University of Southern California Dissertations and Theses

Characterization of a new chromobox protein 8 (CBX8) antagonist in a model of human colon cancer

(USC Thesis Other)

Characterization of a new chromobox protein 8 (CBX8) antagonist in a model of human colon cancer

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Characterization of a New Chromobox Protein 8 (CBX8)
Antagonist in a Model of Human Colon Cancer

By

Yibo Si

A Thesis Presented to the
FACULTY OF THE KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(Biochemistry and Molecular Medicine)

August 2020

Copyright 2020 Yibo Si
ii

Acknowledgements
I would like to thank mt committee members, Dr. Oliver Bell, Dr. Michael Stallcup, Dr.
Peggy Farnham and Dr. Shannon Mumenthaler, for their supports in my project and thesis.
I would like to specially thank Dr. Oliver Bell for his patient guidance and generous support
as a mentor in both my academics and personal life. His dedication led me to the palace
of science.

I would like to thank all members of the Bell’s laboratory being very kind and helpful during
my time here. I would like to thank Silvia Golumbeanu, Qing Zhao, Saiyyada Mohsina
Shahid. I would like to specifically thank Daniel Bsteh for his generous help with the next
generation sequencing data analysis. I would also like to thank Roy Lau from Dr.
Mumenthaler’s laboratory for his contribution to the generation of proliferation data for the
figure 3.1.

iii

Contents
Acknowledgements ..........................................................................................................ii
List of Tables ...................................................................................................................iv
List of Figures .................................................................................................................. v
Abstract ...........................................................................................................................vi
Chapter 1: Introduction .................................................................................................... 1
1.1 Chromatin and Gene Expression Regulation ..................................................... 1
1.2 Polycomb Repressive Complexes (PRCs) ......................................................... 3
1.3 Chromobox Proteins (CBXs) .............................................................................. 7
Chapter 2: Characterizing UNC7040 with reporter cell line ..........................................11
2.1 Introduction .......................................................................................................11
2.2 Results ........................................................................................................... 14
2.3 Discussion ........................................................................................................ 15
Chapter 3: Treatment of human colon cancer cell lines with CBX8 antagonist UNC7040
...................................................................................................................................... 19
3.1 Introduction ...................................................................................................... 19
3.2 Results ............................................................................................................. 20
3.3 DIscussion........................................................................................................ 26
Chapter 4: UNC7040 displaces CBX8-containing PRC1 in LoVo cells ........................ 27
4.1 Introduction ...................................................................................................... 27
4.2 Results ............................................................................................................. 28
4.2 Discussion ........................................................................................................ 35
Chapter 5: Summary and Future Steps ....................................................................... 37
5.1 Summary .......................................................................................................... 37
5.2 Future Steps..................................................................................................... 38
Materials and Methods ................................................................................................ 40
Cell Culturing.......................................................................................................... 40
Flow Cytometry ...................................................................................................... 41
Proliferation Assays ................................................................................................ 41
Western Blot ........................................................................................................... 42
Chromatin Immunoprecipitation ............................................................................. 43
qPCR Analysis .............................................................................................................. 44
Supplementary Materials ............................................................................................... 45
References .................................................................................................................... 47
iv

List of Tables
Table 1 .......................................................................................................................... 48
Table 2 .......................................................................................................................... 48
Table 3 .......................................................................................................................... 49

v

List of Figures
Figure 1 ......................................................................................................................... 10
Figure 2.1 ...................................................................................................................... 13
Figure 2.2 ...................................................................................................................... 19
Figure 3.1 ...................................................................................................................... 23
Figure 3.2 ...................................................................................................................... 27
Figure 4.1 ...................................................................................................................... 33
Figure 4.2 ...................................................................................................................... 36

vi

Abstract

The synergy between polycomb repressive complexes (PRCs), PRC1 and PRC2, is
essential for gene expression silencing during the early development. PRC2 is recruited
in a sequence-specific manner to target genes where it catalyzes H3K27me3. In turn,
PRC1 can recognize H3K27me3 and catalyzes H2AK119ub1. Polycomb modifications
enforce chromatin compaction and genes nearby are repressed (Laugesen et al., 2019).
In mammals, PRC1 can be classified into canonical and variant complexes (cPRC1 and
vPRC1, respectively) based on the presence of Chromobox-domain (CBX) proteins. CBX
proteins are specific for cPRC1 and confer the capacity to recognize H3K27me3 and
direct chromatin binding of PRC1 at PRC2 target genes. CBXs are differentially
expressed and form mutually exclusive cPRC1 (Kaustov et al., 2011). However, even
when co-expressed in the same tissue, incorporation of individual CBX proteins endows
cPRC1 complexes with different regulation and functions (Morey et al., 2012). Recently,
overexpression of CBX8 was associated with different cancer types, including colon
cancer (Zhang et al., 2019) and breast cancer (Chung et al., 2016). Therefore, effective
and specific CBX8 antagonists might benefit basic research in elucidating the functions
of cPRC1 in development and potential translational applications in the clinic. In
collaboration with the laboratory of Stephen Frye at UNC we have leveraged a Polycomb
vii

reporter assay to screen and identify novel CBX8-specific small molecule antagonists.
Here, we characterized the CBX8 antagonist, UNC7040. In mouse embryonic stem cells,
we show that this novel peptidomimetic compound is highly efficacious in inhibiting CBX8-
dependent transcriptional reporter gene silencing. Furthermore, using a model of CBX8-
dependent colon cancer, we demonstrate efficient displacement of CBX8-containing
cPRC1 from genome which results in impaired proliferation. These results demonstrate
the UNC7040 offers a powerful chemical tool to investigate the molecular mechanism of
CBX8 function and explore inhibition of cPRC1 as a potential therapeutic option in
colorectal cancer
1

Chapter 1: Introduction
1.1 Chromatin and Gene Expression Regulation
In eukaryotes, DNA is wrapped around histone proteins forming a nucleo-protein complex
called the nucleosome (Figure 1A and B). Four core histones form an octamer consisting
of two copies each of the core histones H2A, H2B, H3, and H4. A fifth histone, H1,
associates with linker DNA between nucleosomes. One nucleosome covers about 147
base-pairs of DNAs. Multiple nucleosomes connecting with each other form the higher-
order structure, called chromatin (Gross DS et al., 2015).

Gene expression is tightly regulated at developmental stages and in tissues to keep the
homeostasis. Transcription is primarily controlled by binding of transcription factors (TFs)
to regulatory sequences within promoters and enhancers. However, in the context of
chromatin packaging, DNA sequence is not readily available for transcription factor
binding. Hence, alterations in chromatin structure superimpose another layer of regulation
(Zhang G et al., 2014).

DNA accessibility for TFs binding can be affected by DNA methylation, posttranslational
histone modifications and nucleosome positioning. DNA methylation at CpG dinucleotides
can obscure DNA sequence motifs in enhancer and promotor regions and prevent
2

recognition by TFs affecting initiation of transcription (Choy MK et al., 2010). In addition,
DNA methylation can be bound by methyl-CpG-binding domain proteins (MBDs) which
further recruit other chromatin remodeling proteins that can modify histones and form
compact chromatin (Nakao M et al., 2001).

Histone modifications have emerged as profound effectors of chromatin accessibility and
gene regulations (Chi P et al., 2010). As proteins, histones can be post-translationally
modified with chemical modifications (PTMs). Up to now, many chemical modifications
are found decorated to the different histones, such as acetylation, methylation,
phosphorylation and ubiquitination (Figure 1B). They can be generally classified into two
groups, active histone modifications and repressive histone modifications. For example,
Histone H3 lysine 27 acetylation (H3K27ac), histone H3 lysine 4 mono-methylation
(H3K4me) and histone H3 lysine 4 di/trimethylation (H3K4me2/3) are associated with
gene activation (Zhang et al, 2015), while histone H3 lysine 27 trimethylation (H3K27me3)
and histone H2A lysine 119 mono-ubiquitination (H2AK119ub1) are related with chromatin
compaction and gene repression (Zhang et al, 2015).

