles in each group. Note that GFP transfection with endogenous TPX2 knockdown already resulted in a significant increase in multipolar spindles (5.4 % increases from control without TPX2 depletion). Compared to cells transfected with an empty GFP vector and GFP-TPX2 WT, the percentage of cells transfected with GFP-TPX2 T72A had an even greater increase in the percentage of cells with multipolar spindles (9.8% increases from an empty GFP vector with TPX2 depletion, 7.5 % increases from GFP-TPX2 WT with TPX2 depletion). N=3, ANOVA test for multipolar spindles of 4 groups: P<0.01. Neuman-Keuls test for each group shows Control (with GFP+ control siRNA transfection) vs GFP:* P<0.05, GFP vs.TPX2 WT: NS, not significant, WT vs.T72A: P<0.05, GFP vs T72A: P<0.05. n= at least 500 cells for each set of experiments; these are from 3 independent experiments. Error bars indicate S.E.M.

The western blot in Figure 4.2A shows that the levels of GFP-TPX2 and T72A mutant expression are similar to the levels of endogenous TPX2 in control HeLa cells treated with control siRNA and an empty GFP vector. We also ensured that the levels of GFP-TPX2 WT and

T72A mutant expression were similar so that differences in the results would not be attributed to differences in protein expression. Interestingly, compared to the cells with control RNAi transfection, transfection of the GFP control vector only in cells with TPX2 knockdown increased the percentage of cells with multipolar spindles, which is consistent with other literature, showing that knockdown of human TPX2 caused cells to have more multipolar spindles [40]. Furthermore, when we compared the percentage of cells transfected with an empty

GFP vector, GFP-TPX2 WT and T72A mutant, consistent with the first experiment in the overexpression setting, T72A mutant transfection significantly increased the frequency of cells with multipolar spindles. P values from ANOVA show that these results were statistically significant. The ANOVA test compared the percentage of cells with multipolar spindles in 4 different groups; Control siRNA co-transfected cells with an empty GFP vector, TPX2 siRNA co-transfected cells with an empty GFP vector, GFP-TPX2 WT and T72A mutant had a P value of <0.01. To compare each group, the Neuman-Keuls test was performed; Control vs. GFP:

P<0.05, GFP vs. GFP-TPX2 WT: Not significant, GFP-TPX2 WT vs. T72A: P<0.05 and GFP vs. T72A: P<0.05 (n= at least 500 cells, these are the results from 3 independent sets of experiments). All of the P values are shown in Appendix Table 3. For the frequency of cells with monopolar or bipolar spindles, p value of ANOVA comparing all 4 groups shows that the results were not significant. In sum, in both settings (in the presence of endogenous TPX2 or in the depletion of endogenous TPX2), GFP-TPX2 T72A transfection into HeLa cells caused a significant increase in the percentage of cells with multipolar spindles.

Taken together, it suggests that Cdk/cyclin-dependent phosphorylation of TPX2 at Thr72 is important for preventing multipolar spindle phenotype, therefore, for the formation of bipolar spindles.

4.3 Discussion

4.3.1 Summary

In this study, we found that transfection with the mutant GFP-TPX2 T72A increased the percentage of cells with multipolar spindle compared to that of GFP-TPX2 WT transfection.

From our data, TPX2 knockdown increased cells with multipolar spindles, which is consistent with the results from other papers [11, 22]. Transfection with the mutant GFP-TPX2 T72A increased the frequency of cells with multipolar spindles even more. An increase in multipolar spindle frequency did not occur when cells were transfected with GFP-TPX2 WT. In other words, when Thr72 phosphorylation does not occur, mitotic spindle assembly is compromised

(or impaired), suggesting TPX2 phosphorylation at Thr72 is important for mitotic spindle assembly.

4.3.2 Critical interpretations in each experiment

In Figure 4.1 and 4.2, we found mutating the Thr72 phosphorylation site to Alanine caused cells to have more multipolar spindles, indicating TPX2 phosphorylation at Thr72 is important for normal spindle formation. In Figure 4.2, theoretically, if Thr72 is important for spindle formation, when endogenous TPX2 is depleted, it should cause cells to have more multipolar spindles (phenotype of abnormal spindle), which was the case in our results.

Conversely, re-expressing GFP-TPX2 in cells after depletion of endogenous TPX2 should rescue the normal phenotype; more cells should have bipolar spindle and fewer should have multipoalar

spindles. Our results show the wild-type phenotype is only partially rescued; there is only a 3% increase in the percentage of cells with bipolar spindles and a 3% decrease in multipolar spindles when cells with TPX2 siRNA knockdown were compared to cells with GFP-TPX2 WT transfection after depletion of endogenous TPX2. When we compared the percentage of cells with an empty GFP vector and TPX2 siRNA to the percentage of cells with the GFP-TPX2

T72A mutant and TPX2 siRNA, we saw a more pronounced effect: 10% increases in multipolar spindles and 5% decreases in bipolar spindles. As shown in Figure 4.2A, endogenous TPX2 knockdown was not completely achieved by TPX2 siRNA treatment. However, the result could have been more precise if there has been more complete endogenous TPX2 knockdown.

In addition, in Figure 4.2, we found that TPX2 depletion itself increased the percentage of cells with multipolar spindles when compared to cells with control siRNA transfection

(without TPX2 knockdown). This result is consistent with the previous reports showing that human TPX2 siRNA treatment in HeLa cells resulted in 40 percent of cells having multipolar spindles [40]. In this paper, together with this effect, they reported TPX2 knockdown arrested cells in mitosis; 30% mitotic index in TPX2 siRNA treated cells, 3% mitotic index in control siRNA treated cells [40]. Because we synchronized cells after transfection, it was hard to notice this effect (arrested cells in mitosis) in our experimental setting. However, it would have been possible to compare the mitotic indices in cells with GFP-TPX2 WT and mutant transfection in a non-synchronized setting in order to see the effect of a phospho-dead mutant. In addition, Garrett et al. showed that multipolar spindles caused by TPX2 siRNA treatment is not due to increased numbers of centrosomes [40], and loss of hTPX2 caused MT dependent-fragmentation of peri- centrosomal materials (like -tubulin) at multipolar spindle poles, as indicated by -tubulin staining. Also, they found this multipolar spindle formation was dependent on the mitotic kinesin

motor protein Eg5. When the activity of Eg5 was inhibited by monastrol, most of the mitotic cells had monopolar spindles. Therefore, the balance between structural support by TPX2 and motor force by Eg5 is essential for normal bipolar spindle formation. Considering this information, we could have further performed two experiments to test the effects of the T72A mutant: 1) testing if the multipolar spindle phenotype is due to the increased number of centrosomes, or spindle pole fragmentation, by staining cells with -tubulin, and 2) testing if treatment with the Eg5 inhibitor, monastrol, would change the percentage of cells with multipolar spindles caused by presence of the T72A mutant.

In Figure 4.1B, even though I stated that the levels of GFP-TPX2 WT and T72A mutant expression were similar so that the results would be comparable, the figure shows that the signals in this western blot are too saturated to accurately compare the protein level in each lane. It would have been better if I had a shorter exposure of the film, or used a digital chemiluminescence imager to get better quantification of the differences.

4.3.3. Caveats and alternate suggestions for the systems used

1) Model systems- HeLa cells

We chose HeLa cells as our model system because this cell line is commonly used for studying TPX2, mitosis and phosphorylation. However, HeLa cells are human cervical cancer cells, which are known to commonly have multipolar spindle formation [155]. To avoid the possibility that the baseline spindle multipolarity of HeLa cells may impact the results of transfection with the T72A mutant, we included a group with control siRNA treatment, which shows a low percentage of cells with multipolar spindles (2%). Therefore, the increase in

100

multipolar spindles in the T72A mutant transfected cells was not due to HeLa cells being cancer cells with common spindle multipolarity. However, it would have been better if we had also used other types of cells which are not cancer cells, such as fibroblasts, to test the effects of the GFP-

TPX2 T72A mutant, in addition to HeLa cells.

2) Transient transfection vs. stable cell lines

We used transient transfection for this section; TPX2 siRNA was co-transfected with GFP-

TPX2 WT or the T72A mutant into HeLa cells. Instead of using transient transfection for the

TPX2 siRNA, it would have been better if we could have used a stable HeLa cell line expressing a doxycycline-inducible exogenous miRNA targeting TPX2 mRNA. In this system, the miRNA equipped with a double-stranded sequence fully complementary to the target mRNA would trigger its RISC (RNA-induced silencing complex)-dependent degradation. RISC is a multiprotein complex which can incorporate one strand of siRNA or miRNA. By turning on

TPX2 RNAi originated from an exogenous miRNA targeting endogenous TPX2, RNAi- mediated gene inactivation would be induced [156]. The advantage of using a doxycycline - inducible HeLa cell line would be to get efficient TPX2 knockdown in an entire population of cells. GFP-TPX2 WT or T72A mutant on mitotic spindle formation could then have been easily assessed in the absence of endogenous TPX2. However, the reasons we chose not to use this system were: 1) In order to synchronize cells in mitosis, we had to use nocodazole and then, if we wanted to also knockdown TPX2 in the resulting mitotic cells, we would have had to use a second drug, doxycycline. We were afraid that if we used two drugs at the same time, it might have caused a secondary effect, thereby confounding interpretation of the results. 2) We wanted

101

to keep using the original GFP-TPX2 and T72A mutant plasmids, and not re-generate the mi-

RNA resistant constructs by using site-directed mutagenesis. However, if we could have generated miRNA-resistant constructs of GFP-TPX2 WT and T72A with a few nucleotides different for silent mutation of amino acids, it would have been better to see the exact effects of the WT and T72A mutant on spindle formation.

CHAPTER FIVE:

DETERMINATION OF THE ROLES OF TPX2 PHOSPHORYLATION AT

THREONINE 72 IN DNA DAMAGE RESPONSE

102

5.1 Rationale

Our laboratory previously found that TPX2 regulates the levels of -H2AX upon treatment with ionizing radiation. TPX2 partially co-localized with -H2AX, which is recruited to IRIF [3]. TPX2 knockdown resulted in a significant increase in the level of ionizing radiation- dependent γ-H2AX at the G0 and G1 phases of the cell cycle and an increase in the number of -

H2AX-positive IRIF with high intensity. On the other hand, when TPX2 was overexpressed in cells, the level of γ-H2AX was reduced after IR [3]. When rescue experiments were performed,

TPX2 depleted cells treated with IR showed an increased level of -H2AX, and reintroduction of

GFP-TPX2 expression in those cells reduced the level of -H2AX signal (Figure 5.1). In addition, the role of TPX2 in the regulation of - H2AX levels during the DDR was proven to be independent of TPX2’s mitotic functions because these observed effects on - H2AX formation were also consistently shown in post-mitotic neurons [3].

