Citation
Daxx and Rassf1 Define a Novel Mitotic Stress Checkpoint That Is Critical for Cellular Taxol Response

Material Information

Title:
Daxx and Rassf1 Define a Novel Mitotic Stress Checkpoint That Is Critical for Cellular Taxol Response
Creator:
Lindsay, Cory R.
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (89 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences
Molecular Cell Biology (IDP)
Committee Chair:
Ishov, Alexander M.
Committee Members:
Robertson, Keith D.
Liao, Daiqing
Swanson, Maurice S.
Bungert, Jorg
Graduation Date:
5/1/2008

Subjects

Subjects / Keywords:
Apoptosis ( jstor )
Breast cancer ( jstor )
Cell cycle ( jstor )
Cell lines ( jstor )
Microtubules ( jstor )
Mitosis ( jstor )
Mitotic spindle apparatus ( jstor )
Small interfering RNA ( jstor )
Taxoids ( jstor )
Tumors ( jstor )
Molecular Cell Biology (IDP) -- Dissertations, Academic -- UF
daxx, mitosis, rassf1, taxol
Genre:
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Medical Sciences thesis, Ph.D.

Notes

Abstract:
Taxanes, a family of compounds which includes taxol, are powerful chemotherapy agents used for breast cancer treatment. Large numbers of patients, however, are resistant to these drugs for unknown reasons. Taxol binds and hyper-stabilizes microtubules, but mutations or alterations in tubulin occur very rarely in cancers and cannot itself explain the majority of Taxol-resistance observed in patients. Currently, it is thought that defects in mitotic proteins may affect Taxol sensitivity in cells. Here, Daxx and Rassf1 are identified as novel regulators of cellular Taxol response. Daxx is a ubiquitously expressed and highly conserved nuclear protein with enigmatic roles in transcription and apoptosis. Increased prometaphase index in Daxx deficient embryos and aneuploidy of Daxx knockout cells was observed which suggested a potential function of Daxx in mitosis and cell division. During interphase, Daxx remains a strictly nuclear associated protein localized to PML bodies or heterochromatin. Upon nuclear envelope breakdown, Daxx was found to co-localize and interact with Rassf1 at mitotic spindles. Rassf1 is a cytoplasmic, microtubule-associated protein that is important for normal mitotic progression and cell division. Daxx was also shown to be important for the proper timing and progression of early mitosis. Together, Daxx and Rassf1 define a novel mitotic stress checkpoint that enables cells to efficiently exit mitosis (and eventually die) when encountered with specific stress stimuli during mitosis, including Taxol. In the absence of Daxx or Rassf1, cells treated with Taxol remain arrested in mitosis due to a sustained mitotic spindle checkpoint. Upon Taxol decay or removal, these cells can resume mitosis and complete cell division: thus being Taxol resistant. Inhibition of the spindle checkpoint using Aurora Kinase inhibitors efficiently abolished Taxol resistance in Daxx and Rassf1 depleted cells. Deregulation of most known mitotic proteins leads to enhanced Taxol response. Absence or depletion of Daxx and Rassf1, in contrast, results in increased drug resistance. In the future, Daxx and Rassf1 may be useful predictive markers for the proper selection of patients for taxane chemotherapy. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2008.
Local:
Adviser: Ishov, Alexander M.
Statement of Responsibility:
by Cory R. Lindsay

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Cory R. Lindsay. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/11/2008
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:

lindsay_c.pdf

lindsay_c_Page_31.txt

lindsay_c_Page_05.txt

lindsay_c_Page_08.txt

lindsay_c_Page_70.txt

lindsay_c_Page_49.txt

lindsay_c_Page_64.txt

lindsay_c_Page_46.txt

lindsay_c_Page_51.txt

lindsay_c_Page_14.txt

lindsay_c_Page_75.txt

lindsay_c_Page_07.txt

lindsay_c_Page_76.txt

lindsay_c_Page_66.txt

lindsay_c_Page_02.txt

lindsay_c_Page_10.txt

lindsay_c_Page_06.txt

lindsay_c_Page_62.txt

lindsay_c_Page_25.txt

lindsay_c_Page_69.txt

lindsay_c_Page_21.txt

lindsay_c_Page_65.txt

lindsay_c_Page_24.txt

lindsay_c_Page_22.txt

lindsay_c_Page_27.txt

lindsay_c_Page_09.txt

lindsay_c_Page_68.txt

lindsay_c_Page_28.txt

lindsay_c_Page_18.txt

lindsay_c_Page_79.txt

lindsay_c_Page_57.txt

lindsay_c_Page_67.txt

lindsay_c_Page_45.txt

lindsay_c_Page_26.txt

lindsay_c_Page_35.txt

lindsay_c_Page_73.txt

lindsay_c_Page_17.txt

lindsay_c_Page_36.txt

lindsay_c_Page_29.txt

lindsay_c_Page_71.txt

lindsay_c_Page_20.txt

lindsay_c_Page_33.txt

lindsay_c_Page_58.txt

lindsay_c_Page_53.txt

lindsay_c_Page_39.txt

lindsay_c_Page_83.txt

lindsay_c_Page_13.txt

lindsay_c_Page_41.txt

lindsay_c_Page_80.txt

lindsay_c_Page_16.txt

lindsay_c_Page_19.txt

lindsay_c_Page_43.txt

lindsay_c_Page_03.txt

lindsay_c_Page_87.txt

lindsay_c_Page_01.txt

lindsay_c_Page_63.txt

lindsay_c_Page_72.txt

lindsay_c_Page_23.txt

lindsay_c_Page_77.txt

lindsay_c_Page_55.txt

lindsay_c_Page_47.txt

lindsay_c_Page_74.txt

lindsay_c_Page_60.txt

lindsay_c_Page_78.txt

lindsay_c_Page_88.txt

lindsay_c_Page_15.txt

lindsay_c_Page_32.txt

lindsay_c_Page_59.txt

lindsay_c_Page_61.txt

lindsay_c_Page_40.txt

lindsay_c_Page_50.txt

lindsay_c_Page_34.txt

lindsay_c_Page_81.txt

lindsay_c_Page_85.txt

lindsay_c_Page_04.txt

lindsay_c_Page_38.txt

lindsay_c_Page_56.txt

lindsay_c_Page_30.txt

lindsay_c_Page_37.txt

lindsay_c_Page_54.txt

lindsay_c_Page_42.txt

lindsay_c_Page_44.txt

lindsay_c_pdf.txt

lindsay_c_Page_12.txt

lindsay_c_Page_86.txt

lindsay_c_Page_48.txt

lindsay_c_Page_89.txt

lindsay_c_Page_11.txt

lindsay_c_Page_52.txt

lindsay_c_Page_82.txt

lindsay_c_Page_84.txt


Full Text





3-7). Accruing evidence suggests that the tight orchestration of events during mitosis combines

seemingly unrelated factors at critical junctures during cellular division (Tsai et al., 2006). A

number of essential interactions occur during pro-metaphase when the nuclear envelope

disintegrates and no longer compartmentalizes the nucleus and cytoplasm. During this time,

Daxx is released from the nucleus where it is associated with ND10/PML bodies and

accumulates at the mitotic spindle apparatus. It is at this spatiotemporal point that Daxx interacts

with Rassfl (Figure 5-4, Figure 6-1 summary). Future studies will reveal whether Daxx

association with the spindle apparatus is dependent on presence of Rassfl.

To accompany observations of Daxx-dependent paclitaxel response, Rassfl was identified as

a novel Daxx-interacting protein that was also confirmed to be important for cellular taxol

response (Figure 5-5). This interaction was mapped to the C spliceform of Rassfl (RassflC) and

interaction between RassflA and RassflC by dimerization was also confirmed (Figure 5-2).

These novel interactions are thought to form a complex during mitosis and may perform critical

regulatory processes including proper cellular response to mitotic stresses such as paclitaxel,

nocodazole and other microtubule inhibiting compounds. These chemicals, moreover, can

induce mitotic stress in several ways, depending on the stage of mitosis in which the stress is

applied. Proper separation of the centrosomes during prophase, correct alignment of

chromosomes during metaphase and faithful segregation of chromosomes during anaphase are

all processes which can be altered or affected when external stress (i.e. paclitaxel) is applied.

Together, Daxx and Rassfl define a unique mitotic stress checkpoint during pro-metaphase

(Figure 6-2, summary). Cells lacking Daxx or RassflA arrest in pro-metaphase during taxol

treatment. From time-lapse studies, it is suggested that Daxx-depleted cells have chromosomes

that remain unable to properly align at the metaphase plate under treatment conditions, thus









Dammann, R., Yang, G., and Pfeifer, G.P. (2001). Hypermethylation of the cpG island of Ras
association domain family 1A (RASSF1A), a putative tumor suppressor gene from the 3p21.3
locus, occurs in a large percentage of human breast cancers. Cancer Res 61, 3105-3109.

Dellaire, G., Eskiw, C.H., Dehghani, H., Ching, R.W., and Bazett-Jones, D.P. (2006). Mitotic
accumulations of PML protein contribute to the re-establishment of PML nuclear bodies in G1. J
Cell Sci 119, 1034-1042.

Ditchfield, C., Johnson, V.L., Tighe, A., Ellston, R., Haworth, C., Johnson, T., Mortlock, A.,
Keen, N., and Taylor, S.S. (2003). Aurora B couples chromosome alignment with anaphase by
targeting BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 161, 267-280.

Donninger, H., Vos, M.D., and Clark, G.J. (2007). The RASSF1A tumor suppressor. J Cell Sci
120, 3163-3172.

Ecsedy, J.A., Michaelson, J.S., and Leder, P. (2003). Homeodomain-interacting protein kinase 1
modulates Daxx localization, phosphorylation, and transcriptional activity. Mol Cell Biol 23,
950-960.

Emelyanov, A.V., Kovac, C.R., Sepulveda, M.A., and Birshtein, B.K. (2002). The interaction of
Pax5 (BSAP) with Daxx can result in transcriptional activation in B cells. J Biol Chem 277,
11156-11164.

Fenton, S.L., Dallol, A., Agathanggelou, A., Hesson, L., Ahmed-Choudhury, J., Baksh, S.,
Sardet, C., Dammann, R., Minna, J.D., Downward, J., et al. (2004). Identification of the E1A-
regulated transcription factor p120 E4F as an interacting partner of the RASSF1A candidate
tumor suppressor gene. Cancer Res 64, 102-107.

Fu, J., Bian, M., Jiang, Q., and Zhang, C. (2007). Roles of Aurora kinases in mitosis and
tumorigenesis. Mol Cancer Res 5, 1-10.

Giannakakou, P., Robey, R., Fojo, T., and Blagosklonny, M.V. (2001). Low concentrations of
paclitaxel induce cell type-dependent p53, p21 and G1/G2 arrest instead of mitotic arrest:
molecular determinants of paclitaxel-induced cytotoxicity. Oncogene 20, 3806-3813.

Gostissa, M., Morelli, M., Mantovani, F., Guida, E., Piazza, S., Collavin, L., Brancolini, C.,
Schneider, C., and Del Sal, G. (2004). The transcriptional repressor hDaxx potentiates p53-
dependent apoptosis. J Biol Chem 279, 48013-48023.

Hannemann, J., Oosterkamp, H.M., Bosch, C.A., Velds, A., Wessels, L.F., Loo, C., Rutgers, E.J.,
Rodenhuis, S., and van de Vijver, M.J. (2005). Changes in gene expression associated with
response to neoadjuvant chemotherapy in breast cancer. J Clin Oncol 23, 3331-3342.

Hari, M., Wang, Y., Veeraraghavan, S., and Cabral, F. (2003a). Mutations in alpha- and beta-
tubulin that stabilize microtubules and confer resistance to colcemid and vinblastine. Mol Cancer
Ther 2, 597-605.




































Apoptosis

25%

20%

15%
10%

5%
0%

Oh 12h 24h 36h

--MDAMB468 0 T47D

Figure 4-4. Response of MDA MB 468 and T47D Breast Cancer Cell Lines to Increased
Concentration ofPaclitaxel. MDA MB 468 and T47D cell lines were treated with
100 nM paclitaxel for 12h, 24h and 36h or mock treated (control). A) DNA was
stained with HOECHST 33342 and cells were characterized as being mitotic, B)
micronuclei or C) apoptotic, based on nuclear morphology. Cellular response to
paclitaxel between these two cell lines remained relatively unchanged despite
increased paclitaxel concentrations. Note: apoptosis does not significantly increase
with elevated concentrations of paclitaxel.


Mitotic Cells

70%
60% -
50%
40%
30%
20% -
10%
0%
Oh 12h 24h 36h

-*-MDAMB468 I T47D


Micronuclei

70%
60% -
50% -
40%
30%
20%
10% -
0% --
Oh 12h 24h 36h

-*-MDAMB468 I T47D








Inberphase & Pro ph ie Prretaphase
Cytoplasm



*t--
RRaflCD3








Nucleus NE breakdown
Figure 6-1. Dynamics of Daxx-Rassfl Interaction Throughout the Cell Cycle. During interphase
and beginning of mitosis prophasee), Daxx and Rassfl are compartmentally separated
in the cytoplasm and nucleus. After nuclear envelope (NE) breakdown, Daxx and
Rassfl can interact through RassflC. This association is maintained through
metaphase, but by late stages of mitosis, is absent.









partners ofRassflA, including microtubule associated protein 1A (MAP1A), MAP1B and

C190RF5 (Dallol et al., 2004). Several reports have shown because of the direct interaction

between RassflA, tubulin and tubulin-related proteins, it served to stabilize microtubules under

stress conditions (i.e. nocodazole treatment) (Liu et al., 2003; Rong et al., 2004). During mitosis,

over-expressed RassflA was observed to co-localize with centrosomes and microtubules during

prophase, the spindle apparatus (spindle poles and spindle fibers) during prometaphase,

metaphase and anaphase and microtubules as they reformed in divided daughter cells (Liu et al.,

2003; Song et al., 2004). The minimal interaction domain that is responsible for RassflA

binding to microtubules and for association with the spindle apparatus was mapped to amino

acids 120-288 of RassflA (Liu et al., 2003).



Functions of RassflC

While extensive studies on the epigenetic regulation of RassflA have been performed

coupled with essential studies into the function of RassflA and its importance in tumor

progression, very little is known about RassflC biology. Armesilla et al. (2004) used a yeast

two-hybrid screen to find novel interactors of plasma membrane Ca2+ pump 4b (PMCA4b)

which identified RassflC. The interaction between RassflC and PMCA4b was narrowed down

to a region that is common to both RassflC and RassflA, presumably showing that this

interaction is shared between both isoforms although this data was not shown. Potential tumor

suppressor functions of RassflC were first described by Li and colleagues demonstrating that

RassflC could substantially reduce anchorage independent growth of tumor cells and elicit cell

cycle arrest similar to RassflA (Li et al., 2004). Amaar et al. performed a similar investigation

showing that upon depletion of RassflC protein in H1299 cells that lack RassflA expression, it

caused a significant decrease in cell proliferation (Amaar et al., 2006). Upon over-expression of

28









Torii, S., Egan, D.A., Evans, R.A., and Reed, J.C. (1999). Human Daxx regulates Fas-induced
apoptosis from nuclear PML oncogenic domains (PODs). EMBO J 18, 6037-6049.

Tsai, M.Y., Wang, S., Heidinger, J.M., Shumaker, D.K., Adam, S.A., Goldman, R.D., and
Zheng, Y. (2006). A mitotic lamin B matrix induced by RanGTP required for spindle assembly.
Science 311, 1887-1893.

Vos, M.D., Martinez, A., Elam, C., Dallol, A., Taylor, B.J., Latif, F., and Clark, G.J. (2004). A
role for the RASSF1A tumor suppressor in the regulation of tubulin polymerization and genomic
stability. Cancer Res 64, 4244-4250.

Wang, L.G., Liu, X.M., Kreis, W., and Budman, D.R. (1999). The effect of antimicrotubule
agents on signal transduction pathways of apoptosis: a review. Cancer Chemother Pharmacol 44,
355-361.

Wang, Y., and Cabral, F. (2005). Paclitaxel resistance in cells with reduced beta-tubulin.
Biochim Biophys Acta 1744, 245-255.

Wassmann, K., and Benezra, R. (2001). Mitotic checkpoints: from yeast to cancer. Curr Opin
Genet Dev 11, 83-90.

Wood, K.W., Comwell, W.D., and Jackson, J.R. (2001). Past and future of the mitotic spindle as
an oncology target. Curr Opin Pharmacol 1, 370-377.

Xia, G., Luo, X., Habu, T., Rizo, J., Matsumoto, T., and Yu, H. (2004). Conformation-specific
binding of p3 l(comet) antagonizes the function of Mad2 in the spindle checkpoint. Embo J 23,
3133-3143.

Xue, Y., Gibbons, R., Yan, Z., Yang, D., McDowell, T.L., Sechi, S., Qin, J., Zhou, S., Higgs, D.,
and Wang, W. (2003). The ATRX syndrome protein forms a chromatin-remodeling complex
with Daxx and localizes in promyelocytic leukemia nuclear bodies. Proc Natl Acad Sci U S A
100, 10635-10640.

Yang, X., Khosravi-Far, R., Chang, H.Y., and Baltimore, D. (1997). Daxx, a novel Fas-binding
protein that activates JNK and apoptosis. Cell 89, 1067-1076.

Yeung, T.K., Germond, C., Chen, X., and Wang, Z. (1999). The mode of action of taxol:
apoptosis at low concentration and necrosis at high concentration. Biochem Biophys Res
Commun 263, 398-404.

Yu, X., Minter-Dykhouse, K., Malureanu, L., Zhao, W.M., Zhang, D., Merkle, C.J., Ward, I.M.,
Saya, H., Fang, G., van Deursen, J., and Chen, J. (2005). Chfr is required for tumor suppression
and Aurora A regulation. Nat Genet 37, 401-406.

Zhao, L.Y., Liu, J., Sidhu, G.S., Niu, Y., Liu, Y., Wang, R., and Liao, D. (2004). Negative
regulation of p53 functions by Daxx and the involvement of MDM2. J Biol Chem 279, 50566-
50579.




















B













Y XZ Y XZ Y XZ




X





YZ YZ YZ
GFP-RasfD A GFP-Ras sC GFP-Da
Tub uIhi| Tub ui~in














GFP-RasEsfA GFP-R3af1C GFP-Daxx
Figure 5-3. Cellular Distribution ofDaxx and Rassfl During Interphase. A) Immunostaining of
endogenous RassflC and Daxx in interphase HEp2 cells. B) Biochemical separation
of nuclear and cytosolic fractions from HEp2 cells. Note nuclear association of Daxx
and cytosolic association of RassflA. C) 3D confocal imaging of transiently over-
expressed GFP-RassflA (far left), GFP-RassflC (middle) and GFP-Daxx (far right).
Note the compartmentally separated expression of Daxx and Rassfl.









CHAPTER 3
DAXX FUNCTION IN MITOSIS

Introduction

Components of the mitotic spindle checkpoint, as well as factors involved in the controlled

regulation of the spindle checkpoint and other mitotic processes, are necessary for proper

partitioning of chromosomes into daughter cells. In the absence of such factors, cells can

mishandle proper chromosomal alignment and segregation during metaphase and anaphase,

resulting in lagging chromosomes and aneuploidy (Ikui et al., 2005). Proper timing and the rate

of mitotic progression can also be affected in the absence of one or more mitotic proteins. For

this reason, the cell cycle has been compared to a series of molecular timers: "clocks" that

control the average duration of each cycle, and "dominoes" that make each step dependent on the

proper completion of a prior step (Meraldi et al., 2004). Cyclin B and Securin are two of the

primary mitotic substrates that drive entry into and exit from mitosis (Pines, 2006). Using

biochemical analysis of the relative stability of these proteins it is possible to measure the

duration of mitosis and cell division. Along with the targeted depletion of proteins and time-

lapse microscopy, many mitotic factors have been shown to affect timing and rate of mitotic

progression-adding leverage to their already established roles as mitotic regulators and adding

the possibility of uncovering novel proteins that may also influence cell cycle progression.



Results and Discussion

Originally identified as a pro-apoptotic Fas-interacting protein and later demonstrated to have

anti-apoptotic activity (Michaelson, 2000), Daxx is a ubiquitously expressed and highly

conserved nuclear protein that is also implicated in transcription regulation. As an almost

exclusively nuclear protein during interphase, Daxx interacts with intrinsic kinetochore protein









BIOGRAPHICAL SKETCH

Cory Lindsay was born in Laurel, Nebraska, a small town located in northeast Nebraska

cornering Iowa and South Dakota. He is the son of Archie and Virginia Lindsay, of Laurel, and

has three siblings Lori, Scott and Michelle. Cory attended Laurel-Concord High School and

graduated in 1998. During his high school years, Cory had several outstanding teachers who

motivated and inspired him to pursue academics. His first experience with science research

came during this time when he studied meiofauna biology and won several local, regional and

national awards for his efforts. He also developed an interest in collecting and breeding snakes

and other reptiles and amphibians, a hobby which he still pursues today.

After high school, Cory pursued a Bachelor of Science degree in Biological Sciences from

Wayne State College, in Wayne, NE and graduated in 2002. While an undergraduate, Cory was

fortunate to have many opportunities to perform independent science research. He conducted

ecological research in the British West Indies studying dwarf-geckos and snakes as part of a

collaborative effort with Avila College, Kansas City, MO; identified novel genes from the

Schistosoma mansoni genome at the Whitney Laboratory, St. Augustine, FL; performed

molecular biology experiments as an intern at the Bermuda Biological Station for Research, St.

Georges, Bermuda; and deciphered genetic disease abnormalities at the world famous Cold

Spring Harbor Laboratory, Long Island, NY.

Cory joined the University of Florida's IDP graduate school program in 2003 and finished

his Ph.D work in the laboratory of Dr. Alexander Ishov studying Daxx function in cellular taxol

response and cell cycle progression. He has presented this work at several national and

international conferences and published his work summarizing these results. Cory will continue

post-doctoral work in cancer research and diagnostics.










CHAPTER 5
DAXX INTERACTS WITH RAS-ASSOCIATION DOMAIN FAMILY 1 (RASSF1) WHICH
COOPERATE IN CELLULAR TAXOL RESPONSE



Introduction

Several lines of evidence suggest Daxx is important for proper timing of mitotic

progression and that paclitaxel (taxol) resistance can be dependent on the level of Daxx protein

in cells. Firstly, cells generated from Daxx-deficient embryos exhibit genomic instability while

tissue sectioning of E9.5 Daxx-deficient mouse embryos shows an accumulation of cells in early

mitosis (specifically pro-metaphase). Targeted depletion of Daxx also results in alteration of

mitotic progression during prophase and the pro-metaphase to anaphase transitions.

Additionally, proper degradation of Cyclin B in synchronized Daxx-depleted cells is altered.

Importantly, cells that are deficient or depleted of Daxx was proven to provide resistance to the

microtubule inhibiting drug taxol by arresting cells in mitosis; cells with a functional Daxx

protein have a shortened mitotic arrest resulting in micronucleation and cell death (Lindsay et al.,

2007).



Discussion and Results

To further investigate Daxx mitotic function and to study the potential mechanism by which

Daxx regulates taxol response during mitosis a search for Daxx-interacting proteins using Daxx

as "bait" in a yeast two-hybrid screen was performed. Several yeast two-hybrid screens have

been used in the past using Daxx as both "bait" and "prey." The majority of these screens have

used, or have identified through their screens, the carboxyl-terminal region of Daxx

(corresponding with PML- and Fas-interacting domains) which also contains a region of Daxx









Junn, E., Taniguchi, H., Jeong, B.S., Zhao, X., Ichijo, H., and Mouradian, M.M. (2005).
Interaction of DJ-1 with Daxx inhibits apoptosis signal-regulating kinase 1 activity and cell
death. Proc Natl Acad Sci U S A 102, 9691-9696.

Kang, D., Chen, J., Wong, J., and Fang, G. (2002). The checkpoint protein Chfr is a ligase that
ubiquitinates Plkl and inhibits Cdc2 at the G2 to M transition. J Cell Biol 156, 249-259.

Karunakaran, S., Diwakar, L., Saeed, U., Agarwal, V., Ramakrishnan, S., Iyengar, S., and
Ravindranath, V. (2007). Activation of apoptosis signal regulating kinase 1 (ASK1) and
translocation of death-associated protein, Daxx, in substantial nigra pars compact in a mouse
model of Parkinson's disease: protection by alpha-lipoic acid. FASEB J 21, 2226-2236.

Keen, N., and Taylor, S. (2004). Aurora-kinase inhibitors as anticancer agents. Nat Rev Cancer
4, 927-936.

Khelifi, A.F., D'Alcontres, M.S., and Salomoni, P. (2005). Daxx is required for stress-induced
cell death and JNK activation. Cell Death Differ 12, 724-733.

Kienitz, A., Vogel, C., Morales, I., Muller, R., and Bastians, H. (2005). Partial downregulation of
MAD1 causes spindle checkpoint inactivation and aneuploidy, but does not confer resistance
towards taxol. Oncogene 24, 4301-4310.

Kiriakidou, M., Driscoll, D.A., Lopez-Guisa, J.M., and Strauss, J.F., 3rd (1997). Cloning and
expression of primate Daxx cDNAs and mapping of the human gene to chromosome 6p21.3 in
the MHC region. DNA Cell Biol 16, 1289-1298.

Kitagawa, D., Kajiho, H., Negishi, T., Ura, S., Watanabe, T., Wada, T., Ichijo, H., Katada, T.,
and Nishina, H. (2006). Release of RASSFIC from the nucleus by Daxx degradation links DNA
damage and SAPK/JNK activation. EMBO J 25, 3286-3297.

Ko, Y.G., Kang, Y.S., Park, H., Seol, W., Kim, J., Kim, T., Park, H.S., Choi, E.J., and Kim, S.
(2001). Apoptosis signal-regulating kinase 1 controls the proapoptotic function of death-
associated protein (Daxx) in the cytoplasm. J Biol Chem 276, 39103-39106.

Kroemer, G., and Martin, S.J. (2005). Caspase-independent cell death. Nat Med 11, 725-730.

Lalioti, V.S., Vergarajauregui, S., Pulido, D., and Sandoval, I.V. (2002). The insulin-sensitive
glucose transporter, GLUT4, interacts physically with Daxx. Two proteins with capacity to bind
Ubc9 and conjugated to SUMO1. J Biol Chem 277, 19783-19791.

Lee, E.A., Keutmann, M.K., Dowling, M.L., Harris, E., Chan, G., and Kao, G.D. (2004).
Inactivation of the mitotic checkpoint as a determinant of the efficacy of microtubule-targeted
drugs in killing human cancer cells. Mol Cancer Ther 3, 661-669.

Lehembre, F., Muller, S., Pandolfi, P.P., and Dejean, A. (2001). Regulation ofPax3
transcriptional activity by SUMO-1-modified PML. Oncogene 20, 1-9.













MDA-MB 468


100%


80%


60%


40%


20%


I I I


o 0 "
(P-


OG1
OS
*l G2/M
OAp/Mic


100%


80%


60%


40%


20%


, In I- I I- -


OG1
OS
* G2/M
SAp/Mic


Figure 4-5. FACS Analysis of Cell Cycle Distribution after Paclitaxel Treatment. A) T47D and
B) MDA MB 468 cells were treated with 10nM paclitaxel for the indicated time and
sorted as G1 phase, S phase, G2/M phase, and apoptotic + micronucleated (Ap/Mic).
While the majority of T47D cells accumulate in G2/M after 36h of treatment, MDA-
MB 468 cells are mostly apoptotic + micronucleated at this time point. Unfortunately,
FACS does not allow discrimination between apoptotic and micronucleated cells.


T47D









high variety in Daxx protein level, intracellular distribution of Daxx is similar in all breast cancer

cell lines tested, showing Daxx co-localizing with PML (not shown). To study a correlation

between Daxx and cellular response to paclitaxel, cell lines with extreme level of Daxx were

chosen: T47D (low level of Daxx, Daxx/actin = 1.0) and MDA MB 468 (high level of Daxx,

Daxx/actin = 14.0) and tested for paclitaxel induced cell death measured by colony formation

assay. Increased survival of T47D cells was observed compared to MDA MB 468 cells (Figure

4-2). 24h of treatment reduces the survival of MDA MB 468 cells almost three fold, and at 48h

of treatment very few colonies formed. 24h treatment had almost no effect on T47D cell survival

rate and 48h treatment reduced the number of colonies only by 30%. Thus, low level of Daxx

correlates with increased resistance to paclitaxel treatment in these breast cancer cell lines.

To address the mechanism of cell death and differential survival of breast cancer cell lines

T47D and MDA MB 468 upon paclitaxel treatment, morphological changes that occurred with

nuclei were observed at different times of drug addition, which is effective means for

discriminating between apoptosis and micronucleation. Cells were categorized based on the

nuclear morphology (Figure 4-2). In mock-treated conditions (control, Figure 4-3) the majority

of cells were in interphase. The rate of accumulation in mitosis at 12h is similar for both cell

lines (Figure 4-3) that reflects almost an identical time of cell cycle progression. Already at 12h

of treatment, 27% of micronucleated cells appear in MDA MB 468 cells and reaches 60% and

79% at 24h and 36h correspondingly (Figure 4-3, top graph), while only small portion of cells

(20 and 8%) remains blocked in mitosis (Figure 4-3, middle graph). An insignificant number of

T47D cells are micronucleated at 12h and 24h, reaching only 22% by 36h of treatment; most

cells remain blocked in pro-metaphase during the course of treatment. The level of apoptosis in

both cell lines was negligible, reaching a maximum of 2% and 7% correspondingly in T47D and









5-6 Cyclin B levels are stabilized in Daxx- and RassflA-depleted cells treated with
tax ol ........... ................................................. .................... ........ ...... 70

5-7 Inactivation of the mitotic spindle checkpoint using Aurora kinase inhibitors
abolishes taxol resistance in Daxx-and RassflA-depleted cells ..................................71

6-1 Dynamics of Daxx-Rassfl interaction throughout the cell cycle .................................77

6-2 Model depicting Daxx-Rassfl-mediated mitotic stress checkpoint during pro-
m e ta p h a se .......................................................................... .. 7 8















-..Petaas
Prometaphase


Prometaphase
Block


sensitive cells
______ cell
death

Milotic Catastrophe:
Micronuclei Fornation
paclitaxel
decay!/ithdrawat


resistant cells


Prometaphase
Block (prolongated)


completion
- of mitosis,
cell division


inhibition of microtubule dynamics
by paclitaxel


recovery of microtubule dynamics


Figure 1-1. Dynamics of Paclitaxel Action in Cells. At pharmacological concentrations,
paclitaxel reversibly inhibits microtubule dynamics blocking cells in pro-metaphase.
Cells that are sensitive to paclitaxel activate mitotic block only transiently, followed
by micronucleation (mitotic catastrophe) and cell death, while resistant cells have a
more prolonged pro-metaphase block and continue proliferation after drug decay/
withdrawal and microtubule dynamics restoration-thus surviving chemotherapy.


r


GI Phase
At ND10PML bodies
-mteractiHn with PML F1


t


S-Phase
At lsn lrma
-inkraction with ATR
(Activew)


tin



G2 Phase
AtN-mVMMrWs
-ina~eracti with PMLSM
(Inative)


I


Mitosis
Disaaemby ofNDIO
-deSUMOlatin of PML
Function ?
Figure 1-2. Localization of Daxx Throughout the Cell Cycle. During Gl and G2 phase, Daxx is
localized to PML bodies where it is presumably inactive. During S phase, Daxx
relocates from PML to condensed chromatin where it interacts with ATRX. The fate
and function of Daxx during mitosis (M phase) is not documented.


Prophase














Daxx Truncation Mutant



41R !
p.....


s Rassil Tnmcation Mutauts
DwK WT (1-740B
Dwu 1-142
Dan 1-2-900
D1K 1-407 -
Dax 1-625
Dax t625-740


RassflA 22O34M
RtMlA 194-3411
RNtiC WFT (I-27O)
LrnC 5-50


(i)
iapuls. 10'i
oiMr GOSTC 2S
RasdlCDau


6XMis Puldown
fo Rassfl C
GST- GST 65hk
Danx RaadIC
I----I IalI


Inputs. 2%
SHin- GST- OST- 0W1-
Ra*iC -a.x DOa= Da= OST
1-142 1-290 1-407


6X-His Pulldown for RafiC
S!T GST OST F 6 -r
Doan DOa Da GST RasfiC
1-142 1-20 1407


(iii)

U OS1- WST- OST- GST-
ns- RaSntARsiA Raln A DMco GST
RaniC if-3M IN-3 1-i14


SX-FIn Pulidown foE Ra rIC
GST- GiT- 6ST- Q$T- C,,p
RWalA RMaUA RaelA Da ST 0 R BlI
12H-340 lu3a 1-14W


Rlfl Dx DID Dn DR= DLa Dun
F!.':Pdhlmi 1-40 1-142 1-290 -40-n 1425 62 -740




RuaflA
120-340 -1ND -dND -ND -dND -ND JND
RamIC
1-270 *i.'* *.'t* I*/ *'. 'ND -lND

I 50 +'-* *4**. **/. *-JND .ID

Figure 5-2. Mapping of Daxx-Rassfl Interaction. A) Diagram of human Daxx, RassflA and
RassflC constructs used for mapping interaction by co-localization (GFP and RFP)
and in vitro pull down assay (GST and 6His). B) (i) 6xHis pull down assay of GST-
Daxx wt and 6xHis-RassflC wt. Immobilized 6xHis-RassflC wt was incubated with
either GST or GST-Daxx wt. GST-Daxx wt but not GST alone binds 6xHis-RassflC
wt. (ii) In similar experimental settings, all Daxx amino terminal constructs including
aa 1-142 retain capacity to bind 6His-RassflC wt. (iii) 6His pull down assay of GST-
RassflA wt and mutants using immobilized 6His-RassflC wt. GST-RassflA wt/
mutants, GST-Daxx 1-142 and GST alone were incubated with 6His-RassflC wt.
GST-Daxx 1-142, GST-RassflA wt and mutants 120-340 and 194-340 bind to
RassflC, while GST does not. C) Table summarizing interaction (+, -, or ND for not
determined) tested by co-localization of GFP and RFP fusions or in vitro pull down
assay (right in cell).









depleted cells to taxol, a known microtubule-interfering compound. Interaction mapping

between Daxx, RassflC and RassflA and deletion mutants was analyzed with proteins in vitro

using a pull-down assay and by co-localization of GFP and RFP fusion proteins in cells (Figure

5-2). Interaction between Daxx and Rassfl was tested by incubation of immobilized 6His-fusion

RassflC with GST-Daxx and GST alone (Figure 5-2). RassflC associates only with GST-Daxx

and does not associate with GST control that indicates specificity of RassflC-Daxx interaction.

All truncation mutants of Daxx, including first 142 amino acids of protein, were co-purified with

RassflC wt suggesting that region of Rassfl interaction is located at the amino terminus of the

Daxx protein, thus confirming yeast two-hybrid data (Figure 5-2). Co-localization was observed

between RassflC-RFP and GFP-fused Daxx wt and Daxx amino terminal deletion mutants but

not with a carboxyl terminus mutant of Daxx (amino acids 625-740), further pointing at the

amino terminus of Daxx as the minimal Rassfl interacting domain.

GFP-Daxx does not co-localize with RFP-RassflA wt or with a RassflA mutant that covers

aa 120-340 which is homologous between RassflA and RassflC, but GFP-Daxx co-localizes

with the first 50 amino acids of RFP-RassflC that are unique for this splice variant. Moreover,

the first 142 amino acids of Daxx can also co-localize with this domain of RassflC. In

conclusion, the interaction between Rassfl and Daxx is facilitated by the first 50 unique amino

acids of RassflC; in combination with a minimal clone purified in a yeast two-hybrid analysis

(which encodes amino acids 30 to 270 of RassflC) the potential minimal region of interaction

can likely be narrowed down to amino acids 30-50 of RassflC. Therefore, Daxx interacts with

RassflC, but not RassflA.