Other than covalent histone modifications, chromatin status can be altered by chromatin
remodeling complexes (remodelers) in an ATP-dependent manner (Clapier CR et al.,
2009). Chromatin remodeling complexes are composed of an ATPase, from SNF2 family,
3

and other proteins that are essential for the complex’s assembly and functions (Clapier
CR et al., 2009). They can be classified into four families: SWI/SNF, CHD, ISWI, and
INO80 (Längst G, et al., 2015). In general, chromatin remodelers use energy from ATP
hydrolysis to remove or translocate histone octamers, which exposes DNA to TFs or other
DNA recognizing proteins (Clapier CR et al., 2009). Chromatin remodelers can be
recruited to nucleosomes through histone modifications, specific DNA sequences, or
RNAs, depending on their different reader domains (Längst G, et al., 2015).

1.2 Polycomb Repressive Complexes (PRCs)
Polycomb repressive complexes were first identified and studied in Drosophila. In
Drosophila, PRCs are required to regulate the HOX gene clusters which are responsible
for the head-tail patterns during the different development stages (Pirrotta V, 1997).
Mutations in genes encoding PRCs led to homeotic transformations as a result of aberrant
activation of HOX gene expression at later stages of development (Entrevan M et al.,
2016). While Polycomb group (PcG) proteins are highly conserved, mammalian PRCs
have massively expanded into multiple related complexes with unique catalytic and non-
catalytic functions. In general, there are two major classes of Polycomb Repressive
Complexes, PRC1 and PRC2 (Chittock EC et al., 2017).
4

1.2.1 Polycomb repressive complex 2 (PRC2)
PRC2 is comprised of three core subunits, SUZ12, EED and EZH1/EZH2. SUZ12 is
responsible for the recognition of target loci and recruitment of the other PRC2 subunits.
EED has an essential role in maintaining the integrity of PRC2. EZH2 and its paralog,
EZH1, catalyze the mono-, di-, and tri-methylation on histone 3 lysine 27 (Cao et al., 2002;
Czermin et al., 2002; Margueron et al., 2008; Shen at al., 2008). Together with the histone-
binding proteins, RBBP4/7, they form the major functional subunits of PRC2. Recently it
was reported that by interacting with different partners, PRC2 could be subtyped into
PRC2.1 (containing PCL1-3 and EPOP/PALI) and PRC2.2 (containing JARID2 and
AEBP2) (Hauri et al., 2016) In general, non-methylated CpG islands are the targets of
PRC2 recognition (Bauer M et al., 2016).

PRC2 plays an important role in the human normal development and diseases (Laugesen
et al., 2014). Similar to that of Drosophila, PRC2 represses different HOX genes at
different stages and tissues. Significantly, PRC2 is involved in the X chromosome
inactivation (Margueron R, 2011). Also, PRC2 is essential for the pluripotency of
embryonic stem cells (ESCs). Mouse ESCs lacking Eed, Suz12 or Ezh2 display
deficiency to differentiate (Laugesen et al., 2014). Mutations on EED and EZH2 relate to
Weaver syndrome (WVS) (Cohen et al., 2015). Dysregulation of EZH2 is related to many
different cancer types. EZH2-mediated tumor suppressor silencing is related with
5

melanoma (Zingg et al., 2015). Overexpression of EZH2 abnormally activates RAF1-beta-
catenin signaling which supports the breast cancer formation and expansion (Lee et al.,
2011; Chang et al., 2011). Similarly, EZH2 can promote bladder cancer in downstream of
Rb-E2F signaling. (Santos et al., 2014). In parallel, EZH2 was shown to act as a co-
activator in supporting prostate cancer oncogenesis (Xu et al., 2012).

1.2.2 Polycomb Repressive Complex 1 (PRC1)
Generally, cPRC1 is composed of a CBX protein, Polycomb group RING finger protein
(PCGF)2 or 4, Polyhomeotic protein (PHC), and the catalytic subunit RING1A/B (Gao Z
et al., 2012). CBXs can bind to the histone modification H3K27me3 and subsequently
recruit the remaining subunits of cPRC1. RING1 is an E3 ubiquitin-protein ligase which
catalyzes the histone modification H2AK119ub which is responsible for the chromatin
compaction (Francis et al., 2004). PRC1 and PRC2 act synergistically to achieve
transcriptional gene silencing. The canonical hierarchical model of PRC2/cPRC1 is that
PRC2 catalyzes the H3K27me3 on target loci and cPRC1 is further recruited to these loci
via the binding of CBXs to the H3K27me3 modification. However, the discovery (Tavares
L et al., 2012) that RING1A/B could work independent of CBXs and RING1B occupied
many regions lacking H3K27me3 suggest that the canonical model could not explain the
PRC1 mechanism entirely.
6

Variant PRC1 lacks CBX proteins. Instead, the heterodimer of RING1A/B and PCGF
proteins associates with either RYBP or YAF2 (Gao Z et al., 2012) (Figure 1C).
Consequently, vPRC1 targeting to chromatin is independent of PRC2-mediated
H3K27me3 (Spahn L et al., 2006). vPRC1 can bind to unmethylated CpG dinucleotides
through interactions with the CXXC-domain protein KDM2B (Blackledge et al., 2014).
vPRC1s are responsible for the majority of H2AK119ub in vivo. It was demonstrated that
the H2A119ub can signal for recruitment of PRC2, suggesting that vPRC1 acts most
upstream in forming repressive Polycomb domains (Blackledge et al., 2014).

1.2.3 PRCs and Chromatin-Remodeling Trithorax Complexes
Trithorax proteins (TrxGs) were first identified as antagonist of PcGs functions in
Drosophila (Ingham PW, 1998). The opposing roles of PcGs and Trithorax proteins and
their competition in chromatin status were suggested and discussed (Geisler SJ, 2015).
Chromatin-remodeling complexes is one of the subgroups among many TrxGs. While
PRCs are responsible for the repressive histone modifications and related with chromatin
compaction, ATP-dependent chromatin remodelers are associated with accessible
chromatin and gene activation. The co-occupancy of the PcGs and TrxGs (Papp B, 2006;
Beisel et al., 2007) indicated that they were keeping the balance of chromatin ‘on/off’ in
7

different cell fates. Dysregulation of the delicate equilibrium between Polycomb and
Trithorax activities can contribute to many diseases, including cancers (Kolybaba A et al.,
2014). For example, the antagonism between PRCs and SWI/SNF complexes, TrxG-
associated complexes, was essential for the transformation of normal tissues into tumors
(Wilson BG et al., 2011). In our researches, we found that the balance of PRCs and
SWI/SNF might regulate the proliferation of a human colon cancer cell, LoVo.
1.3 Chromobox Proteins (CBXs)
In canonical PRC1, chromobox proteins are the subunits recognizing H3K27me3 with
their chromodomains and recruiting the other subunits of PRC1. In mammals, eight CBX
proteins exist. Among them, CBX2/4/6/7/8 form mutually exclusive canonical PRC1 in
vivo (Kaustov et al., 2011).

CBX8 is one of the CBXs. Unlike its paralog, CBX7, CBX8 has lower affinity for
H3K27me3 binding (Connelly et al., 2019). Interestingly, several recent reports suggested
that CBX8 may also function independently of cPRC1, especially in cancers. It is showed
that CBX8 could support MLL-AF9-induced acute myeloid leukemia (AML) independent
of cPRC1 (Tan et al., 2011). Leukemia cells proliferated much slower after CBX8
knockdown, and CBX8 interacted with TIP60 to promote the cell migration and
proliferation. CBX8 is also reported to relate to colon cancer cell growth and metastasis
8

(Zhang et al., 2019). CBX8 knockdown in colon cancer cell lines led to reduced
proliferation and migration both in vitro and in vivo (Zhang et al., 2019). Further
investigation found that m6A-mediated upregulation of CBX8 promoted LGR5 expression
in colon cancer cells, which improved their stemness and decreased the chemosensitivity.
Together with the report discussing CBX8 in breast cancer (Chung et al., 2016), they
provide evidence for CBX8 in formation and maintenance of multiple tumor types.

Considering the oncogenic role of CBX8, it is desirable to generate a CBX8-specific
inhibitor/antagonist for either scientific researches or future clinical applications. Here, in
collaboration with the lab from University of North Carolina, we successfully selected an
effective CBX8 antagonist, UNC7040, using our unique reporter cell line. Further test on
colon cancer cell lines identified a cell line, LoVo, that showed sensitivity to the UNC7040
treatment, which motivated us take advantage of it to characterize UNC7040 in human
cells as well as evaluate the role of CBX8 in LoVo cells.

