Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Functional characterization of colon cancer risk enhancers
(USC Thesis Other)
Functional characterization of colon cancer risk enhancers
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Functional Characterization of Colon Cancer Risk Enhancers
By
Yuli Hung
A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR BIOLOGY)
August 2014
i
ACKNOWLEDGMENTS
I must begin by expressing my utmost gratitude towards Dr. Peggy Farnham, my wonderful mentor
and committee chair. Your intelligence and creativity has guided me to success that I could never
have accomplished on my own. Thank you for constantly inspiring and motivating me to reach a
little further. In the past two years, you have taught me valuable skills that I will appreciate for life.
To my committee members Dr. Gerry Coetzee and Dr. Zoltan Tokes, thank you for giving me the
opportunity to share my passion on the subject with you.
I would also like to thank all my colleagues in the Farnham Lab - Adam, Albert, Heather, Jordan,
Lijing, Malaina, Matt, and Phoebe. Thank you for creating a fun and motivating environment that
makes me look forward to being in the laboratory. Esther, you have been patient and supportive
from day one until today. Without you, finishing this project would be impossible.
To my friends and the USC badminton team - thank you, for being my family away from home.
To my family - Dad, Mom, and Andrew - my gratitude simply cannot be described by words. You
are and always will be the best.
ii
TABLE OF CONTENTS
ACKNOWLEDGMENTS ....................................................................................................................i
LIST OF FIGURES AND TABLES .................................................................................................. iii
ABBREVIATIONS .............................................................................................................................iv
ABSTRACT ........................................................................................................................................vi
CHAPTER 1 ........................................................................................................................................ 1
CHAPTER 2 ........................................................................................................................................ 8
2.1 Guide RNA and Cas9 Preparation- ................................................................................................ 8
2.2 Cell Culture and Transfection ...................................................................................................... 12
2.3 Cell Sorting and Colony Selection ............................................................................................... 14
2.4 Genomic DNA and RNA isolation ............................................................................................... 16
2.5 Methylated DNA Immunoprecipitation (MeDIP) ........................................................................ 17
2.6 Histone H3K27Ac Chromatin Immunoprecipitation ................................................................... 19
2.7 Experimental controls .................................................................................................................. 22
2.8 RNA Sequencing (RNA-Seq) ...................................................................................................... 22
2.9 Primer design and preparation ..................................................................................................... 23
CHAPTER 3 ...................................................................................................................................... 24
3.1 Specific Aim 1.............................................................................................................................. 24
3.2 Specific Aim 2.............................................................................................................................. 30
3.3 Specific Aim 3.............................................................................................................................. 37
CHAPTER IV .................................................................................................................................... 43
CHAPTER V ..................................................................................................................................... 51
CHAPTER VI .................................................................................................................................... 52
REFERENCES .................................................................................................................................. 54
iii
LIST OF FIGURES AND TABLES
Figure 1. UCSC genome browser snapshot of the 400 kb window around index SNPs rs7136702
and rs11169552
Figure 2. UCSC genome browser snapshot of the risk enhancer cluster in DIP2B and E16
Figure 3. Enhancer deletion using the CRISPR technology
Figure 4. Workflow for isolating deletion clones
Figure 5. Overlap of down-regulated genes from inner deletion and outer deletion clones
Figure 6. Down-regulated genes following enhancer deletion
Figure 7. Fold change vs distance in megabase of all down-regulated genes on chromosome 12, in
relation to E16
Figure 8. Down-regulated genes exclusively in outer CRISPR deletion set
Figure 9. Enhancer 15, the target for forced DNA methylation by dCas9-DNMT3A
Figure 10. Preliminary results from MeDIP for methylating enhancer15 using dCas9-DNMT3A
Figure 11. Enhancer 11, the target for forced H3K27 acetylation by dCas9-CBP
Figure 12. Transfection efficiency of HCT116 cells
Table 1. Guide RNA plasmid clones
Table 2. Cas9 plasmid clone number, name, and description
Table 3. Name and description of primers
Table 4. Name and description of viable cell stocks frozen in liquid nitrogen
Table 5. Name and description of all primers used for MeDIP experiments
Table 6. Name and description of all primers used for ChIP experiments
Table 7. SNP ID, chromosomal location, and index SNP information for the three risk SNPs
identified in E16
Table 8. Chromosomal location, gene name, and average fold change of the 20 most
down-regulated genes
Table 9. Number of genes down-regulated by more than 2 fold or more than 3 fold
Table 10. Number of genes down-regulated by more than 2 fold or more than 3 fold for genes
exclusive to outer CRISPR deletion set
Table 11. Chromosomal location, gene name, and average fold change of the 20 most
down-regulated genes exclusive to outer CRISPR deletion set
Table 12. SNP ID, chromosomal location, and index SNP information for the risk SNPs identified
in enhancer 11 and 15
iv
ABBREVIATIONS
CBP: CREB-binding protein
ChIP: Chromatin immunoprecipitation
CpG: -C-phosphate-G-
CRC: colorectal cancer
CRISPR: Clustered Regularly Interspaced Short Palindromic Repeats
DIP2B: disco-interacting protein 2 homolog B
DMAP: DNA methyltransferase-associated protein
DNMT3A: DNA (cytosine-5)-methyltransferase 3 alpha
FACS: fluorescence-activated cell sorting
GFP: green fluorescence protein
GWAS: genome-wide association study
H3K27Ac: histone H3 lysine 27 acetylation
H3K4Me3: histone H3 lysine 4 tri-methylation
LD: linkage disequilibrium
lncRNA: long non-coding RNA
MeDIP: methylated DNA immunoprecipitation
PCR: polymerase chain reaction
v
RNA-Seq: RNA sequencing
RPKM: reads per kilobase per million mapped reads
SNP: single nucleotide polymorphism
TALEN: transcription activator-like effector nucleases
UCSC: University of California, Santa Cruz
ZNF: zinc finger nuclease
vi
ABSTRACT
Colorectal cancer (CRC) is among the top three cancer incidence and death rates in the United
States. To further understand the disease, it is essential to develop a mechanistic comprehension of
cancer initiation and progression. To characterize a CRC associated risk enhancer in the DIP2B
intron, I reversed the presence of this enhancer using CRISPRs (Clustered Regularly Interspaced
Short Palindromic Repeats) by deleting the enhancer from a colon cancer cell line. I subsequently
performed RNA-Seq to identify the candidate genes regulated by the risk enhancer. Furthermore, I
demonstrated the feasibility of using modified Cas9 proteins as epigenetic toggle switches for
enhancer repression and activation.
1
CHAPTER 1
INTRODUCTION
In the past several decades, incidence and death rates of colorectal cancer (CRC) have been
reduced drastically due to lifestyle changes and medical improvements. However, CRC still
remains as the third most common cancer in the United States. Although patients have a 90%
survival rate if the disease is treated in the local stage, only 40% of all CRC patients in the US are
diagnosed in the early stages. Twenty percent (20%) of patients are not diagnosed until the cancer
has metastasized, which drastically decreases the chance of survival to 12.5% (Siegel, Desantis, &
Jemal, 2014). Improvement in early detection programs could reduce the burden of CRC, but the
emotional, physical, and psychological toll on patients and their families will remain immeasurable.
Most colorectal cancer cases develop from three main pathogenic pathways (Pancione, Remo,
& Colantuoni, 2012). The first is microsatellite instability, which is responsible for a large fraction
of the sporadic CRC occurrences. This is caused by inactivation of DNA mismatch repair
mechanisms. The second pathway is through chromosomal instability, most commonly observed in
familial cases. This is caused by deregulation of common tumour suppressor genes and oncogenes
including APC, KRAS, TP53, and SMAD4. The third pathogenic pathway is known as the CpG
island methylator phenotype, in which certain promoters are hyper-methylated at a global level in
2
colon cancers cells. CpG islands are defined loosely as a stretch of DNA with high CpG (Cytosine
followed by Guanine) content. The majority of CpG islands are at promoter regions and they are
normally hypomethylated. However, some of these promoters become hypermethylated in colon
cancer, resulting in transcriptional repression by interfering with transcription factor binding (Goel
& Boland, 2012).
Approximately one third of all CRC cases occur in individuals who have genetic
predispositions for increased risk (Burt, 2000). In some cases, individuals inherit mutations in genes
that have a large effect size. Most genes in this category are tumour suppressors such as APC and
STK11 or DNA repair-associated genes such as MUTYH or MSH2. However, these high-impact
mutations occur very rarely in the population (the allele frequency is less than 0.1%). The more
frequently identified susceptibility loci have much smaller effect sizes but higher allele frequencies
(around 5%). These loci were identified through genome-wide association studies (GWAS) (Peters
et al., 2012; Schumacher, unpublished). Although this method is powerful and informative, there is
a commonly neglected caveat - the susceptibility loci identified through GWAS studies are often
linked to their nearest gene. In reality, a majority of these loci lie in non-coding regions or outside
the gene bodies. Without performing functional studies to identify the relevant disease-associated
genes, results from GWAS studies can be misleading (Zhang, Bailey, & Lupien, 2014).
GWAS is a comprehensive method to identify genomic variations between individuals with
3
and without a disease. Specifically, GWAS analyzes single nucleotide polymorphisms (SNPs),
which are DNA sequence variations at the level of a single nucleotide. At every SNP location, an
individual can be either homozygous for the common (major) allele, heterozygous, or homozygous
for the rare (minor) allele. Although the human genome consists of approximately 12 million SNPs,
due to technological limitations with the array platforms, usually only a portion of these SNPs are
surveyed in any GWAS study. Nonetheless, large blocks of DNA segregate together during meiosis.
