Citation
CBX2 stably associates with mitotic chromosomes via a PRC2 or PRC1- independent mechanism and is needed for recruiting PRC1 complex to mitotic chromosomes

Material Information

Title:
CBX2 stably associates with mitotic chromosomes via a PRC2 or PRC1- independent mechanism and is needed for recruiting PRC1 complex to mitotic chromosomes
Creator:
Zhen, Chao Yu ( author )
Language:
English
Physical Description:
1 electronic file : ;

Subjects

Subjects / Keywords:
Chromatin ( lcsh )
Epigenetics ( lcsh )
Cell division ( lcsh )
Cell division ( fast )
Chromatin ( fast )
Epigenetics ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Polycomb group (PcG) proteins are composed of a multiplicity of transcriptional regulatory factors that maintain cellular identity through epigenetic mechanism. Often in most cases, once transcriptional silencing is imposed on genes, it can be transmitted from mother to daughter cells. Surprisingly, during mitotic event, chromatin goes through a dramatic structural reorganization and is compacted that would limit most transcription factors and chromatin protein access. Thus, it is unclear how the PcG-mediated transcriptional program is maintained during mitosis. Here, using quantitative live-cell imaging in mouse ES cells and tumor cells, we demonstrate that, although Polycomb repressive complex (PRC) 1 proteins (Cbx-family proteins, Ring1b, Mel18, and Phc1) exhibit variable capacities of association with mitotic chromosomes, Cbx2 overwhelmingly binds to mitotic chromosomes. The recruitment of Cbx2 to mitotic chromosomes is independent of PRC1 or PRC2, and Cbx2 is needed to recruit PRC1 complex to mitotic chromosomes. Quantitative fluorescence recovery after photobleaching analysis indicates that PRC1 proteins rapidly exchange at interphasic chromatin. On entry into mitosis, Cbx2, Ring1b, Mel18, and Phc1 proteins become immobilized at mitotic chromosomes, whereas other Cbx-family proteins dynamically bind to mitotic chromosomes. Depletion of PRC1 or PRC2 protein has no effect on the immobilization of Cbx2 on mitotic chromosomes. We find that the N-terminus of Cbx2 is needed for its recruitment to mitotic chromosomes, whereas the C-terminus is required for its immobilization. Thus these results provide fundamental insights into the molecular mechanisms of epigenetic inheritan.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
Chao Yu Zhen.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
982698604 ( OCLC )
ocn982698604
Classification:
LD1193.L46 2916m Z54 ( lcc )

Downloads

This item has the following downloads:


Full Text
CBX2 STABLY ASSOCIATES WITH MITOTIC CHROMOSOMES VIA A PRC2 OR PRC1-INDEPENDENT MECHANISM AND IS NEEDED FOR RECRUITING PRC1 COMPLEX TO
MITOTIC CHROMOSOMES By
CHAO YU ZHEN
B.S., University of Colorado Denver, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Chemistry Program
2016


2016
CHAO YU ZHEN ALL RIGHTS RESERVED
11


This thesis for the Master of Science degree by Chao Yu Zhen has been approved for the Chemistry Program by
Xiaojun Ren, Chair Christopher Phiel Hai Lin
April 21, 2016
m


Chao Yu Zhen (M.S., Chemistry)
Cbx2 Stably Associates with Mitotic Chromosomes via a PRC2 or PRC1-independentMechanism and is needed for Recruiting PRC1 Complex to Mitotic Chromosomes
Thesis directed by Assistant Professor Xiaojun Ren
ABSTRACT
Polycomb group (PcG) proteins are composed of a multiplicity of transcriptional regulatory factors that maintain cellular identity through epigenetic mechanism. Often in most cases, once transcriptional silencing is imposed on genes, it can be transmitted from mother to daughter cells. Surprisingly, during mitotic event, chromatin goes through a dramatic structural reorganization and is compacted that would limit most transcription factors and chromatin protein access. Thus, it is unclear how the PcG-mediated transcriptional program is maintained during mitosis. Here, using quantitative live-cell imaging in mouse ES cells and tumor cells, we demonstrate that, although Polycomb repressive complex (PRC) 1 proteins (Cbx-family proteins, Ringlb, Mel 18, and Phcl) exhibit variable capacities of association with mitotic chromosomes, Cbx2 overwhelmingly binds to mitotic chromosomes. The recruitment of Cbx2 to mitotic chromosomes is independent of PRC 1 or PRC2, and Cbx2 is needed to recruit PRC1 complex to mitotic chromosomes. Quantitative fluorescence recovery after photobleaching analysis indicates that PRC1 proteins rapidly exchange at interphasic chromatin. On entry into mitosis, Cbx2, Ringlb, Mel 18, and Phcl proteins become immobilized at mitotic chromosomes, whereas other Cbx-family proteins dynamically bind to mitotic chromosomes. Depletion of PRC 1 or PRC2 protein has no effect on the immobilization of Cbx2 on mitotic chromosomes. We find that the N-terminus of Cbx2 is needed for its recruitment to mitotic chromosomes, whereas the C-terminus is required for its immobilization. Thus these results provide fundamental insights into the molecular mechanisms of epigenetic inheritance.
IV


The form and content of this abstract are approved. I recommend its publication.
Approved: Xiaojun Ren
v


ACKNOWLEDGEMENTS
First, I would like to express my sincere gratitude for my advisor, Prof. Xiaojun Ren, for his continuous supports and thoughtful advice of my masters work. The door to Prof. Ren office was always open at the time of need and he always steers me to ask the correct questions. His guidance helped me in all the time of my research. Beside research, Prof Rens life advices are invaluable. Also, I would like to further express my sincere gratitude for my advisor Prof. Ren for his patient, motivation, and above and beyond efforts to help me. I truly had an amazing research experience and I could not have imagined a better advisor.
Beside my advisor, I would like to thank my committees: Prof. Hai Lin and Prof. Christopher Phiel for their advices and time commitment to help improve this thesis. Also, I would like to thank them for being a great teacher to me.
My sincere further goes to Prof. Jefferson Knight and Prof. Christopher Phiel, who gave me access to their laboratory equipment and the wonderful scientific advices.
Also, I would like to thank my fellow Ren lab colleague at University of Colorado Denver and especially: Huy Due, Roubina Tatavosian, and Marko Kokotovic for their stimulating discussion and encouragement during time of my master research.
Lastly to my family, I would like express my deepest appreciation for their conditional love, support, pain, and sacrifice to shape my life.
Note: The work presented in this thesis has been published in Molecular Biology of Cell. The chromatin and PRC1 plasmids were made by Huy Due and the Cbx plasmids were made by Marko Kokotovia.
vi


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION................................................................1
II. RESULTS....................................................................4
PRC1 proteins vary in association with mitotic chromosomes.................4
The mitotic chromosomal association of Cbx2 is independent of PcG proteins ... 10
Cbx2 is required for recruiting the canonical PRC1 proteins to mitotic chromosomes ..............................................................13
Cbx2 directly recruits the canonical PRC1 proteins to mitotic chromosomes.15
The Cbx2-PRC1 complex is immobilized at mitotic chromosomes, but other Cbx family proteins rapidly exchange at mitotic chromosomes ..................18
The immobilization of Cbx2 at mitotic chromosomes is independent of PcG proteins .......................................................................22
The recruitment and immobilization of Cbx2 to mitotic chromosomes requires its distinct regions..........................................................24
III. DISCUSSION..................................................................28
IV. MATERIALS AM) METHODS........................................................34
Cell lines................................................................34
Plasmid...................................................................34
Generation of stable cell lines by lentivirus infection...................35
Confocal microscope imaging of live cells and quantification of mitotic fraction 36
FRAP imaging and quantification ..........................................37
FRET imaging .............................................................38
Epifluorescence imaging of live cells.....................................38
Cell synchronization and fractionation....................................38
vii


Immunoblottin
39
Immunofluorescence .............................................40
REFERENCES.............................................................42
viii


LIST OF FIGURES
Figures
1. Mitotic chromosomal association of PRC 1 fusion PROTEINS........................5
2. Mitotic chromosomal association of YFP-Cbx2 fusion protein in HEK293 and
HeLa cells......................................................................6
3. Effects of doxycycline concentration on level of YFP-Cbx2 association with mitotic
Chromosomes ....................................................................7
4. Western blot analysis of levels of endogenous and fusion proteins...............7
5. Analysis of mitotic chromatin binding of PRC 1 proteins by chromatin Fractionation .................................................................9
6. The association of Cbx2 fusion protein with mitotic chromosomes in PRC1 and PRC2
gene knockout ES cells.........................................................12
7. Western blot analysis of Ringlb in Ring la-/-/Ring lbfl/fl ES cells treated with or without
OHT ...........................................................................13
8. The mitotic chromosomal association of PRC1 fusion proteins Ringlb, Phcl, and Mell8
in Cbx2-/- ES cell.............................................................14
9. Immunostaining of Phcl and Ringlb proteins in Cbx2+/+ and Cbx2-/- ES cells.....15
10. Directly recruiting PRC1 fusion proteins to mitotic chromosomes by YFP-Cbx2, but not
by YFP-Cbx21-498...............................................................18
11. FRAP analysis of PRC1 fusion proteins binding to interphasic and mitotic
Chromatins ....................................................................20
12. FRAP analysis of YFP-Cbx2 fusion protein binding to interphasic and mitotic
chromatins in HeLa and HEK293 cells...........................................22
13. FRAP analysis of YFP-Cbx2 fusion protein binding to chromatins in Eed-/-, Ringla-/-
/Ringlb-/-, and Bmil-/-/Mell8-/- ES cells ....................................23
14. Analysis of structural elements of YFP-Cbx2 fusion protein required for its targeting
and immobilizing .............................................................26
15. Hypothetic model for the interaction of canonical PRC1 complex with interphasic and
mitotic chromatin ............................................................33
IX


LIST OF ABBREVIATIONS
BiFC Bimolecular fluorescence complementation
BSA Bovine Serum Albumin
Cbx2"' Knock out of Cbx2 gene
Cbx2 Chromobox Homolog 2
Cbx4 Chromobox Homolog 4
Cbx6 Chromobox Homolog 6
Cbx7 Chromobox Homolog 7
Cbx8 Chromobox Homolog 8
Cbox Chromobox
CHD Chromodomain
DMEM Dulbeccos Modified Eagles Medium
DTT Dithiothreitol
EDTA Ethylenediaminetetraacetic acid
Eed Embryonic Ectoderm Development
EGFP Enhanced Green Fluorescent Protein
EGTA Aminopolycarboxylic acid
EMCCD Electron Multiplying Charge Coupled Device
ES Cel Embryonic Stem Cell
FITC Fluorescein Isothiocyanate
fl/fl Flox/flox (conditional knockout)
FRAP Fluoresence Recovery After Photobleaching
FRET Fluoresence Resonence Energy Transfer
x


H3K27me3 Trimethylation of Lysine 27 on Histone 3
H2A Histone Family, Member Z
HEK293 Human Embryonic Kidney 293
HEPES 4-(2-hydroxyethyl)-1 -piperazineethanesulfonic acid
KC1 Potassium Chloride
KO Knockout
LIF Leukemia Inhibitor Factor
MgCl2 Magnesium Chloride
M-Phase Metaphase
NaCl Sodium Chloride
OHT 4-hyroxytamoxifen
PBS Phosphate Buffered Saline
PcG Polycomb Group
Pcgfl Poly comb Group Ring Finger 1
Pcgf2 Poly comb Group Ring Finger 2
Pcgfi Poly comb Group Ring Finger 3
Pcgf4 Poly comb Group Ring Finger 4
Pcgf5 Poly comb Group Ring Finger 5
Pcgf6 Poly comb Group Ring Finger 6
Phcl Polyhomeotic Homolog 1
Phc2 Polyhomeotic Homolog 2
Phc3 Polyhomeotic Homolog 3
PMSF Phenylmethylsulfonyl fluoride
XI


PRC1 Polycomb Repressive Complex 1
PRC2 Polycomb Repressive Complex 2
SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis
RING1B Really Interesting New Gene 1 B
WT Wild Type
YFP Yellow Fluorescence Protein
Xll


CHAPTER I
INTRODUCTION
The transcriptional programming of genes occurs during early embryonic developmental stages and is a major influence for cellular identity and diversity. Transcriptional activation or repression (silencing) of genes will often be retained during development, although it remains plasticity. Polycomb Group (PcG) proteins are epigenetic transcriptional regulators that controls hundreds of key developmental homeotic genes by biochemical modification through the mediation of trimethylation on lysine 27 of histone (H3K27me3) (Kerppola, 2009; Di Croce and Helin, 2013; Sharif .,2013; Simon and Kingston, 2013). Depletion of PcG genes can results in dysfunction of mouse embryonic stem (ES) cell (Park IK, et al 2003) or impairment and terminal of mouse embryonic development (Voncken JW, et al 2003). PcG mediated epigenetic repression mark or memory, in most case, can be transmitted from mother to daughter cells through mitotic event. PcG proteins action is especially vital during cell division to preserve cells identities. However, during mitosis, chromatin undergoes disruption and subsequent restoration, which challenge the integrity of genetic and epigenetic memory and also presents a window of opportunity for epigenetic change. The vast majority of transcription factors and chromatin bind proteins dissociate from mitotic chromosomes, although, some mitotic chromosomes are open for traffics (Chen et al., 2005; Hemmerich et al., 2011). Thus, to better understand epigenetic memory and flexibility, it is essential to study the behavior of PcG-mediate gene repression action on mitotic chromosomes.
Initially discovered in drosophila melanogaster (fly), (PcG) protein associate with two transcriptional repressive complexes, Polycomb repressive complex (PRC) 1 and 2 (Kerppola, 2009; Di Croce and Helin, 2013; Sharif et al., 2013; Simon & Kingston, 2013). Both PRC 1 and
1


2 process histone catalytic activity. PRC2 is a methlytransferase that primarily enzymatically trimethlyate at histone H3 tails on lysine 27 (H3K27me3) (Cao et al, 2002), and recruits PRC1, the enzymatic function of ubiquitin ligase, to monoubquitylation of lysine 119 on H2A tails (H2Akl 19ubl) (de Napoles et al, 2004; Wang et al, 2004a; Simon and Kingston, 2013). PRC1 comprises of 5 major core subunits families proteins, which comprise of poly comb group of ring finger (PcgFl/2/3/4/5/6), Phc (Phcl/2/3), Cbx (cbx2/4/6/7/8), Rybp/YaF2, and Ringl (A/B), the assemblage of PRC 1 complexes (Simon & Kingston, 2009; Turner & Bracken, 2013). In mammal, the multiple variations of each core PRC1 family protein attributed to a large library of uniquely individual PRC1 complexes. An attempt to establish a more simplified classification of PRC1 complexes was drafted, which was based on the concept of canonical and noncanonical PRC1 complexes (Gao et al, 2012; Tavares et al, 2012; Morey etal., 2013). Cbx and RybP are mutually exclusive associated with PRC1 complexes are defined as canonical and noncanonical, respectively. Canonical PRC1 has been reported to be recruited to chromatin by H3K23me3 a PRC2 dependent mechanism (Wang et al, 2004b). Even as now, the functions of each unique individual PRC1 complexes remain to be resolved.
Extensive research has been done on the recruitment of canonical PRC1 complexes to interphasic chromatin (Cao et al., 2002; Ogawa et al., 2002; Wang et al., 2004b; Gearhart et al., 2006; Sanchez et al., 2007; Boukarabila et al., 2009; Ren and Kerppola, 2011; Vandamme et al., 2011; Gao et al., 2012; Tavares et al., 2012; Morey et al., 2013; Cheng et al., 2014), yet some fundamental epigenetic inheritance role of PRC 1 in mitotic chromosomes remains to be answer. (1) Does PRC1 complexes interact with mitotic chromosomes, and if so, (2) what is the mechanism that dictate PRC1 for this interaction. Studies have reached a division of views on how vertebrate PRC1 proteins interact with mitotic chromosomes (Wang et al., 1997; Saurin et
2


al., 1998; Koga et al., 1999; Voncken et al., 1999; Akasaka et al., 2002; Suzuki et al., 2002; Miyagishima et al., 2003; Aoto et al., 2008; Vincenz and Kerppola, 2008). In earlier studies, evidence was provided that a small subpopulation of PRC 1 proteins associated with mitotic chromosomes (Follmer etal., 2012; Fonseca etal., 2012; Steffen etal., 2013). Surprisingly, PRC1 binding dynamic was found to be 300-fold longer comparing mitotic to interphasic chromatin (Fonseca et al., 2012). The key mechanism for epigenetic inheritance of PRC1 is the physical interaction with mitotic chromosomes, which the interaction are retention and immobilization, which should be conserved across organisms that contain PRC1.
Here we show, using quantitative live cell imaging, Cbx2 is the dominant Cbx family protein that overwhelmingly associate to mitotic chromosomes. The association of Cbx2 is independent of PRC 1 and PRC2 mechanisms. Cbx2 is the key protein that is required for the recruitment of PRC 1 proteins to mitotic chromosomes. Utilizing quantitative FRAP, we demonstrated that PRC1 proteins is highly dynamic with interphasic chromatin, as reported previously (Ren et al., 2008), and identify that a special Cbx2-PRC1 complex is selectively immobilized at mitotic chromosomes. The immobilization of cbx2 protein at mitotic chromosomes is independent of PRC2 or PRC1. Finally, we reveal that cbx2 retain uniquely different mechanisms for recruitment and immobilization on mitotic chromosomes.
3


CHAPTER II
RESULTS
PRC1 proteins vary in association with mitotic chromosomes
Early studies of the association of mammalian PRC1 proteins with mitotic chromosomes reach divergent opinions (Wang et al., 1997; Saurin etal., 1998; Koga etal., 1999; Voncken et al., 1999; Akasaka et al., 2002; Suzuki et al., 2002; Miyagishima et al., 2003; Aoto et al., 2008; Vincenz and Kerppola, 2008). To fully appreciate whether the Cbx-containing PRC1 complexes associate with mitotic chromosomes, we established ES cell lines that inducibly express PRC1 protein fused with Cerulean or YFP under doxycycline-controlled manner. To facilitate live cell imaging and to mark mitotic chromosomes, histone H2A fused with Cerulean or mCherry was co-expressed with PRC1 fusion protein. We performed multicolor Z-scan imaging of live cells at 37 C with a confocal laser-scanning microscope. The PRC1 fusion proteins (Cbx family proteins (Cbx2, Cbx4, Cbx6, Cbx7, Cbx8), Ringlb, Phcl and Mel 18) exhibited variable capacities of association with mitotic chromosomes. YFP-Cbx2 fusion was the primary protein accumulated at mitotic chromosomes (Fig. 1A-H). These PRC1 fusion proteins were granularly distributed at mitotic chromosomes. Quantitative analysis of Z-stack images showed that (96 5)% of YFP-Cbx2 protein associates with mitotic chromosomes, while YFP-Cbx4 (22 5)%, YFP-Cbx6 (26 7)%, YFP-Cbx7 (44 4)%, Cbx8 (40 9)%, Cerulean-Ringlb (63 10)%, Cerulean-Phcl (84 10)%, and Cerulean-Mell8 (84 7)% all showed reduced association with mitotic chromosomes (Fig. II). These results are consistent with previous BiFC analysis of Cbx2 and Cbx6 association with mitotic chromosomes (Vincenz and Kerppola, 2008), but the current studies provide the quantitative insights of the entire Cbx family proteins and other PRC1
4


proteins. Thus, these data indicate that the Cbx-containing PRC1 complexes associate with mitotic chromosomes and Cbx2 is the primary protein enriched at mitotic chromosomes.
A YFP mCherry B YFP mC hefty C YFP mCheay
F Cerulean mCherry ^ Cerulean mCherry ^ Cerulean mCherry
-Ringl b -H2A Overlay -Phc1 -H2A Overlay -Mel 18 -H2A Overtay
Figure 1. Mitotic chromosomal association of PRC1 fusion proteins. (A-H) Confocal fluorescence images of PRC 1 fusion proteins tagged with either YFP or Cerulean expressed in PGK12.1 ES cells in various phases of mitosis. The mCherry-H2A fusion protein was used to mark mitotic chromosomes. The images of prometaphase (top panel), metaphase (middle panel), and anaphase/telophase (bottom panel) are presented. The representative image from Z-scan stack is presented in the Figures. Scale bar is 5 pm. (I) Quantitative analysis of mitotic binding fraction of PRC 1 fusion proteins at mitotic chromosomes. The mitotic binding fraction is average of individual slices of Z-scan stack. The data represents at least ten cells analyzed. Error bars indicate standard deviation of the mean.
5


Since Cbx2 is the primary protein associated with mitotic chromosomes, we wished to determine whether the mitotic chromosomal association of Cbx2 protein is cell-type specific. Therefore, we established HEK293 and HeLa cell lines that stably and inducibly express YFP-Cbx2 and mCherry-H2A fusion proteins. Z-scan imaging of live HEK293 and HeLa cells showed that YFP-Cbx2 fusion protein was profoundly accumulated at mitotic chromosomes in both cell lines (Fig. 2). Consistent with the YFP-Cbx2 fusion protein localization at mitotic chromosomes in ES cells, quantitative analysis of Z-stack images showed that (96 4)% of YFP-Cbx2 is associated with mitotic chromosomes in both HEK293 and HeLa cells, indicating the mitotic chromosomal association of Cbx2 protein is not restricted to ES cells.
Cerulean
YFP-Cbx2 -H2A Overlay
LU
X
03
_l
0
X
Figure 2. Mitotic chromosomal association of YFP-Cbx2 fusion protein in HEK293 and HeLa cells. YFP-Cbx2 and Cerulean-H2A fusion proteins were expressed in HEK293 and HeLa cells. The Cerulean-H2A was used to mark mitotic chromosomes. The confocal images of metaphase are presented. Scale bar is 5 pm.
To ask whether the expression level of Cbx2 fusion protein affects the mitotic chromosomal association, we administrated a wide range of concentrations of doxycycline, from 0.1 to 1.0 pM, into HEK293 cells stably and inducibly expressing YFP-Cbx2 and Cerulean-H2A. Epifluorescence imaging of live HEK293 cells showed that the fraction of mitotic chromosomal association of Cbx2 fusion protein was similar to one another under varying concentrations of
6


doxycycline (Fig. 3). Unless otherwise indicated, we used doxycycline at the concentration of 0.20-0.50 gM throughout subsequent studies. At 0.5 pM of doxycycline, the expression levels of the PRC1 proteins tested were either similar to or 2-3-fold higher than that of their endogenous counterparts (Fig. 4).
YFP-Cbx2 Cerulean-H2A Overlay
Figure 3. Effects of doxycycline concentration on level of YFP-Cbx2 association with mitotic chromosome. The expression of YFP-Cbx2 fusion protein in HEK293 cells was induced by using a range of doxycycline concentrations. Cerulean-H2A was used to mark mitotic chromosomes. The epifluorescence images of metaphase are presented. Scale bar is 5 pm.
YFP-Cbx2 I Cbx2 I
Cerulean-Ring 1b Ring 1b
Cerulean-Phc1
Phc1
Figure 4. Western blot analysis of levels of endogenous and fusion proteins.
Western blots were performed using antibodies against endogenous proteins. The cell extracts were prepared from PGK12.1 ES cells and PGK12.1 ES cells expressing either YFP-Cbx2 (PGK12.1 + YFP-Cbx2), Cerulean-Ringlb (PGK12.1 + Cerulean-Ringlb), or Cerulean-Phcl (PGK12.1 + Cerulean-Phcl). The fusion proteins were induced to express by 0.5 pM of doxycycline for 2 days.
7