9

10

Figure 1
Chromatin Modifications and Polycomb Repressive Complexes. A. DNA
organization in eukaryotes. Image modified from the online resource, ThoughtCo. B. Core
histones with diverse posttranslational N-terminal tail modifications. Image adopted from
the Thermo Fisher Scientific. C. Upper panel is showing the general mechanism of the
synergy of PRC2/canonical PRC1 in chromatin compaction and gene silencing. Lower
panel shows the PRC2-independent variant PRC1 functions in gene silencing.

11

Chapter 2: Characterizing UNC7040 with reporter cell line
2.1 Introduction
Previously, our lab developed a GFP reporter locus in mESCs which enable reversible
tethering of core subunits of PRC1 and PRC2 to investigate their contribution to
epigenetic gene silencing (Moussa & Bsteh et al., 2019). In this Tet-Off system (Figure
2.1A), the tetracycline repressor (TetR) protein is fused with CBX7. TetR can bind to
tetracycline operators (TetO) facilitating rapid CBX7 recruitment to the GFP reporter gene.
CBX7 tethering was sufficient to nucleate a functional PRC1 complex and establishment
of a repressive Polycomb chromatin domain. Notably, TetR interaction with TetO is fully
reversible upon addition of Doxycycline providing an opportunity to investigate the
capacity of the synthetic Polycomb domain to maintain silencing through DNA replications
and cell divisions. We showed that CBX7-initiated silencing could be heritably maintained
for many cell generations in the absence of the TetR-CBX7. In this current study, we
generated a new reporter cell line that enabled investigation of CBX8 function in
transcriptional gene silencing (Figure 2.1B). The reporter consisted of a green
fluorescence protein (GFP) reporter controlled by the mouse PGK promoter. Besides the
PGK promoter, seven TetO were located upstream of the reporter cassette. In the
absence of TetR fusion gene expression, this reporter could robustly express GFP (Figure
12

2.1C). Then, TetR-CBX8 along with a mCherry reporter was introduced into CP15 via
lentivirus infection. With screening and selection of the successfully infected cells, we
confirmed the capacity of the TetR-CBX8 fusion protein in repressing the GFP reporter
gene. In contrast to CBX7, GFP reporter suppression was reversible upon treatment with
doxycycline, which means our inducible system was effective (Figure 2.1C).

Our collaborator from University of North Carolina developed a new allosteric antagonist
of CBX8, UNC7040 (Figure 2.2A). As a peptide, UNC7040 was designed to bind to the
chromobox of CBX8 and thus blocking CBX8 from recognizing its target, the H3K27me3
histone modification. Among many compounds synthesized, UNC7040 stood out for its
effect in reversing the GFP repression induced by TetR-CBX8 (Figure 2.2B). In our lab,
we characterized the UNC7040 using the CP15-TetR-CBX8 reporter cell line.
13

Figure 2.1
Inducible Reporter Mouse Embryonic Stem Cell Lines. A. Tetracycline-Off system.
TRE, tetracycline (Tet) response element. tTA, Tet repressor (TetR) + transcription
activator. Image adopted from Addgene. B. Anticipated working model of the Tet-Off-
based GFP reporter system. TetR is fused with CBX8 (Moussa & Bsteh et al., 2019). C.
Flow cytometry analysis for the GFP signals of CP15 and CP15-TetR-CBX8. For
doxycycline treatment, CP15-TetR-CBX8 was treated with 2ug/mL doxycycline for 4 days.
Y axis is cell count, X axis is GFP signal. Bars are the gate set for GFP negative/positive;
Numbers are the percentages. GFP-, GFP negative. GFP+, GFP positive.

14

2.2 Results
2.2.1 TetR-CBX8 could repress GFP reporter gene expression
After the infection of reporter mESCs with lentivirus encoding a TetR-CBX8-mCherry,
mCherry-positive reporter cells were isolated by fluorescence-activated cell sorting
(FACS) and single cells were picked and expanded to generate a TetR-CBX8 reporter cell
line. Flow cytometry analysis revealed that GFP expression was largely suppressed in
the absence of doxycycline. In contrast, after we treated the CP15-TetR-CBX8 with
2ug/mL doxycycline for 4 days, nearly 100% of the cells turned on their GFP expression
(Figure 2.1C). Therefore, recruiting CBX8 alone to the reporter gene was enough to
trigger the GFP suppression, and the suppression could be controlled by the
absence/presence of tetracycline/doxycycline.

2.2.2 UNC7040 could de-repressed the GFP reporter in a dose-dependent manner
With the confidence that our reporter system worked well, we treated the TetR-CBX8
reporter cell line with different concentrations of UNC7040 to investigate its effect on the
reporter expression. After 48 hours of treatment with UNC7040, flow cytometry analysis
showed that 10%, 40% and 53% of the CP15 re-expressed the GFP reporter at 0.1uM,
1uM and 10uM treatment, respectively (Figure 2.2C). Interestingly, further increasing
concentration of UNC7040 could not improve the re-expression of GFP. It could be
15

partially explained by the phenomenon that high concentration (above 20uM) of UNC7040
would easily form aggregates in the medium, which potentially impaired its capability in
penetrating cells or decreased its actual concentration. Nevertheless, we demonstrated
that UNC7040 could reverse the suppression of GFP in a dose-dependent manner.

2.3 Discussion
The fact that UNC7040 can antagonize TetR-CBX8-induced GFP repression supported
its anticipated function in inhibiting CBX8. We previously reported that TetR-Cbx7 reporter
could maintain the GFP expression in the presence of doxycycline. It indicated that the
Polycomb repressive chromatin domain initiated by Cbx7 could be inherited through cell
generations in mESCs (Figure 2.2D) (Moussa & Bsteh et al., 2019). In contrast, GFP
repression induced by the TetR-CBX8 could not be maintained after doxycycline addition.
We have previously shown that propagation by cPRC1 required interaction between
CBX7 and H3K27me3. Since CBX8 has a lower affinity for H3K27me3 (Connelly et al.,
2019), we speculate that the feedback mechanism between PRC1 and PRC2 and their
histone modifications is less robust in case of CBX8-containing cPRC1. Nevertheless, the
lack of heritability facilitated testing of UNC7040 in our reporter system.
16

One could debate that UNC7040 had lower efficacy (about 50% at working concentration)
in reversing GFP suppression compared with that of doxycycline (88%). One explanation
is that CBX8 is not endogenously expressed in mESCs which means the initiation and
spreading of the transcription inactive chromatin modifications (e.g. H2AK119ub and
H3K27me) are operated by the ectopically expressed TetR-CBX8. UNC7040 can only
prevent the TetR-CBX8 from binding to the regions marked with H3K27me3 other than
TetO sites around the transcription starting sites (TSS), while doxycycline could remove
TetR-CBX8 from TetO-GFP. Because of the incapability of CBX8 in maintaining epigenetic
memory, the clearance of chromatin marker would be much more complete. Also,
considering that seven TetO were placed upstream of the reporter, it is understandable
that the de-repression induced by the UNC7040 was not as robust as that of doxycycline.

In the study of TetR-Cbx7, the interactions of TetR-Cbx7 between essential canonical
PRC1 subunits (e.g. RING1B) were demonstrated, and the chromatin status was also
investigated by chromatin immunoprecipitation assay. Although we can conclude that
TetR-CBX8 induce the repression via canonical PRC1, further experiments are still
required to evaluate the efficacy of our new inducible reporter system. Despite these, we
showed that TetR-CBX8 can establish repression of the GFP reporter and this is
reversible by treatment with UNC7040 in a dose-dependent manner.