This non-random co-segregation forms linkage disequilibrium (LD) blocks. By knowing the alleles
of a few representative “index” SNPs, it is possible to infer the alleles of all other SNPs within their
LD blocks with a certain degree of confidence by a value termed r
2
(Visscher, Brown, Mccarthy, &
Yang, 2012).
Investigators studying CRC compared a large number of individuals with and without the
disease and identified 25 risk SNPs where one of the SNP alleles is enriched in CRC patients over
healthy controls (Yao, Tak, Berman, & Farnham, 2014).
Yao et al. used these 25 high CRC susceptibility index SNPs to perform functional annotation
analysis. The study utilized funciSNP, an R/bioconductor bioinformatic program, to identify
candidate functional SNPs. FunciSNP allows its users to identify SNPs which may have functional
significance by integrating SNP information with functional features in the genome. Taking all
index SNPs and their correlated SNPs with r
2
> 0.5 within a 400 kb window, the study first
4
identified 13 SNPs inside coding regions and 233 SNPs at transcription start sites of transcribed
genes.
To determine whether any CRC index SNPs were correlated with SNPs in functional regions
far from promoters, the study then searched for SNPs that overlapped with histone H3K27
acetylation (acetylation of lysine 27 on histone H3). This histone mark is a signature for active
regulatory regions. When far (e.g. more than 2 kb) from a transcription start site, these regulatory
regions are called enhancers. Enhancers play a role in augmenting expression of their target genes
by recruiting co-activators during transcription. The study used H3K27Ac marks for normal colon
(sigmoid) as well as colorectal cancer cells (HCT116). Using a lenient r
2
cut-off of 0.1, the study
identified 746 SNPs which fell into 199 distinct H3K27Ac marks. These were categorized as
tumour-specific, normal-specific, or common regions. Since H3K27Ac marks are also present in
promoters, only the distal H3K27Ac marks (i.e. Enhancers) were retained. When a more stringent r
2
cut-off of 0.5 was applied, only 68 enhancers remained. Finally, after filtering out enhancers for
which the SNPs fall near the tails of H3K27Ac signal, the study identified 13 normal unique
enhancers, 1 tumour unique enhancer, and 13 common enhancers present in both the HCT116 and
sigmoid colon cell line.
Enhancers identified through GWAS studies are unexplored targets for functional studies.
Enhancers are cell-type specific and play a role in differential gene expression (Calo & Wysocka,
5
2013). Dysfunctional enhancers can therefore contribute to altered expression profiles that may lead
to increased CRC susceptibility. In fact, in silico studies have shown that risk SNPs in enhancers
often disrupt transcription factor binding motifs (Yao et al., 2014). This could increase or decrease
transcription factor binding and disturb expression levels of the target genes. The presence of
tumour-specific and normal-specific enhancers also indicates that single nucleotide differences may
contribute to the enhancer creation or abolition, which ultimately increases CRC susceptibility.
However, compared to transcription start sites and exons, the target genes of enhancers are largely
unknown; less than one third of all enhancers interact with its nearest promoter. Furthermore, each
enhancer interacts with 2.5 promoters on average (Andersson et al., 2014).
One of the most intriguing regions identified from the Yao et al. study was a cluster of
enhancers in the intronic region of DIP2B. DIP2B is a poorly characterized gene that is postulated
to play a role in DNA methylation. The gene potentially interacts with DMAP-1, a DNA
methyltransferase associated protein, through its DMAP-1 binding domain (Winnepenninckx et al.,
2007). In total, the two CRC index SNPs rs7136702 and rs1169552 identified six different risk
enhancers in the introns of DIP2B. Among the six risk enhancers, three were unique to normal
colon, one was unique to tumour, and two were common between both cell types. If these
enhancers are regulating DIP2B, they may be contributing to the cancerous phenotype by disrupting
normal DNA methylation in the genome. DIP2B is also directly upstream of ATF1, a transcription
6
factor that has been identified as cancer-related (Jean et al., 2000; Yamada, Ohno, Hara, Yamada, &
Shimizu, 2013). If the enhancers are regulating nearby genes, ATF1 expression may also be altered
to favor cancer development or progression.
Figure 1. UCSC Genome Browser snapshot of the 400 kb window around index SNPs
rs7136702 and rs11169552. One index SNP is upstream of the DIP2B gene and the other is near
the ATF1 gene. The correlated SNPs with r
2
> 0.1 and r
2
> 0.5 are shown; most of the SNPs
correlated with these two index SNPs lie within the DIP2B or ATF1 genes. Risk enhancers unique
to normal sigmoid cells are shown in green blocks; common and tumour-unique enhancers are
shown in red blocks. H3K27Ac tracks identify both promoters and enhancers, whereas H3K4Me3
marks only identify promoters. The LD block track is from the UCSC Preview Genome Browser.
http://genome.ucsc.edu
Index SNP Index SNP
7
Darker shades of red represent higher r
2
.
The goal of my project was to determine the genes regulated by risk enhancers in the DIP2B
gene. I aimed to reverse the presence of a risk enhancer by deleting it from the genome of HCT116,
a colon cancer cell-line. Subsequently, I aimed to identify differential gene expression between
regular HCT116 cells and those that have the enhancer deletion. The last objective of my project
was to determine whether epigenetic toggle switches can effectively silence or activate enhancers to
act as an alternative approach for characterizing risk enhancers.
8
CHAPTER 2
MATERIALS AND METHODS
All sequences and genomic locations within this study are in reference to the human genome
Hg19 assembly (The Genome Sequencing Consortium, 2001). Sequences for guide RNA and
primer design were retrieved from the UCSC Genome Browser (Kent et al., 2002).
2.1 Guide RNA and Cas9 Preparation- To search for potential CRISPR binding sites in the region
of interest, the genomic sequence was entered into the CRISPR Design website (crispr.mit.edu).
Results with the highest scores from the web tool were entered into NCBI BLAST website
(http://blast.ncbi.nlm.nih.gov/Blast.cgi) to check for specificity. Guide RNA synthesis was based on
the Church Lab gRNA synthesis protocol, version 01-14-2013 option B with slight modifications
(Mali et al., 2013). The guide RNA sequences and their reverse complements were incorporated
into 60-mer oligonucleotides and synthesized by Integrated DNA Technologies. Oligos were
resuspended in nuclease free water and diluted to a working concentration of 5 uM.
To make a double stranded DNA fragment, the two 60-mer oligonucleotides were combined to
produce a double stranded guide RNA insert. PCR reaction consisted of 5 ul of forward and reverse
oligos each, 2x Phusion High-Fidelity Master Mix (NEB), and nuclease-free water for a total
volume of 50 ul. PCR was programmed as follows (all PCR reactions were performed using the
Bio-Rad T100 Thermal Cycler):
9
1. 98 °C for 2:00
2. 98 °C for 0:10
3. 53 °C for 0:20
4. 72 °C for 0:30
5. Go to 2, 3 times
6. 72 °C for 5:00
Qiagen MinElute PCR Purification kit (Qiagen) was used to purify the plasmid. In brief, 5
times the sample volume of PB was added. The solution was passed through MinElute Spin
Columns by centrifugation at maximum speed for 1 minute. The flow-through was discarded and
the column was washed with PE. Again, the column was centrifuged and flow-through was
discarded. The column was centrifuged for 1 extra minute to remove all the solution from the
column. DNA was eluted with 10 ul of EB into a microcentrifuge tube.
Plasmid Clone
Number
Plasmid
Clone
Name
Target Genomic
Coordinates
Target Genomic Sequence
PJF_JH002 E16
L_I
Chr12: 51038978-
51039000
AAAGTAATAGCCCAACAGTTTGG
PJF_JH003 E16
R_I
Chr12: 51041310-
51041332
CAACGACTAACTTCTGAAATAGG
PJF_JH004 E16
R_O
Chr12: 51041497-
51041519
TCCTAACAAGTAAGGTCACTTGG
PJF_JH005 E15
OFF-1
Chr12: 51013029-
51013051
GCAGTACCTTGAGCAATTAGGGG
PJF_JH006 E15
OFF-2
Chr12: 51013403-
51013425
GCGACATAATTGTCCCCACTTGG
PJF_JH007 E11 ON-1 Chr12: 50908822-
50908844
CTCCCTCTCTCGAGCCTGAAAGG
10
PJF_JH008 E11 ON-2 Chr12: 50909304-
50909326
CCAGTCCACAGTGTAGGGCTGGG
PJF_JH009 E11 ON-3 Chr12: 50909702-
50909724
AAATGGATGGAGATAGTAAGTGG
PJF_JH010 E11 ON-4 Chr12: 50910208-
50911230
GAACTAACACTGGAATGATATGG
PJF_JH011 E11 ON-5 Chr12: 50911656-
50911678
GGAACGAAACAGAAGTTAACTGG
PJF_JH012 E11 ON-6 Chr12: 50911911-
50911933
TTTCCCATAAGGTAGCATTCTGG
Table 1. Guide RNA plasmid clones.
To prepare linear guide RNA empty vector, the guide RNA cloning vector (Addgene Plasmid
41824) was digested with AFIII (NEB). 1 ul of enzyme was added to 1.5 ug of DNA in addition to
10X NEBuffer. Nuclease free water was added to a total volume of 50 ul and the sample was
incubated for 1 hour at 37 °C. The solution was run next to the Tridye 2-Log DNA Ladder (New
England Biolabs) on a 1% agarose gel stained with ethidium bromide and visualized under UV
using the Gel Doc EZ System (Bio-Rad). (All gel electrophoresis performed in this experiment
follow this preparation method).