Consistent with the findings from our laboratory, there are other papers showing evidence that TPX2 phosphorylation is involved in the regulation of DDR [65]. A large proteomic study performed by Matsuoka et al. showed that TPX2 is phosphorylated by ATM or

ATR in response to IR [65]. Furthermore, there is increasing evidence that the Cdk-related pathway is involved in DDR [87, 157, 158]. It was generally believed that Cdk1 and Cdk2 are downregulated in DNA damage checkpoint signaling, however, more and more evidence shows that they play important roles in initiating checkpoint control and DNA repair [86]. For example,

Xu et al. recently reported that Cdk1 and Cdk2 mediate the phosphorylation of Checkpoint kinase 1 (Chk1), a kinase required for efficient activation and proficiency of checkpoints in response to DNA damage [159, 160]. Furthermore, it has been reported that some of the proteins

103

involved in the DNA repair pathway, such as replication protein A (RPA), RAD51, BRAC2 and

RAP80 are phosphorylated by the Cdk-cyclin complex and these proteins play an essential role in the DNA repair pathway through Cdk-dependent phosphorylation [161, 162]. Notably, receptor-associated protein 80 (RAP80), as a component of the BRCA1-A complex, plays important roles in cell cycle checkpoint activation and DNA damage repair. RAP80 is phosphorylated by Cdk1/cyclin B at Ser677 in M phase when Cdk1 is active [163]. Interestingly, one of the RAP80 phosphorylation sites, Ser677, also appeared in a large-scale database generated from several phosphoproteomic screenings [111, 131, 133, 164, 165]. Also, there is increasing evidence that many proteins originally playing important roles in mitotic events, such as centrosome amplification and mitotic spindle assembly, are found to have previously undescribed functions in DDR. Aster-associated protein (ASAP), also called MAP9 is a centrosome- and spindle-associated protein [166, 167]. Phosphorylation of ASAP by Aurora A is essential for bipolar spindle assembly and required for correct mitotic progression. Its deregulation results in severe mitotic defects. Recently, Basbous et al. found that ASAP is involved in DDR by transiently accumulating at double strand breaks in response to IR, interacts with p53, and induces p53 stabilization for increasing its transcriptional activity during DDR

[167].

When Matsuoka et al. performed screening using the antibodies against the ATM/ATR consensus motif, they found that TPX2 is phosphorylated in response to IR at one particular site of TPX2, Serine 634 [65]. However, there has been no further study supporting the importance of this site in the literature. Based on their finding, we originally aimed to study this putative site,

Ser634, in the function of TPX2 in DDR, so we generated a phospho-dead mutant, S634A

(Serine is replaced to Alanine), and used it in rescue experiments; we knocked down endogenous

104

TPX2 by siRNA and transfected GFP-WT TPX2 or the S634A mutant to see the effect of this mutant on the level of -H2AX signal upon IR, but we did not observe any differences between

WT TPX2 and the S634A mutant (Data not shown). We also experienced some problems with the phospho-specific antibodies against Ser634 that we had generated, so we decided not to continue the characterization of TPX2 phosphorylation at Ser634. Instead, we found Thr72 phosphorylation of TPX2 is Cdk-dependent, as Cdk-dependent phosphorylation is involved in the DDR, we decided to focus on Thr72 phosphorylation in the function of TPX2 in DDR.

In this section, we hypothesized that TPX2 phosphorylation at Thr72 can mediate the function of TPX2 in DDR through regulating the level of -H2AX in response to IR.

105

Figure 5.1 Overexpression of GFP-TPX2 reverses the increase of -H2AX levels triggered by the absence of endogenous TPX2 in HeLa cells in response to IR [3] Transient transfection of TPX2 siRNA significantly increased the levels of -H2AX signal after 10 Gy of IR. Exogenous GFP-TPX2 expression in HeLa cells in the absence of endogenous TPX2 diminished the increase in -H2AX expression caused by depletion of TPX2 after 10 Gy of IR, as shown by western blot analysis.

5.2 Results

5.2.1 Analyzing the formation of IR-induced foci positive for pThr-72 upon IR treatment

When irradiated cells were stained with pThr-72 antibodies, we could see foci-like shapes within cells, so we wanted to test whether they were indeed IR-induced foci (IRIF), and whether they would co-localize with IRIF positive for -H2AX. For this experiment, we used the U2OS

(osteosarcoma cell line) because the formation of IRIF in U2OS cells is more pronounced than in

HeLa cells. To analyze the formation of IRIF, U2OS cells grown on coverslips were irradiated with 5 Gy of IR. After 1 hour recovery, cells were fixed with 4% PFA and immunostained with the pThr-72 antibodies, -H2AX (Ser139) antibodies and DAPI. As shown in Figure 5.2, cells treated with 5 Gy of IR show the formation of IRIF positive for pThr72-TPX2, which partially co-localized with IRIF positive for -H2AX signal. In un-irradiated U20S cells, very low expression of pThr-72, or low numbers of pThr-72-positive IRIF were detected. The specificity

106

of pThr-72 antibodies to pThr-72-positive IRIF signal was validated using antigen preabsorption with the corresponding blocking peptides (LQQAI VpT72PLKPVD, 66 to 78 amino acid region of TPX2 used as an antigen for raising the antibody). The different ratios of antibody-blocking peptide mixture (from 1:5 to 1:50) successfully blocked IRIF positive for the pThr-72 signal of

IRIF. This result which shows the antibodies are specific to pThr-72 IRIF signal (Figure 5.2).

Next, we tested whether these IRIF were phosphorylation-dependent. U2OS cells on coverslips were irradiated with 5 Gy of IR and incubated at 37 °C for 1 hour recovery. Cells were fixed, permeabilized and treated with lambda protein phosphatase (-PPase,100U/coverslip) at 30 °C for 30 minutes in the presence of 2 mM MnCl2. Control cells on coverslips were incubated in the same buffer containing phosphatase inhibitors (10 mM NAF, 10 mM -glycerophosphate) without -PPase. After treatment, cells were stained with the pThr-72 and the -H2AX antibodies. In control cells without -PPase treatment, IRIF of pThr-72, which partially co- localized with -H2AX were present, however, -PPase treatment abolished both IRIF of pThr-

72 and -H2AX (Figure 5.3). This proves that this IRIF positive for pThr-72 signal is phosphorylation-dependent.

107

108

Figure 5.2 Formation of IRIF positive for pThr-72 in irradiated U2OS cells and specificity of pThr-72 antibodies against pThr-72-positive IRIF signal using blocking peptides.

U20S cells on coverslips were irradiated with 5 Gy of IR and 1 hour after recovery, cells were fixed and stained with pThr-72 (FITC) and -H2AX antibodies (Cy3). DAPI staining shows

109

DNA. Cells treated with 5 Gy of IR show the formation of IRIF positive for pThr-72. In un- irradiated U20S cells, very low expression of pThr-72 or only few pThr-72-positive IRIF were detected. The specificity of pThr-72 antibodies to pThr-72-positive IRIF signal was validated using antigen preabsorption with pThr-72 blocking peptides. The blocking peptide blocked IRIF positive for the pThr-72 signal, which shows the antibodies are specific to pThr-72-positive IRIF signal.

110

Figure 5.3 -PPase treatment on pThr-72-positive IRIF shows that this IRIF formation is phosphorylation-dependent. U20S cells on coverslips were irradiated with 5 Gy of IR and 1 hour after recovery, cells were fixed, and permeabilized. Then, cells were incubated with -PPase at 30 °C for 30 minutes in the presence of 2 mM MnCl2. After -PPase treatment, cells were stained with pThr-72 (FITC) and -H2AX antibodies (Cy3). DAPI staining shows DNA. Irradiated U2OS cells demonstrated pThr-72 and -H2AX-positive IRIF partially co-localized. -PPase treatment abolished the IRIF formation of both pThr-72 and -H2AX.

111

5.2.2 Determination of the role of TPX2 phosphorylation at Thr72 in the amplification of -

H2AX signals upon exposure to IR.

In the previous section, we confirmed the formation of IR-induced foci positive for pThr-

72, which partially co-localized with -H2AX-positive IRIF. This result suggested that TPX2 phosphorylation at Thr72 might be important for the regulation of TPX2 in DDR. To test this possibility, we examined the effect of GFP-TPX2 WT and phospho-mutants on the levels of -

H2AX upon IR.

First, we tested the effect of the phospho-mutant of GFP-TPX2 T72A on the levels of -

H2AX signal upon IR treatment compared to the effect of GFP-TPX2 WT. Control siRNA or

TPX2 siRNA was co-transfected with an empty GFP vector, GFP-TPX2 WT or T72A mutant into U2OS cells. After 24 hours of transfection, cells were irradiated with 10 Gy of IR and incubated at 37 C for 1 hour of recovery. Cells were harvested and lysed in NETN buffer.

Chromatin fractionation was performed to get the soluble fraction as well as the chromatin insoluble fraction. The western blot result demonstrated that TPX2 depletion significantly increased the level of -H2AX in HeLa cells after 10 Gy of IR and 1 hour of recovery. When we transfected GFP-TPX2 WT or GFP-TPX2 T72A into TPX2-depleted HeLa cells, we found that

GFP-TPX2 expression reduced the increased level of phosphorylation of H2AX caused by the

TPX2 knockdown (Figure 5.4). On the other hand, GFP-TPX2 T72A expression in TPX2- depleted HeLa cells did not cause a reduction of -H2AX levels to the same extent as seen with

GFP-TPX2 expression (Figure 5.4).

Statistical analysis showed that the P value from ANOVA comparing 6 groups (Control siRNA no IR, Control siRNA IR, TPX2 siRNA + GFP vector only no IR, TPX2 siRNA + GFP

112

vector only IR, TPX2 siRNA + GFP-TPX2 IR and TPX2 siRNA + GFP-T72A IR) is <0.01, indicating there is a difference among the means of 6 groups. However, when the Neuman-Keuls test was performed to determine which means of each group in 6 groups differ for multiple comparison, P value was bigger than 0.05 for the groups, TPX2 siRNA+ GFP vector only IR vs.

TPX2 siRNA + GFP-TPX2 IR, or TPX2 siRNA + GFP-TPX2 IR vs. TPX2 siRNA + GFP-TPX2

T72A IR. This is probably due to the large variance in the results of 3 independent experiments.

But, we can see that there might be a trend that GFP-TPX2 WT and T72A mutant act differently in H2AX phenotype rescue (see the difference in the last two lanes in Figure 5.4A). However, more experiments with more accurate and similar amount of total H2AX may help to clarify these results. Also, the previous results of our laboratory show that regulation of -H2AX levels by TPX2 is cell cycle-dependent, particularly during the G0 and G1 phases of the cell cycle [3].

Therefore, if we perform this experiment using non-synchronized cells, the effect on the regulation of -H2AX might not be pronounced. It would be worth doing experiments with G1 phase-synchronized cells.

Taken together, these data suggest that TPX2 phosphorylation at Thr72 might regulate the function of TPX2 on -H2AX amplification in response to IR, however, cell cycle-dependent experiments and more functional assays should be performed in order to determine the role of

Thr72 phosphorylation in DDR.

113

114

Figure 5.4 Effect of the Thr72 phospho-mutant (T72A) on the levels of -H2AX in response to IR A) Control siRNA or TPX2 siRNA was co-transfected with an empty GFP vector, GFP-TPX2 WT or T72A mutant into U2OS cells. After treatment with 10 Gy of IR and recovery at 37 C for 1 hour, cells were harvested and lysed in NETN buffer for the chromatin fractionation. The samples of soluble fractions were used for analysis of TPX2 and actin blots and the samples of insoluble chromatin fractions were used for analysis of -H2AX (S139) and total H2AX blots. Transient transfection of GFP-TPX2 T72A mutant into HeLa cells in the absence of endogenous TPX2 did not show the reduction of H2AX as much as reduction caused by GFP-TPX2 WT expression upon IR, as indicated by western blot analysis. These are the representative western blot figures from 3 independent experiments. B) Bar charts showing the relative quantification of -H2AX signal from 3 independent experiments as shown in A. The relative quantification of -H2AX signals in control RNAi without IR were normalized to 1. The other values were compared to the control value. Control RNAi - IR: 1+/- 3.17209E-08, Control RNAi + IR: 11.223 +/- 2.2750, TPX2 UTR siRNA+ GFP - IR: 12.3473 +/- 3.7133, TPX2 UTR siRNA + GFP + IR: 93.587 +/- 17.953, TPX2 UTR siRNA + GFP-TPX2 + IR: 54.0478 +/- 6.24, TPX2 UTR siRNA + GFP-TPX2 T72A + IR: 88.3804 +/- 26.03. Group: Mean +/- S.E.M. n=3, P value of ANOVA comparing all groups was P<0.01. However, each P value from Neuman-Keuls test shows P>0.05 in TPX2 siRNA+ GFP vector + IR vs.TPX2 siRNA+ GFP-TPX2 + IR and in TPX2 siRNA+ GFP-TPX2 + IR vs.TPX2 siRNA + T72A + IR, NS: Not significant.