17

18

Figure 2.2
UNC7040 Characterization with the Inducible Reporter Mouse Embryonic Stem Cell
Line. A. Chemical structure of the UNC7040. Image adopted from Junghyun Suh (Frye
Lab, University of North Carolina). B. Evaluating UNC7040 in CP15-TetR-CBX8 reporter
cell line. Data contributed by Junghyun Suh (Frye Lab, University of North Carolina). C.
Dose-dependent de-repression of GFP induced by UNC7040. CP15-TetR-CBX8 were
treated with 0.1-10uM UNC7040 for 2 days before sent for flow cytometry analysis. D.
Maintenance of reporter repression in TetR-Cbx7 reporter cells. Doxycycline was added
as 1ug/mL (Moussa & Bsteh et al., 2019).
19

Chapter 3: Treatment of human colon cancer cell lines with
CBX8 antagonist UNC7040
3.1 Introduction
We sought to use a CBX8-dependent cancer cell model to test the efficacy of UNC7040
in CBX8 inhibition. Knockdown of CBX8 has been shown to affect proliferation of several
human colon cancer cell lines (Zhang et al., 2019), but neither the specific oncogenic
mechanism nor the outcome of CBX8 inhibition were clear. Therefore, we decided to test
the impact of UNC7040 on proliferation of a panel of colon cancer cells.

Five human colon cancer cell lines, Caco2, DLD-1, LoVo, HCT116 and HT-29, were
selected for this study. All these cell lines are patient-derived colorectal adenocarcinomas
with different genetic backgrounds. A recent report demonstrated that CBX8 knockdown
reduced proliferation of DLD-1 and LoVo while HCT116 remained unaffected. First, we
used UNC7040 treatment of different colon cancer cell lines to test its ability to inhibit
CBX8-dependent proliferation. Second, we sought to identify UNC7040-sensitive cell
lines to further characterize the impact of chemical treatment on Polycomb binding and
activity. At last, the sensitive cell lines could be used as models to investigate the
mechanisms of how CBX8 leads to tumor proliferation or metastasis.
20

3.2 Results
3.2.1 UNC7040 inhibits proliferation of colon cancer cells
In 96-well plates, Caco2, LoVo, HCT116 and HT-29 were treated with different
concentrations of either UNC7040 or negative control compound, UNC4219. Cells were
imaged with Operetta cell imager and counted every day from day 0 to day 3. Caco2,
HCT116 and DLD-1 showed no reduced proliferation at different concentrations of
UNC7040 (Figure 3.1B). In contrast, LoVo cells showed reduced cell count in a dose-
dependent manner (Figure 3.1A & B). Reduced cell count of LoVo cell did not reflect
reduced viability. LoVo cells showed only a mild increase in their percentage of dead cells
after UNC7040 treatment. Repeated LoVo proliferation assay using FACS showed about
40% proliferation reduction in UNC7040-treated group verses control group (Figure 3.1C).
The experiments indicated that among five tested cell lines, LoVo cells specifically
responded to UNC7040 treatment.

21

Figure 3.1
22

UNC7040 reduced LoVo proliferation. A. LoVo cells responded to UNC7040 treatment
in a dose-dependent manner as proliferation defect. Data contributed by Roy Lau
(Mumenthaler Lab, University of Southern California). B. Statistical analysis of
proliferation differences at day 3. **p<0.01, ***p<0.001, ****p<0.0001. C. Proliferation
assays investigating longer treatment time. LoVo cells were treated with 20uM UNC7040
for 4days and 7days. **p<0.01.

23

3.2.2 Colon cancer cell lines express CBX8 at similar level
To investigate the mechanism of differential UNC7040 sensitivity, we interrogated the
endogenous expression levels of CBX8 among five cell lines. Western blot results
indicated that Caco2, DLD-1, LoVo, HCT116 and HT-29 expressed CBX8 at similar level
(Figure 3.2B), which meant the discrepancy in their response to UNC7040 treatment was
not linked to differences in endogenous protein expression.

Since CBX8 expression levels were not associated with the specific response observed
in LoVo cells, we then were interested in identifying additional mutations that could create
a vulnerability to CBX8 antagonism. Since CBX8 contributes to the normal function of the
polycomb repressive complexes, we focused on the mutations that might disrupt PRC
regulation. Using online database, Catalogue Of Somatic Mutations In Cancer (COSMIC),
we found out that LoVo cells possess many somatic mutations related to PRCs functions.
ARID1A is one of the mutations we hypothesized to be relevant. As a core subunit of
SWI/SNF complexes, ARID1A supports the SWI/SNF normal function which serves as an
ATP-dependent chromatin remodeler (Kelso TWR et al., 2017). In contrast to PRCs,
SWI/SNF complexes induce accessible chromatin. The deletion in the ARID1A gene
shifts the open reading frames, which might abrogate SWI/SNF function. This might affect
the Polycomb-Trithorax balance favoring transcriptional repression by PRC1 at common
target genes. However, the endogenous expression of mutated ARID1A is unknown in
24

LoVo cells. We probed ARID1A expression in five colon cancer cell lines. Notably, LoVo
cells had lower expression of ARID1A compared with the other cell lines suggesting that
the mutation might affect protein stability and SWI/SNF complex (Figure 3.2A). We also
probed for PRC2 and PRC1-dependent histone modifications. H3K27me3 and
H2AK119ub were elevated in LoVo, HCT116 and HT-29 compared to Caco2 and DLD-1.
However, probing for total histone H3 revealed that the core histones were also elevated
in LoVo, HCT116 and HT-29. This argues that Polycomb-dependent modifications are not
increased but instead these cell lines have significant aneuploidy (Figure 3.2A).

25

Figure 3.2
Western Blot in colon cancer cell lines. A. Western Blot showing endogenous
expressions of ARID1A and H3K27me3/H2AK119ub levels in colon cancer cells. Histone
3 was probed for histone levels and beta-actin was used as loading control. Asterisks
indicate for unspecific bands. B. Endogenous CBX8 expressions in LoVo, DLD-1,
HCT116, HT29 and Caco2 detected by western blot. Beta-actin was used as loading
control.
26

3.3 Discussion
Our experiments identified a specific colon cancer cell line, LoVo cells, as a potential
human model for UNC7040 characterization. On one hand, we could further test the effect
of UNC7040 in human cells using LoVo; on the other hand, the phenotype alternation
upon UNC7040 treatment of LoVo allowed us to investigate the outcomes of the CBX8
chemical repression in a cancer cell line.
One pending question is that why LoVo but not the other cell lines tested responded to
the UNC7040 treatment? We hypothesized that the genetic background of LoVo cells led
to their susceptibility of UNC7040. The exploration on that identified one potential
candidate, ARID1A, and its different endogenous expression levels in cell lines tested
supported our hypothesis. However, more direct evidence is needed (e.g., restoration of
the wildtype of ARID1A allele) to test the causal connection between to the specific
response of LoVo cells to UNC7040.

Nevertheless, we confirmed that LoVo cells could serve as a model allowing us to
characterize UNC7040 in human cells while assessing the effect of CBX8 inhibition at the
same time.

27

Chapter 4: UNC7040 displaces CBX8-containing PRC1 in LoVo
cells
4.1 Introduction
As previously described, UNC7040 was designed to bind to the chromobox of CBX8 and
it would disrupt its interaction with H3K27me3. To link the proliferation defect to the effect
of UNC7040, the evidence of CBX8 displacement after UNC7040 treatment in LoVo cells
were required. Chromatin immunoprecipitation (ChIP) is an effective method to detect
proteins occupancy on certain DNA loci or whole genome. Usually, reagents, like
formaldehyde, are used to crosslink DNA/proteins to capture dynamic regulatory
interactions. After crosslinking, physical methods, such as sonication, can shred the large
chromatin into smaller pieces while the DNA bound by proteins/histones can be protected
from this procedure. Then, high affinity/specificity antibodies for target protein/histone
modifications will be added to the solution for enrichments of target proteins/DNA
complexes. After extraction of antibody and isolation of enriched DNA, DNA pieces are
purified as the final products. Depending on the goals of the different projects, ChIP can
provide different information. Using next generation sequencing (NGS) methods,
researchers can find out what regions are bound by target proteins or enriched for certain
histone modifications. Meanwhile, by comparing groups with different treatments, the
28

frequency of detecting the same regions can serve as the parameter for the protein
occupancy among specific loci or whole genome. One caveat is that when comparing two
treatments with NGS methods, the differences in reads between two conditions could be
generated due to the technical issues, such overamplification of sequencing variations.
Therefore, to count in the errors brought by sequencing itself, chromatin spike-in is a new
modification to the traditional ChIP experiments (Grzybowski et al., 2015). By adding
chromatin from a different species (e.g., mouse chromatin spiked in human samples) to
both the control and experimental samples, researchers can normalize the reads in
different conditions based on the reads detected for the spike-in chromatin. Here, we used
spike in of mouse chromatin to quantify changes in human PRC1, PRC2 and histone
modifications in response to treatment of LoVo cells with UNC7040 or control compound.

