Cytoskeleton-Associated Protein 2 Is Required for the Maintenance of Chromosomal Stability by Tethering Spindle Microtubules to Their Poles

Cytoskeleton-Associated Protein 2 is Required for the Maintenance of Chromosomal Stability by Tethering Spindle Microtubules to their Poles

by Chanelle M. Case

B.S. in Biology, May 2006, Villanova University

A Dissertation submitted to

The Faculty of The Columbian College of Arts and Sciences of The George Washington Unviersity in partial fulfillment of the requirements for the degree of Doctor of Philosophy

January 31, 2013

Dissertation directed by

Thomas Ried Chief, Section of Cancer Genomics, National Cancer Institute, National Institutes of Health

Norman Lee Professor of Pharmacology and Physiology

The Columbian College of Arts and Sciences of The George Washington University certifies that Chanelle M. Case has passed the Final Examination for the degree of Doctor of Philosophy as of January 31, 2013. This is the final and approved form of the dissertation.

Cytoskeleton-Associated Protein 2 is Required for the Maintenance of Chromosomal Stability by Tethering Spindle Microtubules to their Poles Chanelle M. Case

Dissertation Research Committee:

Thomas Ried, Chief of the Section for Cancer Genomics, National Cancer Insitute, National Institutes of Health, Dissertation Co- Director

Norman Lee, Professor of Pharmacology and Physiology, Dissertation Co-Director

Dan Sackett, Staff Scientist, Section on Cell Biophysics, National Institute of Child Health and Human Development, National Institutes of Health, Committee Member

Susan Ceryak, Associate Research Professor of Pharmacology and Physiology, Committee Member

iii

Dedication

I dedicate my dissertation to my parents and grandparents, for their love, prayers and support. It is on their shoulders I stand, and I would not have completed this work without their unwavering support of my passion for science.

I also dedicate this project to my best friend and wonderful husband, John

Borden, for encouraging me throughout my academic career to be the best scientist I can be. And although my grandfather, John E. Couch, Jr. did not live to see me earn my doctorate, I thank him for always believing me, and I know he is truly proud of his granddaughter.

Acknowledgements

I would like to thank Dr. Janice Knepper and Dr. Mary Desmond for nurturing my love for science, and shepherding me through my first research experience at Villanova University. Equally important to my development were

Dr. Carl June and Dr. Jim Riley, and the entire June lab, who gave me the opportunity to continue developing my skills as a Research Associate in their lab at the University of Pennsylvania. I would like to especially thank Dr. Angel

Varela-Rohena and Dr. Samik Basu for being great mentors and friends.

I owe a special debt of gratitude to Dr. Thomas Ried, and the Ried lab for opening their minds and hearts to me as I pursed my doctoral degree. In particular, I am grateful for the wisdom and mentorship of Dr. Jordi Camps, who helped me brainstorm, develop, edit and perfect my ideas. I would also like to acknowledge the current and former members of the Ried lab, especially Dr.

Danny Wangsa, Dr. Dara Wangsa, Dr. Michael Difillippantonio and Dr. Kundan

Sengupta, for their assistance and encouragement as I completed my dissertation. Dr. Dan Sackett deserves special acknowledgment for his advice, assistance and patience in answering the many questions I had throughout this project.

I would like to thank my co-mentor, Dr. Normal Lee, for his encouragement, patience and guidance. I also would like to acknowledge my dissertation committee members and participants, Dr. Dan Sackett, Dr. Stan

Lipkowitz, Dr. Susan Ceryak, Dr. Travis O’Brien, and Dr. Daniela Cimini. Their

v insightful questions and comments were invaluable in rightly guiding my finished product. I am deeply appreciative of Dr. Anne Chiaramello for chairing my dissertation, and the other faculty and staff members that were instrumental in making this a remarkable and memorable experience, including Dr. Linda

Werling, Marc Witliff, Amanda Page, and the members of the Department of

Pharmacology and Physiology. I would also like to thank the OITE staff at the

National Institutes of Health for their support assistance throughout this process, especially Dr. Sharon Milgram.

Abstract of Dissertation

Cytoskeleton-Associated Protein 2 is Required for the Maintenance of Chromosomal Stability by Tethering Spindle Microtubules to their Poles Errors in chromosome segregation lead to aneuploidy. Integrity of the microtubule spindle apparatus and intact cell division checkpoints are essential to ensure the fidelity of chromosome distribution into daughter cells. Cytoskeleton- associated protein 2, CKAP2, is a microtubule-associated protein that colocalizes with spindle poles and aids in microtubule stabilization, but the exact function and mechanism of its action are poorly understood. In the present study, RNA interference was utilized to determine the extent to which the expression of

CKAP2 plays a role in chromosome segregation in colorectal cancer cells.

CKAP2-depleted cells showed a significant increase of multi-polar mitoses and other spindle pole aberrations. Notably, when interrogated for microtubule nucleation capacity, CKAP2-depleted cells showed a very unusual phenotype as early as two minutes after release from mitotic block, consisting of dispersal of newly polymerized microtubule filaments through the entire chromatin region.

Nevertheless, spindle poles were formed after one hour of mitotic release, suggesting that centrosome-mediated nucleation remained dominant.

Kinetochore-driven microtubule nucleation was not implicated, as there was no colocalization of nascent microtubule filaments with Hec1. Interestingly, there was no effect on the localization of nuclear mitotic apparatus (NuMA) protein in

CKAP2-depleted cells. Finally, we showed that suppression of CKAP2 results in a higher incidence of merotelic attachments, anaphase lagging, and

vii chromosomal instability. In conclusion, CKAP2 is involved in tethering the microtubule-minus ends to the spindle pole in early mitosis. Delays in this process may alter the mitotic spindle tension, ultimately promoting merotelic kinetochore- microtubule attachments that result in chromosome lagging and increased chromosomal instability.

viii

Table of Contents

Dedication………………………………………………………………………...... iv

Acknowledgements……………………………………………………………………..v

Abstract of Dissertation………………………………………………………………...vi

List of Figures…………………………………………………………………………....x

List of Tables…………………………………………………………………………….xi

List of Abbreviations……………………………………………………………………xii

Chapter 1: Introduction…………………………………………………………………1

Section I. Components of the Mitotic Spindle……..……….……………...2 Section II. Importance of Microtubule Dynamics in Chromosome Segregation…………...... 4 Section III. MAPs Influence Spindle Pole Integrity and Chromosome Missegregation…………………………………………...……..6 Section IV. Microtubule-Associated Protein, CKAP2, May Influence Microtubule Dynamics and Maintenance of the Genome.....7 Section V. Dissertation goals………………………………………………..12

Chapter 2: Materials and Methods

Chapter 3: Results

Section I. CKAP2 Expression and Localization in Wild-Type DLD1…...24 Section II. Establishment of shCKAP2 Model……………………………..28 Section III. Evaluation of CKAP2 Function……...………………………….38 Section IV. Differential Expression of CKAP2 in Human and Mouse Cancer Cell Lines…………………………………………….. 65

Chapter 4: Discussion

Section I. Depletion of CKAP2 Increases Spindle Pole Defects…..……73 Section II. CKAP2 Plays a Role in Tethering the Centrosome to the Spindle Pole……………………………………………………75 Section III. Loss of Spindle Pole Integrity Results in Chromosome Missegregation………………………………………………...78 Section IV. CKAP2 in the context of cancer………………………………..81

References……………………………………………………………………………..84

List of Figures

Figure 1……....Diagram of pGIPZ-shCKAP2 plasmid……………………………. 14 Figure 2.……...Schema of Apoptosis Assays……………………………………... 17 Figure 3………CKAP2 Expression is Restricted to Mitosis……………………… 26 Figure 4………CKAP2 Localizes to and is Associated with the Mitotic Spindle. 27 Figure 5.……...CKAP2 is Not Essential for Cell Viability………………………… 30 Figure 6.……...Generation of Stable shCKAP2 Cell Lines .……………………. 31 Figure 7.……...Validation of CKAP2 Knock-down………………………………... 32

Figure 8………Stable CKAP2 Depletion does not Influence Cell Proliferation or Viability…………………………………………….. 33 Figure 9.……...Depletion of CKAP2 Expression does not Affect Cell Cycle Distribution………………………………………………………….. 36

Figure 10.…….CKAP2 Depletion Results in a Decrease in the Length of Mitosis……………………………………………………………….. 37 Figure 11.…….Reduction of CKAP2 Expression Results in an Increase in Multipolar Spindles………………………………………………… 41

Figure 12.…….CKAP2 Depletion Results in a Dispersal of γ-tubuling Away from the Centrosome……………………………………………… 42

Figure 13……..CKAP2 Depletion Results in an Increase in Centrosome Dislocation………………………………………………………….. 43 Figure 14……..Spindle Pole Defects Result in an Increase in Spindle Tension And Chromosome Misalignment in Metaphase………………… 44

Figure 15……..CKAP2 is Required for the Anchoring of Centrosome- Nucleated Microtubules to the Spindle Pole……………………. 47

Figure 16…….. CKAP2 is Required for the Anchoring of Centrosome- Nucleated Microtubules to the Spindle Pole…………………… 48

Figure 17……..Depletion of CKAP2 Causes a Delay in Spindle Formation…... 49 Figure 18……..CKAP2-Depleted Cells are Capable of Bipolar Spindle Formation…………………………………………………………… 50

Figure 19……..Total Nucleation Capacity is not Reduced in CKAP2 Depleted Cells……………………………………………………………….... 51

Figure 20……..Reduction of CKAP2 Expression does not Enhance Chromosome-Directed Nucleation……………………………… 55

Figure 21……..Depletion of CKAP2 does not Affect NuMA Expression and Localization………………………………………………………… 56 Figure 22……..CKAP2 and NuMA Both Localize to the Mitotic Spinde Pole... 57 Figure 23……..Localization of NuMA is not Affected by CKAP2 Depletion….. 58

Figure 24……..Reduction of CKAP2 Results in an Increase in Chromosome Missegregation……………………………………………………. 61

Figure 25……..Reduction of CKAP2 Results in an Increase in Merotelic Attachments……………………………………………………….. 62

Figure 26……..Increased Chromosome Lagging in CKAP2-depleted Cells Results in Abnormal Nuclei Morphology………………………. 63

Figure 27……..Increased Chromosome Missegregation Ultimately Results In Chromosomal Instability………………………………………. 64

Figure 28…….Spectral Karyotyping of CKAP2-depleted Cells………………... 65

Figure 29…….Differential Expression of CKAP2 in Colorectal Cancer Cell Lines…………………………………………………………….. 68

Figure 30…….MIN+ CRC Cell Lines Arrest in Mitosis after Nocodazole Treatment…………………………………………………………… 69

Figure 31…….CIN+ CRC Cell Lines Arrest in Mitosis after Nocodazole Treatment…………………………………………………………… 70

Figure 32…….CKAP2 Expression in MIN+ and CIN+ CRC Cell Lines in Mitosis……………………………………………………………... 71

xii

List of Tables

Table 1.………CKAP2 Expression and Relative Ploidy of Colorectal Cancer Cell Lines…………………………………………………………… 72

xiii

List of Abbreviations

APC Adenomatus polyposis coli

APC/C anaphase-promoting complex/cyclosome cDNA complementary DNA

CDK cyclin dependent kinase

CENP centromere-protein

CIN chromosomal instability

CKAP2 cytoskeleton-associated protein 2

CKAP5 cytoskeletal-associated protein 5

CLIP70 class II-associated invariant chain peptide 70

CT cycle threshold

CMV cytomegalovirus

DAPI 4’,6-diamidino-2-phenylindole

EB1 end binding 1

FACS fluorescence-activated cell sorting

GFP green fluorescent protein

GTP guanosine-triphosphate

Hice1 hec1-interacting centrosome associated protein 1 mRNA messenger RNA

MCAK microtubule-modulating protein

MIN microsatellite instability

NEB nuclear envelope breakdown

xiv

NuMA nuclear mitotic apparatus

PCR polymerase chain reaction

PUMA p53 upregulated modulator of apoptosis

RT-PCR real-time polymerase chain reaction

SAC spindle assembly checkpoint siRNA short-interfering RNA shRNA short-hairpin RNA shCKAP2 cells transfected with shRNA against CKAP2

TMAP tumor microtubule-associated protein

TPX2 targeting protein for Xklp2

VC vector control for CKAP2 plasmid

ZW10 zeste white 10

7-AAD 7-amino-actinomycin D

Chapter 1: Introduction

Chromosomal segregation in mitosis is governed by a complex microtubule-based structure arranged in a symmetrical bipolar spindle with centrosomes located at the spindle poles. Normally, centrosomes nucleate microtubules and these converge into two spindle poles as a consequence of a minus end-directed microtubule sliding activity present in the spindles (Merdes and Cleveland, 1997). The proper process of distributing the correct number of chromosomes into two daughter cells during mitosis depends on numerous proteins, including those responsible for the organization of the spindle poles.

Often, depletion or malfunctioning of any of these proteins results in activation of mitotic checkpoints whose function is to arrest cell cycle progression when chromosomes are not properly aligned or attached to the spindles (Elledge,

1996). The consequences of impairment of this process is aneuploidy, a hallmark of many cancers, in particular of tumors of epithelial origin, i.e., carcinomas (Ried et al., 2012). The increased rate of gains and losses of whole chromosomes constitutes a phenomenon referred as chromosomal instability (Lengauer et al.,

1997). Chromosomal instability frequently correlates with the presence of multiple centrosomes and increased rates of chromosome lagging (Ghadimi et al., 2000; Thompson et al., 2010). Ultimately, chromosomal instability seems to enable cells to adapt chromosome content to improve their fitness (Heselmeyer-

Haddad et al., in press). Thus, it is important to further elucidate mechanisms of chromosomal instability in order to fully understand tumor progression.

Section I. Components of the Mitotic Spindle

The mitotic spindle is the essential component of mitosis that ensures proper chromosome segregation. It is composed of two centrosomes and microtubules that form a spindle apparatus, which surrounds the chromosomes.