115

5.3 Discussion

5.3.1 Summary

We observed that pThr-72 formed IRIF in response to IR, which only associates with a fraction of IRIF positive for γ-H2AX. Using a pThr-72 blocking peptide, the specificity of pThr-

72 antibodies to pThr-72-positive foci signal was confirmed. In addition, -PPase treatment on irradiated cells demonstrated that these foci are dependent on phosphorylation of TPX2 at Thr72.

Furthermore, the trend from the results of the rescue experiments using TPX2 siRNA and GFP-

TPX2 WT or GFP-TPX2 T72A mutant show that GFP-TPX2 WT can rescue the phenotype of increased levels of -H2AX back to that of the control cells, while the GFP-TPX2 T72A mutant cannot rescue its phenotype. This result suggests that TPX2 phosphorylation at Thr72 might be important for regulating the function of TPX2 in -H2AX amplification upon DNA damage.

However, more experiments need to be performed to clarify this statement.

5.3.2 Alternative interpretations for each experiment.

In Figure 5.2 and 5.3, we confirmed the antibody specificity for pThr-72-positive IRIF formation with a blocking peptide and -PPase treatment. Antibody incubation with the blocking peptide showed that the pThr-72 antibodies are specific to pThr-72-positive IRIF. Furthermore,

-PPase treatment confirmed that these foci are phosphorylation-dependent. Additionally, our data would be stronger if we could have used phospho-mutant transfections (GFP-TPX2 T72A and T72E) to test if the pThr-72 antibodies only recognize pThr-72-positive IRIF on GFP-TPX2

WT transfected cells, keeping in mind the caveat that the cells would have endogenous levels of pThr-72 when irradiated.

116

In Figure 5.4, we compared six test groups (Control RNAi no IR, Control RNAi after IR,

TPX2 UTR siRNA+ an empty GFP vector no IR, TPX2 UTR siRNA + an empty GFP vector after IR, TPX2 UTR siRNA+ GFP-TPX2 after IR, TPX2 UTR siRNA + GFP-TPX2 T72A after

IR) to evaluate the levels of -H2AX upon IR. Even though the result was significant when comparing the means of all 6 groups using ANOVA, the more specific after Newman-Keuls test comparing between each pair of two groups, GFP and GFP-TPX2 WT or GFP-WT and GFP-

TPX2 T72A showed no statistical differences. However, the T72A mutant did not rescue the phenotype of -H2AX levels in cells depleted of TPX2 to the same extent as GFP-TPX2 WT. In this experiment, we compared only irradiated cells with TPX2 siRNA + GFP, TPX2 siRNA+

GFP-TPX2 WT and TPX2 siRNA + GFP-TPX2 T72A mutant. However, two more groups (un- irradiated cells with TPX2 siRNA + GFP-TPX2 and TPX2 siRNA + GFP-TPX2 T72A) could have been added to get a more detailed comparison. In addition, when cells are treated with 10

Gy of IR, DSBs are generated and histone H2AX is rapidly phosphorylated into -H2AX at S139 around DSB sites. This event is fast and abundant, so in western blots, the level of -H2AX phosphorylated at S139 increases and peaks within 1 hour, and the level of -H2AX in irradiated cells is significantly comparable to that of un-irradiated control cells [168]. From the result of our western blots, the levels of -H2AX on irradiated-control cells should have been higher than those seen in Figure 5.4. It would have been better if we used controls to ensure our chromatin fractionation worked properly, making our results more reliable.

In our previous studies, when we proved that TPX2 regulates the levels of -H2AX amplification on DNA damage [3], several different approaches of TPX2 knockdown were used; a TPX2-doxycycline-inducible miRNA HeLa cell line and two different siRNAs (TPX2 cds

117

siRNA and TPX2 UTR siRNA). However, I only used TPX2 UTR siRNA for the experiments in my thesis, and the results would have been more clear and substantiated if the effects of GFP-

TPX2 WT and T72A mutant on the levels of -H2AX amplification were tested in several different settings as in the previous studies [3].

118

CHAPTER SIX:

GENERAL CONCLUSION AND FUTURE DIRECTIONS

6.1 General Discussion and Future directions

6.1.1 Considering other potential phosphorylation sites of TPX2

Among these putative Cdk sites, Thr59 is located quite closed to Thr72. Therefore, there is the possibility that mutation of Thr72 may result in a conformational change of the TPX2 protein, which could possibly influence the phosphorylation of other phosphorylation sites, especially Thr59. We have verified the specificity of the pThr-72 antibodies using IP experiments with GFP-TPX2 WT and GFP-TPX2 T72A mutant and showed the antibodies only recognize WT TPX2 phosphorylated at Thr72, not the mutant. However, because Thr59 is also a

Cdk site, there is possibility that the pThr-72 antibodies also recognize phosphorylation of TPX2 at Thr59. Also, T72A mutation could have affected the phosphorylation of Thr59.

To avoid this possibility, there are several things we should have done for proof.

First, we should have performed the mass spectrometry analysis using WT TPX2 and mutant TPX2 (T72A) IP from cells and compared the phosphorylation status of the Thr59 site. In the IP sample of WT-TPX2, both Thr72 and Thr59 residues have to be phosphorylated (which we have confirmed from our mass spectrometry data). In our mass spectrometry analysis, we used endogenous TPX2 IP from cells, but we still detected the phosphorylation at both Thr72 and Thr59. If we had immunoprecipitated mutant TPX2 (T72A) from cells, we should have gotten the result that Thr72 is unphosphorylated, and phosphorylation at Thr59 should be detected by mass spectrometry. Then, we would be able to say the mutation of Thr72 would not affect the phosphorylation status of Thr59. However, if the result from mass spectrometry

119

analysis of the T72A mutated sample shows that phosphorylation at Thr59 does not occur, which means T72A mutation would affect the phosphorylation of Thr59. We should perform further experiments to test how these two phosphorylation sites reciprocally affect the phosphorylation status of each other.

Second, we could generate the Alanine mutant of Thr59 (T59A) and test this mutant for the specificity of our phospho-Thr72 TPX2 antibodies against Thr59. After we transfect GFP-

TPX2 WT, two mutants forms of GFP-TPX2, T72A and T59A into cells, IP experiments could be performed from each cell lysate and western blot analysis using IP samples can be performed to see the specificity of the pThr-72 antibodies. If the antibodies still recognize TPX2 IP bands from the T59A mutant, this would mean the pThr-72 antibodies are specific to Thr72. However, if the pThr-72 antibodies could not recognize TPX2 IP from T59A, it would indicate that either the pThr-72 antibodies are not specific or mutation of the Thr59 site affects the phosphorylation of TPX2 at Thr72. In these ways, we could also test other Cdk phosphorylation sites besides

Thr59 to see whether mutation of Thr72 would affect any other phosphorylation sites.

Third, we could generate the double mutant of both Thr72 and Thr59 (T72A/T59A) and compare this mutant with each single mutant (T72A or T59A). After we transfect a T72A/T59A double mutant or a single mutant (T72A or T59A) into cells, we could perform IP experiments with TPX2 or GFP antibodies. And then, western blot analysis could be performed with pThr-72 antibodies to check whether there is any difference between a double mutant and a single mutant.

120

6.1.2 TPX2 and cancers

TPX2 is often overexpressed in various human cancers such as lung, cervix, liver, prostate, pancreas, and ovary, etc. [114-118]. Furthermore, expression of TPX2 positively correlated with progression of several cancers such as malignant astrocytoma and squamous cell lung cancer carcinoma, and TPX2 overexpression was associated with decreased patient survival rate [121-123]. According to the results from my thesis, it might be possible for us to use phosphorylation of TPX2 to target the overexpressed TPX2 in cancers, if we can further prove that TPX2 phosphorylation at Thr72 is important for the function of TPX2 in mitotic spindle assembly and spindle organization. By inactivating or abolishing TPX2 phosphorylation at

Thr72, particularly in cancer cells, it would be possible to make cancer cells have more multipolar spindles; possibly, it would kill cancer cells. However, it would be difficult to find a way to inactivate or abolish specifically Thr72 phosphorylation using a pharmaceutical approach.

It would be important to elucidate the exact mechanism of this pathway. On the other hand, if we can determine if pThr-72 TPX2 phosphorylation or any other putative phosphorylation site is important for the function of TPX2 in DDR and we are able to reveal the mechanism of TPX2 phosphorylation in response to DNA damage, we would be able to use the TPX2-regulated DDR pathway for treatment of cancers, perhaps by using radiotherapy. There is increasing research suggesting that TPX2 can be used as a potential target for cancer therapeutics [116, 169-171].

For example, Warner et al. validated TPX2 as a target in pancreatic cancer [116]. The copy number of TPX2 is amplified in pancreatic cancer cells and they found that TPX2 expression by measuring the levels of mRNA and protein was much higher in pancreatic cancer cells compared to normal pancreatic cells. They found that TPX2 siRNA treatment significantly reduced cancer cell growth and led to cell death. Therefore, they suggested that TPX2 can be used as a potential

121

target in pancreatic cancer therapy [116]. Therefore, it would be worth doing translational experiments using particular cancer cells or tissue where overexpression of TPX2 can be found, such as liver, lung, cervix and ovary, in the context of targeting Thr72 phosphorylation [114-

118]. It would be important to examine phosphorylation of TPX2 in those cancers and study the possible correlation of TPX2 phosphorylation and progression of cancer, or differences in the levels of TPX2 phosphorylation between particular cancer tissues or cells and their normal counterparts. Tissue microarray experiments from normal and cancer tissues would also give us clues to the differences in TPX2 phosphorylation between normal and cancer cells [172].

6.1.3 Further characterization of Thr72 TPX2 phosphorylation in mitotic spindle assembly

In chapter 4, we found that TPX2 phosphorylation at Thr72 might be important for the function of TPX2 in spindle formation. We found that when TPX2 phosphorylation at Thr72 did not occur, cells had more multipolar spindles. This finding is only an initial step, as we need to further characterize the mechanism causing this phenotype. There are two hypothetical mechanisms we can consider based on our results and the literature in order to explain the increase in the frequency of multipolar spindles in phospho-deficient mutants of TPX2 at Thr72 and to prove the role of Thr72 phosphorylation in spindle formation.

1) TPX2 phosphorylation at Thr72 may be important for Aurora A activation.

There is one paper showing that Aurora A inactivation results in multipolar spindles and spindle poles fragment, due to MT hyperstabilization [173]. The authors believed that Aurora A inactivation caused an imbalance in directional MT forces, which led to spindle pole fragmentation. They showed that localization of both the MT stabilizing protein ch-TOG

(colonic and hepatic tumor over-expressed gene), and the MT-depolymerizing motor protein,

122

MCAK (mitotic centromere-associated kinesin), at the spindle poles was changed by Aurora A inactivation; increased levels of ch-TOG and decreased MCAK at the spindle poles [173, 174].