4.2 Results
4.2.1 Chromatin immunoprecipitation (ChIP) in LoVo cells
LoVo cells were crosslinked by formaldehyde, and the chromatin were prepared following
standard ChIP protocol (see Methods and Materials). Antibodies specific for CBX8 and
the catalytic subunit RING1B/RNF2 were used to investigate changes in PRC1
29

occupancy. In addition, we performed ChIP for SUZ12 and H3K27me3 to evaluate the
impact on PRC2 binding and function. To determine relative enrichment of PRC1 and
PRC2 in LoVo cells, we employed quantative polymerase chain reaction (qPCR) on
purified DNA. Since the CBX8-binding profile in LoVo cells was not reported before, we
first assessed CCND2 and GATA6 locus using primers from previous report (Pemberton
H et al., 2014). LMNB2 was used as negative control of CBX8 binding (Figure 4.1C). To
find out more CBX8 positive binding sites in colon tissue, we used UCSC genome browser
to identify loci enriched in H3K27me3 signals, which are potential CBX8 targets, and
designed qPCR primers for them. We further investigated the occupancy of CBX8 among
BMP5 and FGF23 in LoVo cells. qPCR demonstrated that BMP5 was another locus bound
by CBX8 in LoVo cells.

4.2.2 UNC7040 displaced CBX8 from CCND2 and BM P 5
LoVo cells were treated with 20uM either UNC7040 or UNC7263 for 6 hours before their
chromatin were extracted for ChIP experiments. Antibodies for CBX8, RING1B/RNF2,
SUZ12 and H3K27me3 were used for immunoprecipitation. Before adding antibodies to
the chromatin solutions, mouse chromatin was spiked in as 1%. qPCR analysis of the
purified DNA showed that CBX8 and RING1B had reduced occupancy on CCND2 and
BMP5 locus in the UNC7040-treated groups, while signals for H3K27me3 remained
30

similar (Figure 4.1A & B). Hence, we confirmed that UNC7040 could displace CBX8 and
its catalytic partner, RING1B/RNF2, from CCND2/BMP5 where they originally bound to.

31

32

Figure 4.1
UNC7040 displaced CBX8 in LoVo cells. A. ChIP-qPCR analysis for UNC7040-treated
LoVo cells on CCND2 locus. For each compound, two replicates were adapted. B. ChIP-
qPCR analysis for UNC7040-treated LoVo cells on BMP5 locus. Two replicates for each
compound. C. ChIP-qPCR analysis for UNC7040-treated LoVo cells on LMNB locus. Two
replicates for each compound.

33

4.2.3 UNC7040 reduced CBX8 occupancy globally in LoVo cells
We then prepared libraries from the same samples for next generation sequencing to
evaluate the global CBX8/RING1B/H3K27me3 binding profiles. ChIP-seq analysis
demonstrated that UNC7040 treatment decreased the binding of CBX8 and RING1B
among the regions where H3K27me3 were enriched (Figure 4.2A). However, the signals
of H3K27me3 between UNC7263-treated groups and UNC7040-treated groups were
unchanged (Figure 4.2A), which indicated that the reduction of CBX8 binding was not due
to the reduction of H3K27me3 level. Several gene loci showed CBX8/RNF2
displacements after UNC7040 treatment (Figure 4.2B). From the ChIP-seq results, we
are confident to conclude that UNC7040 can displace CBX8 and canonical PRC1 from
their binding sites in LoVo cells.

34

Figure 4.2
UNC7040 displaced CBX8 globally in LoVo cells. A. ChIP-seq analysis for UNC7040-
treated LoVo cells. Signals were normalized by spiked-in mouse chromatin (see Materials
and Methods). Two replicates were conducted for each compound. Each line in heatmaps
indicates for a locus enriched for H3K27me3. Data analyzed by Daniel Bsteh (Bell Lab,
University of Southern California). B. Example loci showing displacements of CBX8/RNF2
in LoVo cells. Black lines indicate enrichment levels of CBX8, RING1B and H3K27me3 in
35

UNC7263-treated LoVo cells. Solid colors show enrichment levels in UNC7040-treated
LoVo cells.

4.3 Discussion
The anticipated effect of UNC7040 is antagonizing CBX8 by preventing it from
recognizing and binding to the PRC2-mediated H3K27me3 modification. By analyzing
LoVo cells treated with UNC7040 and control compound, UNC7263, we demonstrated
that UNC7040 did displace CBX8 and canonical PRC1 globally in human cells. The
signals of CBX8 were interrogated among the regions where H3K27me3 were peaked,
because the function of CBX8 in canonical PRC1 is basing on the upstream PRC2
catalyzed H3K27me3. Still, it would be interesting to ask that whether CBX8 was also
reduced in those regions independent of H3K27me3, even if it was not anticipated.
However, due to the noisy background of IP-CBX8 sequencing results, it was hard to call
peaks with CBX8 alone. Nevertheless, canonical PRC1-related CBX8 binding was
confirmed reduced after UNC7040 treatment. In addition, the relationship between
displacement of CBX8 and LoVo cell proliferation defect is still unclear. Although it was
reported that CBX8 knockdown would reduce the in vivo growth of LoVo cells, the role of
the canonical PRC1 was not investigated in LoVo cells. Here, we not only demonstrated
the displacement of CBX8 upon UNC7040 treatment, but also showed disrupted
36

canonical PRC1 interactions with their original targets in LoVo cells. It raised the
possibility that for LoVo cells, dampening CBX8 in PRCs pathway only is enough to trigger
phenotype changes because of its inner vulnerability to PRCs disruption. It can also
explain why the other colon cancer cells are not sensitive to the UNC7040 treatment:
CBX8 contributes to the proliferations of different cancer cells in different major pathways.

37

Chapter 5: Summary and Future Steps
5.1 Summary
Taking advantage of our reporter system, we and our collaborators successfully
generated and evaluated a new CBX8 antagonist, UNC7040. We identified that a human
colon cancer cell line, LoVo, was sensitive to the treatment of UNC7040 in a dose-
dependent manner. In addition, LoVo cells not only could be a model to test and
characterize UNC7040 but also provided us an example for investigating the role of CBX8
in oncogenesis.

Proliferation assays demonstrated that UNC7040 could inhibit the proliferation of LoVo
cells. Chromatin immunoprecipitation assays and following NGS analysis showed that
UNC7040 could displace CBX8 and its partner in canonical PRC1, which provided
evidence for the anticipated effect of UNC7040.

38

5.2 Future Steps
We hypothesized that the specific response of LoVo cells to the UNC7040 treatment was
the consequence of the susceptibility of LoVo to the PRCs dysfunction. ARID1A was one
of the potential candidates, because it contributes to the normal activation of gene and it
is mutated in LoVo cells. However, direct evidence is required to address it, such as
knockdown/knockout of ARID1A in other colon cancer cell lines or rescuing it in LoVo cells.
Also, to further understand the consequences of UNC7040 treatment in LoVo cells, RNA-
sequencing for treated cells is necessary to examine the transcriptome changes. It was
revealed that CBX8-knockdown was related with lower proliferation of colon cancer cells,
but whether it is PRC-dependent is unknown. From LoVo cells, we generated cell lines
for doxycycline-inducible shRNA knockdown of CBX8/RNF2/SUZ12 to investigate the role
of canonical PRC1 and PRC2 in the proliferation of LoVo cells.

Since CBX8 was reported important in other human cancers, such as breast cancer and
leukemia, UNC7040 can also be tested in these cell lines. Interestingly, CBX8 was
involved in the oncogenesis in a PRC-independent way according to reports. Considering
that UNC7040 was designed to interrupt the interaction between CBX8 and chromatin, it
can potentially dissect the canonical function (chromatin binding) and non-canonical
function of CBX8. Therefore, by treating CBX8-sensitive cancer cell lines with UNC7040,
39

we can distinguish them by how CBX8 affect the oncogenesis, which can provide more
insights to the functions of CBX8.