To isolate the linearized gRNA cloning vector, the 600 bp band was excised from the gel. DNA
was purified using the QiaQuick Gel Extraction Kit (Qiagen). 3 volumes of buffer QG was added to
1 volume (by weight) of gel and incubated at 50 °C for 10 minutes. After complete digestion, 1 gel
volume of isopropanol was added to the sample. The sample was transferred to a MinElute column
and centrifuged. After discarding flowthrough, 500 ul of QG buffer was added and centrifuged. The
11
column was washed once with PE, then placed in a microcentrifuge tube. DNA was eluted in 10 ul
of EB. DNA concentration was measured on a Nanodrop (Thermo Scientific).
The Gibson Assembly (NEB) step for guide RNA vector synthesis was carried out using 50 ng
of gRNA cloning vector and 250 ng of insert. After addition of 5 ul 2x Gibson Assembly Master
Mix, the total reaction volume was brought to 10 ul with nuclease free water. The reaction mixture
was incubated at 50 °C for 1 hour.
For transformation, 3ul of the Gibson Assembly product was added to 20 ul of E. Cloni 10G
Chemically Competent Cells (Lucigen) and plated on LB-kanamycin agar plates overnight (USC
Bioreagent and Cell Culture Core). Colonies were inoculated in 5 ml of LB and incubated overnight
in a 37 °C shaking incubator. Bacterial stocks were frozen at -80 °C for future inoculations (400 ul
of the bacteria in LB was mixed with 400 ul of glycine).
The QiaPrep Miniprep Kit (Qiagen) was used to isolate the plasmid. The cells were pelleted by
centrifugation and resuspended in buffer P1 buffer and subsequently lysed by buffer P2. After 5
minutes, the reaction was terminated by the addition of buffer N3. The lysate was centrifuged and
the supernatant was transferred to a QiaPrep spin column. The column was washed with buffer PB
followed by buffer PE. Plasmid DNA was eluted in 50 ul of buffer EB.
12
The plasmids were sequenced by Genewiz Inc to confirm proper ligation. The reaction mixture
consists of 2x GoTaq Green Master Mix (Promega), 600 ng DNA, and 5 uM each of T7-F and
Sp6-R primers (see Table 3). Nuclease free water was added to bring the total volume to 15 ul.
The Cas9 nuclease vector is commercially available (Addgene Plasmid 44719).
dCas9-DNMT3A and dCas9-CBP vectors were constructed and supplied by the David Segal Lab of
the UC Davis Genome Center. LB solution and LB agar plates were supplemented with ampicillin
during transformation and inoculation.
Plasmid Clone
Number
Plasmid Clone Name Description
PJF_JH001 Cas9-GFP Cas9 nuclease (GFP+)
DS001 dCas9-DNMT3A dCas9 with mutated nuclease tethered to
DNMT3A active domain
DS002 dCas9-CBP:HAT dCas9 with mutated nuclease tethered to
the histone acetyltransferase domain of
CBP.
Table 2. Cas9 plasmid clone number, name, and description. Plasmid clone numbers beginning
with DS were designed and cloned by the David Segal Lab of the UC Davis Genome Center.
2.2 Cell Culture and Transfection- The human colon cancer cell line HCT116 (ATCC #CCL-247)
was obtained from the Peter Jones Lab of the USC Norris Cancer Center. Cells were cultured in
DMEM (Corning Cellgro) supplemented with 10% FBS (Gibco by Life Technologies) and 2%
13
penicillin/streptinomycin (USC Bioreagent and Cell Culture Core facility) at 37 °C in 5.0% CO
2
.
The trypsinization method was used for detaching adherent cells from the growth surface to seed,
passage, or prepare for FACS sorting. Briefly, cells were washed with PBS, and then with trypsin.
Cells were placed in the incubator for approximately two minutes. DMEM media was added to the
plate to detach and disaggregate the cells.
Transfections were carried out using Lipofectamine LTX or Lipofectamine 3000 Transfection
Reagent (Invitrogen by Life Technologies) using the manufacturer’s protocol. Opti-MEM solution
and Plus reagent were mixed with the DNA. Lipofectamine reagent was diluted in Opti-MEM, and
then added to the DNA mixture. After incubation at room temperature, the solution was added to
the cells. For transfections carried out in 6-well tissue culture plates (growth area 9.6 cm
2
), cells
were seeded at a density of approximately 6.25x10
5
per well. Transfection of cells in 6-well plates
used 2.5 ug of DNA and 6 ul of Lipofectamine reagent. For transfections carried out in 10cm tissue
culture dishes (growth area 55cm
2
), cells were seeded at a density of approximately 3.5x10
6
per
plate and transfected with 12.5 ug DNA and 30 ug of Lipofectamine reagent.
For enhancer deletion, Cas9-GFP and gRNAs were transfected at a ratio of 1:1 or 1:9 (in ug).
Equal amounts of left side and right side guide RNAs were transfected to achieve deletion of the
region of interest. To confirm CRISPR activity, transfection was first carried out in HEK293 cells.
Genomic DNA was isolated without sorting and tested for the presence of deletions.
14
For enhancer methylation using dCas9-DNMT3A and enhancer histone H3K27 acetylation
using dCas9-CBP:HAT, GFP was co-transfected because the dCas9 was not tagged with GFP. The
transfection ratio of GFP to dCas9 to gRNA was 1:1:9.
2.3 Cell Sorting and Colony Selection- Fluorescence-activated cell sorting (FACS) was carried
out by the Flow Cytometry Core of the USC Norris Comprehensive Cancer Center on the BD
FACS Aria II cell sorter (BD Biosciences). FACS was performed 24 to 48 hours post transfection.
Prior to FACS, cells were collected in 150 ul of 5% FBS in PBS in microcentrifuge tubes. GFP
positive cells were sorted one cell per well into 96-well tissue culture plates, or into 5ml tubes
containing DMEM growth media and subsequently plated at low density (approximately 4,000 to
8,000 cells per plate) in 10 cm tissue culture plates. Cells in 96-well plates were harvested for
genomic DNA isolation when they reached confluency. Colonies on 10 cm plates were picked from
the plate and transferred to a 24-well plate when they became visible by eye, and then cultured until
colonies are big enough for genomic DNA isolation. When collecting cells for genomic DNA
isolation, a small number of cells were transferred to a new well in 24-well plates.
To determine whether the target region was successfully deleted, PCR reactions were carried
out on 50 to 100 ng of genomic DNA. Reaction mixtures consisted of 2 ul of the forward (F) and
reverse (R) primers (10mM), 2x GoTaq Green Master Mix, and nuclease-free water for a total
15
volume of 20 ul. PCR conditions were as follows:
1. 95 °C for 3:00
2. 94 °C for 0:20
3. 60 °C for 0:30
4. 72 °C for (1:00 per 1kb expected product size)
5. Go to 2, 31 times
6. 72 °C for 5:00
Primer Name Genomic
Coordinates
Forward Primer
Sequence
Reverse Primer
Sequence
E16 DE
F/R
Chr12: 51038866-
51042001
TGCAATCGTG
ATTCAACTAAGTAAA
CCACCAAGAT
AAGGATAACACTGA
E16 IN
F/R
Chr12: 51040776-
51042029
GAGAAGCATA
GATGAAAACACAGG
CCACCAAGAT
AAGGATAACACTGA
T6-F/Sp6-R (For gRNA
sequencing)
TAATACGACT
CACTATAGGG
CGCCAAGCTA
TTTAGGTGACA
Table 3. Name and description of primers.
For colonies confirmed to have bi-allelic enhancer deletion, cell pellets were collected to
perform RNA extraction. Two independent passages of clone C43 (outside CRISPR set) were
collected. Viable frozen cell stocks were made by freezing down cells in 10% DMEM in FBS.
These cells were stored in liquid nitrogen for future use.
Cell Stock Name Status Deleted Genomic Coordinate
B15 Bi-allelic deletion clone of E16 Chr12: 51038978-51041310
16
Table 4. Name and description of viable cell stocks frozen in liquid nitrogen.
2.4 Genomic DNA and RNA isolation- Genomic DNA, required for enhancer deletion
confirmation and MeDIP, was isolated using the QiaAmp DNA Mini Kit (Qiagen) as per
manufacturer protocol (Version 06/2012). Cells were harvested and pelleted by centrifugation at
1,000 rpm for 10 minutes. Cells were then resuspended in PBS with proteinase K and buffer AL.
The sample was incubated at 56 °C for 10 minutes. Ethanol was added and the sample was
transferred to a QiaAmp Mini Spin Column. The column was washed with buffer AW1 and AW2.
DNA was eluted in 100 ul of EB.
RNA was isolated using the TRIzol Reagent (Ambion by Life Technologies). Steps that were
indicated as optional in the manufacturer protocol (version 13 Dec 2012) were omitted. All labware
were sanitized with RNAse AWAY to eliminate surface contaminants. Briefly, cells were harvested
and TRIzol reagent was added to lyse cells. To perform phase separation, Chloroform was added
and the sample was centrifuged to separate the organic and aqueous phases. The aqueous phase was
transferred into a new tube for RNA isolation. RNA was precipitated via addition of isopropanol
using inside CRISPR set (2332 bp)
B38 Bi-allelic deletion clone of E16
using inside CRISPR set
Chr12: 51038978-51041310
(2332 bp)
C43 Bi-allelic deletion clone of E16
using outside CRISPR set
Chr12: 51038978-51041519
(2541 bp)
17
followed by centrifugation. The pellet was washed with 75% ethanol and left to air dry. RNA was
then resuspended in DEPC-treated water, incubated at 55 °C and subsequently stored at -80 °C.
2.5 Methylated DNA Immunoprecipitation (MeDIP)- Methylated DNA immunoprecipitation
(MeDIP) was performed to confirm forced methylation by dCas9-DNMT3A at the enhancer of
interest. To verify successful transfection, cells were observed under fluorescent microscope for
GFP expression 24 hours post transfection. Cells were collected for genomic DNA isolation
approximately 48 hours after transfection.