Ch-TOG and MCAK are two main proteins regulating the overall dynamics of MTs in human cells [175, 176]. Interestingly, silencing of ch-TOG in Aurora A-inactivated cells reduced the percentage of cells with multipolar spindles. Aurora A inactivation changed the localization of ch-TOG and MCAK at spindle poles, which led to the disruption of MT dynamics (MCAK- mediated depolymerization was decreased and ch-TOG-mediated stabilization was increased).

As a consequence of this, MT depolymerization was diminished at the minus end of MTs, which caused extra pressure at the poles and led to spindle pole fragmentation. Therefore, when they depleted ch-TOG in Aurora A-inactivated cells, more cells had normal bipolar spindle structure.

In another approach, they inhibited the motor protein Eg5 to decrease centrosome-directed forces, which caused the generation of multipolar spindle to be abolished. As a result of Eg5 inhibition, the percentage of cells with monopolar spindles was increased [173]. Based on information from this paper, it would be interesting to test whether multipolar spindle formation caused by T72A mutation in our experiments is also resulting from spindle pole fragmentation.

The phosphorylation sites of TPX2 that cause Aurora A inactivation is known from the studies of

Eyers et al [32]; Tyr8 and Ty10 in Xenopus TPX2. Eyers et al. showed that mutating Tyr 8 and

10 to Alanine blocked TPX2 phosphorylation and inactivated Aurora A in vitro by preventing the binding of TPX2 to Aurora A and therefore, blocking Aurora A activation. These are in vitro phosphorylation sites and have never been confirmed in human cells. Even though Thr72 is a

Cdk site, it is close to the Aurora A binding domain of TPX2. So, it is possible for Thr72 to be an important site for Aurora A activation. In order to test this hypothesis, we could first start performing immunofluorescence (IF) staining with ch-TOG or MCAK antibodies in GFP-TPX2

123

WT or T72A mutant-transfected cells to get clues as to whether this mutation causes any change in localization of ch-TOG and MCAK. Second, to check the activity of Aurora A of mutants directly, we could perform Aurora A kinase assays using cell lysates from GFP-TPX2 WT and

T72A mutants to see whether there is inactivation of Aurora A in T72A mutant transfected cells.

2) Thr72 phosphorylation might be responsible for TPX2 function on Eg5 activity.

It is already known that the C-terminus of human TPX2 is essential for bipolar spindle formation [11]. The C-terminus of TPX2 is important for the role of TPX2 during mitosis to regulate the localization and the activity of a kinesin-5 motor protein, Eg5 for spindle organization. Ma N et al. have shown that in cells lacking the TPX2-Eg5 interaction domain, fragmentation of spindle poles and multipolar spindles are observed [20]. In addition, TPX2-Eg5 interaction was important for localization of Eg5 to spindle MTs. Eg5and TPX2 co-localized on mitotic spindles. When the Eg5-TPX2 interaction was perturbed by absence of the c-terminus domain of TPX2, Eg5 localization to MTs was greatly reduced [20].

In this section, our data show that TPX2 knockdown causes cells to have the increased frequency of multipolar spindles and this increase was even more pronounced when there is no

TPX2 phosphorylation at Thr72. Therefore, based on the phenotype we observed and literature, we will be able to carefully hypothesize that TPX2 phosphorylation at Thr72 might be important for the function of TPX2 in regulating the activity of Eg5 or Eg5 localization for spindle organization or formation. However, we have not done any experiment to prove this interaction, therefore we have to do further experiments to get clues to start this investigation. In order to test this hypothesis, we can perform MT isolation and purification after transfection of GFP-TPX2

124

WT and T72A mutant into HeLa cells to check whether there is any change in the levels of Eg5 protein in mitotic spindles of T72A mutant compared to TPX2 WT transfected cells. If there is any change, we can further examine whether Thr72 phosphorylation might be important for the regulation of Eg5 activity. As mentioned in the introduction section, 1.2.3.1.1-(4), TPX2 is important for regulating the motor activity of Eg5. Therefore, by using MT gliding and MT sliding assay [20], if we can test whether TPX2 WT protein and T72A mutant protein demonstrate any difference in Eg5 dependent-MT gliding and sliding activity, we would be able to tell whether TPX2 phosphorylation at Thr72 is important for regulating Eg5 motor activity.

3) The effect of WT TPX2 and the T72A mutant on mitotic progression

During mitosis, cells have developed mechanisms to activate the spindle assembly checkpoint (SAC) or spindle checkpoint. SAC delays cell entry into anaphase until two kinetochores on chromosomes are properly attached to the mitotic spindle, preventing chromosome missegregation [177]. Thus, one could justly hypothesize that the T72A mutant may delay mitotic progression (particularly at anaphase onset) because cells transfected with expression of T72A mutant increases the number of cells with multipolar spindles. In this scenario, SAC in those cells would be activated in order to repair the defect before progression into anaphase. In order to see the effect on the progression of mitosis, we can perform time lapse microscopy analysis to measure the kinetics of GFP-TPX2 WT and mutants during mitotic progression.

125

6.1.4 Further characterization of TPX2 phosphorylation at Thr72 in DNA damage response

In chapter 5, we found that Thr72 might be important for the regulation of -H2AX amplification in response to IR. We confirmed that pThr-72 antibodies are specific to IRIF positive for pThr-72. Furthermore, these foci were shown to be Thr72 phosphorylation- dependent by -PPase treatment. The rescue experiments using TPX2 siRNA and GFP-TPX2

WT or GFP-TPX2 T72A mutant show a trend that WT and T72A mutant act differently.

Therefore, if additional experiments suggested below clearly show the importance of this phosphorylation site in DDR, we will be able to clearly make conclusion that TPX2 phosphorylation at Thr72 is important for -H2AX regulation upon DNA damage.

1) Inhibitor treatment on irradiated cells

To start with, we could use a pharmaceutical approach to find out what kinase mediates

Thr72-TPX2 phosphorylation. As we already confirmed, Cdks are responsible for Thr72 phosphorylation in mitosis, therefore, it is also possible that Cdks would be responsible for TPX2 phosphorylation upon IR treatment. There is increasing evidence that Cdk activity regulates various steps in the DDR pathway and Cdks mediate phosphorylation of several proteins involved in DDR such as Chk1, RPA, NBS1 and BRACA2. To test that, we could use Cdk inhibitors such as roscovitine and alsterpaullone which we used to test Thr72 phosphorylation in mitosis. We could treat irradiated cells with Cdk inhibitors to test whether they would be able to abolish pThr-72-positive IRIF.

Alternatively, even though Thr72 is a substrate of Cdks, it is possible that other kinases might regulate this phosphorylation during interphase, such as Polo-like kinase, Aurora A, ATM,

ATR or DNA-PK. First, we could use a broad PI-3 kinase inhibitor, such as wortmannin 126

[178, 179]. If any effect is observed on the formation of IRIF positive for pThr-72, we could try other inhibitors selective for ATM, ATR or DNA-PK. Also, there are other kinases known to be important in mitotic phosphorylation and DDR such as Plk [157, 180]. We could also include this possibility in our experiments. Therefore, by treating cells with appropriate inhibitors to find out which kinase is responsible for the formation of IRIF positive for pThr-72, we could further resolve the mechanism in the context of DDR pathways.

2) Analysis pThr-72-positive IRIF

In order to further characterize TPX2 phosphorylation at pThr-72-positive IRIF, we could perform experiments to test the kinetics of pThr72-positive-IRIF using different recovery times and different irradiation doses. Also, we already know that expression of TPX2 and TPX2 phosphorylation at Thr72 are cell cycle-dependent, so it would be interesting to analyze the pThr-72 positive IRIF in each cell cycle stage.

3) Cell cycle-dependent analysis of the effect of GFP-TPX2 WT and T72A mutant on -H2AX levels upon IR treatment.

We only used non-synchronized cells to see the effect of GFP-TPX2 WT and T72A mutant in cells depleted of endogenous TPX2 on the levels of -H2AX in response to DNA damage. We observed that their effects are different, but we did not get clear results. Therefore, it would be better to synchronize cells in each stage of the cell cycle, G1, S, G2 and M phase and analyze the effect of WT TPX2 and T72A mutant in each stage.

4) Analysis of other putative phosphorylation sites of TPX2

127

If it turns out that Thr72 is not responsible for the function of TPX2 in regulating the level of -H2AX, we could try to characterize other putative phosphorylation sites of TPX2; possibly, some of the ATM/ATR consensus sites from my own spectrometry analysis or, as shown in Matsuoka’s screening that TPX2 is phosphorylated in response to IR. It would be necessary that we perform another mass spectrometry analysis using cell lysates treated with IR to find out which phosphorylation sites of TPX2 get activated in response to IR. Then we could choose candidate sites we want to characterize.

Taken together, in order to reveal whether Thr72 phosphorylation is important for the function of TPX2 in DDR, we would have to perform more experiments for proof, and then study the mechanism of how this phosphorylation is regulated by other factors.

6.2 Conclusion

In my thesis, I characterized TPX2 phosphorylation at one particular phosphorylation site, Thr72 in human cells. I found Thr72 phosphorylation is cell cycle-dependent and Cdk/cylin mediates this phosphorylation in mitosis. Interestingly, I found Thr72 phosphorylation might be important for the functions of TPX2 in mitotic spindle, particularly, spindle organization. A detail mechanism for regulating this phosphorylation remains to be determined. Overexpression of TPX2 and amplification of TPX2 locus have been associated with various cancers, which suggested TPX2 as a potential oncogene [114, 115, 117, 119, 122]. It will be important to analyze phosphorylation levels of TPX2 in those cancers and seek for the possibility whether phosphorylation of TPX2 can be used as a potential therapeutic target. Our findings that TPX2 is phosphorylated at Thr72 in mitosis and the mutation of this site has affected the normal bipolar spindle formation are quite important and original. Also, our findings have the potential to make

128

big contribution to TPX2 research fields. However, the further studies need to be done to make this as a complete story. As mentioned, TPX2 is a phosphoprotein with over 40 putative phosphorylation sites. Here, I characterized only one phosphorylation site. Therefore, it will be valuable to map all of the phosphorylation sites of TPX2 in mitosis and gets phosphorylated in response to DNA damage, validate each site and characterize the functions of each site of TPX2.

Understanding the significance of TPX2 phosphorylation not only at this one site, Thr72, but also at other sites may provide new insights into the roles of TPX2 in healthy tissues or disease conditions, such as cancer. Possibly, TPX2 phosphorylation can be used as a new therapeutic target for certain cancers.

129

CHAPTER SEVEN:

CRITICAL EVALUATION OF MY WORK

The hypothesis of my thesis was that TPX2 phosphorylation at Thr72 is important for regulating the functions of TPX2 in mitotic spindle assembly and DNA damage response. Herein,

I evaluated my work based on the strengths and weaknesses of each section of the results.

1. The results in Chapter Three: Characterization of TPX2 phosphorylation at Thr72 in human cancer cells

Chapter 3 contains the strongest collection of data and constitutes the foundation of my thesis. Therein, I conclusively showed that TPX2 is phosphorylated at Thr72 in vivo. TPX2 phosphorylation at Thr72 had only been shown in vitro from our previous data and reports in the literature describing high throughput proteomic screening. The strength of my data is its reliability. I clearly showed that the phospho-TPX2 antibodies against Thr72 are specific to

TPX2 phosphorylated at Thr72 as shown by western blots and immunostaining of cells, as well as by siRNA treatment, wherein positive staining was removed. With these antibodies, I could accurately and confidently show that Thr72 phosphorylation is cell cycle-dependent, and my own mass spectrometry analysis undoubtedly demonstrates that Thr72 is an in vivo phosphorylation site of TPX2 in mitotic HeLa cells. Cdk inhibitor treatments significantly decreased the levels of TPX2 phosphorylation at Thr72 during mitosis. For the most part, the correct controls were used to show that the experiments worked properly.