The centrosome, also known as the microtubule-organizing center (MTOC) in animal cells, is the major spindle organizer, recognized for its MT nucleation capacity. The centrosome is composed of a pair of barrel-shaped centrioles surrounded by amorphous pericentriolar material (PCM). The major component of the centrosome is γ-tubulin, which is recruited from a cytoplasmic pool of soluble γ-tubulin during centrosome maturation in G2 (Wiese and Zheng, 2006).

Unlike α/β,, γ-tubulin is not incorporated into microtubules. Rather it is associated with additional proteins to form complexes of two different sizes, γ- tubulin small complex (γ-TuSC) and γ-tubulin ring complex (γ-TuRC) that are recruited to the centrosome and nucleate microtubule polymers (Raynoud-

Messina and Merdes, 2007). The PCM consists of pericentrin and additional anchoring proteins, which form a lattice-like structure (Blagden and Glover,

2003). Embedded in the PCM are the centrioles, which consist of triplets of microtubules, centrin, and other proteins, and are primarily responsible for the organization of the PCM. Together, the components of PCM are responsible for microtubule nucleation, where the microtubule array grows with the plus ends outward.

Microtubules are composed of α-tubulin and β-tubulin subunits, which form a α/β-tubulin heterodimer that assembles into a protofilament in a head to tail (αβ-αβ) fashion. Both α- /β-tubulin possess guanine-nucleotide-binding capacity, which alters the conformational shape of the dimer. This structure confers polarity, with slow-growing α-tubulin “minus” ends and faster-growing β- tubulin “plus” ends (Nogales,1999; Valiron et al, 2001; Kline-Smith and Walczak,

2004). The dynamic and structural properties of microtubules give insight into the mechanism of mitosis (Walczak and Heald, 2008). The conformation of each tubulin molecule facilitates microtubule polymerization and depolymerization

(Wiese and Zheng, 2008). As a result, microtubules undergo dynamic instability, in which individual microtubules go through phases of growth and shrinkage, with transitions known as “catastrophe” (growth to shrinkage) and “rescue” (shrinkage to growth) (Walker et al., 1988). However, there are also microtubule-associated proteins (MAPs), some of which enhance and some of which decrease the stability of the spindle, as well as specific regulation of the spindle.

These properties govern the formation of the mitotic spindle. The less dynamic microtubule minus ends gather at the spindle poles, while the more dynamic plus ends extend outward (Kline-Smith and Walczak, 2004). There are three distinct types of microtubules that comprise the mitotic spindle: astral microtubules, interpolar microtubules, and kinetochore microtubules. Astral and interpolar microtubules are primarily responsible for the proper orientation and positioning of the mitotic spindle. Astral microtubules are highly dynamic and form radial arrays around the centrosome (Kline-Smith and Walczak, 2004).

Interpolar microtubules extend from the spindle poles, some into the spindle midzone aiding in stabilization of the spindle apparatus (Scholey et al., 2003;

Kline-Smith and Walczak, 2004). Kinetochore microtubules, or K-fibers, are primarily responsible for attaching to sister chromatids via the kinetochore, allowing chromosomes to align and segregate (Rieder and Salmon, 1998; Kline-

Smith and Walczak, 2004). The dynamic and structural properties of microtubules give insight into the mechanism of mitosis. Microtubule polarity is essential to the movement of microtubule motors, such as dynein and kinesin family proteins, which play a major role in centrosome separation and the formation of a bipolar spindle pole (Walczak and Heald, 2008).

Section II: Importance of Microtubule Dynamics in Chromosome Segregation

The rate of microtubule turnover is significantly higher in mitosis when compared to interphase. The cell utilizes increased dynamics to promote microtubule attachments to the chromosomes in a process known as “search and capture”. In early mitosis, centrosome-nucleated microtubules rapidly depolymerize and regrow until they are stabilized by attachment to a kinetochore on a sister chromatid. This mono-oriented pair of chromatids oscillates until the other kinetochore is successfully attached to microtubules from the opposite spindle pole (Walczak and Heald, 2008). Once bi-orientation is established, the chromosome aligns to the metaphase plate. Chromosome alignment or congression is largely due to polymerization and depolymerization of microtubule

4 plus ends at the kinetochore, which creates a push-pull system that moves chromosomes toward the opposite poles, and motor proteins which move chromosomes along spindle microtubules to the metaphase plate. (Skibbens et al, 1993; Kline-Smith and Walczak, 2004; Walczak and Heald, 2008). This occurs until all of the chromosomes are successfully attached to the spindle poles by kinetochore microtubules. This process is supervised by a tension- monitored spindle assembly checkpoint (SAC). The entire process is mediated by proteins within the kinetochore, namely structural proteins (CENP-A, CENP-B, and CENP-C), motor proteins (dynein and CENP-E), microtubule-modulating protein MCAK, checkpoint signaling proteins (Mad2 and Bub1), and additional proteins including CENP-F, ZW10, and CLIP170 (Compton, 2000).

Microtubule-associated proteins, MAPs, and motor proteins regulate microtubule dynamics. MAPs can be categorized into three categories, microtubule-stabilizing proteins, microtubule-destabilizing proteins, and microtubule end-binding proteins. Microtubule-stabilizing proteins bind to microtubules and promote tubulin assembly, however they do not prevent disassembly. These proteins tend to be regulated by phosphorylation, and many have been shown to be substrates of mitotic protein kinases such as Aurora kinases and Plk. An example of a microtubule-stabilizing protein is CKAP5 or

TOGp, which increases the microtubule polymerization seven-fold during mitosis

(Charrrase et al, 1998; Spittle et al, 2000). Microtubule-destabilizing proteins, on the other hand, bind to microtubules and increase the frequency of tubulin catastrophe or depolymerization, alter the GTP conformation on the tips of

5 microtubules, or induce a conformational change in the microtubule lattice, ultimately inducing depolymerization. An example of microtubule-destabilizing proteins are members of the Kinesin-13 protein family, which induce depolymerization of kinetochore microtubules. Microtubule end-binding proteins associate with microtubules and enhance their dynamics, but they also serve as a scaffold for additional protein interactions. An example of this type of protein is

EB1, which influences microtubule dynamics at the microtubule tips but is also responsible for the recruitment of adenomatus polyposis coli (APC), an established tumor suppressor (Valiron et al, 2001).

Section III. MAPs Influence Spindle Pole Integrity and Chromosome Missegregation

In addition to plus ends, microtubule minus ends also play an important role in proper distribution of chromosomes during mitosis. As mentioned previously, as centrosome-nucleated microtubules are polymerized, the minus ends are polymerized at the spindle pole and grow outward to form spindle microtubules. A delicate balance of forces generated by microtubule dynamics directly influences chromosome congression and segregation. Thus, it is of the utmost importance that the integrity of the spindle pole is maintained. The spindle pole is defined as the site of microtubule minus end convergence that is governed by the centrosome (or spindle assembly factors in acentrosomal cells

(Compton, 1998). Centrosome nucleation generally results in a focused microtubule array (due to proteins in the PCM such as TOGp and CEP90) (Kim and Rhee, 2010; De Luca et al, 2008). However, studies have shown that

NuMA (nuclear mitotic apparatus), TPX2 (targeting protein for Xklp2), and Hice1

(Hec1-interacting and centrosome-associated 1) are largely responsible for the maintenance of spindle pole integrity. NuMA is a MAP that, in a complex with dynein and dynactin cross-links microtubule minus ends, focusing them and tethering them to the spindle pole (Silk et al., 2009; Haren et al., 2009;

Radulescu and Cleveland, 2010). TPX2 is a MAP that is responsible for proper targeting of Aurora A, the formation of kinetochore microtubules, and maintaining proper bipolar spindle assembly (Kufer et al, 2002; Gruss et al, 2002; Garret et al, 2002; Ma et al, 2011). Hice1 is a newly discovered MAP that possesses microtubule bundling and stabilizing capacities that are required for mitotic spindle integrity (Wu et al, 2008). Disruptions in any of these proteins, many of which were the result of protein depletion, resulted in spindle pole defects

(namely multipolar spindle formation and centrosome defects) and ultimately increased chromosome missegregation (Wittman et al, 1998; Wu et al, 2008; Silk et al., 2009; Haren et al., 2009; Radulescu and Cleveland, 2010). These studies suggest that decreases in spindle integrity disrupt the forces generated by spindle microtubules required for proper chromosome segregation.

Section IV. Microtubule-Associated Protein CKAP2 May Influence Microtubule Dynamics and Maintenance of the Genome

CKAP2 was first identified through a characterization of differentially expressed genes (based on histological subtypes) in diffuse large B cell lymphoma. Differential display-reverse transcription and cloning led to a 683

7 amino acid protein, which the authors called LB1, that shared no significant homology to known proteins, but was shown to localize to interphase microtubules. Tissue analysis demonstrated that it was highly expressed in proliferative tissues (e.g. testes, thymus, and tumor-derived cell lines), while it was largely absent in the liver, prostate and kidney (Maouche-Chretien et al.,

1998). CKAP2 was also found to be differentially expressed in a study of human gastric cancer. In the study by Bae et al., the authors showed that CKAP2 was upregulated in 23 out of 42 tumor samples and was associated with microtubule networks. Interestingly, they also found that CKAP2 had a significantly higher expression in adenocarcinomas compared to tubular adenomas, 48.8% to 9.1% respectively, and suggested that CKAP2 could be used as a marker to discriminate between the two cancer subtypes (Bae et al., 2003).

In a study by Jin et al., a mouse homolog of CKAP2 was identified, where its localization and differential expression was confirmed in mouse (colocalization with microtubules and centrosomes in interphase cells). Moreover, CKAP2 over- expression in NIH3T3 cells was shown to have microtubule stabilizing properties in interphase cells (Jin et al., 2004). Using mouse erythroleukemia cell lines (one normal and one with a temperature-sensitive p53) and HCT116 cells (WT and p53 -/- ), CKAP2 was found to be upregulated in a p53-dependent manner.

Expression of the mouse recombinant protein was shown to stabilize and bundle microtubules, induce abnormal nuclear morphology (including multiple nuclei formation), and induce centrosome amplification. The authors suggested that the centrosome amplification and multiple nuclei formation was largely due to

8 aborted cell division, which ultimately gave rise to tetraploidy. Tetraploidization has been shown to activate p53, and CKAP2 over-expression was shown to induce downstream p53 proteins, p21 and PUMA. However, majority of the study focused on the role CKAP2 during interphase, specifically the effect of

CKAP2 on proliferation and apoptosis (Tsuchihara et al., 2005), likely due to over-expression of the protein. Later CKAP2 studies have shown that expression of CKAP2 is confined to late G2 and mitosis.

However, endogenous CKAP2 in HeLa was not found to colocalize with interphase microtubules as was shown in cells that over-expressed CKAP2.

CKAP2 was only present (colocalizing with γ-tubulin and centrosome-proximal microtubules) after centrosome duplication and separation in late G2. CKAP2 colocalized with the mitotic spindle (both the spindle pole and spindle microtubules) from metaphase through anaphase B, however CKAP2 was not present during cytokinesis, which is consistent with degradation of CKAP2 following exit from mitosis. According to Seki and Fang, this finding suggests that CKAP2 may control spindle structure and function (Seki and Fang, 2007).

Endogenous CKAP2 in NIH3T3 cells confirmed that its expression is cell cycle dependent, and is associated with spindle microtubules and centrosomes in mitosis (Hong et al., 2007).

Reduction of CKAP2 by siRNA in HeLa cells showed no effect on the spindle structure (fixed cells) or mitotic chromosome congression and segregation (live-cell imaging). Moreover, there was no effect on microtubule stability or dynamics, as shown by microtubule depolymerization and regrowth

9 assay, which suggests that CKAP2 function is likely redundant to other spindle- associated proteins (Seki and Fang, 2007). Reduction of CKAP2 in C2C12

(mouse myoblast) has little effect on the formation and structure of the mitotic spindle, but leads to an increase in nuclear size and morphology (due to chromosome missegregation). CKAP2 depletion in C2C12 cells results in the formation of anaphase bridges and an increase in chromosomal instability. The authors show that lagging chromosomes resulted from a reduction of spindle checkpoint activity and ultimately resulted in significant retardation in cell growth

(Hong et al., 2008). Alternatively, ectopic over-expression of CKAP2 (~15-fold) induced a mitotic arrest, likely due to mitotic spindle defects (bundled microtubules and monopolar spindles). There was no obvious defect in cytokinesis, and there was no increase in bi-nucleated cells (Seki and Fang,

2007). In both HeLa and NIH3T3 cells, over-expression resulted in thick microtubule bundles and abnormal spindle morphology. Moreover, the cells are more resistant to microtubule depolymerization with nocodazole. The authors suggest that CKAP2 influences microtubule dynamics. Over-expression of

CKAP2 in HeLa and HEK293 also resulted in cell cycle delay or mitotic arrest, likely due to the 9.1-fold increase in spindle defects. The most prevalent spindle defect in HEK293 was the formation of monopolar spindles (single centrosomal

γ-tubulin mass). The increase in monopolar spindles delayed cells in prometaphase, resulting in a failure of cells to complete mitosis, e.g. 11% cell death, 24% cytokinesis failure, 37% asymmetrical division, 29% cytokinesis

10 failure. The authors suggest that CKAP2 inhibits centrosome separation or induces centrosome fusion at the onset of mitosis (Hong et al., 2007).

Phosphorylation of CKAP2 is specific during mitosis (Hong et al., 2009;

Kang et al., 2012). Recent efforts have concentrated on determining how phosphorylation affects association of CKAP2 with microtubules and, ultimately,

CKAP2 function. Previous studies showed that CKAP2 is hyperphosphorylated at the C-terminus during mitosis. It has been suggested that at least four different residues play a role in CKAP2 function in mitosis. Of these, Thr-622 has been shown to be phosphorylated by CDK1-cyclin B1 and to directly regulate spindle dynamics (Hong et al., 2009). Nevertheless, the exact cellular mechanism by which these observations occur and the role of CKAP2 in the maintenance of mitotic spindle poles and the stability of the genome remains elusive.