Considering these data, it was hypothesized that Daxx may cooperate with RassflA

functioning in mitosis via an interaction with RassflC. This model assumes that two alternative









CHAPTER 4
DAXX IS A TRIGGER OF CELLULAR TAXOL RESPONSE

Introduction

The mitotic spindle checkpoint is a very fluid and dynamic apparatus designed to monitor

microtubule-kinetochore attachment and proper microtubule tension. In the event of an error

occurring during pro-metaphase and metaphase, the spindle checkpoint will elicit a "wait" signal

that stalls chromosomal segregation and anaphase onset until errors are corrected. Microtubule-

inhibiting drugs such as taxol and nocodazole (among others) can initiate a prolonged "wait"

signal that if uncorrected, initiates micronucleation (mitotic catastrophe) and cell death (Jiang et

al., 2006; Mansilla et al., 2006a; Mansilla et al., 2006b; Ricci and Zong, 2006). Thus, sub-

populations of cancer cells with deranged regulation of mitotic spindle checkpoint proteins and

other associated factors can respond differently to these compounds offering a therapeutic

molecular target for treating cancers. However, in most instances, inactivation or down-

regulation of mitotic checkpoint proteins leads to increased sensitivity to microtubule inhibiting

drugs. In cases of BubR1 or Mad2 depletion, cells treated with taxol respond in a much stronger

and robust way: the number of cells arrested in mitosis dramatically decreased and cell death

was more rapid (Sudo et al., 2004). Cells lacking Madl, an important factor involved in the

assembly of the spindle checkpoint, also exhibit increased sensitivity to taxol, however cells

treated with nocodazole became more resistant (Kienitz et al., 2005). Thus, identification of

novel regulators of taxol sensitivity-which may alter cellular response to these drugs by

increasing resistance-is required.









hDaxx (sense GATGAAGCAGCTGCTCAGC; anti sense GCTCAGCAGCTGCTTCATC);

control siRNA was directed against base pairs 1262-1284 of SETDB1 (sense

TCCTCTTTCTTATCCTCGTATGT, anti-sense ACATACGAGGATAAGAAAGAGGA).



Western Blotting

Cell and protein extracts were ran on pre-made 4-20% Tris-HC1, SDS/Acrylamide gels

(Biorad), separated by electrophoresis and transferred onto nitrocellulose and PVDF membranes

(Biorad). Membranes were then blocked using 3% milk in 0.1% PBS-Tween for 30 minutes at

room temperature. Primary antibodies were added and incubated overnight at 4C. The next

day, membranes were washed 3X with 0.1% PBS-Tween and secondary antibodies (mouse and

rabbit HRP conjugates, Cell Signaling Technologies) were added and membranes were incubated

at room temperature for 1 hr. After 3X wash in PBS-Tween, membranes were then exposed with

ECL (Amersham) and developed.



Yeast Two-Hybrid Assay

The bait vector pGBDC1-mDaxx wt, -mDaxx C term and -mDaxx AC were transformed

into yeast strain PJ-69a and PJ-69 alpha along with mouse E11.5 cDNA library plasmid cloned

into pGADC1. Briefly, frozen yeast competent cells were vortexed until melted completely and

then incubated at room temperature for 45 min in 1 mL solution B (40% polyethylene glycol

1000, 200 mM bicine pH 8.35). After pelleting cells, supernatant was removed and cells were

resuspended in 1 mL solution C (150 mM NaC1, 10 mM bicene pH 8.35) and repeated with a

second wash in solution C. Cells were then resuspended in 100 pL solution C and plated on

selective media (CM media tryptophan,- leucine and CM media tryptophan,- leucine,-

histidine).










CHAPTER 2
MATERIALS AND METHODS

Antibodies

Antibodies to Daxx (M112), BubR1 (8G1), Bubl (14H5), Cdc20 (H-175), Cdc27 (AF3.1),

Cyclin B (GNS1) and Mad2 (17D10) were from Santa Cruz Biotechnology; GST monoclonal

antibody was from Sigma; 6X His monoclonal antibody was from Invitrogen; RassflC mouse

polyclonal antibody was from UT Southwestern; RassflA monoclonal antibody was from

Abcam; RassflA/C polyclonal antibody was a gift from Dae Sik Lim, Korea Advanced Institute

of Technology; PML rabbit polyclonal antibody was from Gerd Maul, Wistar Institute; Phospho

Histone H3 (SerlO) rabbit polyclonal antibody was from Upstate; a-Tubulin monoclonal

antibody was from Sigma; Daxx monoclonal antibody 514 and rabbit polyclonal antibody

2133/2134 were developed as described in (Ishov et al., 2004).



P-Gal Reporter Assay

Starter cultures of yeast were grown overnight in CM selective media and OD600 readings

were measured the next day. Cells were resuspended in breaking buffer (100 mM Tris-HCl pH

8.0, 1 mM DTT, 10% glycerol, 40 mM PMSF) and glass beads were added and suspensions were

vortexed for approximately 5 min at 40C. After pelleting cell debris, a 1:10 mixture of

superatant/Z buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KC1, 1 mM MgSO4) was

made and incubated for 5 min at 280C. 200 [tL ONPG (Sigma) was added to the mixtures and

the time taken for reactions to turn yellow was recorded. Afterwards, 400 [iL of Na2CO3 was

added to stop reactions. P-Gal activity was measured at OD420 by making 1:10 mixtures of

ONPG reaction/water and recording spectrophotometer readings. Protein concentration was

determined by OD595 spectrophotemeter readings afterterwards.

31









LIST OF FIGURES


Figure Page

1-1 Dynamics of paclitaxel action in cells.. ............................. ...... ............... 30

1-2 Localization of Daxx throughout the cell cycle ................................................30

3-1 Characterization of Daxx- mouse embryos and cells ............................................... 42

3-2 Western blot analysis of Daxx protein level in HEp2-H3-GFP cells expressing
control-siRNA or Daxx-siRNA. ................................. ....................................... 43

3-3 Fluorescence time-lapse microscopy images of HEp2-H3-GFP cells expressing
either control-siRNA or Daxx-siRNA.. ....................................................... ....................43

3-4 Daxx depletion stabilizes cyclin B during mitosis................................. ............... 44

3-5 Western blot analysis of mitotic proteins in wild type (parental), control-siRNA and
tw o independent D axx-siRN A cell lines....................................... ......................... 44

3-6 Western blot analysis of Daxx protein levels throughout the cell cycle. ..........................45

3-7 Dynamics of Daxx localization in mitotic MPEF cells. ..................................................45

4-1 Differential response of Daxx/+ and Daxx-/- MEFs to microtubule inhibitors
nocodazole and paclitaxel ........................................... ....................51

4-2 Colony formation of breast cancer cells after paclitaxel treatment is Daxx-dependent. ...52

4-3 Response to paclitaxel treatment in breast cancer cell lines with different Daxx level.....53

4-4 Response of MDA MB 468 and T47D breast cancer cell lines to increased
concentration of paclitaxel........................................................... .. ............... 54

4-5 FACS analysis of cell cycle distribution after paclitaxel treatment..............................55

4-6 Paclitaxel response is Daxx-dependent..................... ...... ........................... 56

5-1 Daxx interacts with tumor suppressor Rassfl in yeast. ...........................................64

5-2 M apping of D axx-R assfl interaction..................................................................... ...... 66

5-3 Cellular distribution of Daxx and Rassfl during interphase..............................67

5-4 Co-localization of endogenous Daxx and Rassfl during mitosis in HEp2 cells ...............68

5-5 Depletion of Daxx or RassflA desensitizes cells to paclitaxel..............................69









2.3 min in control siRNA) than control cells (Figure 3-3). Overall, condensation of

chromosomes in Daxx-depleted cells occurred more rapidly and the onset of pro-metaphase

occurred sooner. The transition from nuclear envelope break down to chromosome segregation

in anaphase was also 20.5% slower in Daxx-depleted cells (37.6 min +/- 10.36 min for Daxx

siRNA, 31.2 min +/- 7.9 min for control siRNA) (Figure 3-3). Hence, absence of Daxx

significantly influences the rate of early mitosis progression in HEp2 cells which emphasizes

initial observations of increased pro-metaphase index in Daxx knockout embryos.

To confirm that Daxx protein has a direct effect on cell cycle progression, HEp2 cells stably

expressing control or Daxx siRNA were synchronized using a double thymidine block and then

released to allow cells to progress through mitosis synchronously so mitotic cyclin B levels could

be analyzed. Defective mitotic progression was seen in Daxx-depleted cells as cyclin B levels

were consistently sustained longer than wild-type (parental) or control siRNA cells (Figure 3-4).

Specifically, during the time point of 9-9.5 hr release from thymidine block in wild type and

control siRNA cells the majority of Cyclin B levels are destroyed, however, Cyclin B stability is

preserved in Daxx siRNA cells past this time point until 10-10.5 hrs after thymidine release.

These data implicate Daxx in the regulation of mitotic progression, either directly as a regulatory

mitotic protein or indirectly through other mechanisms.

To rule out possible indirect mechanisms of Daxx regulation of mitosis, protein expression of

several known mitotic proteins, including Mad2, a key mitotic checkpoint protein; Cdc20 the

activator of the anaphase promoting complex (APC)/cyclosome and Cdc27, a major subunit of

the APC was analyzed and compared in control- and two independent Daxx-depleted HEp2 cell

lines (Fig 3-5). By comparison, no significant changes in protein expression were observed in

the presence or absence of Daxx, which supports microarray analysis from Daxx++ and Daxx-/-









MDA MB 468 after 36h of treatment (Figure 4-3, bottom graph); thus, apoptosis is not the main

mechanism of cell death for these cell lines at this drug concentration. Indeed, the process of

apoptosis may occur as a secondary event or be more prominent (and is sometimes observed) in

larger paclitaxel concentrations-but this may be outside of any clinical relevance (Hernandez-

Vargas et al., 2006; Wang et al., 1999). Furthermore, upon exposure of MDA MB 468 and

T47D cells to increased paclitaxel concentrations (100 nM and greater) sizeable increase in

apoptotic levels or change in mitotic index/micronucleation could not be detected (Figure 4-4).

This is in contrast to some reports showing that larger paclitaxel concentrations induce a more

prominent mitotic checkpoint arrest and hence a stronger mitotic block in other cell lines

(Giannakakou et al., 2001; Ikui et al., 2005). Similar results were observed by FACS analysis

(Figure 4-5): the majority of T47D cells accumulate in G2/M at 24h and 36h of treatment, while

MDA MB 468 produce an extensive sub-G1 population. Unfortunately, FACS does not allow

discrimination between apoptotic and micronucleated cells as clearly as microscopic analysis

because both types of cell death result in fragmentation of the nucleus (recognized as a sub-G1

population by FACS and thus counted together). Therefore, microscopy was the most useful

approach to determine type of cell death after paclitaxel treatment. Thus, a major difference in

paclitaxel response between these cell lines is the ability to maintain prometaphase block-

which is extended in T47D, but is brief in MDA MB 468 and is followed by micronucleation.

Daxx was recently shown to interact with and regulate stability of p53, one of the key players

in cell growth arrest and apoptosis (Tang et al., 2006; Zhao et al., 2004). Moreover, Daxx seems

to differentially regulate p53 dependent transcription under DNA damage conditions in transient

transfection assay, thus affecting the balance between cell cycle arrest and apoptosis (Gostissa et

al., 2004). Stability of p53 and p53-dependent transcription regulation, however, were unaffected









efficiently exit mitosis (and eventually die) when encountered with specific stress stimuli during

mitosis, including Taxol. In the absence of Daxx or Rassfl, cells treated with Taxol remain

arrested in mitosis due to a sustained mitotic spindle checkpoint. Upon Taxol decay or removal,

these cells can resume mitosis and complete cell division-thus being Taxol resistant. Inhibition

of the spindle checkpoint using Aurora Kinase inhibitors efficiently abolished Taxol resistance in

Daxx and Rassfl depleted cells. Deregulation of most known mitotic proteins leads to enhanced

Taxol response. Absence or depletion ofDaxx and Rassfl, in contrast, results in increased drug

resistance. In the future, Daxx and Rassfl may be useful predictive markers for the proper

selection of patients for taxane chemotherapy.









Zhong, S., Salomoni, P., Ronchetti, S., Guo, A., Ruggero, D., and Pandolfi, P.P. (2000).
Promyelocytic leukemia protein (PML) and Daxx participate in a novel nuclear pathway for
apoptosis. J Exp Med 191, 631-640.









serves as a trigger for cellular taxol response in combination with Rassfl which adds to our

understanding of mechanisms linking cell division, genome instability and breast cancer

progression.



Table 6-1. Alteration of Several Known Mitotic Proteins and Resultant Cellular Paclitaxel
Response. Most mitotic proteins, when mutated or down-regulated in cells, display
increased sensitivity to paclitaxel. In the absence ofDaxx or Rassfl, cells display
increased drug resistance.
Protein Paclitaxel Response
Bub 1 Increased Sensitivity
(Lee et al., 2004; Sudo et al., 2004)
BubR1 Increased Sensitivity
(Lee et al., 2004; Sudo et al., 2004)
CHFR Increased Sensitivity
(Satoh et al., 2003)
Madl Increased Sensitivity
(Kienitz et al., 2005)
Mad2 Increased Sensitivity
(Niikura et al., 2007)
Survivin Increased Sensitivity
(Carvalho et al., 2003)
Daxx Increased Resistance
(Lindsay et al., 2007)
Rassfl Increased Resistance




























To my parents Archie and Virginia Lindsay

















B







C







Figure 5-4. Co-localization of Endogenous Daxx and Rassfl During Mitosis in HEp2 Cells.
HEp2 cells were immunostained with monoclonal Daxx 514 antibody and polyclonal
Rassfl antibody which detects both endogenous RassflA and RassflC. Rassfl and
Daxx co-localized during pro-metaphase and metaphase A) & B). C) By anaphase
and later stages of mitosis, this association is absent.









Cell Cycle Synchronization

HEp2 and MDA MB 468 cells were synchronized at G1/S using a standard double

thymidine block protocol. Briefly, asynchronous cells were set up at 20-30% confluency and

treated with 2 mM thymidine for 18-20 hrs, washed 2X with PBS and released into normal

media without thymidine for 6 hrs. 2 mM thymidine was then added for an additional 18-20 hrs

and cells were washed 2X with PBS and released from G1/S block for experiments.



Colony Formation Assay

Treated and non-treated cells were set up for colony formation using a series of cell

dilutions ranging from 1:100 to 1:1000 to optimize colony growth. On average, cells were

grown for 5-7 days after set up. Colonies were then washed, fixed with methanol and stained

with crystal violet (Fisher) in order to visibly count colonies. Each colony formation assay was

set up in triplicate in order to obtain representable statistics for each drug treatment and its

duration of exposure to cells.



Confocal Microscopy and Subcellular Localization

HEp2 cells were cultured on cover dishes for 24 hrs and then transfected with GFP-Daxx,

GFP-RassflA, or GFP-RassflC. 24-48 hrs later, live cell confocal images were taken using a

Leica TCS SP5 microscope and chamber and Leica imaging software.



Drug Treatment

Standard concentration of paclitaxel (Calbiochem) used throughout these studies is 10 nM,

unless noted otherwise. Paclitaxel was used to treat both asynchronously and synchronously

growing cells. Standard concentration of nocodazole (Calbiochem) was 10 pM and was used to











Biochemical Fractionation

HEp2 cells were separated into nuclear and cytosolic fractions using a biochemical

fractionation buffer consisting of 250 mM sucrose, 20 mM HEPES-KOH pH 7.4, 10 mM KC1,

1.5 mM MgC12, 1 mM EDTA, 1 mM EGTA. Cells were grown on 100 mm dishes and washed

2X with PBS and then placed on ice. Ice cold fractionation buffer was immediately added to

dishes and cells were incubated at 4C with gentle rocking for 15 minutes. Cells were then

collected into tubes and dounced five times with a B-type pestle. Afterwards, cell extract was

collected into microcentrifuge tubes and centrifuged for 2-3 min at 800 RPM to separate intact

nuclei from soluble cytoplasm fraction. Nuclei were washed 2X with fractionation buffer and

resuspended in 1 mL fractionation buffer. 5 M NaCl was then added to the buffer to reach a final

concentration of 450 mM NaC1. Nuclei were then placed at 4C and subjected to gentle

inverting for 5-10 minutes. Nuclear extract was then centrifuged at 13,000 RPM for 5 minutes at

4C to pellet insoluble nuclear material.



Cell Culture and Transfections

Daxx +/+ MEFs, Daxx -/- MEFs, HEp2, MDA MB 468 and T47D cells were maintained in

Dulbecco's Modified Eagle's Medium with 10% fetal bovine serum and 1% penicillin/

streptomycin and maintained in 5% CO2 at 37C. For transient transfection in mammalian cells,

all cDNAs were transfected using the Lipofectamine protocol (Invitrogen) according to the

manufacturer's instructions.









remaining in pro-metaphase by definition. In contrast, with normal primary cells and some

tumor cell lines, Daxx and RassflA may be essential for proper mitotic exit in response to

uncorrectable errors during pro-metaphase. This is suggested by the very robust response of

wild type cells which only transiently arrest in pro-metaphase and then undergo micronucleation

leading to cell death (Figure 4-3, Figure 4-6). One example of a mitotic stress checkpoint is

already known with a protein named checkpoint with FHA and ring finger (CHFR). In the

presence of CHFR, wild type cells exhibit a transient prophase arrest which temporarily prevents

cells from entering metaphase under mitotic stress (Scolnick and Halazonetis, 2000). Cells that

do not have a functional CHFR protein were shown to enter metaphase without delay and

exhibited problems in proper centrosome separation. Thus, CHFR defines a prophase-specific

mitotic stress checkpoint at a stage in mitosis earlier than Daxx and Rassfl.

Evidence accumulated in these studies suggests Daxx and Rassfl are triggers for cellular

taxol response. In the future, Daxx and Rassfl may serve as ideal molecular markers for the

proper selection of breast cancer patients (and other malignancies) for taxane chemotherapy. In

order to achieve this goal, clinical studies will be required examining the status of Daxx and

Rassfl expression in tumors before and after taxane chemotherapy as well as in patients with an

established history of taxane resistance. Altered Daxx or Rassfl expression may be reminiscent

of the differential protein expression that may exist in the original tumor cells from which each

cell line was derived. It is already known that Daxx expression in some breast cancer cell lines

is quite variable (Figure 4-1), but the extent of Daxx down-regulation or mutation in tumor cell

lines has not been addressed. RassflA expression in tumor cell lines, conversely, has been

extensively studied and shown to be altered in a majority of cases (Agathanggelou et al., 2005).

Ultimately, these studies have established new roles for Daxx as a mitotic regulator that also









Hari, M., Yang, H., Zeng, C., Canizales, M., and Cabral, F. (2003b). Expression of class III beta-
tubulin reduces microtubule assembly and confers resistance to paclitaxel. Cell Motil
Cytoskeleton 56, 45-56.

Haruki, N., Saito, H., Harano, T., Nomoto, S., Takahashi, T., Osada, H., and Fujii, Y. (2001).
Molecular analysis of the mitotic checkpoint genes BUB1, BUBR1 and BUB3 in human lung
cancers. Cancer Lett 162, 201-205.

Henderson, I.C., Berry, D.A., Demetri, G.D., Cirrincione, C.T., Goldstein, L.J., Martino, S.,
Ingle, J.N., Cooper, M.R., Hayes, D.F., Tkaczuk, K.H., et al. (2003). Improved outcomes from
adding sequential Paclitaxel but not from escalating Doxorubicin dose in an adjuvant
chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol 21,
976-983.

Hernandez-Vargas, H., Palacios, J., and Moreno-Bueno, G. (2006). Molecular profiling of
docetaxel cytotoxicity in breast cancer cells: uncoupling of aberrant mitosis and apoptosis.
Oncogene.

Hollenbach, A.D., McPherson, C.J., Mientjes, E.J., Iyengar, R., and Grosveld, G. (2002). Daxx
and histone deacetylase II associate with chromatin through an interaction with core histones and
the chromatin-associated protein Dek. J Cell Sci 115, 3319-3330.

Ikui, A.E., Yang, C.P., Matsumoto, T., and Horwitz, S.B. (2005). Low concentrations oftaxol
cause mitotic delay followed by premature dissociation ofp55CDC from Mad2 and BubR1 and
abrogation of the spindle checkpoint, leading to aneuploidy. Cell Cycle 4, 1385-1388.

Ishov, A.M., Sotnikov, A.G., Negorev, D., Vladimirova, O.V., Neff, N., Kamitani, T., Yeh, E.T.,
Strauss, J.F., 3rd, and Maul, G.G. (1999). PML is critical for ND10 formation and recruits the
PML-interacting protein daxx to this nuclear structure when modified by SUMO-1. J Cell Biol
147, 221-234.

Ishov, A.M., Vladimirova, O.V., and Maul, G.G. (2004). Heterochromatin and ND10 are cell-
cycle regulated and phosphorylation-dependent alternate nuclear sites of the transcription
repressor Daxx and SWI/SNF protein ATRX. J Cell Sci 117, 3807-3820.

Iwao-Koizumi, K., Matoba, R., Ueno, N., Kim, S.J., Ando, A., Miyoshi, Y., Maeda, E., Noguchi,
S., and Kato, K. (2005). Prediction of docetaxel response in human breast cancer by gene
expression profiling. J Clin Oncol 23, 422-431.

Jiang, N., Wang, X., Yang, Y., and Dai, W. (2006). Advances in mitotic inhibitors for cancer
treatment. Mini Rev Med Chem 6, 885-895.

Jung, Y.S., Kim, H.Y., Kim, J., Lee, M.G., Pouyssegur, J., and Kim, E. (2008). Physical
interactions and functional coupling between Daxx and sodium hydrogen exchanger 1 in
ischemic cell death. J Biol Chem 283, 1018-1025.

Jung, Y.S., Kim, H.Y., Lee, Y.J., and Kim, E. (2007). Subcellular localization of Daxx
determines its opposing functions in ischemic cell death. FEBS Lett 581, 843-852.









a long time, apoptosis was considered the main mechanism of cell death in response to taxane

treatment. Currently, a more distinct model of therapy response is considered, wherein different

modes of tumor cell death are likely determined by drug concentration and genetic background

of the cells within a tumor (Morse et al., 2005). At pharmacological concentrations, taxanes

reversibly bind to tubulin heterodimers that form microtubules-this accelerates polymerization

and inhibits depolymerization of tubulin, thus disrupting microtubule dynamics. This event, in

turn, activates the spindle checkpoint, which invokes mitotic arrest that in normal, untreated

conditions ensures proper chromosomal attachment and alignment and ensures faithful

chromosomal segregation preventing aneuploidy (Chan and Yen, 2003; Cleveland et al., 2003).

This mitotic arrest does not persist indefinitely. After some period of time, cells usually undergo

an aberrant exit from mitosis, characterized by the lack of metaphase, anaphase and cytokinesis.

The nuclear envelope is reformed around individual or groups of chromosomes producing large

nonviable cells with multiple micronuclei, which are easily distinguishable morphologically

from apoptotic cells. Apoptotic cells will have small, highly condensed chromatin with

fragmented nuclei and a diminished cytoplasm, whereas micronucleated cells are much larger

with uncondensed chromatin in a pattern reminiscent of normal nuclei. This type of cell death

which results in micronucleated cells is known as mitotic catastrophe and is activated during

mitosis as a result of deranged spindle formation coupled with blocks of different checkpoint

mechanisms that activate aberrant chromosome segregation and nuclear fragmentation (Kroemer

and Martin, 2005). Taxane-sensitive human breast cancer cells are blocked in mitosis only

transiently, followed by nuclear fragmentation and mitotic catastrophe, while resistant breast

cancer cells have more prolonged mitotic block and continue proliferation after drug decay and

microtubule dynamics restoration, thus surviving chemotherapy (Figure 1-1).









PML++ compared to PML- cells (Zhong et al., 2000). Thus, the localization of Daxx to ND10

and not to the cytoplasm was critical for Daxx-enhanced apoptosis.



Daxx and Transcription

The identification of PML interacting with and sequestering Daxx into nuclear domains

would become as important a discovery as any study demonstrating the functionality of Daxx.

Ishov and colleagues showed that in situations where PML was absent, Daxx would be relocated

to condensed heterochromatin where it could potentially be involved in some biochemical

function (Ishov et al., 2004). Subsequent studies provided some evidence of what Daxx could be

doing at these sites by showing it could interact not only with core histones, but histone

deacetylase II (HDACII), Dek (Hollenbach et al., 2002) and the SWI/SNF chromatin remodeling

protein ATRX (Xue et al., 2003). These interactions, among others, brought forth the idea that

Daxx could be acting as a regulator of transcription, not only on the level of repression but

activation as well. Among a few of the many genes Daxx has been implicated in regulating

include p53 target genes (Gostissa et al., 2004; Zhao et al., 2004), Pax transcription factor family

members (Emelyanov et al., 2002; Hollenbach et al., 2002; Lehembre et al., 2001) and Smad4

(Chang et al., 2005a)

A more dynamic role of Daxx became appreciated when Ishov and colleagues showed a

cell cycle dependent localization of Daxx between ND 10 and heterochromatin (Figure 1-2)

(Ishov et al., 2004). They found that during G1 and G2 phase, Daxx could be found in its

characteristic location at ND10, while during S phase, Daxx would relocate to condensed

heterochromatin. Interestingly during mitosis, ND10 is disassembled, PML de-sumoylated and

Daxx no longer associated with the remnants of these nuclear domains (Ishov et al., 2004). What

became the fate of Daxx after this set of events was not addressed.

19









CHAPTER 6
SUMMARY AND CONCLUSIONS

Taxane chemotherapy is considered among the most active treatment options for many breast

cancer patients, either alone or as adjuvant in combination with anthracyclins (O'Shaughnessy,

2005). Nevertheless, a large number of patients are resistant to taxanes or become resistant to

this therapy during treatment. The response rate of docetaxel is -50% even after the first-line

chemotherapy administration and it decreases to 20-30% in the second- or third-line

administration (Bonneterre et al., 1999; Crown et al., 2004). A number of studies have been

carried out to determine a genomic profile that could be predictive to taxane treatment (Chang et

al., 2003; Chang et al., 2005b; Iwao-Koizumi et al., 2005; Mauriac et al., 2005; Miyoshi et al.,

2004). However, an alternative approach to understand selective resistance to taxane treatment

is to study mechanisms by which cells can respond to these drugs. Several molecular targets

were reported, starting with mutations in a- and P-tubulin that affect drug binding, increased

expression of tubulin genes, and changes in the synthesis or activity of tubulin interacting

proteins (Hari et al., 2003a; Hari et al., 2003b; Wang and Cabral, 2005). Recently, a new class

of potential targets are being studied after it was suggested that inactivation of mitotic proteins

can contribute to the selective response of taxane treatment in vivo (Wassmann and Benezra,

2001). Thus, development of new genomic prognosis factors and in-depth understanding of drug

activity on both a cellular and organismal level are needed for optimization of adjuvant therapy

and proper patient stratification.

To this end, Daxx was identified as a novel regulator of paclitaxel response in cell culture

conditions. Primary mouse and human fibroblasts with experimentally regulated levels of Daxx

show both strong and divergent responses to taxol (Figure 4-1). In the absence of Daxx, cells

remain in a prolonged mitotic block, while wild type cells undergo a transient arrest in mitosis









TABLE OF CONTENTS

A CK N O W LED G M EN T S .............................................................................. ................... 4

L IS T O F T A B L E S ...................................... ..................................... ................ .. 7

LIST OF FIGURES .................................. .. ..... ..... ................. .8

1 INTRODUCTION ............... ............................ .............................. 12

Breast Cancer and Chemotherapy Resistance ............................................. ............... 12
T axanes and their A activity .............................................................................. ........ 12
Taxanes and M itotic Checkpoints ..............................................................................14
Cancer Cell Line Response to Taxane Treatment ..........................................................15
Predictive M arkers for Taxane Treatment.................................................................... 15
Daxx: The Story of an Enigmatic Protein.................................................... ................17
D axx and A poptosis .................................. .. .......... .. ............17
D axx and Transcription ............................................................ .................... 19
Cellular Localization of Daxx .......................................................... ............... 21
Ras-Association Domain Family-1 (Rassfl) and Cancer.....................................................24
Preferential Alteration of RassflA in Cancer.....................................................25
RassflA and Cell Cycle Control .................................. .....................................25
Cellular Localization of R assflA ................... ....... .............................................. 27
Functions of R assflC .............................................. ... .... ........ ......... 28
Cellular Localization of R assflC ............................ ......................... ............... 29

2 M A TER IA L S A N D M ETH O D S ........................................ .............................................31

A n tib o d ie s .................................................................................................................. 3 1
P-G al R reporter A ssay ................................................. ........ .. ........ .... 31
B iochem ical F ractionation ...................................................................... ...................32
C ell Culture and Transfections......... .................. .................................. ............... 32
C ell C ycle Synchronization ......... ................. ................. ................... ............... 33
Colony Form action A ssay ......... ........................................................ .................... 33
Confocal Microscopy and Subcellular Localization ................................................. 33
D ru g T re atm e n t ...............................................................................................................3 3
E m bryo Isolation and C culture .............................................................. .....................34
F A C S A naly sis .......................................................................... 34
Fluorescence Tim e-Lapse M icroscopy....................................... ......................... 34
Im m unofluorescence ............................................ .. .. ........... ......... 35
In vitro Pull-dow n A ssay ........................................................... .. ............... 35
Plasm id Constructs ................................... ..... .. ...... ............... 36
Stable siR N A Infection ........... .............................................................. .................. 36
W western B lotting................................................... 37
Y east T w o-H ybrid A ssay ........................................................................ .................. 37

3 D A X X FU N C TIO N IN M ITO SIS .............................................................. .....................38









preventing genome instability (Dallol et al., 2004; Vos et al., 2004). Liu and colleagues also

described that overexpression of RassflA induced aberrant mitotic arrest at metaphase in a

similar manner to the microtubule stabilizing drug taxol and how it affects cells (Liu et al.,

2003). This became the first study to link RassflA to possible mitotic functions in cells.

Additional insight into the role of RassflA function in mitosis and cell cycle progression

came in the seminal study by (Song et al., 2004). RassflA was shown to influence the stability

of both Cyclin A and Cyclin B as a result of direct interaction with Cdc20 and negative

regulation of the anaphase promoting complex (APC). Without RassflA, cells were proven to

progress through early mitosis (specifically pro-metaphase) faster as a result of premature

activation of APC. Absence of RassflA by siRNA depletion also caused centrosome

abnormalities and multipolar spindles. Details of interaction of RassflA with Cdc20 would later

become a subject of controversy as Liu and colleagues showed that RassflA was not capable of

interacting with Cdc20 in vitro and immunoprecipitation of Cdc20 with RassflA could not be

detected in synchronous or asynchronous cells (Liu et al., 2007). In the future, additional studies

into the precise role of RassflA regulation of mitosis will be required.



Cellular Localization of RassflA

Not until the Pfeifer group described co-localization of RassflA with microtubules did the

scientific community know about the distribution of the protein in cells (Liu et al., 2003). While

this study showed ample evidence of the microtubule-associated network that over-expressed

RassflA will form in cells, the Vos group were the first to describe the endogenous interaction of

RassflA with polymerized tubulin (Vos et al., 2004). Depolymerized tubulin was unable to

interact with RassflA. The same group performed a yeast two-hybrid screen using RassflA as

"bait" and identified and confirmed several known microtubule-associated proteins as interacting

27











LIST OF TABLES

Table Page

3-1 Statistical analysis of mitotic progression in control- and Daxx-depleted H3-GFP-
H E p 2 c e lls ...................................... ..................................... ................ 4 2

6-1 Alteration of several known mitotic proteins and resultant cellular paclitaxel
respond se........ ................................ ................................................76









From the cell cycle-dependent localization of Daxx model which Ishov and colleagues

proposed, ND10 could be considered a site of inactivation of Daxx function-a potential

"storage depot" for Daxx and numerous other proteins until specific times when they are needed

and become active again (Ishov et al., 2004). Although at the time this was not a novel concept,

a study conducted by (Li et al., 2000) and similarly by (Lin et al., 2003) suggested this notion

showing that when Daxx was bound to increasing amounts of PML, the transcriptional

repression activity of Daxx-as measured by a luciferase reporter assay-was relieved. A

possible mechanism which could regulate the localization of Daxx to ND10 or to

heterochromatin was shown by Ecsedy and colleagues when they demonstrated a physical

interaction between Daxx and the serine/threonine kinase HIPK1 (Ecsedy et al., 2003). This

interaction was capable of displacing Daxx from PML and re-localizing it elsewhere in the

nucleus. In addition, the Ecesedy group found that upon phosphorylation of Daxx by HIPK1,

Daxx transcriptional repression activity was modified (Ecsedy et al., 2003). The investigators

could not, however, definitively show a relocalization of Daxx to heterochromatin but rather an

association with HDAC1. Additionally, they showed that phosphorylation of Daxx by HIPK1

diminished the transcriptionally repressive activity of Daxx rather than enhanced it. Other

studies, which focused on the condition-dependent localization of Daxx (and other nuclear body

associated proteins) were found to be dependent on the sumoylation status of PML as well as

cellular stresses such as heat shock and heavy metal exposure (Nefkens et al., 2003). The small

ubiquitin-like modifier (SUMO), moreover, is a post-translational modification added to proteins

which affects their function and localization. SUMO bears a 20% identity to ubiquitin and is

covalently linked to a wide range of proteins whose functions are commonly implicated in

chromatin organization, transcription and genomic stability (Hay, 2005). Although other









of nuclear proteins is observed in these cases, it is possible that these phenomena are less

attributable to Daxx function and more explainable as a general cellular stress response. Reports

such as (Nefkens et al., 2003) for example, may describe more functional stress-induced Daxx

redistribution in that the activity of several nuclear- and ND 10-associated proteins were

documented in parallel after response to specific stress stimuli. Specifically, Daxx and Spl00

but not PML would disperse from ND10 into the nucleoplasm due to rapid desumoylation of

PML during heat shock, while heavy metal exposure would release Daxx and PML with Spl00

retained at ND10 (Nefkens et al., 2003). While there is some tantalizing evidence to suggest

existence and function of Daxx in the cytoplasm, more extensive studies of endogenous protein

trafficking are required.



Ras-Association Domain Family-1 (Rassfl) and Cancer

The Rassfl gene locus comprises approximately 11,151 base pairs of the human genome

and consists of eight exons (Agathanggelou et al., 2005). It is found on the short arm of human

chromosome 3 (3p21.3). Differential promoter use and alternative splicing creates seven

transcripts RassflA-RassflG. Of these transcripts, isoforms RassflA and RassflC are

predominately expressed in all tissues while RassflB is expressed in the hemopoeitic system

only, and RassflD or RassflE expression is restricted to cardiac or pancreatic cells, respectively.

Homologues of RassflA exist in rodents, fish, nematodes and fruit flies ranging from 38% to

85% identity (Agathanggelou et al., 2005). A RassflC homologue exists in Xenopus with no

apparent homologue for RassflA.













100%


80%

60% -

40%1

20%

0%
Daxx siRNA RassflA siRNA Control siRNA

Taxol D Taxol + ZM i Taxol + AKIII

Figure 5-7. Inactivation of the Mitotic Spindle Checkpoint Using Aurora Kinase Inhibitors
Abolishes Taxol Resistance in Daxx-and RassflA-Depleted Cells. Percentage of
colonies that were formed from control, Daxx- and RassflA-depleted HEp2 cells
which were synchronized using a double thymidine block, released and exposed to
taxol alone, or taxol in combination with ZM447439 (ZM) or Aurora kinase inhibitor
III (AKIII) for six hours is shown. Note taxol resistance is abolished in Daxx- and
RassflA-depleted cells treated in combination with taxol and Aurora kinase
inhibitors.