However, there are some weaknesses in the first section; some caveats to my experiments have to be addressed. As mentioned in the discussion, I identified 16

130

phosphorylation sites in TPX2 from mitotic HeLa cells. In addition to Thr72, there were other

CDK sites that remained uncharacterized. The possibility that mutating Thr72 can change the phosphorylation status of the other sites in TPX2 was not examined. In addition, for mass spectrometry, a control IP with IgG antibody was not used to allow subtraction of non-specific proteins within the TPX2 IP sample. Even though I got the result we expected, I have not conclusively ruled out the possibility that the same band could have been obtained non- specifically.

2. The results in Chapter Four: Determination of the roles of TPX2 phosphorylation at Thr72 in mitotic spindle assembly

While less complete than chapter 3, chapter 4 is the next strongest to prove my hypothesis. In this section, I showed that mutating Threonine to Alanine at Thr72 increased the frequency of cells having multipolar mitotic spindles compared to the frequency of cells with wild-type TPX2.

The originality of this section is that TPX2 phosphorylation at Thr72 was implicated as being important for TPX2 function in mitotic spindle assembly, possibly normal spindle formation. In the experiments testing the effects of GFP-TPX2 WT or the phospho-dead mutant of TPX2 (T72A) on the morphology of mitotic spindle in HeLa cells, we could reproduce the result with TPX2 knockdown from other groups; increase in the percentage of cells with multipolar spindles [11, 22]. Furthermore, T72A mutation increased the frequency of cells with multipolar spindle even more. This is a novel finding.

Conversely, the weaknesses of this section is that even though we saw the phenotype caused by mutating the Thr72 phosphorylation site, we did not continue performing experiments

131

to fully characterize the impact of the mutation. The functions of TPX2 in mitotic spindle assembly are multi-fold. My data would be more convincing if I had investigated more specific functions of TPX2 in spindle assembly, namely MT nucleation, MT bundling, Aurora A activation and regulation of Eg5 localization/motor activity using appropriate functional assays.

3. The results in Chapter Five: Determination of the roles of TPX2 phosphorylation at Thr72 in

DNA damage response

The third section of the results, chapter 5 is the weakest. In this section, I showed unconvincingly that TPX2 phosphorylation at Thr72 might be important for TPX2 regulation of the levels of -H2AX amplification in response to DNA damage. I showed that pThr-72-positive

IRIF signals co-localized with IRIF incorporating -H2AX. And I also demonstrated that the

T72A mutant of GFP-TPX2 could not rescue the levels of -H2AX back to the levels seen with control GFP-TPX2 WT.

The novelty of this section is that my results implicate Thr72 might be important in the

DDR pathway. Blocking peptide experiments obviously showed that pThr-72 antibodies are specific to pThr-72-positive-IRIF. Next, using -PPase treatment on irradiated cells, I clearly verified that these IRIF are phosphorylation-dependent.

However, there are several limitations in chapter 5. First, no further functional assay was performed to fully characterize these observed phenotypes. Next, due to the caveats that

Thr72 phosphorylation site could possibly be affected by other phosphorylation sites (for example, Thr59), the results in control experiments with a blocking peptide and -PPase treatment in the IRIF experiments is not hundred percent convincing. The last limitation in this

132

chapter is that I could not prove that TPX2 phosphorylation at Thr72 in DDR is indeed important for regulating the levels of -H2AX. Adding several more experiments for confirmation could have made this section stronger. For examples, by transfecting the T72A mutant and GFP-TPX2

WT into cells and immunostaining of irradiated cells with -H2AX and pThr-72 antibodies, we could check whether mutant overexpression in these cells has any effect on the -H2AX-positive

IRIF formation, in comparison to the irradiated cells overexpressing WT-TPX2. In addition, we could also check whether the pThr-72 signal would be decreased or abolished by T72A mutant transfection. The result from this experiment would prove that pThr-72-positive IRIF are indeed specific to TPX2 phosphorylated at Thr72 after IR.

133

REFERENCES

1. Cassimeris, L. and C. Spittle, Regulation of microtubule-associated proteins. Int Rev Cytol, 2001. 210: p. 163-226. 2. Cassimeris, L. and R.V. Skibbens, Regulated assembly of the mitotic spindle: a perspective from two ends. Curr Issues Mol Biol, 2003. 5(3): p. 99-112. 3. Neumayer G, H.A., Shim SY, Le HT, Lundin C, Belzil C, Chansard M, Yu Y, Lees- Miller SP, Gruss OJ, van Attikum H, Helleday T, Nguyen MD, Targeting protein for xenopus kinesin-like protein 2 (TPX2) regulates γ-histone 2AX (γ-H2AX) levels upon ionizing radiation. J Biol Chem. , 2012. 287(50): p. 42206-22. 4. Hornbeck, P.V., et al., PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res, 2012. 40(Database issue): p. D261-70. 5. Heidebrecht, H.J., et al., p100: a novel proliferation-associated nuclear protein specifically restricted to cell cycle phases S, G2, and M. Blood, 1997. 90(1): p. 226-33. 6. Wittmann, T., et al., Localization of the kinesin-like protein Xklp2 to spindle poles requires a leucine zipper, a microtubule-associated protein, and dynein. J Cell Biol, 1998. 143(3): p. 673-85. 7. Gennerich A, V.R., Walking the walk: how kinesin and dynein coordinate their steps. Curr Opin Cell Biol., 2009. 21(1): p. 59-67. 8. Wittmann, T., et al., TPX2, A novel xenopus MAP involved in spindle pole organization. J Cell Biol, 2000. 149(7): p. 1405-18. 9. Boleti, H., E. Karsenti, and I. Vernos, Xklp2, a novel Xenopus centrosomal kinesin-like protein required for centrosome separation during mitosis. Cell, 1996. 84(1): p. 49-59. 10. Perez de Castro, I. and M. Malumbres, Mitotic Stress and Chromosomal Instability in Cancer: The Case for TPX2. Genes Cancer, 2012. 3(11-12): p. 721-30. 11. Goshima, G., Identification of a TPX2-like microtubule-associated protein in Drosophila. PLoS One, 2011. 6(11): p. e28120. 12. Ozlu, N., et al., An essential function of the C-elegans ortholog of TPX2 is to localize activated aurora A kinase to mitotic spindles. Developmental Cell, 2005. 9(2): p. 237- 248. 13. Vos, J.W., et al., The Plant TPX2 Protein Regulates Prospindle Assembly before Nuclear Envelope Breakdown. Plant Cell, 2008. 20(10): p. 2783-2797. 14. Hebras, C. and A. McDougall, Urochordate Ascidians Possess a Single Isoform of Aurora Kinase That Localizes to the Midbody via TPX2 in Eggs and Cleavage Stage Embryos. Plos One, 2012. 7(9). 15. Zeng, K., et al., Protein phosphatase 6 regulates mitotic spindle formation by controlling the T-loop phosphorylation state of Aurora A bound to its activator TPX2. J Cell Biol, 2010. 191(7): p. 1315-32. 16. Bayliss, R., et al., Structural basis of Aurora-A activation by TPX2 at the mitotic spindle. Molecular Cell, 2003. 12(4): p. 851-862.

134

17. Giesecke, A. and M. Stewart, Novel binding of the mitotic regulator TPX2 (target protein for Xenopus kinesin-like protein 2) to importin-alpha. J Biol Chem, 2010. 285(23): p. 17628-35. 18. Brunet, S., et al., Characterization of the TPX2 domains involved in microtubule nucleation and spindle assembly in Xenopus egg extracts. Mol Biol Cell, 2004. 15(12): p. 5318-28. 19. Schatz, C.A., et al., Importin alpha-regulated nucleation of microtubules by TPX2. EMBO J, 2003. 22(9): p. 2060-70. 20. Ma, N., et al., TPX2 regulates the localization and activity of Eg5 in the mammalian mitotic spindle. J Cell Biol, 2011. 195(1): p. 87-98. 21. Eckerdt, F., et al., Spindle pole regulation by a discrete Eg5-interacting domain in TPX2. Curr Biol, 2008. 18(7): p. 519-25. 22. Gruss, O.J., et al., Chromosome-induced microtubule assembly mediated by TPX2 is required for spindle formation in HeLa cells. Nature Cell Biology, 2002. 4(11): p. 871- 879. 23. Stewart, S. and G. Fang, Anaphase-promoting complex/cyclosome controls the stability of TPX2 during mitotic exit. Mol Cell Biol, 2005. 25(23): p. 10516-27. 24. Kahana, J.A. and D.W. Cleveland, Cell cycle - Some importin news about spindle assembly. Science, 2001. 291(5509): p. 1718-1719. 25. Trieselmann, N., et al., Ran modulates spindle assembly by regulating a subset of TPX2 and Kid activities including Aurora A activation. J Cell Sci, 2003. 116(Pt 23): p. 4791-8. 26. Dasso, M. and R.T. Pu, Nuclear transport: run by Ran? Am J Hum Genet, 1998. 63(2): p. 311-6. 27. Dasso, M., RCC1 in the cell cycle: the regulator of chromosome condensation takes on new roles. Trends Biochem Sci, 1993. 18(3): p. 96-101. 28. Bischoff, F.R. and H. Ponstingl, Catalysis of guanine nucleotide exchange on Ran by the mitotic regulator RCC1. Nature, 1991. 354(6348): p. 80-2. 29. Gruss, O.J. and I. Vernos, The mechanism of spindle assembly: functions of Ran and its target TPX2. J Cell Biol, 2004. 166(7): p. 949-55. 30. Fu, J., et al., Roles of Aurora kinases in mitosis and tumorigenesis. Mol Cancer Res, 2007. 5(1): p. 1-10. 31. Kufer, T.A., et al., Human TPX2 is required for targeting Aurora-A kinase to the spindle. Journal of Cell Biology, 2002. 158(4): p. 617-623. 32. Eyers, P.A. and J.L. Maller, Regulation of Xenopus Aurora A activation by TPX2. J Biol Chem, 2004. 279(10): p. 9008-15. 33. Petry, S., et al., Branching microtubule nucleation in Xenopus egg extracts mediated by augmin and TPX2. Cell, 2013. 152(4): p. 768-77. 34. Gadde, S. and R. Heald, Mechanisms and molecules of the mitotic spindle. Curr Biol, 2004. 14(18): p. R797-805. 35. Heald, R., et al., Self-organization of microtubules into bipolar spindles around artificial chromosomes in Xenopus egg extracts. Nature, 1996. 382(6590): p. 420-425. 36. Caudron, M., et al., Spatial coordination of spindle assembly by chromosome-mediated signaling gradients. Science, 2005. 309(5739): p. 1373-6. 37. Scholey, J.M., I. Brust-Mascher, and A. Mogilner, Cell division. Nature, 2003. 422(6933): p. 746-52.