Antibodies used for MeDIP include 5-methylcytocine monoclonal antibody (Active Motif),
and rabbit-anti-mouse IGG secondary antibody (Cappell). The protocol for MeDIP was developed
with reference to Active Motif’s MeDIP kit manual as well as protocols by Thu et al. (2009) and
Weng, Huang, & Yan (2009). Genomic DNA was diluted to 100 ng/ul and sonicated at 4 °C for 30
minutes using cycles of 30 seconds ON/30 seconds OFF to yield product size of 200-600bp
(Bioruptor by Diagenode, high setting). 15 ul of sonicated material were used for gel
electrophoresis using 10X Orange Loading Dye (LI-COR Biosciences) to confirm efficient
sonication. 10% of sonication product was stored in -20 °C as input. 3 ul of sonicated product was
placed in a 96 °C water bath for 10 minutes to denature DNA into single strands. Samples were
kept on ice for 10 minutes before addition of 10ul of ice cold 10x IP buffer, 2 ul each of primary
18
and secondary antibody, and ice cold deionized water to a final volume of 100 ul. The sample was
rotated for two hours in 4 °C. 25 ul of Protein A/G magnetic beads (VWR) were washed with 1x ice
cold IP buffer and added to the sample followed by overnight rotating incubation at 4 °C overnight.
To elute, the sample was placed on a magnetic rack and washed twice with 150ul of 1x IP buffer.
The beads were resuspended in 150 ul of 1x IP buffer, transferred to a new microtube, and washed
again. Beads were then resuspended in 250 ul of digestion buffer and 2 ul of Proteinase K (Qiagen)
and incubated at 50 °C for two hours. The supernatant containing immunoprecipitated DNA was
isolated. DNA cleanup was performed using the MinElute PCR Purification Kit (Qiagen) followed
by elution into 16 ul of EB.
PCR was performed for semi-quantitative analysis of the input DNA and MeDIP results.
Primers include positive control, negative control, and target region primers. PCR reaction mixture
consisted of 2 ul of immunoprecipitated DNA, 10 um each of the reverse and forward primer, 2x
GoTaq Green Master Mix, and nuclease-free water for a total volume of 20 ul. PCR was run under
the following conditions and subsequently loaded on agarose gel to determine target region
enrichment.
1. 95 °C for 3:00
2. 94 °C for 0:20
3. 60 °C for 0:30
4. 72 °C for 0:30
5. Go to 2, 39 times
19
6. 72 °C for 5:00
Primer Name Genomic Coordinates Forward Primer
Sequence
Reverse Primer
Sequence
MeDIP Positive
Control 1 F/R
Chr12: 123187714-
123187898
GATAAACTCC
AGCCCCAACA
CAGACACACA
CCTCCTTGCT
MeDIP Negative
Control 1 F/R
Chr10:114710738-
114710911
TATCTGTTTC
CTGGGCTTGG
GGAGAAGGGG
GAGAAAAAGA
E15 MeDIP-2
F/R
Chr12:51013233-
51013340
CCAGACCCAA
ACCAAATGTC
CGTACACAAA
TGCAGCACCT
Table 5. Name and description of all primers used for MeDIP experiments.
2.6 Histone H3K27Ac Chromatin Immunoprecipitation- H3K27Ac chromatin
immunoprecipitation (ChIP) was performed to confirm the histone acetylation effect of dCas9-CBP
on the target enhancer. To verify successful transfection, cells were examined using a fluorescent
microscope for GFP expression 24 hours post transfection.
Cells were cross-linked approximately 48 hours post-transfection at room temperature in 1%
formaldehyde on a rotating platform for 10 minutes before the reaction was terminated by addition
of glycine to a final concentration of 0.125 M. After 5 minutes, cells were washed twice with
ice-cold PBS, scraped into a tube, and centrifuged in 10ml PBS at 430 rcf for 5 minutes at 4 °C.
Cells were then resuspended in cold cell lysis buffer, homogenized, and centrifuged at 4 °C to
isolate cell nuclei.
20
In preparation for sonication, cells were resuspended with 5x the volume of the pellet in
ice-cold nuclei lysis buffer and then incubated on ice. Cells were sonicated to achieve an average
chromatin length of 200-500 bp, which usually requires a 45 minutes of sonication (30 second pulse
at high setting with 90 seconds of pause between each pulse). Chromatin size was verified on 1.5%
agarose gel prior to proceeding with immunoprecipitation.
For immunoprecipitation, 500 ng of chromatin was saved as input and stored in -20 °C. 10 ug
of chromatin was diluted with 5-fold ice-cold IP dilution buffer with protease inhibitors. The
chromatin was incubated on a rotating platform at 4 °C overnight after addition of 6 ul of H3K27Ac
antibody (Active Motif).
15 ul of magnetic protein A/G beads were washed with 15 ul of ice-cold 1x IP buffer added to
the sample and incubated on a rotating platform at 4 °C for 2 hours. The beads with captured
chromatin were washed three times with IP wash buffer 1 and then twice with IP wash buffer 2. The
mixture was transferred to a new tube and the beads were washed again with IP wash buffer 2.
To elute the complex, the wash buffer was discarded and replaced by adding 75 ul of IP elution
buffer. The complex was placed on a vortexer at low setting for 30 minutes. Supernatant containing
the chromatin was transferred to a new tube after allowing beads to settle in a magnetic rack. The
elution process was repeated, resulting in a combined total of 150 ul of material. 20 ul of 5 M NaCl
21
was added to 150 ul of elution buffer for each ChIP as well as the input. All samples were incubated
at 67 °C overnight to reverse cross-link.
DNA was purified using the MinElute PCR Purification Kit (Qiagen) and eluted in 35 ul of EB.
ChIP results for the sample and input were evaluated by PCR using positive control primer,
negative control primer, and target region primer.
For PCR, input DNA was diluted by a factor of 1:50. The PCR reaction consisted of 1 ul DNA
(input or sample), 2X GoTaq Green Mater Mix, 2 ul of reverse/forward primer mix (10 mm), and
water to a final volume of 20 ul. PCR was run for under the following conditions:
1. 95 °C for 3:00
2. 94 °C for 0:20
3. 60 °C for 0:30
4. 72 °C for 0:30
5. Go to 2, 31 times
6. 72 °C for 5:00
Primer Name Genomic
Coordinates
Forward Primer
Sequence
Reverse Primer
Sequence
LSR F/R
(Positive Control)
Chr19: 35758370-
35758510
CCACTACGAC
GACTTCAGGT
CTCACAGCCT
CCTCCAGTAG
ZNF333 F/R
(Negative Control)
Chr19: 14829617-
14829751
TGAAGACACA
TCTGCGAACC
TCGCGCACTC
ATACAGTTTC
E11 ChIP1 F/R Chr12: 50908774-
50908916
GCCACAGTGT
TGCATGAGAA
TTAATGAGGC
AGGGGAAATG
E11 ChIP2 F/R Chr12: 50909308-
50909442
GCCCTACACT
GTGGACTGGT
CCCTCATCAG
GAGAAAGAAAGA
22
E11 ChIP-3 F/F Chr12: 50911841-
50911985
TGAGGGAACT
GAGGTAGCAAA
ACCCAATTCA
GCATGACTCC
Table 6. Name and description of all primers used for ChIP experiments.
2.7 Experimental controls- For RNA-Seq, the controls for enhancer deletion were HCT116 cells
transfected with empty vectors. These cells were transfected in the same conditions with the
exception that guide RNAs vectors were empty vectors. The cells were also selected for using
FACS and cultured in the same condition as the deletion clones. Two clonal populations were
selected after FACS sorting. A third control was a pooled population of cells that were collected
from FACS sorting.
Controls for MeDIP and H3K27Ac ChIP were HCT-116 cells grown under the same condition
in parallel with those that undergo transfection. The transfected cells and controls were harvested
and cross-linked at the same time.
2.8 RNA Sequencing (RNA-Seq)- Two different clones with inner CRISPR set deletion were used
as replicates. For outer CRISPR set deletion, two different passages of the same clone were used as
replicates.
RNA sequencing was performed by the USC NCCC Molecular Genomics Sequencing Facility.
RNA samples were first analyzed using the Experion Automated Electrophoresis System (Bio-Rad)
23
for quality control. RNA library preparation and single read RNA sequencing resulted in
approximately 26 million reads per sample.
Results from sequencing were analyzed using the Partek Flow software. Reads were aligned
using TopHat2 version 2.0.8 with reference to Gencode V19. Gene-level differential gene
expression was analyzed for inner deletion clones and outer deletion clones against the empty
vector control. For quality control, each gene must have a p-value of less than 0.05, total RPKM
greater than 1, and fold change greater than -2 to be considered as significant.
2.9 Primer design and preparation- All primers were designed using the Primer3 software (Rozen
& Skalestsky, 1998). Primers were validated for specificity using the NCBI BLAST and the UCSC
in silico PCR online software (genome.ucsc.edu). Oligos were ordered from Integrated DNA
Technologies and diluted to 10 mm working concentrations in nuclease free water. The specificity
of each primer set was verified in HCT116 genomic DNA.
For MeDIP, methylation status of the control regions were confirmed using HCT116 whole
genome methylation data (Blattler et al., 2014). Control primers were validated by performing
MeDIP on HCT116 genomic DNA. Positive and negative primers for histone H3K27Ac ChIP were
previously validated primers from the Farnham Lab.