Wild type,'" Mitotic Sbrs

cells ,, (ii Taxel

Daxx E ONE 0 0 IN
Rassfl EU EK

Uncorredable errors.
Mitosis aborted, micronulede
formahon

Cancer ,.
cellsMi,osis abrtedie Sb










Figure 6-2. Model Depicting Daxx-Rassfl-Mediated Mitotic Stress Checkpoint During Pro-
metaphase. In wild type cells upon mitotic stress (i.e. taxol or nocodazole) the
spindle checkpoint will be activated. In rare cases, the cell may correct these errors,
complete mitosis and resume cell proliferation. In most cases, however, these errors
are uncorrected and the cell aborts mitosis, undergoes micronuclei formation and cell
death. In subpopulations of cancer cells which encounter mitotic stress and which do
not have a functional Daxx or Rassfl protein, these cells remain in a prolonged
mitotic block, unable to efficiently abort mitosis, and remain in pro-metaphase until
drug removal or decay. This model may partially explain inherent and acquired taxol
resistance in some breast tumors.









Table 3-1. Statistical Analysis of Mitotic Progression in Control- and Daxx-Depleted H3-GFP-
HEp2 Cells. Peak time designates the most frequently occurring time of completion
for each stage of mitosis.
Peak Standard
Stage RNAi Average time deviation Min Max
Prophase


Control 10.2 min 10 min 2.3 min


Daxx


7.5 min 6 min


6 min 18 min


2.64 min 2 min 14 min


Prometaphase-
anaphase


Control
Daxx


31.2 min
37.6 min


28 min
32 min


7.9 min
10.36 min


22 min
22 min


58 min
78 min


A










B
Daxx+/+ Daxx-/-
I 1 It 11 1111 "uli II ei iii
U Ei |( |t aia fI |I| 1Wt at:| *,I

ii i I .. t ta <. aa ,


Figure 3-1. Characterization of Daxx- Mouse Embryos and Cells. A) HOECHST
immunohistochemical staining of Daxx and Daxx E9.5 mouse embryos. Mitotic
stages are marked with small arrows. P/M=pro-metaphase, M=metaphase,
A=anaphase, T=telophase. Apoptotic cells are marked with large arrowheads. B)
Karyotype analysis of Daxx and Daxx MPEFs generated from E9.5 embryos.
Daxx- MPEFs exhibit aneuploidy.












clone 1 clone 2 C
C
0

I



t
0







Figure 5-1. Continued.


jlo


RAsUNfCkne I Rms CIrne 2 DIAPi:marn.i
*lecor II2axx WT Man G C130ax c









embryos which showed no apparent changes in mitotic protein expression (data not shown).

Thus, there is little evidence to suggest Daxx can regulate transcription of known mitotic

proteins. The expression of Daxx protein throughout different stages of the cell cycle was also

analyzed and found to change insignificantly (Figure 3-6). Thus, Daxx is a very stable protein

throughout the cell cycle that has direct involvement in the regulation of mitotic progression.

Cell cycle-dependent protein localization of endogenous Daxx was analyzed in Daxx+

MEFs as they progressed into and through mitosis (Figure 3-7). A striking mitotic spindle-like

association of Daxx was consistently observed in Daxx+/+ MEFs which was not present in Daxx/-

cells beginning in pro-metaphase, upon nuclear envelope breakdown. Daxx was observed to

localize in the nucleus in prophase typically at PML bodies, but by early pro-metaphase, Daxx

protein localization was redistributed to spindle structures as they were being formed (Figure 3-

7). By late pro-metaphase, the majority of Daxx was distributed to the spindle apparatus, away

from PML. Biochemically, it is known that during mitosis, PML bodies are de-sumoylated and

the bulk of PML-associated proteins, including Daxx, leave during this time (Dellaire et al.,

2006; Ishov et al., 2004). The spindle-like localization of Daxx is maintained through

metaphase, but by the later stages of mitosis anaphasee to cytokinesis) this association is absent.

Thus, Daxx is a spindle associated protein that is important for the correct timing of early mitosis

progression in cells.









DAXX AND RASSF1 DEFINE A NOVEL MITOTIC STRESS CHECKPOINT THAT IS
CRITICAL FOR CELLULAR TAXOL RESPONSE




















By

CORY R. LINDSAY


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008









Michaelson, J.S., Bader, D., Kuo, F., Kozak, C., and Leder, P. (1999). Loss of Daxx, a
promiscuously interacting protein, results in extensive apoptosis in early mouse development.
Genes Dev 13, 1918-1923.

Michaelson, J.S., and Leder, P. (2003). RNAi reveals anti-apoptotic and transcriptionally
repressive activities ofDAXX. J Cell Sci 116, 345-352.

Miyoshi, Y., Kim, S.J., Akazawa, K., Kamigaki, S., Ueda, S., Yanagisawa, T., Inoue, T.,
Taguchi, T., Tamaki, Y., and Noguchi, S. (2004). Down-regulation of intratumoral aromatase
messenger RNA levels by docetaxel in human breast cancers. Clin Cancer Res 10, 8163-8169.

Morrow, C.J., Tighe, A., Johnson, V.L., Scott, M.I., Ditchfield, C., and Taylor, S.S. (2005).
Bub 1 and aurora B cooperate to maintain BubRI-mediated inhibition of APC/CCdc20. J Cell Sci
118, 3639-3652.

Morse, D.L., Gray, H., Payne, C.M., and Gillies, R.J. (2005). Docetaxel induces cell death
through mitotic catastrophe in human breast cancer cells. Mol Cancer Ther 4, 1495-1504.

Muromoto, R., Sugiyama, K., Takachi, A., Imoto, S., Sato, N., Yamamoto, T., Oritani, K.,
Shimoda, K., and Matsuda, T. (2004). Physical and functional interactions between Daxx and
DNA methyltransferase 1-associated protein, DMAP1. J Immunol 172, 2985-2993.

Musacchio, A., and Salmon, E.D. (2007). The spindle-assembly checkpoint in space and time.
Nat Rev Mol Cell Biol 8, 379-393.

Nefkens, I., Negorev, D.G., Ishov, A.M., Michaelson, J.S., Yeh, E.T., Tanguay, R.M., Muller,
W.E., and Maul, G.G. (2003). Heat shock and Cd2+ exposure regulate PML and Daxx release
from ND10 by independent mechanisms that modify the induction of heat-shock proteins 70 and
25 differently. J Cell Sci 116, 513-524.

O'Shaughnessy, J. (2005). Extending survival with chemotherapy in metastatic breast cancer.
Oncologist 10 Suppl 3, 20-29.

Perlman, R., Schiemann, W.P., Brooks, M.W., Lodish, H.F., and Weinberg, R.A. (2001). TGF-
beta-induced apoptosis is mediated by the adapter protein Daxx that facilitates JNK activation.
Nat Cell Biol 3, 708-714.

Pines, J. (2006). Mitosis: a matter of getting rid of the right protein at the right time. Trends Cell
Biol 16, 55-63.

Pluta, A.F., Earnshaw, W.C., and Goldberg, I.G. (1998). Interphase-specific association of
intrinsic centromere protein CENP-C with HDaxx, a death domain-binding protein implicated in
Fas-mediated cell death. J Cell Sci 111 (Pt 14), 2029-2041.

Ravdin, P., Erban, J., and Overmoyer, B. (2003). Phase III comparison of docetaxel and
paclitaxel in patients with metastatic breast cancer. Eur J Cancer suppll:32.








MDA MDA
T47D SKBR3 BT-20 MB 231 MB468 MCF7





actin .
Daxx/actin 1.0 1.6
B control 2,


0Il


5.4 1.1 14.0 5.1


cel 48h of oa


mmm
C
120%
on 100% 100%,-


t 40%3
1.5%
MDA MB-468 T47D
Control M 24h pachtaxel d48h paclitaxel
Figure 4-2. Colony Formation of Breast Cancer Cells After Paclitaxel Treatment is Daxx-
Dependent. A) differential Daxx expression among breast cancer cell lines. Daxx
accumulation normalized by actin (bottom). B) & C): MDA MB 468 (high Daxx) and
T47D (low Daxx) were treated with 10 nM paclitaxel for 24h or 48h. Colonies were
fixed and stained with crystal violet and calculated 5 days after drug withdrawal.
Note differential taxol response (% of survival) between MDA MB 468 and T47D.


Daxx


-- 0









CENP-C (Pluta et al., 1998); depletion of Ams2, a Daxx-like motif-containing GATA factor in S.

pombe results in chromosome mis-segregation (Chen et al., 2003). Extensive aneuploidy was

also observed in Daxx-- MEFs, a typical manifestation of chromosomal mis-segregation in cells

(Figure 3-1, bottom). This condition was observed in three independent Daxx-- MEF cell lines,

suggesting Daxx may be important for accurate chromosomal separation and/or cell division.

Furthermore, upon paraffin-sectioning of E9.5 Daxx-deficient embryos and HOECHST staining

of cells, an increased number of Daxx-deficient cells were observed in early mitosis, specifically

pro-metaphase (Figure 3-1, top). In contrast, a smaller number of cells were observed in later

stages of mitosis anaphasee, telophase and cytokinesis) compared to corresponding wild type

Daxx embryos. These findings suggest a potential involvement of Daxx in mitosis progression.

To understand the extent of Daxx influence on the rate and timing of mitosis, the effect of

Daxx-depletion on mitosis progression was analyzed on HEp2 cells which were stably

transfected with GFP-Histone H3 and a control siRNA (mouse Daxx) or human Daxx siRNA

using fluorescence time-lapse microscopy. Cells expressing human Daxx siRNA were

efficiently depleted of Daxx-protein levels, compared to control (Figure 3-2). Photo-toxicity

associated with fluorescence microscopy limited the time interval of movies to one frame every 2

minutes. Using the precedent of Meraldi and colleagues, key events in mitosis were monitored

and timed based on chromosome condensation to nuclear envelope breakdown prophasee to pro-

metaphase), nuclear envelope breakdown to central chromosome alignment (pro-metaphase to

metaphase) and chromosome segregation anaphasee) (Meraldi et al., 2004). Results of this

analysis are formulated in Table 3-1. Extending from the beginning of chromosomal

condensation to nuclear envelope breakdown, completion of prophase in Daxx-depleted cells

was on average 31.4% faster (7.05 min +/- 2.64 min in Daxx siRNA, compared to 10.2 min +/-









known to interact with sumoylated proteins (Pluta et al., 1998; Yang et al., 1997). Since yeast

themselves frequently use sumoylation for protein-protein interactions, this can potentially

explain the frequent occurrence of Daxx-false positives from yeast two hybrid screens

(Michaelson, 2000). For this analysis, the amino-terminal region of Daxx was used which has

previously not been adapted for yeast two-hybrid screens. This novel approach identified several

proteins with functional implications in the regulation of cell division and mitosis. Sequence

analysis of two particularly strong interaction clones revealed homology with amino acids 5-270

and 30-270 of mouse RAS associated domain family 1 splice form C (RassflC) (Figure 5-1).

Retransformation and P-gal reporter assay confirmed the specificity of this interaction in yeast,

pointing at the amino -terminus of Daxx as a potential region of Rassfl interaction (Figure 5-1).

DMAP1 (DNA methyltransferase associated protein 1) that was recently shown to interact with

Daxx (Muromoto et al., 2004) was used as positive control in these experiments. Rassfl is a

highly conserved throughout species and its locus is frequently altered in human cancer. Among

several alternative splice forms, RassflA and RassflC are the most abundantly expressed. Both

RassflA and RassflC are tubulin-associated proteins which influence the overall stability and

dynamics of microtubules (Liu et al., 2003). While RassflC function is relatively unknown,

RassflA has recently been shown to regulate early mitosis progression, particularly during

prometaphase (Song et al., 2004). In addition, RassflA over-expression leads to mitotic arrest

and inhibition of colony growth (Liu et al., 2003).

Daxx-'/ embryos showed an increased accumulation of cells in pro-metaphase, suggesting

that Daxx could cooperate with RassflA function during pro-metaphase. In light of the known

properties of both RassflA and RassflC on the influence of microtubules, the potential interplay

of Daxx and Rassfl binding could also partially explain alterations in sensitivity of Daxx-









Preferential Alteration of RassflA in Cancer

Allelic loss of the short arm of human chromosome 3 (3p) is one of the most frequently

occurring events in lung cancers (90% of small cell lung cancers 50-80% of non-small cell lung

cancers) (Dammann et al., 2000). Specifically, the region 3p21.3, where the Rassfl locus

resides, displays regular loss of heterozygosity and homozygous deletions in cancer cells.

Promoter methylation and loss of protein expression have been directly correlated specifically to

RassflA in many different tumor cell lines and this methylation has been confirmed in at least 37

different tumor types (Agathanggelou et al., 2005). Methylation of the RassflA promoter does

not affect expression of RassflC and in numerous instances RassflC is used as a control for

RNA integrity and loading when studying RassflA expression. Therefore, epigenetic

inactivation of RassflC is much rarer than that of RassflA. In one study, RassflA promoter

methylation occurred with a frequency of 62% in forty-five breast carcinomas that were analyzed

and in many instances treatment of cells with the DNA methylation inhibitor 5-aza-2'-

deoxycytidine reactivated RassflA transcript expression (Dammann et al., 2001). Because of

this very strong correlation between loss of RassflA expression and tumor-specific cell lines it is

highly regarded as a candidate molecular marker for tumor diagnosis. Differential loss of

RassflC expression, however, has been documented in some tumor cell lines and may be

regulated by more significant post-transcriptional mechanisms than RassflA (Donninger et al.,

2007).



RassflA and Cell Cycle Control

Because of the intense focus of RassflA inactivation in cancers, the majority of functional

studies about Rassfl have swayed largely to the A isoform of Rassfl because of its potential

tumor suppressor roles. Indeed, a seminal observation that has been repeatedly confirmed for

25



































2008 Cory R. Lindsay









in primary human fibroblasts depleted of Daxx and seems to be influenced instead by the JNK

pathway during UV and H202 treatment (Khelifi et al., 2005). Both T47D and MDA-MB 468

express mutant, transcriptionally inactive p53 (Concin et al., 2003); thus, differences in

paclitaxel response between these cell lines are most likely p53-independent.

To further confirm Daxx dependent cell survival upon paclitaxel treatment, Daxx protein

levels were depleted in MDA MB 468 breast cancer cells, in HEp2 human epithelial carcinoma

cells (Figure 4-6) and in human fibroblast cell line WI38 by stable expression of anti-Daxx

siRNA (data not shown). In the case of both anti-Daxx siRNA, we observed a marked decrease

and slower rate of micronucleation and an increase in mitotic index (Figure 4-6). In all cases, the

levels of apoptosis throughout this ongoing process were negligible (data not shown).

Importantly, the decline in micronucleation and increase of mitotic index observed in anti-Daxx

siRNA cell lines correlated with an increased survival of cells under colony formation conditions

(Figure 4-6). Thus, depletion of Daxx by siRNAs targeted against Daxx message in a variety of

human cell lines reproduces the original finding that level of Daxx is critical for paclitaxel

response.

The combination of these data allows the proposition of a model in which cells follow one

of two paths in response to taxol and which is dependent on Daxx protein level: Daxx positive

and taxol-sensitive cells will block in mitosis only transiently, followed by micronucleation,

while Daxx deficient and taxol-resistant cells have prolonged mitotic block and continue

proliferation after drug decay and microtubule dynamics restoration, thus surviving

chemotherapy. This model emphasizes Daxx as a trigger for cellular taxol response-particularly

in apoptosis-reluctant cells; cells lacking a functional Daxx protein display an increased

resistance to taxane exposure.









localization showed Daxx to be a strictly nuclear associated protein, as expected. In stark

contrast, both GFP-RassflA and GFP-RassflC displayed cytoplasmic distribution in a

microtubule-like network with no GFP-fluorescence emanating from the nucleus. Thus, Daxx

and Rassfl are compartmentally separated proteins during interphase. Co-localization of

endogenous Daxx and Rassfl, however, was observed in HEp2 cells progressing through

mitosis-demonstrating the cell cycle regulated interaction between these proteins (Figure 5-4).

Co-localization of Daxx and Rassfl was observed beginning in pro-metaphase and metaphase,

but by later stages of mitosis this association could not be detected.

To understand the cell cycle regulated interaction of Daxx and Rassfl during mitosis and

cellular paclitaxel response, stable expressing RassflA-siRNA was introduced into HEp2 cells

which efficiently deplete RassflA protein level (Figure 5-5). While RassflA-depleted cells

could be easily generated, over-expression of RassflA leads to cell toxicity and inhibition of cell

proliferation independent of RassflA-functioning in cells. Therefore these studies focused

primarily on protein-depleted cell lines. In combination with control and anti-Daxx siRNA HEp2

lines, anti-RassflA siRNA HEp2 cells were exposed to 10 nM paclitaxel for 6-18 hrs and then

related for colony formation assay (Figure 5-5). Strikingly, in the case of both Daxx- and

RassflA-depleted cells, paclitaxel resistance was similar. Both cell lines exhibited a strong

paclitaxel resistant phenotype with the majority of treated cells (75-80%) capable of dividing and

forming colonies after removal of taxol. To gain a biochemical understanding of how mitotic

cells respond to paclitaxel in the absence of Daxx or RassflA, control-, Daxx- and RassflA-

depleted cells were synchronized using a double thymidine block and then released and exposed

to taxol for 6-18 hrs and collected for Western-blot evaluation of Cyclin B protein levels (Figure

5-6). Wild type (parental) and control-siRNA cells revealed an accumulation of Cyclin B as









A Transcriptlron factors bindlng
Fas. PML (ND10) binding
Daxx
aal 177 91 625 7/40 (74
ATRX MtlerolcTiro rnan asoiabon
SRA domain SA RAH domain

S00 94-. R assflA
C,20 DAG M I (rowih uleFas5o.o ati on
1 7
R assf lC
aal 0 124 21B 240R,7tA
I I
S lhred inndl ogy

Yeast Two-Hybr id CI ons
RassfIC Clow I
Rassf10 CCon2 aa30 270
aa30 270
Figure 5-1. Daxx Interacts with Tumor Suppressor Rassfl in Yeast. A) Schematics ofDaxx and
Rassfl isoforms RassflA and RassflC along with mapping of clones identified
through yeast two hybrid screening (RassflC clone 1 corresponding to amino acids
(aa) 5-270 of mouse RassflC, RassflC clone 2 corresponding to aa 30-270).
Homology between RassflA and RassflC is also shown. B) Retransformation assay
of Daxx and RassflC constructs. 1 = RassflC + pGBDC1 (empty vector), 2 =
RassflC + Daxx wt, 3 = RassflC + Daxx C term, 4 = RassflC + Daxx AC. C) P-Gal
reporter assay measuring strength of interaction between individual Rassfl clones and
Daxx wt and deletion mutants. Strong interaction between Rassfl clone 2 is
observed. DMAP (positive control for Daxx interaction) was used as an evaluation
for strength of interaction in this system.









Discussion and Results

Compounds that affect the mitotic spindle apparatus, including microtubule inhibitors such as

paclitaxel (taxol) and nocodazole, are known to differentially affect cells that lack one or more

mitotic spindle associated proteins compared to wild type cells. To further explore the potential

role(s) of Daxx in mitosis and cell division, the microtubule inhibitor taxol, which is a

microtubule hyper-stabilizing compound and the tubulin-destabilizing drug nocodazole were

used to examine response of Daxx+/+ and Daxx-/- MEFs. A very striking and divergent response

was observed between Daxx+/+ and Daxx-- mouse fibroblasts treated with taxol and nocodazole

(Figure 4-1), which was not recapitulated when cells were treated with drugs such as roscovitine,

adriamycin and etoposide (data not shown). The number of mitotic cells in Daxx+/+ MEFs was

much lower compared to Daxx-/- MEFs. Conversely, the occurrence of micronucleated cells in

Daxx+/+ mouse fibroblasts was much higher in comparison to Daxx-deficient cells. This evidence

implies a Daxx-specific function which is inhibited when cells are treated with taxol or

nocodazole. The decrease in micronuclei and elevated mitotic index also correlated with cell

survival in Daxx deficient cells, producing 90% of colonies after 24h of taxol treatment

compared to untreated control, while the survival rate of Daxx+/+ cells was only 15%--

approximately 6 times lower compared to Daxx-/- cells (Figure 4-1). Given the similar time of

cell cycle progression for both cell lines (not shown), the difference in colony formation most

likely reflects a differential survival rate of cells upon drug exposure.

Given that taxol is a very potent chemotherapy agent used to treat breast cancers and other

malignancies, the role of taxol-induced cell death in breast cancer cells was examined with how

this may correlate with the level of protein Daxx. High heterogeneity of Daxx protein level

among breast cancer cell lines was observed (Figure 4-2, normalized on actin). Daxx is a nuclear

protein with obvious ND10/PML body association in interphase cells (Ishov et al., 1999); despite

47









(Ishov et al., 2004). Thus, if Daxx protein resides in the cytoplasm at any period of time, it

would most likely occur as a result of specific relocation as part of signaling pathways.

Several reports describe detailed mechanisms of Daxx re-localization under various stress

conditions (Jung et al., 2008; Jung et al., 2007; Junn et al., 2005; Karunakaran et al., 2007; Song

and Lee, 2003). In many cases, this change in distribution of Daxx was shown to be critical for

cell survival under stress. During glucose deprivation, Daxx is re-located from the nucleus to the

cytoplasm (Song and Lee, 2003, 2004). Mutation of Trpytophan 621 and Serine 667 of human

Daxx, moreover, was sufficient to block nuclear export in these stress conditions, which relied

on stable adenoviral expression of Daxx in adenocarcinoma DU-145 cells. Chemical hypoxia-

induced Daxx relocalization to the cytoplasm was eloquently shown by (Jung et al., 2008) using

detailed confocal imaging analysis of endogenous Daxx in Chinese hamster ovary cell line

PS120. Oxidative stress was also reported to influence the localization of Daxx to the cytoplasm

in DU-145 cells, while over-expression of catalase inhibited nuclear export of Daxx and its

glucose deprivation-induced cytotoxicity (Song and Lee, 2003). A contradictory report by

(Khelifi et al., 2005) however, showed via biochemical separation that Daxx remains in the

nucleus after exposure to hydrogen peroxide or UV treatment. In the majority of studies,

however, the primary means of determining stress induced Daxx localization was accomplished

by immunofluorescence staining of transiently over-expressed protein. In some cases, these

localization patterns may be a result of artifacts-created either by conditions of treatment or

methods of fixing and staining of cells. Alternatively, many observed properties of Daxx in

response to stress stimuli may be cell-line specific. Can stress-induced Daxx re-localization be

explained by the shuttling of other nuclear proteins into the cytoplasm? To date, few studies

have effectively incorporated these controls into their investigations. If a general re-distribution











Plasmid Constructs

GFP-hDaxx wt and deletion mutants were cloned into the BamHI site of pEGFP-C1.

GST-hDaxx wt and deletion mutants were cloned into the BamHI site of pGEX-2T and pGEX-

4T3. pEGFPC2-hRassflC was a gift from Gerd Pfeifer, Beckman Research Institute. RFP-

mRassflC wt and deletion mutants were cloned into the EcoRI/HindIII site of pDS-RedN1.

6His-mRassflC was cloned into pQE-30. GFP-hRassflA was a gift from Dae Sik Lim, Korea

Advanced Institute of Technology. RFP-hRassflA wt and deletion mutants were cloned into the

EcoRI/HindIII site of pDS-RedN1. GST-hRassflA wt and deletion mutants were cloned into the

BamHI site of pGEX-2T and pGEX-4T3.



Stable siRNA Infection

MDA MB 468 and HEp2 cells were transduced with recombinant lentivirus supernatants

encoding hairpin siRNA hDaxx, hRassflA and control expression constructs which were

collected from -5 x 104 transfected 293T cells used for multiple rounds of infection in the

presence of 4 ug/ml polybrene. The lentiviral expression system was provided by Peter M.

Chumakov (Lerner Research Institute, Cleveland; (Sablina et al., 2005). This lentiviral system

comprises a targeting envelope expression vector pCMV-VSV-G, a generic packaging

expression vector pCMV-deltaR8.2 and the expression cassette for custom siRNA pLSL-GFP

that contains a minimal histone H4 promoter that drives transcription of a GFP gene allowing

fluorescent cell sorting. Candidate siRNAs for Daxx were designed according to the Dharmacon

siDESIGN algorithm (http://www.dharmacon.com/sidesign/). Anti-Daxx siRNA 1 was targeted

against base pairs 1552-1570 of hDaxx (sense CTACAGATCTCCAATGAAA; anti-sense

TTTCATTGGAGATC TGTAG); anti-Daxx siRNA 2 was targeted against base pairs 100-118 of

36













MDA MB 468









% of Daxx 100% 3% 4% 112%
onrmalized
on acin


Microulel MDA MB 468


untlad 12itM 15B 244h
-*-MDA MB 46 -U-Convh imNRA ziRNA 1 sRnA 2


HEp3

"- *






lc n I f sI i 7I- I
100% S% 5% 1297%'


Mitotic C4bll HEp2



0% O -----C=-- J --



Lnlread'd 1"n IIll "*h
-- HEp2 --ConMral siLA siNAh1 iNA2


Mironuclei HEp2


4D%



LWSPWA i 2h 18h 241


-HW-Ep2 --Co ntrld sRMA


untreated 12h 18h 24h umnreared 12h ISh 24h
--Mn Us4Ea --CaWMri iRNA sJrWA I sRNA2 [-4-Epa-Contsd iRMA iRNA 1 sIMA 2
Figure 4-6. Paclitaxel Response is Daxx-Dependent. A) Western blot analysis of Daxx depletion
in MDA MB 468 and HEp2 cells. B) Mitotic index of parental MDA MB 468 (left)
and HEp2 cells (right) expressing two independent anti-Daxx siRNAs or control
siRNA which were synchronized via double thymidine block and then released and
treated with 10 nM paclitaxel for 12, 18 and 24 hrs. C) Same as (B) except
micronucleated cells are shown. D) Reduction of Daxx increases cell survival during
paclitaxel treatment. Cells treated in (B) and (C) were set up for colony formation
assay and allowed to grow for 5 days. Colonies were then counted after fixation and
staining with crystal violet. Apoptosis, as a result of paclitaxel exposure, was
negligible across all of these different cell lines (data not shown).


Mitotic Cells MDA MB 468







nrftell 12h 14h 24h
6,-MDA MB 48 -a- Co-IMo ainA siRNA 1 a*RNA 2


siRNAI sIRNA2









predominately nuclear protein. Beginning with the study by Pluta and colleagues, which

characterized the interaction between centrosome component CENP-C and the human form of

Daxx from HeLa cells, large-scale biochemical separation into cytosolic, nuclear and mitotic

chromosome fractions would show that Daxx was a protein associated largely with nuclear

isolated fractions. Immunofluorescence of endogenous Daxx was described as a "punctuate

staining pattern" in interphase nuclei (Pluta et al., 1998). This characteristic Daxx-staining

pattern emphasizes its association with PML bodies. Subsequent studies would attempt to

validate Daxx interaction with apoptosis signal-regulating kinase 1 (ASK1) and show co-

localization and interaction of the two proteins in the cytoplasm, but the bulk of these

experiments were based on transient over-expression and this may not necessarily represent the

behavior of endogenous proteins (Ko et al., 2001). One report by Lalioti et al. showed very

detailed cellular fractionation of NIH-3T3 fibroblasts into nuclear, cytosolic, low-density

microsome, high-density microsome and plasma membrane fractions, with the majority of Daxx

accumulating in the nuclear fraction and a small percentage appearing in low-density

microsomes (Lalioti et al., 2002). Using human Daxx antibody directed against endogenous

protein, the Lalioti group observed along with nuclear staining, a very faint speckle-like

cytoplasmic Daxx pattern in human fibroblasts which presumed there may be two intracellular

pools of Daxx that exist in cells. Strong endogenous interaction between Daxx and other

nuclear-associated proteins including chromatin remodeling proteins ATRX and HDACII,

nuclear sub-domain constituent PML, nuclear protein kinase HIPK1 among others, suggests

Daxx is predominately a nuclear protein. In the absence of PML, the major Daxx housing

domain in interphase, Daxx adopts a primarily chromatin-based localization in the nucleus









REFERENCES


Aapro, M.S. (2001). Neoadjuvant therapy in breast cancer: can we define its role? Oncologist 6
Suppl 3, 36-39.

Agathanggelou, A., Cooper, W.N., and Latif, F. (2005). Role of the Ras-association domain
family 1 tumor suppressor gene in human cancers. Cancer Res 65, 3497-3508.

Agnese, V., Bazan, V., Fiorentino, F.P., Fanale, D., Badalamenti, G., Colucci, G., Adamo, V.,
Santini, D., and Russo, A. (2007). The role of Aurora-A inhibitors in cancer therapy. Ann Oncol
18 Supply 6, vi47-52.

Amaar, Y.G., Minera, M.G., Hatran, L.K., Strong, D.D., Mohan, S., and Reeves, M.E. (2006).
Ras association domain family 1C protein stimulates human lung cancer cell proliferation. Am J
Physiol Lung Cell Mol Physiol 291, L1185-1190.

Bonneterre, J., Spielman, M., Guastalla, J.P., Marty, M., Viens, P., Chollet, P., Roche, H.,
Fumoleau, P., Mauriac, L., Bourgeois, H., et al. (1999). Efficacy and safety of docetaxel
(Taxotere) in heavily pretreated advanced breast cancer patients: the French compassionate use
programme experience. Eur J Cancer 35, 1431-1439.

Bothos, J., Summers, M.K., Venere, M., Scolnick, D.M., and Halazonetis, T.D. (2003). The Chfr
mitotic checkpoint protein functions with Ubcl3-Mms2 to form Lys63-linked polyubiquitin
chains. Oncogene 22, 7101-7107.

Brown, J.M., and Wouters, B.G. (1999). Apoptosis, p53, and tumor cell sensitivity to anticancer
agents. Cancer Res 59, 1391-1399.

Burbee, D.G., Forgacs, E., Zochbauer-Muller, S., Shivakumar, L., Fong, K., Gao, B., Randle, D.,
Kondo, M., Virmani, A., Bader, S., et al. (2001). Epigenetic inactivation ofRASSF1A in lung
and breast cancers and malignant phenotype suppression. J Natl Cancer Inst 93, 691-699.

Cahill, D.P., Lengauer, C., Yu, J., Riggins, G.J., Willson, J.K., Markowitz, S.D., Kinzler, K.W.,
and Vogelstein, B. (1998). Mutations of mitotic checkpoint genes in human cancers. Nature 392,
300-303.

Carvalho, A., Carmena, M., Sambade, C., Earnshaw, W.C., and Wheatley, S.P. (2003). Survivin
is required for stable checkpoint activation in taxol-treated HeLa cells. J Cell Sci 116, 2987-
2998.

Chabalier, C., Lamare, C., Racca, C., Privat, M., Valette, A., and Larminat, F. (2006). BRCA1
downregulation leads to premature inactivation of spindle checkpoint and confers paclitaxel
resistance. Cell Cycle 5, 1001-1007.

Chan, G.K., and Yen, T.J. (2003). The mitotic checkpoint: a signaling pathway that allows a
single unattached kinetochore to inhibit mitotic exit. Prog Cell Cycle Res 5, 431-439.











Roleire from tlJeAmidn
DT <6 7 1fhr r 8 IV 1t0 r llhf
Danxr





Figure 3-6. Western Blot Analysis of Daxx Protein Levels Throughout the Cell Cycle. HEp2
cells were synchronized in G1/S using a double thymidine block (DT) and then
released from thymidine block for the indicated times. Daxx protein levels change
insignificantly throughout each cell cycle stage.

A














B


. xI


Figure 3-7. Dynamics of Daxx Localization in Mitotic MPEF Cells. A) Localization of Daxx in
prophase (at PML), pro-metaphase (spindles), metaphase (spindles) and anaphase-
cytokinesis (no association). B) Change in localization of Daxx during early pro-
metaphase (top picture) from PML bodies to the forming spindle apparatus. By late
pro-metaphase (bottom picture), the majority of Daxx is localized to the spindle
apparatus.


x ,x IDNA1




early proJ~fmetaphm









due to uncertainties about the reproducibility of methods between several groups (Mauriac et al.,

2005).



Daxx: The Story of an Enigmatic Protein

Daxx and Apoptosis

Daxx is a 120 kDa ubiquitously expressed protein with a high degree of similarity

between mice and humans (72% identical by amino acid sequence) (Kiriakidou et al., 1997).

Daxx was initially identified through yeast two-hybrid screens as a Fas-interacting protein (Yang

et al., 1997). In this initial study, mouse Daxx was found to potentiate apoptosis through a novel

pathway involving activation of the jun N-terminal kinase (JNK) and not through an interaction

with the Fas-associated death domain (FADD). A follow-up investigation published a year later,

showed mDaxx could activate the JNK kinase kinase (ASK1) by binding ASK1 and

subsequently relieving an inhibitory intramolecular interaction between the N & C-termini of the

protein (Chang et al., 1998).

Through these first two studies, ideas towards the role of Daxx were predominately

shifted towards activation of apoptosis and promoting cell death. Perlman et al. (2001) were able

to add further weight to the notion of Daxx as a pro-apoptotic molecule by showing it could both

physically and biochemically interact with transforming growth factor- P receptor (TGF-P) and

aid its apoptotic response by inducing JNK activation. Correspondingly, when antisense Daxx

oligo-nucleotides were transfected into AML-12 cells, subsequent TGF-P treatment did not

induce apoptosis (Perlman et al., 2001).

Seemingly contradictory evidence towards the role of Daxx in vivo began accumulating a

couple of years after its initial discovery. The Leder group developed a Daxx knockout showing

a phenotype of extensive apoptosis and lethality by embryonic day 8.5-9, rather than a mouse

17









Cancer Cell Line Response to Taxane Treatment

Breast cancers are often resistant to the therapeutic induction of apoptosis which is likely

due to inactivation of apoptotic pathways (Brown and Wouters, 1999). Therefore, therapies that

promote other types of death, such as mitotic catastrophe, may be preferential for use in treating

breast cancer. Apoptosis was commonly regarded as the major mechanism of cell death in

response to taxanes (Wang et al., 1999); lately, both paclitaxel and docetaxel are observed to

induce dose- and cell line-specific apoptotic or mitotic cell death (Roninson et al., 2001). In

breast cancer cells and non-small cell lung carcinoma cell lines, each of which originate from

supposedly primary targets of taxane therapy in vivo-paclitaxel has a concentration-dependent,

biphasic response. At low, pharmacologically relevant concentrations, mitotic catastrophe is

observed, whereas at high concentrations terminal cell cycle arrest followed by necrosis are

documented (Yeung et al., 1999). Taxane-sensitive human breast cancer cells are blocked in

mitosis only transiently, followed by nuclear fragmentation and mitotic catastrophe, while

resistant breast cancer cells have more prolonged mitotic block and continue proliferation after

drug decay and restoration of microtubule dynamics, thus surviving chemotherapy. The factors

that determine prolongation of mitotic block and, thus, resistance to treatment by taxanes, still

need to be characterized (Wood et al., 2001).