135

38. Wadsworth, P., et al., Variations on theme: spindle assembly in diverse cells. Protoplasma, 2011. 248(3): p. 439-46. 39. Tulu, U.S., et al., Molecular requirements for kinetochore-associated microtubule formation in mammalian cells. Current Biology, 2006. 16(5): p. 536-541. 40. Garrett, S., et al., hTPX2 is required for normal spindle morphology and centrosome integrity during vertebrate cell division. Curr Biol, 2002. 12(23): p. 2055-9. 41. Aguirre-Portoles, C., et al., Tpx2 controls spindle integrity, genome stability, and tumor development. Cancer Res, 2012. 72(6): p. 1518-28. 42. Goshima, G., et al., Augmin: a protein complex required for centrosome-independent microtubule generation within the spindle. Journal of Cell Biology, 2008. 181(3): p. 421- 429. 43. Uehara, R., et al., The augmin complex plays a critical role in spindle microtubule generation for mitotic progression and cytokinesis in human cells. Proc Natl Acad Sci U S A, 2009. 106(17): p. 6998-7003. 44. Neumayer G, B.C., Gruss OJ, Nguyen MD, The long and winding road of the map TPX2. 45. Fu, J.Y., et al., Roles of aurora kinases in mitosis and tumorigenesis. Molecular Cancer Research, 2007. 5(1): p. 1-10. 46. Jeng, Y.M., et al., Overexpression and amplification of Aurora-A in hepatocellular carcinoma. Clinical Cancer Research, 2004. 10(6): p. 2065-2071. 47. Kallioniemi, A., et al., Detection and mapping of amplified DNA sequences in breast cancer by comparative genomic hybridization. Proc Natl Acad Sci U S A, 1994. 91(6): p. 2156-60. 48. Marumoto, T., et al., Aurora-A kinase maintains the fidelity of early and late mitotic events in HeLa cells. J Biol Chem, 2003. 278(51): p. 51786-95. 49. Glover, D.M., et al., Mutations in Aurora Prevent Centrosome Separation Leading to the Formation of Monopolar Spindles. Cell, 1995. 81(1): p. 95-105. 50. Sardon, T., et al., Dissecting the role of Aurora A during spindle assembly. EMBO J, 2008. 27(19): p. 2567-79. 51. Walter, A.O., et al., The mitotic serine/threonine kinase Aurora2/AIK is regulated by phosphorylation and degradation. Oncogene, 2000. 19(42): p. 4906-4916. 52. Bird, A.W. and A.A. Hyman, Building a spindle of the correct length in human cells requires the interaction between TPX2 and Aurora A. J Cell Biol. , 2008. 182(2): p. 289- 300. 53. Wyman, C. and R. Kanaar, DNA double-strand break repair: All's well that ends well. Annual Review of Genetics, 2006. 40: p. 363-383. 54. Liang, Y., et al., DNA damage response pathways in tumor suppression and cancer treatment. World J Surg, 2009. 33(4): p. 661-6. 55. Ciccia, A. and S.J. Elledge, The DNA damage response: making it safe to play with knives. Mol Cell, 2010. 40(2): p. 179-204. 56. van Attikum, H. and S.M. Gasser, Crosstalk between histone modifications during the DNA damage response. Trends in Cell Biology, 2009. 19(5): p. 207-217. 57. Myers, J.S. and D. Cortez, Rapid activation of ATR by ionizing radiation requires ATM and Mre11. J Biol Chem, 2006. 281(14): p. 9346-50. 58. Mukherjee, B., et al., DNA-PK phosphorylates histone H2AX during apoptotic DNA fragmentation in mammalian cells. DNA Repair (Amst), 2006. 5(5): p. 575-90.

136

59. Stiff, T., et al., ATM and DNA-PK function redundantly to phosphorylate H2AX after exposure to ionizing radiation. Cancer Res, 2004. 64(7): p. 2390-6. 60. Paull, T.T., et al., A critical role for histone H2AX in recruitment of repair factors to nuclear foci after DNA damage. Curr Biol, 2000. 10(15): p. 886-95. 61. Ward IM, M.K., Jorda KG, Chen J, Accumulation of checkpoint protein 53BP1 at DNA breaks involves its binding to phosphorylated histone H2AX. J Biol Chem. 2003 May 30;278(22):19579-82., 2003. 278(22): p. 19579-82. 62. Stucki, M. and S.P. Jackson, gammaH2AX and MDC1: anchoring the DNA-damage- response machinery to broken chromosomes. DNA Repair (Amst), 2006. 5(5): p. 534-43. 63. Podhorecka, M., A. Skladanowski, and P. Bozko, H2AX Phosphorylation: Its Role in DNA Damage Response and Cancer Therapy. J Nucleic Acids, 2010. 2010. 64. Celeste A, F.-C.O., Kruhlak MJ, Pilch DR, Staudt DW, Lee A, Bonner RF, Bonner WM, Nussenzweig A., Histone H2AX phosphorylation is dispensable for the initial recognition of DNA breaks. Nat Cell Biol. 2003 Jul;5(7):675-9., 2003. 5(7): p. 675-9. 65. Matsuoka, S., et al., ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science, 2007. 316(5828): p. 1160-1166. 66. Mori, D., et al., An essential role of the aPKC-Aurora A-NDEL1 pathway in neurite elongation by modulation of microtubule dynamics. Nat Cell Biol, 2009. 11(9): p. 1057- 68. 67. Kastan, M.B. and J. Bartek, Cell-cycle checkpoints and cancer. Nature, 2004. 432(7015): p. 316-23. 68. Morgan, D., The cell cycle : principles of control. 2007: New science press Ltd. 69. Perkins, D.N., et al., Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis, 1999. 20(18): p. 3551-3567. 70. Kamb, A., Cell-cycle regulators and cancer. Trends Genet, 1995. 11(4): p. 136-40. 71. Schafer, K.A., The cell cycle: a review. Vet Pathol, 1998. 35(6): p. 461-78. 72. Malumbres, M. and M. Barbacid, Cell cycle, CDKs and cancer: a changing paradigm. Nat Rev Cancer, 2009. 9(3): p. 153-66. 73. Collins, K., T. Jacks, and N.P. Pavletich, The cell cycle and cancer. Proc Natl Acad Sci U S A, 1997. 94(7): p. 2776-8. 74. John, P.C.L., M. Mews, and R. Moore, Cyclin/Cdk complexes: their involvement in cell cycle progression and mitotic division. Protoplasma, 2001. 216(3-4): p. 119-142. 75. Nigg, E.A., Cyclin-Dependent Protein-Kinases - Key Regulators of the Eukaryotic Cell- Cycle. Bioessays, 1995. 17(6): p. 471-480. 76. Paglini, G. and A. Caceres, The role of the Cdk5--p35 kinase in neuronal development. Eur J Biochem, 2001. 268(6): p. 1528-33. 77. Dhavan, R. and L.H. Tsai, A decade of CDK5. Nat Rev Mol Cell Biol, 2001. 2(10): p. 749-59. 78. Tang, D. and J.H. Wang, Cyclin-dependent kinase 5 (Cdk5) and neuron-specific Cdk5 activators. Prog Cell Cycle Res, 1996. 2: p. 205-16. 79. Hisanaga, S. and T. Saito, The regulation of cyclin-dependent kinase 5 activity through the metabolism of p35 or p39 Cdk5 activator. Neurosignals, 2003. 12(4-5): p. 221-229. 80. Cruz, J.C., et al., Aberrant Cdk5 activation by p25 triggers pathological events leading to neurodegeneration and neurofibrillary tangles. Neuron, 2003. 40(3): p. 471-83.

137

81. Angelo, M., F. Plattner, and K.P. Giese, Cyclin-dependent kinase 5 in synaptic plasticity, learning and memory. J Neurochem, 2006. 99(2): p. 353-70. 82. Tomizawa, K., et al., Cdk5/p35 regulates neurotransmitter release through phosphorylation and downregulation of P/Q-type voltage-dependent calcium channel activity. Journal of Neuroscience, 2002. 22(7): p. 2590-2597. 83. Shea, T.B., et al., Cdk5 regulates axonal transport and phosphorylation of neurofilaments in cultured neurons. Journal of Cell Science, 2004. 117(6): p. 933-941. 84. Ko, J., et al., p35 and p39 are essential for cyclin-dependent kinase 5 function during neurodevelopment. J Neurosci, 2001. 21(17): p. 6758-71. 85. Humbert, S., R. Dhavan, and L.H. Tsai, p39 activates cdk5 in neurons, and is associated with the actin cytoskeleton. Journal of Cell Science, 2000. 113(6): p. 975-983. 86. Johnson N, S.G., Cyclin-dependent kinases (cdks) and the DNA damage response: rationale for cdk inhibitor-chemotherapy combinations as an anticancer strategy for solid tumors. Expert Opin Ther Targets. 2010 Nov;14(11):1199-212, 2010. 14(11): p. 1199-212. 87. Johnson, N. and G.I. Shapiro, Cyclin-dependent kinases (cdks) and the DNA damage response: rationale for cdk inhibitor-chemotherapy combinations as an anticancer strategy for solid tumors. Expert Opinion on Therapeutic Targets, 2010. 14(11): p. 1199- 1212. 88. Ira, G., et al., DNA end resection, homologous recombination and DNA damage checkpoint activation require CDK1. Nature, 2004. 431(7011): p. 1011-7. 89. Macarulla, T., F.J. Ramos, and J. Tabernero, Aurora kinase family: a new target for anticancer drug. Recent Pat Anticancer Drug Discov, 2008. 3(2): p. 114-22. 90. Vader, G., A.F. Maia, and S.M.A. Lens, The chromosomal passenger complex and the spindle assembly checkpoint: kinetochore-microtubule error correction and beyond. Cell Division, 2008. 3. 91. Tang, C.J.C., C.Y. Lin, and T.K. Tang, Dynamic localization and functional implications of Aurora-C kinase during male mouse meiosis. Developmental Biology, 2006. 290(2): p. 398-410. 92. Carmena, M., S. Ruchaud, and W.C. Earnshaw, Making the Auroras glow: regulation of Aurora A and B kinase function by interacting proteins. Current Opinion in Cell Biology, 2009. 21(6): p. 796-805. 93. Iyer, J. and M.Y. Tsai, A novel role for TPX2 as a scaffold and co-activator protein of the Chromosomal Passenger Complex. Cellular Signalling, 2012. 24(8): p. 1677-1689. 94. Golsteyn, R.M., et al., Cell-Cycle Regulation of the Activity and Subcellular-Localization of Plk1, a Human Protein-Kinase Implicated in Mitotic Spindle Function. Journal of Cell Biology, 1995. 129(6): p. 1617-1628. 95. Petronczki M, L.P., Peters JM., Polo on the rise—From mitotic entry to cytokinesis with Plk. Dev Cell. , 2008 14(5): p. 646-59. 96. Smits VA, K.R., Arnaud L, Rijksen G, Nigg EA, Medema RH., Polo-like kinase-1 is a target of the DNA damage checkpoint. Nat Cell Biol. 2000 Sep;2(9):672-6., 2000. 2(9): p. 672-6. 97. Fu, J., Jiang, Q. Zhang, C., Collaboration of Mitotic Kinases in Cell Cycle Control. Nat Educ, 2010. 3(9): p. 82.