In the present study, using the human diploid, karyotypically stable colorectal cancer cell line DLD-1, I intend to elucidate the role of CKAP2 in the formation of the spindle pole, correlate its expression with partners that are known to play a role in the spindle formation and finally to approach the mechanism by which chromosomal instability arises in cells with altered CKAP2 expression.

Section V. Dissertation goals

1. Establish endogenous expression and localization of CKAP2 in wild-type

DLD1 cells

a. Hypothesis: CKAP2 localizes and binds to the mitotic spindle

2. Evaluate the function of CKAP2 using RNAi technology

a. Hypothesis: Depletion of CKAP2 causes mitotic spindle defects

3. Determine the mechanism of action of CKAP2

a. Hypothesis: CKAP2 interacts with other known spindle pole

proteins to ensure spindle pole integrity

b. Hypothesis: CKAP2 plays a role in maintaining genomic integrity

Chapter 2: Materials and Methods

Cell culture and synchronization

DLD1 cells obtained from ATCC (American Type Culture Collection,

Manassas, VA) were cultured in RPMI-1640 supplemented with antibiotics and

10% FBS at 37°C in 5% CO2. To enrich for mitotic cells, DLD1 cells were treated with 100 ng/mL of nocodazole for 16 hours and synchronized in prometaphase.

For all immunofluorescence experiments, cells were grown on sterilized 22 mm coverslips inside 6-well plates. Microtubule depolymerization in interphase cells was performed by treating the cells with 10 μg/mL nocodazole for 30 minutes.

Cells were released from the nocodazole block by washing with 1X PBS and incubating in fresh media for the desired time points at 37°C.

RNAi experiments

Two different target sequences were selected against CKAP2, siCKAP2_5 (5’-

GCA UUU GUU ACU AAC UGA ATT-3’) and siCKAP2_6 (5’-CAC GAU UGU

AGA UAU UCU ATT-3’) (Qiagen, Germantown, MD). Additionally, AllStars

Negative Control siRNA (scrambled sequence for control), and AllStars Hs Cell

Death siRNA (blend of highly potent siRNAs against several mitotic kinases used as a positive control) were used for RNAi experiments (Qiagen). The siRNAs were transfected into 2,500 plated DLD1 cells at a final concentration of 5nM using Lipofectamine™ RNAiMAX Reagent (Life Technologies). Target specific

13 transfection efficiency was corroborated at the mRNA level by QRT-PCR and at the protein level by immunoblot.

Figure 1. Diagram of the pGIPZ-shCKAP2 vector. Acquired from Open Biosystems, the shRNA (mir30) is regulated by a cytomegalovirus (CMV) promoter for enhanced expression. To improve stability of the transcript, there is a WRE sequence located 3’ to the shRNA. The region of interest is flanked by a 5’ long terminal repeat sequence and a 3’ self inactivating long terminal repeat sequence. The vector is also tagged with turboGFP, which allows for the positive identification of cells containing the plasmid.

The target sequence for stably silencing CKAP2 expression using an shRNA expression plasmid was 5’-AGG AAA CAT GTA TTC CTT TAA-3’

(OpenBiosystems, Lafayette, CO) (Figure 1). Plasmids were transfected into

DLD1 cells using Fugene HD (Promega, Madison, WI). Three days after transfection, positive cells were selected with 2 μg/mL of puromycin (Sigma-

Aldrich, St. Louis, MO) for two weeks. In order to enrich for transfected cells, positive cells were separated by fluorescence activated cell sorting (FACS)

(FACSCalibur, BD Biosciences, Franklin Lakes, NJ) based on green fluorescence protein (GFP) intensity, regrown in a 100 mm dish, and single-cell sorted into 96-well plates. Each well was individually monitored to ensure that only one cell was plated and single-cell clones were generated accordingly.

Quantification of gene expression

Gene expression levels were validated by quantitative reverse transcription-PCR

(RT-PCR) using Power SYBR Green technology (Applied Biosystems, Inc.,

Foster City, CA). Gene specific PCR primers were obtained from Operon

Technologies, Inc. (Huntsville, AL). The gene YWAHZ was used for normalization. Briefly, five µg of total RNA was reverse transcribed using

Superscript II (Invitrogen, Carlsbad, CA), the resulting cDNA was diluted 1:5 and

3 µl were used in each PCR reaction. PCR was performed with the default variables of the Applied Biosystems Prism 7000 sequence detector, except for a

15 total reaction volume of 25 µL. Each sample was analyzed in triplicate, and each data point was calculated as the median of the three measured CT values.

Cytotoxicity assays: CellTiter-Blue and Annexin V

For the CellTiter-Blue® Cell Viability Assay (Promega) (Figure 2), 2,500 cells were transfected with siRNA in 96-well plates and incubated at 37ºC for 96 hours. To measure cell viability, transfected cells were incubated with 20 μl

CellTiter-Blue reagent for 1 hour at 37ºC and the fluorescence measured by

SpectraMax M2 (Molecular Devices, Sunnyvale, CA) and analyzed using the software SoftMax Pro (Molecular Devices).

For the Annexin-V staining (Figure 2), 96 hours after siRNA transfection, DLD1 cells were harvested with CellStripper (Corning, CellGro), washed with binding buffer, and stained with Annexin-V and 7-AAD provided with the Annexin V-PE

Apoptosis Detection Kit (BD Biosciences). Transfection with siRNA against mitotic kinases (AllStar Death,Qiagen) was used as a positive control. Stained cells were loaded to FACSCalibur (BD Bioscience) and analyzed by flow cytometry. Apoptotic cells were determined using the software Cell Quest Pro

(BD Biosciences) and FlowJo (TreeStar, Inc., Ashland, OR).

A B

Figure 2. Schema of Apoptosis Assays. A) Cell Titer Blue assay determines cell viability by measuring the metabolic activity of cells in a particular well. The reagent consists of resazurin, which is metabolized by viable cells into resorufin, which emits fluorescence at 590nm and is readily detectable by a spectrofluorometer. B) Annexin V assay detects the number of apoptotic cells as indicated by the binding of Annexin V antibody to phosphatidyl serine that is present in the interior membrane of the cell. During early apoptosis the membrane flips, exposing phosphatidyl serine on the cell surface. 7-AAD intercalates into the DNA and only penetrates dying or dead cells. The combination of Annexin-V positive and 7-AAD negative or positive cells distinguishes early and late apoptotic cells.

Immunofluorescence, fluorescence microscopy and analysis

In order to preserve the cellular structure, DLD1 cells were cultured on a 22 mm coverslip and fixed with a 50:50 ratio of ice-cold acetone: methanol for 15 minutes at -20°C or 4% PFA with 0.1% Triton X-100 for 15 minutes at room temperature. For analysis of the mitotic spindle, cells were treated with 0.1%

Triton X-100 in PHEM for 5 minutes at room temperature to induce permeabilization. Fixed and permeabilized cells were blocked with 5% BSA with

1% normal goat serum in 0.1% PBST (PBS + 0.1% Triton-X) for 30 minutes.

Antibodies included: rabbit anti-CKAP2 (Sigma-Aldrich, 1:100), rat anti--tubulin

(YL1/2) (Accu-Specs, 1:400), mouse anti--tubulin, (DM1A) (Sigma-Aldrich,

1:400), mouse anti-NuMA (BD Biosciences, 1:100), mouse anti-g-tubulin (GTU-

88) (Sigma-Aldrich, 1:800), rabbit anti-pericentrin (Abcam, Cambridge, MA,

1:250), and mouse anti-Hec1 (Abcam, 1:1,000). Alexa Fluor 488 and Alexa Fluor

568 dyes (Molecular Probes, 1:500) were used as secondary antibodies for labeling. Antibodies were diluted in 0.1% PBST with 5% BSA and 1% normal goat serum. The incubation time was overnight at 4°C for primary antibodies and

1 hour at room temperature for fluorescence-conjugated secondary antibodies.

DAPI solution, ProLong Gold Antifade (Invitrogen) was used at the final step for

DNA staining. Cells were mounted onto a glass slide for subsequent microscopic observation.

Microtubule depolymerization/regrowth assay

To depolymerize microtubules, cells were incubated with the microtubule destabilizer nocodazole (10 μg/ml) for 30min. At the end of treatment cells were washed 4 times with 1X PBS and microtubule regrowth was triggered by transfer to drug-free medium at 37°C. Cells were released for 2, 30, and 60 minutes at

37°C, for a total of 1hr release. Slides were then rinsed once in 1X PBS, once with PHEM buffer and then fixed in -20ºC methanol. Tubulin structures were detected by incubating cells with a monoclonal -tubulin (Sigma-Aldrich, 1:1,000)

18 and rabbit polyclonal γ-tubulin (Sigma-Aldrich, 1:2,000) antibody at three different time points after drug release (Difilippantonio, et.al., 1999).

Image acquisition, processing, and analysis

Images were acquired with the Applied Precision Delta Vision Core System

(Applied Precision, Issaquah, WA). This system is based on an Olympus inverted

IX70 fluorescence microscope (Olympus America, Inc., Melville, NY) equipped with an automatically controlled stage that allows precise movement in XYZ directions. Data collection is controlled by SoftWoRx software installed on a

Linux-based workstation. Images were collected by a 12-bit camera (CoolSnap

HQ, Photometrics, Roper Scientific, Tucson, AZ). Cells were examined with a

86000 Sedat Quadruple Filter Set which included a FITC filter set (Ex 490/20;

Em 528/38; Polychroic mirror), a RD-TR-PE filter set (Ex 555/28; Em 617/73;

Polycroic mirror), and DAPI filter set (Ex 360/40; Em 457/50, Polychroic mirror)

(Chroma Technology Corp.). Images were analysed using Metamorph software, version 7.7.4 (Universal Imaging Corporation, Downington, PA). For the microtubule and spindle pole analysis, z-series stacks of images were analyzed and processed using a Delta Vision image processing workstation (Applied

Precision). For each slide 200 prometaphase and metaphase cells were analyzed based on microtubule morphology and recorded according to the number of spindle poles with centrosomes and characterized as normal or disorganized. A two-tailed t-test was performed to differentiate between the vector control and shCKAP2 cells. Mitotic spindle length was characterized by the distance between spindle poles, and analyzed by a two-tailed t-test. For the

19 total tubulin measurements, 100 cells were analyzed and tubulin measurements were determined by first calculating the total fluorescence in a cell and then subtracting the background intensity from the total intensity. The average intensity and standard deviations were calculated for both the vector control and shCKAP2. A two-tailed paired t-test was utilized to determine the statistical difference between the two groups. To analyze merotelic kinetochore orientation in metaphase and anaphase, each slide was analyzed and recorded as normal or as showing lagging chromosomes.

Immunoblotting

DLD1 cells were harvested by trypsinization and incubated in SDS lysis buffer

(1% SDS, 10 mM Tris-Hcl, pH 7.4, with protease inhibitors). The lysates were sonicated and boiled for 5 minutes with LDS Sample Buffer (Invitrogen). Protein samples were resolved by 4-12% SDS-PAGE and electroblotted onto a PVDF membrane. The membrane was blocked by soaking in TBS, 0.1% Tween 20, and

5% milk for 1 hour, incubated with primary antibody with blocking solution overnight at 4°C, washed three times with TBST (TBS, 0.1% Tween 20), incubated with HRP-linked secondary antibodies for 1 hour at room temperature, and washed three times with TBST. The antibodies used were mouse anti-

CKAP2 (Abcam, 1:1,000), rabbit anti-CKAP2 (Sigma-Aldrich, 1:1,000), rat anti-α- tubulin (YL1/2) (Accu-Specs, 1:1,000), mouse anti-phospho-histone H3 (Ser10,

6G3) (Cell Signaling, Danvers, MA, 1:1,000), mouse anti-cyclin B1 (V152) (Cell

Signaling, 1:2,000), α/β-tubulin (Cell Signaling, 1:2,000), GAPDH (Sigma-Aldrich,

1:40,000), and NuMA (BD Biosciences, 1:2,000). Anti-mouse IgG, HRP linked, anti-rabbit IgG, HRP linked, and anti-rat IgG, HRP linked antibodies (Cell

Signaling) were used for secondary labeling. For the detection of signals, Super

Signal West Pico (ThermoScientific, Rockford, IL) was used according to the manufacturer’s recommendations. Films were developed on a Kodak X-OMAT

2000A device.

Spindle binding assay

Cells were arrested in mitosis as previously described (see section cell culture and synchronization). Mitotic cells were harvested by mitotic shake-off and re- suspended in a hypotonic lysis buffer (1 mM MgCl2, 2 mM EGTA, 20 mM Tris-

HCl, pH 6.8, 0.5% NP-40, 3 μM taxol, 10 μM trichostatin A, and protease inhibitors). The lysate was then centrifuged at 15,000X g for 10 minutes to separate the mitotic spindle and DNA from the remainder of the lysate. The supernatant and pellet (10 μL per each fraction) were analyzed by Western blot.

Metaphase harvesting and SKY

Metaphase chromosomes for SKY were prepared after exposure of the wild-type and CKAP2 depleted DLD1 cells to colcemid (Roche, Indianapolis, IN) for 1–1.5 hours at a final concentration of 0.1 µg/ml. The cells were lysed in hypotonic

21 solution (0.075 M KCl), and the nuclei were fixed in methanol and acetic acid

(3:1). SKY was performed as described by Schrock et al. (Schrock E et al 1996).

For protocol details, please refer to Padilla-Nash et al. (Padilla-Nash et al 2007).

Differentially labeled chromosome-specific painting probes were hybridized simultaneously onto metaphase chromosomes. Images were acquired with a custom-designed triple-pass filter using the SpectraCube SD200 (Applied

Spectral Imaging, Vista, CA) connected to an epifluorescence microscope

(DMRXA, Leica Microsystems, Wetzlar, Germany). At least 20 metaphases and corresponding inverted 4,6-diamidino 2-phenyl-indole (DAPI) images were analyzed for each clone, and karyotypes were defined using standard nomenclature rules.

At least 100 metaphases previously stained with DAPI were assessed to investigate the variability in the number of chromosomes.