To provide independent evidence of PRC 1 protein association with mitotic chromosomes, we performed biochemical fractionation followed by Western blot analysis. ES cells were synchronized at M-phase by treating sequentially with thymidine and nocodazole based on published report (Ballabeni et al., 2011). At least 85% of ES cells were mitotic as revealed by cell morphology and immunostaining of H3 SI Op (Fig. 5 A -B). The non-synchronized (control) and synchronized (mitotic) cells were fractioned according to the scheme showed in Fig. S4C based on previous reports (Mendez and Stillman, 2000; Follmer et al., 2012). In control and mitotic cells, the histone H3 was primarily found in the chromatin fraction (P3), while a-tubulin was found primarily in the cytosolic fraction (S2) (Fig. 5D). We tested the chromatin association of 4 PRC1 proteins (Ringlb, Phcl, Cbx2 and Cbx7) in both control and mitotic cells. In mitotic cells, we found that Cbx2 protein is primarily in the fraction P3, while Cbx7 is primarily in the cytosolic fraction S2 (Fig. 5D). Thus, these data are consistent with live cell imaging data that show that Cbx2 is the primary protein enriched at mitotic chromosomes.
8


B H3-S10p Hoechst Overlay
C Cell lysis with non-ionic detergent 5 min on ice
Centrifuge for 5 min at 1,300 g, 4 C
S1 P1
Centrifuge for 10 min at 13,000 g, 4 C
(Soluble
proteins)
no salt buffer for 30 min on ice Centrifuge for 5 min at 1,700 g, 4 C
P3
(Chromatin-associated fraction)
D
Control cells TCE S2 S3 P3
Mitotic cells TCE S2 S3 P3
9


Figure 5. Analysis of mitotic chromatin binding of PRC1 proteins by chromatin fractionation. (A) Representative images of synchronized PGK12.1 ES cells stably expressing mCherry-H2A. The circular morphology of cells in phase contrast image indicates cell at mitotic stages. The mCherry-H2A was used to mark chromatin. Scale bar is 20 pm. (B) Epifluorescence images of synchronized PGK12.1 ES cells immunostained by anti-Histone H3 phospho S10 (H3PS10) antibody. DNAs were stained by Hoechst. Over 85% of cells show positive-H3PS10 staining. Scale bar is 20 pm. (C) Scheme of PRC 1 protein fractionation, adapted from (Mendez and Stillman, 2000; Follmer eta/., 2012). (D) Western blot analysis of fractions from control and mitotic cells. H3 and a-tubulin are present in the expected fractions. Western blots showed that Ringlb, Phcl, Cbx2, and Cbx7 fractionate with mitotic chromatin
The mitotic chromosomal association of Cbx2 is independent of PcG proteins
The Cbx-containing PRC1 complexes are recruited to interphasic chromatin via Cbx family protein interactions with H3K27me3 mediated by PRC2 (Cao et al., 2002; Wang et al., 2004b; Bernstein et al., 2006; Tavares et al., 2012; Morey et al., 2013). Therefore, we asked whether PRC2 is required for the accumulation of Cbx2 protein at mitotic chromosomes. To test the hypothesis, both YFP-Cbx2 and Cerulean-H2A fusion proteins were stably and inducibly expressed in Eed knockout (KO) ES cell lines, which lack the H3K27me3 modification. Z-scan of confocal images showed that YFP-Cbx2 fusion protein was granularly distributed at mitotic chromosomes, consistent with its distribution in wild type ES cells. Quantitative analysis of Z-stack images showed that (97 5)% of Cbx2 protein was enriched at mitotic chromosomes in Eed KO ES cell lines (Fig. 6B and 6D), consistent with YFP-Cbx2 associating with mitotic chromosomes in wild type ES cells (Fig. 6A and 6D), indicating that a core component of the PRC2 complex, Eed, is not required for the accumulation of Cbx2 protein at mitotic chromosomes.
To determine whether the formation of integral canonical PRC1 complexes is required for the accumulation of Cbx2 protein at mitotic chromosomes, YFP-Cbx2 and Cerulean-H2A fusion proteins were stably and inducibly expressed in Ringla constituently KO and Ringlb conditionally KO ES cell lines (Ringla '/Ringlb). We administrated OHT and doxycycline for
10


three days. The addition of OHT induced depletion of endogenous Ringlb protein (Fig. 7). We performed Z-scan of live cell imaging of these cells using a confocal microscope. Quantitative analysis of Z-stack images showed that (97 4)% of Cbx2 fusion protein is accumulated at mitotic chromosomes (Fig. 6C and 6D), consistent with localization of the YFP-Cbx2 fusion protein in wild type and EedKO ES cells. To further investigate whether other canonical PRC1 subunits affect the mitotic chromosomal accumulation of Cbx2, YFP-Cbx2 and Cerulean-H2A fusion proteins were stably and inducibly expressed in Bmil and Mell8 (Bmil/Mell8) double KO ES cell lines. Quantitative analysis of Z-stack images indicated that (95 3)% of Cbx2 fusion protein is accumulated at mitotic chromosomes, consistent with that in wild type ES cells (Fig. 6C and 6D). Thus, these data show that the mitotic chromosomal accumulation of Cbx2 protein is independent of PRC1 proteins Ringla/Ringlb and Bmil/Mell8.
11


A YFP-Cbx2 Cerulean-H2A Overlay ^ YFP-Cbx2 Cerulean-H2A Overlay
Figure 6. The association of Cbx2 fusion protein with mitotic chromosomes in PRC1 and PRC2 gene knockout ES cells. (A) Confocal fluorescence images of YFP-Cbx2 fusion protein expressed in PGK12.1 ES cells (WT-ES). The different version of YFP-Cbx2 of Fig. 1A is presented here in order to compare with Fig. 2B and 2C. Scale bar is 5 pm.
(B) Confocal fluorescence images of YFP-Cbx2 fusion protein expressed inAA/ES cells in various phases of mitosis. The images of prometaphase (top panel), metaphase (middle panel), and anaphase/telophase (bottom panel) are presented. Scale bar is 5 pm. (C) Confocal fluorescence images of YFP-Cbx2 fusion protein expressed in Ringin' Ringlb^77 and Bmi1" Mell8" ES cell lines in various phases of mitosis. The images of prometaphase (top panel), metaphase (middle panel), and anaphase/telophase (bottom panel) are presented. The Ringin Ringlbnfl was treated with OHT to induce depletion of Ringlb for three days before imaging. The protein level of Ringlb in Ringlet Ringlb^ cells with or without OHT was presented in Figure S5. Scale bar is 5 pm. (D) Quantitative analysis of mitotic binding fractions of YFP-Cbx2 fusion protein at mitotic chromosomes. The data represents at least ten cells analyzed. Error bars indicate standard deviation of the mean.
12


cT
j$>
/TV

Figure 7. Western blot analysis of Ringlb in RinglaA/Ringlb^1 ES cells treated with or without OHT.
Cbx2 is required for recruiting the canonical PRC1 proteins to mitotic chromosomes
Because (1) Cbx2 is the primary Cbx family protein that is accumulated at mitotic chromosomes, and (2) the mitotic chromosomal accumulation of Cbx2 protein is independent of PRC1 or PRC2 complex proteins, we reasoned that Cbx2 protein is required for recruiting canonical PRC1 proteins to mitotic chromosomes. To this end, Cerulean-PRCl protein (either Cerulean-Ringlb, or Cerulean-Phcl, or Cerulean-Mell8) and mCherry-H2A fusion protein were inducibly expressed in Cbx2'~ ES cell lines. Z-scan imaging of live cells by using confocal microscopy revealed that fluorescence intensities of Cerulean-Ringlb, Cerulean-Phcl, and Cerulean-Mell8 fusion proteins at mitotic chromosomes in Cbx2 KO ES cells was greatly reduced in comparison to that seen in wild type ES cells (compare Fig. 1 F-H and 8A). Quantitative analysis of Z-stack images revealed that (29 8)% of Cerulean-Ringlb protein associated with mitotic chromosomes, (17 7)% of Cerulean-Phcl protein, and (27 5)% of Cerulean-Mell8 protein (Fig. 8B). To test whether Cbx2 gene knockout affects the level of Phcl, Mell8 and Ringlb proteins, we performed immunoblotting using extracts from Cbx2+ + and Cbx2'~ ES cells. Western blots showed that the level of Phcl, Mell8 and Ringlb proteins in Cbx2 '~ ES cells was similar to that seen in Cbx2+ + cells (Fig. 8C). Thus, these data suggest that
13


Cbx2 protein is required for the accumulation of canonical PRC1 proteins Ringlb, Phcl, and Mel 18 at mitotic chromosomes.
A
-Q
O)
c
CC
00
oj
Cbx2*~
Cerulean- mCherry-PRC1 subunit H2A Overlay
B
Mei18 Mte*
Ponceau S staining
Figure 8. The mitotic chromosomal association of PRC1 fusion proteins Ringlb, Phcl, and Mell8 in Cbx2 A ES cells. (A) Confocal fluorescence images of Cerulea-Ringlb, Cerulean-Phcl, and Cerulean-Mell8 fusion proteins expressed in Cbx2'~ ES cells in metaphase (top panel) and anaphase (bottom panel). Scale bar is 5 pm. (B) Quantitative comparison of mitotic chromosomal association of Cerulea-Ringlb, Cerulean-Phcl, and Cerulean-Mell8 fusion proteins in Cbx2+ + and Cbx2 '~ ES cells. The data represents average of at least 10 cells analyzed. Error bars indicate standard deviation of the mean. (C) Western blots of cell extracts from Cbx2+ + and Cbx2 '~ ES cells. The Ponceau S staining indicates the loading control.
To further investigate whether Cbx2 is required for recruiting Phcl and Ringlb to mitotic chromosomes, we performed immunostaining in Cbx2+ + and Cbx2 '~ ES cells by using antibodies that detect endogenous Phcl and Ringlb (Fig. 9). Immunostaining showed that endogenous Phcl and Ringlb associate with mitotic chromosomes in wild type ES cells.
14


However, endogenous Phcl and Ringlb were excluded from mitotic chromosomes in Cbx2 knockout ES cells. These data suggest that Cbx2 is required to recruit Phcl and Ringlb to mitotic chromosomes.
Figure 9. Immunostaining of Phcl and Ringlb proteins in Cbx2+/+ and Cbx2 A ES cells.
Cbx2+ + and Cbx2 '~ ES cells were fixed and immunostained with antibodies that detect Phcl and Ringlb (green) at different phases of mitosis. DNAs were stained with Hoechst (blue). Overlay images were shown. Scale bar is 5 pm
Cbx2 directly recruits the canonical PRC1 proteins to mitotic chromosomes
Since Cbx2 protein affects the accumulation of canonical PRC1 proteins at mitotic chromosomes, we reasoned that Cbx2 protein directly recruits canonical PRC1 proteins to mitotic chromosomes. To this end, we co-expressed the three fusion proteins in Cbx2 KO ES cell lines: YFP-Cbx2, Cerulean-PRCl subunit (either Cerulean-Ringlb, or Cerulean-Phcl, or Cerulean-Mell8), and mCherry-H2A. We expected that the introduction of an YFP-Cbx2 fusion protein to the Cbx2 null background ES cells would restore the mitotic chromosomal association
15


of the three PRC1 proteins (Cerulean-Ringlb, Cerulean-Phcl and Cerulean-Mell8). We performed three-color Z-scan imaging of live cells by using confocal microscope. Quantitative analysis of Z-stack images from three ES cell lines expressing Ringlb, Phcl, and Mell8 fusion proteins showed an average of (95 7)% of YFP-Cbx2 associated with mitotic chromosomes, consistent with YFP-Cbx2 localization in wild type ES cells (Fig. 10A and IOC), indicating that mitotic chromosomal association of YFP-Cbx2 fusion protein is independent of endogenous Cbx2 protein. Notably, quantitative image analysis revealed that (63 8)% of Cerulean-Ringlb, (71 12)% of Cerulean-Phcl, and (74 10)% of Ceruelan-Mell8 also associated with mitotic chromosomes (Fig. 10A and 10D). The fraction of retention of the three fusion proteins at mitotic chromosomes in Cbx2 KO ES cell lines complemented with YFP-Cbx2 was similar to that seen in wild type ES cells, indicating YFP-Cbx2 fusion protein recruits the three canonical PRC1 fusion proteins to mitotic chromosomes.
To test whether the Cbx2 interaction with Ringlb is required for the recruitment of Ringlb protein to mitotic chromosomes, the three fusion proteins mCherry-H2A, Cerulean-Ringlb, and YFP-Cbx21-498 were expressed in Cbx2 KO ES cells. The YFP-Cbx21-498 fusion protein lacks of the Chromobox (Cbox) domain required for interaction with Ringlb (Satijn et al., 1997; Schoorlemmer et al., 1997; Bardos eta/., 2000). We expected that the Cbx2 mutant fusion protein would not be able to recruit Cerulean-Ringlb to mitotic chromosomes. Quantitative image analysis showed that (92 7)% of YFP-Cbx21-498 fusion protein was accumulated at mitotic chromosomes, indicating the deletion of the Cbox domain of Cbx2 protein does not affect its mitotic chromosomal association (Fig. 10B and 10C, also see Fig. 14B). However, only (32 8)% of Cerulean-Ringlb fusion protein associated with mitotic chromosomes (Fig. 10B and 10D). The fraction of mitotic retention of Ringblb fusion protein in
16


Cbx2 KO ES cell lines complemented with Cbx21-498 was similar to that observed in Cbx2 KO ES cells. These data indicate that the Cbx2 interaction with Ringlb is required for the recruitment of Cerulean-Ringlb fusion protein to mitotic chromosomes.
The direct recruitment of the canonical PRC1 proteins to mitotic chromosomes by Cbx2 implies that there is direct interaction between Cbx2 and PRC1 subunits at mitotic chromosomes. To test this hypothesis, we performed photobleaching fluorescence resonance energy transfer (FRET) analysis between YFP-Cbx2 and Cerulean-Ringlb at mitotic chromosomes (Fig. 10E). Fluorescence of YFP-Cbx2 fusion protein at half of mitotic chromosomes was photobleached. The ratio of fluorescence intensity of Cerulean-Ringlb in photobleached versus non-photobleached areas was calculated and compared before and after photobleaching. Quantitative image analysis indicated that the fluorescence intensity of Cerulean-Ringlb fusion protein was increased (1.5 0.1)-fold by photobleaching YFP-Cbx2 fusion protein, indicating that there is energy transfer between YFP-Cbx2 and Cerulean-Ringlb. As a control, we photobleached the YFP-Cbx21-498 fusion protein and quantified the fluorescence change of Cerulean-Ringlb fusion protein. Image analysis revealed that the fluorescence intensity of Cerulean-Ringlb fusion protein after photobleaching YFP-Cbx21-498 fusion protein was (1.1 0.2)-fold of that before photobleaching, indicating that there is no energy transfer between YFP-Cbx21-498 and Cerulean-Ringlb. Thus, these data demonstrate that YFP-Cbx2 fusion protein interacts with Cerulean-Ringlb fusion protein at mitotic chromosomes.
17


Figure 10. Directly recruiting PRC1 fusion proteins to mitotic chromosomes by YFP-Cbx2,
but not by YFP-Cbx21-498. (A) Confocal fluorescence images of Cerulean-Ring lb, Cerulean-Phcl, and Cerulean-Mell8 fusion proteins co-expressed with YFP-Cbx2 and mCherry-H2A fusion proteins in Cbx2'~ ES cells in metaphase. Scale bar is 5 pm. (B) Confocal fluorescence images of Cerulean-Ringlb fusion protein co-expressed with YFP-Cbx21-498 and mCherry-H2A fusion protein in Cbx2'~ ES cells in metaphase. Scale bar is 5 pm. (C) Quantification of mitotic chromosomal association of YFP-Cbx2 and YFP-Cbx21-498 fusion proteins in Cbx2 '~ ES cells. The data represents average of at least 10 cells analyzed. Error bars indicate standard deviation of the mean. (D) Quantitative analysis of mitotic chromosomal association of Cerulea-Ringlb, Cerulean-Phcl, and Cerulean-Mell8 fusion proteins co-expressed with YFP-Cbx2 and of Cerulea-Ringlb co-expressed with YFP-Cbx21-498 in Cbx2 '~ ES cells in metaphase. The data represents average of at least 10 cells analyzed. Error bars indicate standard deviation of the mean. (E) Photobleaching FRET images of Cerulean-Ringlb interaction with YFP-Cbx2 and YFP-Cbx21-498 at mitotic chromosomes. The YFP-Cbx2, YFP-Cbx21-498, and Cerulean-Ringlb fusion proteins as indicated above images expressed in Cbx2 '~ ES cells. Half area of fluorescence of YFP-Cbx2 or YFP-Cbx21-498 fusion proteins at mitotic chromosomes was photobleached. Z-scan imaging of live cells by confocal laser microscope was performed before (top panel) and after (bottom panel) photobleaching. The arrowheads indicate the bleaching areas.
The Cbx2-PRC1 complex is immobilized at mitotic chromosomes, but other Cbx family proteins rapidly exchange at mitotic chromosomes
Several studies have demonstrated that mammalian PRC1 proteins are highly dynamic during interphase in cells (Hernandez-Munoz etal., 2005; Ren etal., 2008; Isono etal., 2013; Vandenbunder et al., 2014). Recent studies of Drosophila Pc and Ph proteins showed that a
18


subpopulation of the two proteins bind to mitotic chromosomes with up to 300-fold longer residence time than during interphase (Fonseca et al., 2012; Steffen el al., 2013). To determine the dynamic properties of mammalian PRC1 proteins binding to chromatin in both interphase and mitosis of ES cells, we performed quantitative fluorescence recovery after photobleaching (FRAP) on the Cbx family of proteins (Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8), as well as the three core components of the canonical PRC1 complex (Ringlb, Phcl, and Mell8) (Fig. 11A-I). The mCherry-H2A fusion protein served as a guide for placing bleach spots at mitotic chromosomes. Comparison of recovery kinetics of the Cbx family fusion proteins binding to mitotic chromosomes revealed striking differences among Cbx proteins. Over 90% of YFP-Cbx2 fusion protein was immobilized at mitotic chromosomes without exchange over a time period of 120 seconds. Conversely, over 85% of the YFP-Cbx4, YFP-Cbx6, YFP-Cbx7 and YFP-Cbx8 fusion proteins rapidly exchanged at mitotic chromosomes, with a residence time of 10-15 seconds (Fig. 11 A-l IF and 11 J-l IK). During interphase, over 90% of the Cbx family fusion proteins showed fluorescence recovery with a residence time of about 10-20 seconds (Fig. 11 A-l IF and 11J-1 IK), which is consistent with previous studies (Ren et al., 2008). Thus, these data reveal that the YFP-Cbx2 fusion protein stably binds to mitotic chromosomes but rapidly exchanges at interphasic chromatin, while other Cbx family proteins dynamically exchange on both interphasic and mitotic chromatin.
19


A _______________Metaphase: YFP-Cbx2_____________ ______________________Interphase: YFP-Cbx2
Pre Post
o OS' 10 s' 60s' 120 s'
Pre Post
O 0 s' 10 s S' 120 S'
I 1 Cerulean-Ringlb
! 0.5-
I 0.0
- Interphase
- Metaphase

60
Time (s)
: 1.0
I 0.5-
YFP-Cbx8
0.0
niinimiiiiiuu
- Interphase
- Metaphase
Cerulean-Phc1
i 1.0-
I 0.5-
! 0.0
120
60 120 Time (s)

i Interphase t Metaphase
0 60 120 Time (s)
|1H
5
Cenjlean-Mel18
0.5-
0.0

- Interphase
- Metaphase
0 60 120
60 Time (s)
Figure 11. FRAP analysis of PRC1 fusion proteins binding to interphasic and mitotic chromatins. (A) Representative FRAP images of YFP-Cbx2 fusion protein at metaphasic and interphasic chromatins of ES cells. The images were taken before (pre) and after (post) photobleaching. The bleaching area is indicated and outlined in white. (B-I) FRAP curves of PRC1 fusion proteins at interphases and metaphases of PGK12.1 ES cells. The FRAP curves are the normalized fluorescence intensities of the bleached areas as a function of time after photobleaching and are average of at least 8 cells. Error bars indicate the standard deviations of means. (J) Immobile faction of PRC 1 fusion protein at interphasic and metaphasic chromatins. The immobile fraction was calculated from FRAP curves by fitting first-order kinetic model. The dash line is used to indicate the contrast between interphase and metaphase. The data are average
20


of at least 8 cells. (K) Residence time of PRC 1 fusion protein at interphasic and metaphasic chromatins. The residence time was calculated from FRAP curves by fitting first-order kinetic model. The bar with NA indicates that the fusion protein is immobilized at mitotic chromatin, thus the residence time is not measureable within time scale of experiments. The dash line indicates the contrast between interphase and metaphase.
To ask whether the binding kinetics of the Cbx2 fusion protein is cell type-specific, we performed FRAP analysis of YFP-Cbx2 fusion protein in both HeLa and HEK293 cells (Fig. 12, and also see Fig. 14C). Analysis of FRAP curves revealed that interphasic and mitotic dynamic properties of YFP-Cbx2 fusion protein in both cell lines were similar to that seen in ES cells. Thus, these data demonstrate that the YFP-Cbx2 fusion protein possesses inherently different properties of interaction with interphasic versus mitotic chromatin. The various dynamic binding properties of Cbx family members to interphasic and mitotic chromatin prompted us to explore other core components of the canonic PRC1 complex. We performed FRAP analysis of Cerulean-Ringlb, Cerulean-Phcl and Cerlean-Mell8 fusion proteins bound to chromatin in both interphase and mitosis in ES cells. Analysis of FRAP curves of mitotic chromosomal binding of Cerulean-Ringlb, Cerulean-Phcl and Cerlean-Mell8 fusion proteins revealed that near 80% of these fusion proteins stably bind to mitotic chromosomes without exchange (Fig. 11G-K), which is similar to YFP-Cbx2, but differs from other Cbx family fusion proteins. Calculation of the recovery kinetics and measured parameters of interphasic FRAP curves of the Cerulean-Ringlb, Cerulean-Phcl and Cerlean-Mell8 proteins revealed that these three fusion proteins showed complete recovery, with residence time of 10-15 seconds, and had no immobile fraction at interphasic chromatin. Thus, these data indicate that the four PRC1 fusion proteins, YFP-Cbx2, Cerulean-Ringlb, Cerulean-Phcl, and Cerulean-Mell8, bind to mitotic chromosomes with similar kinetic characteristics, yet differ from other Cbx family proteins.
21