40

Materials and Methods
Cell Culturing
CP15 and CP15-TetR-CBX8 were maintained in DMEM (Corning) supplemented with 15%
fetal bovine serum (FBS), 1% penicillin streptomycin (PS, Sigma Millipore), 1% HEPES
(Thermo Scientific), 1% 100mM Sodium-Pyruvate (Sigma Millipore), 1% 200mM
GlutaMax (Thermo Scientific), 1x MEM non-essential amino acid (Thermo Scientific), 1%
50mM beta-mercaptoethanol (Thermo Scientific) and 2uL LIF (IMBA core facility) / 500mL
DMEM. Cells were split every 2 days as 1: 10.

LoVo cells were maintained in F-12K (Corning) supplemented with 10% FBS and 1% PS.
Cells were split when 80-90% confluent.

For treatments, CP15-TetR-CBX8 was treated with 2ug/mL doxycycline for 4 days to
remove TetR-CBX8 from reporter locus. CP15-TetR-CBX8 was treated with 0.1uM, 1uM
and 10uM UNC7040 (Generous gift from Stephen Frye’s lab at University of North
Carolina. Same as UNC4219 and UNC7263) to evaluate the effect of UNC7040. LoVo
cells were treated with 0uM, 0.62uM, 1.85uM, 5.56uM, 16.7uM and 50uM UNC7040 or
41

UNC4219 at the first-round proliferation assay. LoVo cells were treated with 20uM
UNC7040 or UNC7263 at the 7-days proliferation assay.

All cells were cultivated at 37 ℃ and 5% CO2.

Flow Cytometry
CP15, CP15-TetR-CBX8 and LoVo cells were collected with centrifuge at 200g before
sending for flow cytometry analysis. Cells were washed twice in PBS (VWR) and
resuspended in corresponding culturing media. Cells were analyzed with Attune NxT Flow
Cytometer (Thermo Scientific) using Attune NxT Software. For fluorescence detection,
EGFP was excited with a 488nm blue laser and emission was detected at 530/30
wavelength. Data was exported as .fcs files and further analyzed by FlowJo to generate
visual plots.

Proliferation Assays
For the proliferation assay, ~2000 LoVo cells treated with different conditions (see Cell
Culturing section) were monitored and counted everyday for 3 days with Operetta cell
imager (PerkinElmer). 3 replicates had been prepared for each compound each
42

concentration. For the 7-days experiments, ~3000 LoVo cells were seeded and treated
with 20uM UNC7040 or UNC7263 for 7days. Total cells were counted by Attune NxT Flow
Cytometer at day 7. 3 replicates had been prepared for each condition.

Western Blot
~10 million Caco2, LoVo, DLD-1, HCT116 and HT-29 cells were collected for nuclear
proteins extraction. Cells were first washed with 10mL PBS and then resuspended in 7mL
Buffer A (25 mM HEPES pH7.6, 5 mM MgCl, 25 mM KCl, 0.05 mM EDTA, 10 % Glycerol,
0.1 % NP40), DTT 1:1000 (100 mM) and PMSF 1x (100x)). Cells were incubated on ice
for 10 min. After pelleting cells at 1500 rpm for 5 min they were resuspended in RIPA
buffer (150 mM NaCl, 5 mM
EDTA pH8.0, 50 mM Tris pH8.0, 1 % NP40, 0.5 % Sodium Deoxycholate, 0.1 % SDS,
Proteinase Inhibitor 1x). DNA was sheared and removed via sonication (Bioruptor Pico,
Diagenode) and centrifugation and then protein concentration was measured with a
Bradford Assay. 15ug protein lysate was mixed with 4x Laemmli Buffer (BioRad)
supplemented with fresh 10 mM DTT and 0.5 % 2-mercaptoethanol (VWR) and boiled for
5 min at 95 °C. Samples were resolved on a NuPAGE 4% -12% gradient Bis-Tris gel
(Thermo Fisher Scientific). After a run of 1 hour at 120 Volt the gel was transferred onto
43

a Polyvinylidene difluoride (PVDF) membrane (Sigma-Aldrich) in Transfer Buffer (100 ml
10 x transferr buffer, 700 ml dH2O and 200ml Methanol (100%) for 1 h at 100 Volt. The
membrane was then blocked in 5% milk in PBST (1x PBS supplemented with 0.1 %
Tween 20 (Sigma Aldrich) for 1 hour at room temperature. Then the primary antibody (see
Supplementary Materials Table 2) was added in 5% milk solution, overnight incubation at
4°C. The next day the membrane was washed three times 5 min with PBST before adding
the secondary antibody (1:10,000 in 5% milk Goat anti-Rabbit IgG (H+L) Secondary
Antibody, DyLight 800, Thermo Fisher Scientific) for 1 hour at room temperature and
another three wash steps with PBST. Then the blot was scanned by ODYSSEY CLx
system (LI-COR).

Chromatin Immunoprecipitation
Prior to cell collection, 11% formaldehyde solution was prepared fresh from 16%
formaldehyde (VWR). ~20 million LoVo cells were collected and washed twice and
resuspended with 10mL PBS. 1mL 11% formaldehyde was added into cell suspension,
and cells were incubated for 7 min at room temperature for crosslinking. Then, 0.5mL
2.5M glycine (VWR) was added, and cells were incubated at RT for 5 min to stop
crosslinking. Nuclei were prepared by washes with NP-Rinse buffer 1 (10 mM Tris pH 8.0,
10 mM EDTA pH 8.0, 0.5mMEGTA, 0.25% Triton X-100) followed by NP-Rinse buffer 2
44

(10mMTris pH 8.0, 1mM EDTA, 0.5mM EGTA, 200mM NaCl). Afterwards the cells were
prepared for shearing by sonication by two washes with Covaris shearing buffer (1mM
EDTA pH 8.0, 10mMTris-HCl pH 8.0, 0.1% SDS) and resuspension of the nuclei in 0.9
mL Covaris shearing buffer (with protease inhibitors mini (VWR)). The nuclei were
sonicated with Biorupter for 7 cycles (30 seconds on/30 seconds off). Prior to the
immunoprecipitation, 1% of mouse chromatin was added into each sample for the internal
control. Lysates were incubated in 1x IP buffer (final: 50 mM HEPES/KOH pH 7.5, 300
mM NaCl, 1mM EDTA, 1% Triton X-100, 0.1% DOC, 0.1% SDS), with 2uL antibodies (see
Supplementary Materials Table 3) at 4°C on a rotating wheel. The next day, samples were
incubated with Protein G magnetic Dynal beads (Thermo) for 3 hours at 4°C. Collected
Beads were washed 5 times with 1x IP buffer (50 mM HEPES/KOH pH 7.5, 300mM NaCl,
1mM EDTA,1%Triton-X100, 0.1% DOC, 0.1% SDS), or 1.66x IP buffer for H3K27me3,
followed by three times with DOC buffer (10mM Tris pH 8, 0.25mM LiCl, 1mM EDTA, 0.5%
NP40, 0.5% DOC) and 1x with TE (with 50mM NaCl).

qPCR Analysis
The PCIA extracted IP DNA was precipitated and quantified using a EvaGreen based
qPCR mix on a CFX Connect Real-Time PCR Detection System (BioRad). qPCR primers
are listed in Supplementary Materials Table 1.
45

Supplementary Materials
Table 1
Name Primer sequence
LMNB2-forward CCGAATCTCTGAAATGAAAGTCCATGC
LMNB2-reverse TTAAAGATCTGAGGGACTCCTCAGTC
CCND2-forward ACTGTCTGAAATGAAGGTGAAGC
CCND2-reverse GATTTGATGGACACTTGGTTTGT
BMP5-forward-1 TGCACCTTGGTGTTGGAATA
BMP-reverse-1 TGCTGGAGGTGGAATTAACA
BMP5-forward-2 GTGTAAGCACAACCCTGCTG
BMP5-reverse-2 TGTCCCATGCATCACTGTTT

Table 2
Protein/Modification
Name
Antibody Source Concentration
Used
CBX8 Rabbit anti-CBX8 Bethyl
Laboratories
1:5000
ARID1A ARID1A/BAF250A
(D2A8U) Rabbit
mAb
Cell Signaling
Tech.
1:1000
Histone 3 Anti-Histone H3
antibody
Abcam 1:2000
Beta-actin Anti-Actin antibody Sigma-Aldrich 1:2000
H3K27me3 Histone H3K27me3
antibody (pAb)
Active Motif 1:2000
H2AK119ub Anti-ubiquityl-
Histone H2A
Antibody
Merck Millipore 1:2000