Analysis of Chromosome Segregation by Live-cell imaging

DLD1 cells transfected with empty vector control and CKAP2-depleted cells, were then co-transfected with H2B-Cherry (gifted from Dr. J. Silvio Gutkind,

National Institute of Dental and Craniofacial Research, National Institutes of

Health) and selected with neomycin. Positively selected cells were grown on glass chamber slides (2-chamber) for 72 hours and then analyzed on the Zeiss

LSM 5 Live confocal microscope (Carl Zeiss, Inc., Oberkochen, Germany) within an incubation chamber XL LSM710 S1 (PeCon GmbH, Germany) with a heating

22 insert P-LabTek S1. Lasers used: Diode 25 mW 405 nm, DPSS 40 mW 561 nm,

Diode 100 mW 488 nm. Time-lapse images were taken from three regions for each sample every 3 minutes for 72 hours. Maximum intensity projections were taken for each sample and analyzed by ImageJ 1.46.

Chapter 3: Results

Section I. CKAP2 Expression and Localization in Wild-Type DLD1

The goal of this section is to establish endogenous expression and localization in the diploid, karyotypically stable colorectal cancer cell line, DLD1.

CKAP2 expression is four-fold higher in DLD1 cells, compared to the normal mucosa based on qRT-PCR analysis. Studies have indicated that ectopic over- expression of this protein resulted in spindle defects, namely the formation of monopolar spindles, and increased chromosome missegregation (Hong et al.,

2007; Seki and Fang, 2007). However, DLD1 is a stable diploid cell line, with no significant spindle defects. As a result, this cell line was used as a model to establish the function of CKAP2 in the context of colorectal cancer.

To investigate the endogenous expression of CKAP2, wild-type DLD1 cells were first synchronized with nocodazole, a microtubule-destabilizing agent, to accumulate cells in mitosis. Presence of phospho-histone H3 was assessed as a control for proper mitotic arrest. Next, the mitotic block was released, and protein was harvested one, three, and five hours post-nocodazole release.

Immunoblot analysis showed that CKAP2 accumulated in mitosis and is degraded within an hour after release (Figure 3A). In addition, the data also suggest that CKAP2 is expressed simultaneously with cyclin B1 and, in agreement with previous reports in HeLa cells (Seki and Fang, 2007), both proteins have similar degradation kinetics. Propidium iodide staining and

24 subsequent fluorescence activated cell sorting (FACS) analysis was used to confirm that cells were synchronized and released from the mitotic block (Figure

3B).

The subcellular localization of CKAP2 was approached by immunofluorescence in mitotic cells in an asynchronous population. Cells were co-stained with α-tubulin as a marker for microtubules and γ-tubulin as a marker to identify centrosomes. Our observation led us to conclude that CKAP2 localizes to the mitotic spindle, in particular the spindle pole, but not to the centrosome (Figure 4A,B). Previous studies have shown that CKAP2 associates with the microtubule network, as nocodazole treatment disrupts CKAP2 localization, however a direct link has yet to be established. To determine whether CKAP2 is directly associated with spindle microtubules, a mitotic spindle-binding assay was performed. Mitotic cells were harvested by mitotic shake-off and the mitotic spindle was stabilized with the addition taxol. The mitotic spindle, DNA, and associated proteins were then separated from the remaining proteins by centrifugation. By Western blot analysis, we showed that

CKAP2 is present in the pellet fraction of the lysate, in both untreated and nocodazole treated cells. These data suggest that CKAP2 associates with the mitotic spindle but does not directly bind to microtubules (Figure 4C).

Figure 3. CKAP2 Expression is Restricted to Mitosis. (A) Wild-type DLD1 cells were synchronized in mitosis with 100 ng/mL nocodazole for 16 hours and released for the indicated time points (1, 3, and 5 hours). The cells were harvested and analyzed by immunoblot with antibodies specific for CKAP2, cyclin B1, phospho-Histone H3, and GAPDH. (B) Progression from mitotic release through the cell cycle was verified by synchronizing wild-type cells with nocodazole as previously noted and released for the indicated time points. Cells were harvested, stained with propidium iodide and analyzed by FACS.

Figure 4. CKAP2 Localizes to and is Associated with the Mitotic Spindle. (A) An asynchronous population of DLD1 cells was co-stained with CKAP2 (green), α-tubulin (red), and merged with DAPI (blue). Mitotic cells were analyzed for colocalization of CKAP2 and the mitotic spindle. (B) An asynchronous population of DLD1 cells were co-stained with γ-tubulin (red), CKAP2 (green), and merged with DAPI (blue). Mitotic cells were analyzed for colocalization of CKAP2 and centrosomes. (C) For untreated panel, mitotic cells were enriched in wild-type DLD1 cells by mitotic shake-off. For nocodazole treated panel (control), cells were treated with 100ng/mL nocodazole for 16hrs. Cells were harvested and the mitotic spindle and DNA separated from the remainder of the proteins by centrifugation. The fractions generated from spindle-binding assay were analyzed by immunoblot with antibodies specific for CKAP2, α/β-tubulin and γ-tubulin. CKAP2 was found primarily in the pellet fraction, which is composed of the heaviest components, namely DNA and the mitotic spindle, indicating that CKAP2 binds either directly or indirectly to the mitotic spindle. α/β tubulin was used a loading control.

Section II. Establishment of shCKAP2 Model

The goal of the experiments in this section is to establish a model system in DLD1 cells in which CKAP2 expression is perpetually reduced. RNAi has been utilized in the study of microtubule-associated proteins (MAPs) to determine the function and role these proteins play in bipolar spindle assembly and chromosome segregation. Studies have shown that inhibition of mitosis-related proteins results in defects in cell proliferation, viability, and cell cycle distribution.

CKAP2 is not Essential for Cell Viability

RNAi technology was utilized to assess the cellular function of CKAP2.

Cells were treated with two different siRNA sequences and harvested after incubating for 96 hours. Decrease of CKAP2 expression at the mRNA level was confirmed by QRT-PCR, showing a reduction of 75-90% (Figure 5A). This was confirmed at the protein level by Western blot analysis (Figure 5B). A metabolic assay, CellTiter-Blue, was used to measure cell viability. The CellTiter Blue reagent contains a compound, resazurin, that is metabolized by viable cells to resorufin, which emits fluorescence that can be detected by a spectrofluorometer. As indicated by the CellTiter-Blue assay, the reduction of

CKAP2 expression had only a very limited effect on cellular viability (5% decrease) (Figure 5C), demonstrating that CKAP2 is not essential for cellular survival. Moreover, analysis of Annexin-V and 7-AAD, a combination of markers that detects apoptotic cells, by FACS did not show any significant increase in

28 apoptosis, 11% in CKAP2 siRNA treated cells compared to 74% in the positive control (siRNA against PLK1) (Figure 5D).

To further assess the long-term effects of CKAP2 loss-of-function, cells were transfected with shRNA, and single cell clones were obtained by selection with puromycin and single-cell sorting. Depletion of CKAP2 protein was verified by synchronizing the cells in mitosis with nocodazole and protein harvested for immunoblot analysis. Quantification of the intensity of the CKAP2 signal showed a 98% decrease in protein expression (Figure 6A,B). Immunofluorescence analysis of CKAP2 confirmed the reduction in protein expression, with a 78% and

88% decrease in signal intensity at the mitotic spindle in clone CKAP2_8 and

CKAP2_12, respectively (Figure 7). Studies in mouse C2C12 cells have showed that knock-down of CKAP2 resulted in a retardation in cell growth (Hong, et al.

2009). Control and CKAP2-depleted cells were counted every day for six days, but no significant difference in cell number was observed (Figure 8A). Similarly to the siRNA analysis of cell viability, in DLD1 cells there is limited effect on cell viability in cells transfected with CKAP2 shRNA as measured by CellTiter-Blue

(Figure 8B). For the purposes of experimentation, two clones were primarily used, CKAP2_8 and CKAP2_12.

A B

C D

PLK1

Figure 5. CKAP2 is not Essential for Cell Viability. (A) DLD1 cells were transfected with control (siCTL) or CKAP2 (siCKAP2) siRNA. Seventy-two hours later, RNA was extracted for qRT-PCR analysis. The histogram represents the percentage of remaining RNA for the control and siRNA depleted cells. (B) Ninety-six hours post siRNA transfection, cells were harvested for immunblot analysis with antibodies specific to CKAP2 and GAPDH. (C) Cell viability was analyzed by measuring the metabolic activity of siCTL and siCKAP2 cells 96 hours post siRNA transfection. The histogram represents the remaining viable cells for each experimental group for six biological replicates. (D) Apoptosis was measured by costaining cells with Annexin-V (x-axis) and 7-AAD (y-axis) and analyzed by FACS [negative control (untreated; top left), positive control (PLK1; top right), siCKAP2 (bottom left and right]. The upper left quadrant (7-AAD positive) represents dead cells, the bottom left represents unstained cells, and the right quadrants represent early apoptotic (Annexin-V positive) and late apoptotic cells (Annexin V positive and 7-AAD positive).

Figure 6. Generation of Stable shCKAP2 Cell Lines. (A) DLD1 cells were transfected with shRNA, selected with puromycin, and separated by FACS sorting based on GFP-positivity. Separated cells were synchronized overnight with 10 μg/mL nocodazole and harvested for immunoblot analysis with antibodies specific for CKAP2 and GAPDH. (B) The histogram represents the quantification of the signal intensity for both the control and CKAP2-depleted cells.

Figure 7. Validation of CKAP2 Knock-Down. Representative images of control (top panel) and shCKAP2 (lower panels) coimmunstained with CKAP2 (green), α- tubulin (red), and merged with DAPI (merge). The data is presented as the mean CKAP2 intensity per experiment group, as measured in Image J software. One hundred images were analyzed per experimental group.

1.5×1007 Untreated

1.2×1007 shCTL

s l

l shCKAP2_8

e 1.0×1007 C

shCKAP2_12

f o 06

r 7.5×10

e b

m 5.0×1006

u N 2.5×1006

0.0 1 2 3 4 5 6 Time (days)

Figure 8. Stable CKAP2 Depletion does not Influence Cell Proliferation or Viability. (A) To measure the affect of CKAP2 reduction on cell proliferation, populations of each control (shCTL) and CKAP2-depleted cells (shCKAP2) were counted every for six days and plotted in the histogram. (B) Cell viability was analyzed by measuring the metabolic activity of shCTL and shCKAP2 cells for 96 hours after plating. The histogram represents the remaining viable cells for each experimental group for six technical replicates.

Cell Cycle is not Affected by CKAP2 Depletion

Several reports indicate that CKAP2 is regulated by APC/Cdh1 and its expression rises in late G2, peaks at M phase, and is degraded post-anaphase

(Hong et al., 2007; Seki and Fang, 2007). To validate the distribution of cell cycle, DNA content was measured in an asynchronous population of cells with propidium iodide for analysis of the number of cells in the various stages of cell cycle. Propidium iodide staining of the clones showed no discernable differences in progression through cell cycle, 50% G1, 30% S, and 20% G2/M in both the control and shCKAP2 (Figure 9A). For a closer analysis of cell cycle, cells were transfected with H2B:Cherry and analyzed by live-cell imaging. From these images, mitosis was measured by the level of DNA condensation, while the entire length of the cell cycle was determined by nuclear envelope breakdown

(NEB). A 3-hour difference was observed in the length of the cell cycle in shCKAP2 cells compared to the controls (P = 0.048) (Figure 9B). A decrease in the in length of mitosis was also observed in shCKAP2 cells, 41 min versus 56 min in controls (P = 0.002) (Figure 10A).

To further distinguish between G2 and M stages, cells were co-stained with cyclin B1 and 7-AAD. Cyclin B1 is a mitotic regulatory protein that complexes with mitotic kinase Cdk1 to regulate the early events of mitosis, such as chromosome condensation, nuclear envelope breakdown, spindle pole assembly. Cyclin B1 is expressed in late G2 through anaphase, and is rapidly

34 degraded after anaphase (Ciciarello et al., 2007). FACS analysis showed that for shCKAP2 clones, there was a dramatic decrease in cyclin B1 expression compared to controls (Figure 10B), despite the fact that in prior experiments more than 20% of cells in asynchronous populations were in G2 and/or mitosis

(i.e. actively dividing). This could suggest that CKAP2-depleted cells exit mitosis at a faster rate than control cells.

Figure 9. Depletion of CKAP2 does not Affect Cell Cycle Distribution. (A) Asynchronous control and shCKAP2-depleted cells were stained with propidium iodide and the DNA content analyzed by FACS. The phases of cell cycle, G1, S, G2/M, were determined based on 2N and 4N DNA content. (B) Control (shCTL) and CKAP2-depleted (shCKAP2) cells were transfected with histone H2B-Cherry constructs, selected with gentamicin, and analyzed with live-cell imaging. The length of cell cylce was determined by measuring the length of time from mitosis to mitosis, indicated by nuclear envelope breakdown.

P < 0.02

Figure 10. CKAP2 Depletion Results in a Decrease in the Length of Mitosis. (A) Control (shCTL) and CKAP2-depleted (shCKAP2) cells were transfected with histone H2B-Cherry constructs, selected with gentamicin, and analyzed with live- cell imaging. The length of mitosis was measured by nuclear envelope breakdown (NEB) to chromosome decondensation in cytokinesis. (B) Cells were stained with cyclin B1 (marker for mitosis), 7-AAD (marker for cell viability), and DAPI and analyzed by FACS. The cyclin B1 positivity discriminates cells with 4N as either in G2 or mitosis. For each sample, the percentage of cells in either G2 (small checkered) or mitosis (large checkered) is given.

Section III. Evaluation of CKAP2 Function

The experiments in this section are designed to evaluate the function of

CKAP2. Alterations in the expression of CKAP2 have been shown to result in defects in spindle morphology and alterations in the cell cycle profile (Seki and

Fang, 2007; Hong et al., 2007). These defects ultimately resulted in an increase in chromosome missegregation, however the mechanism is poorly understood.

Moreover, depending on the cellular context (e.g. p53 inhibition), CKAP2 led to an increase abortive cell divisions, ultimately resulting in tetraploidy. However, there is a discrepancy between the results of many of the papers, namely the precise effect of altering CKAP2 in various cell lines by RNAi or over-expression.