138

98. Hammarsten, O. and G. Chu, DNA-dependent protein kinase: DNA binding and activation in the absence of Ku. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(2): p. 525-530. 99. Reddy, Y.V.R., et al., Non-homologous end joining requires that the DNA-PK complex undergo an autophosphorylation-dependent rearrangement at DNA ends. Journal of Biological Chemistry, 2004. 279(38): p. 39408-39413. 100. Zhou, X.Y., et al., Ku affects the ATM-dependent S phase checkpoint following ionizing radiation. Oncogene, 2002. 21(41): p. 6377-81. 101. Wang, H., et al., Ku affects the ataxia and Rad 3-related/CHK1-dependent S phase checkpoint response after camptothecin treatment. Cancer Res, 2002. 62(9): p. 2483-7. 102. Wang S, G.M., Ouyang H, Li X, Cordon-Cardo C, Kurimasa A, Chen DJ, Fuks Z, Ling CC, Li GC., The catalytic subunit of DNA-dependent protein kinase selectively regulates p53-dependent apoptosis but not cell-cycle arrest. Proc Natl Acad Sci U S A. 2000 Feb 15;97(4):1584-8., 2000. 97(4): p. 1584-8. 103. Brown, N. and V. Costanzo, An ATM and ATR dependent pathway targeting centrosome dependent spindle assembly. Cell Cycle, 2009. 8(13): p. 1997-2001. 104. Shang, Z.F., et al., Inactivation of DNA-Dependent Protein Kinase Leads to Spindle Disruption and Mitotic Catastrophe with Attenuated Checkpoint Protein 2 Phosphorylation in Response to DNA Damage. Cancer Research, 2010. 70(9): p. 3657- 3666. 105. Wettenhall RE, A.R., Hood LE., Solid-phase sequencing of 32P-labeled phosphopeptides at picomole and subpicomole levels. Methods Enzymol., 1991. 201: p. 186-99. 106. Campbell, D.G. and N.A. Morrice, Identification of protein phosphorylation sites by a combination of mass spectrometry and solid phase Edman sequencing. J Biomol Tech, 2002. 13(3): p. 119-30. 107. Palm-Forster, M.A., L. Eschen-Lippold, and J. Lee, A mutagenesis-based screen to rapidly identify phosphorylation sites in mitogen-activated protein kinase substrates. Anal Biochem, 2012. 427(2): p. 127-9. 108. Bonilla, L., et al., The evolution of tools for protein phosphorylation site analysis: from discovery to clinical application. Biotechniques, 2008. 44(5): p. 671-9. 109. Nousiainen, M., et al., Phosphoproteome analysis of the human mitotic spindle. Proc Natl Acad Sci U S A, 2006. 103(14): p. 5391-6. 110. Chi, Y., et al., Identification of CDK2 substrates in human cell lysates. Genome Biology, 2008. 9(10). 111. Olsen, J.V., et al., Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis. Sci Signal, 2010. 3(104): p. ra3. 112. Carson, J.P., et al., Pharmacogenomic identification of targets for adjuvant therapy with the topoisomerase poison camptothecin. Cancer Res, 2004. 64(6): p. 2096-104. 113. Eckerdt, F., et al., Phosphorylation of TPX2 by Plx1 enhances activation of Aurora A. Cell Cycle, 2009. 8(15): p. 2413-2419. 114. Tonon, G., et al., High-resolution genomic profiles of human lung cancer. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(27): p. 9625-9630.

139

115. Wang, Y., et al., Large scale identification of human hepatocellular carcinoma- associated antigens by autoantibodies. Journal of Immunology, 2002. 169(2): p. 1102- 1109. 116. Warner, S.L., et al., Validation of TPX2 as a Potential Therapeutic Target in Pancreatic Cancer Cells. Clinical Cancer Research, 2009. 15(21): p. 6519-6528. 117. Ramakrishna, M., et al., Identification of Candidate Growth Promoting Genes in Ovarian Cancer through Integrated Copy Number and Expression Analysis. Plos One, 2010. 5(4). 118. Scotto, L., et al., Identification of copy number gain and overexpressed genes on chromosome arm 20q by an integrative genomic approach in cervical cancer: Potential role in progression. Genes Chromosomes & Cancer, 2008. 47(9): p. 755-765. 119. Zhang, Y., et al., Assignment of human proliferation associated p100 gene (C20orf1) to human chromosome band 20q11.2 by in situ hybridization. Cytogenetics and Cell Genetics, 1999. 84(3-4): p. 182-183. 120. Beroukhim, R., et al., The landscape of somatic copy-number alteration across human cancers. Nature, 2010. 463(7283): p. 899-905. 121. Ma Y, L.D., Sun W, Xiao T, Yuan J, Han N, Guo S, Feng X, Su K, Mao Y, Cheng S, Gao Y., Expression of targeting protein for xklp2 associated with both malignant transformation of respiratory epithelium and progression of squamous cell lung cancer. Clin Cancer Res. , 2006. 12(4): p. 1121-7. 122. Shigeishi, H., et al., Expression of TPX2 in salivary gland carcinomas. Oncol Rep, 2009. 21(2): p. 341-4. 123. Li, B., et al., Expression of targeting protein for Xenopus kinesin-like protein 2 is associated with progression of human malignant astrocytoma. Brain Research, 2010. 1352: p. 200-207. 124. Cordes, C., et al., Prognostic relevance of the proliferation marker REPP86 for laryngeal cancer. Anticancer Res, 2010. 30(9): p. 3541-7. 125. Wei P, Z.N., Xu Y, Li X, Shi D, Wang Y, Li D1, Cai S., TPX2 is a novel prognostic marker for the growth and metastasis of colon cancer. J Transl Med. , 2013. 11(313). 126. Carter, S.L., et al., A signature of chromosomal instability inferred from gene expression profiles predicts clinical outcome in multiple human cancers. Nature Genetics, 2006. 38(9): p. 1043-1048. 127. Cahill, D.P., et al., Mutations of mitotic checkpoint genes in human cancers. Nature, 1998. 392(6673): p. 300-303. 128. Lassmann, S., et al., Predictive value of Aurora-A/STK15 expression for late stage epithelial ovarian cancer patients treated by adjuvant chemotherapy. Clinical Cancer Research, 2007. 13(14): p. 4083-4091. 129. Sotillo, R., et al., Mad2 overexpression promotes aneuploidy and tumorigenesis in mice. Cancer Cell, 2007. 11(1): p. 9-23. 130. Chen, R.Q., et al., CDC25B Mediates Rapamycin-induced Oncogenic Responses in Cancer Cells. Cancer Research, 2009. 69(6): p. 2663-2668. 131. Dephoure, N., et al., A quantitative atlas of mitotic phosphorylation. Proc Natl Acad Sci U S A, 2008. 105(31): p. 10762-7. 132. Malik, R., et al., Quantitative analysis of the human spindle phosphoproteome at distinct mitotic stages. J Proteome Res, 2009. 8(10): p. 4553-63.

140

133. Rigbolt, K.T.G., et al., System-Wide Temporal Characterization of the Proteome and Phosphoproteome of Human Embryonic Stem Cell Differentiation. Science Signaling, 2011. 4(164). 134. Weber, C., T.B. Schreiber, and H. Daub, Dual phosphoproteomics and chemical proteomics analysis of erlotinib and gefitinib interference in acute myeloid leukemia cells. J Proteomics, 2012. 75(4): p. 1343-56. 135. Shiromizu, T., et al., Identification of Missing Proteins in the neXtProt Database and Unregistered Phosphopeptides in the PhosphoSitePlus Database As Part of the Chromosome-Centric Human Proteome Project. Journal of Proteome Research, 2013. 12(6): p. 2414-2421. 136. Mayya, V., et al., Quantitative phosphoproteomic analysis of T cell receptor signaling reveals system-wide modulation of protein-protein interactions. Sci Signal, 2009. 2(84): p. ra46. 137. Dephoure, N., et al., A quantitative atlas of mitotic phosphorylation. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(31): p. 10762- 10767. 138. Delom, F. and E. Chevet, Phosphoprotein analysis: from proteins to proteomes. Proteome Science, 2006. 4. 139. Havlicek, L., et al., Cytokinin-derived cyclin-dependent kinase inhibitors: synthesis and cdc2 inhibitory activity of olomoucine and related compounds. J Med Chem, 1997. 40(4): p. 408-12. 140. De Azevedo, W.F., et al., Inhibition of cyclin-dependent kinases by purine analogues: crystal structure of human cdk2 complexed with roscovitine. Eur J Biochem, 1997. 243(1-2): p. 518-26. 141. Meijer, L., et al., Biochemical and cellular effects of roscovitine, a potent and selective inhibitor of the cyclin-dependent kinases cdc2, cdk2 and cdk5. Eur J Biochem, 1997. 243(1-2): p. 527-36. 142. Mapelli, M., et al., Mechanism of CDK5/p25 binding by CDK inhibitors. J Med Chem, 2005. 48(3): p. 671-9. 143. Schultz, C., et al., Paullones, a series of cyclin-dependent kinase inhibitors: synthesis, evaluation of CDK1/cyclin B inhibition, and in vitro antitumor activity. J Med Chem, 1999. 42(15): p. 2909-19. 144. Skoufias, D.A., et al., Mitosis persists in the absence of Cdk1 activity when proteolysis or protein phosphatase activity is suppressed. Journal of Cell Biology, 2007. 179(4): p. 671- 685. 145. Wu CF, W.R., Liang Q, Liang J, Li W, Jung SY, Qin J, Lin SH, Kuang J., Dissecting the M Phase–specific Phosphorylation of Serine–Proline or Threonine–Proline Motifs. Mol Biol Cell., 2010. 21(9): p. 1470-81. 146. Ma N, M.S., Takata H, Ono-Maniwa R, Uchiyama S, Fukui K., Nucleolin functions in nucleolus formation and chromosome congression. J Cell Sci. , 2007. 120(Pt12): p. 2091- 105. 147. Losiewicz MD, C.B., Kaur G, Sausville EA, Worland PJ, Potent inhibition of CDC2 kinase activity by the flavonoid L86-8275. Biochem Biophys Res Commun, 1994. 201(2)(2): p. 589-95.

141

148. Potapova, T.A., et al., The reversibility of mitotic exit in vertebrate cells. Nature, 2006. 440(7086): p. 954-958. 149. Niiya, F., et al., Inhibition of cyclin-dependent kinase 1 induces cytokinesis without chromosome segregation in an ECT2 and MgcRacGAP-dependent manner. Journal of Biological Chemistry, 2005. 280(43): p. 36502-36509. 150. Gray, N., et al., ATP-site directed inhibitors of cyclin-dependent kinases. Current Medicinal Chemistry, 1999. 6(9): p. 859-875. 151. Hong, K.U., et al., Cdk1-cyclin B1-mediated phosphorylation of tumor-associated microtubule-associated protein/cytoskeleton-associated protein 2 in mitosis. J Biol Chem, 2009. 284(24): p. 16501-12. 152. Koivomagi, M., et al., Dynamics of Cdk1 substrate specificity during the cell cycle. Mol Cell, 2011. 42(5): p. 610-23. 153. Ubersax, J.A., et al., Targets of the cyclin-dependent kinase Cdk1. Nature, 2003. 425(6960): p. 859-64. 154. Eyers, P.A. and J.L. Maller, Regulation of Xenopus aurora A activation by TPX2. Journal of Biological Chemistry, 2004. 279(10): p. 9008-9015. 155. Silkworth WT, N.I., Scholl LM, Cimini D., Multipolar Spindle Pole Coalescence Is a Major Source of Kinetochore Mis-Attachment and Chromosome Mis-Segregation in Cancer Cells. PLoS One, 2009. 4(8): p. e6564. 156. Berger, S.M., et al., Quantitative analysis of conditional gene inactivation using rationally designed, tetracycline-controlled miRNAs. Nucleic Acids Research, 2010. 38(17). 157. van Vugt, M.A.T.M., et al., A Mitotic Phosphorylation Feedback Network Connects Cdk1, Plk1, 53BP1, and Chk2 to Inactivate the G2/M DNA Damage Checkpoint. Plos Biology, 2010. 8(1). 158. Summers, K.C., et al., Phosphorylation: the molecular switch of double-strand break repair. Int J Proteomics, 2011. 2011: p. 373816. 159. Xu, N., et al., Cdk-mediated phosphorylation of Chk1 is required for efficient activation and full checkpoint proficiency in response to DNA damage. Oncogene, 2012. 31(9): p. 1086-1094. 160. Xu, N., et al., Cdk phosphorylation of Chk1 regulates efficient Chk1 activation and multiple checkpoint proficiency. Biochem Biophys Res Commun, 2011. 413(3): p. 465- 70. 161. Oakley, G.G., et al., RPA phosphorylation in mitosis alters DNA binding and protein- protein interactions. Biochemistry, 2003. 42(11): p. 3255-64. 162. Esashi, F., et al., CDK-dependent phosphorylation of BRCA2 as a regulatory mechanism for recombinational repair. Nature, 2005. 434(7033): p. 598-604. 163. Cho HJ, O.Y., Han SH, Chung HJ, Kim CH, Lee NS, Kim WJ, Choi JM, Kim H., Cdk1 protein-mediated phosphorylation of receptor-associated protein 80 (RAP80) serine 677 modulates DNA damage-induced G2/M checkpoint and cell survival. J Biol Chem. 2013 Feb 8;288(6):3768-76., 2013. 288(6): p. 3768-76. 164. Beausoleil, S.A., et al., Large-scale characterization of HeLa cell nuclear phosphoproteins. Proc Natl Acad Sci U S A, 2004. 101(33): p. 12130-5.