Predictive Markers for Taxane Treatment

Significant numbers of patients are resistant to taxanes or become resistant to this therapy

during treatment. The response rate of docetaxel is -50% even in the first-line chemotherapy and

it decreases to 20-30% in the second- or third-line therapy (Bonneterre et al., 1999; Crown et al.,

2004; Ravdin et al., 2003). Together with side effects, which includes peripheral neurotoxicity

(Rowinsky et al., 1993), it is of vital importance to select responsive patients using prognosis and











Taxanes and Mitotic Checkpoints

Several mitotic checkpoint proteins, including MPS1, Survivin, Chfr, and members of Mad

and Bub protein families (Madl, Mad2, Mad3 (or BubR1) Bub 1, Bub3), sense improper tension

between kinetochores and microtubules of the mitotic spindle and transmit a signal to inhibit

mitotic progression. Inactivation of most of these checkpoint proteins increases sensitivity to

taxane treatment (Carvalho et al., 2003; Lee et al., 2004; Lens et al., 2003). The factors that

determine prolongation of mitotic block and, thus, resistance to treatment by taxanes, remain

incompletely characterized. To date, only few examples are known when inactivation of mitotic

checkpoint proteins leads to reduced sensitivity to taxanes. Inactivation of Chfr, a mitotic-

associated E3 ubiquitin ligase (Bothos et al., 2003; Chaturvedi et al., 2002; Kang et al., 2002)

which degrades the mitotic kinase Aurora A (Yu et al., 2005) leads to decreased sensitivity to the

microtubule depolymerizing drug nocadozole (Scolnick and Halazonetis, 2000). Down-

regulation of breast cancer susceptibility gene 1 (BRCA1) by siRNA leads to increased taxane

resistance in breast cancer cell line MCF-7 (Chabalier et al., 2006). Another report describes

nocodazole-induced delay in mitotic exit upon depletion of p31comet in HeLa cells. p31comet acts

in mitosis by counteracting spindle checkpoint function of Mad2 (Xia et al., 2004). Thus, recent

efforts have started to link sensitivity of tumor cells to taxane treatment with genetic defects in

the cell cycle checkpoints in association with cancer chemotherapy. It has been suggested that

inactivation of mitotic checkpoint proteins can contribute to the selective response of taxane

treatment in vivo (Wassmann and Benezra, 2001). However, mutations in known checkpoint

proteins occur rather rarely (Cahill et al., 1998; Haruki et al., 2001); thus broader studies are

necessary to search for novel molecular targets of taxane therapy.

















]A-

No Treatment 10 uM Noc 10 nM Pac
O Daxx +/+ 0 Daxx -/-


100%
80%
60%
40%
20%
0%
No Treatment 10uMNoc 10 nM Pac
O Daxx +/+ 0 Daxx -/-


U.sh-H N


Phnsphn-H3 ONA
0



Da)cx+/+ control





Daxx+/+ paditaxel





Daxx-/- control


Figure 4-1. Differential Response of Daxx+ and Daxx-- MEFs to Microtubule Inhibitors
Nocodazole and Paclitaxel. A) Mitotic index of MEFs treated with 10 M nocodazole
or 10 nM paclitaxel for 24 hrs. Cells were fixed and stained with phospho-H3
antibody to characterize mitosis. B) Corresponding percentage of micronuclei
formation in cells treated with nocodazole or paclitaxel for 24 hrs. C)
Immunostaining of Daxx+/+ and Daxx-- MEFs treated with paclitaxel using mitotic
markers phospho-H3 and lamin. Note occurrence of mitotic cells (big arrowheads) in
Daxx- cells and micronuclei (small arrows) in Daxx++ cells.


60%
50%
40%
30%
20%
10%
0%



C









ACKNOWLEDGMENTS

I thank my parents Archie and Virginia for always providing the love, support and

encouragement I have needed to pursue my dreams. I thank Mr. Ed Brogie, Dr. Ed Rosa-

Molinar and Dr. Russ Rassmussen who, knowingly or not, motivated, inspired and challenged

me. I thank Dr. Alexander Ishov for his excellent mentorship, guidance and knowledge.









spliced proteins, RassflA and RassflC, can interact with each other and have similar localization

during mitosis or throughout the cell cycle. It was found that GST-RassflA could bind 6His-

RassflC in a pull-down assay (Figure 5-2B). Furthermore, a RassflA mutant (aa 194-340) could

still efficiently bind RassflC (Figure 5-2B). Additional evidence of interaction was obtained by

co-localization of GFP- and RFP-fused RassflA/RassflC. Co-localization between RassflC and

amino acids 194-258 of RassflA indicates smallest region of interaction. Fluorescence time-

lapse microscopy of HEp2 cells expressing GFP-RassflA wt or GFP-RassflC wt also revealed

similar localization throughout the cell cycle implying a potentially constant association of

RassflA-RassflC proteins (data not shown). Thus, the minimal amino acids necessary for

RassflA-RassflC interaction likely resides within the common Ras-Association (RA) domain

shared by both splice forms of Rassfl, which confirms hetero-dimer formation of RassflA and

RassflC and also opens the possibility of homo-dimer formation between each of these isoforms.

To confirm the endogenous interaction of Daxx and Rassfl, double immunofluorescent

staining of Daxx and RassflC was performed in HEp2 cells where these two proteins were found

to be distinctly separated during interphase (Figure 5-3) with Daxx staining relegated strictly to

the nucleus (at PML bodies) and RassflC staining localized strictly to a microtubule-network

pattern. This is in contrast to some reports suggesting RassflC is a nuclear protein that may

interact with Daxx at PML (Kitagawa et al., 2006). To confirm the differential localization of

Daxx and Rassfl in cells, biochemical separation of HEp2 cells into nuclear and cytosolic

fractions was performed as well as 3D-confocal imaging of transiently over-expressed GFP-

RassflA, GFP-RassflC and GFP-Daxx (Figure 5-3 B & C). Daxx was found to be largely a

nuclear associated protein while RassflA appeared strictly in the cytosolic fraction by

biochemical separation. Three-dimensional confocal analysis ofDaxx and Rassfl cellular









2006). Current compounds, including ZM447439, hesperadin and VX680, have been engineered

to target the ATP binding site of Aurora kinases which abolishes their kinase activity. In

addition to paclitaxel, using these compounds has been shown in cell-based assays to alter taxol

response, even when spindle checkpoint proteins were absent (Morrow et al., 2005). Thus, the

potentials of abrogating taxol resistance in combination with other compounds which target the

mitotic spindle checkpoint are promising.

In order to determine if Daxx- and RassflA-mediated taxol response can be altered by

inhibition of Aurora kinase activity, control-, Daxx- and RassflA-depleted HEp2 cells were

treated with taxol in combination with two independent Aurora kinase inhibitors (Figure 5-7).

Compounds used in this study were ZM447439, which targets Aurora A and Aurora B kinase

activity and Aurora kinase inhibitor III, which targets Aurora A kinase activity. After

synchronization with a double thymidine block and a six hour release, control and experimental

cell lines were exposed to paclitaxel alone or paclitaxel in combination with ZM447439 or

Aurora kinase inhibitor III for a period of six hours. After completion of drug exposure, cells

were then related for colony formation assay. Treatment of Daxx- or RassflA-depleted HEp2

cells with paclitaxel alone typically resulted in a very robust taxol resistance, as evidenced by the

75-80% survival rate of these cell lines compared to only 46% of control cells (Figure 5-7).

Strikingly, however, taxol resistance was abolished in Daxx- or RassflA-depleted HEp2 cells

when treated in combination with ZM447439 or Aurora kinase inhibitor III, resulting in

comparable cell survival with control cells (Daxx-siRNA 31%-42%, RassflA siRNA 38%-40%

and control siRNA 29%-37%). Thus, in the absence of Daxx or RassflA, functional inactivation

of the mitotic spindle checkpoint using Aurora kinase inhibitors can change cellular taxol

response and abolish resistance.









Chang, C.C., Lin, D.Y., Fang, H.I., Chen, R.H., and Shih, H.M. (2005a). Daxx mediates the
small ubiquitin-like modifier-dependent transcriptional repression of Smad4. J Biol Chem 280,
10164-10173.

Chang, J.C., Wooten, E.C., Tsimelzon, A., Hilsenbeck, S.G., Gutierrez, M.C., Elledge, R.,
Mohsin, S., Osborne, C.K., Chamness, G.C., Allred, D.C., and O'Connell, P. (2003). Gene
expression profiling for the prediction of therapeutic response to docetaxel in patients with breast
cancer. Lancet 362, 362-369.

Chang, J.C., Wooten, E.C., Tsimelzon, A., Hilsenbeck, S.G., Gutierrez, M.C., Tham, Y.L.,
Kalidas, M., Elledge, R., Mohsin, S., Osborne, C.K., et al. (2005b). Patterns of resistance and
incomplete response to docetaxel by gene expression profiling in breast cancer patients. J Clin
Oncol 23, 1169-1177.

Chaturvedi, P., Sudakin, V., Bobiak, M.L., Fisher, P.W., Mattern, M.R., Jablonski, S.A., Hurle,
M.R., Zhu, Y., Yen, T.J., and Zhou, B.B. (2002). Chfr regulates a mitotic stress pathway through
its RING-finger domain with ubiquitin ligase activity. Cancer Res 62, 1797-1801.

Chen, E.S., Saitoh, S., Yanagida, M., and Takahashi, K. (2003). A cell cycle-regulated GATA
factor promotes centromeric localization of CENP-A in fission yeast. Mol Cell 11, 175-187.

Chi, Y.H., and Jeang, K.T. (2007). Aneuploidy and cancer. J Cell Biochem 102, 531-538.

Cleveland, D.W., Mao, Y., and Sullivan, K.F. (2003). Centromeres and kinetochores: from
epigenetics to mitotic checkpoint signaling. Cell 112, 407-421.

Concin, N., Zeillinger, C., Tong, D., Stimpfl, M., Konig, M., Printz, D., Stonek, F.,
Schneeberger, C., Hefler, L., Kainz, C., et al. (2003). Comparison of p53 mutational status with
mRNA and protein expression in a panel of 24 human breast carcinoma cell lines. Breast Cancer
Res Treat 79, 37-46.

Crown, J., O'Leary, M., and Ooi, W.S. (2004). Docetaxel and paclitaxel in the treatment of breast
cancer: a review of clinical experience. Oncologist 9 Suppl 2, 24-32.

Croxton, R., Puto, L.A., de Belle, I., Thomas, M., Torii, S., Hanaii, F., Cuddy, M., and Reed, J.C.
(2006). Daxx represses expression of a subset of antiapoptotic genes regulated by nuclear factor-
kappaB. Cancer Res 66, 9026-9035.

Dallol, A., Agathanggelou, A., Fenton, S.L., Ahmed-Choudhury, J., Hesson, L., Vos, M.D.,
Clark, G.J., Downward, J., Maher, E.R., and Latif, F. (2004). RASSF1A interacts with
microtubule-associated proteins and modulates microtubule dynamics. Cancer Res 64, 4112-
4116.

Dammann, R., Li, C., Yoon, J.H., Chin, P.L., Bates, S., and Pfeifer, G.P. (2000). Epigenetic
inactivation of a RAS association domain family protein from the lung tumour suppressor locus
3p21.3. Nat Genet 25, 315-319.









that is soon followed by micronucleation and cell death. Importantly, these observations were

also recapitulated in MDA MB 468 and T47D breast cancer cell lines with contrasting levels of

Daxx expression (Figure 4-3). These divergent responses in taxol response are usually seen in

cells deficient in mitotic checkpoint proteins or other regulators of cell division and until these

current studies, Daxx has not been implicated as a regulatory protein of cell cycle progression.

Table 6-1 summarizes a growing list of mitotic proteins and the cellular response to taxol when

these proteins are absent or deregulated in cells. To date, loss of function of the majority of

mitotic proteins, including Madl, Mad2, Bub and BubR1, has shown enhanced response to

paclitaxel in cell culture conditions. Identification of factors which may increase drug resistance

are largely uncharacterized. Among the conclusions of this study is that early mitosis

progression is altered in cells lacking a functional Daxx protein. These observations came from

tissue sectioning of Daxx-/- mouse embryos showing an increased number of cells in early

mitotic stages (Figure 3-1) and were manifested from time-lapse microscopy studies analyzing

mitotic progression in Daxx-depleted cell lines (Figure 3-3, Table 3-1) and Cyclin B protein

stability studies showing an altered rate of cyclin degradation in anti-Daxx siRNA cells (Figure

3-4). Theoretically, alterations of this kind in cell cycle progression could, in part, explain the

genomic instability that is observed in Daxx-deficient mouse cells (Figure 3-1) as improper

mitosis frequently leads to unfaithful chromosome segregation (Chi and Jeang, 2007).

Daxx localization is already understood to be a very dynamic process during the cell cycle.

During G1 and G2, Daxx localizes to PML bodies, while during S phase, it relocates to

condensed heterochromatin where it interacts with ATRX, a chromatin remodeling protein. To

accompany this dynamic protein trafficking, these studies have also revealed that Daxx, a known

transcriptional regulator, can associate with the mitotic spindle apparatus during mitosis (Figure









RassflA as a tumor suppressor is the reintroduction of the protein by over-expression reduces

colony formation in soft agar assays, suppresses growth and reduces independence of anchorage-

free cell growth (Burbee et al., 2001; Dammann et al., 2000). Along with reports of growth

regulation was the first observations that RassflA could block cell cycle progression. Over-

expression of RassflA was found not to impact apoptosis, but to block cells strongly in the G1

stage of interphase and inhibit accumulation of cyclin Dl (Shivakumar et al., 2002).

Importantly, these findings also showed that RassflA transcript variants (identified from breast

and lung tumor samples) could not block cells in G1 compared to wild-type RassflA

(Shivakumar et al., 2002). Supporting the evidence of RassflA function in Gl-arrest was a study

by the Fenton group showing interaction of p20E4F with RassflA that was necessary for G1-

arrest (Fenton et al., 2004). Furthermore, this function became even more intricate when Song

and colleagues described a dynamic cell cycle-dependent regulation of RassflA protein levels by

Skp2 ubiquitin ligase complex (Song et al., 2007). When RassflA levels were degraded by

targeted ubiquitination, it was shown that cells were able to sufficiently progress through G1 into

S phase. Subsequent studies would also begin to analyze sub-cellular localization of RassflA

and associate its localization in cells with function.

The Pfeifer group were the first to describe RassflA as a protein that co-localizes with

microtubules during interphase and associates with the spindle apparatus during mitosis (Liu et

al., 2003). By using Rassfl-/- cells and over-expression studies, it was shown that RassflA

provided stability to microtubules and the region of tubulin interaction was mapped to a specific

169 amino acid stretch of the carboxy-terminus of RassflA (Liu et al., 2003). The Vos and

Dallol groups would confirm RassflA-association with microtubules and microtubule-associated

proteins and that these interactions were important for microtubule stability, dynamics and









In tro d u c tio n ............................................................................................................................. 3 8
R e su lts an d D iscu ssio n .................................................................................. ................ .. 3 8

4 DAXX IS A TRIGGER OF CELLULAR TAXOL RESPONSE .......................................46

In tro d u c tio n ............................................................................................................................. 4 6
Discussion and Results .................... ........................... ........ 47

5 DAXX INTERACTS WITH RAS-ASSOCIATION DOMAIN FAMILY 1 (RASSF1)
WHICH COOPERATE IN CELLULAR TAXOL RESPONSE....................................57

Introduction ..... ........................................................57
Discussion and Results .................... ........................... ........ 57

6 SUMMARY AND CONCLUSIONS.......................................................... ............... 72

R E F E R E N C E S ..........................................................................79

B IO G R A PH IC A L SK E T C H ............................................................................... .....................89


































6











Cydin B
Aclr



Cydin B
Actin


HED2 (parental
T 6hr 7hr hr 8.5 hr 9 h 95 hr 10hr 105hr 11 hr
i _


Control siRNA
T 6hr 7 hr 8 hr 85 hr 9h 95 hr 10hr 10.5 hr 11 hr
wo-o


Daxx siRNA
DT 6 hr 7 hr 8 hr 8.5hr 9hr 9.5hr 10 hr 10.5 hr 11 hr
CydinB -
Actin r
Figure 3-4. Daxx depletion stabilizes Cyclin B during mitosis. Mitotic progression of
synchronized wild type (parental) HEp2 cells, control-siRNA and Daxx-siRNA cell
lines. Cells were synchronized using a double thymidine block and released and
probed for Cyclin B protein levels at the indicated time points.




^' ^


Cdc27

Cdc20


Mad2 aO

Actin a e em* 1

Figure 3-5. Western Blot Analysis of Mitotic Proteins in Wild type (Parental), Control-siRNA
and Two independent Daxx-siRNA Cell Lines. Protein expression did not differ
significantly in the presence or absence of Daxx.









conditions may be found to regulate Daxx localization, it remains intuitive that cellular factors)

and protein modifications play an important role in the regulation of Daxx inside of the nucleus.

To date, Daxx is a protein that has been identified numerous times through yeast two-

hybrid screens with various other proteins both as "prey" or "bait." A list of proteins which have

been found or used in this way is steadily growing. In some instances, this may be an indication

that Daxx could be a false positive of the experimental system. Yet the truly diverse function of

Daxx-mediated protein interactions has made elucidating the role of Daxx and its biological

significance difficult. The first Daxx-deficient mouse model developed by Michaelson and

colleagues still showed the transcription of a mutant form of Daxx, specifically the C-terminal

479 amino acids of the protein (Michaelson et al., 1999). At least theoretically, this C terminal

fragment could be responsible for the observed levels of apoptosis and other phenomenon

associated with Daxx-deficiency. A more comprehensive Daxx knockout was generated by the

Ishov Lab, however, which showed a similar phenotype (Ishov et al., 2004). By embryonic day

8, Daxx-'/ mice were developmentally retarded and by day 11.5-12.5, embryos distenegrated

completely (Ishov et al., 2004). As we continue to learn more of Daxx biology, we will

continually add more to what we already know as a truly unique protein with diverse cellular

functions.



Cellular Localization of Daxx

The sub-cellular localization of Daxx has been a controversy since it was discovered as a

factor involved in Fas-induced apoptosis. Daxx was identified via yeast-two hybrid screening

using Fas as "bait." While Daxx was not completely characterized from this screening, these

findings thrust forward the ideology that Daxx would be found as a cytoplasmic-oriented protein

near the cell membrane. A subsequent paradox would ensue when Daxx was discovered as a

21









cells entered mitosis, but a marked drop-off in Cyclin B protein level ensued indicating cell exit

from mitosis in the later stages of paclitaxel treatment. In contrast, Daxx- and RassflA-siRNA

cells showed a similar accumulation of Cyclin B but these levels were maintained throughout the

course of the experiment, indicating cells were still arrested in mitosis. Thus, Daxx and RassflA

are necessary for efficient cellular response to paclitaxel which includes entry into and exit from

mitosis during treatment.

Sustained Cyclin B protein levels in response to mitotic stresses like paclitaxel is an

indication of prolonged spindle checkpoint activation (Musacchio and Salmon, 2007). During

normal cellular response to paclitaxel, cells will transiently arrest in mitosis due to an activated

spindle checkpoint but will exit mitosis by degrading mitotic substrates (i.e. Cyclin B) because

taxol-generated errors (i.e. microtubular tension, unattached kinetechores) cannot be corrected.

In the absence of Daxx or RassflA, cells remain in a prolonged mitotic block as evidenced by

sustained Cyclin B protein levels. Many different regulatory proteins are involved in proper

spindle checkpoint operation, including the Aurora kinases, a family of serine/threonine kinases

that are highly conserved phylogenetically. Specifically, Aurora A and Aurora B are involved in

the proper placement and localization of key mitotic checkpoint proteins (Ditchfield et al., 2003)

and absence or depletion of Aurora kinases causes spindle checkpoint-override (Fu et al., 2007).

Therapeutically, it would be advantageous to target Aurora kinases in tumors because Aurora A

and Aurora B are frequently up-regulated in cancers (Keen and Taylor, 2004). As a result,

several Aurora kinase inhibitors are in phase I & II clinical trials to evaluate their efficacy as

chemotherapeutic agents (Agnese et al., 2007; Keen and Taylor, 2004). Clinical strategies for

enhancing paclitaxel response are continuously being studied and developed and one potential

method involves the use of taxol in combination with Aurora kinase inhibitors (Malumbres,









Lens, S.M., Wolthuis, R.M., Klompmaker, R., Kauw, J., Agami, R., Brummelkamp, T., Kops,
G., and Medema, R.H. (2003). Survivin is required for a sustained spindle checkpoint arrest in
response to lack of tension. Embo J 22, 2934-2947.

Li, H., Leo, C., Zhu, J., Wu, X., O'Neil, J., Park, E.J., and Chen, J.D. (2000). Sequestration and
inhibition of Daxx-mediated transcriptional repression by PML. Mol Cell Biol 20, 1784-1796.

Li, W., Zhu, T., and Guan, K.L. (2004). Transformation potential ofRas isoforms correlates with
activation of phosphatidylinositol 3-kinase but not ERK. J Biol Chem 279, 37398-37406.

Lin, D.Y., Lai, M.Z., Ann, D.K., and Shih, H.M. (2003). Promyelocytic leukemia protein (PML)
functions as a glucocorticoid receptor co-activator by sequestering Daxx to the PML oncogenic
domains (PODs) to enhance its transactivation potential. J Biol Chem 278, 15958-15965.

Lindsay, C.R., Scholz, A., Morozov, V.M., and Ishov, A.M. (2007). Daxx shortens mitotic arrest
caused by paclitaxel. Cell Cycle 6, 1200-1204.

Liu, L., Amy, V., Liu, G., and McKeehan, W.L. (2002). Novel complex integrating mitochondria
and the microtubular cytoskeleton with chromosome remodeling and tumor suppressor RASSF 1
deduced by in silico homology analysis, interaction cloning in yeast, and colocalization in
cultured cells. In Vitro Cell Dev Biol Anim 38, 582-594.

Liu, L., Baier, K., Dammann, R., and Pfeifer, G.P. (2007). The tumor suppressor RASSF1A does
not interact with Cdc20, an activator of the anaphase-promoting complex. Cell Cycle 6, 1663-
1665.

Liu, L., Tommasi, S., Lee, D.H., Dammann, R., and Pfeifer, G.P. (2003). Control of microtubule
stability by the RASSF1A tumor suppressor. Oncogene 22, 8125-8136.

Malumbres, M. (2006). Therapeutic opportunities to control tumor cell cycles. Clin Transl Oncol
8, 399-408.

Mansilla, S., Bataller, M., and Portugal, J. (2006a). Mitotic catastrophe as a consequence of
chemotherapy. Anticancer Agents Med Chem 6, 589-602.

Mansilla, S., Priebe, W., and Portugal, J. (2006b). Mitotic catastrophe results in cell death by
caspase-dependent and caspase-independent mechanisms. Cell Cycle 5, 53-60.

Mauriac, L., Debled, M., and MacGrogan, G. (2005). When will more useful predictive factors
be ready for use? Breast 14, 617-623.

Meraldi, P., Draviam, V.M., and Sorger, P.K. (2004). Timing and checkpoints in the regulation
of mitotic progression. Dev Cell 7, 45-60.

Michaelson, J.S. (2000). The Daxx enigma. Apoptosis 5, 217-220.









CHAPTER 1
INTRODUCTION

Breast Cancer and Chemotherapy Resistance

Breast cancer is the most frequently diagnosed malignancy for women in the United States.

In 2005, an estimated 211,000 new cases of invasive breast cancer were expected to occur

(Cancer Facts and Figures 2005, American Cancer Society). The mortality rate from breast

cancer declined approximately 2.3% per year from 1990 to 2001, mostly due to earlier detection

and improved therapies. Nonetheless, an estimated 40,000 women would die of breast cancer in

the United States in 2005 (Cancer Facts and Figures 2005, American Cancer Society).

Chemotherapy is a very popular treatment option for many breast cancer patients. There are a

number of agents used in adjuvant therapy with established cytotoxic activity, with the taxanes

considered some of the most active (O'Shaughnessy, 2005).



Taxanes and their Activity

Taxanes, the group of cytotoxic drugs that includes paclitaxel (taxol) and docetaxol

(taxotere), are among the most powerful anticancer agents for breast cancer chemotherapy.

Increasing numbers of patients have been treated with these drugs along, or in combination, with

other chemotherapeutic agents. The successful entry of paclitaxel into clinical trials in 1986

boosted an interest in understanding the mechanism of taxane-induced cell death and in studying

pathways and proteins targeted by this treatment, including tubulin as an immediate target

(Schiff et al., 1979) and downstream targets of taxanes, including mitotic proteins (Wood et al.,

2001).

Although taxanes are successful in selective killing of tumor cells in clinical settings,

current understanding of the molecular basis of this therapy is controversial and incomplete. For











Immunofluorescence

Cells were cultured on glass coverslips in 24-well plates (Corning), washed with KM

buffer (100 mM MES, 100 mM NaC1, 120 mM MgCl2, 50% glycerol) and fixed with 1%

paraformaldehyde or ice cold methanol at room temperature for 20 min. After treatment with

0.4% Triton X-100 in PBS (paraformaldehyde-fixed cells only), the cells were incubated with

primary antibody for 1 hr at room temperature and then washed 2X with PBS and secondary

antibodies (FITC or Texas Red-conjugated, Vector Labs) were then applied sequentially for 45

min each at room temperature in the dark. Cells were then stained with HOECHST (Vector

Labs) and mounted on slides using Fluormount G (Southern Biotech), dried and analyzed using a

Leica fluorescent microscope and imaged using Openlab software.



In vitro Pull-down Assay

Daxx and Rassfl constructs were cloned into pGEX-2T (Invitrogen), pGEX-4T3

(Invitrogen), and pQE-30 (Qiagen) plasmids, respectively. Constructs were then transformed

into a Rosetta strain of E. coli. Protein expression was induced using 50 mM IPTG (Fisher)

(Daxx constructs) and 100 mM IPTG (Rassfl contracts) at room temperature for 6 hrs (Daxx) or

18 C overnight (Rassfl). Cells were then pelleted and lysed using a lysis buffer consisting of

0.1% Triton X-100, 2 mM phenyl-methyl-sulfonyl fluoride (PMSF) (Calbiochem), 1 [tg/mL

aprotinin (Sigma), 1 [tg/mL leupeptin (Sigma), 1 [tg/mL pepstatin (Sigma) and 50 [tg/mL 2-

mercapto-ethanol (Sigma) in TBS. A GST- and 6X-His pulldown kit (Pierce Biotechnology)

was then used to determine binding capability as per the manufacturer's instructions. Protein

samples were analyzed on 4-20% Tris-HC1, sodium dodecyl sulfate (SDS)/arcrylamide gels

(Biorad).




















RassfA -
Actin~


100%


75%


50%


25%


0%


HEp2
(parental)


Control Daxx siRNA RassflA
siRNA siRNA


m No Treatment i 6 hr Taxol m 12 hr Taxol Ei 18 hr Taxol

Figure 5-5. Depletion ofDaxx or RassflA Desensitizes Cells to Paclitaxel. A) Stable expression
of anti-RassflA siRNA in HEp2 cells. Note depletion of RassflA protein compared
to parental cell lines. B) Percentage of colonies formed from parental, control, Daxx-
and RassflA-depleted HEp2 cells which were synchronized using a double thymidine
block and then released and exposed to paclitaxel for the indicated time periods (6,
12, 18hrs). After treatment, cells were related for colony formation assay. Note
increased survival (paclitaxel resistance) ofDaxx- and RassflA-depleted cells
compared to control and parental cell lines.


r
r









with proliferation abnormalities indicative of a pro-apoptotic gene (Michaelson et al., 1999).

These results suggested that Daxx supported a role in an anti-apoptotic function. Indeed,

knockdown of Daxx expression by RNA interference revealed increased levels of apoptosis as

measured by FACS analysis (Michaelson and Leder, 2003). Conversely, it had been known for

years that over-expression of Daxx would lead to induction of apoptosis as well (Torii et al.,

1999; Yang et al., 1997). What could be the true function of Daxx, in relation to apoptosis, in

vivo?

Stronger evidence towards the role Daxx could be playing in apoptosis came from studies

focusing on the endogenous localization of Daxx in cells. Ishov and colleagues and other groups

afterwards (Croxton et al., 2006; Ishov et al., 1999; Ishov et al., 2004; Zhong et al., 2000), found

Daxx to interact with the promyelocytic leukemia (PML) tumor suppressor protein and could be

subsequently recruited to sub-nuclear domains called ND 10 (PML bodies, PODs, or Kraemer

bodies) upon sumoylation of PML. An apparent nuclear localization of Daxx, as would be

consistently shown by biochemical fractionation and immunofluoresence experiments, raised

concern on how Daxx could be involved in Fas-induced apoptosis ifFas was anchored to the cell

membrane. A study published shortly after the discovery of a Daxx/PML interaction and ND10

localization showed that human Daxx, although a potent enhancer of Fas-induced apoptosis

when over-expressed, did not associate with human Fas in cells and maintained its nuclear

localization (at ND10) even upon stimulation of Fas-induced apoptosis (Torii et al., 1999).

Moreover, the localization of Daxx to ND10 seemed to be critical for enhancing apoptosis as a

Daxx mutant lacking its nuclear localization sequence (and hence its association with PML) was

not as effective at promoting cell death (Torii et al., 1999). Zhong and colleagues also supported

this claim by showing a larger induction of apoptosis (as measured by TUNEL assay) by Daxx in









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DAXX AND RASSF1 DEFINE A NOVEL MITOTIC STRESS CHECKPOINT THAT IS
CRITICAL FOR CELLULAR TAXOL RESPONSE

By

Cory R. Lindsay

May 2008

Chair: Alexander Ishov
Major: Medical Sciences-Molecular Cell Biology


Taxanes, a family of compounds which includes taxol, are powerful chemotherapy agents used

for breast cancer treatment. Large numbers of patients, however, are resistant to these drugs for

unknown reasons. Taxol binds and hyper-stabilizes microtubules, but mutations or alterations in

tubulin occur very rarely in cancers and cannot itself explain the majority of Taxol-resistance

observed in patients. Currently, it is thought that defects in mitotic proteins may affect Taxol

sensitivity in cells. Here, Daxx and Rassfl are identified as novel regulators of cellular Taxol

response. Daxx is a ubiquitously expressed and highly conserved nuclear protein with enigmatic

roles in transcription and apoptosis. Increased prometaphase index in Daxx deficient embryos

and aneuploidy of Daxx knockout cells was observed which suggested a potential function of

Daxx in mitosis and cell division. During interphase, Daxx remains a strictly nuclear associated

protein localized to PML bodies or heterochromatin. Upon nuclear envelope breakdown, Daxx

was found to co-localize and interact with Rassfl at mitotic spindles. Rassfl is a cytoplasmic,

microtubule-associated protein that is important for normal mitotic progression and cell division.

Daxx was also shown to be important for the proper timing and progression of early mitosis.

Together, Daxx and Rassfl define a novel mitotic stress checkpoint that enables cells to









predictive markers. Overcoming resistance or incomplete response to these agents would

represent a major advantage in the clinical treatment of breast cancer (Aapro, 2001; Henderson et

al., 2003).

A number of studies have been carried out to determine a genomic profile that could be

predictive for taxane treatment. In 2003, the group of Dr. Chang published a study of gene

expression profiles of 24 patients before and after four cycles of docetaxel treatment in

correlation with differential response to chemotherapy (Chang et al., 2003). They observed a

differential pattern of 92 genes correlating with docetaxel response allowing the predictive

classification of tumor sensitivity. Later this group observed, in the same cohort of patients,

chemotherapy-driven positive selection of resistant cells populated by genes involved in G2/M

arrest (Chang et al., 2005b). Dr. Kato's group performed a similar study analyzing the expression

profile of 44 breast tumor specimens before treatment with docetaxel in combination with

clinical response to therapy, and developed a diagnostic algorithm to differentiate between

responders and non-responders. They also described elevated expression of redox controlling

genes in non-responding patients (Iwao-Koizumi et al., 2005). The same group described down-

regulation of aromatase in docetaxel response patients thus connecting this type of chemotherapy

with suppression of intra-tumoral estradiol synthesis (Miyoshi et al., 2004). A study comparing

gene profiles before and after chemotherapy by either doxorubicin/cyclophosphamide or

doxorubicin/docetaxel treatment could not detect a significant profile for the prediction analysis

of this combination of chemotherapies probably due to a relatively small group of patients

involved (Hannemann et al., 2005). Despite extensive studies trying to identify predictive

markers for taxane treatment, the clinical application of results have, so far, been limited, partly













A B


% of Micronuclei
90% Q Q-----------L0!.
80% 7 *
90%
70%0 60"00 .-
60%
50%
40% *
30% -
20% 0' "I
0%



-- T47D- MDA-MB 468

% of Mitotic Cells
70% 62%5
60%
50% 3
40%
30% .
20% -''
100 i3 .20".- '*4-
0%



-- T47D MDA-MB 468

% of Apoptotic Cells


0 -ouce


10
7% -


2%
1% 0




-*T47D -MDA-MB 468

Figure 4-3. Response to Paclitaxel Treatment in Breast Cancer Cell Lines with Different Daxx
Level. MDA MB 468 (high Daxx) and T47D (low Daxx) were treated with 10 nM
paclitaxel for 12h, 24h or 36h or mock-treated (control). DNA was stained with
HOECHST 33342. A) Cells were categorized as interphase, mitotic, micronucleated,
and apoptotic based on the nuclear morphology. B) Mitotic cells in control: asterisks;
micronucleated cells: big arrowheads; pro-metaphase cells: arrows; apoptotic cells:
small arrowheads. While micronuclei appear in MDA MB 468 already after 12h of
treatment, majority of T47D cells remain in pro-metaphase after 36h of treatment. C)
Relative accumulation of mitotic cells, micronucleated and apoptotic cells in MDA
MB 468 and T47D cells. The majority of MDA MB 468 cells execute
micronucleation, while T47D are accumulated in pro-metaphase. For each time point
one thousand cells were counted.
















Cyclr B-


HEp2 parentala) Control siRNA

Release into Taxol Release into Taxol




Cydr B ---- 81 s-t % ry -aft,


anti-Daxx siRNA ant-RassflA siRNA
Figure 5-6.Cyclin B Levels are Stabilized in Daxx- and RassflA-Depleted Cells Treated with
Taxol. Wild type (parental) HEp2, control and Daxx- and RassflA-depleted cell lines
were synchronized with a double thymidine block, released and then exposed to taxol
for 6-18hrs. Cells were harvested at the indicated time points and probed for Cyclin
Bl levels using anti-Cyclin B antibody (Santa Cruz). Protein levels were
normalized within each siRNA cell line. Note increased relative stability of cyclin B
(normalized to actin) in both Daxx and RassflA-depleted cell lines.


Release into Taxol


Release into Taxol















Daxx a I



Figure 3-2. Western blot Analysis of Daxx Protein Level in HEp2-H3-GFP Cells Expressing
Control-siRNA or Daxx-siRNA. Note efficient depletion of Daxx using Daxx-
specific siRNAs compared to control siRNAs.


A




ontrol-siRNA or Daxx-siRNA. A) Prophase progression in control and Daxx-





B

C *ol





siBNA

Figure 3-3. Fluorescence time-lapse microscopy images of HEp2-H3-GFP cells expressing either
control-siRNA or Daxx-siRNA. A) Prophase progression in control and Daxx-
depleted cell lines. B) Mitotic progression from nuclear envelope break-down (pro-
metaphase) to chromosomal segregation anaphasee).









treat both asynchronously and synchronously growing cells. Standard concentration of Aurora

kinase inhibitors ZM447439 (Tocris Bioscience) and Aurora kinase inhibitor III (Calbiochem)

was 10 aM. Aurora kinase inhibitors were used to treat only synchronously growing cells.



Embryo Isolation and Culture

Daxx +/+ and Daxx -/- embryos were collected at E9.5 from timed matings of Daxx +/-

mice. Presence of a sperm plug was designated as +0.5 days. Approximately nine days later,

female mice were euthanized and embryos (if present) were collected from the uterus. E9.5

embryos were cultured on glass coverslips for immunostaining or paraffin-embedded for tissue

sectioning. In each case, material from individual embryos was collected for genotyping to

confirm presence or absence of Daxx alleles.