According to a study by Seki and Fang, the discrepancy between these studies is likely due to cell line- or tissue-specific differences among each of the cell lines used, suggesting that CKAP2 may have cell line- or tissue-specific functions

(Seki and Fang, 2007). RNAi technology was utilized to assess the cellular function of CKAP2 in a stable diploid cell line, DLD1.

Suppression of CKAP2 results in an increase of spindle pole defects

Data obtained from CKAP2 expression and localization analyses suggested that the effects of silencing this gene would be most obvious in mitotic cells, particularly affecting the mitotic spindle. Thus, the integrity of the mitotic spindle was assessed by immunofluorescence with α-tubulin as a marker for the mitotic spindle and γ-tubulin as a marker for centrosomes. Reduction of CKAP2 expression resulted in a significant increase in cells with multipolar spindles (from

3% to 13% of cells, P<0.05), yet with little effect on bipolar spindle formation or other defects in spindle organization (Figure 11A,B). Consistent with this finding, an increase in supernumerary centrosomes was observed however, but this increase was not statistically significant (P<0.05) (Figure 11C). An interesting observation was a lack of centrosome clustering in CKAP2-depleted cells that was evident in control cells (data not shown).

Using γ-tubulin as a marker for centrosomes, the integrity of spindle poles was assessed in the absence of CKAP2. A significant increase of cells where the γ-tubulin signal was dispersed along the mitotic spindle (from 5% to 40%,

P<0.01) was observed in CKAP2-depleted cells (Figure 12A,B). A small percentage of cells where one centrosome was dislocated from the spindle pole

(from 3% to 8%, P>0.05) was also detected (Figure 13A,B). In order to examine spindle pole function, spindle tension was assessed by measuring the distance between the two spindle poles and found that in shCKAP2_4 and shCKAP2_8 cells there was an increase of 5 μm in the distance between spindle poles when comparing with vector control-transfected cells (P<0.05) (Figure 14A).

Alterations in spindle tension have been linked to errors in chromosome segregation, largely due to inefficiencies in properly aligning chromosomes at the metaphase plate and subsequent movement to either pole. To determine levels of chromosome misalignment, metaphase cells were analyzed for complete chromosome congression or the presence of chromosomes not at the metaphase plate. The data show that the decrease in the spindle tension is the result of an increase in misalignment of metaphase chromosomes (from 6% to

18%, P< 0.02) (Figure 14B). Overall, these data suggest that despite spindle abnormalities, cells remained able to form functional mitotic spindles.

B C

Figure 11. Reduction of CKAP2 Expression Results in an Increase in Multipolar Spindles. (A) Mitotic cells in asysnchronous populations of control (shCTL) and shCKAP2-depleted (shCKAP2) cells were analyzed for mitotic defects by coimmunostaining with γ-tubulin (green) and α-tubulin (red). Representative images of multipolar spindles (top panel) were observed CKAP2- depleted cells, Over 100 spindles per experimental group were analyzed in two independent experiments. (B) The results are presented as mean ± s.d. P- values were determined using the Student’s t-test. (C) The number of cells with supernumerary centrosomes, indicated by γ-tubulin staining, were counted and grouped as 3 or 4+ γ-tubulin foci. Over 200 cells per experimental group were analyzed. The results are represented as mean ± s.d.

Figure 12. CKAP2 Depletion Results in Dispersal of γ-tubulin Away from the Centrosome. (A) Mitotic cells in an asynchronous population of control (shCTL) and CKAP2-depleted (shCKAP2) were coimmunostained with γ-tubulin (green), α- tubulin (red) and merged with DAPI (merge). The dispersal of γ-tubulin away from the centrosome were analyzed in 200 cells per experimental group in two independent experiments. Representative images for each experimental group and the mitotic defect are shown. (B) The results are presented as mean ± s.d. P-values were determined using the Student’s t-test.

Figure 13. CKAP2 Depletion Results in an Increase in Centrosome Dislocation. (A) Mitotic cells in an asynchronous population of control (shCTL) and CKAP2-depleted (shCKAP2) were coimmunostained with γ-tubulin (green), α- tubulin (red) and merged with DAPI (merge). The dislocation of the centrosome were analyzed in 200 cells per experimental group in two independent experiments. Representative images for each experimental group and the mitotic defect are shown. (B) The results are presented below as mean ± s.d. P-values were determined using the Student’s t-test.

Figure 14. Spindle Pole Defects Result in an Increase in Spindle Tension and Chromosome Misalignment in Metaphase. (A) Spindle length measured in mitotic cells with bipolar spindles for 50 spindles in both controls and CKAP2- depleted cells. P-value was determined by Students T-test. (B) The number of misaligned chromosomes in the bipolar metaphase cells was counted for controls (shCTL) and CKAP2-depleted cells (shCKAP2). More than 200 cells were counted per condition.

CKAP2 is crucial for tethering the spindle pole in early mitosis

Because of the observed defects in γ-tubulin, in particular those resulting in signal dispersal along the mitotic spindle, the extent that CKAP2 was required for centrosome-nucleated microtubule formation and microtubule stability was examined. Asynchronous cells were treated with nocodazole for 30 minutes to depolymerize microtubules. Cells were released at 2, 30, and 60 minutes for subsequent analysis.

Strikingly, as early as two minutes after release a very unusual phenotype consisting of dispersal of newly polymerized microtubule filaments was observed that, in normal conditions, should be properly bound to the spindle pole. In

CKAP2-depleted cells, around 60% of mitotic cells showed this cage-like pattern of dispersed nascent microtubules while we only identified this phenomena in less than 20% of cells when transfected with an empty vector control (P<0.01)

(Figure 15). Immunostaining of pericentrin, an integral protein component of the centrosome, showed two prominent foci in the majority of CKAP2-depleted cells

(Figure 16), demonstrating that, despite the spindle pole dispersal, the centrosome remained intact. Nevertheless, 30 minutes post-nocodazole release, the percentage of cage-like spindles decreased to 40% and microtubule filaments began forming distinct poles (Figure 17), and overall, successful bipolar spindles were formed at 60 minutes post-nocodazole release (Figure 18).

Furthermore, total tubulin measurements showed that there was a slight increase in the amount of total polymerized tubulin at 2min and 30min in the shCKAP2 compared to the control, which indicates that the total nucleation capacity of the

45 centrosome is not reduced, but disorganized at early time points (Figure 19).

The mechanism by which this increase in total polymerized tubulin might occur has yet to be fully elucidated.

Figure 15. CKAP2 is Required for Anchoring of Centrosome-Nucleated Microtubules to the Spindle Pole. (A,B) Nucleation was induced by treating the cells with 10 ug/mL nocodazole for 30 minutes and released for two minutes into fresh media. shCTL and shCKAP2-depleted cells were coimmunostained with γ- tubulin (green), α-tubulin (red) and merged with DAPI (blue). A cage-like structure was often observed two minutes post-nocodazole release. Representative images for each experimental group are shown. (C) The number of cells with non- centrosomal tubulin staining was measured 100 cells per experimental group and plotted.

Figure 16. CKAP2 is Required for the Anchoring of Centrosome-Nucleated Microtubules to the Spindle Pole. (A,B) Nucleation was induced by treating the cells with 10 ug/mL nocodazole for 30 minutes and released for two minutes into fresh media. shCTL and shCKAP2-depleted cells were coimmunostained with DM1A (green), pericentrin (red) and merged with DAPI (blue). Similarly, a cage- like structure was often observed two minutes post-nocodazole release. Representative images for each experimental group are shown.

Figure 17. Depletion of CKAP2 Causes a Delay in Spindle Formation (A) Nucleation was induced by treating the cells with 10 ug/mL nocodazole for 30 minutes and released for 30 minutes into fresh media. shCTL and shCKAP2- depleted cells were coimmunostained with γ-tubulin (green), α-tubulin (red) and merged with DAPI (blue). Representative images for each experimental group are shown. (B) The number of cells with non-centrosomal tubulin staining was measured 100 cells per experimental group.

Figure 18. CKAP2-depleted Cells are Capable of Bipolar Spindle Formation (A) Nucleation was induced by treating the cells with 10 ug/mL nocodazole for 30 minutes and released for 30 minutes into fresh media. shCTL and shCKAP2- depleted cells were coimmunostained with γ-tubulin (green), α-tubulin (red) and merged with DAPI (blue). Representative images for each experimental group are shown. (B) The number of cells with non-centrosomal tubulin staining was measured 100 cells per experimental group.

Figure 19. Total nucleation capacity is not reduced in CKAP2 depleted cells. Asynchronous populations were stained with DM1A and pericentrin. The signal intensity of DM1A was measured in mitotic cells. Total tubulin was analyzed for 35 cells 2, 30, and 60 minutes post-nocodazole release by measuring the mean fluorescence intensity of DM1A staining.

CKAP2 Depletion does not Increase Chromosome-Directed Nucleation

Multiple mechanisms for microtubule nucleation exist, of which centrosomes are the most potent. However, acentrosomal cells are capable of nucleating and assembling microtubules using alternative mechanisms, such as kinetochore or chromatin-driven microtubule synthesis. I hypothesize that the disorganization of early nucleated microtubules is due to a delay in centrosome nucleation, and subsequent increase in alternative nucleation pathways. In order to investigate the possibility that CKAP2-depleted cells were utilizing a kinetochore and/or chromatin-driven microtubule nucleation mechanism to overcome centrosomal nucleation difficulties, the localization of the dispersed microtubule filaments in relation to kinetochores was assessed. For this, Hec1, a kinetochore component, and α-tubulin were visualized using immunofluorescence. The results indicate that nascent microtubules were not associated with kinetochores, as there was no colocalization of -tubulin signal to Hec1 (Figure 20). Although some α- tubulin overlapped with chromatin staining with DAPI, the presence of microtubule filaments that positioned outside chromatin boundaries disproved the hypothesis of a chromatin-directed microtubule generation. Therefore, CKAP2- depleted cells did not utilize this alternative mechanism to promote microtubule polymerization. The data suggests that the cage phenotype is the result of a dispersal of newly polymerized tubulin.

Expression and Localization of Spindle Pole Protein, NuMA, is Unaffected in CKAP2-depleted Cells The disorganization of newly nucleated microtubules suggests that other spindle pole proteins responsible for anchoring tubulin to the centrosome and spindle pole may also be mislocalized or absent from the spindle pole. One of the major proteins involved in cross-linking and positioning of minus-end microtubules to the spindle poles is NuMA (Nuclear Mitotic Apparatus)

(Radulescu and Cleveland, 2010). To assess whether depletion of CKAP2 affected the positioning of NuMA, mitotic cells in asynchronous cell populations were analyzed for NuMA localization and expression, using α-tubulin as a control. NuMA localized to the centrosomes and spindle poles in both control and depleted cells (Figure 21A), which was corroborated by Western blot analysis that showed that the abundance of NuMA protein is maintained in synchronized CKAP2-depleted cells (Figure 21B). Considering the function of

NuMA in mitotic cells, colocalization of CKAP2 and NuMA was analyzed by immunofluoresence. Although NuMA and CKAP2 both localize to the spindle pole, there was no apparent colocalization between the two proteins (Figure 22).

To further validate proper NuMA localization and function, NuMA expression was determined 2 minutes after nocodazole block and release. As expected, both control and shRNA transfected cells showed a scattered distribution of NuMA, indicating movement of NuMA from the nucleus to the spindle pole. In CKAP2- depleted cells where spindle pole dispersal was evident, we observed colocalization of nascent microtubules and NuMA two minutes post-nocodazole

53 release (Figure 23), confirming the finding that the centrosome nucleated microtubules are dispersed throughout the entire chromatin region, and supporting the interpretation that CKAP2 might be involved in the maintenance and recruitment of microtubules at the spindle poles.

Figure 20. Reduction of CKAP2 Expression does not Enhance Chromosome-Directed Nucleation. Nucleation was induced by treating the cells with 10 ug/mL nocodazole for 30 minutes and released for two minutes into fresh media. Two minutes post-nocodazole release, shCTL and shCKAP2- depleted cells were coimmunostained with kinetochore protein Hec1 (green), α- tubulin (red) and merged with DAPI (merge) to determine the presence of chromosome-directed nucleation. Colocalization of Hec1 and α-tubulin signals was analyzed in control and CKAP2-depleted cells. Representative images for each experimental group are shown.

A ! A

B ! A

Figure 21. Depletion of CKAP2 does not Affect NuMA Expression and Localization. (A) shCTL and shCKAP2-depleted cells were coimmunostained with NuMA (green) and α-tubulin (red), and the images are merged with DAPI (merge). The expression of NuMA was analyzed at the spindle pole. Representative metaphase images are shown. (B) Transfected cells were synchronized overnight with 100 ng/mL nocodazole. Cells were harvested for immunoblot analysis with antibodies specific for NuMA and GAPDH.

Figure 22. CKAP2 and NuMA Both Localize to the Mitotic Spindle Pole. (A) Mitotic cells in an asynchronous population of shCTL and shCKAP2-depleted cells were coimmunostained with NuMA (green), CKAP2 (red) and merged with DAPI (merge). Representative images from each experimental group are shown.

Figure 23. Localization of NuMA is not Affected by CKAP2 Depletion. (A) shCTL and shCKAP2-depleted cells were treated with 10 μg/mL nocodazole for 30 minutes and released for two minutes into fresh media. Cells were comimmunostained NuMA (green), α-tubulin (red), and merged with DAPI (merge). Representative images for each experimental group are shown.

Knockdown of CKAP2 results in a higher incidence of merotelic attachments, anaphase lagging, and chromosomal instability Consequences of the mitotic defects generated by CKAP2 depletion were examined carefully. Using an asynchronous population of cells, we sought evidence of chromosome missegregation, including anaphase lagging, micronuclei, and nuclear blebs. By performing live-cell imaging using a histone H2B-Cherry construct, we first assessed the anaphase lagging events. Analysis of time-lapse sequences of cells undergoing mitosis demonstrated a higher incidence of lagging chromosomes in CKAP2-depleted cells and showed how anaphase bridges resolved in the formation of micronuclei in the daughter cells (Figure 24A). As a consequence of the elevated rate of missegregation errors, a significant increase of micronuclei, nuclear blebs and abnormal nuclei shape was identified in a drug-free asynchronous cell population analysis when comparing CKAP2-depleted cells with empty vector control (P<0.05) (Figure 24B). It is important to note that there was no evidence of cytokinesis failure or mitotic slippage in CKAP2 depleted cells by live-cell imaging.