24
CHAPTER 3
RESULTS
3.1 Specific Aim 1: Reverse enhancer presence of enhancer 16
Enhancer 16 (E16) in the intronic regions of DIP2B is the only tumour-specific enhancer
identified in the study by Yao et al. as shown in Figure 2, the H3K27Ac mark is robust in HCT116,
a colorectal cancer cell line, but minimal in normal sigmoid cells. Three risk SNPs (rs13378012,
rs12231317, and rs12231309) were identified in this enhancer by two index SNPs (rs7136702 and
rs11169552). The two latter risk SNPs are in prominent regions of the H3K27Ac mark and overlap
with transcription factor binding sites according to ENCODE ChIP-Seq studies. Furthermore, in
silico analysis performed by Yao et al. also predicted altered transcription factor binding strength in
the presence of disease-associated alleles. The SNP rs13378012 decreases PRDM1 and RXRA
binding but increases affinity for RUNX1, while rs12231309 increases the strength of E2F4 binding.
The UA5 motif also decreases in strength in the presence of the disease-associated allele of
rs12231309.
Risk SNP ID Chromosomal
Location
Index SNP ID (r
2
) Distance from Index
SNP
rs13378012 Chr12: 50646614
rs7136702 (0.59) 160,181 bp
rs12231317 Chr12: 51067222 rs7136702 (0.59);
rs11169552 (0.59)
160,789 bp;
114,658 bp
25
rs12231309 Chr12: 50647279 rs11169552 (0.59) 114,601 bp
Table 7. SNP ID, chromosomal location, and index SNP information for the three risk SNPs
identified in E16.
I hypothesized that this enhancer may increase the risk for development of CRC. This novel
enhancer may be disrupting the balance of gene expression by up-regulating genes favoring tumour
development.
26
Figure 2. UCSC Genome Browser snapshot of the risk enhancer cluster in DIP2B and E16. A.
The DIP2B intron has six risk enhancers identified by two index SNPs (Indicated by red arrows).
Risk enhancers unique to normal sigmoid cells are shown in green blocks; common and
tumour-unique enhancers are shown in red blocks. From left to right, the enhancers are identified as
E11, E12, E13, E15, E14, and E16, respectively.
B. E16, a tumour-unique enhancer, harbours three risk SNPs identified by two different index SNPs.
These risk SNPs either increase (+) or decrease (-) the strength of transcription factor binding as
well as the UA5 motif (shown in blue). The second and third SNP reside in the prominent portion
of the H3K27Ac mark and overlaps with TF binding sites identified by ENCODE ChIP
27
experiments. Parenthesis behind each TF indicates the cell-line used for the particular ChIP
experiment.
To reverse the presence of this tumour-specific enhancer, I applied the CRISPR (Clustered
Regularly Interspaced Short Palindromic Repeats) genome engineering technology to delete the
enhancer from HCT-116. Each CRISPR consists of a Cas9 nuclease and a guide RNA (gRNA).
gRNAs have an invariant portion and a variable portion that binds to complementary DNA. The
Cas9 nuclease subsequently cleaves both strands of the DNA (Hsu, Lander, & Zhang, 2014). To
achieve a deletion, two gRNAs targeting each side of the enhancer are required. A deletion will
occur when two simultaneous cuts occur, followed by non-homologous end joining DNA repair.
Figure 3. Enhancer deletion using the CRISPR technology.
CRISPRs consist of a Cas9 nuclease and a guide RNA. The guide RNA has an invariable
component and a variable component which binds to its complementary DNA. The Cas9 nuclease
is GFP-tagged to facilitate selection. To delete an enhancer, cells must be transfected with Cas9,
gRNA(L), and gRNA(R).
The CRISPR technology is superior compared to its predecessors ZNFs (Zinc Finger
28
Nucleases) and TALENs (Transcription Activator-Like Effector Nucleases). Compared to ZNFs,
CRISPRs show higher specificity. ChIP-Seq experiments showed that CRISPRs have an almost
perfect specificity for its intended target while ZNFs have thousands of off-target binding sites.
Compared to TALENs, CRISPRs have less restriction in terms of target sites, and are much easier
to construct (Segal & Meckler, 2013).
To isolate clonal population with bi-allelic deletions, HCT116 cells were transfected with Cas9
nuclease and two gRNAs surrounding the enhancer. To eliminate false positive and off-target
effects on gene expression, two different gRNAs were designed downstream of the enhancer. Using
the first and second gRNAs will create an “inner CRISPR deletion”, whereas using the first and
third gRNAs will create an “outer CRISPR deletion” (Figure 4B). The true positive expression
changes would be observed in both types of enhancer-deleted clones. The outer CRISPR deletion
removes more transcription factor binding site than the inner CRISPR deletion. This must be taken
into consideration when analyzing the consequences of enhancer deletion. Since the Cas9 nuclease
was fused to GFP, transfected cells were FACS sorted to isolate cells that have high GFP expression.
Clonal population of cells were cultured at low density and tested for bi-allelic deletion. At a
Cas9:gRNA ratio of 1:1, the percentage of cells with mono-allelic and bi-allelic deletion were both
5.3%. Altering the ratio to 1:9 resulted in a slight increase in deletion frequency; both mono-allelic
and bi-allelic deletions were found in 6.3% of clones tested for deletion. I designed two sets of
29
primers to confirm successful deletion. The first set flanks the enhancer and the second set staggers
a cut site. In total, three clones with a bi-allelic deletion were identified from 51 clones - two clones
for the “inner CRISPR deletion set” and one clone for the “outer CRISPR deletion set”.
Figure 4. Workflow for isolating deletion clones. A. Cells were transfected with Cas9-GFP,
gRNA(L), and gRNA(R). Cells that were successfully transfected with the Cas9-GFP plasmid
showed green fluorescence. Following FACS, cells were grown at low density. Each clone had no
deletion, mono-allelic deletion, or bi-allelic deletion. B. Schematics of double-stranded cuts and
primer locations. Two different gRNAs, inner (I) and outer (O), were designed on the downstream
side of the enhancer. Deletion (DE) primers and inside (IN) primers were used to test for deletions.
C. Left: PCR results for control (untransfected) and three different deletion clones were visualized
on a polyacrylamide gel stained with ethidium bromide. For the DE primer set, the expected size of
30
the control (3164bp) was too big to detect using these PCR conditions.
3.2 Specific Aim 2: Determine gene expression changes resulting from enhancer deletion
After isolating three clones with a bi-allelic deletion of E16, I performed RNA sequencing
(RNA-Seq) to determine differential gene expression. For the inner CRISPR set, the two different
deletion clones were used as replicates. For the outer CRISPR set, independent passages of the
deletion clone were used as replicates. Single end RNA sequencing resulted in approximately 26
million reads per sample. Using the Partek Flow Pipeline that applies the tophat2 read aligner, I
identified genes which had expression changes. Only down-regulated genes were considered for the
analysis. This is because enhancers regulate target genes by increasing gene expression. Thus,
deletion of an enhancer will result in decreased expression of its target. Up-regulated genes are
considered as secondary effects of enhancer deletion. For the analysis, I used a cut-off of 2 fold
down.
Overall, the inner CRISPR set deletion and the outer CRISPR set deletion resulted in 1862 and
2644 down-regulated genes respectively. Among these genes, 1719 were common between the two
sets of deletions. For the inner set deletion, 118 genes were down-regulated by more than 3 fold.
This number increased to 541 for the outer set deletion samples. The overlap between the two
samples was 101 (Figure 5). This indicates that almost all the genes down-regulated in the inner
31
CRISPR deletion were also down-regulated in the outer CRISPR set deletion. Furthermore, the
impact of the outer deletion was more profound in comparison to the inner deletion.
Figure 5. Overlap of down-regulated genes from inner deletion and outer deletion clones. A.
Genes that were more than 2 fold down-regulated. B. Genes that were more than 3 fold
down-regulated. Almost all down-regulated genes resulting from the inner CRISPR set deletion
were also down-regulated in the outer CRISPR set deletion. More genes were down-regulated
following the outer CRISPR set deletion in comparison to the inner CRISPR set deletion.
Initially, I limited my analysis to genes that were down-regulated more than 2 fold in both sets. The
largest average fold change was -10.2 for the gene RP11-161H23.5, a long non-coding RNA
(lncRNA) on chromosome 12 approximately 1.4 Mb upstream from E16. Within the top 20 most
32
down-regulated genes, 5 genes were on chromosome 12. Among these 5 genes, 2 genes were within
2 Mb of the enhancer (RP11-161H23.5 and RP11-386G11.10).
Chr. Gene Name Average FC
12 RP11-161H23.5 -10.2
16 MT1E -7.5
20 RP5-977B1.11 -7.0
12 RP11-386G11.10 -5.7
17 TLCD1 -5.1
14 RP11-649E7.5 -4.8
2 BCYRN1 -4.6
2 ODC1 -4.5
11 CST6 -4.4
17 RP11-498C9.3 -4.4
20 RP5-850E9.3 -4.4
16 MT2A -4.3
12 RP5-940J5.9 -4.3
16 GINS2 -4.3
1 HMGN2 -4.3
17 ALYREF -4.3
12 RP11-603J24.17 -4.2
6 GTF3C6 -4.2
16 RP11-343C2.12 -4.2
12 EMG1 -4.2
Table 8. Chromosomal location, gene name, and average fold change of the 20 most
down-regulated genes. Genes on chromosome 12 are shown in bold.
33
Figure 6. Down-regulated genes following enhancer deletion. A. Genes were ranked by fold
change and plotted against average fold change. Genes on chromosome 12 are highlighted as red
dots. B. Circos plot showing the top ten down-regulated genes. The location of the enhancer is
indicated by “E” on the outside of the circos plot. The two genes on chromosome 12 are around 1.3
kb apart and therefore appear overlapped on the plot. Blue lines link each gene to the location of the
enhancer.