FACS Analysis

Cells were trypsonized and re-suspended in DMEM + 10% FBS + 1% penicillin/

streptomycin, centrifuged and washed 2X with PBS and then fixed in 95% ethanol. After

fixation, cells were treated with 0.5 [tg/mL RNase for 20 minutes and propidium iodide was

added to a final concentration of 20 ag/mL afterwards. Cells were then analyzed via flow

cytometry for cell cycle distribution.



Fluorescence Time-Lapse Microscopy

Control and Daxx-depleted HEp2 cells stably transfected with GFP-Histone H3 (gift from

Duane Compton, Dartmouth University) were cultured on cover dishes for 24 hrs and then

grown at 37C in a microscope chamber supplying 5% CO2. Mitotic cells were imaged using a

confocal Leica TCS SP5 microscope and Leica imaging software.









Ricci, M.S., and Zong, W.X. (2006). Chemotherapeutic approaches for targeting cell death
pathways. Oncologist 11, 342-357.

Rong, R., Jin, W., Zhang, J., Sheikh, M.S., and Huang, Y. (2004). Tumor suppressor RASSF1A
is a microtubule-binding protein that stabilizes microtubules and induces G2/M arrest. Oncogene
23, 8216-8230.

Roninson, I.B., Broude, E.V., and Chang, B.D. (2001). If not apoptosis, then what? Treatment-
induced senescence and mitotic catastrophe in tumor cells. Drug Resist Updat 4, 303-313.

Rowinsky, E.K., Eisenhauer, E.A., Chaudhry, V., Arbuck, S.G., and Donehower, R.C. (1993).
Clinical toxicities encountered with paclitaxel (Taxol). Semin Oncol 20, 1-15.

Sablina, A.A., Budanov, A.V., Ilyinskaya, G.V., Agapova, L.S., Kravchenko, J.E., and
Chumakov, P.M. (2005). The antioxidant function of the p53 tumor suppressor. Nat Med 11,
1306-1313.

Schiff, P.B., Fant, J., and Horwitz, S.B. (1979). Promotion of microtubule assembly in vitro by
taxol. Nature 277, 665-667.

Scolnick, D.M., and Halazonetis, T.D. (2000). Chfr defines a mitotic stress checkpoint that
delays entry into metaphase. Nature 406, 430-435.

Shivakumar, L., Minna, J., Sakamaki, T., Pestell, R., and White, M.A. (2002). The RASSF1A
tumor suppressor blocks cell cycle progression and inhibits cyclin Dl accumulation. Mol Cell
Biol 22, 4309-4318.

Song, J.J., and Lee, Y.J. (2003). Role of the ASK1-SEK1-JNK1-HIPK1 signal in Daxx
trafficking and ASK1 oligomerization. J Biol Chem 278, 47245-47252.

Song, J.J., and Lee, Y.J. (2004). Tryptophan 621 and serine 667 residues of Daxx regulate its
nuclear export during glucose deprivation. J Biol Chem 279, 30573-30578.

Song, M.S., Song, S.J., Ayad, N.G., Chang, J.S., Lee, J.H., Hong, H.K., Lee, H., Choi, N., Kim,
J., Kim, H., et al. (2004). The tumour suppressor RASSF1A regulates mitosis by inhibiting the
APC-Cdc20 complex. Nat Cell Biol 6, 129-137.

Song, M.S., Song, S.J., Kim, S.J., Nakayama, K., Nakayama, K.I., and Lim, D.S. (2007). Skp2
regulates the antiproliferative function of the tumor suppressor RASSF1A via ubiquitin-mediated
degradation at the G(1)-S transition. Oncogene.

Sudo, T., Nitta, M., Saya, H., and Ueno, N.T. (2004). Dependence of paclitaxel sensitivity on a
functional spindle assembly checkpoint. Cancer Res 64, 2502-2508.

Tang, J., Qu, L.K., Zhang, J., Wang, W., Michaelson, J.S., Degenhardt, Y.Y., El-Deiry, W.S.,
and Yang, X. (2006). Critical role for Daxx in regulating Mdm2. Nat Cell Biol 8, 855-862.









RassflC, however, cell proliferation was actually increased compared to cells over-expressing

RassflA. This became some of the first tantalizing evidence to suggest that RassflA and

RassflC could have different effector targets. The microtubule binding and stabilizing functions

of RassflC were also characterized and shown to have identical properties to RassflA-mediated

microtubule stability, although these findings were largely an oversight (Rong et al., 2004)

Interestingly, a study purporting the interaction of RassflC and Daxx in the nucleus was

described by (Kitagawa et al., 2006). This study suggested that upon degradation of Daxx,

RassflC would be released from the nucleus where it can activate the SAPK/JNK pathway to

trigger apoptosis during stress conditions. If proven correct, this evidence and others like it

(Amaar et al., 2006) suggest that although RassflA and RassflC have several properties in

common they may also have many divergent roles in the cell.



Cellular Localization of RassflC

The localization of RassflC in cells has not been a subject of critical attention. Several

reports suggested microtubule-associated localization of over-expressed RassflC was similar to

RassflA localization although these similarities were not emphasized (Liu et al., 2002; Liu et al.,

2003; Vos et al., 2004). Some reports have suggested over-expressed RassflC adopts a nuclear

localization, perhaps in an effort to separate the tumor suppressor-based roles of RassflA from

its lesser isoform (Song et al., 2004). Moreover, the description of both over-expressed and

endogenous RassflC as a component of PML bodies was described by Kitagawa, et al. (2006).

In this study, RassflC was shown to be exported from the nucleus when Daxx was degraded by

ubiquitination or siRNA depletion, implying a Daxx-dependent nuclear localization of RassflC

(Kitagawa et al., 2006). More extensive studies into the partitioning of RassflC inside of the cell

are required.




Full Text

PAGE 1

DAXX AND RASSF1 DEFINE A NOVEL MITO TIC STRESS CHECKPOINT THAT IS CRITICAL FOR CELLULAR TAXOL RESPONSE By CORY R. LINDSAY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Cory R. Lindsay 2

PAGE 3

To my parents Archie and Virginia Lindsay 3

PAGE 4

ACKNOWLEDGMENTS I thank my parents Archie and Virginia for always providing the love, support and encouragement I have needed to pursue my dreams. I thank Mr. Ed Brogie, Dr. Ed RosaMolinar and Dr. Russ Rassmussen who, knowingly or not, motivated, inspired and challenged me. I thank Dr. Alexander Ishov for his ex cellent mentorship, guidance and knowledge. 4

PAGE 5

TABLE OF CONTENTS ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .........................................................................................................................8 1 INTRODUCTION................................................................................................................. .12 Breast Cancer and Chemotherapy Resistance ........................................................................12 Taxanes and their Activity ...............................................................................................12 Taxanes and Mitotic Checkpoints ...................................................................................14 Cancer Cell Line Response to Taxane Treatment ...........................................................15 Predictive Markers for Taxane Treatment .......................................................................15 Daxx: The Story of an Enigmatic Protein ..............................................................................17 Daxx and Apoptosis ........................................................................................................17 Daxx and Transcription ...................................................................................................19 Cellular Localization of Daxx .........................................................................................21 Ras-Association Domain Family-1 (Rassf1) and Cancer .......................................................24 Preferential Alteration of Rassf1A in Cancer ..................................................................25 Rassf1A and Cell Cycle Control .....................................................................................25 Cellular Localization of Rassf1A ....................................................................................27 Functions of Rassf1C......................................................................................................28 Cellular Localization of Rassf1C ....................................................................................29 2 MATERIALS AND METHODS...........................................................................................31 Antibodies ........................................................................................................................31 -Gal Reporter Assay ......................................................................................................31 Biochemical Fractionation...............................................................................................32 Cell Culture and Transfections ........................................................................................32 Cell Cycle Synchronization .............................................................................................33 Colony Formation Assay .................................................................................................33 Confocal Microscopy and Subcellular Localization .......................................................33 Drug Treatment ...............................................................................................................33 Embryo Isolation and Culture .........................................................................................34 FACS Analysis ................................................................................................................34 Fluorescence Time-Lapse Microscopy ............................................................................34 Immunofluorescence .......................................................................................................35 In vitro Pull-down Assay .................................................................................................35 Plasmid Constructs ..........................................................................................................36 Stable siRNA Infection ....................................................................................................36 Western Blotting ..............................................................................................................37 Yeast Two-Hybrid Assay ................................................................................................37 3 DAXX FUNCTION IN MITOSIS.........................................................................................38 5

PAGE 6

Introduction .............................................................................................................................38 Results and Discussion ...........................................................................................................38 4 DAXX IS A TRIGGER OF CE LLULAR TAXOL RESPONSE..........................................46 Introduction .............................................................................................................................46 Discussion and Results ...........................................................................................................47 5 DAXX INTERACTS WITH RAS-ASSOCIAT ION DOMAIN FAMILY 1 (RASSF1) WHICH COOPERATE IN CELL ULAR TAXOL RESPONSE............................................57 Introduction .............................................................................................................................57 Discussion and Results ...........................................................................................................57 6 SUMMARY AND CONCLUSIONS.....................................................................................72 REFERENCES ..............................................................................................................................79 BIOGRAPHICAL SKETCH .........................................................................................................89 6

PAGE 7

LIST OF TABLES Table Page 3-1 Statistical analysis of mitotic progre ssion in controland Daxx-depleted H3-GFPHEp2 cells.. ........................................................................................................................42 6-1 Alteration of several known mitotic proteins and re sultant cellular paclitaxel response.. ............................................................................................................................76 7

PAGE 8

LIST OF FIGURES Figure Page 1-1 Dynamics of paclitaxel action in cells.. .............................................................................30 1-2 Localization of Daxx throughout the cell cycle.. ...............................................................30 3-1 Characterization of Daxx-/mouse embryos and cells. .....................................................42 3-2 Western blot analysis of Daxx protein level in HEp2-H3-GFP cells expressing control-siRNA or Daxx-siRNA. .......................................................................................43 3-3 Fluorescence time-lapse microscopy im ages of HEp2-H3-GFP cells expressing either control-siR NA or Daxx-siRNA.. .............................................................................43 3-4 Daxx depletion stabilizes cyclin B during mitosis.. ...........................................................44 3-5 Western blot analysis of mitotic proteins in wild type (parental), control-siRNA and two independent Daxx-siRNA cell lines. ...........................................................................44 3-6 Western blot analys is of Daxx protein levels throughout the cell cycle. ..........................45 3-7 Dynamics of Daxx localization in mitotic MPEF cells. ....................................................45 4-1 Differential response of Daxx+/+ and Daxx-/MEFs to microtubule inhibitors nocodazole and paclitaxel.. ................................................................................................51 4-2 Colony formation of breast cancer cells after paclitaxel treatment is Daxx-dependent. ...52 4-3 Response to paclitaxel treatment in breas t cancer cell lines with different Daxx level. ....53 4-4 Response of MDA MB 468 and T47D breast cancer cell lin es to increased concentration of paclitaxel.. ...............................................................................................54 4-5 FACS analysis of cell cycle distribution after pa clitaxel treatment. ..................................55 4-6 Paclitaxel response is Daxx-dependent ..............................................................................56 5-1 Daxx interacts with tumo r suppressor Rassf1 in yeast. .....................................................64 5-2 Mapping of Daxx-Rassf1 interaction.................................................................................66 5-3 Cellular distribution of Daxx and Rassf1 during interphase..............................................67 5-4 Co-localization of endogenous Daxx and Rassf1 during mitosis in HEp2 cells ...............68 5-5 Depletion of Daxx or Rassf1A desensitizes cells to paclitaxel. .........................................69 8

PAGE 9

5-6 Cyclin B levels are stabilized in Da xxand Rassf1A-depleted cells treated with taxol.. ..................................................................................................................................70 5-7 Inactivation of the mito tic spindle checkpoint using Aurora kinase inhibitors abolishes taxol resistance in Daxx-and Rassf1A-depleted cells ........................................71 6-1 Dynamics of Daxx-Rassf1 interaction throughout the cell cycle.. ....................................77 6-2 Model depicting Daxx-Rassf1-mediated mitotic stress checkpoint during prometaphase ...........................................................................................................................78 9

PAGE 10

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DAXX AND RASSF1 DEFINE A NOVEL MITO TIC STRESS CHECKPOINT THAT IS CRITICAL FOR CELLULAR TAXOL RESPONSE By Cory R. Lindsay May 2008 Chair: Alexander Ishov Major: Medical Sciences -Molecular Cell Biology Taxanes, a family of compounds which includes taxol, are powerful chemotherapy agents used for breast cancer treatment. Large numbers of pa tients, however, are resist ant to these drugs for unknown reasons. Taxol binds and hype r-stabilizes microtubules, but mutations or alterations in tubulin occur very rarely in cancers and cannot itself expl ain the majority of Taxol-resistance observed in patients. Currently, it is thought that defects in mitotic proteins may affect Taxol sensitivity in cells. Here, Daxx and Rassf1 are id entified as novel regulato rs of cellular Taxol response. Daxx is a ubiquitously expressed and highly conserved nuclear protein with enigmatic roles in transcription and apopt osis. Increased prometaphase index in Daxx deficient embryos and aneuploidy of Daxx knockout cells was observed which suggested a potential function of Daxx in mitosis and cell division. During interpha se, Daxx remains a strictly nuclear associated protein localized to PML bodies or heteroch romatin. Upon nuclear envelope breakdown, Daxx was found to co-localize and interact with Rassf1 at mitotic spindles. Rassf1 is a cytoplasmic, microtubule-associated protein that is important for normal mito tic progression and cell division. Daxx was also shown to be important for the prop er timing and progression of early mitosis. Together, Daxx and Rassf1 define a novel mitoti c stress checkpoint that enables cells to 10

PAGE 11

efficiently exit mitosis (and eventually die) when encountered with specific stress stimuli during mitosis, including Taxol. In the absence of Da xx or Rassf1, cells treated with Taxol remain arrested in mitosis due to a su stained mitotic spindle checkpoint. Upon Taxol decay or removal, these cells can resume mitosis and complete cell divisionthus being Taxol resistant. Inhibition of the spindle checkpoint using Aurora Kinase inhi bitors efficiently abolished Taxol resistance in Daxx and Rassf1 depleted cells. Deregulation of most known mitotic proteins leads to enhanced Taxol response. Absence or depletion of Daxx a nd Rassf1, in contrast, results in increased drug resistance. In the future, Daxx and Rassf1 may be useful predictive markers for the proper selection of patients for taxane chemotherapy. 11

PAGE 12

CHAPTER 1 INTRODUCTION Breast Cancer and Chemotherapy Resistance Breast cancer is the most frequently diagnosed malignancy for women in the United States. In 2005, an estimated 211,000 new cases of invasi ve breast cancer were expected to occur (Cancer Facts and Figures 2005, American Cancer Society). The mortality rate from breast cancer declined approximately 2.3% per year from 1990 to 2001, mostly due to earlier detection and improved therapies. Nonetheless, an esti mated 40,000 women would die of breast cancer in the United States in 2005 (Cancer Facts a nd Figures 2005, American Cancer Society). Chemotherapy is a very popular treatment option for many breast cancer patients. There are a number of agents used in adj uvant therapy with established cy totoxic activity, with the taxanes considered some of the mo st active (O'Shaughnessy, 2005). Taxanes and their Activity Taxanes, the group of cytotoxic drugs that includes paclitaxel (taxol) and docetaxol (taxotere), are among the most powerful antican cer agents for breast cancer chemotherapy. Increasing numbers of patients have been treated with these drugs along, or in combination, with other chemotherapeutic agents. The successful en try of paclitaxel into clinical trials in 1986 boosted an interest in understanding the mechanism of taxane-induced cell death and in studying pathways and proteins targeted by this treatm ent, including tubulin as an immediate target (Schiff et al., 1979) and downstream targets of ta xanes, including mitotic proteins (Wood et al., 2001). Although taxanes are successful in selective kill ing of tumor cells in clinical settings, current understanding of the molecular basis of this therapy is controversial and incomplete. For 12

PAGE 13

a long time, apoptosis was considered the main mechanism of cell death in response to taxane treatment. Currently, a more di stinct model of therapy response is considered, wherein different modes of tumor cell death are likely determin ed by drug concentrati on and genetic background of the cells within a tumor (M orse et al., 2005). At pharmacol ogical concentrations, taxanes reversibly bind to tubulin heterodimers that form microtubulesthis accelerates polymerization and inhibits depolymerization of tubulin, thus disrupting microtubule dynamics. This event, in turn, activates the spindle checkpoint, which invo kes mitotic arrest that in normal, untreated conditions ensures proper chromosomal att achment and alignment and ensures faithful chromosomal segregation preventing aneuploi dy (Chan and Yen, 2003; Cleveland et al., 2003). This mitotic arrest does not pers ist indefinitely. After some pe riod of time, cells usually undergo an aberrant exit from mitosis, characterized by th e lack of metaphase, anaphase and cytokinesis. The nuclear envelope is reformed around individual or groups of chromosomes producing large nonviable cells with multiple micronuclei, wh ich are easily distingui shable morphologically from apoptotic cells. Apoptotic cells will have small, highly condensed chromatin with fragmented nuclei and a diminished cytoplasm, whereas micronucleated cells are much larger with uncondensed chromatin in a pattern reminisc ent of normal nuclei. This type of cell death which results in micronucleated cells is known as mitotic catastrophe and is activated during mitosis as a result of deranged spindle formati on coupled with blocks of different checkpoint mechanisms that activate aberrant chromosome segregation and nuclear fragmentation (Kroemer and Martin, 2005). Taxane-sensi tive human breast cancer cells are blocked in mitosis only transiently, followed by nuclear fragmentation and mitotic cata strophe, while resistant breast cancer cells have more prolonged mitotic block a nd continue proliferation after drug decay and microtubule dynamics restoration, thus surviving chemotherapy (Figure 1-1). 13

PAGE 14

Taxanes and Mitotic Checkpoints Several mitotic checkpoint prot eins, including MPS1, Survivin, Chfr, and members of Mad and Bub protein families (Mad1, Mad2, Mad3 (or BubR1) Bub1, Bub3), sense improper tension between kinetochores and microt ubules of the mitotic spindle and transmit a signal to inhibit mitotic progression. Inactivation of most of these checkpoint proteins increases sensitivity to taxane treatment (Carvalho et al., 2003; Lee et al., 2004; Lens et al., 2003). The factors that determine prolongation of mitotic block and, thus, resistance to treatment by taxanes, remain incompletely characterized. To date, only few ex amples are known when inactivation of mitotic checkpoint proteins leads to re duced sensitivity to taxanes. Inactivation of Chfr, a mitoticassociated E3 ubiquitin ligase (Bothos et al., 2003; Chaturvedi et al., 2002; Kang et al., 2002) which degrades the mitotic kinase Aurora A (Yu et al., 2005) leads to decreased sensitivity to the microtubule depolymerizing drug nocadozole (Scolnick and Halazonetis, 2000). Downregulation of breast cancer susceptibility gene 1 (BRCA1) by siRNA leads to increased taxane resistance in breast cancer cell line MCF-7 (Chabalier et al., 2006). Another report describes nocodazole-induced delay in m itotic exit upon depletion of p31 comet in HeLa cells. p31 comet acts in mitosis by counteracting spindle checkpoint function of Mad2 (Xia et al., 2004). Thus, recent efforts have started to link sensit ivity of tumor cells to taxane tr eatment with genetic defects in the cell cycle checkpoints in association with can cer chemotherapy. It has been suggested that inactivation of mitotic ch eckpoint proteins can contribute to the selective response of taxane treatment in vivo (Wassmann and Benezra, 2001). However, mutations in known checkpoint proteins occur rather rarely (Cahill et al., 1998 ; Haruki et al., 2001); t hus broader studies are necessary to search for novel mol ecular targets of taxane therapy. 14

PAGE 15

Cancer Cell Line Response to Taxane Treatment Breast cancers are often resistant to the ther apeutic induction of apoptosis which is likely due to inactivation of apoptotic pathways (Brown and Wouters, 1999). Therefore, therapies that promote other types of death, such as mitotic catastrophe, may be preferential for use in treating breast cancer. Apoptosis was co mmonly regarded as the major mechanism of cell death in response to taxanes (Wang et al., 1999); lately, both paclitaxel and docetaxel are observed to induce doseand cell linespecifi c apoptotic or mitotic cell d eath (Roninson et al., 2001). In breast cancer cells and nonsmall cell lung carcinoma cell lines, each of which originate from supposedly primary targets of taxane therapy in vivopaclitaxel has a co ncentration-dependent, biphasic response. At low, pharm acologically relevant concentr ations, mitotic catastrophe is observed, whereas at high concentrations termin al cell cycle arrest followed by necrosis are documented (Yeung et al., 1999). Taxane-sensitive human breast cancer cells are blocked in mitosis only transiently, followed by nuclear fr agmentation and mitotic catastrophe, while resistant breast cancer cells have more prolonged mitotic block and contin ue proliferation after drug decay and restoration of microtubule dynami cs, thus surviving chemotherapy. The factors that determine prolongation of m itotic block and, thus, resistance to treatment by taxanes, still need to be characterized (Wood et al., 2001). Predictive Markers for Taxane Treatment Significant numbers of patients ar e resistant to taxanes or become resistant to this therapy during treatment. The response rate of docetaxel is ~50% even in the first-line chemotherapy and it decreases to 20-30% in the s econdor third-line th erapy (Bonneterre et al., 1999; Crown et al., 2004; Ravdin et al., 2003). Together with side e ffects, which includes pe ripheral neurotoxicity (Rowinsky et al., 1993), it is of vital importance to select responsive patients using prognosis and 15

PAGE 16

predictive markers. Overcoming resistance or incomplete response to these agents would represent a major advantage in th e clinical treatment of breast cancer (Aapro, 2001; Henderson et al., 2003). A number of studies have been carried out to determine a genomic profile that could be predictive for taxane treatment. In 2003, the group of Dr. Chang published a study of gene expression profiles of 24 patients before and af ter four cycles of do cetaxel treatment in correlation with differential response to chem otherapy (Chang et al., 2003). They observed a differential pattern of 92 genes correlating wi th docetaxel response allowing the predictive classification of tumor sensitivit y. Later this group observed, in the same cohort of patients, chemotherapy-driven positive selection of resistant cells populated by genes involved in G2/M arrest (Chang et al., 2005b). Dr. Katos group pe rformed a similar study an alyzing the expression profile of 44 breast tumor specimens before tr eatment with docetaxel in combination with clinical response to th erapy, and developed a diagnostic al gorithm to differentiate between responders and non-responders. They also descri bed elevated expressi on of redox controlling genes in non-responding patients (Iwao-Koizumi et al., 2005). The same group described downregulation of aromatase in docetax el response patients thus connec ting this type of chemotherapy with suppression of intra-tumoral estradiol synthesis (Miyoshi et al., 2004). A study comparing gene profiles before and after chemothera py by either doxorubicin/cyclophosphamide or doxorubicin/docetaxel treatment could not detect a significant profile for the prediction analysis of this combination of chemotherapies probably due to a relatively small group of patients involved (Hannemann et al., 2005). Despite exte nsive studies trying to identify predictive markers for taxane treatment, the clinical applica tion of results have, so far, been limited, partly 16

PAGE 17

due to uncertainties about the reproducibility of methods between several groups (Mauriac et al., 2005). Daxx: The Story of an Enigmatic Protein Daxx and Apoptosis Daxx is a 120 kDa ubiquitously expressed pr otein with a high degree of similarity between mice and humans (72% identical by amino acid sequence) (Kiriakidou et al., 1997). Daxx was initially identified thr ough yeast two-hybrid sc reens as a Fas-intera cting protein (Yang et al., 1997). In this initial study, mouse Da xx was found to potentiate apoptosis through a novel pathway involving activation of the jun N-terminal kinase (JNK) and not through an interaction with the Fas-associated death domain (FADD). A follow-up investigation pub lished a year later, showed mDaxx could activate the JNK kina se kinase (ASK1) by binding ASK1 and subsequently relieving an inhibitory intramolecular interaction between the N & C-termini of the protein (Chang et al., 1998). Through these first two studies, ideas toward s the role of Daxx were predominately shifted towards activation of apoptosis and promoting cell death. Perlman et al. (2001) were able to add further weight to the notion of Daxx as a pro-apoptotic mo lecule by showing it could both physically and biochemically interact with transforming growth factorreceptor (TGF) and aid its apoptotic response by inducing JNK ac tivation. Correspondingly, when antisense Daxx oligo-nucleotides were transfected into AML-12 cells, subsequent TGFtreatment did not induce apoptosis (Perlman et al., 2001). Seemingly contradictory evidence towards the role of Daxx in vivo began accumulating a couple of years after its initi al discovery. The Leder group developed a Daxx knockout showing a phenotype of extensive apoptos is and lethality by embryonic day 8.5-9, rather than a mouse 17

PAGE 18

with proliferation abnormalities indicative of a pro-apoptotic gene (Michaelson et al., 1999). These results suggested that Daxx supported a role in an an ti-apoptotic function. Indeed, knockdown of Daxx expression by RNA interference re vealed increased levels of apoptosis as measured by FACS analysis (Michaelson and Leder, 2003). Conversel y, it had been known for years that over-express ion of Daxx would lead to induction of apoptosis as well (Torii et al., 1999; Yang et al., 1997). What could be the true function of Daxx, in relation to apoptosis, in vivo ? Stronger evidence towards the role Daxx could be playing in apoptosis came from studies focusing on the endogenous localization of Daxx in cells. Ishov and colleagues and other groups afterwards (Croxton et al., 2006; Ishov et al., 1999; Ishov et al., 2004; Zhong et al., 2000), found Daxx to interact with the promyelocytic leukemi a (PML) tumor suppressor protein and could be subsequently recruited to sub-nuclear domai ns called ND10 (PML bodies, PODs, or Kraemer bodies) upon sumoylation of PML. An apparent nuclear localization of Daxx, as would be consistently shown by biochemical fractionati on and immunofluoresence experiments, raised concern on how Daxx could be involved in Fas-indu ced apoptosis if Fas was anchored to the cell membrane. A study published shortly after the discovery of a Daxx/PML interaction and ND10 localization showed that huma n Daxx, although a potent enhancer of Fas-induced apoptosis when over-expressed, did not associate with huma n Fas in cells and maintained its nuclear localization (at ND10) even upon stimulation of Fas-induced apoptosis (Torii et al., 1999). Moreover, the localization of Daxx to ND10 seemed to be critical for enhancing apoptosis as a Daxx mutant lacking its nuclear localization sequence (and hence it s association with PML) was not as effective at promoting cell death (Torii et al., 1999). Zhong and colleagues also supported this claim by showing a larger induction of apoptosis (as measured by TUNEL assay) by Daxx in 18

PAGE 19

PML +/+ compared to PML -/cells (Zhong et al., 2000). Thus the localization of Daxx to ND10 and not to the cytoplasm was critic al for Daxx-enhanced apoptosis. Daxx and Transcription The identification of PML interacting with and sequestering Daxx into nuclear domains would become as important a discovery as any study demonstrating the f unctionality of Daxx. Ishov and colleagues showed that in situations where PML was ab sent, Daxx would be relocated to condensed heterochromatin where it could potentially be involved in some biochemical function (Ishov et al., 2004). Subs equent studies provided some ev idence of what Daxx could be doing at these sites by showing it could interact not only with core histones, but histone deacetylase II (HDACII), Dek (Hollenbach et al ., 2002) and the SWI/SNF chromatin remodeling protein ATRX (Xue et al., 2003). These interact ions, among others, brought forth the idea that Daxx could be acting as a regulat or of transcription, not only on the level of repression but activation as well. Among a few of the many genes Daxx has been implicated in regulating include p53 target genes (Gostissa et al., 2004; Zhao et al., 2004), Pa x transcription factor family members (Emelyanov et al., 2002; Hollenbach et al., 2002; Lehembre et al., 2001) and Smad4 (Chang et al., 2005a) A more dynamic role of Daxx became apprec iated when Ishov and colleagues showed a cell cycle dependent localizati on of Daxx between ND10 and heterochromatin (Figure 1-2) (Ishov et al., 2004). They found that during G1 and G2 phase, Daxx could be found in its characteristic location at ND10, while during S phase, Daxx would relocate to condensed heterochromatin. Interestingly during mitosis, ND10 is disass embled, PML de-sumoylated and Daxx no longer associated with the remnants of these nuclear domains (Ishov et al., 2004). What became the fate of Daxx after this se t of events was not addressed. 19

PAGE 20

From the cell cycle-dependent localizati on of Daxx model whic h Ishov and colleagues proposed, ND10 could be considered a site of inactivation of Da xx functiona potential storage depot for Daxx and numerous other prot eins until specific times when they are needed and become active again (Ishov et al., 2004). Although at the time this was not a novel concept, a study conducted by (Li et al., 2000) and similarl y by (Lin et al., 2003) suggested this notion showing that when Daxx was bound to increa sing amounts of PML, the transcriptional repression activity of Daxxas measured by a luciferase reporter assaywas relieved. A possible mechanism which could regulate the localization of Daxx to ND10 or to heterochromatin was shown by Ecsedy and coll eagues when they demonstrated a physical interaction between Daxx and the serine/threoni ne kinase HIPK1 (Ecsedy et al., 2003). This interaction was capable of disp lacing Daxx from PML and re-localizing it elsewhere in the nucleus. In addition, the Ecesedy group f ound that upon phosphorylation of Daxx by HIPK1, Daxx transcriptional repression act ivity was modified (Ecsedy et al., 2003). The investigators could not, however, definitively show a relocalization of Daxx to heterochromatin but rather an association with HDAC1. Additionally, they sh owed that phosphorylation of Daxx by HIPK1 diminished the transcriptionally repressive activ ity of Daxx rather than enhanced it. Other studies, which focused on the condition-dependent localization of Daxx (and other nuclear body associated proteins) were found to be dependent on the sumoylation status of PML as well as cellular stresses such as heat shock and heavy metal exposure (Nefkens et al., 2003). The small ubiquitin-like modifier (SUMO), mo reover, is a post-translational modification added to proteins which affects their function and localization. SUMO bears a 20% identit y to ubiquitin and is covalently linked to a wide range of proteins whose functions are commonly implicated in chromatin organization, transcription and ge nomic stability (Hay, 2005). Although other 20

PAGE 21

conditions may be found to regula te Daxx localization, it remains in tuitive that cellular factor(s) and protein modifications play an important role in the regulati on of Daxx inside of the nucleus. To date, Daxx is a protein that has been identified numerous tim es through yeast twohybrid screens with various other proteins both as p rey or bait. A list of proteins which have been found or used in this way is steadily growin g. In some instances, this may be an indication that Daxx could be a false positive of the experime ntal system. Yet the truly diverse function of Daxx-mediated protein interactio ns has made elucidating the role of Daxx and its biological significance difficult. The fi rst Daxx-deficient mouse model developed by Michaelson and colleagues still showed the transcription of a mutant form of Daxx, specifically the C-terminal 479 amino acids of the protein (Michaelson et al., 1999). At least theoreti cally, this C terminal fragment could be responsible for the obser ved levels of apoptosis and other phenomenon associated with Daxx-deficiency. A more comprehensive Daxx knockout was generated by the Ishov Lab, however, which showed a similar phenotype (Ishov et al., 2004). By embryonic day 8, Daxx -/mice were developmentally retarded and by day 11.5-12.5, embryos distenegrated completely (Ishov et al., 2004). As we continue to learn more of Daxx biology, we will continually add more to what we already know as a truly unique protein with diverse cellular functions. Cellular Localization of Daxx The sub-cellular localization of Daxx has been a controversy since it was discovered as a factor involved in Fas-induced apoptosis. Da xx was identified via yeast-two hybrid screening using Fas as bait. While Daxx was not completely characterized from this screening, these findings thrust forward the ideology that Daxx w ould be found as a cytoplasmic-oriented protein near the cell membrane. A subsequent para dox would ensue when Daxx was discovered as a 21

PAGE 22

predominately nuclear protein. Beginning w ith the study by Pluta and colleagues, which characterized the interaction between centrosome component CENP-C and the human form of Daxx from HeLa cells, large-scale biochemical se paration into cytosolic, nuclear and mitotic chromosome fractions would show that Daxx wa s a protein associated largely with nuclear isolated fractions. Immunofluorescence of endogenous Daxx was described as a punctuate staining pattern in interphase nuclei (Pluta et al., 1998). Th is characteristic Daxx-staining pattern emphasizes its association with PML bod ies. Subsequent stud ies would attempt to validate Daxx interaction with apoptosis sign al-regulating kinase 1 (ASK1) and show colocalization and interact ion of the two proteins in the cy toplasm, but the bulk of these experiments were based on transi ent over-expression and this may not necessarily represent the behavior of endogenous pr oteins (Ko et al., 2001). One repor t by Lalioti et al. showed very detailed cellular fractionation of NIH-3T3 fibr oblasts into nuclear, cy tosolic, low-density microsome, high-density microsome and plasma me mbrane fractions, with the majority of Daxx accumulating in the nuclear fraction and a sm all percentage appearing in low-density microsomes (Lalioti et al., 2002). Using huma n Daxx antibody directed against endogenous protein, the Lalioti group observed along with nuclear staining, a very faint speckle-like cytoplasmic Daxx pattern in human fibroblasts which presumed there may be two intracellular pools of Daxx that exist in cells. Strong endogenous interaction between Daxx and other nuclear-associated proteins including chroma tin remodeling proteins ATRX and HDACII, nuclear sub-domain constituent PML, nuclear protein kinase HIPK1 among others, suggests Daxx is predominately a nuclear protein. In the absence of PML, the major Daxx housing domain in interphase, Daxx adopts a primarily chromatin-based localiz ation in the nucleus 22

PAGE 23

(Ishov et al., 2004). Thus, if Da xx protein resides in the cytopl asm at any period of time, it would most likely occur as a result of specific relocation as part of signaling pathways. Several reports describe detailed mechanisms of Daxx re-localizati on under various stress conditions (Jung et al., 2008; Jung et al., 2007; J unn et al., 2005; Karuna karan et al., 2007; Song and Lee, 2003). In many cases, this change in distribution of Daxx was shown to be critical for cell survival under st ress. During glucose deprivation, Daxx is re-located from the nucleus to the cytoplasm (Song and Lee, 2003, 2004). Mutation of Trpytophan 621 and Serine 667 of human Daxx, moreover, was sufficient to block nuclear export in these stress conditions, which relied on stable adenoviral expression of Daxx in adenocarcinoma DU-145 cells. Chemical hypoxiainduced Daxx relocalization to the cytoplasm was eloquently shown by (Jung et al., 2008) using detailed confocal imaging analysis of e ndogenous Daxx in Chinese hamster ovary cell line PS120. Oxidative stress was also reported to infl uence the localization of Daxx to the cytoplasm in DU-145 cells, while over-expression of catal ase inhibited nuclear export of Daxx and its glucose deprivation-induced cy totoxicity (Song and Lee, 2003). A contradictory report by (Khelifi et al., 2005) however, showed via bioc hemical separation that Daxx remains in the nucleus after exposure to hydrogen peroxide or UV treatment. In the majority of studies, however, the primary means of determining stress induced Daxx localization was accomplished by immunofluorescence staining of transiently over-expressed pr otein. In some cases, these localization patterns may be a result of artifacts created either by conditions of treatment or methods of fixing and staining of cells. Alte rnatively, many observed properties of Daxx in response to stress stimuli may be cell-line spec ific. Can stress-induced Daxx re-localization be explained by the shuttling of other nuclear proteins into the cytoplasm? To date, few studies have effectively incorporated these controls into their investigations. If a general re-distribution 23