In order to further understand the cause of these segregation errors, immunostaining with Hec1 and -tubulin demonstrated a higher incidence of lagging chromosomes in CKAP2-depleted cells compared to control cells. In fact, cells with suppressed CKAP2 had an average of 12% of lagging chromosomes, which represented up to three times more than the control cells. Interestingly, merotelic kinetochore-microtubule attachments occurred at a higher frequency in lagging chromosomes when expression of CKAP2 was suppressed (Figure 25). As a consequence of the elevated rate of missegregation, a significant increase of abnormal nuclear shapes, including micronuclei and nuclear blebs, was identified in

59 a drug-free asynchronous cell population analysis when comparing CKAP2-depleted cells with empty vector control (P<0.05) (Figure 26). Therefore, I concluded that delays in tethering the microtubule spindles at the pole promote geometric defects that increase merotelic kinetochore-microtubule attachments that result in chromosome lagging and increased chromosomal instability or CIN.

To further analyze chromosomal content, spectral karyotype (SKY) analysis was performed. SKY is a cytogenetic tool in which mitotic chromosomes are hybridized with “chromosome paints” specific for each chromosome, numerical aberrations and structural rearrangements can be observed in the genome.

Analysis of the CKAP2-depleted cells revealed that the chromosomal heterogeneity was substantially higher in the clones than in the control cells. As shown in the radial plots in Figure 27, karyotypes of control cells showed a near-perfect distribution of the expected modal number around diploid (2N) circle, however suppression of

CKAP2 in the clones resulted in an increase in the number of cells with abnormal chromosomal content. Interestingly, an unusual increase of aneuploidy (22.6% of polyploid cells for clone CKAP2_8 and 36.2% for clone CKAP2_12 compared to

2.2% for control cells) was also observed (Figure 27). Spectral karyotype analysis for a representative subset of cells in each clone showed ongoing patterns of aneuploidy as well as de novo clonal structural chromosome aberrations, confirming the elevated incidence of CIN in these cells (Figure 28).

Figure 24. Reduction of CKAP2 Results in an Increase in Chromosome Missegregation. (A) Control (shCTL) and CKAP2-depleted (shCKAP2) cells were transfected with histone H2B-Cherry constructs, selected with gentamicin, and analyzed with live-cell imaging (image taken every three minutes). The movie shows shCKAP2-depleted histone H2B:Cherry positive cells undergoing aberrant mitosis with chromosome missegregation resulting into two daughter nuclei with micronuclei. Arrowheads indicate lagging chromosomes and resultant micronuclei. (B) The histogram represents the number of chromosome missegregation events for each H2B-Cherry positive experimental group.

Figure 25. Reduction of CKAP2 Results in an Increase in Merotelic Attachments. Mitotic cells shCTL and shCKAP2-depleted populations were enriched by treating the cells with 100 μM monastrol for four hours. Cells were immunostained for Hec1 (green), α-tubulin (red), and DAPI (blue). Cells with lagging chromosomes in anaphase and telophase were analyzed for merotelic attachments. Each image represents an individual mitotic event. The magnified views in the inset show individual merotelic attachments, where each chromosome is attached to both poles.

Figure 26. Increased Chromosome Lagging in CKAP2-depleted Cells Results in Abnormal Nuclei Morphology. (A) Asynchronous shCTL and shCKAP2-depleted cells were co-stained with α-tubulin and γ-tubulin and analyzed for evidence of chromosome missegregation, including micronuclei, nuclear blebs, and anaphase bridges. The image on the right refects a nuclei with abnormal morphology, whereas the image on the left refect a cell with multiple micronuclei (indicated by the arrow). (B) Approximatley 100 cells for each experimental group was analyzed and the results are presented as the mean ± S.D.

A B

C D

Figure 27. Increased Chromosome Missegregation Ultimately Results in Chromosomal Instability. Cells were treated with colcemid in order to obtain metaphase spreads. Chromosome content was determined by counting the individual chromosomes in at least 100 metaphases. The results are presented as a circos plot, where the concentric circle represents the relative ploidy and each point represents an individual cell. (A) shCTL (B) shCKAP2_4 (C) shCKAP2_8 (D) shCKAP2_12.

Figure 28. Spectral Karyotyping of CKAP2-depleted Cells. Mitotic cells were treated with colcemid in order to obtain metaphase spreads. Karyotypes analyzed by spectral karyotyping (SKY) showed the increased level of aneuploidy and chromosome instability in the CKAP2-depleted cells. (A) shCTL (B) shCKAP2_12, where a clonal rearrangement of chromosomes 5 and 6 was observed, as well as deletion of chromosome 3 and 21 (C) shCKAP2_8, where an aneuploid cell with several chromosome deletions and insertions was observed.

Section IV. Differential expression of CKAP2 in human colorectal cancer cell lines Chromosomal instability, CIN, is a major factor in the progression of many malignancies. Understanding the mechanism(s) by which CIN develops, particularly in the early stages, is crucial to understanding of cancer progression.

Mechanisms of chromosomal instability have been established, such as the defective signaling of spindle checkpoint proteins, sister-chromatid cohesion defects, or hyper-stable microtubule attachments to chromosomes, which ultimately results in elevated frequency of chromosome lagging (Thompson et al.,

2010; Bakhoum and Compton, 2012). Chromosomal instability has been shown to be a hallmark of solid malignancies. Colorectal cancer in particular, has been shown to overexpress CKAP2 and in general is highly aneuploid.

To determine if CKAP2 is differentially expressed in colorectal cancer cells and if its expression correlates with the level of aneuploidy, RNA was harvested from 25 established cell lines and CKAP2 expression analyzed by RT-PCR. The data show that CKAP2 is differentially expressed in colorectal cancer lines compared to normal colon mucosa (Figure 29A). This finding was supported by immunoblot analysis in seven CRC cell lines, including two diploid, mismatch repair deficient (HCT116 and DLD1) and five aneuploidy, mismatch repair proficient (CaCO2, HT29, SW480, SW620 and SW837) (Figure 29B). However, mRNA levels did not always directly correlate to protein levels for all of the cell lines. Next, to determine whether differential expression was due to differences in cell cycle distribution, the selected cell lines were synchronized and then analyzed by immunoblot. Each cell line was treated with nocodazole overnight

66 and harvested protein analysis. Synchronization was determined by a rounded cell morphology and increase in 4N DNA content after nocodazole treatment

(Figure 30 & 31). Both MIN and CIN positive cells successfully arrested in mitosis, which was congruent with previous findings that CRC cells have an intact spindle checkpoint. Immunoblot analysis of synchronized cells showed a marked increase in CKAP2 expression, and the differences congruent with those seen in unsynchronized cells (Figure 32).

Lastly, CKAP2 RNA expression was correlated with chromosome content and relative ploidy in the 25 CRC cell lines using previously published cytogenetic analysis. The data show that increased CKAP2 expression does not correlate with increased aneuploidy (Table 1).

Figure 29. Differential Expression of CKAP2 in Colorectal Cancer Cell Lines. (A) RNA was harvested from colorectal cell lines and analyzed by RT-PCR. The data is presented to as the relative fold-change compared to normal colon mucosa. (B) Asynchronous populations from colorectal cell lines, DLD1, CaCo-2, HCT116, SW480, SW620, SW837, and HT29 and harvested for immunoblot analysis with antibodies specific to CKAP2 and GAPDH.

Figure 30. MIN+ CRC Cell Lines Arrest in Mitosis after Nocodazole Treatment. Diploid CRC cell lines (A) DLD1 and (B) HCT116 were treated with nocodazole for 15 hours, cells stained with propidium iodide, and DNA content analyzed by FACS. Each panel contains a picture of asynchronous (left) and synchronous (right) populations of each cell type and histogram displaying the DNA content of asynchronous (red) and synchronous (blue) populations by FACS.

Figure 31. CIN+ CRC Cell Lines Arrest in Mitosis after Nocodazole Treatment. Aneuploid CRC cell lines (A) HT29, (B) SW480, and (C) SW620 were treated with nocodazole for 15 hours, cells stained with propidium iodide, and DNA content analyzed by FACS. Each panel contains a picture of asynchronous (left) and synchronous (right) populations of each cell type and histogram displaying the DNA content of asynchronous (red) and synchronous (blue) populations by FACS.

Figure 32. CKAP2 Expression in MIN+ and CIN+ CRC Cell Lines in Mitosis. Mitotic cells in DLD1, CaCO2, HTC116, SW480, SW620, SW837, and HT29 cell lines were harvested for immunoblot analysis with antibodies specific for CKAP2 and GAPDH. Differences in protein amounts correlated with respective mRNA level.

Table 1. Comparison of Relative Expression of CKAP2 RNA and Relative Ploidy. Colorectal cancer cell lines, both MIN+ and CIN+ are listed with relative CKAP2 expression (compared to normal mucosa), number of chromosomes and relative ploidy.

Chapter 4: Discussion

The experiments undertaken in this dissertation were designed to determine the function of microtubule-associated protein, CKAP2. MAPs have been shown to play an important role in bipolar spindle assembly and maintaining genomic stability. Alteration in the expression of several MAPs has been associated with alterations in microtubule dynamics, which results in defects in spindle structure, increased chromosome missegregation and ultimately aneuploidy (Wittman et al., 1998; Wu et al., 2008; Radulescu and

Cleveland 2010). Aneuploidy, a hallmark of many cancers, in particular of tumors of epithelial origin (Ried et al., 2012), constitutes a phenomenon referred as chromosomal instability (Lengauer et al., 1997). Approximately 65-80% of colorectal cancers are characterized by aneuploidy and chromosomal instability, which is often associated with enhanced tumorgenicity and decreased overall patient survival (Pino and Chung, 2010; Knutsen et al., 2010). On the other hand, 5-10% of colorectal cancers show a stable, near-diploid karyotype, but a mismatch repair deficient phenotype. This paradigm makes the colorectal cancer model a perfect model to study chromosomal instability.

CKAP2 is located in a locus on chromosome 13 that is commonly gained in colorectal cancer, whereas in other cancers, such as retinoblastoma and leukemia, it is frequently lost (Heim and Mitelman, 2009; Camps et al., in press).

However, CKAP2 expression has limited effects on cellular proliferation and viability, thus it is not likely that CKAP2 confers a growth advantage in cancers

73 where it is gained, in particular colorectal cancer. As a result, we sought to elucidate the function of CKAP2 and its role in the development of aneuploidy in the context of karyotypically stable colorectal cancer cell line DLD1.

Section I. Depletion of CKAP2 increases spindle pole defects The initial insights into the function of CKAP2 were the result of studies that utilized CKAP2 over-expression constructs, however, little information about

CKAP2 function was gained from these analyses. Further, depletion of CKAP2 in mammalian cell lines has yielded contradicting results. Localization of CKAP2 was interrogated by immunofluorescence staining of α- and γ-tubulin. In wild- type DLD1 cells, CKAP2 localized at the mitotic spindle, in particular at the spindle pole. A mitotic spindle-binding assay was then utilized to determine if

CKAP2 was indeed associated with microtubules. The presence of CKAP2 in the pellet fraction in untreated and nocodazole treated cells suggests that CKAP2 associates with the spindle, but does not directly bind microtubules. To elucidate the function of CKAP2, a karyotypically stable, diploid colorectal cancer cell line, DLD1, was transfected with shRNA and analyzed for mitotic defects, particularly defects in cell cycle timing and mitotic spindle defects.

Initial findings with RNAi in this study demonstrated that CKAP2 was not required for cell viability, as shown by Annexin-V or metabolic assay with

CellTiter blue. Previous studies that alter CKAP2 expression levels have shown that silencing CKAP2 yielded little effects on cell cycle whereas ectopic over- expression of CKAP2 resulted in mitotic arrest due to an increase in cells with monopolar spindles (Seki and Fang, 2007; Hong et al., 2007). In this study,

74 reduction of CKAP2 expression did not result in a discernable mitotic arrest, however a slight decrease in the length of the cell cycle (approximately two hours) and the length of mitosis (approximately 15 minutes) was observed.

While the shortening of mitosis does not account for the decrease in the length of overall cell cycle length, it does suggest defects in proper cell cycle regulation.

The dysregulation of mitosis with no apparent mitotic arrest, typically indicative of gross mitotic defects, prompted a further analysis of the mitotic spindle, in particular the spindle poles where CKAP2 is localized. A major function of the spindle pole is to bundle and anchor together new centrosome- nucleated mitotic microtubules after the nuclear envelope breakdown at the onset of mitosis. CKAP2-depleted cells showed an increase in spindle defects, most notably including an increase in multipolar spindles, a dispersion of γ-tubulin, and the detachment of the centrosome from the spindle pole. The increase in multipolar spindles could be due to a decrease in centrosome clustering and/or fragmentation of the pericentriolar matrix. Abnormal size and shape of γ-tubulin foci, an indicator of PCM fragmentation, was observed in a small percentage of

CKAP2-depleted cells. However, a more glaring observation was the differences in centrosome clustering in control and CKAP2-depleted cells. A slight increase in the number of γ-tubulin foci was observed in CKAP2-depleted cells, but it is important to note that a majority of the cells with supernumerary centrosomes in control cells clustered to form bipolar spindle poles, where as a majority of the

CKAP2-depleted cells with supernumerary centrosomes formed multipolar spindles. Centrosome clustering is one of several mechanisms utilized by cancer

75 cells to prevent multipolar asymmetrical cell divisions that drastically reduce cellular viability. Centrosome clustering is dependent on microtubule-based motors and microtubule-bundling proteins, such as Ncd, TACC3, ch-TOG,

MCAK, Eg5, and dynein, many of which are required for focusing of microtubule minus-ends and centrosome attachment to the spindle pole (Kwon et al., 2008;

Kramer et al., 2010).