142

165. Christensen, G.L., et al., Quantitative Phosphoproteomics Dissection of Seven- transmembrane Receptor Signaling Using Full and Biased Agonists. Molecular & Cellular Proteomics, 2010. 9(7): p. 1540-1553. 166. Saffin, J.M., et al., ASAP, a human microtubule-associated protein required for bipolar spindle assembly and cytokinesis. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(32): p. 11302-11307. 167. Basbous, J., et al., Induction of ASAP (MAP9) contributes to p53 stabilization in response to DNA damage. Cell Cycle, 2012. 11(12): p. 2380-90. 168. Rogakou, E.P., et al., DNA double-stranded breaks induce histone H2AX phosphorylation on serine 139. J Biol Chem, 1998. 273(10): p. 5858-68. 169. Chang H, W.J., Tian Y, Xu J, Gou X, Cheng J., The TPX2 gene is a promising diagnostic and therapeutic target for cervical cancer. Oncol Rep. , 2012. 27(5): p. 1353-9. 170. Grover, A., et al., Ashwagandha derived withanone targets TPX2-Aurora A complex: computational and experimental evidence to its anticancer activity. PLoS One, 2012. 7(1): p. e30890. 171. Vainio, P., et al., High-Throughput Transcriptomic and RNAi Analysis Identifies AIM1, ERGIC1, TMED3 and TPX2 as Potential Drug Targets in Prostate Cancer. Plos One, 2012. 7(6). 172. Ayala, G., et al., High levels of phosphorylated form of Akt-1 in prostate cancer and non- neoplastic prostate tissues are strong predictors of biochemical recurrence. Clinical Cancer Research, 2004. 10(19): p. 6572-6578. 173. Asteriti, I.A., et al., Aurora-A inactivation causes mitotic spindle pole fragmentation by unbalancing microtubule-generated forces. Mol Cancer, 2011. 10: p. 131. 174. De Luca, M., et al., Aurora-A and ch-TOG act in a common pathway in control of spindle pole integrity. Oncogene, 2008. 27(51): p. 6539-49. 175. Barr, A.R. and F. Gergely, MCAK-independent functions of ch-Tog/XMAP215 in microtubule plus-end dynamics. Mol Cell Biol, 2008. 28(23): p. 7199-211. 176. Cassimeris, L., B. Becker, and B. Carney, TOGp regulates microtubule assembly and density during mitosis and contributes to chromosome directional instability. Cell Motil Cytoskeleton, 2009. 66(8): p. 535-45. 177. Silva, P., et al., Monitoring the fidelity of mitotic chromosome segregation by the spindle assembly checkpoint. Cell Prolif, 2011. 44(5): p. 391-400. 178. Nakanishi, S., et al., Wortmannin, a microbial product inhibitor of myosin light chain kinase. J Biol Chem, 1992. 267(4): p. 2157-63. 179. Sarkaria, J.N., et al., Inhibition of phosphoinositide 3-kinase related kinases by the radiosensitizing agent wortmannin. Cancer Res, 1998. 58(19): p. 4375-82. 180. Lee, H.J., H.I. Hwang, and Y.J. Jang, Mitotic DNA damage response Polo-like kinase-1 is dephosphorylated through ATM-Chk1 pathway. Cell Cycle, 2010. 9(12): p. 2389-2398. 181. Blom, N., et al., Prediction of post-translational glycosylation and phosphorylation of proteins from the amino acid sequence. Proteomics, 2004. 4(6): p. 1633-49. 182. Malik R, L.R., Santamaria A, Ries A, Nigg EA, Körner R., Quantitative analysis of the human spindle phosphoproteome at distinct mitotic stages. J Proteome Res., 2009. 8(10): p. 4553-63.

143

183. Molina, H., et al., Global proteomic profiling of phosphopeptides using electron transfer dissociation tandem mass spectrometry. Proceedings of the National Academy of Sciences of the United States of America, 2007. 104(7): p. 2199-2204. 184. Daub, H., et al., Kinase-selective enrichment enables quantitative phosphoproteomics of the kinome across the cell cycle. Molecular Cell, 2008. 31(3): p. 438-448. 185. Cantin, G.T., et al., Combining protein-based IMAC, peptide-based IMAC, and MudPIT for efficient phosphoproteomic analysis. J Proteome Res, 2008. 7(3): p. 1346-51. 186. Wang, Z.H., et al., Extensive Crosstalk Between O-GlcNAcylation and Phosphorylation Regulates Cytokinesis. Science Signaling, 2010. 3(104). 187. Beausoleil, S.A., et al., A probability-based approach for high-throughput protein phosphorylation analysis and site localization. Nat Biotechnol, 2006. 24(10): p. 1285-92. 188. Blethrow, J.D., et al., Covalent capture of kinase-specific phosphopeptides reveals Cdk1- cyclin B substrates. Proc Natl Acad Sci U S A, 2008. 105(5): p. 1442-7. 189. Brill, L.M., et al., Phosphoproteomic analysis of human embryonic stem cells. Cell Stem Cell, 2009. 5(2): p. 204-13. 190. Hegemann, B., et al., Systematic Phosphorylation Analysis of Human Mitotic Protein Complexes. Science Signaling, 2011. 4(198). 191. Hsu, P.P., et al., The mTOR-Regulated Phosphoproteome Reveals a Mechanism of mTORC1-Mediated Inhibition of Growth Factor Signaling. Science, 2011. 332(6035): p. 1317-1322. 192. Iliuk, A.B., et al., In-depth Analyses of Kinase-dependent Tyrosine Phosphoproteomes Based on Metal Ion-functionalized Soluble Nanopolymers. Molecular & Cellular Proteomics, 2010. 9(10): p. 2162-2172. 193. Lowery, D.M., et al., Proteomic screen defines the Polo-box domain interactome and identifies Rock2 as a Plk1 substrate. Embo Journal, 2007. 26(9): p. 2262-2273. 194. Oppermann, F.S., et al., Large-scale Proteomics Analysis of the Human Kinome. Molecular & Cellular Proteomics, 2009. 8(7): p. 1751-1764. 195. Tsai, C.F., et al., Immobilized metal affinity chromatography revisited: pH/acid control toward high selectivity in phosphoproteomics. J Proteome Res, 2008. 7(9): p. 4058-69.

144

APPENDIX

Appendix Figure 1 Sequence coverage and phosphorylation sites identified by LC-MS/MS analysis. Amino acid sequences of hTPX2 showing 44% sequence coverage. Peptides identified from LC- MS/MS analysis are underlined in bold and phosphorylation sites of TPX2 identified are marked in bold red with the exact phosphorylation site number. The site marked in bold blue (S102) is a novel site.

145

146

147

148

Appendix Figure 2 Multiple sequence alignment of TPX2 protein from different species and the location of phosphorylation sites of TPX2 from Phosphosite [4] & my own mass spectrometry analysis data. Human TPX2 protein sequences were aligned with the sequences of TPX2 protein orthologs from other species (mouse, rat and frog) using DNAMAN software (Lynnon Corperation.). The identical amino acid sequences of all four species are shaded in light blue. The phosphorylation sites identified from my own mass spectrometry analysis were underlined. The amino acids shaded in gray and lilac color show that the sequences of amino acids among different species are not identical, but have similar characteristic (called as amino acid similarity).One phosphorylation site underlined and shaded in yellow (S102) is a novel phosphorylation site that never been previously reported in any literature or proteomic screenings. The alignment clearly shows that Thr72 is well conserved among different species.

149

Appendix Figure 3 The localization of pThr-72 protein during mitosis pThr-72 is localized on a layer around mitotic spindle, which might be localized at the chromosome periphery. HeLa cells were co-stained with pThr-72 and nucleolin antibodies.

150

Appendix Table 1 Matching kinase motif information of the phosphorylation sites of TPX2 identified from my own mass spectrometry data analysis (based on the phosphopeptide sequences) [4, 181]

Sequence Site (Phospho Matching kinase motif References peptide) T59 KTPLR Cdk [131]

T72 VTPLK Cdk2, [109-111, 130, 131,

P38MAPK/Cdk5 134, 136, 182]

S102 SSLEV Cdk1 (Cdc2) A novel site

S121 RSLRL PKC,PKA, GSK3 [109, 111, 133, 165,

Aurora A 183]

S125 LSAQ PKC,PKA,CK1, [111, 130, 132, 133,

NEK6 165]

S185 ASSPE CK2 [184]

S186 SSPEK CK1 [109]

S257 KSVDF PKC [131, 185]

S293 SSPAR Cdks (1,2 and 5) [109, 111, 165, 184]

MAPK, ERK,GSK3

S310 LSQGk ATM/ATR, [131]

PKC, DNA-PK

S359 SSVTK Aurora A, PKC [109, 185]

151

T369 QTPVL MAPK, ERK, [111, 130, 185]

P38MAPK

S486 KSPAF MAPK/Cdk [109, 111, 134, 186,

Cdk,GSK3,CK1, 187]

ERK/MAPK

S652 LSGSL n/a [109]

S654 GSLVQ n/a [109]

S738 VSPKF MAPK/Cdk [109, 111, 132-134,

CK1,GSK3, ERK, 136, 164, 165, 186-

ERK/MAPK 195]

152

Appendix Table 2 P values of ANOVA and Neuman-Keulus tests in the experiments in

Figure 4.1

P values of P values of Neuman-Keuls test # of Mitotic ANOVA TEST spindle comparing Three GFP vector GFP-TPX2 vs. GFP vector vs. groups vs. GFP-T72A GFP-T72A TPX2 mutant mutant Monopolar P<0.05 P>0.05 (NS) P>0.05 (NS) P<0.05

Bipolar P<0.01 P<0.05 P>0.05 (NS) P<0.01

Multipolar P<0.001 P>0.05 (NS) P<0.001 P<0.001

153

Appendix Table 3 P values of ANOVA and Neuman-Keulus tests in the experiments in

Figure 4.2

P values of Neuman-Keuls test P values of ANOVA GFP vector + GFP-TPX2 + # of Mitotic GFP vector TEST Control TPX2 siRNA TPX2 siRNA spindle vs. comparing siRNA vs. vs. GFP-T72A four groups vs GFP TPX2 GFP-T72A mutant + GFP+TPX2 WT+TPX2 mutant + TPX2 siRNA siRNA siRNA TPX2 siRNA Monopolar P>0.05 (NS) NS NS NS NS

Bipolar P>0.05 (NS) NS NS NS NS

Multipolar P<0.01 P<0.05 P>0.05 (NS) P<0.05 P<0.05

154