PAGE 24

of nuclear proteins is observe d in these cases, it is possible that these phenomena are less attributable to Daxx function and more explainable as a general cellular stress response. Reports such as (Nefkens et al., 2003) for example, may describe more functional stress-induced Daxx redistribution in that the activity of severa l nuclearand ND10-associated proteins were documented in parallel after res ponse to specific stress stimu li. Specifically, Daxx and Sp100 but not PML would disperse from ND10 into th e nucleoplasm due to rapid desumoylation of PML during heat shock, while heavy metal e xposure would release Daxx and PML with Sp100 retained at ND10 (Nefkens et al ., 2003). While there is some ta ntalizing evidence to suggest existence and function of Daxx in the cytoplasm, more extensive studies of endogenous protein trafficking are required. Ras-Association Domain Family-1 (Rassf1) and Cancer The Rassf1 gene locus comprises approximately 11,151 base pairs of the human genome and consists of eight exons (Aga thanggelou et al., 2005). It is found on the short arm of human chromosome 3 (3p21.3). Differential promoter use and alternative sp licing creates seven transcripts Rassf1A-Rassf1G. Of these transcripts, isofor ms Rassf1A and Rassf1C are predominately expressed in all tissues while Rassf1B is expressed in the hemopoeitic system only, and Rassf1D or Rassf1E expres sion is restricted to cardiac or pancreatic cells respectively. Homologues of Rassf1A exist in rodents, fish, nematodes and fruit flies ranging from 38% to 85% identity (Agathanggelou et al., 2005). A Ra ssf1C homologue exists in Xenopus with no apparent homologue for Rassf1A. 24

PAGE 25

Preferential Alteration of Rassf1A in Cancer Allelic loss of the short arm of human chromo some 3 (3p) is one of the most frequently occurring events in lung cancers (90% of small cell lung cance rs 50-80% of non-small cell lung cancers) (Dammann et al., 2000). Specificall y, the region 3p21.3, where the Rassf1 locus resides, displays regu lar loss of heterozygosity and homoz ygous deletions in cancer cells. Promoter methylation and loss of protein expression have been dire ctly correlated specifically to Rassf1A in many different tumor cell lines and this methylation has been confirmed in at least 37 different tumor types (Agathanggelou et al., 2005) Methylation of the Rassf1A promoter does not affect expression of Rassf1C and in numerous instances Rassf 1C is used as a control for RNA integrity and loading when studying Rassf1A expression. Therefore, epigenetic inactivation of Rassf1C is much rarer than th at of Rassf1A. In one study, Rassf1A promoter methylation occurred with a frequency of 62% in forty-five breast carcinomas that were analyzed and in many instances treatment of cells with the DNA methylati on inhibitor 5-aza-2`deoxycytidine reactivated Rassf1A transcript expression (Dammann et al., 2001). Because of this very strong correlation betw een loss of Rassf1A expression and tumor-specific cell lines it is highly regarded as a candidate molecular marker for tumor diagnosis. Differential loss of Rassf1C expression, however, has been documented in some tumor cell lines and may be regulated by more significant pos t-transcriptional mechanisms than Rassf1A (Donninger et al., 2007). Rassf1A and Cell Cycle Control Because of the intense focus of Rassf1A inactiv ation in cancers, the ma jority of functional studies about Rassf1 have swayed largely to the A isoform of Rassf1 because of its potential tumor suppressor roles. Indeed, a seminal obser vation that has been repeatedly confirmed for 25

PAGE 26

Rassf1A as a tumor suppressor is the reintroduction of the pr otein by over-expression reduces colony formation in soft agar assays, suppresses growth and re duces independence of anchoragefree cell growth (Burbee et al., 2001; Dammann et al., 2000). Along with reports of growth regulation was the first observations that Ra ssf1A could block cell cy cle progression. Overexpression of Rassf1A was found not to impact a poptosis, but to block cells strongly in the G1 stage of interphase and inhibit accumulation of cyclin D1 (Shivakumar et al., 2002). Importantly, these findings also showed that Rassf1A transcript variants (identified from breast and lung tumor samples) could not block cells in G1 compared to wild-type Rassf1A (Shivakumar et al., 2002). Supporting the eviden ce of Rassf1A function in G1-arrest was a study by the Fenton group showing interaction of p120 E4F with Rassf1A that was necessary for G1arrest (Fenton et al., 2004). Furthermore, this function became even more intricate when Song and colleagues described a dynamic cell cycle-depe ndent regulation of Rassf1A protein levels by Skp2 ubiquitin ligase complex (Song et al., 2007) When Rassf1A levels were degraded by targeted ubiquitination, it was shown that cells we re able to sufficiently progress through G1 into S phase. Subsequent studies would also begin to analyze sub-cellular localization of Rassf1A and associate its localization in cells with function. The Pfeifer group were the first to describe Ra ssf1A as a protein that co-localizes with microtubules during interphase and associates with the spindle apparatus during mitosis (Liu et al., 2003). By using Rassf1 -/cells and over-expression studies it was shown that Rassf1A provided stability to microtubules and the region of tubulin interaction was mapped to a specific 169 amino acid stretch of the carboxy-terminus of Rassf1A (Liu et al., 2003). The Vos and Dallol groups would confirm Rassf 1A-association with microtubules and microtubule-associated proteins and that these intera ctions were important for microtubule stability, dynamics and 26

PAGE 27

preventing genome instability (Dallol et al., 2004 ; Vos et al., 2004). Liu and colleagues also described that overexpression of Rassf1A induced aberrant mitotic arrest at metaphase in a similar manner to the microtubule stabilizing dr ug taxol and how it affects cells (Liu et al., 2003). This became the first study to link Rassf1A to possible mitotic functions in cells. Additional insight into the role of Rassf1A function in mito sis and cell cycle progression came in the seminal study by (Song et al., 2004). Rassf1A was shown to influence the stability of both Cyclin A and Cyclin B as a result of direct interaction w ith Cdc20 and negative regulation of the anaphase promoting complex (APC). Without Rassf1A, cells were proven to progress through early mitosis (specifically prometaphase) faster as a result of premature activation of APC. Absence of Rassf1A by siRNA depletion also caused centrosome abnormalities and multipolar spindles. Details of interaction of Rassf1A with Cdc20 would later become a subject of controversy as Liu and coll eagues showed that Rassf1A was not capable of interacting with Cdc20 in vitro and immunoprecipitation of Cdc20 with Rassf1A could not be detected in synchronous or asynchronous cells (Liu et al., 2007). In the future, additional studies into the precise role of Rassf1A re gulation of mitosis will be required. Cellular Localization of Rassf1A Not until the Pfeifer group described co-locali zation of Rassf1A with microtubules did the scientific community know about the distribution of the protein in cells (Liu et al., 2003). While this study showed ample evidence of the microtubule-associated network that over-expressed Rassf1A will form in cells, the Vos group were the first to describe the endogenous interaction of Rassf1A with polymerized tubulin (Vos et al., 2004). Depolymerized tu bulin was unable to interact with Rassf1A. The sa me group performed a yeast two-hybrid screen using Rassf1A as bait and identified and confir med several known microtubule-assoc iated proteins as interacting 27

PAGE 28

partners of Rassf1A, including microtubule a ssociated protein 1A (MAP1A), MAP1B and C19ORF5 (Dallol et al., 2004). Several reports have shown because of the direct interaction between Rassf1A, tubulin and tubulin-related prot eins, it served to stabilize microtubules under stress conditions (i.e. nocodazole treatment) (L iu et al., 2003; Rong et al ., 2004). During mitosis, over-expressed Rassf1A was obser ved to co-localize with centr osomes and microtubules during prophase, the spindle apparatu s (spindle poles and spindle fibers) during prometaphase, metaphase and anaphase and microtubules as they reformed in divided daughter cells (Liu et al., 2003; Song et al., 2004). The mi nimal interaction domain that is responsible for Rassf1A binding to microtubules and for association with the spindle apparatus was mapped to amino acids 120-288 of Rassf1A (Liu et al., 2003). Functions of Rassf1C While extensive studies on the epigenetic re gulation of Rassf1A have been performed coupled with essential studies into the function of Rassf1A and its importance in tumor progression, very little is known about Rassf1C biology. Armesilla et al. (2004) used a yeast two-hybrid screen to find novel inte ractors of plasma membrane Ca 2+ pump 4b (PMCA4b) which identified Rassf1C. The interaction be tween Rassf1C and PMCA4b was narrowed down to a region that is common to both Rassf1C and Rassf1A, presumably showing that this interaction is shared between bot h isoforms although this data was not shown. Potential tumor suppressor functions of Rassf1C were first described by Li and colleagues demonstrating that Rassf1C could substantially reduce anchorage inde pendent growth of tumor cells and elicit cell cycle arrest similar to Rassf1A (Li et al., 2004). Amaar et al. performed a similar investigation showing that upon depletion of Rassf1C protein in H1299 cells that lack Rassf1A expression, it caused a significant decrease in cell proliferation (Amaar et al., 2006). Upon over-expression of 28

PAGE 29

Rassf1C, however, cell prolifera tion was actually increased comp ared to cells over-expressing Rassf1A. This became some of the first tant alizing evidence to suggest that Rassf1A and Rassf1C could have different effector targets. The microtubule binding an d stabilizing functions of Rassf1C were also characteri zed and shown to have identical properties to Rassf1A-mediated microtubule stability, although thes e findings were largely an oversight (Rong et al., 2004) Interestingly, a study purporting the interacti on of Rassf1C and Daxx in the nucleus was described by (Kitagawa et al., 2006). This study suggested that upon degradation of Daxx, Rassf1C would be released from the nucleus where it can activate the SAPK/JNK pathway to trigger apoptosis during stress conditions. If pr oven correct, this evidence and others like it (Amaar et al., 2006) suggest that although Rassf 1A and Rassf1C have several properties in common they may also have many divergent roles in the cell. Cellular Localization of Rassf1C The localization of Rassf1C in cells has not b een a subject of critic al attention. Several reports suggested microtubule-asso ciated localization of over-expressed Rassf1C was similar to Rassf1A localization although these si milarities were not emphasized (Liu et al., 2002; Liu et al., 2003; Vos et al., 2004). Some re ports have suggested over-expr essed Rassf1C adopts a nuclear localization, perhaps in an effort to separate the tumor suppressor-based roles of Rassf1A from its lesser isoform (Song et al., 2004). Moreover, the descript ion of both over-expressed and endogenous Rassf1C as a component of PML bodies was described by Kitagawa, et al. (2006). In this study, Rassf1C was shown to be exporte d from the nucleus when Daxx was degraded by ubiquitination or siRNA deplet ion, implying a Daxx-dependent nuc lear localization of Rassf1C (Kitagawa et al., 2006). More exte nsive studies into the partitioning of Rassf1C inside of the cell are required. 29

PAGE 30

Figure 1-1. Dynamics of Paclitaxel Action in Cells. At pharmacological concentrations, paclitaxel reversibly inhibits microtubule dynamics blocking cells in pro-metaphase. Cells that are sensitive to paclitaxel activ ate mitotic block only transiently, followed by micronucleation (mitotic catastrophe) and ce ll death, while resistant cells have a more prolonged pro-metaphase block and continue proliferation after drug decay/ withdrawal and microtubule dynamics re storationthus surviving chemotherapy. Figure 1-2. Localization of Da xx Throughout the Cell Cycle. Du ring G1 and G2 phase, Daxx is localized to PML bodies where it is pr esumably inactive. During S phase, Daxx relocates from PML to condensed chromatin where it interacts with ATRX. The fate and function of Daxx during mitosi s (M phase) is not documented. 30

PAGE 31

CHAPTER 2 MATERIALS AND METHODS Antibodies Antibodies to Daxx (M112), BubR1 (8G1), B ub1 (14H5), Cdc20 (H-175), Cdc27 (AF3.1), Cyclin B (GNS1) and Mad2 (17D10) were from Santa Cruz Biotechnology; GST monoclonal antibody was from Sigma; 6X His monoclonal antibody was from Invitrogen; Rassf1C mouse polyclonal antibody was from UT Southweste rn; Rassf1A monoclonal antibody was from Abcam; Rassf1A/C polyclonal antib ody was a gift from Dae Sik Li m, Korea Advanced Institute of Technology; PML rabbit polyclonal antibody was from Gerd Maul, Wistar Institute; Phospho Histone H3 (Ser10) rabbit polycl onal antibody was from Upstate; -Tubulin monoclonal antibody was from Sigma; Daxx monoclona l antibody 514 and rabbi t polyclonal antibody 2133/2134 were developed as descri bed in (Ishov et al., 2004). -Gal Reporter Assay Starter cultures of yeast were grown overnight in CM selective media and OD 600 readings were measured the next day. Cells were resu spended in breaking buffer (100 mM Tris-HCl pH 8.0, 1 mM DTT, 10% glycerol, 40 mM PMSF) and gl ass beads were added and suspensions were vortexed for approximately 5 min at 4C. Af ter pelleting cell debris, a 1:10 mixture of supernatant/Z buffer (60 mM Na 2 HPO 4 40 mM NaH 2 PO 4 10 mM KCl, 1 mM MgSO 4 ) was made and incubated for 5 min at 28C. 200 L ONPG (Sigma) was added to the mixtures and the time taken for reactions to turn yello w was recorded. Afterwards, 400 L of Na 2 CO 3 was added to stop reactions. Gal activity was measured at OD 420 by making 1:10 mixtures of ONPG reaction/water and recordi ng spectrophotometer readings. Protein concentration was determined by OD 595 spectrophotemeter read ings afterterwards. 31

PAGE 32

Biochemical Fractionation HEp2 cells were separated into nuclear a nd cytosolic fractions using a biochemical fractionation buffer consisting of 250 mM sucrose, 20 mM HEPES-KOH pH 7.4, 10 mM KCl, 1.5 mM MgCl 2 1 mM EDTA, 1 mM EGTA. Cells we re grown on 100 mm dishes and washed 2X with PBS and then placed on ice. Ice cold fractionation buffer was immediately added to dishes and cells were incubated at 4 C with gentle rocking for 15 minutes. Cells were then collected into tubes and dounced five times with a B-type pestle. Afterwards, cell extract was collected into microcentrifuge tubes and centrifuged for 2-3 min at 800 RPM to separate intact nuclei from soluble cytoplasm fraction. Nuclei were washed 2X with fractionation buffer and resuspended in 1 mL fractionation buffer. 5 M NaCl was then added to the buffer to reach a final concentration of 450 mM NaCl. Nuclei were then placed at 4 C and subjected to gentle inverting for 5-10 minutes. Nucl ear extract was then centrifuged at 13,000 RPM for 5 minutes at 4 C to pellet insoluble nuclear material. Cell Culture and Transfections Daxx +/+ MEFs, Daxx -/MEFs, HEp2, MDA MB 468 and T47D cells were maintained in Dulbeccos Modified Eagles Medium with 10% fetal bovine serum and 1% penicillin/ streptomycin and maintainted in 5% CO 2 at 37 C. For transient transfection in mammalian cells, all cDNAs were transfected using the Lipofect amine protocol (Invitrogen) according to the manufacturers instructions. 32

PAGE 33

Cell Cycle Synchronization HEp2 and MDA MB 468 cells were synchr onized at G1/S using a standard double thymidine block protocol. Briefly, asynchronous cells were set up at 20-30% confluency and treated with 2 mM thymidine for 18-20 hrs, washed 2X with PBS and released into normal media without thymidine for 6 hrs. 2 mM thym idine was then added for an additional 18-20 hrs and cells were washed 2X with PBS and released from G1/S bl ock for experiments. Colony Formation Assay Treated and non-treated cells were set up for colony formation using a series of cell dilutions ranging from 1:100 to 1:1000 to optimize colony growth. On average, cells were grown for 5-7 days after set up. Colonies were then washed, fixed with methanol and stained with crystal violet (Fisher) in order to visibly count colonies. Each colony formation assay was set up in triplicate in order to obtain represen table statistics for each drug treatment and its duration of exposure to cells. Confocal Microscopy and Subcellular Localization HEp2 cells were cultured on cover dishes fo r 24 hrs and then tran sfected with GFP-Daxx, GFP-Rassf1A, or GFP-Rassf1C. 24-48 hrs later, live cell confocal images were taken using a Leica TCS SP5 microscope and chamber and Leica imaging software. Drug Treatment Standard concentration of paclitaxel (Calbioc hem) used throughout these studies is 10 nM, unless noted otherwise. Paclitaxel was used to treat both asynchronously and synchronously growing cells. Standard concentratio n of nocodazole (Calbiochem) was 10 M and was used to 33

PAGE 34

treat both asynchronously and synchronously growing cells. Standard concentration of Aurora kinase inhibitors ZM447439 (Tocris Bioscience) and Aurora kinase inhibitor III (Calbiochem) was 10 M. Aurora kinase inhibitors were used to treat only synchronously growing cells. Embryo Isolation and Culture Daxx +/+ and Daxx -/embryos were collected at E9.5 from timed matings of Daxx +/mice. Presence of a sperm plug was designated as +0.5 days. Approximately nine days later, female mice were euthanized and embryos (if pr esent) were collected from the uterus. E9.5 embryos were cultured on glass coverslips fo r immunostaining or paraffin-embedded for tissue sectioning. In each case, mate rial from individual embryos was collected for genotyping to confirm presence or abse nce of Daxx alleles. FACS Analysis Cells were trypsonized and re-suspended in DMEM + 10% FBS + 1% penicillin/ streptomycin, centrifuged and washed 2X with PBS and then fixed in 95% ethanol. After fixation, cells were treated with 0.5 g/mL RNase for 20 minutes and propidium iodide was added to a final concentration of 20 g/mL afterwards. Cells were then analyzed via flow cytometry for cell cycle distribution. Fluorescence Time-Lapse Microscopy Control and Daxx-depleted HEp2 cells stably tr ansfected with GFP-Histone H3 (gift from Duane Compton, Dartmouth University) were cultured on cover dishes for 24 hrs and then grown at 37 C in a microscope chamber supplying 5% CO 2. Mitotic cells were imaged using a confocal Leica TCS SP5 microsc ope and Leica imaging software. 34

PAGE 35

Immunofluorescence Cells were cultured on glass coverslips in 24-well plates (Corning), washed with KM buffer (100 mM MES, 100 mM NaCl, 120 mM MgCl 2 50% glycerol) and fixed with 1% paraformaldehyde or ice cold methanol at room temperature for 20 min. After treatment with 0.4% Triton X-100 in PBS (paraformaldehyde-fixed cells only), the cells were incubated with primary antibody for 1 hr at room temperature and then washed 2X with PBS and secondary antibodies (FITC or Texas Red-conjugated, Vector Labs) were then applied sequentially for 45 min each at room temperature in the dark. Ce lls were then stained with HOECHST (Vector Labs) and mounted on slides using Fluormount G (S outhern Biotech), dried and analyzed using a Leica fluorescent microscope and imaged using Openlab software. In vitro Pull-down Assay Daxx and Rassf1 constructs were cloned into pGEX-2T (Invitrogen), pGEX-4T3 (Invitrogen), and pQE-30 (Qiagen) plasmids, resp ectively. Constructs were then transformed into a Rosetta strain of E. coli. Protein expression was induced using 50 mM IPTG (Fisher) (Daxx constructs) and 100 mM IPTG (Rassf1 contructs) at room temperature for 6 hrs (Daxx) or 18 C overnight (Rassf1). Cells were then pellet ed and lysed using a lysis buffer consisting of 0.1% Triton X-100, 2 mM phenyl-methyl-sulfonyl fluoride (PMSF) (Calbiochem), 1 g/mL aprotinin (Sigma), 1 g/mL leupeptin (Sigma), 1 g/mL pepstatin (Sigma) and 50 g/mL 2 mercapto-ethanol (Sigma) in TBS. A GSTand 6X-His pulldown kit (Pierce Biotechnology) was then used to determine binding capability as per the manufacturers instructions. Protein samples were analyzed on 4-20% Tris-HCl, sodium dodecyl sulfate (SDS)/arcrylamide gels (Biorad). 35

PAGE 36

Plasmid Constructs GFP-hDaxx wt and deletion mutants were cloned into the BamHI s ite of pEGFP-C1. GST-hDaxx wt and deletion mutants were cloned into the BamHI site of pGEX-2T and pGEX4T3. pEGFPC2-hRassf1C was a gift from Gerd Pfeifer, Beckma n Research Institute. RFPmRassf1C wt and deletion mutants were cloned into the EcoRI/HindIII site of pDS-RedN1. 6His-mRassf1C was cloned into pQE-30. GFP-hRassf1A was a gift from Dae Sik Lim, Korea Advanced Institute of Technology. RFP-hRassf1A wt and deletion mutants were cloned into the EcoRI/HindIII site of pDS-RedN1. GST-hRassf1A wt and deletion mutants were cloned into the BamHI site of pGEX-2T and pGEX-4T3. Stable siRNA Infection MDA MB 468 and HEp2 cells were transduced with recombinan t lentivirus supernatants encoding hairpin siRNA hDaxx, hRassf1A and control expression constructs which were collected from ~5 x 10 4 transfected 293T cells used for multiple rounds of infection in the presence of 4 ug/ml polybrene. The lentivir al expression system wa s provided by Peter M. Chumakov (Lerner Research Institute, Cleveland; (Sablina et al., 2005). Th is lentiviral system comprises a targeting envelope expression vector pCMV-VSV-G, a generic packaging expression vector pCMV-deltaR8.2 and the expr ession cassette for custom siRNA pLSL-GFP that contains a minimal histone H4 promoter th at drives transcription of a GFP gene allowing fluorescent cell sorting. Candidate siRNAs for Daxx were designe d according to the Dharmacon si DESIGN algorithm ( http://www.dharmacon.com/sidesign/ ) Anti-Daxx siRNA 1 was targeted against base pairs 1552-1570 of hDaxx (sense CTACAGATCTCCAATGAAA; anti-sense TTTCATTGGAGATC TGTAG); anti-Da xx siRNA 2 was targeted ag ainst base pairs 100-118 of 36

PAGE 37

hDaxx (sense GATGAAGCAGCTGCTCAGC; anti sense GCTCAGCAGCTGCTTCATC); control siRNA was directed against ba se pairs 1262-1284 of SETDB1 (sense TCCTCTTTCTTATCCTCGTATGT, anti-se nse ACATACGAGGAT AAGAAAGAGGA). Western Blotting Cell and protein extracts were ran on pre-made 4-20% Tris -HCl, SDS/Acrylamide gels (Biorad), separated by electrophores is and transferred onto nitroc ellulose and PVDF membranes (Biorad). Membranes were then blocked using 3% milk in 0.1% PBS-Tween for 30 minutes at room temperature. Primary antibodies were added and incubated overnight at 4 C. The next day, membranes were washed 3X with 0.1% PB S-Tween and secondary antibodies (mouse and rabbit HRP conjugates, Cell Signaling Technologies) were added and membranes were incubated at room temperature for 1 hr. After 3X wash in PBS-Tween, membranes we re then exposed with ECL (Amersham) and developed. Yeast Two-Hybrid Assay The bait vector pGBDC1-mDaxx wt, -mDaxx C term and -mDaxx C were transformed into yeast strain PJ-69a and PJ-69 alpha al ong with mouse E11.5 cDNA library plasmid cloned into pGADC1. Briefly, frozen yeast competent cells were vortexed unti l melted completely and then incubated at room temperature for 45 min in 1 mL solution B (40% polyethylene glycol 1000, 200 mM bicine pH 8.35). Afte r pelleting cells, supernatant was removed and cells were resuspended in 1 mL solution C (150 mM NaCl 10 mM bicene pH 8.35) and repeated with a second wash in solution C. Cells were then resuspended in 100 L solution C and plated on selective media (CM media tryptophan,leuc ine and CM media tr yptophan,leucine,histidine). 37

PAGE 38

CHAPTER 3 DAXX FUNCTION IN MITOSIS Introduction Components of the mitotic spindle checkpoint, as well as factors involved in the controlled regulation of the spindle checkpoint and other mitotic processes, are necessary for proper partitioning of chromosomes into daughter cells. In the absence of such factors, cells can mishandle proper chromosomal alignment and segregation during metaphase and anaphase, resulting in lagging chromosomes and aneuploidy (I kui et al., 2005). Proper timing and the rate of mitotic progression can also be affected in the absence of one or more mitotic proteins. For this reason, the cell cycle has been compared to a series of molecular timers: clocks that control the average duration of each cycle, and d ominoes that make each step dependent on the proper completion of a prior step (Meraldi et al., 2004). Cyclin B and Securin are two of the primary mitotic substrates that drive entry into and exit from mitosis (Pines, 2006). Using biochemical analysis of the rela tive stability of these proteins it is possible to measure the duration of mitosis and cell division. Along with the targeted depleti on of proteins and timelapse microscopy, many mitotic fact ors have been shown to affect timing and rate of mitotic progressionadding leverage to th eir already established roles as mitotic regulators and adding the possibility of uncovering novel proteins that may also influence cell cycle progression. Results and Discussion Originally identified as a pro-apoptotic Fas-inte racting protein and later demonstrated to have anti-apoptotic activity (Michaelson, 2000), Da xx is a ubiquitously expressed and highly conserved nuclear protein that is also implicated in transcription regulation. As an almost exclusively nuclear protein duri ng interphase, Daxx inte racts with intrinsic kinetochore protein 38

PAGE 39

CENP-C (Pluta et al., 1998); depletion of Ams2, a Daxx-like motif-containing GATA factor in S. pombe results in chromosome mis-segregation (C hen et al., 2003). Extensive aneuploidy was also observed in Daxx -/MEFs, a typical manifestation of ch romosomal mis-segregation in cells (Figure 3-1, bottom). This condition wa s observed in three independent Daxx -/MEF cell lines, suggesting Daxx may be important for accurate ch romosomal separation and/or cell division. Furthermore, upon paraffin-sectioning of E9.5 Daxx-deficient embryos and HOECHST staining of cells, an increased number of Daxx-deficient cells were observed in early mitosis, specifically pro-metaphase (Figure 3-1, top). In contrast, a smaller number of cells were observed in later stages of mitosis (anaphase, telophase and cyt okinesis) compared to corresponding wild type Daxx embryos. These findings suggest a potential involvement of Daxx in mitosis progression. To understand the extent of Daxx influence on the rate and timing of mitosis, the effect of Daxx-depletion on mitosis progression was an alyzed on HEp2 cells which were stably transfected with GFP-Histone H3 and a control siRNA (mouse Daxx) or human Daxx siRNA using fluorescence time-lapse microscopy. Cells expressing huma n Daxx siRNA were efficiently depleted of Daxx-prot ein levels, compared to control (Figure 3-2). Photo-toxicity associated with fluorescence microscopy limited the time interval of movies to one frame every 2 minutes. Using the precedent of Meraldi and coll eagues, key events in mitosis were monitored and timed based on chromosome condensation to nuclear envelope breakdown (prophase to prometaphase), nuclear envelope breakdown to centr al chromosome alignment (pro-metaphase to metaphase) and chromosome segregation (anaphase) (Meraldi et al., 2004). Results of this analysis are formulated in Table 3-1. Extending from the beginning of chromosomal condensation to nuclear envel ope breakdown, completion of propha se in Daxx-depleted cells was on average 31.4% faster (7.05 min +/2.64 min in Daxx siRNA, compared to 10.2 min +/39

PAGE 40

2.3 min in control siRNA) than control cells (Figure 3-3). Over all, condensation of chromosomes in Daxx-depleted cells occurred more rapidly and the ons et of pro-metaphase occurred sooner. The transition from nuclear envelope break down to chromosome segregation in anaphase was also 20.5% slower in Daxx-depleted cells (37.6 min +/10.36 min for Daxx siRNA, 31.2 min +/7.9 min for control siRNA) (Figure 3-3) Hence, absence of Daxx significantly influences the rate of early mito sis progression in HEp2 cells which emphasizes initial observations of increased pro-metaphase index in Daxx knockout embryos. To confirm that Daxx protein has a direct eff ect on cell cycle progression, HEp2 cells stably expressing control or Daxx siR NA were synchronized using a double thymidine block and then released to allow cells to progr ess through mitosis synchronously so mitotic cyclin B levels could be analyzed. Defective mitotic progression was seen in Daxx-deple ted cells as cyclin B levels were consistently sustained longer than wild-type (parental) or c ontrol siRNA cells (Figure 3-4). Specifically, during the time point of 9-9.5 hr re lease from thymidine block in wild type and control siRNA cells the majority of Cyclin B leve ls are destroyed, however, Cyclin B stability is preserved in Daxx siRNA cells past this time point until 10-10.5 hrs after thymidine release. These data implicate Daxx in the re gulation of mitotic progression, ei ther directly as a regulatory mitotic protein or indirec tly through other mechanisms. To rule out possible indirect mechanisms of Da xx regulation of mitosis, protein expression of several known mitotic proteins including Mad2, a key mitotic checkpoint protein; Cdc20 the activator of the anaphase promoting complex (APC)/cyclosome and Cdc27, a major subunit of the APC was analyzed and compared in contro land two independent Daxx-depleted HEp2 cell lines (Fig 3-5). By comparison, no significant ch anges in protein expression were observed in the presence or absence of Daxx, which supports micr oarray analysis from Daxx +/+ and Daxx -/40

PAGE 41

embryos which showed no apparent changes in mitotic protein expression (data not shown). Thus, there is little evidence to suggest Daxx can regulate transcri ption of known mitotic proteins. The expression of Daxx protein throughout different stag es of the cell cycle was also analyzed and found to change insi gnificantly (Figure 3-6). Thus, Daxx is a very stable protein throughout the cell cycle that has direct involvement in the re gulation of mitotic progression. Cell cycle-dependent protein localization of endogenous Daxx was analyzed in Daxx +/+ MEFs as they progressed into and through mitosis (Figure 3-7). A striking mitotic spindle-like association of Daxx was consistently observed in Daxx +/+ MEFs which was not present in Daxx -/cells beginning in pro-metaphase, upon nuclear envelope breakdown. Daxx was observed to localize in the nucleus in prophase typically at PML bodies, but by ear ly pro-metaphase, Daxx protein localization was redistribu ted to spindle structures as th ey were being formed (Figure 37). By late pro-metaphase, the majority of Daxx was distributed to the sp indle apparatus, away from PML. Biochemically, it is known that dur ing mitosis, PML bodies are de-sumoylated and the bulk of PML-associated proteins, including Daxx, leave during this time (Dellaire et al., 2006; Ishov et al., 2004). The spindle-like lo calization of Daxx is maintained through metaphase, but by the later stages of mitosis (anaphase to cytokinesis) this association is absent. Thus, Daxx is a spindle associated protein that is important for the correct timing of early mitosis progression in cells. 41

PAGE 42

Table 3-1. Statistical Analysis of Mitotic Pr ogression in Controla nd Daxx-Depleted H3-GFPHEp2 Cells. Peak time designates the most frequently occurring time of completion for each stage of mitosis. Stage RNAi Average Peak time Standard deviation Min Max Prophase Control 10.2 min 10 min 2.3 min 6 min 18 min Daxx 7.5 min 6 min 2.64 min 2 min 14 min Prometaphaseanaphase Control 31.2 min 28 min 7.9 min 22 min 58 min Daxx 37.6 min 32 min 10.36 min 22 min 78 min Figure 3-1. Characterization of Daxx -/Mouse Embryos and Cells. A) HOECHST immunohistochemical staining of Daxx +/+ and Daxx -/E9.5 mouse embryos. Mitotic stages are marked with small arrows. P/M=pro-metaphase, M=metaphase, A=anaphase, T=telophase. Apoptotic cells are marked with large arrowheads. B) Karyotype analysis of Daxx +/+ and Daxx -/MPEFs generated from E9.5 embryos. Daxx -/MPEFs exhibit aneuploidy. A B 42

PAGE 43

Figure 3-2. Western blot Analys is of Daxx Protein Level in HEp2-H3-GFP Cells Expressing Control-siRNA or Daxx-siRNA. Note efficient depletion of Daxx using Daxxspecific siRNAs compared to control siRNAs. A B Figure 3-3. Fluorescence time-lapse microscopy im ages of HEp2-H3-GFP cells expressing either control-siRNA or Daxx-siRNA. A) Pr ophase progression in control and Daxxdepleted cell lines. B) Mitotic progressi on from nuclear envel ope break-down (prometaphase) to chromosomal segregation (anaphase). 43

PAGE 44

Figure 3-4. Daxx depletion stabilizes Cyclin B during mitosis. Mitotic progression of synchronized wild type (parental) HEp2 cells, control-siRNA and Daxx-siRNA cell lines. Cells were synchronized using a double thymidine block and released and probed for Cyclin B protein levels at the indicated time points. Figure 3-5. Western Blot Analysis of Mitotic Proteins in Wild type (Parental), Control-siRNA and Two independent Daxx-siRNA Cell Line s. Protein expression did not differ significantly in the presence or absence of Daxx. 44

PAGE 45

Figure 3-6. Western Blot Analys is of Daxx Protein Levels Th roughout the Cell Cycle. HEp2 cells were synchronized in G1/S using a double thymidine block (DT) and then released from thymidine block for the indi cated times. Daxx pr otein levels change insignificantly throughout each cell cycle stage. A B Figure 3-7. Dynamics of Daxx Lo calization in Mitotic MPEF Cells. A) Localization of Daxx in prophase (at PML), pro-metaphase (spindles ), metaphase (spindles) and anaphasecytokinesis (no association). B) Change in localization of Daxx during early prometaphase (top picture) from PML bodies to the forming spindle apparatus. By late pro-metaphase (bottom picture), the major ity of Daxx is localized to the spindle apparatus. 45

PAGE 46

CHAPTER 4 DAXX IS A TRIGGER OF CELLULAR TAXOL RESPONSE Introduction The mitotic spindle checkpoint is a very fl uid and dynamic apparatus designed to monitor microtubule-kinetochore attachment and proper microtubule tension. In the event of an error occurring during pro-metaphase and metaphase, the spindle checkpoint will elicit a wait signal that stalls chromosomal segrega tion and anaphase onset until e rrors are corrected. Microtubuleinhibiting drugs such as taxol and nocodazole (among others) can initia te a prolonged wait signal that if uncorrected, init iates micronucleation (mitotic cata strophe) and cell death (Jiang et al., 2006; Mansilla et al., 2006a; Mansilla et al., 2006b; Ricci and Zong, 2006). Thus, subpopulations of cancer cells with deranged regulation of mitotic spindle checkpoint proteins and other associated factors can respond differen tly to these compounds offering a therapeutic molecular target for treating cancers. Howeve r, in most instances, inactivation or downregulation of mitotic checkpoint proteins leads to increased sensitivity to microtubule inhibiting drugs. In cases of BubR1 or Mad2 depletion, cell s treated with taxol re spond in a much stronger and robust way: the number of cells arrested in mitosis dramatically decreased and cell death was more rapid (Sudo et al., 2004). Cells lackin g Mad1, an important factor involved in the assembly of the spindle checkpoint, also exhib it increased sensitivity to taxol, however cells treated with nocodazole became mo re resistant (Kienitz et al., 2005). Thus, identification of novel regulators of taxo l sensitivitywhich ma y alter cellular response to these drugs by increasing resistanceis required. 46