Moreover, the increased spindle length suggests that there was an imbalance of forces on the spindle. This imbalance could put pressure on the centrosomes, which causes PCM fragmentation and formation of extra spindle poles (Lee and Rhee, 2010; Manning and Compton, 2007). A balance of microtubule polymerization and depolymerization controls relative spindle length.

Inhibition of proteins that play a role in MT depolymerization, e.g. Kinesin-8 and

Kinesin-13, as well as sliding motors and cohesion proteins can increase spindle length (Goshima et al., 2005). Spindle tension is also required for clustering, which is brought about by positioning of centrosomes by microtubule tension- dependent forces ultimately resulting in bipolar spindle formation. Spindle tension can be disrupted by a variety of mechanisms, including reduced microtubule formation and disturbed microtubule bundling and centrosome attachment (Leber et al., 2010).

The presence of γ-tubulin along the mitotic spindle is not uncommon. In interphase, γ-tubulin is recruited to microtubule-organizing centers from a soluble cytoplasmic pool of γ-tubulin by a process governed by mitotic kinases Aurora A and Plk1. During mitosis however, γ-tubulin complexes have been shown to

76 localize both to the centrosome and along the mitotic spindle (Doxsey et al.,

2005; Raynoud-Messina and Merdes, 2007). In addition, although γ-tubulin was dispersed from the centrosome, staining with pericentrin suggests that centrosome distribution and structure remained intact. Perturbation of well- known spindle proteins, such as NuMA or ch-TOG, have demonstrated a decrease in the integrity of the spindle pole, ultimately resulting in microtubule disorganization, multipolar spindles, and increased aberrant microtubule- kinetochore attachments (De Luca et al., 2008; Haren et al., 2009; Silk et al.,

2009).

Section II. CKAP2 Plays a Role in Tethering the Centrosome to the Spindle Pole The observation of spindle defects despite functional centrosomes prompted us to interrogate whether CKAP2 is required for microtubule nucleation. Interestingly, we identified a very unusual pattern of microtubule distribution after release from a nocodazole block. As early as two minutes after mitotic release, we observed a cage-like phenotype showing nascent α-tubulin staining distributed across the entire chromatin region in CKAP2-depleted cells.

This “cage” has been shown to consist of polymerized tubulin, but we also observed colocalization of γ-tubulin. Non-centrosomal sites of γ-tubulin can represent novel sites of nucleation or capped microtubule minus ends that were released from the centrosome (Raynaud-Messina and Merdes, 2007). We failed to observe microtubule nucleation at or near the kinetochores/chromatin, which suggests that in the observed cage-like structure, nucleation sites that normally

77 localize together at the centrosome are dispersed. Similar to chromosome- mediated nucleation, however, the microtubule filaments gathered into specific poles by 30 minutes after mitotic block release, albeit with an increase in the number of spindle poles. Furthermore, at 60 minutes, in both control and CKAP2- depleted cells, bipolar mitotic spindles were fully formed and functional. We interpreted these observations as indicating that CKAP2 is involved in the early anchoring of microtubules at the spindle pole. The increase in spindle intensity may indicate a slight increase in microtubule stabilization or enhanced microtubule nucleation, potentially via other proteins such as TACC or TPX2, both known to be potent microtubule nucleators in addition to γ-tubulin.

Microtubules formed from the centrosome do not remain tightly bound in early mitosis, but, nonetheless, they are anchored in the vicinity of the pericentriolar material (PCM) by a group of proteins that form the spindle pole matrix. Whether these microtubules come from newly polymerized microtubules detaching from the pole or whether they are fragmented minus-end microtubules is still undetermined. In CKAP2 depleted cells, microtubules nucleated in early mitosis are not held in the vicinity of the PCM and thus disperse across the nucleus. The presence of this phenotype corroborates the role of CKAP2 in focusing centrosomal microtubules and suggests its subcellular localization within the spindle pole matrix.

Early microtubule nucleation correlates with a significant increase in the recruitment of γ-tubulin. In early prophase, MT minus-ends are stabilized at the spindle pole by the γ-tubulin complex, γ-TuRC, cross-linking motors and MAPs

78 that localize at the spindle pole, as well as in the vicinity of the nuclear envelope,

The stabilization of microtubules from the developing asters helps microtubules from both centrosomes find each other and form overlapping associations. Upon nuclear envelope breakdown, MT dynamics change as interphase MTs are destabilized and centrosome nucleation of MTs is enhanced. In order to maintain spindle pole integrity, the formation of new MTs and organization of pre- existing MTs must be coordinated (Kline-Smith and Walczak, 2004).

One of the principal molecular components in tethering and focusing spindle microtubules at the poles is NuMA (Merdes et al., 1996). NuMA localizes to the spindle pole and has been shown to directly cross-link microtubules. Depletion of NuMA results in aberrant spindles, pole fragmentation, and dissociation of the centrosome from already assembled spindles, resulting in splayed microtubule ends (Silk et al., 2009; Haren et al., 2009; Radulescu and Cleveland, 2010).

CKAP2 showed functional similarities with NuMA and, like NuMA, CKAP2 mitotic activity has also shown to be regulated by phosphorylation and dephosphorylation residues near the C-terminus (Hong et al., 2008; Hong et al.,

2009b). However, depletion of CKAP2 does not compromise the localization and expression of NuMA. In fact, the distribution of NuMA after microtubule depolymerization and regrowth was unaltered. Therefore, we suggest that

CKAP2 and NuMA, although they might be functionally similar, do not act as a complex. This observation, along with unaltered localization and amount of

TPX2, led us to hypothesize that CKAP2 does not affect dynein-dependent transport of spindle pole organizing proteins.

Section III. Reduction of Spindle Pole Integrity Results in Chromosome Missegregation Inhibition of CKAP2 in C2C12 cells resulted in cells with increased nuclear size and abnormal morphology (e.g. invaginations, protrusions/blebs, and donut- like shapes) (Hong et al., 2009). Moreover, the same authors noted that these cells typically had a compromised nuclear lamina, as indicated by an uneven distribution and punctate staining of Lamin A/C. Abnormal nuclear morphology has been demonstrated to be a marker of aberrant chromosome segregation

(Gisselson, 2008). Further analysis of depleted cells showed an increase in anaphase bridge formation and lagging chromosomes. However, live-cell imaging of CKAP2-depleted cells showed no defects in cell cycle or cytokinesis

(Hong et al., 2009).

In DLD1, CKAP2-depleted cells resulted in a marked increase in anaphase lagging events, which resulted in micronuclei, nuclear blebs, and abnormal nuclei morphology. Further analysis with Hec1 immunostaining demonstrated that cells with suppressed CKAP2 had a higher frequency of merotelic kinetochore-microtubule attachments. Although most chromosomes with merotelic attachments segregate properly, a fraction of them remain at the spindle equator while the other chromosomes move to the poles (Cimini et al.,

2001). Recently, it has been suggested that segregation errors are a consequence of cells passing through a transient multipolar spindle intermediate that results in the formation of merotelic attachments (Ganem et al., 2009). Thus we hypothesize that suppression of CKAP2 results in aberrant anchoring of the

80 minus-ends at the pole, which increases multipolar spindle formation and subsequently increases aberrant microtubule-kinetochore attachments and chromosome missegregation.

Colon cancer DLD1 cells are mismatch repair positive, diploid cells with a stable karyotype (Knutsen et al., 2010). However, in CKAP2-depleted cells, there was an increased number of cells with abnormal DNA content, or aneuploidy, consistent with an increased rate of chromosome missegregation. Moreover, in clone CKAP2_12 we identified the formation of clonal chromosome rearrangements most likely as a consequence of the high level of chromosomal instability (Camps et al., 2005). Interestingly, a high frequency of polyploid cells was observed in both CKAP2 clones, with a majority having near-triploid chromosome content. There are several causes of polyploidy, most notably cell fusion, mitotic slippage, and failure to undergo cytokinesis (Storchova and

Pellman, 2004). Ectopic CKAP2 over-expression has been shown to result in cytokinesis failure, however there was little evidence failure of cytokinesis, such as an increase in binucleate cells or cleavage furrow regression in CKAP2- depleted cells. Thus, it is likely that cells become near-tetraploid via mitotic slippage, in which a persistent error causes the cell to escape checkpoint arrest and exit mitosis without undergoing anaphase or cytokinesis (Storchova and

Kuffer, 2008). Since asymmetrical cell division generated by multipolar cell divisions are usually not viable, tetraploidy might represent an intermediate on the route to aneuploidy as cells can maintain important components of the genomes and remain viable (Shi and King, 2005; Storchova and Kuffer, 2008).

Genome integrity relies on the equal distribution of replicated chromosomes to daughter cells during cell division. Chromosome segregation is a process that requires structural integrity of the spindle pole and a checkpoint signaling system, the spindle assembly checkpoint, that ensures the fidelity of the process. The spindle assembly checkpoint prevents chromosome segregation errors by controlling the onset of anaphase until all chromosomes are attached spindle microtubules, a process monitored by tension. The spindle checkpoint remains active as long as unattached kinetochores are present. In cells with merotelic attachments, all kinetochores are attached to microtubules generating spindle tension. Thus, this type of abnormal kinetochore-microtubule attachment is not readily detected by the spindle checkpoint machinery (Cimini et al., 2001;

Lampson et al., 2004). As a consequence, these incorrect attachments persist, resulting in an increase of lagging chromosomes. Studies have shown that the spindle assembly checkpoint in DLD1, as well as other colorectal cancer cell lines, is in fact functional (Tighe et al., 2001).

In addition to the spindle assembly checkpoint, there is a set of non- essential proteins that act as chaperones or as a scaffold whose absence might not be sufficient to threaten cellular viability. CKAP2 appears to be largely dispensable for the ultimate establishment of bipolar spindle poles, and depletion of CKAP2 does not result in checkpoint activation, as demonstrated by the lack of an accumulation of cells in mitosis. Previous studies in mouse and human fibroblasts showed that centrosome separation and establishment of the spindle apparatus were not noticeably affected in CKAP2-depleted cells (Hong et al.,

2009a). The same authors also concluded that congression of chromosomes at the metaphase plate appeared largely unaffected and suggested that functional aspects of the microtubule spindle apparatus were largely unaffected. The lack of

CKAP2 does not confer a dominant phenotype and indicates that CKAP2 may not be essential to maintain cellular viability, suggesting that there must be other proteins that compensate for its absence. The data in this study suggest that, although the mitotic spindle remains functional, CKAP2-depleted cells present a delay in tethering the microtubules to the spindle poles in early mitosis and this imbalance is sufficient to promote merotelic attachments, and consequently increase chromosome misalignment and lagging.

Section IV. CKAP2 in the Context of Cancer

A major issue that arises in the present study is the increase in CIN as a consequence of CKAP2 depletion, whereas cancers are generally associated with CKAP2 overexpression. Little is known regarding the expression of CKAP2 in cancer cells and its role in tumorigenesis. The first report that identified differential expression of CKAP2 in cancer was in large B cell lymphomas, where

CKAP2 was shown to present in highly proliferative tissues (Maouche-Chretien et al., 1998). In a later study, Bae et al., reported the upregulation of CKAP2 in primary gastric adenocarcinomas, which could then be exploited to differentiate between adenocarcinomas and tubular adenomas from the stomach (Bae et al.,

2003).

Data generated in our laboratory also indicated that CKAP2 is overexpressed in colon primary tumors when compared to normal mucosa. This trend was confirmed in in vitro mouse models of colorectal cancer. Overall,

CKAP2 is upregulated regardless of the level of aneuploidy, as both near-diploid and polyploid cell lines showed high levels of CKAP2 mRNA. In addition, this gene was overexpressed in spontaneously transformed mouse epithelial cells generated in our laboratory when compared to normal tissues. Altogether, although CKAP2 is governing the fidelity of chromosome segregation, our data suggest that expression of CKAP2 does not necessarily correlate with the level of aneuploidy.

Dissertation Summary

In summary, the cellular mechanism by which CKAP2 regulates proper chromosome segregation has been elucidated. CKAP2 is involved in tethering and anchoring microtubules in early mitosis to accurately form the spindle pole.

Although there is apparently enough redundancy to ensure spindle formation and chromosome segregation, CKAP2 depletion increases the formation of multipolar spindles, likely due to an inability to cluster centrosomes. This geometric defect allows for greater accessibility to kinetochores, resulting in an increase in aberrant microtubule-kinetochore attachments, a higher frequency of chromosome missegregation and ultimately chromosomal instability.

Concluding Remarks

Mitosis is the process by which replicated DNA is evenly distributed into two distinct daughter cells. The mitotic spindle is the essential component of mitosis that allows for proper chromosome segregation. Alterations in mitotic proteins, in particular MAPs, often result in defects in proper assembly of the mitotic spindle and chromosome segregation. Increased rates of chromosome missegregation lead to aneuploidy, which can provide cells with selective growth advantage (by the gain of oncogenes or the loss of tumor suppressors), and ultimately promote tumor initiation and progression. In order to further characterize CKAP2 function, it will be important to study CKAP2 in

85 untransformed cells to determine exactly how it contributes to proper spindle assembly and aneuploidy.

References

Bae, C.-D., Sung, Y.-S., Jeon, S.-M., Suh, Y., Yang, H.-K., Kim, Y.I., Park, K.-H., Choi, J., Ahn, G., and Park, J. (2003). Up-regulation of cytoskeletal-associated protein 2 in primary human gastric adenocarcinomas. J. Cancer Res. Clin. Oncol. 129, 621–630.

Bakhoum, S.F., and Compton, D.A. (2012). Chromosomal instability and cancer: a complex relationship with therapeutic potential. J. Clin. Invest. 122, 1138–1143.

Blagden, S.P., and Glover, D.M. (2003). Polar expeditions--provisioning the centrosome for mitosis. Nat Cell Biol 5, 505–511.