Within chromosome 12, 122 genes were down-regulated more than 2 fold. Of these genes, 16 were
within 5 Mb of the enhancer and 2 were within 2Mb (RACGAP1, -2.4 fold; COX14, -2.3 fold). 3
out of the 16 genes within 5 Mb of the enhancer showed more than a 3 fold decrease in expression.
34
More than 2 fold
down-regulated
More than 3 fold
down-regulated
All 1719 284
Chr12 107 15
< 5Mb from E16 16 3
<1Mb from E16 2 0
Table 9. Number of genes down-regulated by more than 2 fold or more than 3 fold.
Figure 7. Fold change vs distance in megabase (Mb) of all down-regulated genes on
chromosome 12, in relation to E16. Many of the down-regulated genes on chromosome 12 were
clustered within 10 Mb of the enhancer.
As mentioned above, the outer CRISPR deletion resulted in loss of more transcription factor
binding sites. Although it is not clear whether these transcription factors bind to the enhancer in
2
3
4
5
6
7
8
9
10
11
-60 -40 -20 0 20 40 60 80
Fold change
Distance from E16 (Mb)
35
HCT116 cells, I analyzed whether there are interesting patterns in genes that were exclusively
down-regulated in the outer deletion sample. For this analysis, I restricted my data to genes that
were down-regulated more than 2 fold in the outer CRISPR deletion sample but excluded from that
set those genes that were also down-regulated more than 2 fold in the inner CRISPR deletion cells.
Applying this filter resulted in 925 genes (refer to Figure 5A).
Of these 925 genes, 51 were located on chromosome 12. 3 of these showed expression
changes greater than 3 fold (RPS2P5, -4.5 fold; PWP1, -4.0 fold; KRT8, -3.2 fold). These 3 genes
rank number 1, 3, and 10 in terms of most down-regulated genes on all chromosomes.
More than 2 fold
down-regulated
More than 3 fold
down-regulated
All 925 15
Chr12 51 3
< 5Mb from E16 6 1
<1Mb from E16 1 0
Table 10. Number of genes down-regulated by more than 2 fold or more than 3 fold for genes
which were exclusive to outer CRISPR deletion set.
Chr. Gene Name Average FC
12 RPS2P5 -4.5
19 CFD -4.1
12 PWP1 -4.0
36
3 ACOX2 -3.6
8 CHCHD7 -3.3
19 TMEM238 -3.3
16 CYBA -3.3
15 MESP1 -3.3
16 NPW -3.2
X PRDX4 -3.2
12 KRT8 -3.2
22 POLDIP3 -3.2
8 PDLIM2 -3.1
1 FBXO6 -3.1
1 ATP6V0B -3.1
13 RASL11A -3.0
18 FAM210A -3.0
1 TOR3A -3.0
19 UBE2M -3.0
1 ZMYM6NB -2.3
Table 11. Chromosomal location, gene name, and average fold change of the 20 most
down-regulated genes exclusive to outer CRISPR deletion set. Genes on chromosome 12 are
shown in bold.
37
Figure 8. Down-regulated genes exclusively in outer CRISPR deletion set. A. Genes were
ranked by fold change and plotted against average fold change. Genes on chromosome 12 are
highlighted as red dots. B. Circos plot showing the top ten down-regulated genes. Blue lines link
the genes on different chromosomes to the enhancer, whereas red lines link the genes on
chromosome 12 to the enhancer.
3.3 Specific Aim 3: Explore the possibility of mutant Cas9 constructs as epigenetic enhancer
toggle switches
Knowing the epigenetic characteristics of silenced and active enhancers, I explored the
possibility of applying epigenetic toggle switches derived from the CRISPR system.
Active enhancers are generally associated with high H3K27Ac, whereas silenced enhancers
lack H3K27Ac (Calo & Wysocka, 2013). This is because transcription factors that bind to
regulatory regions, such as p300, often have histone acetyltransferase activity (The ENCODE
consortium, 2012). We postulated that forced acetylation at H3K27 may transform silenced
enhancers into active enhancers if the histone mark plays a role in recruiting factors to augment
enhancer activity. To test this hypothesis, we used a mutant Cas9 (dCas9) construct with abolished
nuclease activity (but still retaining RNA binding activity). The protein is tethered to the acetylase
domain of CBP (CREB-binding protein). By designing gRNAs to target a silenced enhancer, it may
38
be possible to reverse the enhancer state and create an active enhancer.
Alternatively, many enhancers in the repressed state have high levels of DNA methylation at
their CpG sites, obstructing transcription factor binding and silencing the enhancer (Lister et al.,
2009). I hypothesized that forced methylation at active enhancers may be sufficient to deactivate an
enhancer. I used dCas9 tethered to the functional domain of DNMT3A (DNA methyltransferase 3A)
to test my hypothesis, since DNMT3A catalyzes the methylation of CpG cytosines.
Risk
Enhancer
Risk SNP ID Chromosomal
Location
Index SNP ID
(r
2
)
Distance from
Index SNP
E11 rs10876045 Chr12: 50909254 rs7136702 (0.59) 29,038 bp
E15 rs10876072 Chr12: 51013559 rs11169552(0.59) 142,102 bp
Table 12. SNP ID, chromosomal location, and index SNP information for the risk SNPs
identified in enhancer 11 and 15. These two enhancers are the subject for epigenetic toggle switch
studies.
To investigate whether dCas9-DNMT3A can be used as an epigenetic toggle switch, we
designed two guide RNAs to target enhancer 15 (E15). Enhancer 15 is active in both tumour and
normal colon cells. DNA methylation data for HCT116 indicates the presence of a methylation
valley which overlaps accurately with the H3K27Ac mark. There is one risk SNP (rs10876072) in
the enhancer identified by the index SNP rs11169552, around 142 kb downstream of the index SNP
and having an r
2
of 0.59. The risk allele increases the UA9 motif (top motif in NANOG hes
39
ChIP-Seq data) strength, but decreases the binding affinity of two transcription factors, EF1 and
ETS1. Methylation status of this region was tested through methylated DNA immunoprecipitation
(MeDIP) on genomic DNA collected 48 hours post-transfection. Cells were co-transfected with
GFP to confirm successful transfection.
Figure 9. Enhancer 15, the target for forced DNA methylation by dCas9-DNMT3A.
E15 is a CRC risk-associated enhancer present in both sigmoid and HCT116 cells. The enhancer
has one risk SNP which alters the binding of EF1 and ETS1. It also increases the strength of UA9
motif. The pink shaded region is the target for forced methylation.
Using the methylated DNA immunoprecipitation technique (MeDIP), I tested the ability of
dCas9-DNMT3A to methylate the target region in E15. I found that cells that were transfected
showed enrichment over input. However, this enrichment was absent in untransfected cells. I used
http://genome.ucsc.edu
40
negative control primers, targeting an unmethylated region of the genome, and positive control
primers, targeting a highly methylated region in the genome to validate the specificity of the
antibody.
Figure 10. Preliminary results from MeDIP for methylating enhancer 15 using
dCas9-DNMT3A. The input (IN) and sample (S) of untransfected HCT116 (U) were compared to
cells transfected with dCas9-DNMT3A and gRNAs (T). The negative control and positive control
primer sets are experimental controls to confirm MeDIP was performed successfully. The enhancer
15 primer set shows the presence of enrichment over the input for transfected sample, but not for
untransfected sample.
Enhancer 11 was the chosen target for testing the feasibility of the dCas9-CBP epigenetic
toggle switch. This normal-unique enhancer has a robust H3K27Ac peak, which is lost in the
cancerous HCT116 cells. The loss of enhancer function may be contributing to the disease by
41
lowering expression of important genes. The enhancer harbours the risk SNP rs10876045 which
was identified by the index SNP rs7136702, approximately 21 kb upstream, with an r
2
of 0.59. I
designed six guide RNAs spanning the entire enhancer region. To determine if dCas9-CBP can
successfully deposit H3K7Ac marks in this region, I transfected HCT116 cells with dCas9-CBP and
the gRNAs. Cells were co-transfected with GFP to confirm successful transfection. At 48 hours
post-transfection, cells were cross-linked for chromatin immunoprecipitation (ChIP).
Figure 11. Enhancer 11, the target for forced H3K27 acetylation by dCas9-CBP. Enhancer 11
(E11) is a normal-unique enhancer with robust H3K27Ac in sigmoid cells and minimal H3K27Ac
in HCT116 cells. Whole genome bisulfite sequencing indicates that every CpG in this region is
highly methylated in HCT116. Six guide RNAs targeting this region were designed to force the
deposition of H3K27Ac marks by CBP tethered to dCas9.
http://genome.ucsc.edu
42
The feasibility of using dCas9-CBP as an epigenetic toggle switch for activating enhancers is
currently under investigation.
43
CHAPTER IV
DISCUSSION
In this study, I have characterized a colorectal cancer risk-associated enhancer. This risk
enhancer was identified because it contains SNPs in high LD with CRC risk SNPs discovered using
GWAS. GWAS is a powerful technique to identify the susceptibility loci of a disease. However, the
lack of post-GWAS analyses result in misleading information by defining the closest gene as the
risk loci, rather than identifying the gene that is truly contributing to the disease. I have applied
several different techniques to identify the targets of a CRC risk-associated enhancer. My results
demonstrate the necessity of post-GWAS characterization because the nearby genes to the
risk-associated enhancer do not appear to be the targets.
My initial approach for this study included using both TALENs (Transcription Activator-Like
Effector Nucleases) and CRISPRs as genome engineering platforms. However, CRISPRs were
quickly proven to be the superior choice for many reasons. Firstly, TALENs require the formation
of a heterodimer to achieve a double stranded cut. This means that for each deletion, four different
TALEN constructs are required. Not only is this tedious, it also creates stringent limits in terms of
feasible target sites in the genome. Secondly, unlike CRISPRs which have gRNAs as its variable
component, sequence recognition for TALENs is intrinsic within the protein. Constructing a gRNA
vector only requires the ligation of a plasmid backbone to an oligonucleotide that contains the target
44
sequence. In comparison, each position in the target sequence is a separate plasmid for TALENs. A
TALEN monomer recognizing 17 bases in the genome therefore requires the fusion of 17 different
plasmids and two backbone plasmids.
Overall, during the course of my studies, I collaborated with Yu Gyoung (Esther) Tak to
construct and confirm the activity of CRISPRs targeting 9 out of 28 of the risk enhancers identified
by the post-GWAS analyses performed by Yao et al. Of these, we have obtained at least one clone
with a bi-allelic deletion for 5 different risk enhancers. For the risk enhancers in 18q21.1 and
19q13.1, we have performed RNA-Seq to identify candidate genes associated with the enhancer.
In this study, I successfully deleted a tumour-specific CRC risk enhancer in the intron of
DIP2B. Prior to sorting cells with FACS, the efficiency of isolating clones with mono-allelic or
bi-allelic deletion was extremely low. For deletions of E16, 50 clones were tested but failed to show
any deletion. I also attempted to delete E15. Although genomic DNA isolated from pooled HCT116
following transfection showed that some cells have successful deletion, 35 clones were tested but
none had mono-allelic or bi-allelic deletion.
I therefore modified my experimental approach and used FACS to sort and select for cells that
express GFP. This was made possible by using a new Cas9 construct, in which the nuclease was
tethered to GFP. Although transfection efficiency appeared to be high (around 70%) under a
fluorescent microscope, FACS sorting consistently results in the selection of less than 5% of total
45
cell population. This may be due to the stringency of the gate I have set to only select for cells with
very high GFP expression.
To further optimize CRISPR deletion efficiency, I altered the transfection ratio of Cas9:gRNA
from 1:1 to 1:9. Since all cells are first pre-selected by GFP expression, I postulated that increasing
the availability of gRNA plasmids and making Cas9 limiting may increase the deletion efficiency. I
observed a slight increase in deletion frequency. At a transfection ratio of 1:1, only 1 clone with a
mono-allelic deletion and 1 clone with a bi-allelic deletion was identified out of the 19 tested clones
(5.3% for both cases). Using a ratio of 1:9, we identified 2 clones with a mono-allelic deletion and
2 clones with a bi-allelic deletion (6.3% each) after surveying 32 clones. Compared to the rate of
0% for clones selected without FACS sorting, these efficiencies were considered as exceptional.
Figure 12. Transfection efficiency of HCT116 cells. A. HCT116 cells transfected with Cas9 and
46
gRNAs, observed under an optical microscope. B. The transfected cells observed under
fluorescence microscope. Cells that were successfully transfected with the Cas9 plasmid appear
green under fluorescent microscope.
Although FACS sorts for cells with successful transfection, this does not guarantee that all
cells will have a deletion. Deletions only occur when the nuclease upstream and downstream of the
target site act synchronously. This is because one double stranded cut can be repaired efficiently by
the cellular repair mechanism. If the repair is erroneous and creates small insertions, deletions, or
base substitutions, the gRNA will no longer be able to recognize the binding site. This prevents
subsequent nuclease action. When both upstream and downstream nucleases act at the same time, it
is also possible that the cell uses homologous recombination to repair the damaged site. Overall,
many random events must occur in a specific order to achieve bi-allelic deletion even when cells
are successfully transfected with the Cas9 and gRNA plasmids.
For this project, I have taken several cautionary steps to ensure the quality of the results.
HCT116 cells for transfection were kept at relatively low passage numbers. To reduce the impact of
batch effects, the empty vector control clones were cultured, prepared and sequenced at the same
time as the samples. Furthermore, I limited my data from RNA-Seq to genes which have a p-value
of less than 0.05, total RPKM greater than 1, and fold change of less than -2 (down-regulated by
47
more than 2 fold). These cut-offs are common for RNA-Seq experiments and consistent with our
analysis for clones with deletions of other risk enhancers.
RNA-Seq results for the clones with a bi-allelic deletion indicated that RP11-161H23.5 (1.2
Mb upstream of enhancer 16) expression decreased by approximately 10 fold. However, I observed
that this long non-coding RNA (lncRNA) was consistently down-regulated in all RNA-Seq samples,
regardless of which enhancer was deleted. This indicates that down-regulation of the lncRNA is
likely not a direct result of the loss of interaction between its promoter and the risk enhancer.
Of note, the expression changes in DIP2B could not be determined with confidence since the
p-value for the gene was greater than the cut-off of 0.05. Since the RPKM value was higher than
the cut-off for DIP2B, it is likely that the high p-value resulted from large variation between the two
sample replicates or the three control triplicates. The gene showed a 1.1 fold increase in expression
with a p-value of 0.2 and 0.5 in the two CRISPR deletion sets respectively. These results do not
show any indication of DIP2B being down-regulated following enhancer deletion. ATF1, directly
downstream of DIP2B, showed a modest decrease in expression of 1.5 fold.
5 out of the 20 most down-regulated genes (20%) are on chromosome 12. The human genome
consists of approximately 58,000 genes. Of these, only around 4,000 (7%) are on chromosome 12.
This suggests that the high percentage of most down-regulated genes on chromosome12 is most
likely not random. Nonetheless, a majority of the most down-regulated genes are outside of
48
chromosome 12, indicating the need for additional experiments, such as circular chromosome
conformation capture (4C) to distinguish the direct target genes from downstream effects.
Each gene can be regulated by more than one enhancer (Feingold et al., 2004). It is therefore
possible that deleting a single enhancer may result in a modest effect on expression or trigger a
compensation mechanism in the cell to increase the use of another enhancer regulating the same
gene. It may be necessary to delete a cluster of enhancers to result in a strong impact on gene
expression.
Comparing the impact of the inner CRISPR set and the outer CRISPR set demonstrated that
the larger deletion resulted in a higher quantity and more profound impact on gene expression.
Most of the down-regulated genes in the inner CRISPR set were also present in the outer CRISPR
set, indicating that gene expression changes were not random. One explanation is that RNA-Seq
results for the outer CRISPR deletion set have more false positive because two passages (as
opposed to two independent clones) of the same clone were used as replicates. Another explanation
is that the more profound impact of the outer CRISPR set may be due to the loss of additional
transcription factor binding sites that were only removed by the larger deletion. Although the ChIP
experimental data displayed on the UCSC genome browser is a combination of many different
cell-types, it is clear that there are binding motifs in this region. For this reason, I analyzed the
RNA-Seq data using only genes which were exclusively down-regulated in the samples with the
49
larger deletion. Two of the top three genes identified by this analysis were on chromosome 12.
These genes may be target genes for the enhancer when transcription factors interact with the
region that was only deleted by the outer CRISPR set, although further studies are required to
confirm this hypothesis.
In aim 3, I demonstrated that an epigenetic toggle switch using dCas9-DNMT3A is capable of
methylating the CpGs in an enhancer. However, I used a transient dCas9-DNMT3A. This has
shortcomings in comparison to using cell lines that stably express the toggle switch. Since the
expression is transient, cells cannot be sorted for using FACS and therefore were tested as a
heterogeneous population when performing methylated DNA immunorecipitation (MeDIP). This
means that the results of MeDIP can vary greatly depending on transfection efficiency. Furthermore,
although transient transfection was sufficient to demonstrate the enrichment of methylation at
enhancer 15 in comparison to untransfected cells, I cannot study the impact on gene expression in
this mixed population. In future studies, cells stably expressing the toggle switch should be used to
accurately quantify the increase in methylation resulting from the action of dCas9-DNMT3A and to
identify the effects on gene expression.
The successful results from MeDIP provides a promising outlook for applying
dCas9-CBP:HAT as a toggle switch to activate an enhancer. Enhancer 11 is highly acetylated in
normal colon, indicating the potential of this genomic region to be a functional regulatory element.
50
If successful, these epigenetic toggle switches may be the key towards understanding whether these
two opposing epigenetic forces are drivers or consequences of functional element activation and
repression.
51
CHAPTER V
CONCLUSIONS
Genome wide association studies, CRISPRs, and enhancers are all topics of high popularity in
recent years. However, very few studies focus on characterizing risk enhancers that were identified
through GWAS studies. Furthermore, the use of CRISPRs is generally limited to introducing small
indels to disrupt target genes. To date, there are no studies that apply CRISPRs as a genetic editing
tool for characterizing disease associated risk enhancers that were identified through GWAS.
Studies using epigenetic toggle switches (employing variants of the Cas9 nuclease) to modify
enhancers are even more restricted.
In this study, I have demonstrated the use of CRISPRs to delete kilobase sized regions of the
genome. I have shown that deletion of a risk enhancer results in changes in gene expression,
although confirmation of the target genes will require further studies. Lastly, I have shown
preliminary results to support the idea that the Cas9 variant, dCas-DNMT3A, can force methylation
at an enhancer.
As a part of a larger project to characterize several colorectal cancer risk enhancers, my results
from this study are new and relevant. My project confirms the feasibility of our post-GWAS
analysis pipeline and serves as stepping stone towards developing an epigenetic approach for
colorectal cancer risk enhancer characterization.
52
CHAPTER VI
FUTURE DIRECTIONS
My project revealed many interesting results that are worthy of further investigation. In aim
one and two, I confirmed that enhancer deletion using CRISPRs is a feasible method to identify
candidate genes that are regulated by the enhancer. However, it is necessary to determine whether
these candidate genes are direct targets of the enhancer. This can be achieved by using circularized
chromatin conformation capture (4C), a technique that can detect DNA looping. The direct targets
of the risk enhancer should also be further characterized to determine their contribution towards
colorectal cancer. Furthermore, the CRISPR enhancer deletion technique should be applied to
characterize all of the risk enhancers within the DIP2B region. Enhancer deletion results may be
more profound if more than one enhancer is deleted in a single clone to study the additive effect of
these enhancers.
In aim three of the study, I demonstrated the application of epigenetic toggle switches as an
alternative approach for studying enhancers. Although preliminary results indicate that this method
is feasible, both dCas9-DNMT3A and dCas-CBP:HAT constructs were transiently expressed in my
experiments. Creating stable cell lines for dCas-DNMT3A and dCas-CBP:HAT will likely result in
more robust and long-term changes in the epigenome at the target region. The effects on gene
expression upon epigenetic modification of an enhancer should be compared with results from
53
enhancer deletion to determine which method is most effective.
54
REFERENCES
An integrated encyclopedia of DNA elements in the human genome: The ENCODE project
consortium. (2012). Nature, 489(7414), 57.
Andersson, R., Suzuki, T., Ntini, E., Arner, E., Valen, E., Li, K., . . . FANTOM Consortium. (2014).
An atlas of active enhancers across human cell types and tissues. Nature, 507(7493), 455-455.
Doi:10.1038/nature12787
Blattler, A., Yao, L., Witt, H., Guo, Y ., Nicolet, C. M., Berman, B. P., & Farnham, P. J. (2014).
Global loss of DNA methylation uncovers intronic enhancers in genes showing expression
changes. Unpublished Manuscript.
Burt, R. W. (2000). Colon cancer screening. UNITED STATES: Elsevier Inc.
Doi:10.1053/gast.2000.16508
Calo, E., & Wysocka, J. (2013). Modification of enhancer chromatin: What, how, and why?
Molecular Cell, 49(5), 825-837. Doi:10.1016/j.molcel.2013.01.038
Feingold, E., Sekinger, E., Urban, A., Cheng, J., Hirsch, H., Ghosh, S., . . . The ENCODE Project
Consortium. (2004). The ENCODE (encyclopedia of DNA elements) project. Science,
306(5696), 636-640. Doi:10.1126/science.1105136
Goel, A., & Boland, C. R. (2012). Epigenetics of colorectal cancer. Gastroenterology, 143(6),
1442-1442. Doi:10.1053/j.gastro.2012.09.032
Hsu, P. D., Lander, E. S., & Zhang, F. (2014). Development and Applications of CRISPR-Cas9 for
Genome Engineering. Cell, 157(6), 1262-1278.
Jean, D., Tellez, C., Huang, S., Davis, D. W., Bruns, C. J., mcconkey, D. J., . . . Bar-Eli, M. (2000).
Inhibition of tumor growth and metastasis of human melanoma by intracellular anti-ATF-1
single chain fv fragment. Oncogene, 19(22), 2721-2730. Doi:10.1038/sj.onc.1203569
Kent, W. J., Sugnet, C. W., Furey, T. S., Roskin, K. M., Pringle, T. H., Zahler, A. M., &
55
Haussler, D. (2002). The human genome browser at UCSC. Genome Research, 12(6), 996-1006.
Doi:10.1101/gr.229102
Lander, E. S., fitzhugh, W., Frazier, M., Funke, R., Gage, D., Harris, K., . . . International Human
Genome Sequencing Consortium. (2001). Initial sequencing and analysis of the human genome.
Nature, 409(6822), 860-921. Doi:10.1038/35057062
Lister, R., Ngo, Q., Edsall, L., Antosiewicz-Bourget, J., Stewart, R., Ruotti, V., . . . Ye, Z. (2009).
Human DNA methylomes at base resolution show widespread epigenomic differences. Nature,
462(7271), 315-322. Doi:10.1038/nature08514
Mali, P., Yang, L., Esvelt, K. M., Aach, J., Guell, M., dicarlo, J. E., . . . Church, G. M. (2013).
RNA-guided human genome engineering via Cas9. Science (New York, N.Y.), 339(6121),
823-826. Doi:10.1126/science.1232033
Pancione, M., Remo, A., & Colantuoni, V. (2012). Genetic and epigenetic events generate multiple
pathways in colorectal cancer progression. Pathology Research International, 2012, 509348.
Doi:10.1155/2012/509348
Peters, U., Zanke, B. W., Lemire, M., Rangrej, J., Vijayaraghavan, R., Chan, A. T., . . . Gallinger, S.
(2012). Meta-analysis of new genome-wide association studies of colorectal cancer risk.
Human Genetics, 131(2), 217-234. Doi:10.1007/s00439-011-1055-0
Rozen, S. & Skaletsky, J. H. (1998). Primer3. Code available at
http://www-genome.wi.mit.edu/genome_software/other/primer3.html.
Segal, D. J., & Meckler, J. F. (2013). Genome engineering at the dawn of the golden age. Annual
Review of Genomics and Human Genetics, 14, 135-158.
Doi:10.1146/annurev-genom-091212-153435
Siegel, R., desantis, C., & Jemal, A. (2014). Colorectal cancer statistics, 2014. CA: A Cancer
Journal for Clinicians, 64(2), 104-117. Doi:10.3322/caac.21220
Thu, K. L., Vucic, E. A., Kennett, J. Y., Heryet, C., Brown, C. J., Lam, W. L., & Wilson, I. M.
(2009). Methylated DNA immunoprecipitation. Journal of Visualized Experiments : jove, (23)
56
Visscher, P. M., Brown, M. A., mccarthy, M. I., & Yang, J. (2012). Five years of GWAS discovery.
American Journal of Human Genetics, 90(1), 7-24. Doi:10.1016/j.ajhg.2011.11.029
Weng, Y., Huang, T. H., & Yan, P. S. (2009). Methylated DNA immunoprecipitation and
microarray-based analysis: Detection of DNA methylation in breast cancer cell lines. Methods
in Molecular Biology (Clifton, N.J.), 590, 165.
Winnepenninckx, B., Debacker, K., Ramsay, J., Smeets, D., Smits, A., fitzpatrick, D. R., & Kooy,
R. F. (2007). CGG-repeat expansion in the DIP2B gene is associated with the fragile site
FRA12A on chromosome 12q13.1. American Journal of Human Genetics, 80(2), 221-231.
Doi:10.1086/510800
Yao, L. J., Tak, Y . G., Berman, B. P., & Farnham, P. J. (2014). Functional Annotation of Colon
Cancer Risk SNPs. Nature Communications, in-press.
Zhang, X., Bailey, S. D., & Lupien, M. (2014). Laying a solid foundation for manhattan--'setting
the functional basis for the post-GWAS era'. Trends in Genetics : TIG, 30(4), 140-149.
Doi:10.1016/j.tig.2014.02.006
Abstract (if available)
Abstract
Colorectal cancer (CRC) is among the top three cancer incidence and death rates in the United States. To further understand the disease, it is essential to develop a mechanistic comprehension of cancer initiation and progression. To characterize a CRC associated risk enhancer in the DIP2B intron, I reversed the presence of this enhancer using CRISPRs (Clustered Regularly Interspaced Short Palindromic Repeats) by deleting the enhancer from a colon cancer cell line. I subsequently performed RNA‐Seq to identify the candidate genes regulated by the risk enhancer. Furthermore, I demonstrated the feasibility of using modified Cas9 proteins as epigenetic toggle switches for enhancer repression and activation.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Functional characterization of colon cancer risk-associated enhancers: connecting risk loci to risk genes
PDF
Functional analysis of a prostate cancer risk enhancer at 7p15.2
PDF
Functional characterization of colorectal cancer GWAS loci
PDF
Functional characterization of a prostate cancer risk region
PDF
The relationship between DNA methylation and transcription factor binding in colon cancer cells
PDF
Breast epithelial cell type specific enhancers and functional annotation of breast cancer risk loci
PDF
Functional DNA methylation changes in normal and cancer cells
PDF
Identification and characterization of cancer-associated enhancers
PDF
Understanding prostate cancer genetic susceptibility and chromatin regulation
PDF
Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics
PDF
Functional role of chromatin remodeler proteins in cancer biology
PDF
Characterization of a new chromobox protein 8 (CBX8) antagonist in a model of human colon cancer
PDF
Identification and characterization of the enhancer elements for lymphatic-specific expression of Prox1
PDF
Application of tracing enhancer networks using epigenetic traits (TENET) to identify epigenetic deregulation in cancer
PDF
Using epigenetic toggle switches to repress tumor-promoting gene expression
PDF
etd-GaddisMala-3657~36
PDF
etd-GaddisMala-3657~90
PDF
etd-GaddisMala-3657~58
PDF
etd-GaddisMala-3657~92
PDF
etd-GaddisMala-3657~28
Asset Metadata
Creator
Hung, Yuli
(author)
Core Title
Functional characterization of colon cancer risk enhancers
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Biochemistry and Molecular Biology
Publication Date
07/15/2014
Defense Date
06/17/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
colon cancer,CRC,CRISPR,DIP2B,enhancers,OAI-PMH Harvest
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Farnham, Peggy J. (
committee chair
), Coetzee, Gerhard (Gerry) A. (
committee member
), Tokes, Zoltan A. (
committee member
)
Creator Email
joyyhung@gmail.com,yulihung@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-441213
Unique identifier
UC11286901
Identifier
etd-HungYuli-2695.pdf (filename),usctheses-c3-441213 (legacy record id)
Legacy Identifier
etd-HungYuli-2695.pdf
Dmrecord
441213
Document Type
Thesis
Format
application/pdf (imt)
Rights
Hung, Yuli
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
Tags
colon cancer
CRC
CRISPR
DIP2B
enhancers