PAGE 47

Discussion and Results Compounds that affect the mitotic spindle apparatus, including mi crotubule inhibitors such as paclitaxel (taxol) and nocodazole, are known to diffe rentially affect cells th at lack one or more mitotic spindle associated proteins compared to wild type cells. To further explore the potential role(s) of Daxx in mitosis and cell division, th e microtubule inhibitor taxol, which is a microtubule hyper-stabilizing compound and the tubulin-destabilizing drug nocodazole were used to examine response of Daxx +/+ and Daxx -/MEFs. A very striking and divergent response was observed between Daxx +/+ and Daxx -/mouse fibroblasts treated with taxol and nocodazole (Figure 4-1), which was not recapitulated when cells were treated with drugs such as roscovitine, adriamycin and etoposide (dat a not shown). The number of mitotic cells in Daxx +/+ MEFs was much lower compared to Daxx -/MEFs. Conversely, the occurrenc e of micronucleated cells in Daxx +/+ mouse fibroblasts was much higher in comp arison to Daxx-deficient cells. This evidence implies a Daxx-specific function which is inhib ited when cells are treated with taxol or nocodazole. The decrease in micronuclei and elev ated mitotic index also correlated with cell survival in Daxx deficient cells, producing 90 % of colonies after 24h of taxol treatment compared to untreated control, while the survival rate of Daxx +/+ cells was only 15%-approximately 6 times lower compared to Daxx -/cells (Figure 4-1). Given the similar time of cell cycle progression for both cell lines (not shown), the difference in colony formation most likely reflects a differential survival rate of cells upon drug exposure. Given that taxol is a very pot ent chemotherapy agent used to treat breast cancers and other malignancies, the role of taxol-i nduced cell death in breast cancer cells was examined with how this may correlate with the level of protei n Daxx. High heterogeneity of Daxx protein level among breast cancer cell lines was observed (Figure 4-2, normalized on actin). Daxx is a nuclear protein with obvious ND10/PML body association in interphase cells (Ishov et al., 1999); despite 47

PAGE 48

high variety in Daxx protein level, intracellular distribution of Daxx is similar in all breast cancer cell lines tested, showing Daxx co-localizing with PML (not shown) To study a correlation between Daxx and cellular response to paclitaxel, cell lines with extreme level of Daxx were chosen: T47D (low level of Daxx, Daxx/actin = 1.0) and MDA MB 468 (high level of Daxx, Daxx/actin = 14.0) and tested fo r paclitaxel induced cell deat h measured by colony formation assay. Increased survival of T47D cells was observed compared to MDA MB 468 cells (Figure 4-2). 24h of treatment reduces the survival of MDA MB 468 cells almost three fold, and at 48h of treatment very few colonies formed. 24h treatm ent had almost no effect on T47D cell survival rate and 48h treatment reduced the number of colonies only by 30%. Thus, low level of Daxx correlates with increased resistance to paclitax el treatment in these breast cancer cell lines. To address the mechanism of cell death and di fferential survival of breast cancer cell lines T47D and MDA MB 468 upon paclitaxel treatment, morphological change s that occurred with nuclei were observed at differe nt times of drug addition, which is effective means for discriminating between apoptosis and micronucl eation. Cells were categorized based on the nuclear morphology (Figure 4-2). In mock-treated conditions (control, Figure 4-3) the majority of cells were in interphase. The rate of accumu lation in mitosis at 12h is similar for both cell lines (Figure 4-3) that reflects almost an iden tical time of cell cycle progression. Already at 12h of treatment, 27% of micronucleated cells appe ar in MDA MB 468 cells and reaches 60% and 79% at 24h and 36h correspondingly (F igure 4-3, top graph), while only small portion of cells (20 and 8%) remains blocked in mitosis (Figure 4-3, middle graph). An insignificant number of T47D cells are micronucleated at 12h and 24h, reaching only 22% by 36h of treatment; most cells remain blocked in pro-metaphase during the course of treat ment. The level of apoptosis in both cell lines was negligible, reaching a maximu m of 2% and 7% correspondingly in T47D and 48

PAGE 49

MDA MB 468 after 36h of treatment (Figure 4-3, bottom graph); thus apoptosis is not the main mechanism of cell death for these cell lines at this drug concentration. Indeed, the process of apoptosis may occur as a secondary event or be more prominent (and is sometimes observed) in larger paclitaxel concentrationsbut this may be outside of any clini cal relevance (HernandezVargas et al., 2006; Wang et al., 1999). Furthermore, upon exposure of MDA MB 468 and T47D cells to increased paclitaxel concentra tions (100 nM and greater) sizeable increase in apoptotic levels or change in m itotic index/micronucleation could not be detected (Figure 4-4). This is in contrast to some reports showing that larger paclitaxel concentrations induce a more prominent mitotic checkpoint arrest and hence a stronger mitotic block in other cell lines (Giannakakou et al., 2001; Ikui et al., 2005). Similar results we re observed by FACS analysis (Figure 4-5): the majority of T47D cells accumula te in G2/M at 24h and 36h of treatment, while MDA MB 468 produce an extensive sub-G1 popul ation. Unfortunately, FACS does not allow discrimination between apoptotic and micronucleated cells as clea rly as microscopic analysis because both types of cell death result in fragme ntation of the nucleus (recognized as a sub-G1 population by FACS and thus counted together). Therefore, microscopy was the most useful approach to determine type of cell death after paclitaxel treatm ent. Thus, a major difference in paclitaxel response between these cell lines is the ability to maintain prometaphase block which is extended in T47D, but is brief in MDA MB 468 and is followed by micronucleation. Daxx was recently shown to interact with and regulate stability of p53, one of the key players in cell growth arrest and apoptosis (Tang et al., 2006; Zhao et al., 2004). Moreover, Daxx seems to differentially regulate p53 de pendent transcription under DNA da mage conditions in transient transfection assay, thus affecting the balance between ce ll cycle arrest and a poptosis (Gostissa et al., 2004). Stability of p53 and p53-dependent tran scription regulation, howe ver, were unaffected 49

PAGE 50

in primary human fibroblasts de pleted of Daxx and seems to be influenced instead by the JNK pathway during UV and H 2 O 2 treatment (Khelifi et al., 2005 ). Both T47D and MDA-MB 468 express mutant, transcriptionally inactive p53 (Concin et al., 2003); thus, differences in paclitaxel response between these cell lin es are most likely p53-independent. To further confirm Daxx dependent cell surv ival upon paclitaxel treatment, Daxx protein levels were depleted in MDA MB 468 breast cancer cells, in HEp2 human epithelial carcinoma cells (Figure 4-6) and in hu man fibroblast cell line WI38 by st able expression of anti-Daxx siRNA (data not shown). In the case of both anti-Daxx siRNA, we observed a marked decrease and slower rate of micronucleation and an increase in mitotic index (Figure 4-6). In all cases, the levels of apoptosis throughout this ongoing process were ne gligible (data not shown). Importantly, the decline in micronucleation and in crease of mitotic index observed in anti-Daxx siRNA cell lines correlated with an increased su rvival of cells under colony formation conditions (Figure 4-6). Thus, depletion of Daxx by siRNAs targeted against Daxx message in a variety of human cell lines reproduces the original finding th at level of Daxx is critical for paclitaxel response. The combination of these data allows the pr oposition of a model in which cells follow one of two paths in response to taxol and which is dependent on Daxx protein level: Daxx positive and taxol-sensitive cells will block in mitosis only transiently, followed by micronucleation, while Daxx deficient and taxol-resistant cells have prolonged mitotic block and continue proliferation after drug decay and microtubu le dynamics restoration, thus surviving chemotherapy. This model emphasizes Daxx as a trigger for cellular taxol responseparticularly in apoptosis-reluctant cells; cells lacking a functional Da xx protein display an increased resistance to taxane exposure. 50

PAGE 51

A B 0% 20% 40% 60% 80% 100%No Treatment10 uM Noc10 nM Pac Daxx +/+ Daxx -/0% 10% 20% 30% 40% 50% 60%No Treatment10 uM Noc10 nM Pac Daxx +/+ Daxx -/C Figure 4-1. Differen tial Response of Daxx +/+ and Daxx -/MEFs to Microtubule Inhibitors Nocodazole and Paclitaxel. A) Mitotic i ndex of MEFs treated with 10 M nocodazole or 10 nM paclitaxel for 24 hrs. Cells were fixed and stained with phospho-H3 antibody to characterize mitosis. B) Corresponding percentage of micronuclei formation in cells treated with nocodazole or paclitaxel for 24 hrs. C) Immunostaining of Daxx +/+ and Daxx -/MEFs treated with paclitaxel using mitotic markers phospho-H3 and lamin. Note occurrence of mitotic cells (b ig arrowheads) in Daxx -/cells and micronuclei (small arrows) in Daxx +/+ cells. 51

PAGE 52

Figure 4-2. Colony Formation of Breast Cancer Cells After Paclitax el Treatment is DaxxDependent. A) differential Daxx expression among breast cancer cell lines. Daxx accumulation normalized by actin (bottom). B) & C): MDA MB 468 (high Daxx) and T47D (low Daxx) were treated with 10 nM paclitaxel for 24h or 48h. Colonies were fixed and stained with crystal violet and calculated 5 days after drug withdrawal. Note differential taxol response (% of survival) between MDA MB 468 and T47D. 52

PAGE 53

C B Ainterphase metaphase micronuclei apoptosis % of Micronuclei0% 1% 4% 22% 0% 27% 60% 79%0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Control 12h 24h 36h T47D MDA-MB 468 % of Mitotic Cells3% 38% 62% 59% 4% 29% 20% 8%0% 10% 20% 30% 40% 50% 60% 70% Co ntrol 12h 24h 3 6h T47D MDA-MB 468 % of Apoptotic Cells0% 0% 0% 2% 0% 3% 5% 7%0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%Con t r o l 1 2h 24h 3 6 h T47D MDA-MB 468 Figure 4-3. Response to Paclitaxel Treatment in Breast Cancer Cell Lines with Different Daxx Level. MDA MB 468 (high Daxx) and T47D (low Daxx) were treated with 10 nM paclitaxel for 12h, 24h or 36h or mock-tr eated (control). DNA was stained with HOECHST 33342. A) Cells were categorized as interphase, mitotic, micronucleated, and apoptotic based on the nuc lear morphology. B) Mitotic cel ls in control: asterisks; micronucleated cells: big arrowheads; pro-me taphase cells: arrows ; apoptotic cells: small arrowheads. While micronuclei appear in MDA MB 468 already after 12h of treatment, majority of T47D cells remain in pro-metaphase after 36h of treatment. C) Relative accumulation of mito tic cells, micronucleated a nd apoptotic cells in MDA MB 468 and T47D cells. The majority of MDA MB 468 cells execute micronucleation, while T47D are accumulate d in pro-metaphase. For each time point one thousand cells were counted. 53

PAGE 54

A B Mitotic Cells0% 10% 20% 30% 40% 50% 60% 70% 0h12h24h36h MDA MB 468 T47D Micronuclei0% 10% 20% 30% 40% 50% 60% 70% 0h12h24h36h MDA MB 468 T47D C Apoptosis0% 5% 10% 15% 20% 25%0h12h24h36h MDA MB 468 T47D Figure 4-4. Response of MDA MB 468 and T47D Breast Cancer Cell Lines to Increased Concentration of Paclitaxel. MDA MB 468 and T47D cell lines were treated with 100 nM paclitaxel for 12h, 24h and 36h or mo ck treated (control). A) DNA was stained with HOECHST 33342 and cells were characterized as being mitotic, B) micronuclei or C) apoptotic, based on nuc lear morphology. Cellular response to paclitaxel between these two cell lines remained relatively unchanged despite increased paclitaxel concentr ations. Note: apoptosis does not significantly increase with elevated concentrations of paclitaxel. 54

PAGE 55

A B T47D0% 20% 40% 60% 80% 100%c o ntro l 12 h 24 h 36 h G1 S G2/M Ap/Mic MDA-MB 4680% 20% 40% 60% 80% 100%control 12 h 24h 3 6h G1 S G2/M Ap/Mic Figure 4-5. FACS Analysis of Ce ll Cycle Distribution af ter Paclitaxel Treatment. A) T47D and B) MDA MB 468 cells were treated with 10nM paclitaxel for the indicated time and sorted as G1 phase, S phase, G2/M phase, and apoptotic + micronucleated (Ap/Mic). While the majority of T47D cells accumu late in G2/M after 36h of treatment, MDAMB 468 cells are mostly apoptotic + micronucle ated at this time point. Unfortunately, FACS does not allow discrimination between apoptotic and micronucleated cells. 55

PAGE 56

A B C D Figure 4-6. Paclitaxel Response is Daxx-Dependent. A) Western blot analysis of Daxx depletion in MDA MB 468 and HEp2 cells. B) Mitotic index of parental MDA MB 468 (left) and HEp2 cells (right) expressing two independe nt anti-Daxx siRNAs or control siRNA which were synchronized via double th ymidine block and then released and treated with 10 nM paclitax el for 12, 18 and 24 hrs. C) Same as (B) except micronucleated cells are shown. D) Reduc tion of Daxx increases cell survival during paclitaxel treatment. Cells treated in (B ) and (C) were set up for colony formation assay and allowed to grow for 5 days. Col onies were then counted after fixation and staining with crystal violet. Apoptosis, as a result of pacl itaxel exposure, was negligible across all of these diffe rent cell lines (d ata not shown). 56

PAGE 57

CHAPTER 5 DAXX INTERACTS WITH RAS-ASSOCIATI ON DOMAIN FAMILY 1 (RASSF1) WHICH COOPERATE IN CELLULAR TAXOL RESPONSE Introduction Several lines of evidence suggest Daxx is important for proper timing of mitotic progression and that paclitaxel (taxol) resistance can be dependent on the level of Daxx protein in cells. Firstly, cells generated from Daxx-defi cient embryos exhibit genomic instability while tissue sectioning of E9.5 Daxx-deficient mouse em bryos shows an accumulation of cells in early mitosis (specifically pro-metaphase). Targeted depletion of Daxx also results in alteration of mitotic progression during prophase and the pr o-metaphase to anaphase transitions. Additionally, proper degradation of Cyclin B in synchronized Daxx-depleted cells is altered. Importantly, cells that are deficient or depleted of Daxx was proven to provide resistance to the microtubule inhibiting drug taxol by arresting cells in mitosis; cells with a functional Daxx protein have a shortened mitotic arrest resulting in mi cronucleation and cell d eath (Lindsay et al., 2007). Discussion and Results To further investigate Daxx m itotic function and to study the potential mechanism by which Daxx regulates taxol response during mitosis a s earch for Daxx-interacting proteins using Daxx as bait in a yeast two-hybrid screen was performed. Several yeast two-hybrid screens have been used in the past using Daxx as both bait and prey. The majority of these screens have used, or have identified through their sc reens, the carboxyl-terminal region of Daxx (corresponding with PMLand Fas-interacting doma ins) which also contains a region of Daxx 57

PAGE 58

known to interact with sumoylat ed proteins (Pluta et al., 1998 ; Yang et al., 1997). Since yeast themselves frequently use sumoylation for prot ein-protein interactions this can potentially explain the frequent occurrence of Daxx-fals e positives from yeast two hybrid screens (Michaelson, 2000). For this analysis, the ami no-terminal region of Daxx was used which has previously not been adapted for yeast two-hybrid screens. This novel approach identified several proteins with functional implications in the re gulation of cell division and mitosis. Sequence analysis of two particularly st rong interaction clones revealed homology with amino acids 5-270 and 30-270 of mouse RAS associated domain family 1 splice form C (Rassf1C) (Figure 5-1). Retransformation and -gal reporter assay confirmed the specifi city of this interaction in yeast, pointing at the amino -terminus of Daxx as a poten tial region of Rassf1 in teraction (Figure 5-1). DMAP1 (DNA methyltransferase associated protein 1) that was recently shown to interact with Daxx (Muromoto et al., 2004) was us ed as positive control in these experiments. Rassf1 is a highly conserved throughout specie s and its locus is frequently altered in human cancer. Among several alternative splice form s, Rassf1A and Rassf1C are the most abundantly expressed. Both Rassf1A and Rassf1C are tubulin-associated protei ns which influence the overall stability and dynamics of microtubules (Liu et al., 2003). Wh ile Rassf1C function is relatively unknown, Rassf1A has recently been shown to regulate early mitosis progression, particularly during prometaphase (Song et al., 2004). In addition, Rassf1A over-expres sion leads to mitotic arrest and inhibition of colony growth (Liu et al., 2003). Daxx -/embryos showed an increased accumulation of cells in pro-metaphase, suggesting that Daxx could cooperate with Rassf1A functi on during pro-metaphase. In light of the known properties of both Rassf1A and Rassf1C on the in fluence of microtubules, the potential interplay of Daxx and Rassf1 binding could also partially explain alterations in sensitivity of Daxx58

PAGE 59

depleted cells to taxol, a known microtubule-interfering compound. Interaction mapping between Daxx, Rassf1C and Rassf1A and deletion mutants was analyzed with proteins in vitro using a pull-down assay and by co-l ocalization of GFP and RFP fusi on proteins in cells (Figure 5-2). Interaction between Daxx and Rassf1 was tested by incubation of immobilized 6His-fusion Rassf1C with GST-Daxx and GST alone (Figure 5-2). Rassf1C associat es only with GST-Daxx and does not associate with GST control that in dicates specificity of Rassf1C-Daxx interaction. All truncation mutant s of Daxx, including first 142 amino acids of protein, were co-purified with Rassf1C wt suggesting that region of Rassf1 interaction is located at the amino terminus of the Daxx protein, thus confirming yeast two-hybrid da ta (Figure 5-2). Co-localization was observed between Rassf1C-RFP and GFP-fused Daxx wt a nd Daxx amino terminal deletion mutants but not with a carboxyl terminus mutant of Daxx (amino acids 625-740), further pointing at the amino terminus of Daxx as the minimal Rassf1 interacting domain. GFP-Daxx does not co-localize w ith RFP-Rassf1A wt or with a Rassf1A mutant that covers aa 120-340 which is homologous between Rassf1A and Rassf1C, but GFP-Daxx co-localizes with the first 50 amino acids of RFP-Rassf1C that are unique for this splice variant. Moreover, the first 142 amino acids of Daxx can also co -localize with this domain of Rassf1C. In conclusion, the interaction between Rassf1 and Da xx is facilitated by the first 50 unique amino acids of Rassf1C; in combination with a minimal clone purified in a y east two-hybrid analysis (which encodes amino acids 30 to 270 of Rassf 1C) the potential minimal region of interaction can likely be narrowed down to amino acids 30-50 of Rassf1C. Therefore, Daxx interacts with Rassf1C, but not Rassf1A. Considering these data, it was hypothesized that Daxx may coope rate with Rassf1A functioning in mitosis via an interaction with Ra ssf1C. This model assumes that two alternative 59

PAGE 60

spliced proteins, Rassf1A and Rassf1C, can interact with each other and have similar localization during mitosis or throughout the cell cycle. It was found that GST-Rassf1A could bind 6HisRassf1C in a pull-down assay (Figure 5-2B). Fu rthermore, a Rassf1A mutant (aa 194-340) could still efficiently bind Rassf1C (Figure 5-2B). A dditional evidence of inte raction was obtained by co-localization of GFPand R FP-fused Rassf1A/Rassf1C. Co-loc alization between Rassf1C and amino acids 194-258 of Rassf1A indicates smallest region of interact ion. Fluorescence timelapse microscopy of HEp2 cells expressing GFP-Ra ssf1A wt or GFP-Rassf1C wt also revealed similar localization throughout the cell cycle implying a potentia lly constant association of Rassf1A-Rassf1C proteins (dat a not shown). Thus, the minimal amino acids necessary for Rassf1A-Rassf1C interaction lik ely resides within the common Ras-Association (RA) domain shared by both splice forms of Rassf1, which conf irms hetero-dimer formation of Rassf1A and Rassf1C and also opens the possibility of homo-d imer formation between each of these isoforms. To confirm the endogenous interaction of Daxx and Ra ssf1, double immunofluorescent staining of Daxx and Rassf1C was performed in HEp2 cells where these two proteins were found to be distinctly separated during interphase (Figure 5-3) with Daxx staining relegated strictly to the nucleus (at PML bodies) and Rassf1C staining localized strictly to a microtubule-network pattern. This is in contrast to some reports suggesting Rassf1C is a nuclear protein that may interact with Daxx at PML (Kitaga wa et al., 2006). To confirm the differential localization of Daxx and Rassf1 in cells, biochemical separation of HEp2 cells into nuclear and cytosolic fractions was performed as well as 3D-confocal imaging of transiently over-expressed GFPRassf1A, GFP-Rassf1C and GFPDaxx (Figure 5-3 B & C). Daxx was found to be largely a nuclear associated protein while Rassf1A app eared strictly in the cytosolic fraction by biochemical separation. Three-dimensional c onfocal analysis of Daxx and Rassf1 cellular 60

PAGE 61

localization showed Daxx to be a strictly nuclear associated protein, as expected. In stark contrast, both GFP-Rassf1A and GFP-Rassf1C displayed cytoplasmic distribution in a microtubule-like network with no GFP-fluorescen ce emanating from the nucleus. Thus, Daxx and Rassf1 are compartmentally separated protei ns during interphase. Co-localization of endogenous Daxx and Rassf1, however, was obs erved in HEp2 cells progressing through mitosisdemonstrating the cell cycl e regulated interaction between these proteins (Figure 5-4). Co-localization of Daxx and Rassf1 was observe d beginning in pro-metaphase and metaphase, but by later stages of mitosis this association could not be detected. To understand the cell cycle regulated intera ction of Daxx and Rassf1 during mitosis and cellular paclitaxel response, stable expressing Rassf1A-siRNA was introduced into HEp2 cells which efficiently deplete Rassf1A protein level (Figure 5-5). While Ra ssf1A-depleted cells could be easily generated, over-exp ression of Rassf1A leads to cell to xicity and inhibition of cell proliferation independent of Rassf1A-functioning in cells. Therefore these studies focused primarily on protein-depleted cel l lines. In combination with control and anti-Daxx siRNA HEp2 lines, anti-Rassf1A siRNA HEp2 cells were expo sed to 10 nM paclitaxel for 6-18 hrs and then replated for colony formation a ssay (Figure 5-5). Strikingly, in the case of both Daxxand Rassf1A-depleted cells, paclitaxel resistance was similar. Both cell lines exhibited a strong paclitaxel resistant phenotype with the majority of treated cells (75-80%) capable of dividing and forming colonies after removal of taxol. To gain a biochemical unders tanding of how mitotic cells respond to paclitaxel in the absence of Daxx or Rassf1A, control-, Daxxand Rassf1Adepleted cells were synchronized using a double t hymidine block and then released and exposed to taxol for 6-18 hrs and collected for Western-blot evaluation of Cyclin B protein levels (Figure 5-6). Wild type (parental) and control-siRNA cells revealed an accumulation of Cyclin B as 61

PAGE 62

cells entered mitosis, but a marked drop-off in Cyclin B protein level ensued indicating cell exit from mitosis in the later stages of paclitaxel treatment. In contrast, Daxxand Rassf1A-siRNA cells showed a similar accumulation of Cyclin B but these levels were maintained throughout the course of the experiment, indicating cells were still arrested in mitosis. Thus, Daxx and Rassf1A are necessary for efficient cellular response to paclitaxel which in cludes entry into and exit from mitosis during treatment. Sustained Cyclin B protein leve ls in response to mitotic st resses like paclitaxel is an indication of prolonged spindle checkpoint ac tivation (Musacchio and Salmon, 2007). During normal cellular response to paclitaxel, cells will tran siently arrest in mitosis due to an activated spindle checkpoint but will exit mitosis by degradi ng mitotic substrates (i.e. Cyclin B) because taxol-generated errors (i.e. microtubular tensi on, unattached kinetechores ) cannot be corrected. In the absence of Daxx or Rassf1A, cells remain in a prolonged mitotic block as evidenced by sustained Cyclin B protein leve ls. Many different regulatory proteins are involved in proper spindle checkpoint operati on, including the Aurora kinases, a fa mily of serine/threonine kinases that are highly conserved phylogenetically. Specifically, Aurora A and Aurora B are involved in the proper placement and localizatio n of key mitotic checkpoint prot eins (Ditchfield et al., 2003) and absence or depletion of Aurora kinases cau ses spindle checkpoint-ove rride (Fu et al., 2007). Therapeutically, it would be advantageous to targ et Aurora kinases in tumors because Aurora A and Aurora B are frequently up-regulated in cancers (Keen and Taylor, 2004). As a result, several Aurora kinase inhibitors are in phase I & II clinical trials to eval uate their efficacy as chemotherapeutic agents (Agnese et al., 2007; K een and Taylor, 2004). Clinical strategies for enhancing paclitaxel response ar e continuously being studied a nd developed and one potential method involves the use of taxol in combination with Aurora kinase i nhibitors (Malumbres, 62

PAGE 63

2006). Current compounds, including ZM447439, hesp eradin and VX680, have been engineered to target the ATP binding site of Aurora kina ses which abolishes thei r kinase activity. In addition to paclitaxel, using these compounds has b een shown in cell-based assays to alter taxol response, even when spindle checkpoint proteins were absent (Morrow et al., 2005). Thus, the potentials of abrogating taxol resistance in comb ination with other compounds which target the mitotic spindle checkpoint are promising. In order to determine if Daxxand Rassf1A -mediated taxol response can be altered by inhibition of Aurora kinase activity, control, Daxxand Rassf1A-depleted HEp2 cells were treated with taxol in combination with two indepe ndent Aurora kinase inhibitors (Figure 5-7). Compounds used in this study were ZM447439, whic h targets Aurora A and Aurora B kinase activity and Aurora kinase inhi bitor III, which targets Aurora A kinase activity. After synchronization with a double thymidine block and a six hour release, cont rol and experimental cell lines were exposed to pacl itaxel alone or paclitaxel in combination with ZM447439 or Aurora kinase inhibitor III for a period of six hours. After completion of drug exposure, cells were then replated for colony formation assay. Treatment of Daxxor Rassf1A-depleted HEp2 cells with paclitaxel alone typica lly resulted in a very robust taxol resistance, as evidenced by the 75-80% survival rate of these cell lines compared to only 46% of control cells (Figure 5-7). Strikingly, however, taxol resistance was abo lished in Daxxor Rassf1A-depleted HEp2 cells when treated in combination with ZM447439 or Aurora kinase inhibito r III, resulting in comparable cell survival with control cells (Daxx-siRNA 31%-42%, Rassf1A siRNA 38%-40% and control siRNA 29%-37%). Thus, in the abse nce of Daxx or Rassf1A, functional inactivation of the mitotic spindle checkpoint using Aurora kinase inhibitors can change cellular taxol response and abolish resistance. 63

PAGE 64

Figure 5-1. Daxx Interacts with Tumor Suppressor Ra ssf1 in Yeast. A) Schematics of Daxx and Rassf1 isoforms Rassf1A and Rassf1C along with mapping of clones identified through yeast two hybrid screening (Rassf 1C clone 1 corresponding to amino acids (aa) 5-270 of mouse Rassf1C, Rassf1C clone 2 corresponding to aa 30-270). Homology between Rassf1A and Rassf1C is also shown. B) Retransformation assay of Daxx and Rassf1C constructs. 1 = Ra ssf1C + pGBDC1 (empty vector), 2 = Rassf1C + Daxx wt, 3 = Rassf1C + Daxx C term, 4 = Rassf1C + Daxx C. C) Gal reporter assay measuring strength of inte raction between individual Rassf1 clones and Daxx wt and deletion mutants. Strong interaction between Rassf1 clone 2 is observed. DMAP (positive control for Daxx in teraction) was used as an evaluation for strength of interaction in this system. 64

PAGE 65

C Figure 5-1. Continued. 65

PAGE 66

A B (i) (ii) (iii) C Figure 5-2. Mapping of Daxx-Rassf1 Interaction. A) Diagram of human Daxx, Rassf1A and Rassf1C constructs used for mapping inte raction by co-localization (GFP and RFP) and in vitro pull down assay (GST and 6His). B) ( i) 6xHis pull down assay of GSTDaxx wt and 6xHis-Rassf1C wt. Immobilized 6xHis-Rassf1C wt was incubated with either GST or GST-Daxx wt. GST-Daxx wt but not GST alone binds 6xHis-Rassf1C wt. (ii) In similar experimental settings, all Daxx amino terminal constructs including aa 1-142 retain capacity to bind 6His-Rassf1C wt. (iii) 6His pull down assay of GSTRassf1A wt and mutants using immobilized 6His-Rassf1C wt. GST-Rassf1A wt/ mutants, GST-Daxx 1-142 and GST alone we re incubated with 6His-Rassf1C wt. GST-Daxx 1-142, GST-Rassf1A wt and mutants 120-340 and 194-340 bind to Rassf1C, while GST does not. C) Table summarizing interaction ( +, or ND for not determined) tested by co-localiza tion of GFP and RFP fusions or in vitro pull down assay (right in cell). 66

PAGE 67

A B C Figure 5-3. Cellular Distribution of Daxx and Rassf1 During Interp hase. A) Immunostaining of endogenous Rassf1C and Daxx in interphase HEp2 cells. B) Biochemical separation of nuclear and cytosolic fractions from HE p2 cells. Note nuclear association of Daxx and cytosolic association of Rassf1A. C) 3D confocal imaging of transiently overexpressed GFP-Rassf1A (far left), GFP-Rassf 1C (middle) and GFPDaxx (far right). Note the compartmentally separa ted expression of Daxx and Rassf1. 67

PAGE 68

A B C d Figure 5-4. Co-localization of Endogenous Da xx and Rassf1 During Mitosis in HEp2 Cells. HEp2 cells were immunostained with monoclonal Daxx 514 antibody and polyclonal Rassf1 antibody which detects both endogenous Rassf1A and Rassf1C. Rassf1 an Daxx co-localized during pro-metaphase and metaphase A) & B). C) By anaphase and later stages of mitosis, this association is absent. 68

PAGE 69

A B 0% 25% 50% 75% 10 0% HEp2 (parental) Control siRNA Daxx siRNARassf1A siRNA No Treatment 6 hr Taxol 12 hr Taxol 18 hr Taxol Figure 5-5. Depletion of Daxx or Ra ssf1A Desensitizes Cells to Pacl itaxel. A) Stable expression of anti-Rassf1A siRNA in HEp2 cells. Note depletion of Rassf1A protein compared e to parental cell lines. B) Pe rcentage of colonies formed from parental, control, Daxxand Rassf1A-depleted HEp2 cells which were synchronized using a double thymidin block and then released and exposed to p aclitaxel for the indicated time periods (6, 12, 18hrs). After treatment, cells were replated for colony formation assay. Note increased survival (paclitaxel resistan ce) of Daxxand Rassf1A-depleted cells compared to control a nd parental cell lines. 69

PAGE 70

Figure 5-6.Cyclin B Levels are Stabilized in Daxxand Rassf1A-Depleted Cells Treated with Taxol. Wild type (parental) HEp2, control and Daxxand Rassf1A-depleted cell lines were synchronized with a double thymidine bl ock, released and then exposed to tax for 6-18hrs. Cells were harvested at the indicated time points and probed for Cyclin B1 levels using anti-Cyclin B1 antibody (Santa Cruz). Protein levels were normalized within each siRNA cell line. Note increased relative stability of cyclin B1 (normalized to actin) in both Da xx and Rassf1A-depleted cell lines. ol 70

PAGE 71

0% 20% 40% 60% 80% 100% Daxx siRNARassf1A siRNAControl siRNA Taxol Taxol + ZM Taxol + AKIII Figure 5-7. Inactivation of the Mitotic Spindle Checkpoint Using Aurora Kinase Inhibitors Abolishes Taxol Resistance in Daxx-and Ra ssf1A-Depleted Cells. Percentage of tor colonies that were formed from control, Daxxand Rassf1A-depleted HEp2 cells which were synchronized using a double t hymidine block, released and exposed to taxol alone, or taxol in combination with ZM447439 (ZM) or Aurora kinase inhibi III (AKIII) for six hours is shown. Note taxo l resistance is abolished in Daxxand Rassf1A-depleted cells treated in combin ation with taxol and Aurora kinase inhibitors. 71

PAGE 72

CHAPTER 6 SUMMARY AND CONCLUSIONS Taxane chemotherapy is c ent options for many breast can e been et lass ug ied as a novel regu lator of paclitaxel res ponse in cell culture con x onsidered among the mo st active treatm cer patients, either alone or as adjuvant in combination with anthracyclins (O'Shaughnessy, 2005). Nevertheless, a large number of patients are resistant to ta xanes or become resistant to this therapy during treatment. Th e response rate of docetaxel is ~50% even after the first-line chemotherapy administration and it decreases to 20-30% in the secondor third-line administration (Bonneterre et al., 1999; Crown et al., 2004). A number of studies hav carried out to determine a genomic profile that co uld be predictive to taxane treatment (Chang al., 2003; Chang et al., 2005b; Iwao-Koizumi et al ., 2005; Mauriac et al., 2005; Miyoshi et al., 2004). However, an alternative approach to understand selective resistance to taxane treatment is to study mechanisms by which cells can resp ond to these drugs. Several molecular targets were reported, starting with mutations in and -tubulin that affect drug binding, increased expression of tubulin genes, and changes in th e synthesis or activity of tubulin interacting proteins (Hari et al., 2003a; Ha ri et al., 2003b; Wang and Cabral 2005). Recently, a new c of potential targets are being studi ed after it was suggested that in activation of m itotic proteins can contribute to the selective response of taxane treatment in vivo (Wassmann and Benezra, 2001). Thus, development of new genomic prognosis factors and in-depth understanding of dr activity on both a cellular and organismal level are needed for optimization of adjuvant therapy and proper patient stratification. To this end, Daxx was identif ditions. Primary mouse and hum an fibroblasts with experiment ally regulated levels of Dax show both strong and divergent responses to taxol (Figure 4-1). In the absence of Daxx, cells remain in a prolonged mitotic block, while wild t ype cells undergo a transi ent arrest in mitosis 72

PAGE 73

that is soon followed by micronucleation and cell death. Importantly, these observations were also recapitulated in MDA MB 468 and T47D breas t cancer cell lines with contrasting levels of Daxx expression (Figure 4-3). These divergent responses in taxol response are usually seen in cells deficient in mitotic checkpoint proteins or other regulators of cell division and until these current studies, Daxx has not been implicated as a regulatory pr otein of cell cycle progression. Table 6-1 summarizes a growing lis t of mitotic proteins and the cellular response to taxol when these proteins are absent or deregulated in cells. To date, loss of function of the majority of mitotic proteins, including Mad1, Mad2, Bub1 and BubR1, has shown enhanced response to paclitaxel in cell culture conditi ons. Identification of factors which may increase drug resistan are largely uncharacterized. Among the conclu sions of this study is that early mitosis progression is altered in cells lacking a functional Daxx protein. These observations ca tissue sectioning of Daxx ce me from ing re cycle. Dur in. To -/mouse embryos showing an incr eased number of cells in early mitotic stages (Figure 3-1) and were manifest ed from time-lapse microscopy studies analyz mitotic progression in Daxx-depleted cell lines (Figure 3-3, Table 3-1) and Cyclin B protein stability studies showing an alte red rate of cyclin degradation in anti-Daxx siRNA cells (Figu 3-4). Theoretically, alterations of this kind in cell cycle progression coul d, in part, explain the genomic instability that is observed in Daxx-de ficient mouse cells (Figure 3-1) as improper mitosis frequently leads to unfaithful chro mosome segregation (Chi and Jeang, 2007). Daxx localization is already unde rstood to be a very dynamic process during the cell ing G1 and G2, Daxx localizes to PML bodies, while during S phase, it relocates to condensed heterochromatin where it interacts w ith ATRX, a chromatin remodeling prote accompany this dynamic protein trafficking, these st udies have also revealed that Daxx, a known transcriptional regulator, can associate with th e mitotic spindle apparatu s during mitosis (Figure 73