Camps, J., Ponsa, I., Ribas, M., Prat, E., Egozcue, J., Peinado, M.A., and Miró, R. (2005). Comprehensive measurement of chromosomal instability in cancer cells: combination of fluorescence in situ hybridization and cytokinesis-block micronucleus assay. Faseb J. 19, 828–830.

Charrasse, S., Schroeder, M., Gauthier-Rouviere, C., Ango, F., Cassimeris, L., Gard, D.L., and Larroque, C. (1998). The TOGp protein is a new human microtubule-associated protein homologous to the Xenopus XMAP215. J. Cell. Sci. 111 ( Pt 10), 1371–1383.

Ciciarello, M., Mangiacasale, R., and Lavia, P. (2007). Spatial control of mitosis by the GTPase Ran. Cell. Mol. Life Sci. 64, 1891–1914.

Cimini, D., Howell, B., Maddox, P., Khodjakov, A., Degrassi, F., and Salmon, E.D. (2001). Merotelic kinetochore orientation is a major mechanism of aneuploidy in mitotic mammalian tissue cells. The Journal of Cell Biology 153, 517–527.

Compton, D.A. (1998). Focusing on spindle poles. J. Cell. Sci. 111 ( Pt 11), 1477–1481.

Compton, D.A. (2000). Spindle assembly in animal cells. Annu. Rev. Biochem. 69, 95–114.

De Luca, M., Brunetto, L., Asteriti, I.A., Giubettini, M., Lavia, P., and Guarguaglini, G. (2008). Aurora-A and ch-TOG act in a common pathway in control of spindle pole integrity. Oncogene 27, 6539–6549.

Doxsey, S., Zimmerman, W., and Mikule, K. (2005). Centrosome control of the cell cycle. Trends in Cell Biology 15, 303–311.

Elledge, S.J. (1996). Cell cycle checkpoints: preventing an identity crisis. Science 274, 1664–1672.

Ganem, N.J., Godinho, S.A., and Pellman, D. (2009). A mechanism linking extra centrosomes to chromosomal instability. Nature 460, 278–282.

Garrett, S., Auer, K., Compton, D.A., and Kapoor, T.M. (2002). hTPX2 is required for normal spindle morphology and centrosome integrity during vertebrate cell division. Curr. Biol. 12, 2055–2059.

Ghadimi, B.M., Sackett, D.L., Difilippantonio, M.J., Schröck, E., Neumann, T., Jauho, A., Auer, G., and Ried, T. (2000). Centrosome amplification and instability occurs exclusively in aneuploid, but not in diploid colorectal cancer cell lines, and correlates with numerical chromosomal aberrations. Genes Chromosomes Cancer 27, 183–190.

Gisselsson, D. (2008). Classification of chromosome segregation errors in cancer. Chromosoma 117, 511–519.

Goshima, G., Wollman, R., Stuurman, N., Scholey, J.M., and Vale, R.D. (2005). Length control of the metaphase spindle. Curr. Biol. 15, 1979–1988.

Gruss, O.J., Wittmann, M., Yokoyama, H., Pepperkok, R., Kufer, T., Silljé, H., Karsenti, E., Mattaj, I.W., and Vernos, I. (2002). Chromosome-induced microtubule assembly mediated by TPX2 is required for spindle formation in HeLa cells. Nat Cell Biol 4, 871–879.

Haren, L., Gnadt, N., Wright, M., and Merdes, A. (2009). NuMA is required for proper spindle assembly and chromosome alignment in prometaphase. BMC Res Notes 2, 64.

Heim, S., and Mitelman, F. (2011). Cancer Cytogenetics (Wiley-Blackwell).

Hong, K.-U., Choi, Y.-B., Lee, J.-H., Kim, H.-J., Kwon, H.-R., Seong, Y.-S., Kim, H.T., Park, J., Bae, C.-D., and Hong, K.-M. (2008). Transient phosphorylation of tumor associated microtubule associated protein (TMAP)/cytoskeleton associated protein 2 (CKAP2) at Thr-596 during early phases of mitosis. Exp. Mol. Med. 40, 377–386.

Hong, K.-U., Kim, E., Bae, C.-D., and Park, J. (2009a). TMAP/CKAP2 is essential for proper chromosome segregation. Cell Cycle 8, 314–324.

Hong, K.-U., Kim, H.-J., Bae, C.-D., and Park, J. (2009b). Characterization of mitosis-specific phosphorylation of tumor-associated microtubule-associated protein. Exp. Mol. Med. 41, 832–840.

Hong, K.-U., Kim, H.-J., Kim, H.-S., Seong, Y.-S., Hong, K.-M., Bae, C.-D., and Park, J. (2009c). Cdk1-cyclin B1-mediated phosphorylation of tumor-associated microtubule-associated protein/cytoskeleton-associated protein 2 in mitosis. J. Biol. Chem. 284, 16501–16512.

Hong, K.-U., Park, Y.S., Seong, Y.-S., Kang, D., Bae, C.-D., and Park, J. (2007). Functional importance of the anaphase-promoting complex-Cdh1-mediated degradation of TMAP/CKAP2 in regulation of spindle function and cytokinesis. Mol. Cell. Biol. 27, 3667–3681.

Jeon, S.-M., Choi, B., Hong, K.-U., Kim, E., Seong, Y.-S., Bae, C.-D., and Park, J. (2006). A cytoskeleton-associated protein, TMAP/CKAP2, is involved in the proliferation of human foreskin fibroblasts. Biochem. Biophys. Res. Commun. 348, 222–228.

Jin, Y., Murakumo, Y., Ueno, K., Hashimoto, M., Watanabe, T., Shimoyama, Y., Ichihara, M., and Takahashi, M. (2004). Identification of a mouse cytoskeleton- associated protein, CKAP2, with microtubule-stabilizing properties. Cancer Sci. 95, 815–821.

Kang, D.-S., Hong, K.-M., Park, J., and Bae, C.-D. (2012). Cyclin A regulates a cell-cycle-dependent expression of CKAP2 through phosphorylation of Sp1. Biochem. Biophys. Res. Commun. 420, 822–827.

Kim, A., Ban, G., Song, M.-S., Bae, C.-D., Park, J., and Lee, S.-W. (2007). Selective regression of cells expressing mouse cytoskeleton-associated protein 2 transcript by trans-splicing ribozyme. Oligonucleotides 17, 95–103.

Kim, H.-J., Kwon, H.-R., Bae, C.-D., Park, J., and Hong, K.U. (2010). Specific primary sequence requirements for Aurora B kinase-mediated phosphorylation and subcellular localization of TMAP during mitosis. Cell Cycle 9, 2027–2036.

Kim, H.-S., Choi, Y.-B., Lee, J.-H., Park, S.-Y., Kim, H.-K., Koh, J.-S., Yi, S.-Y., Kim, K.-T., Hong, K.-U., Park, J., et al. (2012). Condensed chromatin staining of CKAP2 as surrogate marker for mitotic figures. J. Cancer Res. Clin. Oncol. 138, 95–102.

Kim, K., and Rhee, K. (2011). The pericentriolar satellite protein CEP90 is crucial for integrity of the mitotic spindle pole. J. Cell. Sci. 124, 338–347.

Kline-Smith, S.L., and Walczak, C.E. (2004). Mitotic spindle assembly and chromosome segregation: refocusing on microtubule dynamics. Mol. Cell 15, 317–327.

Knutsen, T., Padilla-Nash, H.M., Wangsa, D., Barenboim-Stapleton, L., Camps, J., McNeil, N., Difilippantonio, M.J., and Ried, T. (2010). Definitive molecular cytogenetic characterization of 15 colorectal cancer cell lines. Genes Chromosomes Cancer 49, 204–223.

Kufer, T.A. (2002). Human TPX2 is required for targeting Aurora-A kinase to the spindle. The Journal of Cell Biology 158, 617–623.

Kwon, M., Godinho, S.A., Chandhok, N.S., Ganem, N.J., Azioune, A., Thery, M.,

89 and Pellman, D. (2008). Mechanisms to suppress multipolar divisions in cancer cells with extra centrosomes. Genes Dev. 22, 2189–2203.

Lampson, M.A., Renduchitala, K., Khodjakov, A., and Kapoor, T.M. (2004). Correcting improper chromosome-spindle attachments during cell division. Nat Cell Biol 6, 232–237.

Leber, B., Maier, B., Fuchs, F., Chi, J., Riffel, P., Anderhub, S., Wagner, L., Ho, A.D., Salisbury, J.L., Boutros, M., et al. (2010). Proteins required for centrosome clustering in cancer cells. Sci Transl Med 2, 33ra38.

Lee, S., and Rhee, K. (2010). CEP215 is involved in the dynein-dependent accumulation of pericentriolar matrix proteins for spindle pole formation. Cell Cycle 9, 774–783.

Lengauer, C., Kinzler, K.W., and Vogelstein, B. (1997). Genetic instability in colorectal cancers. Nature 386, 623–627.

Ma, N., Titus, J., Gable, A., Ross, J.L., and Wadsworth, P. (2011). TPX2 regulates the localization and activity of Eg5 in the mammalian mitotic spindle. The Journal of Cell Biology 195, 87–98.

Manning, A.L., and Compton, D.A. (2007). Mechanisms of spindle-pole organization are influenced by kinetochore activity in mammalian cells. Curr. Biol. 17, 260–265.

Maouche-Chrétien, L., Deleu, N., Badoual, C., Fraissignes, P., Berger, R., Gaulard, P., Roméo, P.H., and Leroy-Viard, K. (1998). Identification of a novel cDNA, encoding a cytoskeletal associated protein, differentially expressed in diffuse large B cell lymphomas. Oncogene 17, 1245–1251.

Merdes, A., and Cleveland, D.W. (1997). Pathways of spindle pole formation: different mechanisms; conserved components. The Journal of Cell Biology 138, 953–956.

Nogales, E. (1999). A structural view of microtubule dynamics. Cell. Mol. Life Sci. 56, 133–142.

Pino, M.S., and Chung, D.C. (2010). The chromosomal instability pathway in colon cancer.

Radulescu, A.E., and Cleveland, D.W. (2010). NuMA after 30 years: the matrix revisited. Trends in Cell Biology 20, 214–222.

Raynaud-Messina, B., and Merdes, A. (2007). Gamma-tubulin complexes and microtubule organization. Curr. Opin. Cell Biol. 19, 24–30.

Ried, T., Hu, Y., Difilippantonio, M.J., Ghadimi, B.M., Grade, M., and Camps, J.

(2012). The consequences of chromosomal aneuploidy on the transcriptome of cancer cells. Biochim. Biophys. Acta 1819, 784–793.

Rieder, C.L., and Salmon, E.D. (1998). The vertebrate cell kinetochore and its roles during mitosis. Trends in Cell Biology 8, 310–318.

Scholey, J.M., Brust-Mascher, I., and Mogilner, A. (2003). Cell division. Nature 422, 746–752.

Seki, A., and Fang, G. (2007). CKAP2 is a spindle-associated protein degraded by APC/C-Cdh1 during mitotic exit. J. Biol. Chem. 282, 15103–15113.

Shi, Q., and King, R.W. (2005). Chromosome nondisjunction yields tetraploid rather than aneuploid cells in human cell lines. Nature 437, 1038–1042.

Silk, A.D., Holland, A.J., and Cleveland, D.W. (2009). Requirements for NuMA in maintenance and establishment of mammalian spindle poles. The Journal of Cell Biology 184, 677–690.

Skibbens, R.V., Skeen, V.P., and Salmon, E.D. (1993). Directional instability of kinetochore motility during chromosome congression and segregation in mitotic newt lung cells: a push-pull mechanism. The Journal of Cell Biology 122, 859– 875.

Spittle, C., Charrasse, S., Larroque, C., and Cassimeris, L. (2000). The interaction of TOGp with microtubules and tubulin. J. Biol. Chem. 275, 20748– 20753.

Storchova, Z., and Kuffer, C. (2008). The consequences of tetraploidy and aneuploidy. J. Cell. Sci. 121, 3859–3866.

Storchova, Z., and Pellman, D. (2004). From polyploidy to aneuploidy, genome instability and cancer. Nat. Rev. Mol. Cell Biol. 5, 45–54.

Tedeschi, A., Ciciarello, M., Mangiacasale, R., Roscioli, E., Rensen, W.M., and Lavia, P. (2007). RANBP1 localizes a subset of mitotic regulatory factors on spindle microtubules and regulates chromosome segregation in human cells. J. Cell. Sci. 120, 3748–3761.

Thompson, S.L., Bakhoum, S.F., and Compton, D.A. (2010). Mechanisms of chromosomal instability. Curr. Biol. 20, R285–R295.

Tighe, A., Johnson, V.L., Albertella, M., and Taylor, S.S. (2001). Aneuploid colon cancer cells have a robust spindle checkpoint. EMBO Rep. 2, 609–614.

Tsuchihara, K., Lapin, V., Bakal, C., Okada, H., Brown, L., Hirota-Tsuchihara, M., Zaugg, K., Ho, A., Itie-Youten, A., Harris-Brandts, M., et al. (2005). Ckap2 regulates aneuploidy, cell cycling, and cell death in a p53-dependent manner.

Cancer Res. 65, 6685–6691.

Valiron, O., Caudron, N., and Job, D. (2001). Microtubule dynamics. Cell. Mol. Life Sci. 58, 2069–2084.

Walczak, C.E., and Heald, R. (2008). Mechanisms of mitotic spindle assembly and function. Int. Rev. Cytol. 265, 111–158.

Wiese, C., and Zheng, Y. (2006). Microtubule nucleation: gamma-tubulin and beyond. J. Cell. Sci. 119, 4143–4153.

Wittmann, T., Boleti, H., Antony, C., Karsenti, E., and Vernos, I. (1998). Localization of the kinesin-like protein Xklp2 to spindle poles requires a leucine zipper, a microtubule-associated protein, and dynein. The Journal of Cell Biology 143, 673–685.

Wu, G., Lin, Y.-T., Wei, R., Chen, Y., Shan, Z., and Lee, W.-H. (2008). Hice1, a novel microtubule-associated protein required for maintenance of spindle integrity and chromosomal stability in human cells. Mol. Cell. Biol. 28, 3652– 3662.