CD
§1H
HeLa: YFP-Cbx2
; 0.5-
! 0.0

Interphase
Metaphase

60
Time (s)
120
HEK293: YFP-Cbx2
Figure 12. FRAP analysis of YFP-Cbx2 fusion protein binding to interphasic and mitotic chromatins in HeLa and HEK293 cells. The FRAP curves are the normalized fluorescence intensities of the bleached areas as a function of time after photobleaching and are average of at least 8 cells. Error bars indicate the standard deviations of means.
The immobilization of Cbx2 at mitotic chromosomes is independent of PcG proteins
Since Cbx2 protein is recruited to mitotic chromosomes by a PRC2-independent mechanism, we reasoned that depletion of PRC2 complex gene Eed would not affect the immobilization of Cbx2 on mitotic chromosomes. To test the hypothesis, we performed FRAP analysis of YFP-Cbx2 protein binding to both mitotic and interphasic chromosomes mEedYJd ES cells. Analysis of mitotic FRAP curves revealed that over 85% of YFP-Cbx2 fusion protein showed no recovery of fluorescence within 120 seconds (Fig. 13A and 13D). Calculation of interphasic FRAP curves of YFP-Cbx2 fusion protein revealed that the residence time for the mobile fraction is 25 seconds, slightly higher than that observed in wild type ES cells, while the immobile fraction is the same as seen in wild type ES cells. Thus, these data indicate that the dynamics of YFP-Cbx2 fusion protein binding to interphase and mitotic chromatin is independent of the PRC2 gene Eed.
To interrogate whether the interphasic and mitotic kinetics of Cbx2 fusion protein binding to chromatin are affected by PRC1 proteins, we performed FRAP analysis of YFP-Cbx2 fusion protein in Ringlet Ring lb double KO and Bmill Mel 18 double KO ES cells. Analysis of mitotic FRAP curves in both double KO ES cell lines revealed that more than 85% of YFP-Cbx2
22


showed no recovery of fluorescence, indicating that the YFP-Cbx2 fusion protein binds to mitotic chromosomes without exchange (Fig. 13B- 13D). Next, we analyzed the interphasic FRAP curves of YFP-Cbx2 fusion protein in the double KO ES cell lines. Calculation of residence time and immobile fraction of YFP-Cbx2 fusion protein in interphase of the two double KO ES cell lines revealed striking differences in comparison to wild type ES cell (Fig. 13B-13D). The YFP-Cbx2 fusion protein was much less dynamic in the double KO ES cells than in wild type ES cells, with residence time 30-35 seconds. A further difference is that the immobile fraction of YFP-Cbx2 fusion protein in interphase of the double KO ES cell lines was 23-30% of total protein, which is over 2.0-fold of that seen in wild type ES cells. Thus, these data suggest that the immobilization of YFP-Cbx2 fusion protein at mitotic chromosomes is independent of the PRC1 proteins Ringla/Ringlb and Mell8/Bmil, while the dynamic behavior of YFP-Cbx2 during interphase of the cell cycle can be affected by the PRC1 proteins.
A Eed-'- C Bmi1//Mel18-/-
.1.0-,
Ring 1 a'/Ringlb'-
Interphase
Metaphase
= 0.5-
'0.0-
MiWii
60
Time (s)
120
!o
o
E
E
0
ill
co i,' W §;
k i iS S s ?? I I t UJ O) O
oiin $ 2 i goa s
23


Figure 13. FRAP analysis of YFP-Cbx2 fusion protein binding to chromatins in Ee(TA,
Ring 1 a A/Ring 1 b~A, and 11 mi 1 /Mel 18A ES cells. (A-C) FRAP curves of interphases and metaphases of YFP-Cbx2 fusion protein in lied ~, Ring la A/Ring lb A, and BmilA/Mell 8"" ES cells. The FRAP curves were normalized and plotted described as in Fig. 5. Over 8 cells were analyzed. Error bars indicate the standard deviations of means. (D) Residence time and immobile fraction of YFP-Cbx2 on chromatins. The residence time and immobile fraction was calculated from FRAP curves by fitting first-order kinetic model. The bars with NA indicate that the residence time is not available due to the fact of immobilization of YFP-Cbx2 on mitotic chromosomes. The dash lines indicate the contrast between interphase and metaphase. WT-ES is denoted as PGK12.1 ES cells. The data are average of at least 8 cells.
The recruitment and immobilization of Cbx2 to mitotic chromosomes requires its distinct regions
To dissect the domains (regions) of Cbx2 required for targeting mitotic chromosomes, we generated a variety of Cbx2 mutants tagged with YFP and introduced them into HeLa cells. The mCherry-H2A protein was used to mark mitotic chromosomes. Imaging of live cells by using a confocal fluorescence microscope showed that the deletion of the C-terminus of Cbx2 protein (Cbx21-498, Cbx21-281, and Cbx21-194) does not affect targeting of the Cbx2 fusion mutants to mitotic chromosomes (Fig. 14A and 14B). On the other hand, deletion of the N-terminus of Cbx2 (Cbx289'532) resulted in a complete loss of the Cbx2 fusion variant from mitotic chromosomes. These data suggest that the N-terminus of Cbx2 protein is required for targeting the Cbx2 fusion protein to mitotic chromosomes.
To explore the molecular basis for the immobilization of Cbx2 at mitotic chromosomes, we performed FRAP assay of the YFP-Cbx2 fusion variants (Cbx21-498, Cbx21-281, and Cbx21-194) during interphase and mitosis of the cell cycle. Kinetic analysis of interphasic FRAP curves of YFP-Cbx2 fusion variants revealed that the residence time of three YFP-Cbx2 variants was 6-9 seconds, which is half of full-length Cbx2 fusion protein. In contrast to the existence of immobile fraction of Cbx2 fusion protein in interphase cells, there was no immobile fraction for the three YFP-Cbx2 fusion variants. Analysis of mitotic FRAP curves of the three YFP-Cbx2
24


variants showed striking kinetic differences in comparison to the YFP-Cbx2 fusion protein. The three YFP-Cbx2 variants rapidly exchanged at mitotic chromosomes, with residence time of 8-11 seconds. In contrast to full-length YFP-Cbx2, YFP-Cbx2 variants became fully recovered at mitotic chromosomes. Thus, these data indicate that the immobilization of Cbx2 fusion proteins at mitotic chromosomes require its extreme C-terminus.
25


A
C
1.0
i
Targeting
CHD (11-63)
Immobilizing
CbOK (503-51) {
ATH (74-86)'

B YFP-Ct2 mChrry
and mutants -H2A Overlay
20.0
0 60 120
T-----'-----I--------1
0 70 140
Tima (s)
Figure 14. Analysis of structural elements of YFP-Cbx2 fusion protein required for its targeting and immobilizing. (A) Diagram of structural domains of Cbx2. The dash rectangles indicate the region required for targeting Cbx2 to mitotic chromosomes (left) and the region required for immobilizing Cbx2 at mitotic chromosomes (right). CHD represents Chromodomain domain. Cbox is Chromobox domain and ATH is AT-hook domain. The number in parentheses indicates the starting and ending of amino acid sequence. (B) Confocal images of YFP-Cbx2 and its variant fusion proteins in metaphase of HeLa cells. YFP-Cbx2 mutant and mCherry-H2A fusion proteins were stably expressed in HeLa cells. The mCherry-H2A was used to mark
26


mitotic chromosomes. Scale bar is 5 pm. (C) FRAP curves of interphases and metaphases of YFP-Cbx2 and its variant fusion proteins expressed in HeLa cells. FRAP analysis was described in Fig. 5. (D) Residence time and immobile fraction of YFP-Cbx2 variant fusion proteins at interphasic and mitotic chromatins. The residence time and immobile fraction were calculated from FRAP curves by fitting first-order kinetic model.
27


CHAPTER III
DISCUSSION
We have used quantitative live cell imaging analysis to investigate the mitotic chromosomal association of the canonical PRC1 proteins and to interrogate the dynamics of these proteins binding to chromatin in both interphase and mitosis. Our results revealed several striking findings, summarized as follows: (1) The canonical PRC1 subunits tested vary at the level of association with mitotic chromosomes, and Cbx2 is the primary protein accumulated at mitotic chromosomes; (2) The mitotic chromosomal association of Cbx2 protein is independent of PRC 1 or PRC2 complex proteins; (3) The Cbx2 protein directly targets the canonical PRC1 proteins to mitotic chromosomes; (4) The Cbx2-containing PRC1 complex is immobilized at mitotic chromosomes, while other Cbx family proteins dynamically exchange at mitotic chromosomes; (5) The immobilization of Cbx2 protein at mitotic chromosomes is independent of PRC1 or PRC2 proteins; (6) The recruitment of Cbx2 protein to mitotic chromosomes requires its N-terminus, while the immobilization of Cbx2 protein at mitotic chromosomes requires its C-terminus. Thus, these data provide insights into the mechanisms underlying how canonical PRC1 proteins interact with interphasic and mitotic chromatin, and also have implications for understanding PRC 1-mediated epigenetic inheritance.
Early studies of mammalian PRC1 proteins by immunofluorescence in fixed cells have provided divergent opinions as to whether PRC1 proteins are substantially retained at mitotic chromosomes (Wang etal., 1997; Saurin etal., 1998; Koga et al., 1999; Voncken etal., 1999; Akasaka et al., 2002; Suzuki et al., 2002; Miyagishima et al., 2003; Aoto et al., 2008; Vincenz and Kerppola, 2008). These variations could be due to inaccessibility of the protein to the antibody, may be due to damage or loss of epitopes during the experimental procedures, or
28


differences in how cells were prepared and imaged, as well as the cell types that were used. Consistent with these notions, we noticed that there were remarkable differences in the mitotic chromosomal association of PRC 1 proteins if subtle experimental variations were applied. For instance, by adding Hoechst to cells before fixing with formaldehyde, we observed that Cbx family proteins are completely excluded from mitotic chromosomes. By fixing cells with formaldehyde before adding Hoechst, we observed that Cbx2 protein now shows a punctate pattern at mitotic chromosomes (data not shown). Thus, we performed quantitative live cell imaging to interrogate mitotic chromosomal association and chromatin binding of PRC 1 proteins. The quantitative live cell imaging requires that PRC1 proteins are fused with fluorescence proteins. Many PRC1 fusion proteins have been reported to function normally in cells or animals. The Cbx family proteins fused with Venus have been documented to be able to form PRC1 complex and bind to PcG target genes (Ren et al., 2008; Ren and Kerppola, 2011). The knockin mice expressing Mel 18 and Ringlb proteins fused with EGFP or YFP have been shown to function normally as their endogenous counterparts (Isono et al., 2013). Drosophila Ph and Pc proteins fused with GFP can fulfill the functions of the endogenous proteins (Fonseca et al., 2012). These data suggest that the PRC1 proteins can tolerate the addition of fluorescence protein tag. The inducible gene deliver vector used in the current studies allows us to control the level of fusion protein expression. The protein expression level of the test fusion protein under doxycycline concentration used is similar or slightly higher than endogenous counterparts. The immunostaining of endogenous Ringlb and Phcl in wild type and Cbx2 knockout ES cells also support that PRC1 proteins associates with mitotic chromosomes and that Cbx2 affects the mitotic chromosomal association of PRC 1 proteins.
29


Although Cbx family proteins share conserved domains (Cbox and CHD) (Simon and Kingston, 2009), accumulating evidence suggests that they have both overlapping and nonoverlapping functions (Core eta/., 1997; Katoh-Fukui eta/., 1998; Vincenz and Kerppola, 2008; Forzati eta/., 2012; Gao eta/., 2012; Morey eta/., 2012; Klauke eta/., 2013). Cbx2 accumulates at mitotic chromosomes, yet other Cbx family proteins show greatly reduced association with mitotic chromosomes. Deletion of the CHD domain causes dissociation of Cbx2 variants from mitotic chromosomes, suggesting that the CHD domain plays a role in recruiting Cbx2 to mitotic chromosomes. Other Cbx family proteins also contain a CHD domain, but display a much reduced association with mitotic chromosomes, indicating that other unknown factors must also contribute to the unique binding properties of Cbx2. These factors may include post-translational modifications of Cbx2, which could lead to a switch in binding platform. A previous report showed that phosphorylation of Cbx2 changes its binding specificity for methylated histone H3 (Hatano et al., 2010). Another recent report also indicated that methylation of Cbx4 switches its binding partners (Yang et al., 2011). Another possibility is that Cbx2 protein on its own has unique physical properties, for example, the intrinsic charge properties (Grau et al., 2011). It is also possible that the accumulation of Cbx2 proteins at mitotic chromosomes is due to changes of recruiting or competing molecules. Finally, Cbx2 may form unique protein complexes at mitotic chromosomes. Further studies will help understanding mechanisms by which mitotic Cbx family proteins are selectively displaced and retained.
We provided several lines of evidence to demonstrate that Cbx2 protein is essential for the recruitment of the canonical PRC1 proteins to mitotic chromosomes. First, we observed that the mitotic fraction of the three PRC1 proteins Ringlb, Phcl, and Mell8 in Cbx2 KO ES cells are reduced at least two-fold in comparison to that observed in wild type ES cells, suggesting
30


that Cbx2 plays a major role in recruiting PRC1 proteins to mitotic chromosomes. Second, the mitotic chromosomal association of the three PRC1 proteins, Ringlb, Phcl, and Mell8, in Cbx2 KO ES cells can be restored by supplementing YFP-Cbx2 fusion protein, but not the YFP-Cbx21_ 498 fusion protein that is unable to interact with Ringlb (Satijn et al., 1997; Schoorlemmer el al., 1997; Bardos et al., 2000). Finally, we observed that the YFP-Cbx2 fusion protein interacts with the Cerulean-Ringlb fusion protein at mitotic chromosomes by FRET imaging. Taken together, these data reveal that Cbx2 directly recruits canonical PRC1 proteins to mitotic chromosomes.
It is interesting to note that the Cbx2-containing PRC1 complex (Cbx2, Ringlb, Phcl, and Mel 18) is immobilized at mitotic chromosomes without exchange, whereas other Cbx family proteins (Cbx4, Cbx6, Cbx7, and Cbx8) dynamically bind to mitotic chromosomes with kinetics similar to their binding to interphasic chromatin. It is not clear which factors dictate the transition between a dynamic and a stable Cbx2-PRC1 complex during different phases of the cell cycle. Since the C-terminus of Cbx2 is required for the immobilization of Cbx2 proteins at mitotic chromosomes, the C-terminus may dictate the dynamic switching between interphase and mitosis. The C-terminus contains the Cbox domain that interacts with Ringlb (Satijn et al., 1997; Schoorlemmer et al., 1997; Bardos et al., 2000), but depletion of Ringla/Ringlb proteins did not alter the immobilization of Cbx2 to mitotic chromosomes, suggesting that other factors play roles in immobilizing the Cbx2-PRC1 complex at mitotic chromosomes. Previous studies of transcription factors, epigenetic regulators and chromosomal structural proteins have shown that most of them either rapidly exchange at or stably bind to chromatin (Phair et al., 2004; Cherukuri et al., 2008; Ueda et al., 2008; Souza et al., 2009; Hemmerich et al., 2011), yet a subset of these factors switch binding dynamics upon signaling stimuli or cell cycle transition (Angus etal., 2003; Schmiedeberg etal., 2004; Chen et al., 2005; Mekhail etal., 2005; Gerlich etal., 2006;
31


Meshorer et al., 2006; Yao et al., 2006; Ren et al., 2008; Giglia-Mari el al., 2009; Hellwig el al., 2011; Hemmerich et al., 2011). We hypothesize that the dynamic switching of the Cbx2-PRC1 complex between interphase and mitosis may be regulated through covalent modifications or additional interacting partners (Fig. 15).
The Ringrose laboratory (Fonseca et al., 2012) and the Francis laboratory (Follmer et al., 2012) identified a fraction of drosophila PRC1 proteins association with mitotic chromosomes and the Ringrose laboratory (Fonseca et al., 2012) also revealed that 0.2-2% of PRC1 proteins (PC and PH) remains stably bound to mitotic chromatin with up to 300-fold longer residence times than in interphase, which supports our findings of the mitotic chromosomal association and the stably binding to mitotic chromosomes of mammalian PRC1 proteins. All mammalian PRC1 proteins test in our research have the capacity of stably binding to mitotic chromosome, however the faction of their stably binding to mitotic chromosomes varies greatly. We revealed that over 85% of the Cbx2-PRC1 complex is selectively and specifically immobilized at mitotic chromosomes, while more than 85% of other Cbx family proteins dynamically exchanges at mitotic chromosomes. Since mammalian PRC1 complexes comprise a multiplicity of variants and are far more biochemically diverse than their drosophila counterparts, the selective immobilization of the Cbx2-PRC1 complex at mitotic chromosomes implies that the PRC1 complexes become functionally divergent during evolution.
32


Ringib Phc1
Figure 15. A hypothetic model for the interaction of canonical PRC1 complex with interphasic and mitotic chromatin. In interphases of cells, Cbx2-PRC1 and Cbx4/6/7/8-PRCl complexes dynamically bind to chromatin. During mitosis, the Cbx2-PRC1 complex is immobilized at mitotic chromosomes, whereas other Cbx family (Cbx4, Cbx6, Cbx7, and Cbx8) rapidly exchange at mitotic chromosomes. Red star implies factors such as covalent modification and protein interactor stabilize the Cbx2-PRC1 complex binding to mitotic chromosomes.
33


CHAPTER IV
MATERIALS AND METHODS
Cell lines
The Cbx2~~ (Katoh-Fukui et al., 1998), Ringlet I Ring I b,l,1\ Rosa26::CreERT2 (Ringla knockout, Ring lb conditional knockout) (Endoh etal., 2008), BmiRTMellR" (Bmil and Mel 18 double knockout) (Elderkin et al., 2007), Eed~ (Endoh et al., 2008), and PGK12.1 mouse ES cell (Penny etal., 1996) lines were maintained in DMEM (Sigma) supplemented with 15% FBS (BioExpress), 2 mM glutamine (Life Technologies), 100 units/ml penicillin G sodium (Life Technologies), 0.1 mg/ml streptomycin sulfate (Life Technologies), 0.1 mM P-mercaptoethanol (Life Technologies), 103 units/ml leukemia inhibitor factor (LIF), and 0.1 mM non-essential amino acids (Life Technologies) at 37 C in 5% CO2. The depletion of Ringlb alleles mRingld ~ IRingllRosa26: :CreERT2 ES cells was achieved by administration of 4-Hydroxytamoxifen (OHT, Sigma) for three days under the concentration of 1.0 pM prior to experiment. HeLa, HEK293, and HEK293T cells were maintained in DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/ml penicillin G sodium, 0.1 mg/ml streptomycin sulfate at 37 C in 5%
CO2. Culture medium was replaced with fresh medium every 24 hours and cells lines were split at 66-75 % confluence during maintenances.
Plasmids
The pTRIPZ shRNAmir Lentivirus vector (Open Biosystems) was engineered to remove both the turboRFP and the regulatory sequences of shRNAmir to produce pTRIPZ(M). The sequences coding Cerulean (Addgene), YFP (Ren etal., 2008), and mCherry (Addgene) fluorescence proteins were amplified by PCR and inserted to the pTRIPZ(M) to produce vectors pTRIPZ(M)-Cerulean, pTRIPZ(M)-YFP, and pTRIPZ(M)-mCherry. The sequences coding
34


Ringlb (Ren etal., 2008), Phcl (Addgene), Mell8 (Addgene), H2A (Addgene), Cbx2 (Ren et al., 2008), Cbx4 (Ren et al., 2008), Cbx6 (Ren et al., 2008), Cbx7 (Ren et al., 2008), and Cbx8 (Ren et al., 2008) were amplified by PCR and inserted downstream of the coding sequence of fluorescence protein in pTRIPZ(M) vector. The same strategy was used to construct Cbx2 variants tagged with YFP. The Cbx2 variants were as follows: (1) Cbx21-498, deletion of amino acids (499-532); (2) Cbx21-281, deletion of amino acids (282-532); (3) Cbx2M94, deletion of amino acids (195-532); (4) Cbx289'532, deletion of amino acids (1-88). The sequences encoding fusion proteins have been verified by DNA sequencing.
Generation of stable cell lines by lentivirus infection
HEK 293T cells density of 3.5-4.0 x 106 were seeded on 10 cm culture dish 24 hours prior at the time of transfection to reach a monolayer of 90 % confluence. Cells were cotransfected by calcium phosphate precipitation. Transfection particles are composed of 21 pg pTRIPZ(M) containing the gene of interest, 21 pg psPAX2, and 10.5 pg pMD2.G were first mixed by vortex at max speed for 30 seconds follow the by addition of cold CaCh drop wise at a final concentration of .128 M and vortex at max speed for 40 seconds. Cold 2xHBSS was added drop wise at 3 seconds interval to the mixture while under slow vortex speed. Final mixture was vortex at max speed for 40 seconds and was incubated at room temperature for 30 mins undisturbed. After incubation, mixture was vortex for 20 seconds at max speed and added drop wise to cell plate. After 12 hours of transfection, culture medium was replaced with fresh DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/mL pencillin G sodium, and 0.1 mg/mL streptomycin sulfate for 48 hours of incubation. Pseudovirus medium was collected and centrifuged at 1,600 g for 10 min at 4 degree. The supernatants were collected and used for infection of mouse ES, HeLa, and HEK293 cells. If co-expression of multiple proteins needed,
35


lentiviruses were produced separately and mixed at the time of infection. Polybrene (sigma) was added at the final concentration of 8 pg/mL and the cells were seeded at -15% confluence on gelatin-coated plates or mitotically inactivated MEF cells. 16 hours after transduction, the medium was replaced with fresh medium. After 2 medium changes at 24 hours interval, infected cells were selected with 1.0-2.0 pg/ml puromycin (Life Technologies). The expression of transgenes was induced with doxycycline (Sigma) at concentration of 0.1-1.0 pg/ml.
Confocal microscope imaging of live cells and quantification of mitotic fraction
Zeiss LSM 700 observer Z1 equipped with a 100 x oil objective (numerical aperture, 1.4) and an EMCCD camera was used for Z-scan imaging. For Cerulean fluorescence, 435 nm excitation and 476 nm emission filters were used. For YFP fluorescence, 514 nm excitation and 527 nm emission filters were used. For mCherry fluorescence, 587 nm excitation and 610 nm emission filters were used. The section size was 1 pm for the three fluorescence proteins Cerulean, YFP and mCherry. The frame size was 512 x 512 pixels. Scan time is 1.56 sec. Average of images was 4. For live cell imaging, cells were incubated with doxycycline (0.1-1.0 pM) to induce protein expression. 48 hours after induction of protein expression, cells were seed to overnight gelatin-coated cover glass dish (Mat Teck Corp) in the presence of doxycycline.
One day after seeding, medium was replaced with either Ring buffer (155 mM NaCl, 5 mM KC1, 2 mM CaCl2, 1 mM MgCl2, 2 mM NaH2P04, 10 mM HEPES, 10 mM glucose, pH7.2) or phenol-free DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/ml penicillin G sodium, 0.1 mg/ml streptomycin sulfate. Cells were maintained at 37 C using Heater controller (Model TC-324, Warner Instrument Corp) during imaging. The grayscale images were converted into pseudo-color, merged, and cropped using Adobe Photoshop.
36