PAGE 74

3-7). Accruing evidence suggests that the tight orchestration of events during mitosis combines seemingly unrelated factors at cr itical junctures during cellular division (Tsai et al., 2006). A number of essential interac tions occur during pro-metaphase when the nuclear envelope disintegrates and no longer compartmentalizes the nucleus and cytoplasm. During this tim Daxx is released from the nucleus where it is associated with ND10/PML bodies and accumulates at the mitotic spindle apparatus. It is at this spatiotemporal point that Dax with Rassf1 (Figure 5-4, Figure 6-1 summary). Future studies will reveal whether Daxx association with the spi ndle apparatus is dependent on presence of Rassf1. To accompany observations of Daxx-dependent p aclitaxel response, Ras e, x interacts sf1 was identified as a no ) and al s hase are s vel Daxx-interacting protein th at was also confirmed to be important for cellular taxol response (Figure 5-5). This interaction was ma pped to the C spliceform of Rassf1 (Rassf1C interaction between Rassf1A and Rassf1C by dime rization was also confirmed (Figure 5-2). These novel interactions are thought to form a complex during mitosis and may perform critic regulatory processes including prope r cellular response to mitotic stresses such as paclitaxel, nocodazole and other microtubule inhibiting compounds. These chemicals, moreover, can induce mitotic stress in several ways, depending on the stage of mitosis in which the stress i applied. Proper separation of the centroso mes during prophase, correct alignment of chromosomes during metaphase and faithful segregation of chromosomes during anap all processes which can be altered or affected when external stress (i.e. paclitaxel) is applied. Together, Daxx and Rassf1 define a unique mitotic stress checkpoint during pro-metaphase (Figure 6-2, summary). Cells lacking Daxx or Rassf1A arrest in pro-metaphase during taxol treatment. From time-lapse studies, it is sugge sted that Daxx-depleted cells have chromosome that remain unable to properly align at the me taphase plate under treatment conditions, thus 74

PAGE 75

remaining in pro-metaphase by definition. In c ontrast, with normal primary cells and some tumor cell lines, Daxx and Rassf1A may be essen tial for proper mitotic exit in response to uncorrectable errors during prometaphase. This is suggested by the very robust response o wild type cells which only tran siently arrest in pro-metaphase and then undergo micronucleati leading to cell death (Figure 4-3, Figure 4-6). One example of a mitotic stress checkpoint is already known with a protein named checkpoint with FHA and ring finger (CHFR). In the presence of CHFR, wild type ce lls exhibit a transient prophase arrest which temporarily prev cells from entering metaphase under mitotic stress (Scolnick and Halazonetis, 2000). Cells that do not have a functional CHFR protein were shown to enter metaphase without delay and exhibited problems in proper centrosome separa tion. Thus, CHFR defines a prophase-spec mitotic stress checkpoint at a stage in mitosis earlier than Daxx and Rassf1. Evidence accumulated in these studies suggest s Daxx and Rassf1 are trig f on ents ific gers for cellular taxo In an 5). l response. In the future Daxx and Rassf1 may serve as ideal molecular markers for the proper selection of breas t cancer patients (and other malignancies) for taxane chemotherapy. order to achieve this goal, clin ical studies will be required examining the status of Daxx and Rassf1 expression in tumors before and after taxane chemotherapy as well as in patients with established history of taxane re sistance. Altered Daxx or Rassf 1 expression may be reminiscent of the differential protein expression that may ex ist in the original tumor cells from which each cell line was derived. It is alr eady known that Daxx expression in some breast cancer cell lines is quite variable (Figure 4-1), but the extent of Daxx down-regul ation or mutation in tumor cell lines has not been addressed. Rassf1A expressi on in tumor cell lines, conversely, has been extensively studied and shown to be altered in a majority of cas es (Agathanggelou et al., 200 Ultimately, these studies have established new ro les for Daxx as a mitotic regulator that also 75

PAGE 76

serves as a trigger for cellular taxol response in combination with Rassf1 which adds to our understanding of mechanisms linking cell divi sion, genome instability and breast cancer progression. Table 6-1. Alteration of Several Known Mitoti c Proteins and Resultant Cellular Paclitaxel Response. Most mitotic proteins, when mu tated or down-regulated in cells, display y increased sensitivity to paclitaxel. In the absence of Daxx or Rassf1, cells displa increased drug resistance. Protein Paclitaxel Response Bub1 Inc reased Sensitivity (Lee et al., 2004; Sud o et al., 2004) n BubR1 Increased Sensitivity (Lee et al., 2004; Sudo et al., 2004) CHFR Increased Sensitivity (Satoh et al., 2003) Mad1 Increased Sensitivity (Kienitz et al., 2005) Mad2 Increased Sensitivity (Niikura et al., 2007) Survivi Increased Sensitivity (Carvalho et al., 2003) Daxx Increased Resistance (Lindsay et al., 2007) Rassf1 Increased Resistance 76

PAGE 77

Figure 6-1. Dynamics of Daxx-Rassf 1 Interaction Throughout the Cell Cycle. During interphase and beginning of mitosis (prophase), Daxx and Rassf1 are compartmentally separated in the cytoplasm and nucleus. After nuc lear envelope (NE) breakdown, Daxx and Rassf1 can interact throu gh Rassf1C. This associa tion is maintained through metaphase, but by late stages of mitosis, is absent. 77

PAGE 78

Figure 6-2. Model Depicting Daxx-Rassf1-Medi ated Mitotic Stress Checkpoint During Prometaphase. In wild type cells upon mitoti c stress (i.e. taxol or nocodazole) the spindle checkpoint will be act ivated. In rare cases, the cell may correct these errors, complete mitosis and resume cell prolifera tion. In most cases, however, these errors are uncorrected and the cell aborts mitosi s, undergoes micronuclei formation and cell death. In subpopulations of cancer cells which encounter mitotic stress and which do not have a functional Daxx or Rassf1 prot ein, these cells remain in a prolonged mitotic block, unable to efficiently abort m itosis, and remain in pro-metaphase until drug removal or decay. This model may partially explain inherent and acquired taxol resistance in some breast tumors. 78

PAGE 79

REFERENCES Aapro, M.S. (2001). Neoadjuvant th erapy in breast cancer: can we define its role? Oncologist 6 Suppl 3, 36-39. Agathanggelou, A., Cooper, W.N., and Latif, F. (2005). Role of the Ras-association domain family 1 tumor suppressor gene in human cancers. Cancer Res 65, 3497-3508. Agnese, V., Bazan, V., Fiorentino, F.P., Fanale, D., Badalamenti, G., Colucci, G., Adamo, V., Santini, D., and Russo, A. (2007). The role of Au rora-A inhibitors in cancer therapy. Ann Oncol 18 Suppl 6, vi47-52. Amaar, Y.G., Minera, M.G., Hatran, L.K., Str ong, D.D., Mohan, S., and Reeves, M.E. (2006). Ras association domain family 1C protein stimul ates human lung cancer cell proliferation. Am J Physiol Lung Cell Mol Physiol 291, L1185-1190. Bonneterre, J., Spielman, M., Guastalla, J.P., Marty, M., Viens, P., Chollet, P., Roche, H., Fumoleau, P., Mauriac, L., Bourgeois, H. et al. (1999). Efficacy and safety of docetaxel (Taxotere) in heavily pretreated advanced brea st cancer patients: the French compassionate use programme experience. Eur J Cancer 35, 1431-1439. Bothos, J., Summers, M.K., Venere, M., Scolnic k, D.M., and Halazonetis, T.D. (2003). The Chfr mitotic checkpoint protein functions with Ub c13-Mms2 to form Lys63-linked polyubiquitin chains. Oncogene 22, 7101-7107. Brown, J.M., and Wouters, B.G. (1999). Apoptosis p53, and tumor cell sensitivity to anticancer agents. Cancer Res 59, 1391-1399. Burbee, D.G., Forgacs, E., Zochbauer-Muller, S., Shivakumar, L., Fong, K., Gao, B., Randle, D., Kondo, M., Virmani, A., Bader, S. et al. (2001). Epigenetic inac tivation of RASSF1A in lung and breast cancers and malignant phe notype suppression. J Natl Cancer Inst 93, 691-699. Cahill, D.P., Lengauer, C., Yu, J., Riggins, G.J ., Willson, J.K., Markowitz, S.D., Kinzler, K.W., and Vogelstein, B. (1998). Mutations of mito tic checkpoint genes in human cancers. Nature 392 300-303. Carvalho, A., Carmena, M., Sambade, C., Earnsh aw, W.C., and Wheatley, S.P. (2003). Survivin is required for stable checkpoint activati on in taxol-treated HeLa cells. J Cell Sci 116 29872998. Chabalier, C., Lamare, C., Racca, C., Privat, M ., Valette, A., and Larminat, F. (2006). BRCA1 downregulation leads to premat ure inactivation of spindle ch eckpoint and confers paclitaxel resistance. Cell Cycle 5, 1001-1007. Chan, G.K., and Yen, T.J. (2003). The mitotic checkpoint: a signaling pathway that allows a single unattached kinetoc hore to inhibit mitotic exit. Prog Cell Cycle Res 5, 431-439. 79

PAGE 80

Chang, C.C., Lin, D.Y., Fang, H.I., Chen, R.H., and Shih, H.M. (2005a). Daxx mediates the small ubiquitin-like modifier-dependent transc riptional repression of Smad4. J Biol Chem 280, 10164-10173. Chang, J.C., Wooten, E.C., Tsimelzon, A., Hils enbeck, S.G., Gutierrez, M.C., Elledge, R., Mohsin, S., Osborne, C.K., Chamness, G.C., A llred, D.C., and O'Connell, P. (2003). Gene expression profiling for the prediction of therapeutic response to docetaxel in patients with breast cancer. Lancet 362, 362-369. Chang, J.C., Wooten, E.C., Tsimelzon, A., Hils enbeck, S.G., Gutierrez, M.C., Tham, Y.L., Kalidas, M., Elledge, R., Mohsin, S., Osborne, C.K. et al. (2005b). Patterns of resistance and incomplete response to docetaxel by gene expressi on profiling in breast cancer patients. J Clin Oncol 23, 1169-1177. Chaturvedi, P., Sudakin, V., Bobiak, M.L., Fisher P.W., Mattern, M.R., Jablonski, S.A., Hurle, M.R., Zhu, Y., Yen, T.J., and Zhou, B.B. (2002). Chfr regulates a mitotic stress pathway through its RING-finger domain with ubiquiti n ligase activity. Cancer Res 62, 1797-1801. Chen, E.S., Saitoh, S., Yanagida, M., and Takaha shi, K. (2003). A cell cycle-regulated GATA factor promotes centromeric localization of CENP-A in fission yeast. Mol Cell 11, 175-187. Chi, Y.H., and Jeang, K.T. (2007). Aneuploidy and cancer. J Cell Biochem 102, 531-538. Cleveland, D.W., Mao, Y., and Sullivan, K.F. (2003). Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell 112 407-421. Concin, N., Zeillinger, C., Tong, D., Stimpf l, M., Konig, M., Printz, D., Stonek, F., Schneeberger, C., Hefler, L., Kainz, C. et al. (2003). Comparison of p53 mutational status with mRNA and protein expression in a panel of 24 human breast carcinoma cell lines. Breast Cancer Res Treat 79 37-46. Crown, J., O'Leary, M., and Ooi, W.S. (2004). Docetax el and paclitaxel in th e treatment of breast cancer: a review of clinical experience. Oncologist 9 Suppl 2 24-32. Croxton, R., Puto, L.A., de Belle, I., Thomas, M., Torii, S., Hanaii, F., Cuddy, M., and Reed, J.C. (2006). Daxx represses expression of a subset of antiapoptotic genes regu lated by nuclear factorkappaB. Cancer Res 66, 9026-9035. Dallol, A., Agathanggelou, A., Fenton, S.L., Ahmed-Choudhury, J., Hesson, L., Vos, M.D., Clark, G.J., Downward, J., Maher, E.R., a nd Latif, F. (2004). RASSF1A interacts with microtubule-associated prot eins and modulates microt ubule dynamics. Cancer Res 64 41124116. Dammann, R., Li, C., Yoon, J.H., Chin, P.L., Bate s, S., and Pfeifer, G. P. (2000). Epigenetic inactivation of a RAS associati on domain family protein from the lung tumour suppressor locus 3p21.3. Nat Genet 25, 315-319. 80

PAGE 81

Dammann, R., Yang, G., and Pfeife r, G.P. (2001). Hypermethylation of the cpG island of Ras association domain family 1A (RASSF1A), a putative tumor suppressor gene from the 3p21.3 locus, occurs in a large percentage of human breast cancers. Cancer Res 61 3105-3109. Dellaire, G., Eskiw, C.H., Dehghani, H., Ching, R.W., and Bazett-Jones, D.P. (2006). Mitotic accumulations of PML protein contribute to the re -establishment of PML nuclear bodies in G1. J Cell Sci 119 1034-1042. Ditchfield, C., Johnson, V.L., Tighe, A., Ellston, R., Haworth, C., Johnson, T., Mortlock, A., Keen, N., and Taylor, S.S. (2003). Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 161, 267-280. Donninger, H., Vos, M.D., and Clark, G.J. ( 2007). The RASSF1A tumor suppressor. J Cell Sci 120, 3163-3172. Ecsedy, J.A., Michaelson, J.S., and Leder, P. (2003). Homeodomain-interac ting protein kinase 1 modulates Daxx localization, phosphorylation, an d transcriptional activity. Mol Cell Biol 23, 950-960. Emelyanov, A.V., Kovac, C.R., Sepulveda, M.A., a nd Birshtein, B.K. (2002). The interaction of Pax5 (BSAP) with Daxx can resu lt in transcriptional activation in B cells. J Biol Chem 277, 11156-11164. Fenton, S.L., Dallol, A., Agathanggelou, A., Hesson, L., Ahmed-Choudhury, J., Baksh, S., Sardet, C., Dammann, R., Minna, J.D., Downward, J. et al. (2004). Identification of the E1Aregulated transcription factor p120 E4F as an interacting partner of the RASSF1A candidate tumor suppressor gene. Cancer Res 64 102-107. Fu, J., Bian, M., Jiang, Q., and Zhang, C. (2007) Roles of Aurora kinases in mitosis and tumorigenesis. Mol Cancer Res 5, 1-10. Giannakakou, P., Robey, R., Fojo, T., and Blagosklonny, M.V. (2001). Low concentrations of paclitaxel induce cell type-dependent p53, p21 a nd G1/G2 arrest instead of mitotic arrest: molecular determinants of paclita xel-induced cytotoxicity. Oncogene 20, 3806-3813. Gostissa, M., Morelli, M., Mantovani, F., Guida, E., Piazza, S., Collavin, L., Brancolini, C., Schneider, C., and Del Sal, G. (2004). The tr anscriptional represso r hDaxx potentiates p53dependent apoptosis. J Biol Chem 279, 48013-48023. Hannemann, J., Oosterkamp, H.M., Bosch, C.A., Veld s, A., Wessels, L.F., Loo, C., Rutgers, E.J., Rodenhuis, S., and van de Vijver, M.J. (2005). Changes in gene expression associated with response to neoadjuvant chemothera py in breast cancer. J Clin Oncol 23, 3331-3342. Hari, M., Wang, Y., Veeraraghavan, S., and Cabral F. (2003a). Mutations in alphaand betatubulin that stabilize microtubules and confer resistance to colcem id and vinblastine. Mol Cancer Ther 2, 597-605. 81

PAGE 82

Hari, M., Yang, H., Zeng, C., Canizales, M., and Ca bral, F. (2003b). Expression of class III betatubulin reduces microtubule assembly and c onfers resistance to paclitaxel. Cell Motil Cytoskeleton 56, 45-56. Haruki, N., Saito, H., Harano, T., Nomoto, S., Ta kahashi, T., Osada, H., and Fujii, Y. (2001). Molecular analysis of the m itotic checkpoint genes BUB1, BU BR1 and BUB3 in human lung cancers. Cancer Lett 162, 201-205. Henderson, I.C., Berry, D.A., Demetri, G.D., Ci rrincione, C.T., Goldstein, L.J., Martino, S., Ingle, J.N., Cooper, M.R., Hayes, D.F., Tkaczuk, K.H., et al. (2003). Improved outcomes from adding sequential Paclitaxel but not from escalating Doxor ubicin dose in an adjuvant chemotherapy regimen for patients with node-p ositive primary breast cancer. J Clin Oncol 21, 976-983. Hernandez-Vargas, H., Palacios, J., and Moreno-Bueno, G. (2006). Molecular profiling of docetaxel cytotoxicity in breas t cancer cells: un coupling of aberrant mitosis and apoptosis. Oncogene. Hollenbach, A.D., McPherson, C.J., Mientjes, E.J., Iyengar, R., and Grosveld, G. (2002). Daxx and histone deacetylase II associate with chromatin through an interaction with core histones and the chromatin-associated protein Dek. J Cell Sci 115, 3319-3330. Ikui, A.E., Yang, C.P., Matsumoto, T., and Horwitz S.B. (2005). Low concentrations of taxol cause mitotic delay followed by premature disso ciation of p55CDC from Mad2 and BubR1 and abrogation of the spindle checkpoint leading to aneuploidy. Cell Cycle 4 1385-1388. Ishov, A.M., Sotnikov, A.G., Negorev, D., Vladimir ova, O.V., Neff, N., Kamitani, T., Yeh, E.T., Strauss, J.F., 3rd, and Maul, G.G. (1999). PML is critical for ND10 form ation and recruits the PML-interacting protein daxx to this nuclear structure when modified by SUMO-1. J Cell Biol 147, 221-234. Ishov, A.M., Vladimirova, O.V., and Maul, G.G. (2004). Heterochromatin and ND10 are cellcycle regulated and phosphorylati on-dependent alternate nuclear sites of the transcription repressor Daxx and SWI/SNF protein ATRX. J Cell Sci 117, 3807-3820. Iwao-Koizumi, K., Matoba, R., Ueno, N., Kim, S.J., Ando, A., Miyoshi, Y., Maeda, E., Noguchi, S., and Kato, K. (2005). Prediction of docetaxel response in human breast cancer by gene expression profiling. J Clin Oncol 23, 422-431. Jiang, N., Wang, X., Yang, Y., and Dai, W. (2006). Advances in mitotic in hibitors for cancer treatment. Mini Rev Med Chem 6, 885-895. Jung, Y.S., Kim, H.Y., Kim, J., Lee, M.G., Pouyssegur, J., and Kim, E. (2008). Physical interactions and functional coupling between Daxx and sodi um hydrogen exchanger 1 in ischemic cell death. J Biol Chem 283, 1018-1025. Jung, Y.S., Kim, H.Y., Lee, Y.J., and Kim, E. (2007). Subcellular localization of Daxx determines its opposing functions in ischemic cell death. FEBS Lett 581, 843-852. 82

PAGE 83

Junn, E., Taniguchi, H., Jeong, B.S., Zhao, X., Ichijo, H., and Mouradian, M.M. (2005). Interaction of DJ-1 with Daxx inhibits apoptosis signal-regulating kina se 1 activity and cell death. Proc Natl Acad Sci U S A 102, 9691-9696. Kang, D., Chen, J., Wong, J., and Fang, G. (2002). The checkpoint protein Chfr is a ligase that ubiquitinates Plk1 and inhibits Cdc2 at the G2 to M transition. J Cell Biol 156, 249-259. Karunakaran, S., Diwakar, L., Saeed, U., Agar wal, V., Ramakrishnan, S., Iyengar, S., and Ravindranath, V. (2007). Activation of apoptos is signal regulating ki nase 1 (ASK1) and translocation of death-associated protein, Daxx, in substantia nigra pars compacta in a mouse model of Parkinson's disease: protection by alpha-lipoic acid. FASEB J 21 2226-2236. Keen, N., and Taylor, S. (2004). Aurora-kinase inhibitors as an ticancer agents. Nat Rev Cancer 4, 927-936. Khelifi, A.F., D'Alcontres, M.S., and Salomoni P. (2005). Daxx is required for stress-induced cell death and JNK activa tion. Cell Death Differ 12, 724-733. Kienitz, A., Vogel, C., Morales, I., Muller, R., and Bastians, H. (2005). Partial downregulation of MAD1 causes spindle checkpoint inactivation and aneuploidy, but does not confer resistance towards taxol. Oncogene 24, 4301-4310. Kiriakidou, M., Driscoll, D.A., Lopez-Guisa, J. M., and Strauss, J.F., 3rd (1997). Cloning and expression of primate Daxx cDNAs and mapping of the human gene to chromosome 6p21.3 in the MHC region. DNA Cell Biol 16 1289-1298. Kitagawa, D., Kajiho, H., Negishi, T., Ura, S., Watanabe, T., Wada, T., Ichijo, H., Katada, T., and Nishina, H. (2006). Release of RASSF1C fr om the nucleus by Daxx degradation links DNA damage and SAPK/JNK activation. EMBO J 25, 3286-3297. Ko, Y.G., Kang, Y.S., Park, H., Seol, W., Kim, J., Kim, T., Park, H.S., Choi, E.J., and Kim, S. (2001). Apoptosis signal-regulating kinase 1 co ntrols the proapoptotic function of deathassociated protein (Daxx) in the cytoplasm. J Biol Chem 276, 39103-39106. Kroemer, G., and Martin, S.J. (2005). Caspase-independent cell death. Nat Med 11, 725-730. Lalioti, V.S., Vergarajauregui, S., Pulido, D., and Sandoval, I.V. (2002). The insulin-sensitive glucose transporter, GLUT4, inte racts physically with Daxx. Two pr oteins with capacity to bind Ubc9 and conjugated to SUMO1. J Biol Chem 277, 19783-19791. Lee, E.A., Keutmann, M.K., Dowling, M.L., Harris, E., Chan, G., and Kao, G.D. (2004). Inactivation of the mitotic checkpoint as a dete rminant of the efficacy of microtubule-targeted drugs in killing human can cer cells. Mol Cancer Ther 3, 661-669. Lehembre, F., Muller, S., Pandolfi, P.P., and Dejean, A. (2001). Regulation of Pax3 transcriptional activity by SUMO-1-modified PML. Oncogene 20, 1-9. 83

PAGE 84

Lens, S.M., Wolthuis, R.M., Klompmaker, R., Ka uw, J., Agami, R., Brummelkamp, T., Kops, G., and Medema, R.H. (2003). Survivin is require d for a sustained spindle checkpoint arrest in response to lack of tension. Embo J 22 2934-2947. Li, H., Leo, C., Zhu, J., Wu, X., O'Neil, J., Par k, E.J., and Chen, J.D. (2000). Sequestration and inhibition of Daxx-mediated transcrip tional repression by PML. Mol Cell Biol 20, 1784-1796. Li, W., Zhu, T., and Guan, K.L. (2004). Transfor mation potential of Ras isoforms correlates with activation of phosphatidylinositol 3kinase but not ERK. J Biol Chem 279, 37398-37406. Lin, D.Y., Lai, M.Z., Ann, D.K., and Shih, H.M. (2003). Promyelocytic leukemia protein (PML) functions as a glucocorticoid receptor co-activator by sequest ering Daxx to the PML oncogenic domains (PODs) to enhance its trans activation potential. J Biol Chem 278, 15958-15965. Lindsay, C.R., Scholz, A., Morozov, V.M., and Ishov, A.M. (2007). Daxx shortens mitotic arrest caused by paclitaxel. Cell Cycle 6, 1200-1204. Liu, L., Amy, V., Liu, G., and McKeehan, W.L. (2002). Novel complex integrating mitochondria and the microtubular cytoskeleton with chromo some remodeling and tumor suppressor RASSF1 deduced by in silico homology analysis, interac tion cloning in yeast, and colocalization in cultured cells. In Vitro Cell Dev Biol Anim 38, 582-594. Liu, L., Baier, K., Dammann, R., and Pfeifer, G.P. (2007). The tumor suppressor RASSF1A does not interact with Cdc20, an activator of the anaphase-promoting complex. Cell Cycle 6 16631665. Liu, L., Tommasi, S., Lee, D.H., Dammann, R., and Pfeifer, G.P. (2003). Control of microtubule stability by the RASSF1A tumor suppressor. Oncogene 22, 8125-8136. Malumbres, M. (2006). Therapeutic opportunities to control tumor cell cycles Clin Transl Oncol 8, 399-408. Mansilla, S., Bataller, M., and Portugal, J. (2006a). Mitotic catastrophe as a consequence of chemotherapy. Anticancer Agents Med Chem 6, 589-602. Mansilla, S., Priebe, W., and Portugal, J. (2006b) Mitotic catastrophe re sults in cell death by caspase-dependent and caspase-independent mechanisms. Cell Cycle 5, 53-60. Mauriac, L., Debled, M., and MacGrogan, G. (2005) When will more useful predictive factors be ready for use? Breast 14 617-623. Meraldi, P., Draviam, V.M., and Sorger, P.K. (2004). Timing and checkpoints in the regulation of mitotic progression. Dev Cell 7, 45-60. Michaelson, J.S. (2000). The Daxx enigma. Apoptosis 5, 217-220. 84

PAGE 85

Michaelson, J.S., Bader, D., Kuo, F., Kozak, C., and Leder, P. (1999). Loss of Daxx, a promiscuously interacting protein, results in ex tensive apoptosis in ea rly mouse development. Genes Dev 13, 1918-1923. Michaelson, J.S., and Leder, P. (2003). RNAi reveals anti-apoptotic and transcriptionally repressive activities of DAXX. J Cell Sci 116, 345-352. Miyoshi, Y., Kim, S.J., Akazawa, K., Kamigaki S., Ueda, S., Yanagisawa, T., Inoue, T., Taguchi, T., Tamaki, Y., and Noguchi, S. (2004) Down-regulation of intratumoral aromatase messenger RNA levels by docetaxel in hum an breast cancers. Clin Cancer Res 10, 8163-8169. Morrow, C.J., Tighe, A., Johnson, V.L., Scott, M. I., Ditchfield, C., and Taylor, S.S. (2005). Bub1 and aurora B cooperate to maintain BubR 1-mediated inhibition of APC/CCdc20. J Cell Sci 118, 3639-3652. Morse, D.L., Gray, H., Payne, C.M., and Gillies, R.J. (2005). Docetax el induces cell death through mitotic catastrophe in human breast cancer cells. Mol Cancer Ther 4, 1495-1504. Muromoto, R., Sugiyama, K., Takachi, A., Imot o, S., Sato, N., Yamamoto, T., Oritani, K., Shimoda, K., and Matsuda, T. (2004). Physical and functional interactions between Daxx and DNA methyltransferase 1-associ ated protein, DMAP1. J Immunol 172, 2985-2993. Musacchio, A., and Salmon, E.D. (2007). The spi ndle-assembly checkpoint in space and time. Nat Rev Mol Cell Biol 8 379-393. Nefkens, I., Negorev, D.G., Ishov, A.M., Michaelson, J.S., Yeh, E.T., Tanguay, R.M., Muller, W.E., and Maul, G.G. (2003). Heat shock and Cd2+ exposure regulate PML and Daxx release from ND10 by independent mechanisms that modify the induction of heat-shock proteins 70 and 25 differently. J Cell Sci 116 513-524. O'Shaughnessy, J. (2005). Extending survival with chemotherapy in metastatic breast cancer. Oncologist 10 Suppl 3 20-29. Perlman, R., Schiemann, W.P., Brooks, M.W., Lodish, H.F., and Wei nberg, R.A. (2001). TGFbeta-induced apoptosis is mediated by the adapte r protein Daxx that faci litates JNK activation. Nat Cell Biol 3, 708-714. Pines, J. (2006). Mitosis: a matte r of getting rid of the right prot ein at the right time. Trends Cell Biol 16, 55-63. Pluta, A.F., Earnshaw, W.C., and Goldberg, I.G. (1998). Interphase-specific association of intrinsic centromere protein CENP-C with HDaxx, a death domain-binding protein implicated in Fas-mediated cell death. J Cell Sci 111 ( Pt 14) 2029-2041. Ravdin, P., Erban, J., and Overmoyer, B. ( 2003). Phase III comparison of docetaxel and paclitaxel in patients with meta static breast cancer. Eur J Cancer suppl1:32 85

PAGE 86

Ricci, M.S., and Zong, W.X. (2006). Chemotherape utic approaches for targeting cell death pathways. Oncologist 11 342-357. Rong, R., Jin, W., Zhang, J., Sheikh, M.S., a nd Huang, Y. (2004). Tumor suppressor RASSF1A is a microtubule-binding protein that stabilizes microtubules a nd induces G2/M arrest. Oncogene 23, 8216-8230. Roninson, I.B., Broude, E.V., and Chang, B.D. ( 2001). If not apoptosis, then what? Treatmentinduced senescence and mitotic catastrophe in tumor cells. Drug Resist Updat 4, 303-313. Rowinsky, E.K., Eisenhauer, E.A., Chaudhry, V ., Arbuck, S.G., and Donehower, R.C. (1993). Clinical toxicities encountered with paclitaxel (Taxol). Semin Oncol 20, 1-15. Sablina, A.A., Budanov, A.V., Ilyinskaya, G.V., Agapova, L.S., Kravchenko, J.E., and Chumakov, P.M. (2005). The antioxidant func tion of the p53 tumor suppressor. Nat Med 11, 1306-1313. Schiff, P.B., Fant, J., and Horwitz, S.B. (1979). Promotion of microtubule assembly in vitro by taxol. Nature 277, 665-667. Scolnick, D.M., and Halazonetis, T.D. (2000). Chfr defines a mitotic stress checkpoint that delays entry into metaphase. Nature 406 430-435. Shivakumar, L., Minna, J., Sakamaki, T., Pest ell, R., and White, M.A. (2002). The RASSF1A tumor suppressor blocks cell cycle progression a nd inhibits cyclin D1 accumulation. Mol Cell Biol 22, 4309-4318. Song, J.J., and Lee, Y.J. (2003). Role of the ASK1-SEK1-JNK1-HIPK1 signal in Daxx trafficking and ASK1 oligomerization. J Biol Chem 278, 47245-47252. Song, J.J., and Lee, Y.J. (2004). Tryptophan 621 and serine 667 residue s of Daxx regulate its nuclear export during glucose deprivation. J Biol Chem 279, 30573-30578. Song, M.S., Song, S.J., Ayad, N.G., Chang, J.S., Lee, J.H., Hong, H.K., Lee, H., Choi, N., Kim, J., Kim, H. et al. (2004). The tumour suppressor RASSF1A regulates mitosis by inhibiting the APC-Cdc20 complex. Nat Cell Biol 6 129-137. Song, M.S., Song, S.J., Kim, S.J., Nakayama, K., Nakayama, K.I., and Lim, D.S. (2007). Skp2 regulates the antiproliferative function of th e tumor suppressor RASSF1A via ubiquitin-mediated degradation at the G(1) -S transition. Oncogene. Sudo, T., Nitta, M., Saya, H., and Ueno, N.T. (2004) Dependence of paclitaxel sensitivity on a functional spindle assembly checkpoint. Cancer Res 64, 2502-2508. Tang, J., Qu, L.K., Zhang, J., Wang, W., Michaels on, J.S., Degenhardt, Y. Y., El-Deiry, W.S., and Yang, X. (2006). Critical role for Daxx in regulating Mdm2. Nat Cell Biol 8, 855-862. 86

PAGE 87

Torii, S., Egan, D.A., Evans, R.A., and Ree d, J.C. (1999). Human Daxx regulates Fas-induced apoptosis from nuclear PML oncoge nic domains (PODs). EMBO J 18, 6037-6049. Tsai, M.Y., Wang, S., Heidinger, J.M., Shum aker, D.K., Adam, S.A., Goldman, R.D., and Zheng, Y. (2006). A mitotic lamin B matrix induced by RanGTP required for spindle assembly. Science 311, 1887-1893. Vos, M.D., Martinez, A., Elam, C., Dallol, A., Taylor, B.J., Latif, F., and Clark, G.J. (2004). A role for the RASSF1A tumor suppressor in the re gulation of tubulin poly merization and genomic stability. Cancer Res 64, 4244-4250. Wang, L.G., Liu, X.M., Kreis, W., and Budman, D.R. (1999). The effect of antimicrotubule agents on signal transduction pathways of apoptosis: a review. Cancer Chemother Pharmacol 44, 355-361. Wang, Y., and Cabral, F. (2005). Paclitaxel re sistance in cells with reduced beta-tubulin. Biochim Biophys Acta 1744 245-255. Wassmann, K., and Benezra, R. (2001). Mitotic checkpoints: from yeast to cancer. Curr Opin Genet Dev 11, 83-90. Wood, K.W., Cornwell, W.D., and Jackson, J.R. (2001) Past and future of the mitotic spindle as an oncology target. Curr Opin Pharmacol 1, 370-377. Xia, G., Luo, X., Habu, T., Rizo, J., Matsumot o, T., and Yu, H. (2004). Conformation-specific binding of p31(comet) antagonizes the function of Mad2 in the spindle checkpoint. Embo J 23, 3133-3143. Xue, Y., Gibbons, R., Yan, Z., Yang, D., McDowell, T.L., Sechi, S., Qin, J., Zhou, S., Higgs, D., and Wang, W. (2003). The ATRX syndrome prot ein forms a chromatin-remodeling complex with Daxx and localizes in promyelocytic leuke mia nuclear bodies. Proc Natl Acad Sci U S A 100, 10635-10640. Yang, X., Khosravi-Far, R., Chang, H.Y., and Ba ltimore, D. (1997). Daxx, a novel Fas-binding protein that activates JNK and apoptosis. Cell 89 1067-1076. Yeung, T.K., Germond, C., Chen, X., and Wang, Z. (1999). The mode of action of taxol: apoptosis at low concentrati on and necrosis at high con centration. Biochem Biophys Res Commun 263, 398-404. Yu, X., Minter-Dykhouse, K., Malureanu, L., Zha o, W.M., Zhang, D., Merkle, C.J., Ward, I.M., Saya, H., Fang, G., van Deursen, J., and Chen, J. (2005). Chfr is require d for tumor suppression and Aurora A regulation. Nat Genet 37 401-406. Zhao, L.Y., Liu, J., Sidhu, G.S., Niu, Y., Liu, Y., Wang, R., and Liao, D. (2004). Negative regulation of p53 functions by Daxx and th e involvement of MDM2. J Biol Chem 279, 5056650579. 87

PAGE 88

Zhong, S., Salomoni, P., Ronchetti, S., Guo, A., Ruggero, D., and Pandolfi, P.P. (2000). Promyelocytic leukemia protein (PML) and Da xx participate in a nove l nuclear pathway for apoptosis. J Exp Med 191, 631-640. 88

PAGE 89

BIOGRAPHICAL SKETCH Cory Lindsay was born in Laurel, Nebraska, a small town located in northeast Nebraska cornering Iowa and South Dakota. He is the son of Archie and Virginia Lindsay, of Laurel, and has three siblings Lori, Scott and Michelle. Cory attended Laurel-Concord High School and graduated in 1998. During his high school years, Cory had several ou tstanding teachers who motivated and inspired him to pursue academics. His first experience with science research came during this time when he studied meiofauna biology and won several local, regional and national awards for his efforts. He also devel oped an interest in colle cting and breeding snakes and other reptiles and amphibians, a hobby which he still pursues today. After high school, Cory pursued a Bachelor of Sc ience degree in Biological Sciences from Wayne State College, in Wayne, NE and graduated in 2002. While an undergraduate, Cory was fortunate to have many opportuni ties to perform independent sc ience research. He conducted ecological research in the British West Indies studying dwarf-geckos and snakes as part of a collaborative effort with Avila College, Kans as City, MO; identified novel genes from the Schistosoma mansoni genome at the Whitney Laborat ory, St. Augustine, FL; performed molecular biology experiments as an intern at the Bermuda Biological Station for Research, St. Georges, Bermuda; and deciphered genetic dise ase abnormalities at the world famous Cold Spring Harbor Laboratory, Long Island, NY. Cory joined the University of Floridas IDP graduate school program in 2003 and finished his Ph.D work in the laboratory of Dr. Alex ander Ishov studying Daxx f unction in cellular taxol response and cell cycle progression. He has pr esented this work at several national and international conferences and published his work su mmarizing these results. Cory will continue post-doctoral work in cancer research and diagnostics. 89