46

Table 3
Protein/Modification
Name
Antibody Source
CBX8 Rabbit anti-CBX8 Bethyl
Laboratories
RNF2 RING1B (D22F2)
XP Rabbit mAb
Cell Signaling
Tech.
SUZ12 SUZ12 (D39F6)
XP Rabbit mAb
Cell signaling
Tech.
H3K27me3 Histone
H3K27me3
antibody (pAb)
Active Motif

47

References
Bauer M, Trupke J, Ringrose L. The quest for mammalian Polycomb response ele
ments: are we there yet? Chromosoma. 2016;125(3):471-496. doi:10.1007/s00412-0
15-0539-4

Blackledge NP, Farcas AM, Kondo T, et al. Variant PRC1 complex dependent H2A
ubiquitylation drives PRC2 recruitment and polycomb domain formation. Cell. 201
4;157(6):1445-1459. doi: 10.1016/j.cell.2014.05.004

Chang C-J, Yang J-Y, Xia W, et al. EZH2 Promotes Expansion of Breast Tumor I
nitiating Cells through Activation of RAF1-β-Catenin Signaling. Cancer Cell. 2011;1
9(1):86-100. doi: 10.1016/j.ccr.2010.10.035

Chi P, Allis CD, Wang GG. Covalent histone modifications — miswritten, misinterpr
eted and mis-erased in human cancers. Nature Reviews Cancer. 2010;10(7):457-4
69. doi:10.1038/nrc2876

Chittock EC, Latwiel S, Miller TCR, Müller CW. Molecular architecture of polycomb
repressive complexes. Biochemical Society transactions. 2017;45(1):193-205. doi:1
0.1042/BST20160173

Choy M-K, Movassagh M, Goh H-G, Bennett MR, Down TA, Foo RSY. Genome-w
ide conserved consensus transcription factor binding motifs are hyper-methylated.
BMC genomics. 2010;11(1):519-519. doi:10.1186/1471-2164-11-519

Christian Beisel, Andreas Buness, Ian M. Roustan-Espinosa, et al. Comparing Acti
ve and Repressed Expression States of Genes Controlled by the Polycomb/Trithor
ax Group Proteins. Proceedings of the National Academy of Sciences of the Unite
d States of America. 2007;104(42):16615-16620. doi:10.1073/pnas.0701538104

Chung C-Y, Sun Z, Mullokandov G, et al. Cbx8 Acts Non-canonically with Wdr5 to
Promote Mammary Tumorigenesis. Cell Reports. 2016;16(2):472-486. doi: 10.1016
/j.celrep.2016.06.002

Clapier CR, Cairns BR. The biology of chromatin remodeling complexes. Annual r
eview of biochemistry. 2009;78(1):273-304. doi:10.1146/annurev.biochem.77.062706.
153223
48

Cohen ASA, Tuysuz B, Shen Y, Bhalla SK, Jones SJM, Gibson WT. A novel muta
tion in EED associated with overgrowth. Journal of human genetics. 2015;60(6):33
9-342. doi:10.1038/jhg.2015.26

Connelly KE, Weaver TM, Alpsoy A, Gu BX, Musselman CA, Dykhuizen EC. Enga
gement of DNA and H3K27me3 by the CBX8 chromodomain drives chromatin ass
ociation. Nucleic Acids Research. 2019;47(5):2289-2305. doi:10.1093/nar/gky1290

Czermin B, Melfi R, McCabe D, Seitz V, Imhof A, Pirrotta V. Drosophila Enhancer
of Zeste/ESC Complexes Have a Histone H3 Methyltransferase Activity that Marks
Chromosomal Polycomb Sites. Cell. 2002;111(2):185-196. doi:10.1016/S0092-8674
(02)00975-3

Entrevan M, Schuettengruber B, Cavalli G. Regulation of Genome Architecture and
Function by Polycomb Proteins. Trends in Cell Biology. 2016;26(7):511-525. doi:1
0.1016/j.tcb.2016.04.009

Gao Z, Zhang J, Bonasio R, et al. PCGF Homologs, CBX Proteins, and RYBP D
efine Functionally Distinct PRC1 Family Complexes. Molecular Cell. 2012;45(3):344
-356. doi:10.1016/j.molcel.2012.01.002

Geisler SJ, Paro R. Trithorax and Polycomb group-dependent regulation: a tale of
opposing activities. Development (Cambridge, England). 2015;142(17):2876-2887. d
oi:10.1242/dev.120030

Gross DS, Chowdhary S, Anandhakumar J, Kainth AS. Chromatin. Current Biology.
2015;25(24):R1158-R1163. doi:10.1016/j.cub.2015.10.059

Grzybowski AT, Chen Z, Ruthenburg AJ. Calibrating ChIP-Seq with Nucleosomal
Internal Standards to Measure Histone Modification Density Genome Wide. Molecu
lar cell. 2015;58(5):886-899. doi: 10.1016/j.molcel.2015.04.022

Hauri S, Comoglio F, Seimiya M, et al. A High-Density Map for Navigating the Hu
man Polycomb Complexome. Cell Reports. 2016;17(2):583-595. doi: 10.1016/j.celre
p.2016.08.096

49

Kaustov L, Ouyang H, Amaya M, et al. Recognition and specificity determinants of
the human cbx chromodomains. The Journal of biological chemistry. 2011;286(1):
521-529. doi:10.1074/jbc.M110.191411

Kelso TWR, Porter DK, Amaral ML, Shokhirev MN, Benner C, Hargreaves DC. Ch
romatin accessibility underlies synthetic lethality of SWI/SNF subunits in ARID1A-m
utant cancers. eLife. 2017;6. doi:10.7554/eLife.30506

Kexin Xu, Zhenhua Jeremy Wu, Anna C. Groner, et al. EZH2 Oncogenic Activity i
n Castration-Resistant Prostate Cancer Cells Is Polycomb-Independent. Science. 2
012;338(6113):1465-1469. doi:10.1126/science.1227604

Kolybaba A, Classen A-K. Sensing cellular states—signaling to chromatin pathway
s targeting Polycomb and Trithorax group function. Cell and Tissue Research. 201
4;356(3):477-493. doi:10.1007/s00441-014-1824-x

Längst G, Manelyte L. Chromatin Remodelers: From Function to Dysfunction. Gen
es. 2015;6(2):299-324. doi:10.3390/genes6020299

Laugesen A, Helin K. Chromatin Repressive Complexes in Stem Cells, Developme
nt, and Cancer. Cell Stem Cell. 2014;14(6):735-751. doi: 10.1016/j.stem.2014.05.0
06

Laugesen A, Højfeldt JW, Helin K. Molecular Mechanisms Directing PRC2 Recruit
ment and H3K27 Methylation. Molecular Cell. 2019;74(1):8-18. doi: 10.1016/j.molce
l.2019.03.011

Lee ST, Li Z, Wu Z, et al. Context-Specific Regulation of NF-κB Target Gene Exp
ression by EZH2 in Breast Cancers. Molecular Cell. 2011;43(5):798-810. doi: 10.10
16/j.molcel.2011.08.011

Margueron R, Li G, Sarma K, et al. Ezh1 and Ezh2 Maintain Repressive Chromat
in through Different Mechanisms. Molecular Cell. 2008;32(4):503-518. doi: 10.1016/
j.molcel.2008.11.004

Margueron R, Reinberg D. The Polycomb complex PRC2 and its mark in life. Nat
ure. 2011;469(7330):343-349. doi:10.1038/nature09784
50

Morey L, Pascual G, Cozzuto L, et al. Nonoverlapping Functions of the Polycomb
Group Cbx Family of Proteins in Embryonic Stem Cells. Cell Stem Cell. 2012;10
(1):47-62. doi:10.1016/j.stem.2011.12.006

Moussa HF, Bsteh D, Yelagandula R, et al. Canonical PRC1 controls sequence-in
dependent propagation of Polycomb-mediated gene silencing. Nature communicatio
ns. 2019;10(1):1931-12. doi:10.1038/s41467-019-09628-6