The Z-stack movies were exported as individual images using Zeiss Zen software. The intensity of each imaging section was quantified using Image J software. The mean fluorescence intensities of a region of interest corresponding to the mitotic chromosomes of the metaphase plate marked with H2A were measured. The mean fluorescence intensities of cell nucleus were also measured. The fluorescence intensities of a region without cells were measured as background. The fluorescence values were sum of intensities of individual sections of 3-D stack. The mitotic fraction RM was calculated asRM = (1^ Ib) / (Inucleus -Ib), where Imitosis is the
fluorescence intensities of PRC 1 fusion protein at mitotic chromosomes; Inudeus is the
fluorescence intensities of PRC 1 fusion protein in cell nucleus; Ib is the background
fluorescence intensities corresponding to region without cells.
FRAP imaging and quantification
FRAP imaging was performed using Zeiss LSM 700 observer. Cells were maintained as described in confocal laser scanning imaging of live cells. The expression of fusion proteins was induced by 0.1-0.2 pM doxycycline for 2 days. The pinhole is fully open for FRAP imaging. The scan speed is 1.56 sec. Images were taken without average. Before photobleaching, four images were taken. Immediately after photobleaching, 30 images were taken with 5 sec intervals. The images were analyzed and fluorescence intensities were quantified using Image J software. To correct for movement in the x-y plane, the images were aligned using TurboReg. The fluorescence intensities were corrected for fluctuations in background and total signal and normalized to the signal prior to bleaching to obtain the fluorescence recovery (IR) as described previously (Ren et al., 2008). IR was plotted as a function of time (t) after bleaching. FRAP
curves were fitted by one binding state kinetic model IR = 1 m- where m is the mobile
37


fraction. The immobile fraction was calculated as (1 ifi). The residence time was calculated as
(1 Ik).
FRET imaging
FRET measurements were performed using Zeiss LSM 700 observer. Two images were acquired in the same field of view in the Cerulean-Ringlb (donor) and YFP-Cbx2 (donor) channels. Half area of YFP-Cbx2 at mitotic chromosomes was bleached with 514 nm laser, and a second set of images of Cerulean-Ringlb and YFP-Cbx2 were acquired. The FRET ratio was
/1post-bleach \ iA post-nonbleach \
calculated as RF = (y---------------)/(y---------------)
L
n /R nt*n /
fluorescence intensities of Cerulean-Rinbglb after and before bleaching at the bleached half area, respectively; Ipost-nonbieach and Ipre-nonbieach are the mean fluorescence intensities of Cerulean-Rinbglb after and before bleaching at the nonphotobleached half area, respectively. Epifluorescence imaging of live cells
The images were acquired using an Axio Observer D1 Microscope (Zeiss) equipped with a 100 x oil objective (numerical aperture, 1.4) and an EMCCD camera. For Cerulean fluorescence, 438/24 nm excitation and 483/32 nm emission filters were used. For YFP fluorescence, 500/24 nm excitation and 542/27 emission filters were used. For mCherry fluorescence, 560/10 nm excitation and 610/35 nm emission filters were used. For Hoechst fluorescence, 387/11 nm excitation and 447/60 emission filters were used. For live cell imaging, cells were maintained as described above. Images were presented as described above.
Cell synchronization and fractionation
ES cells synchronization was carried out as described in previous report (Ballabeni etal., 2011). Briefly, ES cell were cultured in the presence of 1.25 mM thymidine (Sigma, T1895-1G)
38


for 14 hrs. After removal of medium and washing with PBS, to the plate fresh ES cell medium with 200 ng/ml nocodazole (Sigma, M1404-2MG) was added and cells were cultured for 7 hrs. Cells were harvested and washed with PBS followed by chromatin isolation or immunofluorescence.
Chromatin fractionation was performed as in (Mendez and Stillman, 2000; Follmer eta/., 2012) with minor modifications. To prepare total cell extracts (TCE), synchronized (mitotic) and non-synchronized (control) cells were directly resuspended in Laemmli buffer followed by sonication. To isolate chromatin, control and mitotic cells were resuspended in (5 x 107 cells/ml) in buffer A (10 mM HEPES, pH7.9, 10 mM KC1, 1.5 mM MgCl2, 0.34 M Sucrose, 10% glycerol, 0.1% Triton-X 100, protein inhibitors (Sigma, P8340), 0.2 mM PMSF, 1.0 mM DTT), and incubated on ice for 5 min. Nuclei (PI) was collected by centrifugation (1,300 g, 4 min, 4 C). The supernatant (SI) was centrifuged (13,000 g, 15 min, 4 C) to give supernatant (S2) and pellet (P2). PI was washed once with buffer A and incubated with buffer B (3 mM EDTA, 0.2 mM EGTA, 0.2 mM PMSF, 1.0 mM DTT and protein inhibitors (Sigma, P8340)). The sample was centrifuged (1,700 g, 4 min, 4 C) to give supernatant (S2) and pellet (P3). P3 was resuspended in Laemmli buffer followed by sonication. The fractions were run on 4-12% SDS-PAGE gels, and immunoblotted.
Immunoblotting
To quantify protein level, cell culture medium was aspirated and was washed with PBS. Citric saline buffer was added to cell and incubated for 10 min for detachment. Buffer was collected with cells for centrifuge at 500 g to pelleted. After centrifuge, the supernatant was removed and was re-washed with PBS. Cell pellet was lysed with buffer (20 mM Tris-HCl, pH 7.4, 2.0% NP-40, 1.0% Triton X-100, 500 mMNaCl, 0.25 mM EDTA, 0.1 mMNa3V04, 0.1
39


mM PMSF, and protease inhibitors (Sigma, P8340)). Lysate was rocked for 30 min and centrifuge with 15000 g for 20 min at 4 C. After centrifugation, supernatant was collected. The protein concentration was quantified and normalized to the same concentration. Dithiothreitol (sigma) and SDS-loading, Laemmli buffer (BioRad), was added to protein and heated at 90 C for 10 min.
Proteins were separated using SDS-PAGE (Novex, NP0322) gel at 120 volts, 400 amp, and 120 min. Separated proteins were transferred to 0.45 pm Immobilon-FL PVDF membrane (Millipore) by Transfer blot (bioRad, SD cell) at 20 volt, 400 amp, and 60 min. PVDF membrane incubated in blocking buffer compose of PBS (sigma) solution supplemented with 5 % non-fat milk (Labscientific, M0841) and .1 % triton x-100 (Sigma, T8787) over night at 4 C. PVDF membrane is washed with PBS + 1 % triton x-100 (Sigma, T8787) 3 times at room temperature with rocking for 20 min each. PVDF membrane probed with anti-Cbx2 (Abeam, ab80044), anti-Phcl (Active motif, 6-1-3), anti-Ringlb (MBL, D139-3), and anti-Mell8 (Santa Cruz, sc-10744) were diluted in anti-body incubation buffer contain 50 % PBS + 1 % triton x-100 (Sigma,
T8787) solution and 50 % Odyssey blocking buffer (LI-COR, 927-4000) for overnight at 4 C with gently rocking. After over night incubation, PVDF membrane is washed with PBS + .1 % triton x-100 (Sigma, T8787) 3 times at room temperature with rocking for 20 min each. Secondary anti-rabbit or mouse diluted in anti-body buffer and incubate with PVDF for one hour at room temperature with gentle rocking. Proteins were detected using ECL Plus detection reagents (GE Healthcare). Membranes were imaged using a ChemiDoc XRS system (BioRad).
Immunofluorescence
Wild-type and Cbx2 knockout ES cells were plated on overnight gelatin coverslips and cultured for 24 hrs. Cell were washed with PBS and fixed with 1.0% paraformaldehyde for 10
40


min. After fix, cells are washed 2 times with PBS and incubated with 0.2% Triton X-100 for 10 min. Two more wash with PBS were performed and cells were incubated with blocking buffer (basic blocking buffer plus 3% goat serum and 3% BSA) for 1 hour at room temperature. After incubation, cells are rinsed with basic blocking buffer (10 mM PBS, pH 7.2, 0.1% Triton X-100, 0.05% Tween 20) 3 times for 5 min each. Follow by wash; cells were incubated with primary antibody that was diluted in blocking buffer for 2 hours at room temperature. After incubation, cells were rinsed with basic blocking buffer 2 times for 5 mins each and incubate with secondary antibody diluted in blocking buffer for 1 hour in dark. Cells were rinsed with PBS 3 times and basic blocking buffer 2 times for 5 min each. Cells are stained with PBS + 1 ug/ml Hoechst for 10 min and rinsed with PBS 2 times in the dark. Coverslip is mounted on slide with ProLong Antifade reagents (Life Technologies). The primary antibodies were used as follows: anti-Phcl (Active motif, 6-1-3) and anti-Ringlb (MBL, D139-3). The primary antibodies were detected using FITC labeled goat anti-mouse antibodies (Sigma).
To immunostain synchronized cells, after trypsinization, mitotic cells were collected by centrifugation and washed with PBS. Cells were spun onto glass slides at 1,000 rpm for 10 minutes in a Shandon Cytospin 2. Cells were fixed with 2.0% formaldehyde at room temperature for 10 min and immunostained as described above. The primary antibody is anti-Histone H3 (phospho S10) antibody (Abeam, ab5176) and was detected using FITC labeled goat anti-rabbit antibodies (Sigma)
41


REFERENCES
Akasaka, T., Takahashi, N., Suzuki, M., Koseki, H., Bodmer, R., and Koga, H. (2002). MBLR, a new RING finger protein resembling mammalian Polycomb gene products, is regulated by cell cycle-dependent phosphorylation. Genes Cells 7, 835-850.
Angus, S.P., Solomon, D.A., Kuschel, L., Hennigan, R.F., and Knudsen, E.S. (2003).
Retinoblastoma tumor suppressor: analyses of dynamic behavior in living cells reveal multiple modes of regulation. Mol Cell Biol 23, 8172-8188.
Aoto, T., Saitoh, N., Sakamoto, Y., Watanabe, S., and Nakao, M. (2008). Polycomb group
protein-associated chromatin is reproduced in post-mitotic G1 phase and is required for S phase progression. J Biol Chem 283, 18905-18915.
Bardos, J.I., Saurin, A.J., Tissot, C., Duprez, E., and Freemont, P.S. (2000). HPC3 is a new
human polycomb orthologue that interacts and associates with RING1 and Bmil and has transcriptional repression properties. J Biol Chem 275, 28785-28792.
Bernstein, E., Duncan, E.M., Masui, O., Gil, J., Heard, E., and Allis, C.D. (2006). Mouse polycomb proteins bind differentially to methylated histone H3 and RNA and are enriched in facultative heterochromatin. Mol Cell Biol 26, 2560-2569.
Boukarabila, H., Saurin, A.J., Batsche, E., Mossadegh, N., van Lohuizen, M., Otte, A.P., Pradel, J., Muchardt, C., Sieweke, M., and Duprez, E. (2009). The PRC1 Polycomb group complex interacts with PLZF/RARA to mediate leukemic transformation. Genes Dev 23, 1195-1206.
42


Cao, R., Wang, L., Wang, H., Xia, L., Erdjument-Bromage, H., Tempst, P., Jones, R.S., and Zhang, Y. (2002). Role of histone H3 lysine 27 methylation in Poly comb-group silencing. Science 298, 1039-1043.
Chen, D., Dundr, M., Wang, C., Leung, A., Lamond, A., Misteli, T., and Huang, S. (2005). Condensed mitotic chromatin is accessible to transcription factors and chromatin structural proteins. J Cell Biol 168, 41-54.
Cheng, B., Ren, X., and Kerppola, T.K. (2014). KAP1 Represses Differentiation-Inducible Genes in Embryonic Stem Cells through Cooperative Binding with PRC1 and Derepresses Pluripotency-Associated Genes. Mol Cell Biol 34, 2075-2091.
Cherukuri, S., Hock, R., Cieda, T., Catez, F., Rochman, M., and Bustin, M. (2008). Cell cycle-dependent binding of HMGN proteins to chromatin. Mol Biol Cell 19, 1816-1824.
Core, N., Bel, S., Gaunt, S.J., Aurrand-Lions, M., Pearce, J., Fisher, A., and Djabali, M. (1997). Altered cellular proliferation and mesoderm patterning in Polycomb-M33-deficient mice. Development 124, 721-729.
de Napoles, M., Mermoud, J.E., Wakao, R., Tang, Y.A., Endoh, M., Appanah, R., Nesterova, T.B., Silva, J., Otte, A.P., Vidal, M., Koseki, H., and Brockdorff, N. (2004). Polycomb group proteins RinglA/B link ubiquitylation of histone H2A to heritable gene silencing and X inactivation. Developmental Cell 7, 663-676.
Di Croce, L., and Helin, K. (2013). Transcriptional regulation by Polycomb group proteins. Nat Struct Mol Biol 20, 1147-1155.
Elderkin, S., Maertens, G.N., Endoh, M., Mallery, D.L., Morrice, N., Koseki, H., Peters, G., Brockdorff, N., and Hiom, K. (2007). A phosphorylated form of mel-18 targets the RinglB histone H2A ubliquitin ligase to chromatin. Molecular Cell 28, 107-120.
43


Endoh, M., Endo, T.A., Endoh, T., Fujimura, Y.I., Ohara, O., Toyoda, T., Otte, A.P., Okano, M., Brockdorff, N., Vidal, M., and Koseki, H. (2008). Polycomb group proteins RinglA/B are functionally linked to the core transcriptional regulatory circuitry to maintain ES cell identity. Development 135, 1513-1524.
Follmer, N.E., Wani, A.H., and Francis, N.J. (2012). A polycomb group protein is retained at specific sites on chromatin in mitosis. PLoS Genet 8, el003135.
Fonseca, J.P., Steffen, P.A., Muller, S., Lu, J., Sawicka, A., Seiser, C., andRingrose, L. (2012).
In vivo Polycomb kinetics and mitotic chromatin binding distinguish stem cells from differentiated cells. Genes Dev 26, 857-871.
Forzati, F., Federico, A., Pallante, P., Abbate, A., Esposito, F., Malapelle, U., Sepe, R., Palma,
G., Troncone, G., Scarfo, M., Arra, C., Fedele, M., and Fusco, A. (2012). CBX7 is a tumor suppressor in mice and humans. J Clin Invest 122, 612-623.
Gao, Z., Zhang, J., Bonasio, R., Strino, F., Sawai, A., Parisi, F., Kluger, Y., and Reinberg, D.
(2012). PCGF Homologs, CBX Proteins, and RYBP Define Functionally Distinct PRC1 Family Complexes. Mol Cell 45, 344-356.
Gearhart, M.D., Corcoran, C.M., Wamstad, J.A., and Bardwell, V.J. (2006). Polycomb group and SCF ubiquitin ligases are found in a novel BCOR complex that is recruited to BCL6 targets. Mol Cell Biol 26, 6880-6889.
Gerlich, D., Hirota, T., Koch, B., Peters, J.M., and Ellenberg, J. (2006). Condensin I stabilizes chromosomes mechanically through a dynamic interaction in live cells. Curr Biol 16, 333-344.
Giglia-Mari, G., Theil, A.F., Mari, P.O., Mourgues, S., Nonnekens, J., Andrieux, L.O., de Wit,
J., Miquel, C., Wijgers, N., Maas, A., Fousteri, M., Hoeijmakers, J.H., and Vermeulen,
44


W. (2009). Differentiation driven changes in the dynamic organization of Basal transcription initiation. PLoS Biol 7, el000220.
Grau, D.J., Chapman, B.A., Garlick, J.D., Borowsky, M., Francis, N.J., and Kingston, R.E.
(2011). Compaction of chromatin by diverse Poly comb group proteins requires localized regions of high charge. Genes Dev 25, 2210-2221.
Hatano, A., Matsumoto, M., Higashinakagawa, T., and Nakayama, K.I. (2010). Phosphorylation of the chromodomain changes the binding specificity of Cbx2 for methylated histone H3. Biochem Biophys Res Commun 397, 93-99.
Hellwig, D., Emmerth, S., Ulbricht, T., Doring, V., Hoischen, C., Martin, R., Samora, C.P., McAinsh, A.D., Carroll, C.W., Straight, A.F., Meraldi, P., and Diekmann, S. (2011). Dynamics of CENP-N kinetochore binding during the cell cycle. J Cell Sci 124, 3871-3883.
Hemmerich, P., Schmiedeberg, L., and Diekmann, S. (2011). Dynamic as well as stable protein interactions contribute to genome function and maintenance. Chromosome Res 19, 131-151.
Hernandez-Munoz, F, Taghavi, P., Kuijl, C., Neefjes, J., and van Lohuizen, M. (2005).
Association of BMI1 with poly comb bodies is dynamic and requires PRC2/EZH2 and the maintenance DNA methyltransferase DNMT1. Molecular and Cellular Biology 25, 11047-11058.
Isono, K., Endo, T.A., Ku, M., Yamada, D., Suzuki, R., Sharif, J., Ishikura, T., Toyoda, T.,
Bernstein, B.E., and Koseki, H. (2013). SAM domain polymerization links subnuclear clustering of PRC 1 to gene silencing. Dev Cell 26, 565-577.
45


Katoh-Fukui, Y., Tsuchiya, R., Shiroishi, T., Nakahara, Y., Hashimoto, N., Noguchi, K., and Higashinakagawa, T. (1998). Male-to-female sex reversal in M33 mutant mice. Nature 393, 688-692.
Kerppola, T.K. (2009). Polycomb group complexesmany combinations, many functions. Trends Cell Biol 19, 692-704.
Klauke, K., Radulovic, V., Broekhuis, M., Weersing, E., Zwart, E., Olthof, S., Ritsema, M.,
Bruggeman, S., Wu, X., Helin, K., Bystrykh, L., and de Haan, G. (2013). Polycomb Cbx family members mediate the balance between haematopoietic stem cell self-renewal and differentiation. Nat Cell Biol.
Koga, EL, Matsui, S., Hirota, T., Takebayashi, S., Okumura, K., and Saya, H. (1999). A human homolog of Drosophila lethal(3)malignant brain tumor (l(3)mbt) protein associates with condensed mitotic chromosomes. Oncogene 18, 3799-3809.
Mekhail, K., Khacho, M., Carrigan, A., Hache, R.R., Gunaratnam, L., and Lee, S. (2005). Regulation of ubiquitin ligase dynamics by the nucleolus. J Cell Biol 170, 733-744.
Mendez, J., and Stillman, B. (2000). Chromatin association of human origin recognition complex, cdc6, and minichromosome maintenance proteins during the cell cycle: assembly of prereplication complexes in late mitosis. Mol Cell Biol 20, 8602-8612.
Meshorer, E., Yellajoshula, D., George, E., Scambler, P.J., Brown, D.T., and Misteli, T. (2006). Hyperdynamic plasticity of chromatin proteins in pluripotent embryonic stem cells. Dev Cell 10, 105-116.
Miyagishima, H., Isono, K., Fujimura, Y., Iyo, M., Takihara, Y., Masumoto, H., Vidal, M., and Koseki, H. (2003). Dissociation of mammalian Polycomb-group proteins, RinglB and
46


Rae28/Phl, from the chromatin correlates with configuration changes of the chromatin in mitotic and meiotic prophase. Histochem Cell Biol 120, 111-119.
Morey, L., Aloia, L., Cozzuto, L., Benitah, S.A., and Di Croce, L. (2013). RYBP and Cbx7
define specific biological functions of polycomb complexes in mouse embryonic stem cells. Cell Rep 3, 60-69.
Morey, L., Pascual, G., Cozzuto, L., Roma, G., Wutz, A., Benitah, S.A., and Di Croce, L. (2012). Nonoverlapping functions of the Polycomb group Cbx family of proteins in embryonic stem cells. Cell Stem Cell 10, 47-62.
Ogawa, H., Ishiguro, K., Gaubatz, S., Livingston, D.M., andNakatani, Y. (2002). A complex with chromatin modifiers that occupies E2F-and Myc-responsive genes in G(0) cells. Science 296, 1132-1136.
Park IK, etal. (2003) Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells. Nature 423:302-305
Penny, G.D., Kay, G.F., Sheardown, S.A., Rastan, S., and Brockdorff, N. (1996). Requirement for Xist in X chromosome inactivation. Nature 379, 131-137.
Phair, R.D., Scaffidi, P., Elbi, C., Vecerova, J., Dey, A., Ozato, K., Brown, D.T., Hager, G., Bustin, M., and Misteli, T. (2004). Global nature of dynamic protein-chromatin interactions in vivo: three-dimensional genome scanning and dynamic interaction networks of chromatin proteins. Mol Cell Biol 24, 6393-6402.
Ren, X., and Kerppola, T.K. (2011). REST interacts with Cbx proteins and regulates polycomb repressive complex 1 occupancy atREl elements. Mol Cell Biol 31, 2100-2110.
47


Ren, X., Vincenz, C., and Kerppola, T.K. (2008). Changes in the Distributions and Dynamics of Polycomb Repressive Complexes During Embryonic Stem Cell Differentiation. Mol Cell Biol 28, 2884-2895
Sanchez, C., Sanchez, I., Demmers, J.A.A., Rodriguez, P., Strouboulis, J., and Vidal, M. (2007). Proteomics analysis of RinglB/Rnf2 interactors identifies a novel complex with the FbxllO/JhdmlB histone demethylase and the Bcl6 interacting corepressor. Molecular & Cellular Proteomics 6, 820-834.
Satijn, D.P., Gunster, M.J., van der Vlag, J., Hamer, K.M., Schul, W., Alkema, M.J., Saurin,
A.J., Freemont, P.S., van Driel, R., and Otte, A.P. (1997). RING1 is associated with the polycomb group protein complex and acts as a transcriptional repressor. Mol Cell Biol 77, 4105-4113.
Saurin, A.J., Shiels, C., Williamson, J., Satijn, D.P., Otte, A.P., Sheer, D., and Freemont, P.S. (1998). The human poly comb group complex associates with pericentromeric heterochromatin to form a novel nuclear domain. J Cell Biol 142, 887-898.
Schmiedeberg, L., Weisshart, K., Diekmann, S., Meyer Zu Hoerste, G., and Hemmerich, P.
(2004). High- and low-mobility populations of HP1 in heterochromatin of mammalian cells. Mol Biol Cell 15, 2819-2833.
Schoorlemmer, J., MarcosGutierrez, C., Were, F., Martinez, R., Garcia, E., Satijn, D.P.E., Otte, A.P., and Vidal, M. (1997). Ringl A is a transcriptional repressor that interacts with the Polycomb-M33 protein and is expressed at rhombomere boundaries in the mouse hindbrain. Embo Journal 16, 5930-5942.
48


Sharif, J., Endo, T.A., Ito, S., Ohara, O., and Koseki, H. (2013). Embracing change to remain the same: conservation of poly comb functions despite divergence of binding motifs among species. Curr Opin Cell Biol 25, 305-313.
Simon, J.A., and Kingston, R.E. (2009). Mechanisms of Poly comb gene silencing: knowns and unknowns. Nat Rev Mol Cell Biol 10, 697-708.
Simon, J.A., and Kingston, R.E. (2013). Occupying chromatin: Poly comb mechanisms for
getting to genomic targets, stopping transcriptional traffic, and staying put. Mol Cell 49, 808-824.
Souza, P.P., Volkel, P., Trinel, D., Vandamme, J., Rosnoblet, C., Heliot, L., and Angrand, P.O. (2009). The histone methyltransferase SUV420H2 and Heterochromatin Proteins HP1 interact but show different dynamic behaviours. BMC Cell Biol 10, 41.
Steffen, P.A., Fonseca, J.P., Ganger, C., Dworschak, E., Kockmann, T., Beisel, C., and Ringrose, L. (2013). Quantitative in vivo analysis of chromatin binding of Polycomb and Trithorax group proteins reveals retention of ASH1 on mitotic chromatin. Nucleic Acids Res 41, 5235-5250.
Suzuki, M., Mizutani-Koseki, Y., Fujimura, Y., Miyagishima, H., Kaneko, T., Takada, Y.,
Akasaka, T., Tanzawa, H., Takihara, Y., Nakano, M., Masumoto, H., Vidal, M., Isono,
K., and Koseki, H. (2002). Involvement of the Polycomb-group gene RinglB in the specification of the anterior-posterior axis in mice. Development 129, 4171-4183.
Tavares, L., Dimitrova, E., Oxley, D., Webster, J., Poot, R., Demmers, J., Bezstarosti, K., Taylor, S., Ura, H., Koide, H., Wutz, A., Vidal, M., Elderkin, S., and Brockdorff, N. (2012). RYBP-PRC1 Complexes Mediate H2A Ubiquitylation at Poly comb Target Sites Independently of PRC2 and H3K27me3. Cell.
49


Ueda, T., Catez, F., Gerlitz, G., and Bustin, M. (2008). Delineation of the protein module that anchors HMGN proteins to nucleosomes in the chromatin of living cells. Mol Cell Biol 28, 2872-2883.
Vandamme, J., Volkel, P., Rosnoblet, C., Le Faou, P., and Angrand, P.O. (2011). Interaction proteomics analysis of poly comb proteins defines distinct PRC 1 complexes in mammalian cells. Mol Cell Proteomics 10, M110 002642.
Vandenbunder, B., Fourre, N., Leray, A., Mueller, F., Volkel, P., Angrand, P.O., and Heliot, L. (2014). PRC1 components exhibit different binding kinetics in Polycomb bodies. Biol Cell 106, 111-125.
Vincenz, C., and Kerppola, T.K. (2008). Different polycomb group CBX family proteins
associate with distinct regions of chromatin using nonhomologous protein sequences. Proc Natl Acad Sci U S A 105, 16572-16577.
Voncken, J.W., Schweizer, D., Aagaard, L., Sattler, L., Jantsch, M.F., and van Lohuizen, M. (1999). Chromatin-association of the Polycomb group protein BMI1 is cell cycle-regulated and correlates with its phosphorylation status. Journal of Cell Science 112, 4627-4639.
Wang, G., Horsley, D., Ma, A., Otte, A.P., Hutchings, A., Butcher, G.W., and Singh, P.B.
(1997). M33, a mammalian homologue of Drosophila Polycomb localises to euchromatin within interphase nuclei but is enriched within the centromeric heterochromatin of metaphase chromosomes. Cytogenet Cell Genet 78, 50-55.
50


Wang, H.B., Wang, L.J., Erdjument-Bromage, H., Vidal, M., Tempst, P., Jones, R.S., and Zhang, Y. (2004a). Role of histone H2A ubiquitination in polycomb silencing. Nature 431, 873-878.
Wang, L., Brown, J.L., Cao, R., Zhang, Y., Kassis, J.A., and Jones, R.S. (2004b). Hierarchical recruitment of polycomb group silencing complexes. Mol Cell 14, 637-646.
Yang, L., Lin, C., Liu, W., Zhang, J., Ohgi, K.A., Grinstein, J.D., Dorrestein, P.C., and
Rosenfeld, M.G. (2011). ncRNA- and Pc2 methylation-dependent gene relocation between nuclear structures mediates gene activation programs. Cell 147, 773-788.
Yao, J., Munson, K.M., Webb, W.W., and Lis, J.T. (2006). Dynamics of heat shock factor association with native gene loci in living cells. Nature 442, 1050-1053
51


Full Text

PAGE 1

CBX2 STABLY ASSOCIATES WITH MITOTIC CHROMOSOMES VIA A PRC2 OR PRC1 INDEPENDENT MECHANISM AND IS NEEDED FOR RECRUITING PRC1 COMPLEX TO MITOTIC CHROMOSOMES By CHAO YU ZHEN B.S., University of Colorado Denver, 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Chemistry Program 2016

PAGE 2

! ii 2016 CHAO YU ZHEN ALL RIGHTS RESERVED

PAGE 3

! iii This thesis for the Master of Science degree by Chao Yu Zhen has been approved for the Chemistry Program by Xiaojun Ren, Chair Christopher Phiel Hai Lin April 21, 2016

PAGE 4

! iv Chao Yu Zhen (M.S., Chemistry) Cbx2 Stably Associates with Mitotic Chromosomes via a PRC2 or PRC1 independentMechanism and is needed for Recruiting PRC1 Complex to Mitotic Chromosomes Thesis directed by Assistant Professor Xiaojun Ren ABSTRACT Polycomb group (PcG) proteins are composed of a multiplicity of transcriptional regulatory factors that maintain cellular identity through epigenetic mechanism Often in most cases, once transcriptional silencing is imposed on genes, it can be transmitted from mother to daughter cells. Surprisingly, during mitotic event, chromatin goes through a dramatic str uctural reorganization and is compacted that would limit most transcription factors and chromatin protein access. Thus, it is unclear how the PcG mediated transcriptional program is maintained during mitosis. Here, using quantitative live cell imaging in mouse ES cells and tumor cells, we demonstrate that, although Polycomb repressive complex (PRC) 1 proteins (Cbx family proteins Ring1b, Mel18, and Phc1) exhibit variable capacities of association with mitotic chromosomes, Cbx2 overwhelmingly binds to mito tic chromosomes. The recruitment of Cbx2 to mitotic chromosomes is independent of PRC1 or PRC2, and Cbx2 is needed to recruit PRC1 complex to mitotic chromosomes. Quantitative fluorescence recovery after photobleaching analysis indicates that PRC1 proteins rapidly exchange at interphasic chromatin. On entry into mitosis, Cbx2, Ring1b, Mel18, and Phc1 proteins become immobilized at mitotic chromosomes, whereas other Cbx family proteins dynamically bind to mitotic chromosomes. Depletion of PRC1 or PRC2 protei n has no effect on the immobilization of Cbx2 on mitotic chromosomes. We find that the N terminus of Cbx2 is needed for its recruitment to mitotic chromosomes, whereas the C terminus is required for its immobilization. Thus these results provide fundamenta l insights into the molecular mechanisms of epigenetic inheritance.

PAGE 5

! v The form and content of this abstract are approved. I recommend its publication. Approved: Xiaojun Ren

PAGE 6

! vi ACKNOWLEDGEMENT S First, I would like to express my sincere gratitude for my advisor Prof. Xiaojun Ren, for his continuous supports and thoughtful advice of my master's work. The door to Prof. Ren office was always open at the time of need and he always ste ers me to ask the correct questions His guidance helped me in all the time of my research. Beside re search, Prof Ren's life advices are invaluable. Also, I would like to further express my sincere gratitude f or my advisor Prof Ren for his patient, motivation, and above and beyond effort s to help me. I truly had an amazing research experience and I could not have imagined a better advisor. Beside my advisor, I would like to thank my committees: Prof. Hai Lin and Prof. C hristopher P h iel for their advice s and time commitment to help improve this the sis. Also, I would like to thank them for being a great teacher to me. My sincere further goes to Prof. Jefferson Knight and Prof. Christopher P h iel who gave me access to their laboratory equipment and the wonderful scientific advices. Also, I would like to thank my fellow Ren lab colleague at University of Colorado Denver and especially: Huy Duc, Roubina Tatavosian, and Marko Kokotovic for their stimulating discussion and encouragement during time of my master resea rch Lastly to my family, I would like express my deepest appreciation for their conditional love, support, pain, and sacrifice to shape my life. Note: The work presented in this thesis has been published in Molecular Biology of Cell. The chromatin and PRC1 plasmids were made by Huy Duc and the Cbx plasmids were made by Marko Kokotovia.

PAGE 7

! vii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ................. 1 II. R ESULTS ................................ ................................ ................................ .............................. 4 PRC1 proteins vary in association with mitotic chromosomes ................................ ...... 4 T he mitotic chromosomal association of Cbx2 is independent of PcG proteins ... 10 Cbx2 is required for recruiting the canonical PRC1 proteins to mitotic chromosomes ................................ ................................ ................................ .............. 13 Cbx2 directly recruits the canonical PRC1 proteins to mitotic chromosomes ........... 15 The Cbx2 PRC1 complex is immobilized at mitotic chromosomes, but other Cbx family proteins rapidly exchange at mitotic chromosomes ................................ ........ 18 The immobilization of Cbx2 at mitotic chromosomes is independent of PcG proteins ................................ ................................ ................................ ................................ ..... 22 The recruitment and immobilization of Cbx2 to mitotic chromosomes requires its distinct regions ................................ ................................ ................................ ............. 24 III. DISCUSSION ................................ ................................ ................................ ................... 28 IV. MATERIALS AND METHODS ................................ ................................ ...................... 34 C ell lines ................................ ................................ ................................ ...................... 34 Plasmid ................................ ................................ ................................ ........................ 34 Generation of stable cell lines by lentivirus infection ................................ ................. 35 Confocal microscope imaging of live cells and quantification of mitotic fraction ..... 36 FRAP imaging and quantification ................................ ................................ .............. 37 FRET imaging ................................ ................................ ................................ ............ 38 Epifluorescence imaging of live cells ................................ ................................ ......... 38 Cell synchronization and fractionation ................................ ................................ ....... 38

PAGE 8

! viii Immunoblottin ................................ ................................ ................................ ............ 39 Immunofluorescence ................................ ................................ ................................ .. 40 REFERENCE S ................................ ................................ ................................ ........................... 42

PAGE 9

! ix LIST OF FIGURES Figures 1. Mitotic chromosomal association of PRC1 fusion PROTEINS ................................ .......... 5 2. Mitotic chromosomal association of YFP Cbx2 fusion protein in HEK293 and HeLa cells ................................ ................................ ................................ ........................... 6 3. Effects of doxycycline concentration on level of YFP Cbx2 association with mitotic Chromosomes ................................ ................................ ................................ ..................... 7 4. Western blot analysis of levels of endogenous and fusion proteins ................................ ... 7 5. Analysis of mitotic chromatin binding of PRC1 proteins by chromatin Fractionation ................................ ................................ ................................ ...................... 9 6. The association of Cbx2 fusion protein with mitotic chromosomes in PRC1 and PRC2 gene knockout ES cells ................................ ................................ ................................ ..... 12 7. Western blot analysis of Ring1b in Ring1a / /Ring1bfl/fl ES cells treated with or w ithout OHT ................................ ................................ ................................ ................................ 13 8. The mitotic chromosomal association of PRC1 fusion proteins Ring1b, Phc1, and Mel18 in Cbx2 / ES cell ................................ ................................ ................................ .............. 14 9. Immunostaining of Phc1 and Ring1b proteins in Cbx2+/+ and Cbx2 / ES cells ............ 15 10. Directly recruiting PRC1 fusion proteins to mitotic chromosomes by YFP Cbx2, but not by YFP Cbx21 498 ................................ ................................ ................................ ............ 18 11. FRAP analysis of PRC1 fusion proteins binding to interphasic and mitotic Chromatins ................................ ................................ ................................ ...................... 20 12. FRAP analysis of YFP Cbx2 fusion protein binding to interphasic and mitotic chromatins in HeLa and HEK293 cells ................................ ................................ ........... 22 13. FRAP analysis of YFP Cbx2 fusion protein binding to chromatins in Eed / Ring1a / /Ring1b / and Bmi1 / /Mel18 / ES cells ................................ ................................ ..... 23 14. Analysis of structural elements of YFP Cbx2 fusion pro tein required for its targeting and immobilizing ................................ ................................ ................................ ........... 26 15. Hypothetic mod el for the interaction of canonical PRC1 complex with interphasic and mitotic chromatin ................................ ................................ ................................ ............ 33

PAGE 10

! x LIST OF ABBREVIATIONS BiFC Bimolecular fluorescence complementation BSA Bovine Serum Albumin Cbx2 / Knock out of Cbx2 gene Cbx2 Chromobox Homolog 2 Cbx4 Chromobox Homolog 4 Cbx6 Chromobox Homolog 6 Cbx7 Chromobox Homolog 7 Cbx8 Chromobox Homolog 8 Cbox Chromobox CHD Chromodomain DMEM Dulbecco's Modified Eagle's Medium DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid Eed Embryonic Ectoderm Development EGFP Enhanced Green Fluorescent Protein EGTA Aminopolycarboxylic acid EMCCD Electron Multiplying Charge Coupled Device ES Cel Embryonic Stem Cell FITC Fluorescein Isothiocyanate fl/fl Flox/flox (conditional knockout) FRAP Fluoresence Recovery After Photobleaching FRET Fluoresence Resonence Energy Transfer

PAGE 11

! xi H3K27me3 Trimethylation of Lysine 27 on Histone 3 H2A Histone Family, Member Z HEK293 Human Embryonic Kidney 293 HEPES 4 (2 hydroxyethyl) 1 piperazineethanesulfonic acid KCl Potassium Chloride KO Knockout LIF Leukemia Inhibitor Factor MgCl 2 Magnesium Chloride M Phase Metaphase NaCl Sodium Chloride OHT 4 hyroxytamoxifen PBS Phosphate Buffered Saline PcG Polycomb Group Pcgf1 Polycomb Group Ring Finger 1 Pcgf2 Polycomb Group Ring Finger 2 Pcgf3 Polycomb Group Ring Finger 3 Pcgf4 Polycomb Group Ring Finger 4 Pcgf5 Polycomb Group Ring Finger 5 Pcgf6 Polycomb Group Ring Finger 6 Phc1 P olyhomeotic Homolog 1 Phc2 P olyhomeotic Homolog 2 Phc3 P olyhomeotic Homolog 3 PMSF P henylmethylsulfonyl fluoride

PAGE 12

! xii PRC1 Polycomb Repressive Complex 1 PRC2 Polycomb Repressive Complex 2 SDS PAGE S odium dodecyl sulfate polyacrylamide gel electrophoresis RING1B Really Interesting New Gene 1 B WT Wild Type YFP Yellow Fluorescence Protein

PAGE 13

! 1 CHAPTER I INTRODUCTION The transcriptional programming of genes occurs during early embryonic developmental stages and is a major influence for cellular identity and diversity. T ranscriptional activation or repression (silencing) of genes will often be retained during development, although it remains plasticity. Polyco mb G roup (PcG) proteins are epigenetic transcriptional regulators that controls hundreds of key developmental homeotic genes by biochemical modification through the mediation of trimethylation on lysine 27 of histone (H3K27me3) (Kerppola, 2009; Di Croce and Helin, 2013; Sharif ., 2013; Simon and Kingston, 2013) Depletion of PcG genes can results in dysfunction of mouse embryonic stem ( ES ) cell (Park IK, et al 2003) or impairment and terminal of mouse embryonic development (Voncken JW, et al 2003). PcG mediated epigenetic repressio n mark or memory, in most case, can be transmitted from mother to daughter cells through mitotic event. PcG proteins action is especially vital during cell division to preserve cells identities However, during mitosis, chromatin undergoes disruption and s ubsequent restoration, which challenge the integrity of genetic and epigenetic memory and also presents a window of opp ortunity for epigenetic change. The vast majority of transcription factors and chromatin bind proteins dissociate from mitotic chromosome s, although, some mitotic chromosomes are open for traffics (Chen et al., 2005; Hemmerich et al., 2011) Thus, to better understand epigenetic memory and f lexibility, it is essential to study the behavior of PcG mediate gene repression action on mitotic chromosomes. Initially discovered in drosophila melanogaster (fly), (PcG) protein associate with two transcriptional repressive complexes, Polycomb repressiv e complex (PRC) 1 and 2 (Kerppola, 2009; Di Croce and Helin, 2013; Sharif et al. 2013 ; Simon & Kingston, 2013). Both PRC 1 and

PAGE 14

! 2 2 process histone catalytic activity. PRC2 is a methlytransferase that primarily enzymatically trimethlyate at histone H3 tail s on lysine 27 (H3K27me3) (Cao et al, 2002), and recruit s PRC1, the enzymatic function of ubiquitin ligase, to monoubquitylation of lysine 119 on H2A tails (H2Ak119ub1) (de Napoles et al, 2004; Wang et al, 2004a ; Simon and Kingston, 2013 ). PRC1 compri ses of 5 major core subunits families proteins which comprise of polycomb group of ring finger (PcgF1/2/3/4/5/6), Phc (Phc1/2/3), Cbx (cbx2/4/6/7/8), Rybp/YaF2, and Ring1 (A/B), t he assemblage of PRC1 complexes (Simon & Kingston, 2009; Turner & Bracken, 2013). I n mammal, the multiple variations of each core PRC1 family protein attributed to a large library of uniquely individual PRC1 complexes. An attempt to establish a more simplified classification of PRC1 co mplexes was drafted, which was based on the concept o f canonical and noncanonical PRC1 complexes (Gao et al, 2012; Tavares et al, 2012; Morey et al. 2013 ). Cbx and RybP are mutually exclusive associated with PRC1 complexes are defined as canonical and noncanonical, respectively. Canonical PRC1 has been rep orted to be recruit ed to chromatin by H3K23me3 a PRC2 dependent mechanism (Wang et al, 2004b). Even as now the functions of each unique individual PRC1 complexes remain to be resolved. Extensive research has been done on the recruitment of canonical PRC1 complexes to interphasic chromatin (Cao et al. 2002; Ogawa et al. 2002; Wang et al. 2004b; Gearhart et al. 2006; Sanchez et al. 2007; Boukarabila et al. 2009; Ren and Kerppola, 2011; Vandamme et al. 2011; Gao et al. 2012; Tavares et al. 2012; Morey et al. 2013; Cheng et al. 2014) yet some fundamental epigenetic inheritance role of PRC1 in mitotic chromosomes remains to be answer. (1) Does PRC1 complexes interact wi th mitotic chromosomes, and if so, (2) what is the mechanism that dictate PRC1 for this interaction. Studies have reached a division of views on how vertebrate PRC1 proteins interact with mitotic chromosomes (Wang et al. 1997; Saurin et

PAGE 15

! 3 al. 1998; Koga et al. 1999; Voncken et al. 1999; Akasaka et al. 200 2; Suzuki et al. 2002; Miyagishima et al. 2003; Aoto et al. 2008; Vincenz and Kerppola, 2008) In earlier studies, evidence was provided that a small subpopulation of PRC1 proteins ass ociated with mitotic chromosomes (Follmer et al. 2012; Fonseca et al. 201 2; Steffen et al. 2013) Surprisingly, PRC1 binding dynamic was found to be 300 fold longer comparing mitot ic to interphasic chromatin (Fonseca et al. 2012) The key mechanism for epigenetic inheritance of PRC1 is the physical interaction with mitotic chromosomes, which the interaction are retention and immobilization, which should be conserved across organisms that contain PRC1. Here we show, using q uan ti tative live cell imaging, Cbx2 is the dominant Cbx family protein that overwhelmingly associate to mitotic chromosomes. The association of Cbx2 is independent of PRC1 and PRC2 mechanism s Cbx2 is the key protein that is required for the recruitment of PRC1 proteins to mitotic chromosomes. Utilizing quantitative FRAP, we demonstrated that PRC1 proteins is highly dynamic with interphasic chromatin, as reported previously (Ren et al. 2008) and identify that a special Cbx2 PRC1 complex is selectively immobilized at mitotic chrom osomes. The immobilization of cbx2 protein at mitotic chromosomes is independent of PRC2 or PRC1. Finally, we reveal that cbx2 retain uniquely different mechanisms for recruitment and immobilization on mitotic chromosomes.

PAGE 16

! 4 CHAPTER II RESULTS PRC1 proteins vary in association with mitotic chromosomes Early studies of the association of mammalian PRC1 proteins with mitotic chromosomes reach divergent opinions (Wang et al. 1997; Saurin et al. 1998; Koga et al. 1999; Voncken et al. 1999; Akasaka et al. 2002; Suzuki et al. 2002; Miyagishima et al. 2003; Aoto et al. 2008; Vincenz and Kerppola, 2008) To fully appreciate wheth er the Cbx containing PRC1 complexes associate with mitotic chromosomes, we established ES cell lines that inducibly express PRC1 protein fused with Cerulean or YFP under doxycycline controlled manner. To facilitate live cell imaging and to mark mitotic ch romosomes, histone H2A fused with Cerulean or mCherry was co expressed with PRC1 fusion protein. We performed multicolor Z scan imaging of live cells at 37 o C with a confocal laser scanning microscope. The PRC1 fusion proteins (Cbx family proteins (Cbx2, C bx4, Cbx6, Cbx7, Cbx8), Ring1b, Phc1 and Mel18) exhibited variable capacities of association with mitotic chromosomes. YFP Cbx2 fusion was the primary protein accumulated at mitotic chromosomes (Fig. 1A H). These PRC1 fusion proteins were granularly distri buted at mitotic chromosomes. Quantitative analysis of Z stack images showed that (96 5)% of YFP Cbx2 protein associates with mitotic chromosomes, while YFP Cbx4 (22 5)%, YFP Cbx6 (26 7)%, YFP Cbx7 (44 4)%, Cbx8 (40 9)%, Cerulean Ring1b (63 10) %, Cerulean Phc1 (84 10)%, and Cerulean Mel18 (84 7)% all showed reduced association with mitotic chromosomes (Fig. 1I). These results are consistent with previous BiFC analysis of Cbx2 and Cbx6 association with mitotic chromosomes (Vincenz and Kerppola, 2008) but the current studies provide the quantitative insights of the entire Cbx family proteins and other PRC1

PAGE 17

! 5 proteins. Thus, these data indicate that the Cbx containing PRC1 complexes associate with mitotic chromosomes and Cbx2 is the prim ary protein enriched at mitotic chromosomes. Figure 1 Mitotic chromosomal association of PRC1 fusion proteins. (A H) Confocal fluorescence images of PRC1 fusion proteins tagged with either YFP or Cerulean expressed in PGK12.1 ES cells in various phases of mitosis. The mCherry H2A fusion protein was used to mark mitotic chromosomes. The images of prometaphase (top panel), metaphase (middle pan el), and anaphase/telophase (bottom panel) are presented. The representative image from Z scan stack is presented in the Figures. Scale bar is 5 !m. (I) Quantitative analysis of mitotic binding fraction of PRC1 fusion proteins at mitotic chromosomes. The m itotic binding fraction is average of individual slices of Z scan stack. The data represents at least ten cells analyzed. Error bars indicate standard deviation of the mean.

PAGE 18

! 6 Since Cbx2 is the primary protein associated with mitotic chromosomes, we wished t o determine whether the mitotic chromosomal association of Cbx2 protein is cell type specific. Therefore, we established HEK293 and HeLa cell lines that stably and inducibly express YFP Cbx2 and mCherry H2A fusion proteins. Z scan imaging of live HEK293 an d HeLa cells showed that YFP Cbx2 fusion protein was profoundly accumulated at mitotic chromos omes in both cell lines (Fig. 2 ). Consistent with the YFP Cbx2 fusion protein localization at mitotic chromosomes in ES cells, quantitative analysis of Z stack im ages showed that (96 4)% of YFP Cbx2 is associated with mitotic chromosomes in both HEK293 and HeLa cells, indicating the mitotic chromosomal association of Cbx2 protein is not restricted to ES cells. Figure 2. Mitotic chromosomal association of YFP Cbx2 fusion protein in HEK293 and HeLa cells YFP Cbx2 and Cerulean H2A fusion proteins were expressed in HEK293 and HeLa cells. The Cerulean H2A was used to mark mitotic chromosomes. The confocal images of metaphase are presented. Scale bar is 5 !m. To ask whether the expression level of Cbx2 fusion protein affects the mitotic chromosomal association, we administrated a wide range of concentrations of doxycycline, from 0.1 to 1.0 !M, into HEK293 cells stably and inducibly expressing YFP Cbx2 and Cerulean H2A. Epifluorescence imaging of live HEK293 cells showed that the fraction of mitotic chromosomal association of Cbx2 fusion protein was similar to one another under varying concen trations of

PAGE 19

! 7 doxycycline (Fig. 3 ). Unless otherwise indicated, we used doxyc ycline at the concentration of 0.20 0.50 !M throughout subsequent studies. At 0.5 !M of doxycycline, the expression levels of the PRC1 proteins tested were either similar to or 2 3 fold higher than that of their endogenous counterparts (Fig. 4 ). Figure 3. Effects of doxycycline concentration on level of YFP Cbx2 association with mitotic chromosome The expression of YFP Cbx2 fusion protein in HEK293 cells was induced by using a range of doxycycline concentrations. Cerulean H2A was used to mark mit otic chromosomes. The epifluorescence images of metaphase are presented. Scale bar is 5 !m. Figure 4. Western blot analysis of levels of endogenous and fusion proteins Western blots were performed using antibodies against endogenous proteins. The cel l extracts were prepared from PGK12.1 ES cells and PGK12.1 ES cells expressing either YFP Cbx2 (PGK12.1 + YFP Cbx2), Cerulean Ring1b (PGK12.1 + Cerulean Ring1b), or Cerulean Phc1 (PGK12.1 + Cerulean Phc1). The fusion proteins were induced to express by 0.5 !M of doxycycline for 2 days.

PAGE 20

! 8 To provide independent evidence of PRC1 protein association with mitotic chromosomes, we performed biochemical fractionation followed by Western blot analysis. ES cells were synchronized at M phase by treating sequentially with thymidine and nocodazole bas ed on published report (Ballabeni et al. 2011) At least 85% of ES cells were mitotic as revealed by cell morphology and immunostaining of H3S10p (Fig. 5 A B). The non synchronized (con trol) and synchronized (mitotic) cells were fractioned according to the scheme showed in Fig. S4C based on previous reports (Mendez and Stillman, 2000; Follmer et al. 2012) In control and mitotic cells, the histone H3 was primarily found in the chromatin fraction (P3), while tubulin was found primarily in the cytosolic fraction (S2) (Fig. 5 D). We tested the chromatin association of 4 PRC1 proteins (Ring1b, Phc1, Cbx2 and Cbx7) in both control and mitotic cells. In mitotic cells, we found that Cbx2 protein is primarily in the fraction P3, while Cbx7 is primarily in th e cy tosolic fraction S2 (Fig. 5 D). Thus, these data are consistent with live cell imaging data that show that Cbx2 is the primary protein enriched at mitotic chromosomes.

PAGE 21

! 9

PAGE 22

"# Figure 5. Analysis of mitotic chromatin binding of PRC1 proteins b y chromatin fractionation (A) Representative images of synchronized PGK12.1 ES cells stably expressing mCherry H2A. The circular morphology of cells in phase contrast image indicates cell at mitotic stages. The mCherry H2A was used to mark chromatin. Scal e bar is 20 !m. (B) Epifluorescence images of synchronized PGK12.1 ES cells immunostained by anti Histone H3 phospho S10 (H3PS10) antibody. DNAs were stained by Hoechst. Over 85% of cells show positive H3PS10 staining. Scale bar is 20 !m. (C) Scheme of PR C1 protein fractionation, adapted from (Mendez and Stillman, 2000; Follmer et al. 2012) (D) Western blot analysis of fractions from control and m itotic cells. H3 and tubulin are present in the expected fractions. Western blots showed that Ring1b, Phc1, Cbx2, and Cbx7 fractionate with mitotic chromatin The mitotic chromosomal association of Cbx2 is independent of PcG proteins The Cbx containing P RC1 complexes are recruited to interphasic chromatin via Cbx family protein interactions with H3K27me3 mediated by PRC2 (Cao et al. 2002; Wang et al. 2004b; Bernstein et al. 2006 ; Tavares et a l. 2012; Morey et al. 2013) Therefore, we asked whether PRC2 is required for the accumulation of Cbx2 protein at mitotic chromosomes. To test the hypothesis, both YFP Cbx2 and Cerulean H2A fusion proteins were stably and inducibly expressed in Eed knockout (KO) ES cell lines, which lack the H3K27me3 modification. Z scan of confocal images showed that YFP Cbx2 fusion protein was granularly distributed at mitotic chromosomes, consistent with its distribution in wild type ES cells. Quantitative analysi s of Z stack images showed that (97 5)% of Cbx2 protein was enriched at mitotic chromosomes in Eed KO ES cell lines (Fig. 6 B and 6 D), consistent with YFP Cbx2 associating with mitotic chromosomes in wild type ES cells (Fig. 6A and 6 D), indicating that a core component of the PRC2 complex, Eed, is not required for the accumulation of Cbx2 protein at mitotic chromosomes. To determine whether the formation of integral canonical PRC1 complexes is required for the accumulation of Cbx2 protein at mitotic chromo somes, YFP Cbx2 and Cerulean H2A fusion proteins were stably and inducibly expressed in Ring1a constituently KO and Ring1b conditionally KO ES cell lines (Ring1a / / Ring1b fl/fl ). We administrated OHT and doxycycline for

PAGE 23

"" three days. The addition of OHT induced depletion of en dogenous Ring1b protein (Fig. 7 ). We performed Z scan of live cell imaging of these cells using a confocal microscope. Quantitative analysis of Z stack images showed that (97 4)% of Cbx2 fusion protein is accumulate d at mitotic chr omosomes (Fig. 6 C and 6 D), consistent with localization of the YFP Cbx2 fusion protein in wild type and Eed KO ES cells. To further investigate whether other canonical PRC1 subunits affect the mitotic chromosomal accumulation of Cbx2, YFP Cbx2 and Cerulean H2A fusion proteins were stably and inducibly expressed in Bmi1 and Mel18 ( Bmi1/Mel18 ) double KO ES cell lines. Quantitative analysis of Z stack images indicated that (95 3)% of Cbx2 fusion protein is accumulated at mitotic chromosomes, consistent with th at in wild type ES cells (Fig. 6 C and 6 D). Thus, these data show that the mitotic chromosomal accumulation of Cbx2 protein is independent of PRC1 proteins Ring1a/Ring1b and Bmi1/Mel18.

PAGE 24

"$ Figure 6 The association of Cbx2 fusion protein with mitotic chromosomes in PRC1 and PRC2 gene knockout ES cells. (A) Confocal fluorescence images of YFP Cbx2 fusion protein expressed in PGK12.1 ES cells (WT ES). The different version of YFP Cbx2 of Fig. 1A is presented here in order to compare with Fig. 2B and 2C. Scale bar is 5 !m. (B) Confocal fluorescence images of YFP Cbx2 fusion protein expressed in Eed / ES cells in various phases of mitosis. The images of prometaphase (top panel), metaphase (middle panel), and anaphase/telophase (bottom panel) are presented. Scale bar is 5 !m. (C) Confocal fluorescence images of YFP Cbx2 fusion protein expressed in Ring1a / /Ring1b fl/fl and Bmi1 / /Mel18 / ES cell lines in various phases of mitosis. The images of prometaphase (top panel), metaphase (middle panel), and anapha se/telophase (bottom panel) are presented. The Ring1a / /Ring1b fl/fl was treated with OHT to induce depletion of Ring1b for three days before imaging. The protein level of Ring1b in Ring1a / /Ring1b fl/fl cells with or without OHT was presented in Figure S5 Scale bar is 5 !m. (D) Quantitative analysis of mitotic binding fractions of YFP Cbx2 fusion protein at mitotic chromosomes. The data represents at least ten cells analyzed. Error bars indicate standard deviation of the mean.

PAGE 25

"% Figure 7. Western blot a nalysis of Ring1b in Ring1a / / Ring1b fl/fl ES cells treated with or without OHT Cbx2 is required for recruiting the canonical PRC1 proteins to mitotic chromosomes Because (1) Cbx2 is the primary Cbx family protein that is accumulated at mitotic chromosomes, and (2) the mitotic chromosomal accumulation of Cbx2 protein is independent of PRC1 or PRC2 complex proteins, we reasoned that Cbx2 protein is required for recru iting canonical PRC1 proteins to mitotic chromosomes. To this end, Cerulean PRC1 protein (either Cerulean Ring1b, or Cerulean Phc1, or Cerulean Mel18) and mCherry H2A fusion protein were inducibly expressed in Cbx2 / ES cell lines. Z scan imaging of live cells by using confocal microscopy revealed that fluorescence intensities of Cerulean Ring1b, Cerulean Phc1, and Cerulean Mel18 fusion proteins at mitotic chromosomes in Cbx2 KO ES cells was greatly reduced in comparison to that seen in wild type ES cells (compare Fig. 1 F H and 8 A). Quantitative analysis of Z stack images revealed that (29 8)% of Cerulean Ring1b protein associated with mitotic chromosomes, (17 7)% of Cerulean Phc1 protein, and (27 5)% o f Cerulean Mel18 protein (Fig. 8 B). To test whet her Cbx2 gene knockout affects the level of Phc1, Mel18 and Ring1b proteins, we performed immunoblotting using extracts from Cbx2 +/+ and Cbx2 / ES cells. Western blots showed that the level of Phc1, Mel18 and Ring1b proteins in Cbx2 / ES cells was similar to that seen in Cbx2 +/+ cells (Fig. 8 C). Thus, these data suggest that

PAGE 26

"& Cbx2 protein is required for the accumulation of canonical PRC1 proteins Ring1b, Phc1, and Mel18 at mitotic chromosomes. Figure 8. The mitotic chromosomal asso ciation of PRC1 fusion proteins Ring1b, Phc1, and Mel18 in Cbx2 / ES cells (A) Confocal fluorescence images of Cerulea Ring1b, Cerulean Phc1, and Cerulean Mel18 fusion proteins expressed in Cbx2 / ES cells in metaphase (top panel) and anaphase (bottom panel). Scale bar is 5 !m. (B) Quantitative comparison of mitotic chromosomal association of Cerulea Ring1b, Cerulean Phc1, and Cerulean Mel18 fusion proteins in Cbx2 +/+ and Cbx2 / ES cells. The data represents average of at least 10 cells analyzed. Error bars indicate standard deviation of the mean. (C) Western blots of cell extracts from Cbx2 +/+ and Cbx2 / ES cells. The Ponceau S staining indicates the loading control. To further investigate wheth er Cbx2 is required for recruiting Phc1 and Ring1b to mitotic chromosomes, we performed immunostaining in Cbx2 +/+ and Cbx2 / ES cells by using antibodies that detect end ogenous Phc1 and Ring1b (Fig. 9 ). Immunostaining showed that endogenous Phc1 and Ring1 b associate with mitotic chromosomes in wild type ES cells.

PAGE 27

"' However, endogenous Phc1 and Ring1b were excluded from mitotic chromosomes in Cbx2 knockout ES cells. These data suggest that Cbx2 is required to recruit Phc1 an d Ring1b to mitotic chromosomes. Figure 9 Immunostaining of Phc1 and Ring1b proteins in Cbx2 +/+ and Cbx2 / ES cells Cbx2 +/+ and Cbx2 / ES cells were fixed and immunostained with antibodies that detect Phc1 and Ring1b (green) at different phases of mitosis. DNAs were stained with Hoech st (blue). Overlay imag es were shown. Scale bar is 5 !m Cbx2 directly recruits the canonical PRC1 proteins to mitotic chromosomes Since Cbx2 protein affects the accumulation of canonical PRC1 proteins at mitotic chromosomes, we reasoned that Cbx2 protein directly recruits canonical PRC1 proteins to mitotic chromosomes. To this end, we co expressed the three fusion proteins in Cbx2 KO ES cell lines: YFP Cbx2, Cerulean PRC1 subunit (either Cerulean Ring1b, or Cerulean Phc1, or Cerulean Mel18), and mCherry H2A. We expected that the introduction of an YFP Cbx2 fusion protein to the Cbx2 null background ES cells would restore the mitoti c chromosomal association

PAGE 28

"( of the three PRC1 proteins (Cerulean Ring1b, Cerulean Phc1 and Cerulean Mel18). We performed three color Z scan imaging of live cells by using confocal microscope. Quantitative analysis of Z stack images from three ES cell lines e xpressing Ring1b, Phc1, and Mel18 fusion proteins showed an average of (95 7)% of YFP Cbx2 associated with mitotic chromosomes, consistent with YFP Cbx2 localizati on in wild type ES cells (Fig. 10A and 10 C), indicating that mitotic chromosomal associatio n of YFP Cbx2 fusion protein is independent of endogenous Cbx2 protein. Notably, quantitative image analysis revealed that (63 8)% of Cerulean Ring1b, (71 12)% of Cerulean Phc1, and (74 10)% of Ceruelan Mel18 also associated with mitotic chromosomes (Fig. 10A and 10 D). The fraction of retention of the three fusion proteins at mitotic chromosomes in Cbx2 KO ES cell lines complemented with YFP Cbx2 was similar to that seen in wild type ES cells, indicating YFP Cbx2 fusion protein recruits the three cano nical PRC1 fusion proteins to mitotic chromosomes. To test whether the Cbx2 interaction with Ring1b is required for the recruitment of Ring1b protein to mitotic chromosomes, the three fusion proteins mCherry H2A, Cerulean Ring1b, and YFP Cbx2 1 498 were expressed in C bx2 KO ES cells. The YFP Cbx2 1 498 fusion protein lacks of the Chromobox (Cbox) domain required for interaction with Ring1b (Satijn et al. 1997; Schoorlemmer et al. 1997; Bardos et al. 2000) We expected that the Cbx2 mutant fusion protein would not be able to recruit Cerulean Ring1b to mitotic chromosomes. Quantitative image analysis showed that (92 7)% o f YFP Cbx2 1 498 fusion protein was accumulated at mitotic chromosomes, indicating the deletion of the Cbox domain of Cbx2 protein does not affect its mitotic chromosomal association (Fig. 10B and 10 C, also see Fig. 14 B). However, only (32 8)% of Cerulean Ring1b fusion protein associated with mitotic chromosomes (Fig. 10B and 10 D). The fraction of mitotic retention of Ringb1b fusion protein in

PAGE 29

") Cbx2 KO ES cell lines complemented with Cbx2 1 498 was similar to that observed in Cbx2 KO ES cells. These data indicate that the Cbx2 interaction with Ring1b is required for the recruitment of Cerulean Ring1b fusion protein to mitotic chromosomes. The direct recruitment of the canonical PRC1 proteins to mitotic chromosomes by Cbx2 implies that there is direct interaction between Cbx2 and PRC1 subunits at mitotic chromosomes. To test this hypothesis, we performed photobleaching fluorescence resonance energy transfer (FRET) analysis between YFP Cbx2 and Cerulean Ring1 b at mitotic chromosomes (Fig. 10 E). Fluorescence of YFP Cbx2 fusion protein at half of mitotic chromosomes was photobleached. The ratio of fluorescence intensity of Cerulean Ring1b in photobleached versus non photobleached areas was calculated and compared before and after photob leaching. Quantitative image analysis indicated that the fluorescence intensity of Cerulean Ring1b fusion protein was increased (1.5 0.1) fold by photobleaching YFP Cbx2 fusion protein, indicating that there is energy transfer between YFP Cbx2 and Cerule an Ring1b. As a control, we photobleached the YFP Cbx2 1 498 fusion protein and quantified the fluorescence change of Cerulean Ring1b fusion protein. Image analysis revealed that the fluorescence intensity of Cerulean Ring1b fusion protein after photobleach ing YFP Cbx2 1 498 fusion protein was (1.1 0.2) fold of that before photobleaching, indicating that there is no energy transfer between YFP Cbx2 1 498 and Cerulean Ring1b. Thus, these data demonstrate that YFP Cbx2 fusion protein interacts with Cerulean Ri ng1b fusion protein at mitotic chromosomes.

PAGE 30

"* Figure 10 Directly recruiting PRC1 fusion proteins to mitotic chromosomes by YFP Cbx2, but not by YFP Cbx2 1 498 (A) Confocal fluorescence images of Cerulean Ring1b, Cerulean Phc1, and Cerulean Mel18 fusion p roteins co expressed with YFP Cbx2 and mCherry H2A fusion proteins in Cbx2 / ES cells in metaphase. Scale bar is 5 !m. (B) Confocal fluorescence images of Cerulean Ring1b fusion protein co expressed with YFP Cbx2 1 498 and mCherry H2A fusion protein in Cbx2 / ES cells in metaphase. Scale bar is 5 !m. (C) Quantification of mitotic chromosomal association of YFP Cbx2 and YFP Cbx2 1 498 fusion proteins in Cbx2 / ES cells. The data represents average of at least 10 cells analyzed. Error bars indicate standa rd deviation of the mean. (D) Quantitative analysis of mitotic chromosomal association of Cerulea Ring1b, Cerulean Phc1, and Cerulean Mel18 fusion proteins co expressed with YFP Cbx2 and of Cerulea Ring1b co expressed with YFP Cbx2 1 498 in Cbx2 / ES cells in metaphase. The data represents average of at least 10 cells analyzed. Error bars indicate standard deviation of the mean. (E) Photobleaching FRET images of Cerulean Ring1b interaction with YFP Cbx2 and YFP Cbx2 1 498 at mitotic chromosomes. The YFP Cbx2 YFP Cbx2 1 498 and Cerulean Ring1b fusion proteins as indicated above images expressed in Cbx2 / ES cells. Half area of fluorescence of YFP Cbx2 or YFP Cbx2 1 498 fusion proteins at mitotic chromosomes was photobleached. Z scan imaging of live cells by c onfocal laser microscope was performed before (top panel) and after (bottom panel) photobleaching. The arrowheads indicate the bleaching areas. The Cbx2 PRC1 complex is immobilized at mitotic chromosomes, but other Cbx family proteins rapidly exchange at mitotic chromosomes Several studies have demonstrated that mammalian PRC1 proteins are highly dynamic during interphase in cells ( Hernandez Munoz et al. 2005; Ren et al. 2008; Isono et al. 2013; Vandenbunder et al. 2014) Recent studies of Drosophila Pc and Ph proteins showed that a

PAGE 31

"+ subpopulation of the two prote ins bind to mitotic chromosomes with up to 300 fold longer residence time than during interphase (Fonseca et al. 2012; Steffen et al. 2013) To determine the dynamic properties of mammalian PRC1 proteins binding to chromatin in both interphase and mitosis of ES cells, we perf ormed quantitative fluorescence recovery after photobleaching (FRAP) on the Cbx family of proteins (Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8), as well as the three core components of the canonical PRC1 complex ( Ring1b, Phc1, and Mel18) (Fig. 11 A I). The mCherry H2 A fusion protein served as a guide for placing bleach spots at mitotic chromosomes. Comparison of recovery kinetics of the Cbx family fusion proteins binding to mitotic chromosomes revealed striking differences among Cbx proteins. Over 90% of YFP Cbx2 fusi on protein was immobilized at mitotic chromosomes without exchange over a time period of 120 seconds. Conversely, over 85% of the YFP Cbx4, YFP Cbx6, YFP Cbx7 and YFP Cbx8 fusion proteins rapidly exchanged at mitotic chromosomes, with a residen ce time of 10 15 seconds (Fig. 11A 11F and 11J 11 K). During interphase, over 90% of the Cbx family fusion proteins showed fluorescence recovery with a residence tim e of about 10 20 seconds (Fig. 11A 11F and 11J 11 K), which is consistent with previous studies (Ren et al. 2008) Thus, these data reveal that the YFP Cbx2 fusion protein stably binds to mito tic chromosomes but rapidly exchanges at interphasic chromatin, while other Cbx family proteins dynamically exchange on both interphasic and mitotic chromatin.

PAGE 32

$# Figure 11 FRAP analysis of PRC1 fusion proteins binding to interphasic and mitotic chromat ins (A) Representative FRAP images of YFP Cbx2 fusion protein at metaphasic and interphasic chromatins of ES cells. The images were taken before (pre) and after (post) photobleaching. The bleaching area is indicated and outlined in white. (B I) FRAP curve s of PRC1 fusion proteins at interphases and metaphases of PGK12.1 ES cells. The FRAP curves are the normalized fluorescence intensities of the bleached areas as a function of time after photobleaching and are average of at least 8 cells. E rror bars indica te the standard deviations of means. (J) Immobile faction of PRC1 fusion protein at interphasic and metaphasic chromatins. The immobile fraction was calculated from FRAP curves by fitting first order kinetic model. The dash line is used to indicate the contrast between interphase and metaphase. The data are average

PAGE 33

$" of at least 8 cells. (K) Residence time of PRC1 fusion protein at interphasic and metaphasic chromatins. The residence time was calculated from FRAP curves by fitting first order kinetic model The bar with NA indicates that the fusion protein is immobilized at mitotic chromatin, thus the residence time is not measureable within time scale of experiments. The dash line indicates the contrast between interphase and metaphase. To ask whether the binding kinetics of the Cbx2 fusion protein is cell type specific, we performed FRAP analysis of YFP Cbx2 fusion protein in bot h HeLa and HEK293 cells (Fig. 12 and also see Fig. 14 C). Analysis of FRAP curves revealed that interphasic and mitotic dynamic properties of YFP Cbx2 fusion protein in both cell lines were similar to that seen in ES cells. Thus, these data demonstrate that the YFP Cbx2 fusion protein possesses inherently different properties of interaction with interphasic versus mitotic chromatin The various dynamic binding properties of Cbx family members to interphasic and mitotic chromatin prompted us to explore other core components of the canonic PRC1 complex. We performed FRAP analysis of Cerulean Ring1b, Cerulean Phc1 and Cerlean Mel18 fu sion proteins bound to chromatin in both interphase and mitosis in ES cells. Analysis of FRAP curves of mitotic chromosomal binding of Cerulean Ring1b, Cerulean Phc1 and Cerlean Mel18 fusion proteins revealed that near 80% of these fusion proteins stably b ind to mitotic chro mosomes without exchange (Fig. 11 G K), which is similar to YFP Cbx2, but differs from other Cbx family fusion proteins. Calculation of the recovery kinetics and measured parameters of interphasic FRAP curves of the Cerulean Ring1b, Cerul ean Phc1 and Cerlean Mel18 proteins revealed that these three fusion proteins showed complete recovery, with residence time of 10 15 seconds, and had no immobile fraction at interphasic chromatin. Thus, these data indicate that the four PRC1 fusion protein s, YFP Cbx2, Cerulean Ring1b, Cerulean Phc1, and Cerulean Mel18, bind to mitotic chromosomes with similar kinetic characteristics, yet differ from other Cbx family proteins.

PAGE 34

$$ Figure 12. FRAP analysis of YFP Cbx2 fusion protein binding to interphasic an d mitotic chromatins in HeLa and HEK293 cells The FRAP curves are the normalized fluorescence intensities of the bleached areas as a function of time after photobleaching and are average of at least 8 cells. Error bars indi cate the standard deviations of means. The immobilization of Cbx2 at mitotic chromosomes is independent of PcG proteins Since Cbx2 protein is recruited to mitotic chromosomes by a PRC2 independent mechanism, we reasoned that depletion of PRC2 complex gene Eed would not affect the immobilization of Cbx2 on mitotic chromosomes. To test the hypothesis, we performed FRAP analysis of YFP Cbx2 protein binding to both mitotic and interphasic chromosomes in Eed KO ES cells. Analysis of mitotic FRAP curves revealed tha t over 85% of YFP Cbx2 fusion protein showed no recovery of fluores cence within 120 seconds (Fig. 13 A a nd 13 D). Calculation of interphasic FRAP curves of YFP Cbx2 fusion protein revealed that the residence time for the mobile fraction is 25 seconds, slight ly higher than that observed in wild type ES cells, while the immobile fraction is the same as seen in wild type ES cells. Thus, these data indicate that the dynamics of YFP Cbx2 fusion protein binding to interphase and mitotic chromatin is independent of the PRC2 gene Eed To interrogate whether the interphasic and mitotic kinetics of Cbx2 fusion protein binding to chromatin are affected by PRC1 proteins, we performed FRAP analysis of YFP Cbx2 fusion protein in Ring1a/Ring1b double KO and Bmil1/Mel18 dou ble KO ES cells. Analysis of mitotic FRAP curves in both double KO ES cell lines revealed that more than 85% of YFP Cbx2

PAGE 35

$% showed no recovery of fluorescence, indicating that the YFP Cbx2 fusion protein binds to mitotic chro mosomes without exchange (Fig. 13B 13 D). Next, we analyzed the interphasic FRAP curves of YFP Cbx2 fusion protein in the double KO ES cell lines. Calculation of residence time and immobile fraction of YFP Cbx2 fusion protein in interphase of the two double KO ES cell lines revealed striki ng differences in compari son to wild type ES cell (Fig. 13B 13 D). The YFP Cbx2 fusion protein was much less dynamic in the double KO ES cells than in wild type ES cells, with residence time 30 35 seconds. A further difference is that the immobile fraction of YFP Cbx2 fusion protein in interphase of the double KO ES cell lines was 23 30% of total protein, which is over 2.0 fold of that seen in wild type ES cells. Thus, these data suggest that the immobilization of YFP Cbx2 fusion protein at mitotic chromosom es is independent of the PRC1 proteins Ring1a/Ring1b and Mel18/Bmi1, while the dynamic behavior of YFP Cbx2 during interphase of the cell cycle can b e affected b y the PRC1 proteins

PAGE 36

$& Figure 13 FRAP analysis of YFP Cbx2 fusion protein binding to chromatins in Eed / Ring1a / /Ring1b / and Bmi1 / /Mel18 / ES cells. (A C) FRAP curves of interphases and metaphases of YFP Cbx2 fusion protein in Eed / Ring1a / /Ring1b / and Bmi1 / /Mel18 / ES cells. The FRAP curves were normalized and plotted described as in Fig. 5. Over 8 cells were analyzed. Error bars indicate the standard deviations of means. (D) Residence time and immobile fraction of YFP Cbx2 on chromatins. The residence time and immobil e fraction was calculated from FRAP curves by fitting first order kinetic model. The bars with NA indicate that the residence time is not available due to the fact of immobilization of YFP Cbx2 on mitotic chromosomes. The dash lines indicate the contrast b etween interphase and metaphase. WT ES is denoted as PGK12.1 ES cells. The data are average of at least 8 cells. The recruitment and immobilization of Cbx2 to mitotic chromosomes requires its distinct regions To dissect the domains (regions) of Cbx2 requi red for targeting mitotic chromosomes, we generated a variety of Cbx2 mutants tagged with YFP and introduced them into HeLa cells. The mCherry H2A protein was used to mark mitotic chromosomes. Imaging of live cells by using a confocal fluorescence microsco pe showed that the deletion of the C terminus of Cbx2 protein (Cbx2 1 498 Cbx2 1 281 and Cbx2 1 194 ) does not affect targeting of the Cbx2 fusion mutant s to mitotic chromosomes (Fig. 14A and 14 B). On the other hand, deletion of the N terminus of Cbx2 (Cbx2 89 532 ) resulted in a complete loss of the Cbx2 fusion variant from mitotic chromosomes. These data suggest that the N terminus of Cbx2 protein is required for targeting the Cbx2 fusion protein to mitotic chromosomes. To explore the molecular basis for t he immobilization of Cbx2 at mitotic chromosomes, we performed FRAP assay of the YFP Cbx2 fusion variants (Cbx2 1 498 Cbx2 1 281 and Cbx2 1 194 ) during interphase and mitosis of the cell cycle. Kinetic analysis of interphasic FRAP curves of YFP Cbx2 fusion variants revealed that the residence time of three YFP Cbx2 variants was 6 9 seconds, which is half of full length Cbx2 fusion protein. In contrast to the existence of immobile fraction of Cbx2 fusion protein in interphase cells, there was no immobile frac tion for the three YFP Cbx2 fusion variants. Analysis of mitotic FRAP curves of the three YFP Cbx2

PAGE 37

$' variants showed striking kinetic differences in comparison to the YFP Cbx2 fusion protein. The three YFP Cbx2 variants rapidly exchanged at mitotic chromosom es, with residence time of 8 11 seconds. In contrast to full length YFP Cbx2, YFP Cbx2 variants became fully recovered at mitotic chromosomes. Thus, these data indicate that the immobilization of Cbx2 fusion proteins at mitotic chromosomes require its extr eme C terminus.

PAGE 38

$( Figure 14 Analysis of structural elements of YFP Cbx2 fusion protein required for its targeting and immobilizing. (A) Diagram of structural domains of Cbx2. The dash rectangles indicate the region required for targeting Cbx2 to mitot ic chromosomes (left) and the region required for immobilizing Cbx2 at mitotic chromosomes (right). CHD represents Chromodomain domain. Cbox is Chromobox domain and ATH is AT hook domain. The number in parentheses indicates the starting and ending of amino acid sequence. (B) Confocal images of YFP Cbx2 and its variant fusion proteins in metaphase of HeLa cells. YFP Cbx2 mutant and mCherry H2A fusion proteins were stably expressed in HeLa cells. The mCherry H2A was used to mark

PAGE 39

$) mitotic chromosomes. Scale bar is 5 !m. (C) FRAP curves of interphases and metaphases of YFP Cbx2 and its variant fusion proteins expressed in HeLa cells. FRAP analysis was described in Fig. 5. (D) Residence time and immobile fraction of YFP Cbx2 variant fusion prot eins at interphasic and mitotic chromatins. The residence time and immobile fraction were calculated from FRAP curves by fitting first order kinetic model.

PAGE 40

$* CHAPTER III DISCUSSION We have used quantitative live cell imaging analysis to investigate the mitotic chromosomal association of the canonical PRC1 proteins and to interrogate the dynamics of these proteins binding to chromatin in both interphase and mitosis. Our results reveal ed several striking findings, summarized as follows: (1) The canonical PRC1 subunits tested vary at the level of association with mitotic chromosomes, and Cbx2 is the primary protein accumulated at mitotic chromosomes; (2) The mitotic chromosomal associati on of Cbx2 protein is independent of PRC1 or PRC2 complex proteins; (3) The Cbx2 protein directly targets the canonical PRC1 proteins to mitotic chromosomes; (4) The Cbx2 containing PRC1 complex is immobilized at mitotic chromosomes, while other Cbx family proteins dynamically exchange at mitotic chromosomes; (5) The immobilization of Cbx2 protein at mitotic chromosomes is independent of PRC1 or PRC2 proteins; (6) The recruitment of Cbx2 protein to mitotic chromosomes requires its N terminus, while the immo bilization of Cbx2 protein at mitotic chromosomes requires its C terminus. Thus, these data provide insights into the mechanisms underlying how canonical PRC1 proteins interact with interphasic and mitotic chromatin, and also have implications for understa nding PRC1 mediated epigenetic inheritance. Early studies of mammalian PRC1 proteins by immunofluorescence in fixed cells have provided divergent opinions as to whether PRC1 proteins are substantially retained at mitotic chromosomes (Wang et al. 1997; Saurin et al. 1998; Koga et al. 1999; Voncken et al. 1999; Akasaka et al. 2002; Suzuki et al. 2002; Miyagishima et al. 2003; Aoto et al. 2008; Vincenz and Kerppola, 2008) These variations could be due to inaccessibility of the protein to the antibody, may be due to damage or loss of epitopes during the experimental procedures, or

PAGE 41

$+ differences in how cells were prepared and imaged, as well as the cell types that were used. Consistent with these notions, we noticed that there were remarkable differences in the mitotic chromosomal association of PRC1 proteins if subtle experimental variations were applied. For instance, by adding Hoechst to c ells before fixing with formaldehyde, we observed that Cbx family proteins are completely excluded from mitotic chromosomes. By fixing cells with formaldehyde before adding Hoechst, we observed that Cbx2 protein now shows a punctate pattern at mitotic chro mosomes (data not shown). Thus, we performed quantitative live cell imaging to interrogate mitotic chromosomal association and chromatin binding of PRC1 proteins. The quantitative live cell imaging requires that PRC1 proteins are fused with fluorescence pr oteins. Many PRC1 fusion proteins have been reported to function normally in cells or animals. The Cbx family proteins fused with Venus have been documented to be able to form PRC1 complex and bind to PcG target genes (Ren et al. 2008; Ren and Kerppola, 2011) The knockin mice expressing Mel18 and Ring1b proteins fused with EGFP or YFP have been shown to function normally as their endogenous counterparts (Isono et al. 2013) Drosophila Ph and Pc protei ns fused with GFP can fulfill the functions of the endogenous proteins (Fonseca et al. 2012) These data suggest that the PRC1 proteins can tolerate the addition of fluorescence protein tag. The inducible gene deliver vector used in the current studies allows us to control the level of fusion prot ein expression. The protein expression level of the test fusion protein under doxycycline concentration used is similar or slightly higher than endogenous counterparts. The immunostaining of endogenous Ring1b and Phc1 in wild type and Cbx2 knockout ES cell s also support that PRC1 proteins associates with mitotic chromosomes and that Cbx2 affects the mitotic chromosomal association of PRC1 proteins.

PAGE 42

%# Although Cbx family proteins share conserved domains (Cbox and CHD) (Simon and Kingston, 2009) accumulating evidence suggests that they have both overlapping and non overlapping functions (Core et al. 1997; Katoh Fukui et al. 1998; Vincenz and Kerppola, 2008; Forzati et al. 2012; Gao et al. 2012; Morey et al. 2012; Klauke et al. 2013) Cbx2 accumulates at mitotic chromosomes, yet other Cbx family proteins show greatly reduced association with mitotic chromosomes. Deletion of the CHD domain causes dissociation of Cbx2 variants from mitotic chromosomes, suggesting that the CHD domain plays a role in recruiting Cbx2 to mitotic chromosomes. Other Cbx f amily proteins also contain a CHD domain, but display a much reduced association with mitotic chromosomes, indicating that other unknown factors must also contribute to the unique binding properties of Cbx2. These factors may include post translational mod ifications of Cbx2, which could lead to a switch in binding platform. A previous report showed that phosphorylation of Cbx2 changes its binding specificity for methylated histone H3 (Hatano et al. 2010) Another recent report also indicated that methylation of Cbx4 switches its binding partners (Yang et al. 2011) Another possibility is that Cbx2 protein on its own has unique physical properties, for example, the intrinsic charge properties (Grau et al. 2011) It is also possible that the accumulation of Cbx2 proteins at mitotic chromosomes is due to changes of recruiting or competing molec ules. Finally, Cbx2 may form unique protein complexes at mitotic chromosomes. Further studies will help understanding mechanisms by which mitotic Cbx family proteins are selectively displaced and retained. We provided several lines of evidence to demonstrate that Cbx2 protein is essential for the recruitment of the canonical PRC1 proteins to mitotic chromosomes. First, we observed that the mitotic fraction of the three PRC1 proteins Ring1b, Phc1, and Mel18 i n Cbx2 KO ES cells are reduced at least two fold in comparison to that observed in wild type ES cells, suggesting

PAGE 43

%" that Cbx2 plays a major role in recruiting PRC1 proteins to mitotic chromosomes. Second, the mitotic chromosomal association of the three PRC1 proteins, Ring1b, Phc1, and Mel18, in Cbx2 KO ES cells can be restored by supplementing YFP Cbx2 fusion protein, but not the YFP Cbx2 1 498 fusion protein that is unable to interact with Ring1b (Satijn et al. 1997; Schoorlemmer et al. 1997; Bardos et al. 2000) Finally, we observed that the YFP Cbx2 fusion protein interacts with the Cerulean Ring1b fusion protein at mitotic chromosomes by FRET imaging. Taken together, these data reveal that Cbx2 directly recruits canonical PRC1 proteins to mitotic chromosomes. It is interesting to note that the Cbx2 containing PRC1 complex (Cbx2, Ring1b, Phc1, and Mel18) is immobilized at mi totic chromosomes without exchange, whereas other Cbx family proteins (Cbx4, Cbx6, Cbx7, and Cbx8) dynamically bind to mitotic chromosomes with kinetics similar to their binding to interphasic chromatin. It is not clear which factors dictate the transition between a dynamic and a stable Cbx2 PRC1 complex during different phases of the cell cycle. Since the C terminus of Cbx2 is required for the immobilization of Cbx2 proteins at mitotic chromosomes, the C terminus may dictate the dynamic switching between i nterphase and mitosis. The C terminus contains the Cbox domain that interacts with Ring1b (Satijn et al. 1997; Schoorlemmer et al. 1997; Bardos et al. 2000) but depletion of Ring1a/Ring1b proteins did not alter the immobilization of Cbx2 to mitotic chromosomes, suggesting that other factors play roles in immobilizing the Cbx2 PRC1 complex at mitotic chromosomes. Previous studi es of transcription factors, epigenetic regulators and chromosomal structural proteins have shown that most of them either rapidly exchange at or stably bind to chromatin (Phair et al. 2004; Cherukuri et al. 2008; Ueda et al. 2008; Souza et al. 2009; Hemmerich et al. 2011) yet a subset of these factors switch binding dynamics upon signaling stimuli or cell cycle transition (Angus et al. 2003; Schmiedeberg et al. 2004; Chen et al. 2005; Mekhail et al. 2005; Gerlich et al. 2006;

PAGE 44

%$ Meshorer et al. 2006; Yao et al. 2006; Ren et al. 2008; Giglia Mari et al. 2009; Hellwig et al. 2011; Hemmerich et al. 2011) We hypothesize that the dynamic switching of the Cbx2 PRC1 complex between interphase and mitosis may be regulated through covalent modifications or additio nal interacting partners (Fig. 15 ). The Ringro se laboratory (Fonseca et al. 2012) and the Francis laboratory (Follmer et al. 2012) identified a fraction of drosophila PRC1 prote ins association with mitotic chromosomes and the Ringrose laboratory (Fonseca et al. 2012) also revealed that 0.2 2% of PRC1 proteins (PC and PH) remains stably bound to mitotic chromatin with up to 300 fold longer residence times than in interphase, which supports our findings of the mitotic chr omosomal association and the stably binding to mitotic chromosomes of mammalian PRC1 proteins. All mammalian PRC1 proteins test in our research have the capacity of stably binding to mitotic chromosome, however the faction of their stably binding to mitoti c chromosomes varies greatly. We revealed that over 85% of the Cbx2 PRC1 complex is selectively and specifically immobilized at mitotic chromosomes, while more than 85% of other Cbx family proteins dynamically exchanges at mitotic chromosomes. Since mamma lian PRC1 complexes comprise a multiplicity of variants and are far more biochemically diverse than their drosophila counterparts, the selective immobilization of the Cbx2 PRC1 complex at mitotic chromosomes implies that the PRC1 complexes become function a lly divergent during evolution.

PAGE 45

%% Figure 15 A hypothetic model for the interaction of canonical PRC1 complex with interphasic and mitotic chromatin. In interphases of cells, Cbx2 PRC1 and Cbx4/6/7/8 PRC1 complexes dynamically bind to chromatin. During mitosis, the Cbx2 PRC1 complex is immobilized at mitotic chromosomes, whereas other Cbx family (Cbx4, Cbx6, Cbx7, and Cbx8) rapidly exchange at mitotic chromosomes. Red star implies factors such as covalent modification and protein interactor stabilize the Cbx2 PRC1 complex binding to mitotic chromosomes.

PAGE 46

%& CHAPTER IV MATERIALS AND METHODS Cell lines The Cbx2 / (Katoh Fukui et al. 1998) Ring1a / / Ring1b fl/fl ; Rosa26 :: CreERT2 ( Ring1a knockout, Ring1b conditional knockout) (Endoh et al. 2008) Bmi1 / /Mel18 / ( Bmi1 and Mel18 double knockout) (Elderkin et al. 2007) Eed / (Endoh et al. 2008) and PGK12.1 mouse ES cell (Penny et al. 1996) lines were maintained in DMEM (Sigma) suppleme nted with 15% FBS (BioExpress), 2 mM glutamine (Life Technologies), 100 units/ml penicillin G sodium (Life Technologies), 0.1 mg/ml streptomycin sulfate (Life Technologies), 0.1 mM mercaptoethanol (Life Technologies), 10 3 units/ml leukemia inhibitor fact or (LIF), and 0.1 mM non essential amino acids (Life Technologies) at 37 o C in 5% CO 2 The depletion of Ring1b alleles in Ring1a / / Ring1b fl/fl ; Rosa26 :: CreERT2 ES cells was achieved by administration of 4 Hydroxytamoxifen (OHT, Sigma) for three days under the concentration of 1.0 !M prior to experiment HeLa, HEK293, and HEK293T cells were maintained in DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/ml penicillin G sodium, 0.1 mg/ml streptomycin sulfate at 37 o C in 5% CO 2 Culture medium wa s replaced w ith fresh medium every 24 hours and cells lines were split at 66 75 % confluence during maintenances. Plasmids The pTRIPZ shRNAmir Lentivirus vector (Open Biosystems) was engineered to remove both the turboRFP and the regulatory sequences of s hRNAmir to produce pTRIPZ(M). The sequences coding Cerulean (Addgene), YFP (Ren et al. 2008) and mCherry (Addgene) fluorescence proteins were amplified by PCR and inserted to the pTRIPZ(M) to produce vectors pTRIPZ(M) Cerulean, pTRIPZ(M) YFP, and pTRIPZ(M) mCherry. The sequences coding

PAGE 47

%' Ring1b (Ren et al. 2008) Phc1 (Addgene), Mel18 (Addgene), H2A (Addgene), Cbx2 (Ren et al. 2008) Cbx4 (Ren et al. 2008) Cbx6 (Ren et al. 2008) Cbx7 (Ren et al. 2008) and Cbx8 (Ren et al. 2008) were amplified by PCR and inserted downstream of the coding sequence of fluorescence protein in pTRIPZ(M) vector. The same strategy was used to construct Cbx2 variants tagged with YFP. The Cbx2 variants were as follows: (1) Cbx2 1 498 deletion of amino acids (499 532); (2) Cbx2 1 281 deletion of amino acids (282 532); (3) Cbx2 1 194 deletion of amino acids (195 532); (4) Cbx2 89 532 d eletion of amino acids (1 88). The sequences encoding fusion proteins have been verified by DNA sequencing. Generation of stable cell lines by lentivirus infection H EK 293T cells density of 3.5 4.0 # 10 6 were seeded on 10 cm culture dish 24 hours prior at the time of transfection to reach a monolayer of 90 % confluence. Cells were co transfected by c alcium phosphate precipitation Transfection particles are composed of 21 !g pTRIPZ(M) containing the gene of interest, 21 !g psPAX2, and 10.5 !g pMD2.G wer e first mixed by vortex at max speed for 30 seconds follow the by addition of cold CaCl 2 drop wise at a final concentration of .128 M and vortex at max speed for 40 seconds. Cold 2 HBSS was added drop wise at 3 seconds interval to the mixture while under slow vortex speed. Final mixture was vortex at max speed for 40 seconds and was incubated at room temperature for 30 mins undisturbed. After incubation, mixture was vortex for 20 seconds at max speed and added drop wise to cell plate. After 12 hours of tra nsfection, culture medium was replaced with fresh DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/mL pencillin G sodium, and 0.1 mg/mL streptomycin sulfate for 48 hours of incubation. Pseudovirus medium was collected and centrifuged at 1,600 g fo r 10 min at 4 de gree. The supernatants were collected and used for infection of mouse ES, HeLa, and HEK293 cells. If co expression of multiple proteins needed,

PAGE 48

%( lentiviruses were produced separately and mixed at the time of infection. Polybrene (sigma) was added at the final concentration of 8 !g/mL and the cells were seeded at ~15% confluence on gelatin coated plates or mitotically inactivated MEF cells. 16 hours after transduction, the medium was replaced with fresh medium. After 2 medium changes at 24 hou rs interval, infected cells were selected with 1.0 2.0 !g/ml puromycin (Life Technologies). The expression of transgenes was induced with doxycycline (Sigma) at concentration of 0.1 1.0 !g/ml. Confocal microscope imaging of live cells and quantification of mitotic fraction Zeiss LSM 700 observer Z1 equipped with a 100 # oil objective (numerical aperture, 1.4) and an EMCCD camera was used for Z scan imaging. For Cerulean fluorescence, 435 nm excitation and 476 nm emission filters were used. For YFP fluorescence, 514 nm excita tion and 527 nm emission filters were used. For mCherry fluorescence, 587 nm excitation and 610 nm emission filters were used. The section size was 1 !m for the three fluorescence proteins Cerulean, YFP and mCherry. The frame size was 512 # 512 pixels. Sca n time is 1.56 sec. Average of images was 4. For live cell imaging, cells were incubated with doxycycline (0.1 1.0 !M) to induce protein expression. 48 hours after induction of protein expression, cells were seed to overnight gelatin coated cover glass di sh (Mat Teck Corp) in the presence of doxycycline. One day after seeding, medium was replaced with either Ring buffer (155 mM NaCl, 5 mM KCl, 2 mM CaCl 2 1 mM MgCl 2 2 mM NaH 2 PO 4 10 mM HEPES, 10 mM glucose, pH7.2) or phenol free DMEM supplemented with 10% FBS, 2 mM glutamine, 100 units/ml penicillin G sodium, 0.1 mg/ml streptomycin sulfate. Cells were maintained at 37 o C using Heater controller (Model TC 324, Warner Instrument Corp) during imaging. The grayscale images were converted into pseudo color, mer ged, and cropped using Adobe Photoshop.

PAGE 49

%) The Z stack movies were exported as individual images using Zeiss Zen software. The intensity of each imaging section was quantified using Image J software. The mean fluorescence intensities of a region of interes t corresponding to the mitotic chromosomes of the metaphase plate marked with H2A were measured. The mean fluorescence intensities of cell nucleus were also measured. The fluorescence intensities of a region without cells were measured as background. The f luorescence values were sum of intensities of individual sections of 3 D stack. The mitotic fraction M R was calculated as R M = ( I m i t o s i s I b ) / ( I n u c l e u s I b ) where itosis m I is the fluorescence intensities of PRC1 fusion protein at mitotic chromosomes; nucleus I is the fluorescence intensities of PRC1 fusi on protein in cell nucleus; b I is the background fluorescence intensities corresponding to region without cells. FRAP imaging and quantification FRAP imaging was performed using Zeiss LSM 700 observer. Cells were maintained as described in confocal laser sc anning imaging of live cells. The expression of fusion proteins was induced by 0.1 0.2 !M doxycycline for 2 days. The pinhole is fully open for FRAP imaging. The scan speed is 1.56 sec. Images were taken without average. Before photobleaching, four images were taken. Immediately after photobleaching, 30 images were taken with 5 sec intervals. The images were analyzed and fluorescence intensities were quantified using Image J software. To correct for movement in the x y plane, the images were aligned using T urboReg. The fluorescence intensities were corrected for fluctuations in background and total signal and normalized to the signal prior to bleaching to obtain the fluorescence recovery (I R ) as described previously (Ren et al. 2008) I R was plotted as a function of time (t) after bleaching. FRAP curves were fitted by one binding state kinetic model 1 ) ( t k R e m I = where m is the mobile

PAGE 50

%* fraction. The immobile fraction was calculated as ). 1 ( m The residence time was calculated as ). / 1 ( k FRET imaging FRET measurem ents were performed using Zeiss LSM 700 observer. Two images were acquired in the same field of view in the Cerulean Ring1b (donor) and YFP Cbx2 (donor) channels. Half area of YFP Cbx2 at mitotic chromosomes was bleached with 514 nm laser, and a second set of images of Cerulean Ring1b and YFP Cbx2 were acquired. The FRET ratio was calculated as ) /( ) ( nonbleach pre nonbleach post bleach pre bleach post F I I I I R ! = where I post bleach and I pre bleach are the mean fluorescence intensities of Cerulean Rinbg1b after and before bleaching at the bleached half area, respectively; I p ost nonbleach and I pre nonbleach are the mean fluorescence intensities of Cerulean Rinbg1b after and before bleaching at the nonphotobleached half area, respectively. Epifluorescence imaging of live cells The images were acquired using an Axio Observer D1 Microscope (Zeiss) equipped with a 100 # oil objective (numerical aperture, 1.4) and an EMCCD camera. For Cerulean fluorescence, 438/24 nm excitation and 483/32 nm emission filters were used. For YFP fluor escence, 500/24 nm excitation and 542/27 emission filters were used. For mCherry fluorescence, 560/10 nm excitation and 610/35 nm emission filters were used. For Hoechst fluorescence, 387/11 nm excitation and 447/60 emission filters were used. For live ce ll imaging, cells were maintained as described above. Images were presented as described above. Cell synchronization and fractionation ES cells synchronization was carried out as described in previous report (Ballabeni et al. 2011) Briefly, ES cell were cultured in the presence of 1.25 mM thymidine (Sigma, T1895 1G)

PAGE 51

%+ for 14 hrs. After removal of medium and washing with PBS, to the plate fresh ES cell medium with 200 ng/ml nocodazole (Sigma, M1404 2MG) was added and cells were cultured for 7 hrs. Cells were harvested and washed with PBS followed by chromatin isolation or immunofluorescence. Ch romatin fractionation was performed as in (Mendez and Stillman, 2000; Follmer et al. 2012) with minor modifications. To prepare total cell extracts (TCE), synchronized (mitotic) and non synchronized (control) cells were directly resuspended in Laemmli buffer followed by sonication. To isolate chromatin, control and mitotic cells were resuspended in (5 # 10 7 cells/ml) in buffer A (10 mM HEPES, pH7.9, 10 mM KCl, 1.5 mM MgCl 2 0.34 M Sucrose, 10% glycerol, 0.1% Triton X 100, protein inhibitors (Sigma, P8340), 0.2 mM PMSF, 1.0 mM DTT), and incubated on ice for 5 min. Nuclei (P1) was c ollected by centrifugation (1,300 g, 4 min, 4 o C). The supernatant (S1) was centrifuged (13,000 g, 15 min, 4 o C) to give supernatant (S2) and pellet (P2). P1 was washed once with buffer A and incubated with buffer B (3 mM EDTA, 0.2 mM EGTA, 0.2 mM PMSF, 1. 0 mM DTT and protein inhibitors (Sigma, P8340)). The sample was centrifuged (1,700 g, 4 min, 4 o C) to give supernatant (S2) and pellet (P3). P3 was resuspended in Laemmli buffer followed by sonication. The fractions were run on 4 12% SDS PAGE gels, and imm unoblotted. Immunoblotting To quantify protein level, c ell culture medium was aspirated and was washed with PBS. Citric saline buffer was added to cell and incubated for 10 min for detachment. Buffer was collected with cells for centrifuge at 500 g to pelleted. After centrifuge, t he supernatant was removed and was re washed with PBS. Cell pellet was lysed with buffer (20 mM Tris HCl, pH 7.4, 2.0% NP 40, 1.0% Triton X 100, 500 mM NaCl, 0.25 mM EDTA, 0.1 mM Na 3 VO 4 0.1

PAGE 52

&# mM PMSF, and protease inhibitors (Si gma, P8340)). L ysate was rocked for 30 min and centrifuge with 15000 g for 20 min at 4 o C. After centrifugation, supernatant was collected. The protein concentration was quantified and normali zed to the same concentration. Dithiothreitol (sigma) and SDS loading, Laemmli buffer (BioRad) was added to protein and heated at 90 o C for 10 min. Proteins were s eparated using SDS PAGE (Novex NP0322 ) gel at 120 volts, 400 amp, and 120 min. Separated proteins were transferred to 0.45 m Immobilon FL PVDF membr ane (Millipore) by Transfer blot (bioRad SD cell ) at 20 volt, 400 amp, and 60 min. PVDF membrane incubated in blocking buffer compose of PBS (sigma) solution supplemented with 5 % non fat milk (Labscientific, M0841) and .1 % triton x 100 (Sigma, T8787) ov er night at 4 o C. PVDF membrane is washed with PBS + .1 % triton x 100 (Sigma, T8787) 3 times at room temperature with rocking for 20 min each. PVDF membrane probed with anti Cbx2 (Abcam, ab80044), anti Phc1 (Active motif, 6 1 3), anti Ring1b (MBL, D139 3), and ant i Mel18 (Santa Cruz, sc 10744) were diluted in anti body incubation buffer contain 50 % PBS + .1 % triton x 100 (Sigma, T8787) solution and 50 % Odyssey blocking buffer (LI COR, 92 7 4000) for overnight at 4 o C with gently rocking. After ove r night incubation, PVDF membrane is washed with PBS + .1 % triton x 100 (Sigma, T8787) 3 times at room temperature with rocking for 20 min each. Secondary anti rabbit or mouse diluted in anti body buffer and incubate with PVDF for one hour at room tempera ture with gentle rocking. Proteins were detected using ECL Plus detection reagents (GE Healthcare). Membranes were imaged using a ChemiDoc XRS system (Bio Rad). Immunofluorescence Wild type and Cbx2 knockout ES cells were plated on overnight gelatin cov erslips and cultured for 24 hrs. Cell were washed with PBS and fixed with 1.0% paraformaldehyde for 10

PAGE 53

&" min. After fix, cells are washed 2 times with PBS and incubated with 0.2% Triton X 100 for 10 min. Two more wash with PBS were performed and cell s were incubated with blocking buffer (basic blocking buffer plus 3% goat serum and 3% BSA) for 1 hour at room temperature. After incubation, cells are rinsed with basic blocking buffer (10 mM PBS, pH 7.2, 0.1% Triton X 100, 0.05% Tween 20) 3 times for 5 mi n each. Follow by wash; cells were incubated with primary antibody that was diluted in blocking buffer for 2 hours at room temperature. After incubation, cells were rinsed with basic blocking buffer 2 time s for 5 mins each and incubate with secondary antib ody diluted in blocking buffer for 1 hour in dark. Cells were rinsed with PBS 3 times and basic blocking buffer 2 times for 5 min each. Cells are stained with PBS + .1 ug/ml Hoechst for 10 min and rinsed with PBS 2 times in the dark. Coverslip is mounted o n slide with ProLong¨ Antifade reagents (Life Technologies). The primary antibodies were used as follows: anti Phc1 (Active motif, 6 1 3) and anti Ring1b (MBL, D139 3). The primary antibodies were detected using FITC labeled goat anti mouse antibodies (Sig ma). To immunostain synchronized cells, after trypsinization, mitotic cells were collected by centrifugation and washed with PBS. Cells were spun onto glass slides at 1,000 rpm for 10 minutes in a Shandon Cytospin 2. Cells were fixed with 2.0% formaldehyd e at room temperature for 10 min and immunostained as described above. The primary antibody is anti Histone H3 (phospho S10) antibody (Abcam, ab5176) and was detected using FITC labeled goat anti rabbit antibodies (Sigma)

PAGE 54

&$ REFERENCES Akasaka, T., Takahashi, N., Suzuki, M., Koseki, H., Bodmer, R., and Koga, H. (2002). MBLR, a new RING finger protein resembling mammalian Polycomb gene products, is regulated by cell cycle dependent phosphorylation. Genes Cells 7 835 850. Angus, S.P., Solomon, D.A., Kuschel, L., Hennigan, R.F., and Knudsen, E.S. (2003). Retinoblastoma tumor suppressor: analyses of dynamic behavior in living cells reveal multiple modes of regulation. Mol Cell Biol 23 8172 8188. Aoto, T., Saitoh, N ., Sakamoto, Y., Watanabe, S., and Nakao, M. (2008). Polycomb group protein associated chromatin is reproduced in post mitotic G1 phase and is required for S phase progression. J Biol Chem 283 18905 18915. Bardos, J.I., Saurin, A.J., Tissot, C., Duprez, E ., and Freemont, P.S. (2000). HPC3 is a new human polycomb orthologue that interacts and associates with RING1 and Bmi1 and has transcriptional repression properties. J Biol Chem 275 28785 28792. Bernstein, E., Duncan, E.M., Masui, O., Gil, J., Heard, E., and Allis, C.D. (2006). Mouse polycomb proteins bind differentially to methylated histone H3 and RNA and are enriched in facultative heterochromatin. Mol Cell Biol 26 2560 2569. Boukarabila, H., Saurin, A.J., Batsche, E., Mossadegh, N., van Lohuizen, M., Otte, A.P., Pradel, J., Muchardt, C., Sieweke, M., and Duprez, E. (2009). The PRC1 Polycomb group complex interacts with PLZF/RARA to mediate leukemic transformation. Genes Dev 23 1195 1206.

PAGE 55

&% Cao, R., Wang, L., Wang, H., Xia, L., Erdjument Bromage, H., T empst, P., Jones, R.S., and Zhang, Y. (2002). Role of histone H3 lysine 27 methylation in Polycomb group silencing. Science 298 1039 1043. Chen, D., Dundr, M., Wang, C., Leung, A., Lamond, A., Misteli, T., and Huang, S. (2005). Condensed mitotic chromatin is accessible to transcription factors and chromatin structural proteins. J Cell Biol 168 41 54. Cheng, B., Ren, X., and Kerppola, T.K. (2 014). KAP1 Represses Differentiation Inducible Genes in Embryonic Stem Cells through Cooperative Binding with PRC1 and Derepresses Pluripotency Associated Genes. Mol Cell Biol 34 2075 2091. Cherukuri, S., Hock, R., Ueda, T., Catez, F., Rochman, M., and Bu stin, M. (2008). Cell cycle dependent binding of HMGN proteins to chromatin. Mol Biol Cell 19 1816 1824. Core, N., Bel, S., Gaunt, S.J., Aurrand Lions, M., Pearce, J., Fisher, A., and Djabali, M. (1997). Altered cellular proliferation and mesoderm pattern ing in Polycomb M33 deficient mice. Development 124 721 729. de Napoles, M., Mermoud, J.E., Wakao, R., Tang, Y.A., Endoh, M., Appanah, R., Nesterova, T.B., Silva, J., Otte, A.P., Vidal, M., Koseki, H., and Brockdorff, N. (2004). Polycomb group proteins Ri ng1A/B link ubiquitylation of histone H2A to heritable gene silencing and X inactivation. Developmental Cell 7 663 676. Di Croce, L., and Helin, K. (2013). Transcriptional regulation by Polycomb group proteins. Nat Struct Mol Biol 20 1147 1155. Elderkin, S., Maertens, G.N., Endoh, M., Mallery, D.L., Morrice, N., Koseki, H., Peters, G., Brockdorff, N., and Hiom, K. (2007). A phosphorylated form of mel 18 targets the Ring1B histone H2A ubliquitin ligase to chromatin. Molecular Cell 28 107 120.

PAGE 56

&& Endoh, M., Endo, T.A., Endoh, T., Fujimura, Y.I., Ohara, O., Toyoda, T., Otte, A.P., Okano, M., Brockdorff, N., Vidal, M., and Koseki, H. (2008). Polycomb group proteins Ring1A/B are functionally linked to the core transcriptional regulatory circuitry to m aintain ES cell identity. Development 135 1513 1524. Follmer, N.E., Wani, A.H., and Francis, N.J. (2012). A polycomb group protein is retained at specific sites on chromatin in mitosis. PLoS Genet 8 e1003135. Fonseca, J.P., Steffen, P.A., Muller, S., Lu, J., Sawicka, A., Seiser, C., and Ringrose, L. (2012). In vivo Polycomb kinetics and mitotic chromatin binding distinguish stem cells from differentiated cells. Genes Dev 26 857 871. Forzati, F., Federico, A., Pallante, P., Abbate, A., Esposito, F., Malap elle, U., Sepe, R., Palma, G., Troncone, G., Scarfo, M., Arra, C., Fedele, M., and Fusco, A. (2012). CBX7 is a tumor suppressor in mice and humans. J Clin Invest 122 612 623. Gao, Z., Zhang, J., Bonasio, R., Strino, F., Sawai, A., Parisi, F., Kluger, Y., and Reinberg, D. (2012). PCGF Homologs, CBX Proteins, and RYBP Define Functionally Distinct PRC1 Family Complexes. Mol Cell 45 344 356. Gearhart, M.D., Corcoran, C.M., Wamstad, J.A., and Bardwell, V.J. (2006). Polycomb group and SCF ubiquitin ligases are found in a novel BCOR complex that is recruited to BCL6 targets. Mol Cell Biol 26 6880 6889. Gerlich, D., Hirota, T., Koch, B., Peters, J.M., and Ellenberg, J. (2006). Condensin I stabilizes chromosomes mechanically through a dynamic interaction in live c ells. Curr Biol 16 333 344. Giglia Mari, G., Theil, A.F., Mari, P.O., Mourgues, S., Nonnekens, J., Andrieux, L.O., de Wit, J., Miquel, C., Wijgers, N., Maas, A., Fousteri, M., Hoeijmakers, J.H., and Vermeulen,

PAGE 57

&' W. (2009). Differentiation driven changes in the dynamic organization of Basal transcription initiation. PLoS Biol 7 e1000220. Grau, D.J., Chapman, B.A., Garlick, J.D., Borowsky, M., Francis, N.J., and Kingston, R.E. (2011). Compaction of chromatin by diverse Polycomb group proteins requires localiz ed regions of high charge. Genes Dev 25 2210 2221. Hatano, A., Matsumoto, M., Higashinakagawa, T., and Nakayama, K.I. (2010). Phosphorylation of the chromodomain changes the binding specificity of Cbx2 for methylated histone H3. Biochem Biophys Res Commun 397 93 99. Hellwig, D., Emmerth, S., Ulbricht, T., Doring, V., Hoischen, C., Martin, R., Samora, C.P., McAinsh, A.D., Carroll, C.W., Straight, A.F., Meraldi, P., and Diekmann, S. (2011). Dynamics of CENP N kinetochore binding during the cell cycle. J Cel l Sci 124 3871 3883. Hemmerich, P., Schmiedeberg, L., and Diekmann, S. (2011). Dynamic as well as stable protein interactions contribute to genome function and maintenance. Chromosome Res 19 131 151. Hernandez Munoz, I., Taghavi, P., Kuijl, C., Neefjes, J., and van Lohuizen, M. (2005). Association of BMI1 with polycomb bodies is dynamic and requires PRC2/EZH2 and the maintenance DNA methyltransferase DNMT1. Molecular and Cellular Biology 25 11047 11058. Isono, K., Endo, T.A., Ku, M., Yamada, D., Suzuki, R., Sharif, J., Ishikura, T., Toyoda, T., Bernstein, B.E., and Koseki, H. (2013). SAM domain polymerization links subnuclear clustering of PRC1 to gene silencing. Dev Cell 26 565 577.

PAGE 58

&( Katoh Fukui, Y., Tsuchiya, R., Shiroishi, T., Nakahara, Y., Hashimoto, N., Noguchi, K., and Higashinakagawa, T. (1998). Male to female sex reversal in M33 mutant mice. Nature 393 688 692. Kerppola, T.K. (2009). Polycomb group complexes -many combinations, many functions. Trends Cell Biol 19 692 704. Klauke, K., Radulovic, V., Broekhuis, M., Weersing, E., Zwart, E., Olthof, S., Ritsema, M., Bruggeman, S., Wu, X., Helin, K., Bystrykh, L., and de Haan, G. (2013). Polycomb Cbx family members mediate the balance between haematopoietic stem cell self renewal and differentiation. Nat Cell Biol. Koga, H., Matsui, S., Hirota, T., Takebayashi, S., Okumura, K., and Saya, H. (1999). A human homolog of Drosophila lethal(3)malignant brain tumor (l(3)mbt) protein associates with condensed mitotic chromosomes. Oncogene 18 3799 3809. Mekhai l, K., Khacho, M., Carrigan, A., Hache, R.R., Gunaratnam, L., and Lee, S. (2005). Regulation of ubiquitin ligase dynamics by the nucleolus. J Cell Biol 170 733 744. Mendez, J., and Stillman, B. (2000). Chromatin association of human origin recognition complex, cdc6, and minichromosome maintenance proteins during the cell cycle: assembly of prereplication complexes in late mitosis. Mol Cell Biol 20 8602 8612. Meshorer E., Yellajoshula, D., George, E., Scambler, P.J., Brown, D.T., and Misteli, T. (2006). Hyperdynamic plasticity of chromatin proteins in pluripotent embryonic stem cells. Dev Cell 10 105 116. Miyagishima, H., Isono, K., Fujimura, Y., Iyo, M., Takihara, Y ., Masumoto, H., Vidal, M., and Koseki, H. (2003). Dissociation of mammalian Polycomb group proteins, Ring1B and

PAGE 59

&) Rae28/Ph1, from the chromatin correlates with configuration changes of the chromatin in mitotic and meiotic prophase. Histochem Cell Biol 120 111 119. Morey, L., Aloia, L., Cozzuto, L., Benitah, S.A., and Di Croce, L. (2013). RYBP and Cbx7 define specific biological functions of polycomb complexes in mouse embryonic stem cells. Cell Rep 3 60 69. Morey, L., Pascual, G., Cozzuto, L., Roma, G., Wu tz, A., Benitah, S.A., and Di Croce, L. (2012). Nonoverlapping functions of the Polycomb group Cbx family of proteins in embryonic stem cells. Cell Stem Cell 10 47 62. Ogawa, H., Ishiguro, K., Gaubatz, S., Livingston, D.M., and Nakatani, Y. (2002). A comp lex with chromatin modifiers that occupies E2F and Myc responsive genes in G(0) cells. Science 296 1132 1136. Park IK et al. (2003) Bmi 1 is required for maintenance of adult self renewing haematopoietic stem cells. Nature 423:302 305 Penny G.D., Kay, G.F., Sheardown, S.A., Rastan, S., and Brockdorff, N. (1996). Requirement for Xist in X chromosome inactivation. Nature 379 131 137. Phair, R.D., Scaffidi, P., Elbi, C., Vecerova, J., Dey, A., Ozato, K., Brown, D.T., Hager, G., Bustin, M., and Misteli, T. (2004). Global nature of dynamic protein chromatin interactions in vivo: three dimensional genome scanning and dynamic interaction networks of chromatin proteins. Mol Cell Biol 24 6393 6402. Ren, X., and Kerppola, T.K. (2011). REST interacts with Cbx proteins and regulates polycomb repressive complex 1 occupancy at RE1 elements. Mol Cell Biol 31 2100 2110.

PAGE 60

&* Ren, X., Vincenz, C., and Kerppola, T.K. (2008). Changes in the Distributions and Dyn amics of Polycomb Repressive Complexes During Embryonic Stem Cell Differentiation. Mol Cell Biol 28 2884 2895 Sanchez, C., Sanchez, I., Demmers, J.A.A., Rodriguez, P., Strouboulis, J., and Vidal, M. (2007). Proteomics analysis of Ring1B/Rnf2 interactors i dentifies a novel complex with the Fbxl10/Jhdm1B histone demethylase and the Bcl6 interacting corepressor. Molecular & Cellular Proteomics 6 820 834. Satijn, D.P., Gunster, M.J., van der Vlag, J., Hamer, K.M., Schul, W., Alkema, M.J., Saurin, A.J., Freemo nt, P.S., van Driel, R., and Otte, A.P. (1997). RING1 is associated with the polycomb group protein complex and acts as a transcriptional repressor. Mol Cell Biol 17 4105 4113. Saurin, A.J., Shiels, C., Williamson, J., Satijn, D.P., Otte, A.P., Sheer, D., and Freemont, P.S. (1998). The human polycomb group complex associates with pericentromeric heterochromatin to form a novel nuclear domain. J Cell Biol 142 887 898. Schmiedeberg, L., Weisshart, K., Diekmann, S., Meyer Zu Hoerste, G., and Hemmerich, P. (2 004). High and low mobility populations of HP1 in heterochromatin of mammalian cells. Mol Biol Cell 15 2819 2833. Schoorlemmer, J., MarcosGutierrez, C., Were, F., Martinez, R., Garcia, E., Satijn, D.P.E., Otte, A.P., and Vidal, M. (1997). Ring1A is a tra nscriptional repressor that interacts with the Polycomb M33 protein and is expressed at rhombomere boundaries in the mouse hindbrain. Embo Journal 16 5930 5942.

PAGE 61

&+ Sharif, J., Endo, T.A., Ito, S., Ohara, O., and Koseki, H. (2013). Embracing change to remain the same: conservation of polycomb functions despite divergence of binding motifs among species. Curr Opin Cell Biol 25 305 313. Simon, J.A., and Kingston, R.E. (2009). Mechanisms of Polycomb gene silencing: knowns and unknowns. Nat Rev Mol Cell Biol 10 697 708. Simon, J.A., and Kingston, R.E. (2013). Occupying chromatin: Polycomb mechanisms for getting to genomic targets, stopping transcriptional traffic, and st aying put. Mol Cell 49 808 824. Souza, P.P., Volkel, P., Trinel, D., Vandamme, J., Rosnoblet, C., Heliot, L., and Angrand, P.O. (2009). The histone methyltransferase SUV420H2 and Heterochromatin Proteins HP1 interact but show different dynamic behaviours. BMC Cell Biol 10 41. Steffen, P.A., Fonseca, J.P., Ganger, C., Dworschak, E., Kockmann, T., Beisel, C., and Ringrose, L. (2013). Quantitative in vivo analysis of chromatin binding of Polycomb and Trithorax group proteins reveals retention of ASH1 on mito tic chromatin. Nucleic Acids Res 41 5235 5250. Suzuki, M., Mizutani Koseki, Y., Fujimura, Y., Miyagishima, H., Kaneko, T., Takada, Y., Akasaka, T., Tanzawa, H., Takihara, Y., Nakano, M., Masumoto, H., Vidal, M., Isono, K., and Koseki, H. (2002). Involvement of the Polycomb group gene Ring1B in the spec ification of the anterior posterior axis in mice. Development 129 4171 4183. Tavares, L., Dimitrova, E., Oxley, D., Webster, J., Poot, R., Demmers, J., Bezstarosti, K., Taylor, S., Ura, H., Koide, H., Wutz, A., Vidal, M., Elderkin, S., and Brockdorff, N. (2012). RYBP PRC1 Complexes Mediate H2A Ubiquitylation at Polycomb Target Sites Independently of PRC2 and H3K27me3. Cell.

PAGE 62

'# Ueda, T., Catez, F., Gerlitz, G., and Bustin, M. (2008). Delineation of the protein module that anchors HMGN proteins to nucleosomes i n the chromatin of living cells. Mol Cell Biol 28 2872 2883. Vandamme, J., Volkel, P., Rosnoblet, C., Le Faou, P., and Angrand, P.O. (2011). Interaction proteomics analysis of polycomb proteins defines distinct PRC1 complexes in mammalian cells. Mol Cell Proteomics 10 M110 002642. Vandenbunder, B., Fourre, N., Leray, A., Mueller, F., Volkel, P., Angrand, P.O., and Heliot, L. (2014). PRC1 components exhibit different binding kinetics in Polycomb bodies. Biol Cell 106 111 125. Vincenz, C., and Kerppola, T. K. (2008). Different polycomb group CBX family proteins associate with distinct regions of chromatin using nonhomologous protein sequences. Proc Natl Acad Sci U S A 105 16572 16577. Voncken, J.W., Schweizer, D., Aagaard, L., Sattler, L., Jantsch, M.F., an d van Lohuizen, M. (1999). Chromatin association of the Polycomb group protein BMI1 is cell cycle regulated and correlates with its phosphorylation status. Journal of Cell Science 112 4627 4639. Wang, G., Horsley, D., Ma, A., Otte, A.P., Hutchings, A., Bu tcher, G.W., and Singh, P.B. (1997). M33, a mammalian homologue of Drosophila Polycomb localises to euchromatin within interphase nuclei but is enriched within the centromeric heterochromatin of metaphase chromosomes. Cytogenet Cell Genet 78 50 55.

PAGE 63

'" Wang, H.B., Wang, L.J., Erdjument Bromage, H., Vidal, M., Tempst, P., Jones, R.S., and Zhang, Y. (2004a). Role of histone H2A ubiquitination in polycomb silencing. Nature 431 873 878. Wang, L., Brown, J.L., Cao, R., Zhang, Y., Kassis, J.A., and Jones, R.S. (20 04b). Hierarchical recruitment of polycomb group silencing complexes. Mol Cell 14 637 646. Yang, L., Lin, C., Liu, W., Zhang, J., Ohgi, K.A., Grinstein, J.D., Dorrestein, P.C., and Rosenfeld, M.G. (2011). ncRNA and Pc2 methylation dependent gene relocati on between nuclear structures mediates gene activation programs. Cell 147 773 788. Yao, J., Munson, K.M., Webb, W.W., and Lis, J.T. (2006). Dynamics of heat shock factor association with native gene loci in living cells. Nature 442 1050 1 053