Nakao M, Matsui S, Yamamoto S, Okumura K, Shirakawa M, Fujita N. Regulation
of transcription and chromatin by methyl-CpG binding protein MBD1. Brain and De
velopment. 2001;23:S174-S176. doi:10.1016/S0387-7604(01)00348-5

Nicole J. Francis, Robert E. Kingston, Christopher L. Woodcock.
Chromatin Compaction by a Polycomb Group Protein Complex. Science. 2004;306
(5701):1574-1577. doi:10.1126/science.1100576

Papp B. Histone trimethylation and the maintenance of transcriptional ONand OFF
states by trxG and PcG proteins. Genes & Development. 2006;20(15):2041-2054.
doi:10.1101/gad.388706

Pemberton H, Anderton E, Patel H, et al. Genome-wide co-localization of Polycom
b orthologs and their effects on gene expression in human fibroblasts. Genome bi
ology. 2014;15(2):R23-R23. doi:10.1186/gb-2014-15-2-r23

Pirrotta V. Chromatin-silencing mechanisms in Drosophila maintain patterns of gen
e expression. Trends in Genetics. 1997;13(8):314-318. doi:10.1016/S0168-9525(97)
01178-5

Ru Cao, Liangjun Wang, Hengbin Wang, et al. Role of Histone H3 Lysine 27 Met
hylation in Polycomb-Group Silencing. Science. 2002;298(5595):1039-1043. doi:10.1
126/science.1076997

Santos M, Martínez-Fernández M, Dueñas M, et al. In vivo disruption of an Rb-E
2F-Ezh2 signaling loop causes bladder cancer. Cancer research. 2014;74(22):6565-
6577. doi: 10.1158/0008-5472.CAN-14-1218

51

Shen X, Liu Y, Hsu Y-J, et al. EZH1 Mediates Methylation on Histone H3 Lysine
27 and Complements EZH2 in Maintaining Stem Cell Identity and Executing Plurip
otency. Molecular Cell. 2008;32(4):491-502. doi: 10.1016/j.molcel.2008.10.016

Spahn L, Kubicek S, Wutz A, et al. Recruitment of PRC1 function at the initiation
of X inactivation independent of PRC2 and silencing. The EMBO Journal. 2006;25
(13):3110-3122. doi:10.1038/sj.emboj.7601187

Tan J, Jones M, Koseki H, et al. CBX8, a Polycomb Group Protein, Is Essential f
or MLL-AF9-Induced Leukemogenesis. Cancer Cell. 2011;20(5):563-575. doi: 10.10
16/j.ccr.2011.09.008

Tavares L, Dimitrova E, Oxley D, et al. RYBP-PRC1 Complexes Mediate H2A Ubi
quitylation at Polycomb Target Sites Independently of PRC2 and H3K27me3. Cell.
2012;148(4):664-678. doi:10.1016/j.cell.2011.12.029

Zhang G, Pradhan S. Mammalian epigenetic mechanisms. IUBMB Life. 2014;66(4):
240-256. doi:10.1002/iub.1264

Zhang T, Cooper S, Brockdorff N. The interplay of histone modifi-cations – writers
that read. EMBO reports. 2015;16(11):1467-1481. doi:10.15252/embr.201540945

Zhang Y, Kang M, Zhang B, et al. m6A modification-mediated CBX8 induction reg
ulates stemness and chemosensitivity of colon cancer via upregulation of LGR5.
Molecular Cancer. 2019;18(1):1-16. doi:10.1186/s12943-019-1116-x

Zingg D, Debbache J, Schaefer SM, et al. The epigenetic modifier EZH2 controls
melanoma growth and metastasis through silencing of distinct tumour suppressors.
Nature communications. 2015;6(1):6051-. doi:10.1038/ncomms7051

Abstract (if available)

Abstract The synergy between polycomb repressive complexes (PRCs), PRC1 and PRC2, is essential for gene expression silencing during the early development. PRC2 is recruited in a sequence-specific manner to target genes where it catalyzes H3K27me3. In turn, PRC1 can recognize H3K27me3 and catalyzes H2AK119ub1. Polycomb modifications enforce chromatin compaction and genes nearby are repressed (Laugesen et al., 2019). In mammals, PRC1 can be classified into canonical and variant complexes (cPRC1 and vPRC1, respectively) based on the presence of Chromobox-domain (CBX) proteins. CBX proteins are specific for cPRC1 and confer the capacity to recognize H3K27me3 and direct chromatin binding of PRC1 at PRC2 target genes. CBXs are differentially expressed and form mutually exclusive cPRC1 (Kaustov et al., 2011). However, even when co-expressed in the same tissue, incorporation of individual CBX proteins endows cPRC1 complexes with different regulation and functions (Morey et al., 2012). Recently, overexpression of CBX8 was associated with different cancer types, including colon cancer (Zhang et al., 2019) and breast cancer (Chung et al., 2016). Therefore, effective and specific CBX8 antagonists might benefit basic research in elucidating the functions of cPRC1 in development and potential translational applications in the clinic. In collaboration with the laboratory of Stephen Frye at UNC we have leveraged a Polycomb reporter assay to screen and identify novel CBX8-specific small molecule antagonists. Here, we characterized the CBX8 antagonist, UNC7040. In mouse embryonic stem cells, we show that this novel peptidomimetic compound is highly efficacious in inhibiting CBX8-dependent transcriptional reporter gene silencing. Furthermore, using a model of CBX8-dependent colon cancer, we demonstrate efficient displacement of CBX8-containing cPRC1 from genome which results in impaired proliferation. These results demonstrate the UNC7040 offers a powerful chemical tool to investigate the molecular mechanism of CBX8 function and explore inhibition of cPRC1 as a potential therapeutic option in colorectal cancer.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Evaluating the therapeutic potential of targeting CBX8 in MLLr leukemia

PDF

Investigating the effects of polycomb repressive complex inhibitors on cancer cell phenotypic plasticity

PDF

Using epigenetic toggle switches to repress tumor-promoting gene expression

PDF

A novel screening approach to identify regulators of polycomb-dependent gene silencing

PDF

Using CRISPR-mediated deletion to study prostate cancer regulatory elements located at loop anchors identified by Hi-C

PDF

Identification of target genes and protein partners of ZNF711 in glioblastoma cells

PDF

Functional role of chromatin remodeler proteins in cancer biology

PDF

Functional characterization of colon cancer risk enhancers

PDF

Functional characterization of a prostate cancer risk region

PDF

Functional characterization of colon cancer risk-associated enhancers: connecting risk loci to risk genes

PDF

Characterization of novel polycomb protein Mbtd1 in hematopoietic system

PDF

The role of microRNAs in cancer

PDF

The role of endoplasmic reticulum chaperone protein GRP78 in breast cancer

PDF

The role of histone H4 lysine 20 monomethylation in gene expression and differentiation

PDF

Understanding the role of APP and DYRK1A in human brain pericytes

PDF

Study of the role of bone morphogenetic proteins in prostate cancer progression

PDF

Defining the functional region of LINC00261 in lung adenocarcinoma

PDF

Differential role of two coactivators, CCAR1 and CARM1, for dysregulated beta-catenin activity in colorectal cancer cell growth and gene expression

PDF

Mechanism of a new CK2 inhibitor triggering senescence in breast cancer cells

PDF

Estrogen receptor-β characterization in breast cancer: development of a reliable assay for measuring expression

Asset Metadata

Creator Si, Yibo (author)

Core Title Characterization of a new chromobox protein 8 (CBX8) antagonist in a model of human colon cancer

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Publication Date 07/28/2022

Defense Date 06/09/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag CBX8,colon cancers,epigenetics,OAI-PMH Harvest,polycomb repressive complexes

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bell, Oliver (committee chair), Farnham, Peggy (committee member), Mumenthaler, Shannon (committee member), Stallcup, Michael (committee member)

Creator Email yibosi@email.unc.edu,yibosi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-348135

Unique identifier UC11663398

Identifier etd-SiYibo-8825.pdf (filename),usctheses-c89-348135 (legacy record id)

Legacy Identifier etd-SiYibo-8825.pdf

Dmrecord 348135

Document Type Thesis

Rights Si, Yibo

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA