Citation
The regulatory role of GSK-3 in DNA and RNA Methylation

Material Information

Title:
The regulatory role of GSK-3 in DNA and RNA Methylation
Creator:
Egelston, Jennifer Nicole ( author )
Language:
English
Physical Description:
1 electronic file (107 pages) : ;

Subjects

Subjects / Keywords:
Proteins -- Methylation ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
In addition to the regulation of DNA methylation by Gsk-3, we continue to investigate whether Gsk-3 regulates the levels of mRNA methylation of adenosines (referred to as m6A). m6A is the most abundant mRNA modification and until recently, the functional significance remained elusive. However, recent findings have suggested that the m6A tag is a mechanism directing mRNA degradation. Mettl3 has been shown to be the methyltransferase enzyme responsible for adding m6A to mRNA, while FTO has been shown to be the demethylase in this reversible process. Recent studies have established a role for m6A in maintaining pluripotency in embryonic stem cells; reduced m6A levels lead to longer mRNA half-lives of pluripotency-associated factors. Due to the phenotypic similarities between Gsk-3 DKO and Mettl3 KO ESCs and their inability to differentiate, we have investigated whether Gsk-3 is regulating m6A levels. Specifically, we have found that deletion of Gsk-3 results in high levels of FTO. The discovery that the m6A modification is controlled by Gsk-3, a multi-faceted regulator involved in numerous pathways, would have profound consequences in understanding the regulatory mechanisms behind the balance between pluripotency and differentiation in ESCs.
Thesis:
Thesis (M.S.) - University of Colorado Denver.
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader.
General Note:
Department of Integrative Biology

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
946450268 ( OCLC )
ocn946450268
Classification:
LD1193.L45 2015m E44 ( lcc )

Downloads

This item has the following downloads:


Full Text
THE REGULATORY ROLE OF GSK-3 IN DNA AND RNA METHYLATION
by
JENNIFER NICOLE EGELSTON
B.S., California Polytechnic State University, 2013
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Integrative Biology Program
2015


2015
JENNIFER NICOLE EGELSTON
ALL RIGHTS RESERVED
11


This thesis for the Master of Science degree by
Jennifer Nicole Egelston
has been approved for the
Integrative Biology Program
by
John Swallow, Chair
Christopher Phiel, Advisor
Amanda Charlesworth
Xiaojun Ren
November 19th, 2015


Egelston, Jennifer Nicole (M.S., Biology)
The Role of Gsk-3 in DNA and RNA Methylation
Thesis directed by Assistant Professor Christopher Phiel.
ABSTRACT
Gsk-3 (Glycogen synthase kinase) is a key protein kinase that has a prominent
role in multiple intracellular pathways. Recently, Gsk-3 activity was found to be
important for the regulation of DNA methylation in mouse embryonic stem cells (ESCs)
when Gsk-3a and Gsk-3(S are genetically deleted. In addition, it was observed that the
addition of lithium, a known inhibitor of Gsk-3 and a common bipolar disorder (BPD)
treatment, also reduces DNA methylation. These studies led us to ask whether DNA
methylation was changed in cells from patients with BPD. This led to the identification of
an uncharacterized long non-coding RNA (IncRNA), LINC00486, whose DNA
methylation patterns are different in BPD patients compared to their unaffected siblings.
Three distinct transcripts have been annotated for LINC00486. Our goal was to
functionally and biochemically characterize these novel IncRNAs. Cytoplasmic and
nucleic localization experiments determined that LINC00486 is present in the nucleus,
which is consistent with other characterized IncRNAs and their role in transcriptional
regulation. RNA immunoprecipitations (RIP) have identified physical interactions
between LINC00486 and both Poly comb Repressive Complexes (PRC1 and PRC2),
which play roles in chromatin remodeling to induce transcriptional repression.
Interestingly, PRC1 subunits are associated with pluripotency maintenance in ESCs.
IV


These findings suggest LINC00486 may play a role in transcriptional silencing which in
turn could sustain stem cell pluripotency.
In addition to the regulation of DNA methylation by Gsk-3, we continue to
investigate whether Gsk-3 regulates the levels of mRNA methylation of adenosines
(referred to as m6A). m6A is the most abundant mRNA modification and until recently,
the functional significance remained elusive. However, recent findings have suggested
that the m6A tag is a mechanism directing mRNA degradation. Mettl3 has been shown to
be the methyltransferase enzyme responsible for adding m6A to mRNA, while FTO has
been shown to be the demethylase in this reversible process. Recent studies have
established a role for m6A in maintaining pluripotency in embryonic stem cells; reduced
m6A levels lead to longer mRNA half-lives of pluripotency-associated factors. Due to the
phenotypic similarities between Gsk-3 DKO and Mettl3 KO ESCs and their inability to
differentiate, we have investigated whether Gsk-3 is regulating m6A levels. Specifically,
we have found that deletion of Gsk-3 results in high levels of FTO. The discovery that
the m6A modification is controlled by Gsk-3, a multi-faceted regulator involved in
numerous pathways, would have profound consequences in understanding the regulatory
mechanisms behind the balance between pluripotency and differentiation in ESCs.
The form and content of this abstract are approved. I recommend its publication.
Approved: Christopher Phi el
v


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.....................................................1
Gsk-3 Overview...................................................1
Specific Aims....................................................4
Characterizing a Novel LncRNA, LINC00486......................4
Involvement of Gsk-3 in Stem Cell Pluripotency................4
II. CHARACTERIZING A NOVEL LNCRNA, LINC00486.........................6
Introduction.....................................................6
Gsk-3 Regulates DNA Methylation...............................6
Non-Coding RNA Overview.......................................7
Role of Long Non-Coding RNA in Gene Regulation................8
LncRNAs Associated with Poly comb Repressive Complexes........9
Materials and Methods...........................................13
Over-Expression of LINC00486 Transcripts in Mammalian Cells..13
Gibson Assembly Cloning of LINC00486 Transcripts 1-3 and
LncRNA HOTAIR into pCAGEN..............................13
N2A and HEK293T Mammalian Cell Transfection Using PEI..15
mESC Transfection Using PEI............................16
RNA Isolation..........................................16
cDNA Synthesis.........................................17
TA Cloning.............................................17
RT-PCR.................................................17
vi


Thermocycler Program......................................17
qPCR......................................................17
RNA Immunoprecipitation.........................................18
Biotinylation of Oligos...................................18
RNA Pull-Down with Biotinylated Oligos....................18
EZH2 Over-Expression and Immunoprecipitation..............20
Cbx Protein Over-Expression and Immunoprecipitation.......22
Results.............................................................22
Over-Expression of LINC00486 in Mammalian Cells.................23
Endogenous Expression In Human Cells and Tissues................28
RNA Immunoprecipitation.........................................39
III. INVOLVEMENT OF GSK-3 IN STEM CELL PLURIPOTENCY......................44
Introduction........................................................44
Pluripotency in Mouse Embryonic Stem Cells......................44
Effect of LIF on Stem Cell Pluripotency.........................45
Effect of Gsk-3 Inhibition on Stem Cell Pluripotency............46
m6A mRNA Modification...........................................48
Materials and Methods...............................................53
RNA Immunoprecipitation and ELISA...............................53
m6A Immunoprecipitation...................................53
m6A ELISA Using Qubit.....................................54
m6A ELISA (Based Off Epigentek Protocol)..................56
Observing Protein Expression....................................57
vii


Western Blotting
57
Results............................................................58
Pluripotency Gene Expression in Immunoprecipitated m6A RNA......59
Effect of LIF on mESC Protein and Gene Expression...............62
Effect of GSk-3 Inhibition on FTO Protein Expression............68
Effect of Vitamin C on Pluripotency.............................70
IV. DISCUSSION.........................................................75
Conclusions........................................................75
Characterizing a Novel LncRNA, LINC00486........................75
Functional Characterization..............................75
Biochemical Characterization.............................78
Involvement of Gsk-3 in Stem Cell Pluripotency..................80
Final Words.....................................................83
V. FUTURE DIRECTIONS..................................................84
REFERENCES...............................................................86
APPENDIX.................................................................92
A. Equations..........................................................92
B. Buffer Reagents....................................................92
C. Cellular Mediums...................................................92
D. Supplemental Data.................................................93
viii


LIST OF TABLES
Table
1. Tissues with robust transcript expression and designated tissue controls.......34
2. Summarized results of Cbx protein immunoprecipitations........................43
IX


LIST OF FIGURES
Figure
1. Verification of Successful Cloning ofLINC00486 Transcripts 1-3...............26
2. Verification of Over-Expression and Localization of LINC00486 Transcripts 1-3 in
N2A Cells.......................................................................27
3. Endogenous LINC00486 Transcript Expression in HEK293T Cells..................32
4. Detecting Presence of LINC00486 Transcript Expression in Human Tissue........33
5. Quantification of Endogenous LINC00486 Transcript Expression in Human Tissues
and Cells.......................................................................35
6. Endogenous LTBP1 and TTC27 Gene Expression in HEK293T and N2A Cells..........36
7. Effect of Gsk-3 Inhibition on Endogenous LINC00486 Transcript Expression in
111X293 I' Cells................................................................37
8. LINC00486 Transcript and Flanking Gene Expression in BPD Patient and Unaffected
Sibling.........................................................................38
9. EZH2 Immunoprecipitation of LINC00486 Transcripts 1-3......................42
10. Cbx Protein Immunoprecipitation of LINC00486 Transcripts 1-3................43
11. m6A mRNA Gene Expression in WT and Gsk-3 DKO ESCs......................61
12. FTO Protein Expression in WT and Gsk-3 DKO ESCs Cultured in Complete ESC
Media in the Presence of LIF....................................................61
13. FTO Expression in ESCs Cultured in Complete ESC Media in the Absence of LIF for
14 Days.........................................................................65
14. Pluripotency of Gene Expression in ESCs Grown in the Absence of LIF for 14
Days............................................................................66
15. Optimizing ESC Media Conditions for Detection of Robust FTO Protein
Expression......................................................................67
x


16. Effect of Gsk-3 Inhibition on FTO Protein Expression in WT ESCs.............69
17. Morphology of WT, Gsk-3 DKO and pi 10* ESCs grown in the presence and absence
of Vitamin C....................................................................72
18. Quantified Gene Expression in Gsk-3 DKO ESCs cultured in the absence of LIF with
the addition of Vitamin C.......................................................73
19. Quantified Gene Expression in WT and pi 10* ESCs Cultured in the Absence of LIF
with the Addition of Vitamin C....................................................74


LIST OF ABBREVIATIONS
2i Dual inhibitor cocktail
3 UTR 3 Untranslated Region
3188 Bipolar Disorder Patient
3189 Unaffected Sibling
AGi Ascorbic Acid and a Gsk-3 P inhibitor supplement
Alkbh5 AlkB Homolog 5
BMP2 Bone Morphogenetic Protein 2
BMP4 Bone Morphogenetic Protein 4
BPD Bipolar Disorder
ChIRP Chromatin Immuno RNA Precipitation
ENCODE Encyclopedia of DNA Elements
Erkl Extracellular Regulated Kinase 1
Erk2 Extracellular Regulated Kinase 2
ESC Embryonic Stem Cells
Esrrb Estrogen-Related Receptor Beta
EZH2 Enhancer of Zeste 2
FGF4 Fibroblast Growth Factor 4
FTO Fat and Obesity Gene
Gsk-3 Glycogen Synthase Kinase
DKO Double Knockout
H2AK119 Histone 2A Lysine 119
H3K27me3 Trimethylation on Histone 3 on Lysine 27
Xll


HEK293T Human Embryonic Kidney Cells\
hESC Human Embryonic Stem Cells
HOTAIR HOX Transcript Antisense RNA
iPSC Induced Pluripotent Stem Cells
Klf4 Kriippel-like factor 4
LC MS/MS Liquid Chromatography Tandem Mass Spectrometry
LiCl Lithium
LIF Leukemia Inhibitory Factor
lincRNA Long Intergenic Non-coding RNA
IncRNA Long Non-coding RNA
m6A N6-methyladensoise
m6A-seq N6-methyladensoise sequencing
MEK Mitogen Activated and Extracellular-Regulated Kinase Inase
mESC Mouse Embryonic Stem Cells
Mettl3 Methyl Transferase-like 3
miRNA MicroRNA
mRNA messenger RNA
N2A Neuroblastoma Cells
ncRNA Non-coding RNA
NSC Neural Stem Cells
pi 10* Myristoylated form of pi 10 subunit
PDK1 Phosophoinositide-Dependent Kinase 1
PDK2 Phosophoinositide-Dependent Kinase 2
Xlll


PEI Polyethylenimine
PI(3,4,5)P3 Phosphatidylinositol-3, 4, 5-triphosphate
PI3K Phosphatidylinositol-3-Kinase
PIC Polymerase II Preinitiation Complex
PRC1 Polycomb Repressive Complex 1
PRC2 Polycomb Repressive Complex 2
qPCR Quantitative PCR
RIP RNA Immunoprecipitation
rRNA Ribosomal RNA
snoRNA Small Nucleolar RNA
snRNA Small Nuclear RNA
STAT3 Signal Transduced and Activator of Transcription 3
Tcfi T-cell factor 3
TdT Terminal Deoxynucleotidyl Transferase
tRNA Transfer RNA
XIV


CHAPTER I
INTRODUCTION
Gsk-3 Overview
Glycogen synthase kinase (Gsk-3) is a constitutively active serine/threonine
protein kinase that is a negative regulator of signal transduction through means of
phosphorylation of downstream proteins (Popkie et al., 2010). Gsk-3 has been defined as
the combined functional redundant activity of both Gsk-3a and Gsk-3 (S isoforms (Popkie
et al., 2010). Both isoforms have a preferred phosphorylation motif (S/T-X-X-X-S/T) and
usually only phosphorylate substrates that are either non-primed (unphosphorylated) or
are primed (have been previously phosphorylated by another kinase) (Fiol, 1987).
Recognition of the same consensus motif suggests that both Gsk-3 isoforms perform
overlapping functions through the phosphorylation of the same proteins, which in turn
allows them to regulate the same signaling pathways (Doble, Patel, Wood, Kockeritz, &
Woodgett, 2007).
Gsk-3 has a large involvement in many signal transduction pathways and
therefore its role in diseases has been observed (Grimes, 2001). One of the most widely
studied roles of Gsk-3 is in the Wnt pathway, in which regulation of Gsk-3 plays an
essential role in many different facets of embryonic development, including brain
development (Caricasole et al., 2003). In addition, Gsk-3 also plays a pivotal role in the
insulin pathway (Frame, 2001). However, while Gsk-3 regulates many pathways, the
pathways do not activate one another (S. S. Ng et al., 2009). Research has shown that
Gsk-3 is involved in a number of intracellular processes such as cell apoptosis, cell
proliferation, gene expression, and regulates differentiation of embryonic stem cells and
1


neural progenitors (Jope, 2003). Finally, altered Gsk-3 activity has been shown to play an
important role in many human diseases such as bipolar disorder, schizophrenia, diabetes,
and Alzheimers disease (Popkie et al., 2010).
Upstream of Gsk-3 in the insulin-signaling pathway lies a lipid kinase,
Phosphatidylinositol-3-Kinase (PI3K) (Osaki, 2004). PI3K is involved in numerous
physiological processes including development and proliferation. Inhibition of PI3K
resulted in the inability to retain pluripotency in mouse embryonic stem cells (mESCs)
(Welham et al., 2011). In response to growth factor stimulation, PI3K generates
Phosphatidylinositol-3, 4, 5-triphosphate (PI (3,4,5) P3), which is essential for activation
of Akt by phosophoinositide-dependent kinase 1 (PDK) and PDK2 (Fang et al., 2000).
Activated Akt phosphorylates Gsk-3 at an N-terminal serine (S21 on Gsk-3 a, S9 on Gsk-
3(3), causing rapid inhibition of Gsk-3 (Popkie et al., 2010). Gsk-3 inhibition results in
numerous downstream effects, including changes in gene and protein expression in
various pathways, epigenetic alterations, and increases in stem cell pluripotency and self-
renewal.
In addition to the insulin-signaling pathway, Gsk-3 plays an important role in the
canonical Wnt signaling pathway (Wray et al., 2011). Many researchers believe this
pathway is the key regulating system in stem cell fate determination due to the
modulation of Gsk-3. In fact, a cocktail of 2 small molecules (termed 2i) is used to
promote stem cell pluripotency (Ying et al., 2008). One of these molecules is a Gsk-3
inhibitor. Gsk-3 is a negative regulator of the Wnt pathway through the phosphorylation
of P-catenin. This phosphorylation results in the ubiquitinylation and proteolysis of P-
catenin (Katoh & Katoh, 2007). However, when Gsk-3 is inhibited, there is an
2


accumulation of P-catenin, which associates with different transcription factors, such as
Tcfi (T-cell factor 3), to activate transcription (Wray et al., 2011). Therefore, inhibition
of Gsk-3 allows for the activation of the Wnt pathway (Hedgepeth et al., 1997). Tcf3 is
predominately found in stem cells and acts as transcription repressor. However, the
accumulation of P-catenin directly interacts with Tcf3 to release this suppression at Tcfi
target genes that are additionally bound by pluripotency factors. This ablation of
inhibition is believed to maintain pluripotency (Wray et al., 2011). A debate between
which pathway, Wnt or insulin, controls the ability for stem cells to stay pluripotent is an
ongoing quest among scientists. However, a possible third mechanism could be that these
two pathways work in parallel when Gsk-3 is inhibited. Only ongoing in-depth Gsk-3
research will help to determine which of these ideas is the most accurate.
In order to functionally study Gsk-3, knocking down or inhibiting Gsk-3 is one
method to provide insight into the role of Gsk-3 in both DNA and RNA epigenetics.
Numerous small molecule inhibitors are available, including insulin (through
phosphorylation by Akt), lithium (LiCl), and SB-415,286. LiCl inhibits Gsk-3 directly by
displacing magnesium, a required cofactor (Klein, 1996). In contrast, SB-415,286 inhibits
through an ATP competitive manner (Coghalin, 2000). While there are many small
molecule inhibitors available to study Gsk-3 function, a drawback is the potential for off
target effects (Phiel, 2001). Therefore, creating a more permanent inhibition of Gsk-3
provides a more confident assessment of Gsk-3. A myristolyated form of a PI3K subunit,
pi 10a (termed pi 10*), promotes a constitutively active insulin pathway, resulting in
constant inhibition of Gsk-3 (Popkie et al., 2010). Additionally, the genetic deletion (also
called knockout) of both Gsk-3 isoforms in mESCs (Gsk-3 double knockout; Gsk-3
3


DKO) allows for numerous downstream experiments permitting the functional study of
Gsk-3.
Specific Aims
The goal of this research project is to functionally and biochemically characterize
a novel long non-coding RNA, LINC00486, which was discovered from analysis of
hypermethylation patterns in BPD patients. This will be accomplished by determining in
which cell types and tissues each human transcript of LINC00486 is expressed.
Additionally, identifying protein-RNA interactions will help to determine the function of
LINC00486.
The second part of this project is to examine the relationship between Gsk-3 and
stem cell pluripotency by a novel mechanism: the regulation of the m6A modification on
mRNA. We hypothesize the m6A tag is regulated by Gsk-3 through controlling the
protein levels FTO, a demethylase. Inhibition of Gsk-3 would result in accumulated FTO
protein levels, resulting in reduced m6A levels and maintenance of stem cell pluripotency.
Identifying the regulatory role of Gsk-3 will provide a mechanistic understanding of stem
cell differentiation as well as directly link m6A modifications with Gsk-3 function.
Characterizing a Novel LncRNA, LINC00486
Aim #1: Characterizing the expression of endogenous LINC00486 transcripts
Aim #2: Biochemical characterization of LINC00486 transcripts
Involvement of Gsk-3 in Stem Cell Pluripotency
Aim #1: Determine if the effect of Gsk-3 on stem cell pluripotency is through the
regulation of mRNA methylation.
4


Aim #2: Investigate if Vitamin C and Gsk-3 inhibition work synergistically to promote
stem cell pluripotency.
5


CHAPTER II
CHARACTERIZING A NOVEL LNCRNA, LINC00486
Introduction
Gsk-3 Regulates DNA Methylation
Gsk-3 activity has been observed to regulate DNA methylation through altered
expression of DNA methyltransferase Dnmt3a2. This was confirmed by knocking out
both Gsk-3 isoforms (Gsk-3 DKO) in mouse ESCs, resulting in the reduction of DNA
methylation and altered gene expression (Meredith et al., 2015; Popkie et al., 2010).
Lithium mimics this reduced Gsk-3 activity in two ways: direct inhibition and indirect
inhibition by increasing inhibitory phosphorylation of Gsk-3 (Jope, 2003). These two
effects work in synergy to influence Gsk-3 regulated functions. Interestingly, lithium is
commonly used as a mood stabilizer treatment for patients with bipolar disorder (BPD)
(Phiel, 2001). The exact mechanism behind how lithium serves to treat BPD is unknown,
but the mechanism may occur through Gsk-3 inhibition, which results in the regulation of
transcription factors and a decrease in DNA methylation (Jope, 2003).
Due to the inhibitory effects lithium has on the DNA methylation patterns in Gsk-
3 DKO ESCs and the clinical use of lithium in BPD patients, mouse neural stem cells
(NSCs) were treated with lithium. Similar reductions in DNA methylation were observed
in these cells. These results demonstrated that the effects of lithium on DNA methylation
extend to neural cells, and provoked an interest in investigating if DNA methylation was
altered in patients with BPD. Due to the difficulties of acquiring neural cells from BPD
patients, we began by using lymphoblast cells obtained from two sets of siblings; in each
set, one sibling had BPD and the other sibling was unaffected. Methyl-seq was performed
6


on genomic DNA isolated from lymphoblasts, providing DNA methylation patterns of
the entire genome for each individual. Several loci displayed hypermethylation in both
BPD patients compared to their unaffected siblings. When the lymphoblast cells were
treated with a therapeutic dose of lithium (ImM), results showed hypomethylation at
several loci (data not shown; A. Popkie; Personal Communications). Among these, one
locus had been hypermethylated in the BPD cells. This differentially methylated locus
was where the novel IncRNA, LINC00486, was located. Sequencing data has indicated
LINC00486 is found in human chromosome 2 and is flanked by genes TTC27 and
LTBP1. A mouse homolog has yet to be discovered. Three distinct RNAs are predicted to
be transcribed from the LINC00486 locus, each of different lengths and containing
different exons.
Non-Coding RNA Overview
Recent advances in genome-wide analyses have revealed that roughly 90% of the
human genome is transcribed, yet less than 3% of the genome consists of protein-coding
genes (Wu et al., 2013). The remaining genes are transcribed as noncoding RNAs
(ncRNAs), which resemble mRNA in length and splicing structures yet do not encode
any proteins (Wu et al., 2013). It has been debated whether all of the ncRNA transcripts
are functional due to their low expression levels and low evolutionary conservation.
However, many functional ncRNA have been identified and are commonly known
(Guttman & Rinn, 2012). The ncRNA family consists of housekeeping ncRNAs and
regulatory ncRNAs. The housekeeping ncRNAs include ribosomal (rRNA), small nuclear
(snRNA), small nucleolar (snoRNA), and transfer RNAs (tRNA). Regulatory ncRNAs
are divided into two classes based on size; those less than 200 nucleotides are short/small
7


ncRNAs and encompass microRNAs (miRNAs) while those greater than 200 nucleotides
are long non-coding RNAs (IncRNAs) (Wu et al., 2013). One subgroup within IncRNAs
are lincRNAs (long intergenic non-coding RNAs) which lie in the intergenic regions of
the genome (L. Yang, Froberg, & Lee, 2014). The ENCODE project, an effort to survey
transcription in entirety, identified more than 9000 genomic loci that are transcribed into
IncRNAs in the nucleus of human cells, yet the majority of these IncRNAs have yet to be
characterized (L. Yang et al., 2014). The pursuit to functionally characterize these
IncRNAs in order to determine their biological relevance has been a popular trend in the
epigenetic realm, but many researchers have faced difficulties due to the lack of protein
coding potential of IncRNAs.
Role of Long Non-Coding RNA in Gene Regulation
Over the past decade, thousands of well-expressed IncRNAs have been identified
to be actively transcribed across the human genome and have shown to have
evolutionarily conserved promoter regions and splice sites as well as unique start and
stop codons (Guttman et al., 2009). These findings indicate that IncRNA transcription
occurs independently from neighboring genes (L. Yang et al., 2014). However, several
IncRNAs have shown to maintain enhancer-like functions, which allows them to work in
cis to activate the expression of their neighboring genes (Guttman & Rinn, 2012).
Uniquely, other IncRNAs work in trans in order to regulate distant genes (Chu, Qu,
Zhong, Artandi, & Chang, 2011). LncRNAs have shown to be key regulators of
numerous biological processes including chromatin regulation, dosage compensation,
imprinting, epigenetic inheritance, and stem cell pluripotency (S. Y. Ng, Johnson, &
Stanton, 2012; Tsai et al., 2010; L. Yang et al., 2014).
8


Despite the challenges in functionally characterizing the ongoing list of identified
IncRNAs, an emerging theme is the ability for IncRNAs to regulate gene expression. One
mechanism for modulating gene expression is through chromatin modifications (L. Yang
et al., 2014). Nucleosomes are the packaged unit of genomic DNA wrapped around
histone proteins, which are then organized into a higher structure called chromatin.
Modifications of these histones regulate the transcription of the DNA wrapped around
them. One of the first and most famous IncRNAs to be discovered is Xist, which is
required for X-inactivation that occurs through IncRNA-mediated chromatin
modifications (Guttman & Rinn, 2012). Xist has been shown to recruit chromatin-
modifying complexes, such as the Polycomb Repressive Complex 2 (PRC2), to
appropriate genomic loci in order to negatively regulate gene expression (L. Yang et al.,
2014). Once the genomic region is silenced, the chromatin condenses to cause the
repression of the entire X-chromosome, resulting in X-inactivation (Guttman & Rinn,
2012). The discovery of this IncRNA and its inhibitory process through PRC2 has opened
numerous doors for characterizing future IncRNAs.
LncRNAs Associated with Polycomb Repressive Complexes
Due to the inability to determine the function of IncRNAs from sequence
information alone, discovering the RNA-protein interactions allows researchers to
functionally characterize these RNAs (L. Yang et al., 2014). Roughly 30% of lincRNAs
tested have been shown to physically associate with at least 1 of 12 distinct chromatin-
regulatory complexes in order to guide chromatin proteins to DNA targets as a method to
silence genomic regions (Guttman & Rinn, 2012). Another characterized and highly
studied IncRNA, HOTAIR (HOX Transcript Antisense RNA), has also been identified to
9


target PRC2. This complex, a histone methyltransferase required for epigenetic silencing,
is localized to thousands of mammalian genes (Cifuentes-Rojas, Hernandez, Sarma, &
Lee, 2014; Davidovich, Zheng, Goodrich, & Cech, 2013). PRC2 is comprised of multiple
subunits including the catalytic H3K27 methyltransferase, EZH2 (Enhancer of Zeste 2), a
protein that binds to RNA with high affinity in order to establish a chromatin structure
that causes repression of transcription (Cifuentes-Rojas et al., 2014; L. Yang et al., 2014).
The exact mechanism behind how PRC2 is targeted to specific loci is yet to be
established, but one possibility is through IncRNA recruitment. LncRNAs in both the cis
and trans formation have been shown to interact with PRC2 and tether the complex to the
site of transcription (Cifuentes-Rojas et al., 2014). HOTAIR, a trans-acting RNA, has
been shown to retrieve EZH2 through RNA immunoprecipitation (RIP). These results
suggest HOTAIR recruits PRC2 to chromatin in order to initiate transcriptional silencing
(Tsai et al., 2010). In addition, a RIP of EZH2 demonstrated the interaction between
PRC2 and Xist RNA during X-chromosome inactivation in order to initiate and maintain
the silencing of one of the X chromosomes (L. Yang et al., 2014). Both PRC2 and EZH2
have become a scientific interest due to their physical interactions with other IncRNAs as
well as their clinical significance in disease and cancer (Cifuentes-Rojas et al., 2014).
Similarly to PRC2, polycomb repressive complex 1 (PRC1) establishes strong
transcriptional repression. Many IncRNAs have been recognized as important participants
in PRC1 function (Margueron & Reinberg, 2011). The biochemical mechanism behind
how PRC1 silences transcription is believed to occur through the dissociation of RNA
Polymerase II Preinitiation Complex (PIC) by blocking the recruitment of Mediator, a
required subunit of PIC (Lehmann et al., 2012). PRC1 has also been shown to stabilize
10


chromatin structures and catalyze the monoubiquitylation of lysine 119 on histone H2A
(H2AK119) (Margueron & Reinberg, 2011; Morey et al., 2012). There are five known
Cbx proteins that associate with PRC1 Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8. Different
PRC1 complexes are associated with distinct Cbx proteins, which is dependent upon
target specificity (Morey et al., 2012). Each Cbx protein contains a chromodomain at the
amino-terminal that binds trimethyl-K27 H3 (H3K27me3), which helps to regulate
pluripotency and differentiation of ESCs through unique, non-redundant functions (Ren,
Vincenz, & Kerppola, 2008). The Cbx subunits are believed to stabilize the PRC1
association with chromatin through direct interaction with H3K27me3 (Morey et al.,
2012). Additionally, researchers have shown that PRC1 is dependent on PRC2 such that
the histone trimethylation performed by PRC2 to initiate genomic silencing is necessary
for PRC1 to be recruited to specific genes in order to maintain the silenced state (Ren et
al., 2008). When PRC2 methylates H3K27 through EZH2, this provides a binding site for
PRC1. Once PRC1 binds, it can ubiquitinate H2AK119 for genomic silencing (Lehmann
et al., 2012). Additionally, research has shown that the synergistic effects of the
Polycomb Repressive Complexes are involved in the regulation of stem cell pluripotency.
Specifically, this stem cell regulation is controlled by the different unique functions of the
Cbx proteins, which maintains the balance between self-renewal and differentiation
(Morey et al., 2012). In general, determining the regulatory interactions and the
mechanistic properties between IncRNA and both Polycomb Repressive Complexes will
improve the understanding of the many regulatory roles of IncRNAs as well as aid in
pharmaceutical design to target disease.
11


In addition to regulating gene expression through chromatin modifications,
characterized IncRNAs can function as transcriptional activators by acting as cofactors
that help to enhance kinase activity in order to modify chromatin architecture. LncRNAs
have also been shown to control the co- and post-translational processes including
splicing and translation of mRNA as well as the subcellular localization of proteins (L.
Yang et al., 2014). Another common function of IncRNAs is to act as modular scaffolds
in order to initiate the assembly of multiple protein complexes and recruit these
complexes to the necessary location to regulate gene expression. These regulatory
complexes not only include interactions with protein, but RNA-DNA and RNA-RNA
interactions as well (Guttman & Rinn, 2012). For example, HOTAIR acts as a scaffold
for PRC2 and a H3K4 demethylase complex, LSD1. It has been shown that HOTAIR
binds to both PRC2 and LSD1, bridging these two complexes together in order to target
this complex to silence multiple HOXD genes as well as other genes on select
chromosomes (Tsai et al., 2010). However, the mechanism behind how HOTAIR guides
PRC2 to its target genes is yet to be understood (Guttman & Rinn, 2012). In general, the
formation of these IncRNA-bridges allows IncRNAs to target specific genomic loci with
high specificity (L. Yang et al., 2014). Due to this identified function of IncRNAs, it is
suggested IncRNAs resemble transcription factors when regulating chromatin states
because the target-gene information is highly selective and resides in the RNA (Chu et
al., 2011)
It is evident that determining the functions of identified IncRNAs is challenging
but will only add to the greater understanding of regulatory mechanisms used by
mammalian cells to control gene expression. Despite the progress epigenetic scientists
12


have made in characterizing IncRNAs, a greater and more in depth analysis is needed to
fully understand IncRNAs and their role in human disease.
The goal of this research project is to functionally and biochemically characterize
a novel long non-coding RNA, LINC00486, which was discovered from analysis of
hypermethylation patterns in BPD patients. This will be accomplished by determining in
which cell types and tissues each human transcript of LINC00486 is expressed.
Additionally, identifying protein-RNA interactions, such as an association between
LINC00486 and PRC1 and/or PRC2, as will help to determine the function of
LINC00486. As stated previously, the specific aims of this project include:
Aim #1: Characterizing the expression of endogenous LINC00486 transcripts
Aim #2: Biochemical characterization of LINC00486 transcripts
Materials and Methods
Over-Expression of LINC00486 Transcripts in Mammalian Cells
Gibson Assembly Cloning of LINC00486 Transcripts 1-3 and LncRNA HOTAIR
into pCAGEN
1. Expression vector of interest, pCAGEN, was linearized using 2uL of EcoRl
restriction enzyme, 2uL EcoRl Buffer, and 2ug of circular pCAGEN. To make
the final volume 20uL, dH20 was added to each sample. The digest reaction was
incubated at 37C for 15 minutes.
2. To remove phosphate groups on pCAGEN in order to avoid plasmid
recircularization, luL of antarctic phosphatase was added to the digest reaction.
Samples were incubated at 37C for 15 minutes.
13


3. The effectiveness of the digest reaction was validated by running the digested
products on a 0.8% agarose gel. The circular plasmid was added as a control.
4. After validation of successful digestion, the remaining linear plasmid was purified
using the QIAquick Gel Extraction Kit from QIAGEN.
a. 1 volume of isopropanol was added to remaining digest reaction. Solution
was added to spin column in a collection tube. Tube was spun at 10,000
rpm for 1 minute and the flow through was discarded.
b. The wash was performed by adding 750uL of PE Buffer to the spin
column. Tube was spun at 10,000 rpm for 1 minute and the flow through
was discarded.
c. Empty column was re-spun at 10,000 rpm for 1 minute.
d. The spin column was placed in a fresh collection tube. 30uL Buffer EB
was added to the spin column and the tube was spun at 10,000 rpm for 1
minute to elute purified digest product.
5. The synthesized gBlock gene fragments (IDT) were diluted with TE buffer (pH
7.5) to a concentration of lOng/uL. Solution was homogenized by vortexing and
was placed on ice for immediate use. (Eventually stored at -20C).
6. A 3-fold molar increase of DNA fragments and 50ng of linear plasmid was
combined for the assembly (up to a 6-fold molar increase in fragments could have
been used if the number of fragments increased above 4). DNA fragments were
converted into pmoles using Equation 1.
14


7. The Gibson assembly was prepared on ice by combining 50ng linear vector, 3-
fold molar excess fragments and lOuL Gibson Assembly Master Mix. dH20 was
added to make a total volume of 20uL.
8. Samples were incubated at 50C for 15 minutes (samples that contained more
than 3 fragments were incubated at 50C for 30 minutes). After incubation,
samples were immediately placed on ice or stored at -20C.
9. NEB 5-alpha competent E. cob cells (included in Gibson Assembly Kit) were
transformed with 2uL of assembly reaction. Overnight cultures were made from
colonies on LB plates. Plasmids were purified through mini-preps (GeneJET
Plasmid Miniprep Kit from Thermo Scientific).
10. To confirm if Transcripts 1-3/HOTAIR was successfully cloned into pCAGEN,
restriction digests were performed and products were run out on a 1% agarose gel.
N2A and HEK293T Mammalian Cell Transfection Using PEI
Day 1: Cells were plated at 5.0x10s cells per well in 6-well plates in cell specific
media (Appendix).
Day 2: Media was not changed in 6-well plates. 2.0ug of total DNA (1800ng
pCAGEN-Transcript or HOTAIR + 200ng pMax-GFP) was added to lOOuL Opti-
MEM. In a separate tube, 5uL of PEI was added to 95uL Opti-MEM per
transfection. Tube was tapped to mix and incubated at room temperature for 5
minutes. lOOuL PEI:Opti-MEM was transferred to DNA:Opti-MEM tubes. Tubes
were tapped to mix, centrifuged at lowest speed for collection, and incubated at
room temperature for 20 minutes. Mixture was added to cells drop-wise. Cells
were incubated at 37C with 5% CO2 overnight.
15


Day 3: To determine if transfection was successful, presence of green
fluorescence from pMax-GFP was observed.
mESC Transfection Using PEI
1. ESCs were trypsinized and resuspended in Opti-MEM. IOuL of resuspension was
added to IOuL trypan blue dye. The total number of ESCs was counted using the
Countess.
2. 5 x 105 cells were aliquotted into 2mL microcentrifuge tubes (the same number as
wells to be plated). Tubes were centrifuged for 2 minutes at 1500rpm. Supernatant
was carefully aspirated.
3. Cell pellets were resuspended in 300uL Opti-MEM.
4. 2ug of total DNA was added to the resuspended cells (200ng pMax-GFP and
1800ng plasmid of interest). Cells were mixed by careful pipetting.
5. IOOuL of PEI was added to the resuspended cells. Cells were mixed by pipetting.
6. Cells were incubated for 30 minutes at room temperature. Mixture was pipetted
every 10 minutes.
7. The transfection reaction was plated on freshly prepared gelatin-coated 6-well
plates with warm and complete ESC media.
8. Transfection was incubated overnight and visualized the next day using the
fluorescence microscope.
RNA Isolation
RNA isolated using mirVana mirRNA Isolation Kit by Life Technologies
followed kit protocol. When RNA was to be isolated using Direct-zol RNA Miniprep by
Zymo, 500uL TRIzol was added per well of a 6-well plate to begin lysing process (lmL
16


of TRIzol was added per 10cm plate). RNA isolation followed kit protocol. All isolated
RNA was quantified using Nanodrop 2000 made by ThermoScientific.
cDNA Synthesis
cDNA was synthesized using High Capacity Reverse Transcriptase kit from
Applied Biosystems. Kit protocol was followed and 100ng-2ug of total RNA was used to
synthesize cDNA. The amount of RNA used was kept constant for each set of RNA.
TA Cloning
Sequences of interest were cloned into pCR2.1 vector using T4 DNA ligase. Kit
protocol was followed.
RT-PCR
To amplify sequences of interest RT-PCR was performed by adding lOx Buffer,
5mM dNTPs, 50mM MgCh, 200ng/uL Forward primer, 200ng/uL Reverse primer, 2u/uL
Platinum Taq, 200ng of cDNA and water, for a final volume of 25uL.
Thermocycler Program
*Note: Thermocycler program was optimized by changing annealing temperature
depending on size of fragment desired for amplification.
10 min 94C
[1 min 94C
x30 1 min 65C
1 min_______72C1
7 min 72C
hold 4C
qPCR
To quantify gene expression, biological replicates were created for each cell or
treatment type. RNA was isolated from cells; cDNA was synthesized using 100ng-2ug
RNA. If 2ug RNA was used, the final concentration of synthesized cDNA was lOOng/uL.
17


cDNA was diluted in sterile water to 8.89ng/uL. A master mix for each probe was made
by combining 2x Taqman Mix (with UNG), 20x Taqman probe, and DEPC H20. Master
mix components were multiplied by total number of wells designated for each probe.
After 19.38uL of master mix was added to each designated well, 5.63 uL of diluted
cDNA was added to plate in triplicates. Plates were spun down to collect all liquid at
bottom of plate, and then placed in StepOne qPCR Machine. Data was collected by
running the Comparative Ct (AACt) program, which followed the cycling parameters:
qPCR Program:
2 min 50C x 1 cycle
10 min_______95C x 1 cycle
[15 sec95C
1 min 60C] x30 cycles
RNA Immunoprecipitation
Biotinylation of Oligos
1. Each primer was diluted to lOOuM to make a stock solution.
2. A serial dilution was performed on the stock solution (lOOuM) to create lOuM
and luM solutions. These serial dilutions were important in order to determine
optimal primer concentration.
3. Biotin was diluted to 5uM (prepared fresh every time).
4. Just before use, a master mix was prepared by diluting TdT to 2U/uL in IX TdT
Reaction Buffer and ultrapure water, resulting in a total volume of 5uL. A master
mix was prepared fresh every time and once made, was kept on ice and used
immediately.
5. The labeling reactions were prepared by adding Ultrapure Water, 10X TdT
Reaction Buffer, luM Oligo, 5uM Biotin-11-UTP, and 2U/uL diluted TdT to
18


make a final volume of 50uL (labeling reactions were prepared individually for
each reaction).
6. Reactions were incubated at 37C for 30 minutes.
7. To stop the reactions, 2.5uL 0.2M-0.5M EDTA was added.
8. In fume hood, 50uL choloroforrmisoamyl alcohol was added to each reaction to
extract TdT.
9. Mixtures were vortexed briefly and tubes were centrifuged for 2 minutes at
14,000 rpm.
10. In fume hood, for each reaction the aqueous layer containing biotinylated oligos
was carefully removed and placed in fresh microcentrifuge tube. Biotinylated
oligos were stored at -20C.
RNA Pull-Down with Biotinylated Oligos
1. To determine the optimal concentration of biotinylated oligos, serial dilutions of
the luM biotinylated oligos were prepared: 1:10, 1:100, and 1:1000.
2. Biotinylated oligos were blocked in lOOuL streptavidin magnetic beads (New
England Bio Labs) for 1 hour at room temperature while rotating. For each
transcript RNA, 3 tubes were prepared:
Tube 1: luL of each 1:10 diluted oligo
Tube 2: luL of each 1:100 diluted oligo
Tube 3: luL of each 1:1000 diluted oligo
3. After 1 hour, 2ug of RNA and water were added to each tube to make a final
volume of 25uL (in addition to lOOuL streptavidin beads).
19


4. Tubes were incubated on ice for 30 minutes and then vortexed and briefly spun
down.
5. A magnetic plate was used to remove any clear liquid from tubes without
disturbing the beads.
6. Beads were washed with lOOuL Wash 2/3 from mirVana mirRNA Isolation Kit
(Life Technologies). Beads were vortexed quickly and spun down. The magnetic
plate was used to remove clear liquid from beads.
7. Wash step was repeated 2 more times and clear liquid was removed and
discarded.
8. cDNA was synthesized by combining entire amount of samples (streptavidin
beads + RNA + Biotinylated oligos) to cDNA master mix (Applied Biosystems).
9. Tubes were spun down to collect beads at bottom and 2uL of clear liquid on top
of beads (the cDNA) was used to perform RT-PCR. Transcript specific primers
were used. Non-pulldown cDNA was used as a negative control.
10. PCR products were observed by running out on a 1% agarose gel.
EZH2 Over-Expression and Immunoprecipitation
1. Two sets of N2A 6-well plates were madeone for HA beads and one for control
beads. For each set, 2ug of total DNA was transfected:
1800ng pCMV-HA-EZH2 + 200ng pMax-GFP (negative control)
900ng pCMV-HA-EZH2 + 900ng pCAGEN-Transcripts 1-3 (or
pCAGEN-HOTAIR) + 200ng pMax-GFP
20


2. Biotinylated oligos were blocked in lOOuL streptavidin magnetic beads for 1 hour
at room temperature while rotating. The previously determined optimal dilution
was used for each set of oligos.
3. In cell hood, transfected cells were washed, trypisinized and pelleted.
4. Cytoplasm were isolated by resuspending the cell pellets in 400uL Buffer A in
order to lyse cell (Appendix). Tubes were left on ice for 15 minutes to allow cells
to swell. To each tube, 25uL of IGEPAL was added. Tubes were vortexed
vigorously for 30 seconds and centrifuged at 4C for 30 seconds at 14,000 rpm.
The supernatant was removed and placed in fresh tube (cytoplasmic isolation).
Isolated cytoplasm was stored at -80C.
5. Nuclei was isolated by resuspending remaining pellet in 50uL ice-cold Buffer B
(Appendix). Tubes were rocked vigorously for 15 minutes at 4C. Tubes were
centrifuged at 4C for 5 minutes at 14,000 rpm. Supernatant was removed and
place in fresh tube (nuclei isolation). Isolated nuclei were stored at -80C.
6. Before immunoprecipitation, the negative controls were acquired by placing lOuL
of the nuclei supernatants into fresh tubes. 3X volume of TRIzol was added to
tubes in fume hood and nuclei were lysed by pipetting up and down. Tubes were
stored at -80C.
7. To the remaining nuclei, 200uL of Lysis/Binding Solution (mirVana RNA
Isolation Kit from Life Technologies) was added to each tube. Nucleic pellets
were resuspended by vigorous vortexing.
8. Resuspended nucleic pellets were separated into two groups: HA magnetic beads
and the Control beads. To the designated HA beads tubes, lOuL of HA magnetic
21


beads were added to the nucleic lysate. To the Control beads tubes, lOuL of the
control beads were added to the nucleic lysate. Tubes were incubated overnight at
4C while rotating.
9. After incubating, both the HA and Control tubes were spun down and a magnetic
rack was used to remove liquid from the beads. Liquid was discarded and beads
were washed 3x with sterile H20 in hood.
10. In the fume hood, 200uL of TRIzol was added to each tube. Tubes were vortexed
and incubated at room temperature for 5 minutes. Tubes were the centrifuged at
12,000 x g for 1 minute and the supernatants were transferred to fresh tubes.
11. RNA was isolated (Direct-zol RNA Miniprep by Zymo) and cDNA was
synthesized (Applied Biosystems). RT-PCR was performed using transcript
specific primers and HOTAIR primers for each sample.
Cbx Protein Over-Expression and Immunoprecipitation
For the Anti-Flag and Anti-GFP immunoprecipitations to determine if Transcripts
1-3 were associated with Cbx proteins, the same protocol as the HA-tagged EZH2
immunoprecipitation was followed, except for a few adjustments. For the Anti-Flag IP,
N2A cells were transfected with each pVenus-Cbx plasmid individually. Once cell pellets
were resuspended, Anti-Flag beads (Sigma-Aldrich) were added for an overnight
incubation. No control beads were used.
Similarly for Anti-GFP IP, the same protocol was followed. Except, N2A cells
were transfected with pTripz-MCSl-Cbx2 (full length) and Anti-GFP beads
(ChromoTek) were added to cell suspension. No control beads were used.
Results
22


Over-Expression of LINC00486 Transcripts in Mammalian Cells
In order to determine the function of LINC00486 transcripts, we decided to start
by overexpressing the IncRNAs in mammalian cells. Therefore, we needed to clone the
three primary predicted LINC00486 transcripts into a mammalian expression vector. The
Gibson Assembly cloning technique (Gibson et al., 2009) was used to individually clone
each transcript and IncRNA HOTAIR into a mammalian expression vector, pCAGEN.
Nucleotide fragments, gBlocks, were synthesized to be identical to the sequences of each
transcript using the PrimerQuest Design Tool provided by Integrative DNA Technologies
(IDT). In addition, because of how Gibson Assembly cloning works, 15 bp corresponding
to pCAGEN were added to the ends of each synthetic DNA fragment. With the concerted
action of the three necessary enzymes, an exonuclease, a polymerase, and a DNA ligase,
the fragments were assembled into a single fragment, which was inserted into the
linearized pCAGEN vector. At the end of the single-tube isothermal reaction, a fully
sealed double-stranded DNA molecule was created that contained each transcript inserted
into pCAGEN (Supplemental Data Figure 1) (Gibson, 2009). A restriction digest using
200ng of Gibson product was performed to ensure the correct insertion into pCAGEN
(Figure 1A).
Next, to confirm that we had successfully cloned the LINC00486 transcripts and
HOTAIR, we performed Sanger sequencing. Each correctly sized RT-PCR product was
TA Cloned into pCR2.1 vector using the TA cloning Kit (The Original TA Cloning Kit
from Invitrogen). The ligations were transformed into E. coli competent cells, overnight
cultures were made from the transformation colonies, and mini-preps (Thermo Scientific)
were performed to isolate cloned plasmids. Restriction digests were performed on
23


isolated plasmids for verification of correct band size (Figure IB). Finally, 200ng of
digest products were sent for sequencing using the DNA sequencing service provided by
Eton Biosciences to validate the correct human transcript sequence.
After successful cloning of each individual transcript into the pCAGEN vector, it
was imperative to determine if the plasmids were capable of successful expression in
mammalian cells. Eising polyethylenimine (PEI), a linear cationic polymer as a
transfection agent, each plasmid was transfected into mouse neuroblastoma (N2A) cells
(Bartman, 2014). We chose N2A cells because murine LINC00486 homologues have not
yet been identified, and therefore the human PCR primers we designed would not amplify
anything in mouse cDNA (No Template Control, Figure 2B). This allowed us to take
advantage of the clean background in N2A cells to assess the expression of human
LINC00486 transcripts. We also cloned the IncRNA HOTAIR into pCAGEN to include
in our experiments as a positive control for a representative IncRNA. In addition, pMax-
GFP was cotransfected with each transcript into the cells to act as a positive control for
successful transfection. LINC00486 Transcripts 1-3 and HOTAIR were transfected into
N2A cells individually. Verification of successful transfection was exhibited through
green fluorescence, indicating the nucleus of the N2A cells successfully acquired the
introduced pMax-GFP and transcript (Figure 2A).
To verify that we were able to overexpress the LINC00486 transcripts and
HOTAIR, RNA was isolated from transfected N2A cells (MirVana mirRNA Isolation Kit
from Life Technologies) and cDNA was synthesized (High Capacity cDNA Reverse
Transcription Kit from Applied Biosystems) using 2ug of total isolated RNA. Next, RT-
PCR was performed to verify expression of each transcript using primers specific to each
24


transcript and HOTAIR. Primers were designed using Primer3Plus
(www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) and each transcript primer
pair was tested on the transfected cells to determine the primer specificity (Figure 2B).
The primers that successfully showed expression of the transfected transcript indicated
high primer specificity; we were able to demonstrate that each primer pair was specific to
each LINC00486 transcript (Transcript 1: 298bp; Transcript 2: 129bp; Transcript 3:
179bp). GAPDH primers were used as a control for RNA integrity (GAPDH: 150bp).
RT-PCR products were validated on a 1% agarose gel to confirm correct band size of
transcripts. The no template controls were used as a negative control to verify that the
primers were not contaminated.
Many IncRNAs are found in the nucleus due to their association with transcription
factors to initiate gene silencing (L. Yang et al., 2014). Therefore, we were curious to see
where in the cell LINC00486 was localized. To do so, we isolated the nuclei and
cytoplasm from the cell lysate in order to focus our downstream experiments on a
specific cellular region. RT-PCR revealed the presence of each transcript in both the
nuclei and cytoplasm. As predicted, LINC00486 transcripts were found in the nucleus,
supporting the idea that LINC00486 may associate with different transcription factors.
Additionally, HOTAIR was localized to the nucleus as predicted. Transcript expression
seen in the cytoplasm was most likely due to the over-expression of each transcript. As
expected, the GAPDH control was only localized to the cytoplasm, validating successful
nucleic and cytoplasmic isolation (Figure 2C).
25


A
Qj
*
JTj

rtf

5?
§
.A
&
c
£
.A
£
c
£
'v
.A
$
c
03
*3
<\
.A
£
c
.Or
Qj
<\
<"V
Oo

rv .A
£
C
/t>
.A
£
C
£
.A
£
c
.A
£
c
£
.A
£
c
.A
£
c
,?
p) aiAiiMtowMj "laJtf
10 kb
Xhol and Not I Digest
B
10,000 bp
|imaaisa..
3000 bp
2000 bp
500 bp
250 bp

*
EcoRI Digest
Figure 1. Verification of Successful Cloning of LINC00486 Transcripts 1-
3. (A) Xhol and Notl double restriction digest on pCAGEN-LINC00486-
Transcript 2. An example of ensuring LINC00486 Transcripts were correctly
inserted into pCAGEN through Gibson Assembly Cloning. (B) EcoRI
restriction digest on pCR2.1-LINC00486-Transcript 2 to determine successful
TA cloning. This is an example of the digests performed on each LINC00486
Transcript for cloning validation.
26


N2A Untransfected N2A Transfected
Primers:
^ c? .
c CL c
_£D re
o fir
-* r\
.A .A
.A
py rv)
.A .A
rv fv?
fcj
.A .9.
t*w r\ rr)
^
A .A -A
200 bp
100 bp
^ ^ o o & ^ £r o' ;? ^ ^ ^
co <-> O ^ c £ a; c ccarc-c c a c c c
A £ N2A Transfection: Transcript 1 Transcript 2 Transcript 3 No Template
r-i C\ ry^
A -A -A
Primers:
3 DO bp
200 bp
100 bp
300 bp | M
200 bp
100 bp
N2A Transfection:
$ t-Z
*S) 'S) <> 'i
c c c 25
fO [0 o
£ 3:
$
3 &
A
6 3:
<% Q
c <5r
ro <7-
O A
'-V
A
£
03
4-j
A
£ ;?
£ £
L? ^ -
b cS ^r
-V t\
A A
£
6 "T
e? ,c ac-
Untransfected
Transcript Transcript Transcript Hotair
12 3
fO
Nuclei
Cytoplasm
Figure 2. Verification of Over-Expression and Localization of LINC00486
Transcripts 1-3 in N2A cells. (A) After co-transfecting N2A cells with individual
LINC00486 transcripts and pMaxGFP, successful transfection is validated by
observing green fluorescence in cells. (B) RT-PCR on over-expressed LINC00486
Transcripts 1-3 throughout entire cell and (C) in nuclei and cytoplasm using
transcript specific primers (Expected sizesTranscript f: 298bp; Transcript 2:
129bp; Transcript 3: 179bp; HOTAIR: 250bp; GAPDH: 150bp. HOTAIR primers
were used as a positive control for a representative IncRNA; GAPDH primers
were used as a control for RNA integrity).
27


Endogenous Expression in Human Cells and Tissues
To determine if LINC00486 Transcripts 1-3 and HOTAIR were endogenously
expressed in human cells, RNA was isolated from untransfected human embryonic
kidney (HEK293T) cells (Life Technologies) and cDNA was synthesized using 2ug total
RNA (Applied Biosystems). Using the transcript-specific primers, RT-PCR was
performed to observe endogenous transcript expression in untransfected HEK293T cells.
Only the Transcript 3 primer showed a single band at the correct size (179bp), indicating
specificity to endogenous Transcript 3 expression (Figure 3 A). Due to the lack of
detection of Transcripts 1 and 2 at the correct sizes (Figure 3 A), new primers of varying
sizes were designed to overlap different exons in each transcript (to prevent the detection
of genomic DNA). RT-PCR was performed on human cDNA using the newly designed
primers. RT-PCR products were ran out on a 1.7% agarose gel (Figure 3B, Transcript 1
not included). The primers with the most robust expression specific to each transcript
were used to look at transcript expression in different human tissues (Figures 4A-C).
After confirming presence of endogenous transcript expression in cultured human
cells, we were curious to see the endogenous expression pattern in human tissues. A
panel of human tissue RNA (Human Total RNA Master Panel II) was obtained from
Clontech Laboratories. cDNA was synthesized from 2ug total RNA for each sample, and
standard RT-PCR was performed. Each transcript showed specific levels of expression in
different tissues. The tissues that showed no expression were chosen as controls for
downstream experiments (Table 1).
Next, we wanted to specifically quantify the expression of each transcript in the
same human tissues by performing quantitative PCR (qPCR). However, pre-made
28


Taqman assays for each transcript were not available for purchase. Therefore, we
designed our own Taqman probes using the sequences of the transcript primers
previously shown to have robust and specific expression in human tissues (Figures 4A-
C). Using these probes, qPCR was performed to quantify transcript expression in each
tissue. The control tissue used for each transcript probe was the tissue lacking transcript
expression seen in the previous RT-PCR results (Table 1). This selected tissue was held
consistent for all tissues while looking at the expression of each transcript. Transcript 1
revealed the highest expression in testes and salivary gland, Transcript 2 in the liver and
salivary gland and Transcript 3 in testes and placenta (Figure 5A). To confirm accurate
results, the qPCR products were run out on a 1.7% agarose gel. Absence of primer dimers
on agarose gel validated specific transcript expression (Figure 5B-D).
After determining the tissues with high transcript expression, we noticed a trend
of relatively high transcript expression in both the testes and placenta. Due to the genetic
similarities between germ cells and ESCs, we were curious to observe endogenous
LINC00486 expression in human embryonic stem cells (H7 hESCs) and induced
pluripotent cells (CWRU205 iPSCs). We received the hESC and iPSC RNA from Dr.
Paul Tesar at Case Western Reserve University. cDNA was synthesized using 500ng
RNA and qPCR was performed to quantify transcript expression in the different cell
types. For each transcript, the same tissue that previously showed no transcript expression
was used as a control. Reduced expression for each transcript was observed in H7 hESCs
and CWRU205 iPSCs (Figure 5E), likely indicating no functional role for LINC00486 in
human ESCs.
29


One common capability of IncRNAs is their enhancer-like functions on
neighboring genes. When this occurs, the flanking genes often show similar IncRNA
expression, which is visualized using qPCR. To determine if LINC00486 had cis
capabilities, we looked at endogenous flanking gene (genes TTC27 and LTBP1)
expression in human cells. Probes were designed for flanking genes LTBP1 and TTC27
(IDT). We performed qPCR on untransfected HEK293T cells and HEK293T cells
transfected with each individual transcript to observe endogenous TTC27 and LTBP1
expression. Low TTC27 expression was observed in HEK293T cells transfected with
LINC00486 Transcripts 1-3. LTBP1 expression remained relatively constant in all
HEK293T cells compared to the untransfected sample (Figure 6A). However, in N2A
cells transfected with each transcript, Transcript 3 revealed an almost 2-fold increase in
LTBP1 expression compared to the other transcripts and untransfected samples (Figure
6B). This suggests there is a correlation in gene expression between the flanking genes
LTPB1 and LINC0486 Transcript 3 in N2A cells. These results are expected due to prior
knowledge of enhancer-capabilities in other characterized IncRNAs.
Since LINC00486 was identified as being hypomethylated in human
lymphoblasts from patients with bipolar disorder (BPD) that had been treated with
lithium, we wanted to examine whether the expression of the three LINC00486
transcripts was changed when cells were treated with lithium or another Gsk-3 inhibitor,
SB-415,286. In order to look at endogenous transcript expression when Gsk-3 is
inhibited, untransfected HEK293T cells were either treated with lOmM LiCl or 20uM
SB-415,286 for 24 hours. HEK293T cells were chosen to act as a baseline for Gsk-3
inhibition in human cells. Results revealed more than a 2-fold increase in Transcript 1
30


expression in the 30uM SB-415,286 treated cells. However, transcript expression in the
other cell treatments remained relatively constant (Figure 7).
After determining the baseline transcript expression in human cells treated with a
Gsk-3 inhibitor, we wanted to see endogenous transcript and flanking gene expression in
lymphoblast cells derived from the BPD individual that were treated with a Gsk-3
inhibitor. qPCR revealed there was no difference in LTBP1 or TTC27 gene expression
between 3188 lymphoblast cells that were untreated or treated with luM LiCl for 21 days
(Figure 8A). Transcript expression was compared between treated and untreated cells for
the same affected individual, revealing a 2-fold increase in 3188 lymphoblast cells treated
with 30uM SB for 2 days compared to untreated and LiCl treated cells (Figure 8B).
To quantify endogenous transcript expression in a (BPD) patient (3188) compared
to their unaffected sibling (3189), qPCR was performed to observe transcript and
flanking gene expression in these individuals. When comparing the BPD patient (3188)
with their unaffected sibling (3189), there was much lower LTBP1 expression in 3189
compared to 3188. TTC27 expression remained relatively constant between the siblings.
Transcript 3 expression revealed a 0.5-fold decrease in 3189 compared to 3188 (Figure
8C). Transcript 3 was chosen due to having the most robust expression in N2A cells in
previous results (Figure 6B).
31


Primers:
400 bp
300 bp
200 bp
£
(J
.A -A
£
8
.A .A
A
C
A?
c
A?
100 bp
Untransfected
HEK293T
No Template
Control
B
Primers:
< £ o Oo 20 O) q; 00 /v oo *-# *-*
i? 8 o' ** rts/ ol rv?7 O* tr7
or) Or) Or) Or)
£>
Or) CO
CU1
rry ry,

xy
O)

c
or)
p)
to N
.^/ .^/
o)
500 bpH
300 bp
200 bp----
100 bp ^
HEK293T cDNA
Figure 3. Endogenous LINC00486 Transcript Expression in HEK293T
Cells. (A) RT-PCR on untransfected HEK293T cells to observe endogenous
transcript expression. (Expected sizesTranscript 1: 298bp; Transcript 2:
129bp; Transcript 3: 179bp; GAPDH: 150bp. GAPDH primers were used as a
control for RNA integrity). (B) Due to the inability of transcript primers to detect
endogenous transcript expression (A), new transcript primers were designed to
enhance detection of endogenous expression. Each primer name indicates
expected band size. Asterisk (*) indicates selected primer used in downstream
experiments (Transcript f primer that was chosen is not shown).
32


^ '-f * - "* *
M .£1 .ft .ft .ft .ft dr O ir> O i'lQioQ < Q i* Q .ft .ft p $ p S .ft .ft .ft 6 i t i t £ Q £ 0 £
Primers: V" P >q- p "T P 'T P 'T P P P "T P <3>-tc3At<3Ai-(3At<3At<3At tf ^ (i ^ y ^
200 bp *
100 bp _
i i Human Tissues: .o 5 Whole Brain Spinal Cord Small Intestines Thymus Kidney 1 .2 n) 0 I Testes Salivary Gland
200 bp *
100 bp mm* **
m C {5 Human Tissues: 2- 0 < Bone Marrow Petal Brain Fetal Liver Lung Placenta 1 Skeletal Muscle Spleen 1 2 1 5 5
B ** 4 <\ t\ i\ f\ t\f r\ <\
-St -£L -5L .fit £ i- ii- i i * * St .ft P 0 P 0 £t .ft -ft C5 <£> iC) <3
Primers: # ij r. 200 bp * * *
100 bp ^
.E | Human Tissues: <£ =5 .Q 01 Whole Brain Spinal Cord Small ntestines Thymus Kidney £ £ .2 2 0 CL I "1 £ = 0
u
200 bp *
100 bp ^ * a

Human Tissues: § -0 < Bone Marrow Fetal Brain Fetal Liver Lung Placenta 1 Skeletal 1 Muscle Spleen 1 1 0 O' 0 5 u
C O) ft) ft) ft) ft) m m m
ft -ft .ft .ft .ft ^ H H U U St St | | St £i p i p S p
Primers: ,P ^ O A- T P rr P TT P T P y (t (3 it <5 At <3 At <3 P ^ P *T At <3 At <3 £ $ £ $ £
200 bp * * *
100 bp ~ mm
s E n a & 1 Testes | Salivary 1 Gland 1
u
200 bp *
100 bp *
Human Tissues: ^ if < Bone Marrow Fetal Brain Fetal Liver Lung Placenta 1 Skeletal 1 Muscle 1 Spleen 1 2 § 1 3
Figure 4. Detecting Presence of Endogenous LINC00486 Transcript
Expression in Human Tissue. Primers selected from Figure 3B were used
to observe (A) Transcript 1 (105bp), (B) Transcript 2 (84bp), and (C)
Transcript 3 (106bp) in 20 human tissues. Asterisk (*) indicates tissue with
robust transcript expression. (GAPDH primers (150bp) were used as a
control for RNA integrity).
33


Table 1. Tissues with robust transcript expression and designated tissue controls.
Expressed Transcript: Robust Tissue Expression: Low Expression Tissues as Chosen Controls:
Transcript 1 Salivary Gland Brain Cerebellum
Colon
Transcript 2 Prostate Whole Brain
Salivary Gland
Spleen
Transcript 3 Whole Brain Small Intestines
Testes
Salivary Gland
Bone Marrow
34


30
25
15
10
Transcript 1 Transcript2
Transcript 3
..flllb..
U
£L,di
E £
1
I
n.oi.dilll.i
SC ^ 00
'= 01 C
Cerebellum
(Control)

Transcriptl Transcript2 Transcripts
Whole Brain
(Control)

Small
Intestines
r*

Small Bone Marrow Fetal Liver
Intestines
(Control)
Figure 5. Quantification of Endogenous LINC00486 Transcript
Expression in Human Tissues and Cells. (A) Quantified
LINC00486 Transcript 1-3 expression in 20 human tissues. Each
transcript was normalized to designated tissue control (Table 1). (B-
D) RT-PCR on qPCR products from (A) using transcript specific
primers. Absence of primer dimers was indicative of accurate results
(GAPDH primers were used as a control for RNA integrity). (E)
LINC00486 Transcript 1-3 expression in hESCs and iPSCs (same
tissue control used for each transcript as shown in Table 1).
35


A
1.2
LTBP1
TTC27
1
0.8
£ 0.6
0.4
HEK293T Untransfected
U
Ml
HEK293T Transcript 1 HEK293T Transcript 2 HEK293T Transcript 3
B
1.8
1.5
1.4
1.2
1
a
££
0.8
0.5
0.4
0.2
0
LTBP1
Figure 6. Endogenous LTBP1 and TTC27 Gene Expression in
HEK293T and N2A Cells. (A) Relative quantification of LTBP1
and TTC27 gene expression in untransfected and transfected
HEK293T cells. (B) Relative quantification of LTBP1 in
untransfected and transfected N2A cells.
36


3
Transcript 1 nTranscript2
Transcript 3
2.5
2
HEK293T Untreated HEK293T ImM LiCI HEK293T 30uM SB
Figure 7. Effect of Gsk-3 Inhibition on Endogenous LINC00486
Transcript Expression in HEK293T Cells. HEK293T cells were
treated with lOmM LiCI or 30uM SB-415,286 for 24 hours. LINC00486
Transcript 1-3 expression was quantified relative to untransfected
HEK293T cells.
37


A 1.6
OLTBP1 TTC27
M
3188 Untreated
Transcript 3
3188 luM LiCI 3188 30uM5B
3188-Untreated 3189-Untreated
Figure 8. LINC00486 Transcript and Flanking Gene Expression in BPD
Patient and Unaffected Sibling. Lymphoblast cells were derived from BPD
individual (3188) and unaffected sibling (3189). (A) Lymphoblast cells from 3188
were treated with luM LiCI for 21 days. Relative quantification of LTPB1 and
TTC27 gene expression in cells treated with Gsk-3 inhibitor. (B) LINC00486
Transcript 3 expression in 3188 lymphoblast cells treated with luM LiCI for 21
days or 30uM SB-415,286 for two days. (C) LTBP1, TTC27 and LINC00486
Transcript 3 expression in 3188 and 3189 lymphoblast cells.
38


RNA Immunoprecipitation
Many characterized IncRNAs have been shown to be associated with chromatin
regulatory complexes in order to initiate genomic silencing. These discoveries have often
occurred through the determination of RNA-protein interactions. Therefore, in order for
us to biochemically characterize LINC00486 we wanted to determine if it physically
associated with any chromatin regulatory complexes. One approach that has been taken
to discover interactions between IncRNAs and histone proteins is ChIRP (chromatin
immuno RNA precipitation) (Chu et al., 2011). We reasoned that a similar approach
should detect interactions between IncRNA and its associated protein. Primers of roughly
20 nucleotides in length were designed to span across the entirety of each transcript.
These primers were designed using the Stellaris Probe Designer by Biosearch
Technologies. Primers were also designed for HOTAIR as a positive control. The first
step was to biotinylate each primer using the Biotin 3 End DNA Labeling Kit from
Thermoscientific. Next, the biotinylated primers were added to the streptavidin beads in
varying concentrations. The concentrations of the biotinylated oligos were varied to
optimize the pull-downs for each transcript. In order to determine if each transcript could
be affinity precipitated from the streptavidin beads, 2ug RNA (from transfected N2A
cells) was added to the biotinylated oligos and streptavidin beads. After a one-hour
incubation, cDNA was synthesized from RNA attached to beads and RT-PCR was
performed. Results indicated the biotinylation of oligos allowed for Transcripts 1-3 to be
affinity precipitated using streptavidin beads. The optimal concentrations for each oligo
were Transcript 11:1000, Transcript 21:10, Transcript 31:10; these concentrations
39


remained constant throughout downstream experiments (Supplemental Figures 2A and
2B).
After validation of successful RNA affinity precipitation from streptavidin beads,
our goal was to determine if the LINC00486 transcripts would associate with EZH2.
EZH2 is the catalytic subunit of PRC2, which many characterized IncRNA have been
known to associate with in order to inhibit genomic loci. N2A cells were transfected with
each transcript individually and RNA was isolated from cells (Life Technologies). This
process was repeated for untransfected HEK293T cells.
An immunoprecipitation was performed to determine if each transcript associated
with HA-tagged EZH2 protein (pCMV-HA-hEZH2 from the Ren lab). EZH2 was
immunoprecipitated using HA-coated magnetic beads to an isolated nucleic lysate from
transfected N2A cells. After an overnight incubation, RNA was isolated from beads and
cDNA was synthesized. Process was repeated using IgG control beads. RT-PCR was
performed to observe immunoprecipitation results. In N2A cells over-expressing each
transcript, HA-tagged EZH2 successfully pulled down Transcript 2. This was confirmed
by the lack of Transcript 2 expression in the IgG control beads set. Transcript 1 and 3
enriched in HA showed non-specific binding (Figure 9A). This immunoprecipitation was
repeated with HEK293T cells transfected with EZH2 (1800ng) and pMax-GFP (200ng)
to determine if endogenous LINC00486 transcripts were capable of co-precipitating HA-
tagged EZH2. Results showed EZH2 pulled down endogenous Transcript 2 from the
nuclei and Transcript 1 from the cytoplasm (Figure 9B).
In addition to the identified IncRNA association with PRC2, many IncRNAs are
also physically associated with PRC1. PRC1 is another chromatin regulatory complex
40


that initiates gene silencing and is involved in maintaining stem cell pluripotency.
Therefore, we wanted to determine if there was any physical interaction between
LINC00486 Transcripts 1-3 and the Cbx protein family. Flag-tagged Cbx proteins were
tested for their capability to immunoprecipitate each transcript. N2A cells were
transfected with individual transcripts (900ng), Flag-tagged pVenus-Cbx (900ng), and
pMax-GFP (200ng). Immunoprecipitation procedure was repeated for Cbx proteins:
Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8 using Anti-Flag magnetic beads instead of HA
magnetic beads (Figures 10A-E). No control beads were used. Results were summarized
in Table 2. Immunoprecipitation was repeated on a GFP-tagged full-length Cbx 2 protein
using Anti-GFP beads (Figure 10F). Results were consistent with the Flag-tagged Cbx2
IP using Anti-Flag beads (Figure 10A).
41


A
Primers:
'V
. .A .A .A £ -t & C? t? .c. tf P £ £? £ n iii .A .§ .A A i 6 £ 6 ... i & .i .A A A o n o £ P P P -I- y /t i .A P J ir £ c? g A .O cj * .A P d ... £ & £ $ Q <3 (5 AT £ A P c? t £ g g a £ *

200 bp
100 bp
1---------1--------,L
N2A Transfection:
Type of Beads Used:
Primers:
300 bp
200 bp
100 bp
N2A Transfection:
Type of Beads Used:
EZH2 E2H2 Transcript 1 Transcript 1
HA beads control beads HA beads control beads
'* <\ m
# <3 P .A 5- b ^ c: g ^ f o <3 £ ¥ o (5 P P Q .A -S Pc? b b .c, C c A A £ SO JV Q C? P P P P
mm **
ll H Jl , I
1 1 Transcript 3 HA beads 1 Transcript 3 Control beads 1 Hotair HA beads T Hotair control beads No Template Control
Transcript 2
HA beads
Transcript 2
control beads
B
200 bp
100 bp
100 bp
HA-EZH2 IP
293 Nuclei Cytoplasm
T
control-EZH2 IP
293 Nuclei
Cytoplasm
No Template
Control
Figure 9. EZH2 Immunoprecipitation of LINC00486 Transcripts 1-3.
(A) RT-PCR of nucleic isolate from N2A cells transfected with EZH2 and
individual transcripts after HA-tagged EZH2 immunoprecipitation. (B) RT-
PCR of endogenous LINC00486 transcripts in HEK293T cells after HA-
tagged EZH2 immunoprecipitation. For A and B, transcript specific
primers were used to amplify any transcripts that were pulled down with
EZH2. (Expected sizesTranscript 1: 105bp; Transcript 2: 84bp;
Transcript 3: 106bp; GAPDH: 150bp, HOTAIR: 250bp. GAPDH primers
were used as a control for RNA integrity and HOTAIR was used as a
positive control for a representative IncRNA).
42


A> liihiiiiin 300 bp __ inob' , , 200 hp |npul
b /// 200bp FlaglP N2ATranrfecbon: cb*4 Cb4 + Cb4- Cb4* Cbx4 / ///A/ // /// ////// //// Z^Z f',r
N2A Transfection: Cb2 Cbx2 4 Cra2. Cbj3+ Cta2- No Template Control N2A Transtecbon: Cbi4 Cbx4 Cbwt Cb*4. Cbx4 No Template Control
zjjlllnliililhii D-~mji/IhihiLihu S-
^niShiSiihiii mills Fiagip 1 1 ' ' 1 11 1 M 1 ru-1* N2A Transfection. Cbx7 Cbx7 Cbx7 d7t Cbx7* fi/Ihih sh sil St iiiii 300 bp
t,;, T;2 Tl"j;is
E p- Jiiflsifsihi/siS ^js/lls ilttliilas
100 tw ^ ^ ^uiilssinisiisis ////in^ --/I//// ih ih ihni f Ils
N2A T ransTect.cn: ctaa r*-,t 1 r,^f2 Trjn^pt 3
Figure 10. Cbx Protein Immunoprecipitation of LINC00486 Transcripts 1-3.
N2A cells were cotransfected with individual Cbx proteins and LINC00486
Transcripts 1-3. A-F) Transcript specific primers were used in RT-PCR to detect
presence of transcript pull-down from Cbx immunoprecipitations. (A-E) Each
Flag-tagged Cbx protein was individually immunoprecipitated from nucleic
isolate. (F) GFP-tagged Cbx2 protein was immunoprecipitated from N2A nucleic
isolate. (Expected sizesTranscript 1: 105bp; Transcript 2: 84bp; Transcript 3:
106bp; GAPDH: 150bp, HOTAIR: 250bp. GAPDH primers were used as a control
for RNA integrity and HOTAIR
Table 2. Summarized results of Cbx protein immunoprecipitations
Transcript 1 Transcript 2 Transcript 3
pVenus-Cbx2 Very Weak None Robust
pVenus-Cbx4 Very Weak None Very Weak
pVenus-Cbx6 Robust None Robust
pVenus-Cbx7 Very Weak None Very Weak
pVenus-Cbx8 None None Robust
43


CHAPTER III
INVOLVEMENT OF GSK-3 IN STEM CELL PLURIPOTENCY
Introduction
Pluripotency in Mouse Embryonic Stem Cells
Mouse embryonic stem cells (mESCs) originate from the inner cell mass of pre-
implantation mouse embryos and are characteristically pluripotent, meaning an individual
cell can give rise to all cell types (Burdon, 2002). Uniquely, mESCs are also capable of
indefinite self-renewing expansion when cultured (Loh et al., 2006). When necessary,
environmental signals can induce differentiation into specific cell types, establishing a
promising model for regenerative disease (Loh et al., 2006; Welham et al., 2011). In
order to establish and maintain pluripotency, extrinsic factors, signaling pathways, and a
regulatory network of transcription factors synergistically work together. The Master
transcription factors (Nanog, Oct4, Klf4, Esrrb, etc.) are essential for establishing
pluripotency (Pei, 2009). These transcription factors contain similar DNA binding and
transactivation domains, which are necessary to interact with important cofactors.
Additionally, these genes are tightly controlled by feedback circuits, which allow them to
self-regulate as well as regulate each other. Data suggests when these genes are
suppressed, pluripotent ESCs begin to differentiate (De et al., 2014). However, when
Gsk-3 is inhibited there is increased expression of the Master transcription factors
allowing maintenance of ESC proliferation and pluripotency (Welham et al., 2011).
A more in-depth focus on these Master transcription factors shows that Nanog is
a homeodomain-containing protein required for the maintenance of pluripotency, which
occurs through the modulation of Oct 4 (Chen et al., 2011; Loh et al., 2006). Oct4 is a
44


POU transcription factor, encoded by Pou5fl. In the absence of Oct4, differentiation
occurs, suggesting that Oct4 is essential for maintaining the identity of pluripotency in
ESCs. Together, Nanog and Oct4 control the downstream genes required for maintaining
pluripotency, such as Esrrb (estrogen-related receptor P) (Loh et al., 2006). Esrrb, a direct
transcriptional target of Nanog, is part of a nuclear hormone receptor superfamily
(Sanchez-Ripoll, 2013). It also lies downstream of Gsk-3 and is required for suppressing
differentiation (Hanna, 2010). During Gsk-3 inhibition, Esrrb gene expression is
increased (Martello et al., 2012). Finally, it is speculated that Klf4 (Kriippel-like factor 4)
helps to regulate differentiation through modifying the chromatin structure of
downstream pluripotent genes. This allows the genes to bind to their targets and activate
the genes that are normally silenced in differentiated cells (Romeo, 2012). These genes
work together in a tightly regulated network to maintain pluripotency in ESCs.
Effect of LIF on Stem Cell Pluripotency
The key factor in mESC pluripotency and self-renewal is LIF (leukemia
inhibitory factor), which activates the STAT3 (signal transducer and activator of
transcription 3) signaling pathway and c-Myc. Once LIF binds to the receptor, it results in
the activation of extracellular regulated kinases (Erkl and Erk2) and PI3K signaling
(Welham et al., 2011). There is a necessary balance between these pathways because
activation of STAT3 and PI3K promote pluripotency and self-renewal through Gsk-3
inhibition, while ERK activation promotes differentiation. This important balance
determines the fate of mESCs (Storm et al., 2007). ESCs will naturally respond to
environmental signals to change from a pluripotent state to a differentiated state, but in
cultured medium, other extrinsic factors such as serum or Bone Morphogenetic Proteins 2
45


or 4 (BMP 2 or BMP4) are supplemented to work synergistically with LIF to tip the
balance in order to sustain pluripotency (Chen et al., 2011; Sanchez-Ripoll, 2013).
Without supplementing LIF, mouse embryo fibroblast (feeders) can be used to co-
culture with mESCs to secrete necessary growth factors and maintain pluripotency
(Tamm, Pijuan Galito, & Anneren, 2013). Ground state pluripotency can also be created
by growing ESCs in serum-free media in the absence of LIF with the simultaneous
inhibition of both Gsk-3 and MEK (Mitogen activated and extracellular-regulated kinase
kinase) through the addition of the two specific inhibitors (CHIR99021 and PD 1843 52m
respectively; 2i) (Sanchez-Ripoll, 2013; Ying et al., 2008). Inhibiting the MEK signaling
pathway results in the inhibition of the differentiation-inducing effects secreted by FGF4
(fibroblast growth factor 4). However, introducing the cells to only MEK inhibitors
caused the cells to grow poorly. When the Gsk-3 inhibitor was added, the cell state was
drastically improved, providing insight into the requirement for Gsk-3 inhibition to
stimulate cellular proliferation and metabolism. This is consistent with the roles of PI3K
and Wnt in inhibition of Gsk-3 and regulating proliferation of ESCs. Additionally, Gsk-3
DKO ESCs mimic effects of Gsk-3 inhibitors and are more resistant to differentiation
than WT ESCs (Welham et al., 2011).
Effect Gsk-3 Inhibition on Stem Cell Pluripotency
Inhibiting Gsk-3 has shown to enhance self-renewal of ESCs and facilitates
resistance to differentiation (Sanchez-Ripoll, 2013). Previous studies have shown that
when Gsk-3 was inhibited through the PI3K pathway, expression of the Master
pluripotency transcription factors were altered at the protein level (Sanchez-Ripoll,
2013). Nanog, a core element of the pluripotency network, is imperative for cell vitality
46


and resisting differentiation (Storm et al., 2007). Gsk-3 inhibition cultured in the absence
of LIF resulted in selective upregulation in Nanog protein levels, which preceded changes
in RNA levels. These results indicate that these changes are translational, but not
transcriptional. However, the stability of these proteins were not altered (Sanchez-Ripoll,
2013) . Despite the numerous studies suggesting Gsk-3 inhibition promotes pluripotency,
the exact mechanism behind this evidence remains unsolved.
Another recent discovery relating to Gsk-3 inhibition is now commonly used to
benefit researchers working with induced pluripotent stem cells (iPSCs). The addition of
ascorbic acid (commonly known as Vitamin C) and a Gsk-3 inhibitor (called AGi), has
been shown to be the most effective combination of treatments to reprogram mouse
somatic cells into iPSCs. Vitamin C acts to alleviate cell senescence (Bar-Nur et al.,
2014) . During the reprogramming process, tumor-suppressive mechanisms (largely
through tumor protein p53) are activated, which initiates cellular senescence to act as a
barrier to iPSC generation (Banito & Gil, 2010). Vitamin C relieves this barrier by
reducing p53 activity to promote the generation of iPSCs (Esteban et al., 2010). To
reprogram somatic cells, Vitamin C and Gsk-3 inhibition work synergistically by
initiating early activation of the key pluripotency genes (Bar-Nur et al., 2014).
Additionally, Vitamin C is a known cofactor for histone demethylases and alpha-
ketoglutarate-dependent dioxygenases. It is suggested that Vitamin C may allow
reprogramming to run more smoothly by aiding in the histone demethylation process
(Esteban et al., 2010). Previously, forcing the cellular change from somatic cells to iPSCs
was only possible through tedious and time consuming genetic manipulation (Bar-Nur et
al., 2014). Today, the addition of AGi improves the efficiency and speed of iPSC
47


formation. It remains to be determined how these mechanisms fully work; yet AGi
provides great potential into making the transition from somatic cells to iPSCs more
efficient.
m6A mRNA Modification
RNA epigenetics plays an important role in stem cell pluripotency. Over a
hundred different post-transcriptional modifications decorate RNA molecules, however
our knowledge of their location and function are limited (Dominissini et al., 2012). N6-
methyladensoise (m6A) is the most abundant posttanscriptional modification found on
thousands of mRNAs, and IncRNAs in somatic cells. It is suggested that m6A abundance
is roughly 0.1%-0.4% of total adenosine residues in mRNA (Meyer et al., 2012). The
novel approach to identify and localize m6A sites, m6A-seq, has exposed the RNA
methylome to improve our understanding of this modification. m6A-seq has identified the
fact that methylation can be found individually or in clusters and are typically enriched
near the stop codon or within the 3 untranslated region (UTR) in mRNA (Linder et al.,
2015). The presence of this modification near stop codons suggests translational control
(Dominissini et al., 2012). Association with 3UTR indicates a role in RNA regulation
because the 3UTR influences RNA stability, subcellular localization and translational
regulation. Additionally, there was no m6A found in the mRNA poly-A tail (Meyer et al.,
2012). Performing m6A-seq on RNA isolated from HEK293T cells, over 12,000 m6A
sites were identified in the transcripts of more than 7,000 human genes, indicating this
modification is very widespread throughout the transcriptome (Dominissini et al., 2012).
mRNA methylation is non-stoichiometric, meaning only a fraction of transcripts are
methylated at a specific position. In addition, some transcripts are completely void of
48


methylation. These observations suggest m6A locations are non-random and most likely
play a regulatory role in translation (Dominissini et al., 2012). This modification is found
on genes that are involved in numerous cellular functions including transcriptional
regulation, intracellular signaling pathways, and have been linked to diseases such as
cancer and obesity. Additionally, m6A is prominently found on transcripts that encode the
major pluripotency transcription factors. Research has shown that this mRNA
modification is directly correlated to stem cell pluripotencyreduced levels of m6A
result in the inability for ESCs to differentiate (Batista et al., 2014). Although this
modification is essential for cellular development and viability, the precise role of the
addition of m6A on mRNA remains to be determined (X. Wang et al., 2014).
The addition of the m6A modification is catalyzed by Mettl3 (Methyl Transferase-
like 3) and a related, but uncharacterized protein, Mettll4 (Y. Wang et al., 2014). These
proteins add methyl groups onto the C6 position of adenosine, which does not hinder its
ability to base pair with thymidine or uracil (Guela, 2015; Meyer et al., 2012). Two
members of the alpha-ketoglutarate-dependent dioxygenases protein family, FTO (Fat
and Obesity Gene) and Alkbh 5 (AlkB homolog 5) are the demethylases that oxidatively
remove the methyl group from adenosines (Zheng et al., 2013). This process is catalyzed
by the glutarate-dependent enzyme co-factor, Vitamin C (Gerken, 2007). FTO is a
regulator of metabolism and energy utilization and SNPs with increased levels of FTO
expression resulted in an elevated body mass and an increased risk for obesity (Meyer et
al., 2012). FTO is present not only in the nucleus, but in the cytoplasm where
posttranscriptional modifications occur, such as the addition of m6A (Gerken, 2007).
When FTO was overexpressed in human cells, there was a decrease in the amount of m6A
49


observed. In contrast, when FTO was knocked down, increased amounts of m6A on
mRNA was observed (Jia et al., 2011). These data validate the role of FTO as a
demethylase and demonstrates RNA methylation is a reversible process that regulates
mRNA metabolism (X. Wang et al., 2014).
Recently, it has been suggested that the addition of m6A directly reduces the
stability of mRNA and therefore decreases translation frequency (Guela, 2015). In
general, a large number of transcripts exhibiting the m6A modification show an inverse
correlation with mRNA stability and gene expression (Y. Wang et al., 2014). The
transcripts marked with m6A showed a reduced half-life and increased rates of decay,
indicating the association between m6A and transcript turnover (Batista et al., 2014). This
is due to the binding of the YTH domain family proteins to m6A sites, which recruit the
transcripts to RNA-decay processing bodies (P-bodies) (Batista et al., 2014). Five
different YTH-domain family members have been identified, two of which have been
highly studied (Y. Yang et al., 2015). YTDHF1 and YTDHF2 play a role in mRNA
translational abilities (X. Wang et al., 2014). The YTH-domain family is known to bind
onto the C-terminal YTH domain of single-stranded RNA. Once bound, each protein has
a unique role to ensure the m6A marked mRNA has efficient gene expression and
controllable production of proteins (Wang et al., 2015). The carboxyl-terminal domain of
YTHDF2 selectively binds the m6A-methylated mRNA while the amino-terminal domain
localizes the YTHDF2-mRNA complex to mRNA P-bodies. This indicates YTHDF2 has
the ability to affect translation and the lifetime of mRNA (X. Wang et al., 2014).
Working synergistically with YTHDF2, YTHDF1 directly promotes the
translation efficiency of methylated mRNAs by trafficking more m6A mRNA transcripts
50


to translational machinery. Additionally, YTHDF1 directly accelerates the translation
initiation rate once the mRNA is bound to the ribosome (Wang et al., 2015). YTHDF1
and YTHDF2 possess very distinct functions, which intuitively result in two opposing
mRNA fates. This suggests a more complex understanding of the overall m6A function.
In order to surpass this issue, data has shown that both proteins have their own individual
sets of target mRNAs as well as a large set of common target mRNAs. For consistent
regulation of the shared targets, YTHDF1 binds to mRNA prior to YTHDF2 in order to
activate the translation of the methylated mRNAs for optimal production of proteins.
Once YTHDF2 binds, the lifetime of the mRNAs is limited. This dynamic gene-
expression regulation allows for a more precise and effective process for protein
production with a decreased response time between the stimulus and the translational
output. This occurs during various biological transformations, one of which is promoting
cellular differentiation (Wang et al., 2015).
Together, Mettl3, FTO/Alkbh5 and YTHDF1/2 act as molecular switches that
regulate the m6A modification on mRNA (Guela, 2015). Using epigenetic terminology,
Mettl3 acts as the writer, FTO/Alkbh5 are the erasers and YTHDF proteins are the
readers of m6A modifications. Despite identifying these epigenetic regulators, the
precise mechanism behind how these enzymes know when to add or remove the m6A tags
is still unknown. However, since the addition of m6A plays a critical role in mRNA
degradation and turnover, it can be suggested that the m6A process is highly regulated
and is capable of being finely tuned in response to varying environmental stimuli.
An interesting observation from m6A RNA reductions is maintenance of
pluripotency in ESCs, even under conditions when differentiation should occur. Mettl3
51


has been shown to terminate pluripotency and promote differentiation in ESCs. When
Mettl3 was knocked out, cells showing normal pluripotent morphology did not start to
differentiate (Y. Wang et al., 2014). Additionally, the expression of Esrrb and Nanog was
not altered. This indicates that ablation of Mettl3 depletes almost all m6A tags on mRNA,
increasing the stability of the pluripotent mRNA transcripts as well as the translational
efficiency, resulting in the maintained pluripotent state (Batista et al., 2014). To take this
one step farther, when the Mettl3 KO ESCs were grown in the absence of LIF, they
became even more resilient to differentiation, further validating the idea that Mettl3
terminates pluripotency (Guela, 2015). This data suggests that the addition of m6A may
play an important role in stem cell differentiation.
Interestingly, the Mettl3 KO and the Gsk-3 DKO ESCs showed a very similar
phenotype due to their inability to differentiate. This observation inspired us to
hypothesize that the addition of the m6A modification on mRNA is regulated by Gsk-3
activity through FTO. We hypothesize that Gsk-3 phosphorylates FTO, resulting in FTO
ubiquitination and degradation. Reduced FTO levels would lead to reduced demethylase
activity, which would cause an increase in m6A mRNA, promoting the degradation of the
mRNA, resulting in the differentiation of ESCs. However, in the absence of Gsk-3
activity, FTO would remain active, resulting in reduced m6A levels and decreased mRNA
degradation, allowing the cells to maintain in the pluripotent state. If our hypothesis is
correct, this discovery of Gsk-3, a multifaceted regulator of multiple signaling pathways,
as the controller of m6A modifications on mRNA, would provide the first mechanistic
explanation into how this RNA epigenetic process is regulated. It would also provide
52


insight into understanding how different Gsk-3 dependent signaling pathways regulate
the fine balance between differentiation and pluripotency in ESCs.
The purpose of this project is to examine the relationship between Gsk-3 and stem
cell pluripotency by a novel mechanism: the regulation of the m6A modification on
mRNA. We hypothesize the m6A tag is regulated by Gsk-3 through controlling the
protein levels FTO, a demethylase. Inhibition of Gsk-3 would result in increased FTO
protein levels, resulting in reduced m6A levels and maintenance of stem cell pluripotency.
Identifying the regulatory role of Gsk-3 will provide a mechanistic understanding of stem
cell differentiation as well as directly link m6A modifications with Gsk-3 function. As
previously stated, the specific aims include:
Aim #1: Determine if the effect of Gsk-3 on stem cell pluripotency is through the
regulation of mRNA methylation.
Aim #2: Investigate if Vitamin C and Gsk-3 inhibition work synergistically to promote
stem cell pluripotency
Materials and Methods
RNA Immunoprecipitation and ELISA:
m6A Immunoprecipitation
Anti-N6-methyladenosine (m6A) (Millipore) was coupled to dynabeads using the
Dynabead Antibody Coupling Kit from Life Technologies. lOmg dynabeads was added
to 50uL m6A antibody (at lug/mL) to make a final concentration of lOmg/mL. The
coupling process was performed by following the kit protocol. Immunoprecipitation
protocol was similar to EZH2 protocol, except 40uL of Anti-m6A beads were added to
untransfected WT and DKO cells resuspensions. No control beads were used. After
53


successful immunoprecipitation of m6A mRNA, qPCR was performed to quantify
pluripotency gene expression.
m6A ELISA Using Qubit
1. Total RNA was isolated (Zymo) from untransfected N2A cells; the additional
DNase restriction step was performed.
2. Per RNA sample, 50uL of Oligo d(T) magnetic beads were added 200uL
Lysis/Binding buffer (mirVana Kit). Mixture was vortexed briefly and then
mixed on rotator for 2 minutes at room temperature.
3. 5ug of total isolated RNA was added to magnetic beads. Tubes were placed on
rotator to incubate overnight at 4C.
4. Using a magnetic plate, the original liquid was taken off beads and saved as
dT Other.
5. Beads were washed with lOOuL RNase-free water and water was discarded.
Wash was repeated 2 more times.
6. To elute RNA off of beads, 20uL cold lOmM Tris-HCl pH 7.5 was added to
RNA and beads. Mixture was incubated at 80C for 2 minutes.
7. After incubation, tube was immediately placed on magnet and eluted RNA
was placed into a new tube on ice.
8. RNA was quantified using Qubit to determine how much poly-A RNA was
pulled down from Oligo dT beads.
9. Desired amount of RNA was added to lOuL, 50uL and lOOuL of m6A beads.
Negative controls with the same volume of m6A beads and no RNA were
created.
54


10. Experimental and control tubes were rotated for 2 hours at 4C.
11. After incubation, tubes were placed on a magnetic plate and original liquid
was removed and saved as m6A Other.
12. Beads were washed with lOOuL RNase-free water and water was discarded.
Wash was repeated 2 more times.
13. To each tube, 300uL Elution Buffer (Appendix) was added. Samples were
incubated at 50C for 1.5 hours to elute RNA off of beads.
14. After incubation, tubes were placed on magnetic plate and eluted RNA was
placed into a new tube on ice and saved as After m6A.
15. RNA was quantified using Qubit to determine how much m6A RNA was
immunoprecipitated from Anti-m6A beads.
16. To purify RNA, 1 volume of isoamyhchloroform was added to 1 volume of
Elution Buffer (300uL). Tubes were vortexed vigorously for 15 seconds.
Tubes were then left to homogenize at room temperature for 3 minutes.
17. Tubes were centrifuged at 12,000 xg for 15 minutes at 4C. Upper aqueous
phase was transferred to a fresh tube. Volume of this phase was determined.
18. 0.1 volume of 3M Sodium Acetate pH 5.2 and 2 volumes of ice-cold Ethanol
was added to each tube. Tubes were mixed.
19. Ethanolic solution was incubated from 1 hour to 18 hours (overnight) at -
20C to allow RNA to precipitate.
20. After incubation, RNA was recovered by centrifugation at 13,000 x g at 4C.
21. All traces of supernatant were removed without disturbing the invisible pellet.
55


22. The pellet was washed with 500uL ice-cold 70% Ethanol. Tubes were
centrifuged at maximum speed for 10 minutes at 4C. Steps 21 and 22 were
repeated.
23. Tubes were left open at room temperature for all traces of Ethanol to
evaporate. However, the RNA pellet was not left to dry completely.
24. After evaporation, RNA pellet was dissolved in 30uL DEPC water. The walls
of the tube were rinsed with the water to obtain as much RNA as possible. To
aid solubilization, RNA pellet was incubated in resuspensions solution for 5
minutes at 65C with intermittent gentle vortexing.
25. Tubes were stored as Final RNA and quantified using the Qubit. All RNA
was stored at -80C.
m6A ELISA (Based Off Epigentek Protocol)
1. Total RNA was isolated (Zymo) from untransfected N2A cells; the additional
DNase restriction step was performed.
2. In a 96-well plate, 80uL DNA/RNA Binding Buffer (Zymo) was added to
each well.
3. 5ug of total isolated RNA was added to designated wells. Solution was mixed
gently by tilting plate side to side, ensuring the solution coated the bottom of
each well evenly.
4. The plate was covered with a lid and incubated at 37C for 90 minutes.
5. After incubation, DNA/RNA Binding Buffer was removed from each well.
Each well was washed with adding 150uL TBST. TBST was then removed
and discarded. Wash was repeated 2 more times.
56


6. m6A Antibody was diluted 1:1000 in TBST. 50uL of diluted m6A Antibody
was added to each well. Plate was covered with lid and incubated at room
temperature for 60 minutes.
7. After incubation, m6A Antibody was removed from each well and each well
was washed with 150uL TBST for a total of three times.
8. Anti-Rabbit antibody was diluted 1:5000 in TBST. 50uL of Anti-Rabbit
Antibody was added to each well. Plate was covered and incubated at room
temperature for 60 minutes.
9. After incubation, Anti-Rabbit antibody was removed from each well and
discarded. Wells were washed with 150uL TBST for a total of four times.
10. PNPP was equilibrated to room temperature and was mixed end over end.
IOOuL of PNPP was added to each well.
11. Plate was gently agitated to mix thoroughly and then incubated at room
temperature until color developed.
12. Reaction was stopped by the addition of IOOuL 2M Sulfuric Acid. Plate was
agitated gently to mix.
13. Plate was placed in illuminometer and the absorbance was measured at
450nm.
Process was repeated with AP-linked Antibody and TMB to determine optimal
protocol.
Observing Protein Expression:
Western Blotting
57


For FTO westerns, membrane was blocked in 5% milk/TBST for 1 hour at room
temperature, while rotating. Primary FTO antibody (PhosphoSolutions Reference#: 597-
FTO) was added to the membrane at 1:1000 in lOmL of 1% milk/TBST. Membrane was
incubated in primary antibody overnight at 4C while rotating. Membrane was then
washed with TBST (lx for 15 min and 2x for 5 min) and then Anti-Mouse secondary
antibody was added at 1:20,000 in lOmL of 1% milk/TBST for 30 minutes at room
temperature while rotating. Membrane was washed with TSBST (3x for 5 minutes) and
then was ECL was added to membrane for 5 minutes. Membrane was blotted dry and
then imaged.
For Alpha-Tubulin westerns, membrane was blocked in 4% BSA/TBST for 1
hour at room temperature, while rotating. Primary Alpha-Tubulin antibody (Cell
Signaling Reference#: A11126) was added at 1:1000 in lOmL of 4% BSA/TBST. For
secondary antibody, Anti-Mouse was added at 1:20,000 in lOmL of 4% BSA/TBST.
Similar procedure was followed as used in FTO westerns.
For Gsk-3 a/p westerns, membrane was blocked in 5% BSA/TBST for 1 hour at
room temperature, while rotating. Primary Gsk-3a/p antibody (CalBioChem Reference #:
368662) was added to membrane at 1:1000 in lOmL 5% BSA/TBST. For secondary
antibody, Anti-Mouse was added at 1:20,000 in lOmL 5% BSA/TBST. Similar procedure
was followed as used in FTO westerns.
Results
Inhibition of Gsk-3 has been shown to maintain stem cell pluripotency. However,
the mechanism behind how this works is yet to be established. Due to the similar
phenotypes between METTL3 KO and Gsk-3 DKO ESCs, we reasoned these genes could
58


be working within the same pathway. Therefore, we hypothesized this resistance to
differentiation in Gsk-3 DKO ESCs may be due to the addition of m6A through the
regulation of Gsk-3. m6A levels on mRNA directly correlate with pluripotency; ESCs
that maintain pluripotency have reduced m6A mRNA levels. As these methylation levels
increase, stem cells begin to differentiate. Due to the common knowledge that Gsk-3 is a
negative regulator, we began by exploring the possibility that Gsk-3 is regulating mRNA
methylation by the phosphorylation and suppression of the demethylase, FTO, possibly
by targeting for degradation. We hypothesize that when Gsk-3 is inhibited, there is an
increase in FTO protein levels, resulting in decreased m6A levels, which leads to
enhanced pluripotency. Fortunately, we had the necessary reagents and tools to begin
dissecting these questions.
Pluripotency Gene Expression in Immunoprecipitated m6A RNA
To determine the relative levels of m6A modifications found on specific mRNAs,
we immunoprecipitated m6A RNA from untransfected WT and Gsk-3 DKO cells grown
in complete ESC media with the addition of LIF. After lysing and adding anti-m6A
magnetic beads, followed by incubation, all liquid was removed from beads (Input) and
m6A RNA was then isolated off of the beads. RNA was isolated from the Input liquid and
cDNA was synthesized from both Input and m6A mRNA. qPCR was performed to
observe GAPDH, Nanog, and Klf4 gene expression (Figure 11). Results revealed
relatively no change in expression levels, which did not support our hypothesis. Next, we
wanted to determine the half-life of the m6A mRNA. This was accomplished by treating
WT and Gsk-3 DKO cells with a-amanatin, a potent inhibitor of RNA polymerase II. By
blocking RNA polymerase II, the cell is unable to enter the next round of synthesis,
59


thereby halting transcription (Brueckner & Cramer, 2008). This allows for the
determination of the stability of mRNA through qPCR. To determine the half-life of
Nanog in m6A mRNA, the m6A immunoprecipitation was repeated in WT and Gsk-3
DKO ESCs grown in complete ESC media with the addition of LIF. These cells were
treated with lug/mL a-amanatin for 0 or 2 hrs. qPCR revealed a higher CT mean for m6A
IP of WT at the 2 hour treatment and the m6A IP of DKO at the 0 hour, indicating
reduced Nanog expression at these time points. The CT mean for the input of both WT
and DKO at all time points remained relatively constant (Figure 1 ID). These results did
not match up with our hypothesis. Next, we were curious about FTO protein levels in
ESCs. In normal ESC culturing conditions with the addition of LIF, FTO protein
expression is high in both WT and Gsk-3 DKO ESCs (a-tubulin western served as a
loading control) (Figure 12). This was anticipated because the addition of LIF maintains
pluripotency in the WT ESCs. Additionally, Gsk-3 DKO ESCs remain pluripotent due to
the ablation of Gsk-3, even in the absence of LIF. As previously mentioned, pluripotency
is correlated with m6A levelspluripotent stem cells have high levels of m6A due to the
suppressed FTO demethylase function.
Unfortunately, the immunoprecipitation and western blot results did not support
our hypothesis. We concluded that we needed to find a way to separate the FTO
mechanism from pluripotency because cells grown in the presence of LIF maintain
pluripotency and mask the mechanistic effects of FTO. Therefore, we decided to
manipulate the cells by removing LIF to see the mechanistic changes on FTO protein
levels.
60


GAPDH Nanog Klf4
input m6A!P
WT 0 hr WT 2 hr DKO 0 hr DK02hr
Nanog
Figure 11. m6A mRNA Gene Expression in WT and Gsk-3 DKO ESCs. m6A
mRNA was immunoprecipitated from total RNA isolated from WT and Gsk-3 DKO
ESCs grown in complete ESC media in the presence of LIF. qPCR was performed to
look at (A) GAPDH, (B) Nanog and (C) Klf4 gene expression in the m6A IP
compared to Input total RNA. (D) Prior to m6A IP, WT and Gsk-3 DKO ESCs were
treated with a-amanatin and total RNA was isolated at zero or two hour time points.
m6A mRNA was immunoprecipitated from RNA and qPCR was performed to look
at Nanog gene expression.
WT DKO
50 kD
IB: FTO
50 kD
IB: a-Tubulin
Figure 12. FTO Protein Expression in WT and Gsk-3 DKO
ESCs Cultured in Complete ESC Media in the Presence of LIF.
Protein band sizes are indicated on left side, a-tubulin western
served as a loading control.
61


Effect of LIF on mESC Protein and Gene Expression
A paper from the Samuel Lunenfeld Research Institute demonstrated that Gsk-3
DKO ESCs remained pluripotent even after removing LIF for 14 days (Doble et al.,
2007). Therefore, we decided to use the same experimental regimen. We grew WT and
Gsk-3 DKO ESCs in normal ESC media in the absence of LIF for 14 days. In addition to
these cell types, we decided to include pi 10* ESCs. These cells have the myristolated
form of the PI3K subunit, pi 10a, which promotes a constitutively active insulin pathway,
resulting in the constant inhibition of Gsk-3. We were curious to see how these cells
would maintain pluripotency in comparison to Gsk-3 DKO ESCs when grown without
LIF. On day 14, protein was isolated and quantified using Qubit. Next, we determined
protein expression by performing FTO, a-tubulin (Figure 13A) and Gsk-3a/p western
blots (Figures 13A-B). FTO expression in Gsk-3 DKO ESCs was much more robust than
in WT or pi 10* ESCs (Figures 13A-B). The a-tubulin western served as a loading
control and the Gsk-3 a/p western was performed to validate sufficient knockout of both
isoforms in the Gsk-3 DKO cells.
In addition to protein, RNA was isolated from the embryonic stem cells (Zymo)
and cDNA was synthesized (Applied Biosystems). We then assessed gene expression for
Nanog, Oct4, and Esrrb (the Master pluripotency genes). Gene expression was
quantified through qPCR in order to observe levels of pluripotency in cells when LIF is
absent (Figure 14). qPCR results revealed almost a 6-fold increase in Nanog expression
in Gsk-3 DKO without LIF for 14 days compared to WT without LIF. The Nanog
expression in WT cells treated without LIF for 14 days was almost identical to both the
WT and Gsk-3 DKO cells with LIF for 14 days (Figure 14A). Esrrb expression in Gsk-3
62


DKO cells treated without LIF for 14 days increased 10 fold compared to WT without
LIF and WT and Gsk-3 DKO plus LIF (Figure 13B). Due to the consistencies in the high
expression of pluripotency genes in Gsk-3 DKO ESCs when grown in the absence of
LIF, we wanted to further investigate exactly how much higher the gene expression
would be in other pluripotency genes. When qPCR results for Nanog, Oct4 and Esrrb
expression in WT and Gsk-3 DKO ESCs without LIF for 14 days were combined, Esrrb
expression in Gsk-3 DKO ESCs without LIF had the greatest increase in expression with
an eleven fold increase compared to WT without LIF for all three genes (Figure 14C).
Next, we were curious about FTO gene expression in WT and Gsk-3 DKO ESCs when
grown without LIF for 14 days. Results indicate a slight decrease in FTO expression in
the Gsk-3 DKO cells, which is consistent with our assumption that regulation by FTO
occurs at the translational level, not at the transcriptional level (Figure 14D).
After seeing successful results that supported our hypothesis, we wanted to see if
we could alter the cellular medium conditions to speed up the time frame. First, we
repeated the no LIF experiment in complete ESC media. However, this time we isolated
protein on various days to compare to protein isolated on day 14. WT and Gsk-3 DKO
ESC protein isolated on day two did not show robust FTO expression in Gsk-3 DKO
ESCs compared to day 14 (Figure 15 A). Therefore, we concluded that optimal FTO
expression is seen in cells grown for 14 days in the absence of LIF in normal ESC media,
a-tubulin was used as a protein loading control.
To take this optimization process one step further, we wanted to determine if we
could alter the medium conditions and receive the same, if not better, results. WT and
Gsk-3 DKO ESCs were grown in N2B27 media supplemented with 10% FBS
63


(Appendix). These cells were grown in the absence of LIF for a total of four days. Each
day, protein was isolated from a subset of cells. At the end of the four-day experiment,
western blots were performed to determine which day had the optimal FTO expression in
WT and Gsk-3 DKO ESCs (Figure 15B). The a-tubulin western was used as a protein
loading control and the Gsk-3a/p western was performed to validate sufficient knockout
of both isoforms in the Gsk-3 DKO cell. Unfortunately, FTO expression in Gsk-3 DKO
ESCs was low in comparison to the Gsk-3 DKO ESCs grown in complete ESC media in
the absence of LIF for 14 days. This allowed us to conclude that the most optimal
medium condition for robust FTO expression is growing cells in complete ESC media in
the absence of LIF for 14 days. By optimizing these cellular conditions, it allowed us to
move forward and proceed on downstream experiments.
64


A
WT
DKO
50 kD _ IB: FTO
50 kD IB: a-Tubulin
WT-VitC WT+VitC DKO-VitC DKO+Vit C P110*-Vit C P110* +Vit C
50 kD
IB: FTO
50 kD
IB: a-Tubulin
a: 51 kD
|3: 47 kD *
IB: Gsk-3a/p
Figure 13. FTO Expression in ESCs Cultured in Complete ESC Media in
the Absence of LIF for 14 Days. FTO expression was observed in (A and B)
WT, Gsk-3 DKO and (B) pi 10* ESCs that were grown in the absence of LIF
for 14 days. (B) A subset of WT, Gsk-3 DKO and pi 10* ESCs were grown
with the addition of 50ug/mL Vitamin C. Protein band sizes are indicated on
the left. The a-tubulin western was used as a protein loading control and the
Gsk-3 a/p western was performed to validate sufficient knockout of both
isoforms in the Gsk-3 DKO cell.
65


A 7
6
5
4
cr
K
3
2
1
0
n+LIFdl4 -LIFdl4
c
12 'I
Esrrb Nanog >Oct4
D
1.2 -
10 -
8
6 -i
4 -i
2 i
WT -LIFd 14
1
0.8
DKO -LIFd 14
WT -LIFdl4 DKO -LIFd 14
FTO
Figure 14. Pluripotency Gene Expression in ESCs Grown in the Absence
of LIF for 14 Days. Comparison of (A) Nanog and (B) Esrrb gene
expression in WT and Gsk-3 DKO cells grown in the presence and absence
of LIF for 14 days. (C) Quantification of Nanog, Oct4 and Esrrb and (D)
FTO gene expression in WT and Gsk-3 DKO cells grown in the absence of
LIF for 14 days.
66


A
Complete ESC Media
-LIF Day 2 -LIF Day 14
WT DKO WT DKO
50 kD IB: FTO
50 kD IB: a-Tubulin
B
N2B27 Media with 10% FBS
WT-UF Gsk-3 DKO-LIF
Day 1 Day 2 Day 3 Day 4 Day 1 Day 2 Day 3 Day 4
50 kD IB: FTO
50 kD mm ------------ IB: a-Tubulin
a: 51 kD
P: 47 kD
IB: Gsk-3a/P
Figure 15. Optimizing ESC Media Conditions for Detection of Robust FTO
Protein Expression. (A) Testing time frame of WT and Gsk-3 DKO ESCs
cultured complete ESC media in the absence of LIF. Protein isolated on day two
of no LIF experiment was compared to cells on day 14. (B) WT and Gsk-3 DKO
ESCs were grown in N2B27 media with 10% FBS. FTO protein levels were
observed in cells that were isolated each day during four-day experiment. The a-
tubulin western was used as a protein loading control and the Gsk-3 a/p western
was performed to validate sufficient knockout of both isoforms in the Gsk-3
DKO cell.
67


Effect of Gsk-3 Inhibition on FTP Protein Expression
Since the addition of a Gsk-3 inhibitor to ESC media helps to promote
pluripotency in the absence of LIF, we wanted to ask whether Gsk-3 inhibitors could
increase FTO levels in WT ESCs similar to the FTO levels in Gsk-3 DKO ESCs. To
determine the effects of Gsk-3 inhibition on FTO levels, WT cells that were grown
without LIF for 14 days were then treated with differing concentrations of SB-415,286 or
LiCl on days 15 and 16. On day 16, protein was isolated from cells and FTO and a-
tubulin western blots were performed. Results indicated that increasing amounts of SB-
415,286 increased FTO expression, with the most robust expression seen in cells treated
with 15uM SB (Figure 16A). The LiCl treated cells showed a similar trend in increasing
FTO expression with increasing LiCl concentrations. The cells treated with lOmM LiCl
showed the most robust FTO expression (Figure 16B). These data show that FTO protein
levels in WT ESCs can be rapidly increased by Gsk-3 inhibition in a dose-dependent
fashion.
68


A
OuMSB 5uM SB lOuM SB 15uM SB
50 kD
IB: FTO
50 kD
IB: a-Tubulin
B
OmM LiCI ImM LiCI 2mM LiCI 5mM LiCI lOmM LiCI
50 kD
IB: FTO
50 kD
' IB: a-Tubulin
Figure 16. Effect of Gsk-3 Inhibition on FTO Protein Expression in WT
ESCs. WT cells were cultured in complete ESC media in the absence of LIF for
16 days. On days 14 and 15, cells were treated with varying concentrations of
(A) SB-415,286 or (B) LiCI. Protein was isolated on day 16 to observe the effect
of Gsk-3 inhibition on FTO expression in a dose-dependent manner. The a-
tubulin westerns were used as a protein loading control.
69


Effect of Vitamin C on Pluripotency
Vitamin C plays a role in promoting pluripotency in the generation of iPSCs from
somatic cells (Bar-Nur et al., 2014). While Vitamin C is best known as an anti-oxidant, it
has also been shown to serve as a co-factor for alpha-ketoglutarate-dependent
dioxygenases, such as FTO. Therefore, we wondered if Vitamin C is promoting
pluripotency by enhancing FTO mRNA demethylase activity. To answer this question,
WT, Gsk-3 DKO, and pi 10* ESCs were grown in the absence of LIF for 14 days. Every
day during the 14-day experiment, a subset of the cells was treated with a relatively low
dose of Vitamin C (50ug/mL). Over the 14-day experiment, the cellular morphology
changed greatly in WT plus Vitamin C compared to without Vitamin C (Figure 17A).
The morphology of Gsk-3 DKO and pi 10* ESCs stayed relatively consistent over the 14-
day experiment (Figures 17B-C).
On day 14, protein was isolated and western blots were performed to observe
FTO expression (Figure 13B). Results indicated there was no change in FTO protein
levels between the WT and Gsk-3 DKO ESCs +/- Vitamin C. However, pi 10* cells
treated with Vitamin C revealed robust FTO expression compared to without Vitamin C.
The robust expression was almost identical to Gsk-3 DKO ESCs +/- Vitamin C. The a-
tubulin western served as a loading control and the Gsk-3 a/p western was performed to
validate sufficient knockout of both isoforms in the Gsk-3 DKO cells.
On day 14, RNA was also isolated from WT, Gsk-3 DKO, and pi 10* ESCs +/-
Vitamin C. cDNA was synthesized and qPCR was performed to quantify Nanog, Oct4,
and Esrrb expression levels. For the DKO cells that were treated with 50ug/mL Vitamin
C for 14 days in the absence of LIF, there was no change between DKO plus and minus
70


Vitamin C in Nanog, Esrrb, or Oct 4 gene expression (Figure 18A). This suggests that
Vitamin C cannot further enhance the pluripotency seen in Gsk-3 DKO ESCs.
Additionally, we were curious to observe FTO expression in Gsk-3 DKO ESCs cultured
in the absence of LIF. Quantitative PCR revealed no change in FTO expression (Figure
18B). However, WT and pi 10* ESCs treated with Vitamin C revealed an increase in
Nanog, Esrrb, and Oct4 expression compared to cells lacking the Vitamin C treatment
(Figure 19). Interestingly, Nanog expression in pi 10* ESCs treated with Vitamin C
revealed a 40-fold increase compared to WT ESCs treated with Vitamin C. This qPCR
data shows that Vitamin C alone can enhance expression of pluripotency factors, yet this
enhancement is not seen in Gsk-3 DKO ESCs.
71


A WT-Vitamin C WT+Vitamin C
B DKO-Vitamin C DKO+VitaminC
C P110* -Vitamin C P110* +Vitamin C
Figure 17. Morphology of WT, Gsk-3 DKO and pi 10* ESCs
Grown in the Presence and Absence of Vitamin C. (A) WT,
(B) Gsk-3 DKO and (C) pi 10* were grown in complete ESC
media without LIF for 14 days. A subset of cells was treated
with 50ug/mL Vitamu^ everyday during 14-day experiment.


A
1.4
Esrrb ENanog HOct4
Figure 18. Quantified Gene Expression in Gsk-3 DKO ESCs
Cultured in the Absence of LIF with the Addition of Vitamin
C. Gsk-3 DKO ESCs were grown in complete ESC media in the
absence of LIF for 14 days. A subset of cells was treated with
50ug/mL Vitamin C everyday during 14-day period. (A) Esrrb,
Nanog and Oct4 expression in Gsk-3 DKO ESCs +/- Vitamin C.
(B) FTO expression in Gsk-3 DKO ESCs +/- Vitamin C.
73


60
Esrrb
Nanog I0ct4
Figure 19. Quantified Gene Expression in WT and pi 10* ESCs
Cultured in the Absence of LIF with the Addition of Vitamin C. WT
and pi 10* ESCs were grown in complete ESC media in the absence of LIF
for 14 days. A subset of cells was treated with 50ug/mL Vitamin C
everyday during 14-day period. Esrrb, Nanog, and Oct4 expression in WT
and pi 10* ESCs +/- Vitamin C.
74


CHAPTER V
DISCUSSION
Conclusions
Characterizing a IncRNA, LINC00486
Functional characterization of LINC00486
Successful cloning into the mammalian expression vector, pCAGEN, allowed for
numerous downstream characterizing experiments. RT-PCR with transcript specific
primers revealed the ability to over express each functional transcript in both N2A and
HEK293T cells. LINC00486 transcript expression was shown to be present in the
nucleus, which was predicted due to the commonly known association between IncRNAs
and transcription factors to initiate gene silencing. Presence of LINC00486 in the
cytoplasm may have been due to the overexpression of each transcript, which caused an
abundant amount of transcript to be found in both parts of the cell. However,
identification of this IncRNA in the nucleus allowed us to narrow our future experiments
on the nuclear isolate.
After validation of successful overexpression of LINC00486 Transcripts 1-3 in
mammalian cells, we were curious if endogenous transcript expression could be detected.
RT-PCR revealed our transcript specific primers could only detect endogenous Transcript
3 expression in HEK293T cells, suggesting the presence of this transcript in human
embryonic kidney tissue. These results led us to question in what other tissues these
LINC00486 transcripts were expressed. However, newly designed primers were required
to detect endogenous expression of Transcripts 1 and 2. Confirmation of successful
detection confirmed our prediction that three transcripts exist, each unique in sequence. A
75


panel of human tissue RNA allowed us to run multiple RT-PCR experiments to
determine the location of each expressed transcript. Robust Transcript 1 expression was
observed in salivary gland and colon tissue; Transcript 2 expression was observed in
prostate, salivary gland and spleen tissues; Transcript 3 expression was see in whole
brain, testes, salivary gland and bone marrow. Each transcript was expressed in different
types of tissues, with the exception of the salivary gland, once again supporting the idea
that three transcripts of LINC00486 exist. Next, we wanted to quantify the expression of
each transcript in each human tissue. Using a tissue that revealed no expression during
RT-PCR as the control for each transcript, quantitative PCR revealed a much deeper
understanding of transcript expression. Transcript 1 and 3 had the greatest amount of
expression in the testes while Transcript 2 had the most in the liver. Transcript 3 was also
high in the placenta and the salivary gland. Transcript 1 revealed high expression in the
salivary gland, the liver and the spinal cord, and Transcript 2 was high in the salivary
gland testes and prostate. By determining what tissues show the greatest amount of
expression of each transcript, we could now focus on specific tissues to continue
functionally characterizing this IncRNA.
Due to the large amount of expression of all three transcripts in the testes and high
Transcript 3 expression in the placenta, we were curious to see if there was any
correlation between LINC00486 in germ cells and stem cells. Germ cells and human
embryonic stem cells (hESCs) have many genetic similarities required for developmental
processes. Therefore, we looked at transcript expression in hESCs and iPSCs. Results
revealed low transcript expression in both hESCs and iPSCs compared to the designated
tissue controls for each transcript, indicating no correlation in LINC00486 transcript
76


expression between hESCs, iPSCs, and human tissues involved in reproduction and
development.
Many IncRNAs have shown to have enhancer-like capabilities, which allows
them to regulate neighboring genes. We were curious to see if LINC00486 Transcripts 1-
3 had the ability to regulate flanking genes, TTC27 and LTBP1. Quantitative PCR
revealed modest changes in LTBP1 and TTC27 endogenous expression in HEK293T
cells transfected with each transcript. However, when each transcript was overexpressed
in N2A cells, LTBP1 expression was almost 2-folds higher in cells transfected with
Transcript 3. This result is suggestive that Transcript 3 has the capability to regulate its
neighbor, LTBP1.
Since the LINC00486 transcripts were originally identified in bipolar patients
through lithium treatment, we were interested in determining transcript expression in
other human cells treated with Gsk-3 inhibitors: LiCl and SB-415,286. HEK293T cells
treated with either ImM LiCl or 30uM SB-415,286 for 24 hours revealed relatively no
change in Transcripts 2 and 3 in comparison the untreated cells. However, Transcript 1
revealed a slight increase by half a fold in cells treated with LiCl and almost 2.5-fold
increase in cells treated with SB-415,286. These results suggest an association between
LINC00486 Transcript 1 and the Gsk-3 pathway, which was predicted due to means by
which LINC00486 was discovered.
After establishing a baseline of transcript expression in mammalian cells treated
with Gsk-3 inhibitors, we were interested in comparing these results to the original
lymphoblast cells from the BPD patient (3188) and their unaffected sibling (3189) that
were treated with either LiCl or SB-415,286. qPCR results indicated a 2-fold increase in
77


Transcript 3 expression in 3188 (BPD) lymphoblast cells treated with 30uM SB-415,286
for 24 hours. Results also indicated no change in LTBP1 and TTC27 expression in 3188
(BPD) cells treated with luM LiCl. Finally, we wanted to take a step backwards and look
at the endogenous expression pattern in untreated lymphoblast cells from 3188 (BPD)
and 3189 (unaffected). LTBP1, TTC27, and Transcript expression were all reduced in
3189 (unaffected) compared to 3188 (BPD). In conclusion, this transcriptional data
allowed us to further characterize LINC00486 by gaining insight into the functional
capabilities of Transcripts 1-3 as well as its regulation in Gsk-3 signaling pathways.
Biochemical characterization of LINC00486
In order to biochemically characterize LINC00486, it was imperative to determine
what proteins are associated with this IncRNA. Research has shown many IncRNAs have
an association with the PRC2 complex to regulate transcriptional inhibition. Therefore,
we wanted to investigate if this novel IncRNA followed the same trend.
Immunoprecipitation of HA-tagged EZH2 in transfected N2A cells revealed the
successful pull-down of Transcript 2. These results indicate a physical association
between EZH2 and Transcript 2. Identical results were seen in the EZH2 IP in HEK293T
cells, which revealed endogenous Transcript 2 immunoprecipitation from HA-tagged
EZH2. Since EZH2 is a protein within PRC2, which binds to RNA with high affinity,
these results imply LINC00486 Transcript 2 is physically associated with PRC2 and
therefore may have a functional role in genomic silencing. Further investigation is
required to determine the mechanisms behind this interaction and transcriptional
regulation. However, one possibility is the ability for LINC00486 to tether PRC2 to the
site of transcription. Once PRC2 reaches the site, the chromatin structure is remodeled to
78


induce genomic inhibition. A more in depth examination of this process is required to
fully understand this mechanism.
Many other characterized IncRNAs have also been shown to be associated with
the PRC1 complex, which maintains genomic silencing through the binding of PRC2.
Additionally, the synergistic activity of PRC 1 and PRC2 has been shown to play a role in
regulating pluripotency in mouse ESCs, specifically through the individual Cbx proteins.
PRC1 is comprised of different Cbx subunits, each of which has a unique function. We
were interested in determining which transcript associates with each Cbx subunit in hope
of determining the involvement of LINC00486 Transcripts 1-3 in the unique Cbx
functions. RT-PCR on Flag-tagged Cbx immunoprecipitations of Transcripts 1-3 revealed
varying amounts of Transcripts 1 and 3 that were immunoprecipitated using different Cbx
proteins. These results suggest Transcripts 1-3 are physically associated with PRC1
through different Cbx proteins. Transcript 1 had the most robust expression after the
Cbx6 IP and weaker expression for Cbx2, Cbx4 and Cbx7. Additionally, Transcript 3 had
the most expression after the Cbx2, Cbx6 and Cbx8 IPs. Cbx7 is considered the main
Cbx subunit that maintains pluripotency. Cbx7 accomplishes this through repression of
Cbx2, Cbx4 and Cbx8, which prevents premature differentiation. Recent findings have
suggested that Cbx6 works alone and does not interact with any other PRC1 subunit. It is
believed to be recruited to genomic targets by different ncRNAs for an independent
function. Therefore, LINC00486 Transcript 1 may recruit Cbx6 to target genes. The
function of Cbx6 remains to be established therefore a further investigation is required to
reveal the purpose of the association between Transcript 1 and Cbx6. However, due to
tight association between Transcript 3 and subunits Cbx2 and Cbx8, it can be suggested
79


that Transcript 3 may have a role in cellular differentiation through PRC 1-dependent
mechanisms, which remains to be determined.
After determining the physical associations between each LINC00486 transcript
and the Polycomb Repressive Complexes, it is important to return back to the BPD
patient and their unaffected sibling to determine if LINC00486 is actually directly related
to BPD.
Involvement of Gsk-3 in Stem Cell Pluripotency
Our data suggests Gsk-3 plays an important role in stem cell differentiation
through the regulation of the m6A modification on mRNA. Due to the similar
morphological phenotypes between Mettl3 KO and Gsk-3 DKO ESCs, we hypothesized
that Gsk-3 activity regulates the mRNA m6A modification process through the
phosphorylation of FTO, the m6A demethylase. Phosphorylation of FTO would cause
inhibition, resulting in an increased presence of m6A on mRNA, resulting in the tagged
pluripotency mRNAs to be degraded. This leads to the differentiation of ESCs. However
when Gsk-3 is inhibited, FTO levels are raised, resulting in reduced m6A levels and a
maintained pluripotent state in ESCs.
When WT ESCs are cultured in LIF, pluripotency is promoted, but FTO levels are
also kept high. Therefore, in order to determine if our hypothesis was correct, we needed
to optimize the cell culturing conditions in order to separate pluripotency from the
mechanistic regulation by FTO. ESCs cultured in complete ESC media in the absence of
LIF for 14 days resulted in the best observation of FTO protein expression. When WT,
Gsk-3 DKO, and pi 10* ESCs were grown in these conditions, FTO expression in WT
cells was greatly reduced compared to the robust expression seen in Gsk-3 DKO cells.
80


These results support our hypothesis, suggesting that when Gsk-3 is knocked out, the
FTO protein is accumulated, resulting in reduced m6A levels and maintained pluripotency
in ESCs. However, when Gsk-3 levels are normal (as seen in WT ESCs), FTO levels are
low due to Gsk-3 phosphorylating FTO and tagging FTO for degradation. The location of
this phosphorylation on FTO remains to be determined. Additionally, pi 10* ESCs
revealed low FTO expression levels, which were almost identical to the levels in the WT
ESCs, indicating this mechanism might be occurring above PI3K. Further investigation is
required.
To confirm the sustained pluripotency in Gsk-3 DKO ESCs, we compared the
gene expression levels of Master pluripotency transcription factors in WT and Gsk-3
DKO ESCs. We observed increased expression of Nanog, Esrrb, and Oct 4 in Gsk-3
DKO ESCs, which validates the pluripotent state in these cells. In addition, FTO
expression in Gsk-3 DKO ESCs grown without LIF for 14 days was reduced compared to
WT. This supports our hypothesis suggesting this effect occurs at the post-translationally,
rather than transcriptionally, hence why we saw robust FTO protein expression in the
western blot results. Additionally, treatment of WT cells with a Gsk-3 inhibitor rescued
FTO protein expression, furthering the notion that Gsk-3 activity regulates FTO levels
post-translationally. Optimal concentrations for FTO rescue were the treatment of lOmM
LiCl or 15uM SB for two days.
To confirm inhibition of Gsk-3 results in increased translational levels of FTO
and therefore reduced m6A modifications on mRNA, total RNA was isolated from WT
and Gsk-3 DKO ESCs cultured in complete ESC media in the absence of LIF for 14
days. mRNA was immunoprecipitated from the total RNA using oligo dT beads.
81


Collected mRNA was digested to retrieve single nucleotides in order to perform Liquid
Chromatography Tandem Mass Spectrometry (LC MS/MS). LC MS/MS data revealed
50% reduction in Gsk-3 DKO ESCs compared to WT, validating our hypothesis that
inhibited forms of Gsk-3 would result in reduced mRNA methylation through the
regulation of FTO (data not shown; K. Faulds, C. Phiel; Personal Communications).
Throughout the 14-day experiment, we observed the phenotypes of the subset of
ESCs that were treated with Vitamin C. There was an obvious morphological change in
WT ESCs treated with Vitamin C compared to those grown in the absence of Vitamin C,
indicating Vitamin C is causing a pluripotent effect on these cells. However, this
pluripotent morphology did not affect FTO expression at the transcriptional or
translational level in Gsk-3 DKO ESCs, which was indicated by no change in FTO
protein or gene expression between the Vitamin C treatments in Gsk-3 DKO ESCs. We
originally hypothesized Vitamin C, a cofactor of FTO, would enhance pluripotency in
Gsk-3 DKO ESCs by assisting the enzymatic reaction of FTO, resulting in high FTO
protein expression. However, these results suggest this antioxidant supplement does not
have an effect on Gsk-3 DKO ESCs. Interestingly, FTO protein expression in pi 10*
ESCs treated with Vitamin C was greatly increased compared to the cells lacking
Vitamin C. This robust FTO expression was almost identical to the amount of FTO
expression observed in Gsk-3 DKO ESCs without LIF for 14 days. The effect of Vitamin
C on pi 10* cells through FTO protein expression is blatantly obvious, yet the mechanism
behind why this is observed is unknown. Additionally, when we quantified pluripotent
gene expression (Nanog, Oct4 and Esrrb) in pi 10* and WT ESCs treated with and
without Vitamin C (no LIF), there was increased gene expression in Nanog, Oct 4 and
82


Esrrb in the WT and pi 10* ESCs treated with Vitamin C. Furthermore, there was no
change in Oct 4 and Esrrb expression levels between WT and pi 10* ESCs with Vitamin
C. However, there was a significant increase in Nanog expression in pi 10* ESCs with
Vitamin C compared to WT ESCs with Vitamin C. These results indicate Vitamin C
alone can enhance expression of pluripotency factors, yet this enhancement is not seen in
Gsk-3 DKO ESCs.
Final Words
The large involvement of Gsk-3 regulatory activity in numerous signaling
pathways indicates how important this kinase is in modulating gene expression, DNA
methylation and m6A mRNA modifications. When Gsk-3 expression is altered, it has
been shown to play a role in numerous human diseases such as diabetes and bipolar
disorder. Therefore, understanding how this kinase works in these systems is extremely
important for understanding disease and creating pharmaceutical therapeutics. In this
thesis, we have uncovered another layer of complexity for the regulatory role of Gsk-3 in
mRNA methylation to sustain stem cell pluripotency. However, the mechanistic
understanding behind these modifications is still relatively unknown and will remain our
main focus as we move forward.
83


CHAPTER VI
FUTURE DIRECTIONS
Characterizing a long non-coding RNA, LINC00486
We propose to perform gain-of-function experiments by over-expressing
LINC00486 transcripts 1-3 in human and mouse cells. A microarray assay will follow
over-expression to observe the effect on downstream genes. This will help to determine
which genes are associated with LINC00486 as well as what cellular pathway
LINC00486 is involved in. Additionally, we propose to perform loss-of-function
experiments to determine what the downstream effect is when the entire LINC00486
gene is knocked down, as well as each individual transcript. siRNAs or CRISPR will be
used to perform the knockdown/knockouts. After each manipulation, a microarray assay
will be performed to observe effect on downstream genes.
In addition to these gain- and loss-of-function experiments, we will determine if
there is a mouse homolog of LINC00486. We will accomplish this by designing mouse
primers to determine if we can detect endogenous LINC00486 in mouse cells through
RT-PCR. We also propose to determine exactly where on each Cbx protein LINC00486
Transcripts 1-3 binds by using Cbx proteins containing different mutations through out its
domain. By determining exactly where each transcript binds onto the Cbx proteins, it will
hopefully give us insight into the function of that transcript and LINC00486 in general.
Role of Gsk-3 in Stem Cell Pluripotency
So far, we have characterized the relationship between m6A, FTO and
pluripotency. To further investigate our findings of Gsk-3 regulating stem cell
pluripotency, we propose to determine the Gsk-3 phosphorylation site on FTO. This will
84


be accomplished by using mass spectrometry on WT and Gsk-3 DKO ESCs to compare
phosphorylation sites and identify those that are Gsk-3 dependent.
In addition, we will perform mRNA half-life studies by treating WT and Gsk-3
DKO ESCs with a-amanatin, followed by qPCR. This experiment needs to be repeated
using the 14-day no LIF paradigm that I have developed. This will allow us to compare
the rates of mRNA degradation between WT and Gsk-3 DKO ESCs and provide insight
into the overall mechanism of Gsk-3 regulated m6A modifications.
Eventually, we will perform m6A-seq to identify the numerous other mRNAs
whose m6A modifications are affected by deletion or inhibition of Gsk-3. Finally, we will
look at m6A levels in different human cells to determine if the Gsk-3 regulation of m6A
modifications is relevant in disease.
85


REFERENCES
Banito, A., & Gil, J. (2010). Induced pluripotent stem cells and senescence: learning the
biology to improve the technology. EMBO Rep, 11(5), 353-359. doi:
10.103 8/embor.2010.47
Bar-Nur, O., Brumbaugh, J., Verheul, C., Apostolou, E., Pruteanu-Malinici, I., Walsh, R.
M., . Hochedlinger, K. (2014). Small molecules facilitate rapid and
synchronous iPSC generation .Nat Methods, 77(11), 1170-1176. doi:
10.103 8/nmeth. 3142
Batista, P. J., Molinie, B., Wang, J., Qu, K., Zhang, J., Li, L., . Chang, H. Y. (2014).
m(6)A RNA modification controls cell fate transition in mammalian embryonic
stem cells. Cell Stem Cell, 75(6), 707-719. doi: 10.1016/j.stem.2014.09.019
Brueckner, F., & Cramer, P. (2008). Structural basis of transcription inhibition by
[alpha]-amanitin and implications for RNA polymerase II translocation. Nat
Struct Mol Biol, 75(8), 811-818.
Burdon, T., Smith, A., Savatier, P. (2002). Signalling, cell cycle and pluripotency in
embroynic stem cells. TRENDS in Cell Biology, 72(9).
Caricasole, A., Copani, A., Caruso, A., Caraci, F., Iacovelli, L., Sortino, M. A., . .
Nicoletti, F. (2003). The Wnt pathway, cell-cycle activation and P-amyloid: novel
therapeutic strategies in Alzheimer's disease? Trends Pharmacol Sci, 24(5), 233-
238. doi: 10.1016/s0165-6147(03)00100-7
Chen, J., Liu, J., Chen, Y., Yang, J., Chen, J., Liu, H., . Pei, D. (2011). Rational
optimization of reprogramming culture conditions for the generation of induced
pluripotent stem cells with ultra-high efficiency and fast kinetics. Cell Res, 21(6),
884-894. doi: 10.1038/cr.2011.51
Chu, C., Qu, K., Zhong, F. L., Artandi, S. E., & Chang, H. Y. (2011). Genomic maps of
long noncoding RNA occupancy reveal principles of RNA-chromatin
interactions .Mol Cell, 44(4), 667-678. doi: 10.1016/j.molcel.2011.08.027
Cifuentes-Rojas, C., Hernandez, A. J., Sarma, K., & Lee, J. T. (2014). Regulatory
interactions between RNA and polycomb repressive complex 2. Mol Cell, 55(2),
171-185. doi: 10.1016/j.molcel.2014.05.009
Coghalin, M. P., Culbert, A., Cross, D., Corcoran, S., Yates, J., Pearce, N., Raush, O.,
Murphy, G., Carter, P., Cox, L., Mills, D., Brown, M., Haigh, D., Ward, R.,
Smith, D., Murray, K., Reith, A., Holder, J. (2000). Selective small molecule
inhibitors of glycogen synthatse kinase-3 modulate glycogen metabolism and
gene trranscription. Chemistry and Biology, 7(10).
86


Full Text

PAGE 1

! THE REGULATORY ROLE OF GSK 3 IN DNA AND RNA METHYLATION by JENNIFER NICOLE EGELSTON B.S. California Polytechnic State University 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillm ent o f the requirements for the degree of Master of Science Integrative Biology Program 2015

PAGE 2

! "" ! 2015 JENNIFER NICOLE EGELSTON ALL RIGHTS RESERVED

PAGE 3

! """ ! This thesis for the Master of Science degree by Jennifer Nicole Egelston has been a pproved for the Integrative Biology Program by John Sw a llow Chair Christopher Phiel Advisor Amanda Charlesworth Xiaojun Ren November 1 9 th 2015

PAGE 4

! "# ! Egelston, Jennifer Nicole (M.S Biology ) The Role of Gsk 3 in DNA and RNA Methylation Thesi s directed by Assistant Professor Christopher Phiel. ABSTRACT Gsk 3 (Glycogen synthase kinase) is a key protein kinase that has a prominent role in multiple intracellular pathways. Recently, Gsk 3 activity was found to be important for the regulation of DNA methylation in mouse embryonic stem cells (ESCs) when Gsk 3 and Gsk 3 are genetically deleted. In addition, it was observed that the addition of lithium, a known inhibitor of Gsk 3 and a common bipolar disorder (BPD) treatment, also reduces DNA methy lation. These studies led us to ask whether DNA methylation was changed in cells from patients with BPD. This led to the identification of an uncharacterized long non coding RNA (lncRNA), LINC00486, whose DNA methylation patterns are different in BPD patie nts compared to their unaffected siblings. Three distinct transcripts have been annotated for LINC00486. Our goal was to functionally and biochemically characterize these novel lncRNAs. Cytoplasmic and nucleic l ocalization experiments determined that LINC0 0486 is present in the nucleus, which is consistent with other characterized lncRNAs and their role in transcriptional regulation. RNA immunoprecipitations (RIP) have identified physical interactions between LINC00486 and both Polycomb Repressive Complexes (PRC1 and PRC2), which play roles in chromatin remodeling to induce transcri ptional repression. Interestingly PRC1 subunits are associated with pluripotency maintenance in ESCs.

PAGE 5

! # ! These findings suggest LINC00486 may play a role in transcriptional silencin g which in turn could sustain stem cell pluripotency. In addition to the regulation of D NA methylation by Gsk 3, we continue to investigate whether Gsk 3 regulates the levels of mRNA methylation of adenosines (referred to as m 6 A). m 6 A is the most abundant mRNA modification and until recently, the functional significance remained elusive. However, recent findings have suggested that the m 6 A tag is a mechanism directing mRNA degradation. Mettl3 has been shown to be the methyltransferase enzyme responsible fo r adding m 6 A to mRNA, while FTO has been shown to be the demethylase in this reversible process. Recent studies have established a role for m 6 A in maintaining pluripotency in embryonic stem cells; reduced m 6 A levels lead to longer mRNA half lives of plurip otency associated factors. Due to the phenotypic similarities between Gsk 3 DKO and Mettl3 KO ESCs and their inability to differentiate we have investigated whether Gsk 3 is regulating m 6 A levels. Specifically, we have found that deletion of Gsk 3 results in high levels of FTO. The discovery that the m 6 A modification is controlled by Gsk 3, a multi faceted regulator involved in numerous pathways, would have profound consequences in understanding the regulatory mechanisms behind the balance between pluripot ency and differentiation in ESCs. The form and content of this abstract are approved. I recommend its publication. Approved: Christopher Phiel

PAGE 6

! #" ! TABLE OF CONTENTS CHAPTER I. INTRODUCTION .1 Gsk 3 Overview . . .. 1 Specific Aims.. ...4 Characterizing a Novel LncRNA, LINC00486...4 Involvement of Gsk 3 in Stem Cell Pluripotency...4 II. CHARACTERIZING A NOVEL LNCRNA, LINC00486............... ... ....... .... 6 Introduction 6 Gsk 3 Regulates DNA Methylation ..... ....... ... ... . 6 Non Coding RNA Overview . ... . .......... 7 Role of Long Non Coding RNA in Gene Regulation .... ...... ...8 LncRNAs Associated with Polycomb Repressive Complex es .. .. 9 Materials and Methods .....13 Over Expression of LINC00486 Transcripts in Mammalian Cells ...13 Gibson Assembly Cloning of LINC00486 Transcripts 1 3 and LncRNA HOTAIR into pCAGEN . 13 N2A and HEK293T Mammalian Cel l Transfection Using PEI.. ...15 mESC Tr ansfection Using PEI.. 16 RNA Isolation 16 cD NA Synthesis.17 T A Cloning 17 RT PCR.17

PAGE 7

! #"" ! Thermocycler Program..17 qPCR ..17 RNA Immun op recipitation18 Biotinyl ati on of Oligos..18 RNA Pull Down with Biotinylated Oligos 18 EZH2 Over Expression and Immunoprecipitation 20 Cbx Protein Over Expression and Immunoprecip i tation ...22 R esult s..22 Over Expression of LINC00486 in Mammalian Cells... .......... ............ .....23 Endogenous Expression In Human Cells and Tissu e s . ..28 RNA I mmunoprecipitation 39 III. INVOLVEMENT OF GSK 3 IN STEM CEL L PLURIPOTENCY...............44 Introd uct ion..44 Pluripotency in Mous e Em bryonic Stem Cells..44 Effect of LIF on Stem Ce ll Pluripotency...45 Effect of Gsk 3 Inhibition on S tem Cell Pluripotency ...46 m 6 A mRNA Mo dification..48 Materials a nd Methods.53 RNA Immunoprec ipitation and ELISA .53 m 6 A Immunoprecipitation .53 m 6 A ELISA Using Qu bit .. 54 m 6 A ELISA (Based Off Epigentek P rotocol ) ... .. ... 56 Observing Protein Expression 57

PAGE 8

! #""" ! Western Blotting 57 R esults .58 Pluripotency Gene Expression in Immunoprecipitated m 6 A RNA 59 Eff ect of LIF on mESC Prot ei n and Gene Expression..62 Effect of GSk 3 Inhibition on FTO Protein Expression68 Effect of Vita min C on Pluripotency.70 IV. DISCUSSION ..75 Conclusions ..75 Characterizing a Novel LncRNA, LINC00486 .75 Functional Characterization ...75 Biochemical Characterization 78 Involvement of Gsk 3 in Stem Cell Pluripotency ..80 Final Words83 V. FUTURE D IRECTIONS .84 REFER EN CES86 APPE NDIX ..92 A. Equa tio ns..92 B. Buffer Reag ents...92 C. Cellular Mediums .92 D. Supplemen tal Da ta...93

PAGE 9

! "$ ! LIST O F TABLES Table 1 Tissues with robust transcript expression and designated tissue controls .....34 2. Summarized results of Cbx protein immunoprecipitations .......43

PAGE 10

! $ ! LIST OF FIGURE S Figure 1 Verification of Successful Cloning of LINC00486 Transcripts 1 3 .. ................... ....... ...26 2 Verification of Over Expression and Localization of LINC00486 Transcripts 1 3 in N2A Cells .. .. ................................................ .. ..................................................... ............... 27 3. Endogenous LINC00486 Transcript Expression in HEK293T Cells .................. .. .. ... ... 32 4. Detecting Presence of LINC00486 Transcript Expression in Human Tissue ... .....33 5. Quantification of Endogenous LINC00486 Transcript Expression in Human Tissues and Cells .35 6. Endogenous LTBP1 and TTC27 Gene Expression in HEK293T and N2A Cells ... .. 36 7. Effect of Gsk 3 Inhibition on Endogenous LINC00486 T ranscript Expression in HEK293T Cells.. 37 8. LINC00486 Transcript and Flanking Gene Expression in BPD Patient and Unaffected Sibling... .38 9. EZH2 Immunoprecipitation of LINC00486 Transcripts 1 3 ... ..42 10 Cbx Protein Immunoprecipitation of LINC00486 Transcripts 1 3 ..............43 11 m 6 A mRNA Gene Expression in WT and Gsk 3 DKO ESCs... .. .6 1 12 FTO Protein Expression in WT and Gsk 3 DKO ESCs Cultured in Complete ESC Media in the Presence of LIF . 61 13 FTO Expression in ESCs Cultured in Complete ESC Media in the Absence of LIF for 14 Days... ...65 14. Pluripotency of Gene Expression in ESCs Grown in the Absence of LIF for 14 Days..................................................... ..... 66 15 Optimizing ESC Media Conditions for Detection of Robust FTO Protein Expression......... .67

PAGE 11

! $" ! 16. Effect of Gsk 3 Inhibition on FTO Protein Expression in WT ESCs. 69 17. Morphology of WT, Gsk 3 DKO and p110* ESCs grown in the presence and absence of Vitamin C ......72 18. Quantified Gene Expression in Gsk 3 DKO ESCs cultured in the absence of LIF with the add ition of Vitamin C.... ..73 19. Quantified Gene Expression in WT and p110* ESCs Cultured in the Absence of LIF with the Addition of Vitamin C .....74

PAGE 12

! $"" ! LIST OF ABBREVIATIONS 2i Dual inhibitor cock tail 3' UTR 3' Untranslated Region 3188 Bipolar Disorder Patient 3189 Unaffected Sibling AGi Ascorbic Acid and a Gsk 3 inhibitor supplement Alkbh5 AlkB Homolog 5 BMP2 Bone Morphogenetic Protein 2 BMP4 Bone Morphogenetic Protein 4 BPD Bipolar Disorder ChIRP Chromatin Immuno RNA Precip it ation ENCODE Encyclopedia of DNA Elements Erk1 Extr acellular Regulated Kinase 1 Erk2 Extracellular Regulated Kinase 2 ESC Embryonic Stem Cells Esrrb Estrogen Related Receptor Beta EZH2 Enhancer of Zeste 2 FGF4 Fibroblast Growth Factor 4 FTO Fat and Obesity Gene Gsk 3 Glycogen Synthase Kinase DKO Double Knockout H2AK119 Histone 2A Lysine 119 H3K27me3 Trimethylation on Histone 3 on Lysine 27

PAGE 13

! $""" ! HEK293T Human Embryonic Kidney Cells \ hESC Human Embryonic Stem Cells HOTAIR HOX Transcript Antisense RNA iPSC Induced Pluripotent Stem Cells Klf4 KrŸppel like factor 4 LC MS/MS Liquid Chromatography Tandem Mass Spectrometry LiCl Lithium LIF Leukemia Inhibitory Factor lincRNA Long Intergenic Non coding RNA lncRNA Long Non coding RNA m 6 A N 6 methyladensoise m 6 A seq N 6 methyladensoise sequencing MEK Mitoge n Activated and Extracellular Regulated Kinase Inase mESC Mouse Embryonic Stem Cells Mettl3 Methyl Transferase like 3 miRNA MicroRNA mRNA messenger RNA N2A Neuroblastoma Cells ncRNA Non coding RNA NSC Neural Stem Cells p110* Myristoylated form of p11 0 subunit PDK1 Phosophoinositide Dependent Kinase 1 PDK2 Phosophoinositide Dependent Kinase 2

PAGE 14

! $"# ! PEI Polyethylenimine PI(3,4,5)P3 Phosphatidylinositol 3, 4, 5 triphosphate PI3K Phosphatidylinositol 3 Kinase PIC Polymerase II Preinitiation Complex PRC 1 Polycomb Repressive Complex 1 PRC2 Polycomb Repressive Complex 2 qPCR Q uantitative PCR RIP RNA Immunoprecipitation rRNA Ribosomal RNA snoRNA Small Nucleolar RNA snRNA Small Nuclear RNA STAT3 Signal Transduced and Activator of Transcription 3 Tcf3 T cell factor 3 TdT Terminal Deoxynucleotidyl Transferase tRNA Transfer RNA

PAGE 15

! % ! CHAPT ER I INTRODUCTION Gsk 3 Overview Glycogen synthase kinase (Gsk 3) is a constitutively active serine/threonine protein kinase that is a negative regulator of signal transduction through means of phosphorylation of downstream proteins (Popkie et al., 2010) Gsk 3 has been defined as the combined functional redundant activity of both Gsk 3 and Gsk 3 isoforms (Popkie et al., 20 10) Both isoforms have a preferred phosphorylation motif (S/T X X X S/T) and usually only phosphorylate substrates that are either "non primed" (unphosphorylated) or are "primed" (have been previously phosphorylated by another kinase) (Fiol, 1987) Recognition of the same cons ensus motif suggest s that both Gsk 3 isoforms perform overlapping functions through the phosphorylation of the same proteins, which in turn allows them to regulate the same signaling pathways (Doble, Patel, Wood, Ko ckeritz, & Woodgett, 2007) Gsk 3 has a large involvement in ma ny signal transduction pathways and therefore its ro le in diseases has been observed (Grimes, 2001) One of the most widely studied roles of Gsk 3 is in the Wnt pathway, in whi ch regulation of Gsk 3 plays an essential role in many different facets of embryonic development, including brain development (Caricasole et al., 2003) In addition, Gsk 3 also plays a pivotal role in the insulin pathway (Frame, 2001) However, while Gsk 3 regulates many pathw ays, the pathways do not activate one another (S. S. Ng et al., 2009) Researc h has shown that Gsk 3 is involved in a number of intracellular processes such as cell apoptosis, cell proliferation, gene expression, a nd regulates differentiation of embryonic stem cells and

PAGE 16

! & ! neural progenitors (Jope, 2003) Finally, altered Gsk 3 activity has been shown to play an important role in many human diseases such as bipolar disorder, schizophrenia, diabetes, and Alzheimer's disease (Popkie et al., 2010) Upstream of Gsk 3 in the insulin signaling pathway lies a lipid kinase, Phosphatidylinositol 3 Kinase (PI3K) (Osaki, 2004) PI3K is involved in numerous physiological processes including develo pment and proliferation. Inhibition of PI3K resulted in the inability to retain pluripotency in mouse embryonic stem cells (mESCs) (Welham et al., 2011) In response to growth factor stimulation, PI3K generates Pho sphatidylinosit ol 3, 4, 5 triphosphate (PI (3,4, 5) P3), which is essential for activation of Akt by phosophoinositide dependent kinase 1 (PDK) and PDK2 (Fang et al., 2000) Activated Akt phosphorylates Gsk 3 at an N terminal serine (S21 on Gsk 3 S9 on Gsk 3 ) causing rapid inhibition of Gsk 3 (Popkie et al., 2010) Gsk 3 inhibition results in numerous downstream effects including changes in gene and protein expression in various pathways, epigenetic alteration s, and increa s es in stem cell pluripotency and self renewal. In addition to the insulin signaling pathway, Gsk 3 plays an important role in the canonical Wnt signaling pathway (Wray et al., 2011) Many researchers believe this pathway is the key regulating system in stem cell fate determination due to the modulation of Gsk 3. In fact, a cocktai l of 2 small molecules (termed 2 i) is used to p romote stem cell pluripotency (Ying et al., 2008) One of the se molecules is a Gsk 3 inhibitor. Gsk 3 is a negative regulator of the Wnt pathway through the phosphorylation of catenin. This phosphorylation results in the ubiquitinylation and proteolysis of catenin (Katoh & Katoh, 2007) However, when Gsk 3 is inhibited, ther e is an

PAGE 17

! ! accumulation of catenin, which associates with different transcription factors, such as Tcf3 (T cell factor 3), to activate transcription (Wray et al., 2011) Therefore, inhibition of Gsk 3 allows for the activation of the Wnt pathway (Hedgepeth et al., 1997) Tcf3 is predominately found in stem cells and acts as tra nscription repressor. However, the accumulation of catenin directly interacts with Tcf3 to release this suppression at Tcf3 target genes that are additionally bound by pluripotency factors. This ablation of inhibition is believed to maintain pluripotency (Wray et al., 2011) A debate between which pathway, Wnt or i nsulin, controls the ability for stem cells to stay pluripotent is an ongoing quest among scientists. However, a poss ible third mechanism could be that these two pathways work in parallel when Gsk 3 is inhibited. Only ongoing in depth Gsk 3 research will help to determine which of these ideas is the most accurate. In order to functionally study Gsk 3, knocking down or i nhibiting Gsk 3 is one method to provide insight into the role of Gsk 3 in both DNA and RNA epigenetics. Numerous small molecule inhibitors are available, including insulin (through phosphorylation by Akt), l ithium (LiCl), and SB 415 286 LiCl inhibits Gsk 3 directly by displacing magnesium, a required cofactor (Klein, 1996) In contrast, SB 415 286 inhibits through an ATP competitive manner (Coghalin, 2000) While there are many small molecule inhibitors available to study Gsk 3 function, a drawbac k is the potential for off target effects (Phiel, 2001) Therefore, creating a more permanent inhibition of Gsk 3 provides a more confident assessment of Gsk 3. A myristolyated form of a PI3K subunit, p110 (termed p110*), promotes a constitutively active insulin pathway, resulting in constant inhibition of Gs k 3 (Popkie et al., 2010) Additionally, the genetic deletion (also called knockout) of both Gsk 3 isoforms in mESCs ( Gsk 3 double knockout; Gsk 3

PAGE 18

! ( ! DKO) allows for numerous downstream experiments permitting the func tional study of Gsk 3. Specific Aims The goal of this research project is to functionally and biochemically characterize a novel long non coding RNA, LINC00486, which was discovered from analysis of hypermethylation patterns in BPD patients This will be accomplished by determining in which cell types and tissues each human transcript of LINC00486 is expressed. Additionally, identifying protein RNA interactions will help to determine the func tion of LINC00486. The second part of this project is to examine the relationship between Gsk 3 and stem cell pluripotency by a novel mechanism: the regulation of the m 6 A modification on mRNA. We hypothesize the m 6 A tag is regulated by Gsk 3 through controlling the protein levels FTO, a demethylase. Inhibition of Gsk 3 would result in accumulated FTO protein levels, resulting in reduced m 6 A levels and maintenance of stem cell pluripotency. Identifying the regulatory role of Gsk 3 will provide a mechanistic understanding of stem cell differen tiation as well as directly li nk m 6 A modification s with Gsk 3 function. Characterizing a Novel LncRNA, LINC00486 Aim #1: Characterizing the expression of endogenous LINC00486 transcripts Aim #2: Biochemical characterization of LINC00486 transcripts Involvement of Gsk 3 in Stem Cell Pl uripotency Aim #1: Determine if the effect of Gsk 3 on stem cell pluripotency is through the regulation of mRNA methylation.

PAGE 19

! ) ! Aim #2: Investigate if Vitamin C and Gsk 3 inhibition work synergistically to promote stem cell pluripotency.

PAGE 20

! ! CHAPTER II CHARACTERIZING A NOVEL LNCRNA, LINC00486 Introduction Gsk 3 Regulates DNA Methylation Gsk 3 activity has been observed to regulate DNA methylation through altered expression of DNA methyltransferase Dnmt3a2. This was confirmed by knocking out both Gsk 3 isoforms (Gsk 3 DKO) in mouse ESCs, resulting in the reduction of DNA methylation and altered gene expression (Meredith et al., 2015; Popkie et al., 2010) Lithium mimics this reduced Gsk 3 activity in t wo ways: direct inhibition and indirect inhibition by increasing inhibitory phosphorylation of Gsk 3 (Jope, 2003) These two effects work in synergy to influence Gsk 3 regulated functions. Interestingly, lithium is commonly used as a mood stabilizer treatment for p atients with bipolar disorder (BPD) (Phiel, 2001) The exact mechanism behind how lithium serves to treat BPD is unknow n, but the mechanism may occur through Gsk 3 inhibition, which results in the regulation of transcription factors and a decrease in DNA methylation (Jope, 2003) Due to the inh ibitory effects lithium has on the DNA methylation patterns in Gsk 3 DKO ESCs and the clinical use of lithium in BPD patients, mouse neural stem cells (NSCs) were treated with lithium S imilar reductions in DNA methylation were observed in these cells The se r esults demonstrated that the effects of l ithium on DNA methylation extend to neural cells, and provoked an interest in investigating if DNA methylation was altered in patients with BPD. Due to the difficulties of acquiring neural cells from BPD patient s, we began by using lymphoblast cells obtained from two sets of siblings; in each set one sibling had BPD and the other sibling was unaffected. Methyl seq was performed

PAGE 21

! + ! on genomic DNA isolated from lymphoblasts, providing DNA methylation patterns of the entire genome for each individual. Several loci displayed hypermethylation in both BPD patients compared to their unaffected siblings. When the lymphoblast cells were treated with a therapeutic dose of lithium (1mM) results showed hypomethylation at sever al loci (data not shown; A. Popkie; Personal Communications) Among these, one locus had been hypermethylated in the BPD cells. This differentially methylated locus was where the novel lncRNA, LINC00486, was located. Sequencing data has indicated LINC00486 is found in human chromosome 2 and is flanked by genes TTC27 and LTBP1. A mouse homolog has yet to be discovered. Three distinct RNAs ar e predicted to be transcribed from the LINC00486 locus, each of different lengths and containing different exons. Non Coding RNA Overview Recent a dvances in genome wide analyse s have revealed that roughly 90% of the human geno me is transcribed, yet less than 3% of the genome consists of protein coding genes (Wu et al., 2013) The remaining genes are transcribed as nonc oding RNAs (ncRNAs), which resemble mRNA in length and splicing structures yet do not encode any proteins (Wu et al., 2013) It has been debated whether all of the ncRNA transcripts are functional due to their low expression levels and low evolutionary conservation. However, many functional ncRNA have been identified and are commonly known (Guttman & Rinn, 2012) The ncRNA family consists of housekeeping ncRNAs and regulatory ncRNAs. The housekeeping ncRNAs in clude ribosomal (rRNA), small nuclear (snRNA), small nucleolar (snoRNA), and transfer RNAs (tRNA). Regulatory ncRNAs are divided into two classes based on size; those less than 200 nucleotides are short/small

PAGE 22

! ! ncRNAs and encompass microRNAs (miRNAs) while t hose greater than 200 nucleotides are long non coding RNAs (lncRNAs) (Wu et al., 2013) One subgroup within lncRNAs are lincRNAs (long intergenic non coding RNAs) which lie in the intergenic regions of the genome (L. Yang, Froberg, & Lee, 2014) The ENCODE project, an effort to survey transcription in entirety, identified more than 9000 genomic l oci that are transcribed into lncRNAs in the nucleus of human cells, yet the majority of these lncRNAs have yet to be characterized (L. Yang et al., 2014) The pursuit to functionally characterize these lncRNAs in order to determine their biological relevance has been a popular trend in the epigenetic realm, but many researchers have faced difficulties due to th e lack of protein coding potential of lncRNAs. Role of Long Non Coding RNA in Gene Regulation Over the past decade, thousands of well expressed lncRNAs have been identified to be actively transcribed across the human genome and have shown to have evolutio narily conserved promoter regions and splice sites as well as unique start and stop codons (Guttman et al., 2009) These findings indicate that lncRNA transcription occurs independently from neighboring genes (L. Yang et al., 2014) However, several lncRNAs have shown to maintain enhancer like functions, which allows the m to work in cis to activate the expression of their neighboring genes (Guttman & Rinn, 2012) Uniquely, other lncRNAs work in trans in order to regulate distant genes (C hu, Qu, Zhong, Artandi, & Chang, 2011) LncRNAs have shown to be key regulators of numerous biological processes including chromatin regulation, dosage compensation, imprinting, epigenetic inheritance, and stem cell pluripotency (S. Y. Ng, Johnson, & Stanton, 2012; Tsai et al., 2010; L. Yang et al., 2014)

PAGE 23

! ! Despite the challenges in functionally characterizing the ongoing list of identified lncRNAs, an emerging theme is the ability for lncRNAs to regulate gene expr ession. One mechanism for modulating gene expression is through chromatin modifications (L. Yang et al., 2014 ) Nucleosomes are the packaged unit of genomic DNA wrapped around histone proteins which are then organized into a higher structure called chromatin Modifications of these histones regulate the transcription of the DNA wrapped around them. One of the f irst and most famous lncRNA s to be discovered is Xist which is required for X inactivation that occurs through lncRNA mediated chromatin modifications (Guttman & Rinn, 2012) Xist has been shown to recruit chro matin modifying complexes, such as the Polycomb Repressive Complex 2 (PRC2), to appropriate genomic loci in order to negatively regulate gene expression (L. Yang et al., 2014) Once the genomic region is silenced, the chromatin condenses to cause the repression of the entire X chromosome, resulting in X inactivation (G uttman & Rinn, 2012) The discovery of this lncRNA and its inhibitory process through PRC2 has opened numerous doors for characterizing future lncRNAs. LncRNAs Associated with Polycomb Repressive Complexes Due to the inability to determine the function of lncRNAs from sequence information alone discovering the RNA protein interactions allows researchers to functionally characterize these RNAs (L. Yang et al., 2014) Roughly 30% of lincRNAs tested have been shown to physically associate with at least 1 of 12 distinct chromatin regulatory complexes in order to guide chromatin proteins to DNA targets as a method to silence genomic regions (Guttman & Rinn, 2012) Another characterized and highly studied lncRNA, HOTAIR ( HOX Transcript Antisense RNA ), has also been identified to

PAGE 24

! %. ! target PRC2. This complex, a histone methyl transferase required for epigenetic silencing, is localized to thousands of mammalian genes (Cifuentes Rojas, Hernandez, Sarma, & Lee, 2014; Davidovich, Zheng, Goodrich, & Cech, 2013) PRC2 is comprised of multiple subunits including the catalytic H3K27 methyltransferase, EZH2 (Enhancer of Zeste 2), a protein that binds to RNA with high affinity in order to establish a chromatin structure that causes repression of transcription (Cifuentes Rojas et al., 2014; L. Yang et al., 2014) The exact mechanism behind how PRC2 is targeted to specific loci is yet to be established, but one possibility is through lncRNA recruitment. LncRNAs in both the cis and trans formation have been sh own to interact with PRC2 and tether the complex to the site of transcription (Cifuentes Rojas et al., 2014) HOTAIR, a trans acting RNA, has been shown to retrieve EZH2 through RNA immunoprecipitation (RIP). These results suggest HOTAIR recruits PRC2 to chromatin in order to initiate transcriptional silencing (Tsai et al., 2010) In addition, a RIP of EZH2 demonstrated the interaction between PRC2 and Xist RNA during X chro mosome inactivation in order to initiate and maintain the silencing of one of the X chromosomes (L. Yang et a l., 2014) Both PRC2 and EZH2 have become a scientific interest due to their physical interactions with other lncRNAs as well as their clinical significance in disease and cancer (Cifuentes Rojas et al., 2014) S imilarly to PRC2, polycomb repressive complex 1 (PRC1) establishes strong transcriptional repression. Many lncRNAs have been recognized as important participants in PRC1 function (Margueron & Reinberg, 2011) The biochemical mechanism behind how PRC1 silences transcription is believed to occur through the dissociation of RNA Polymerase II Preinitiation Complex (PIC) by blocking the recruitment of Mediator, a requir ed subunit of PIC (Lehmann et al., 2012) PRC1 has also been shown to stabilize

PAGE 25

! %% ! chromatin structures and catalyze the monoubiquitylation of lysine 119 on histone H2A (H2AK119) (Margueron & Reinberg, 2011; Morey et al., 2012) There are five known Cbx proteins that associate with PRC1 Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8. Different PRC1 complexes are associated with distinct Cbx proteins, which is dependent upon target specific ity (Morey et al., 2012) Each Cbx protein contains a chromodomain at the amino terminal that binds tri methyl K27 H3 (H3K27me3), which helps to regulate pluripotency and differentiation of ESCs through unique, non redundant functions (Ren, Vincenz, & Kerppola, 2008) The Cbx subunits are believed to stabilize the PRC1 association with chromatin through direct interaction with H3K27me3 (Morey et al., 2012) Additionally, researchers have shown that PRC1 is dependent on PRC2 such that the histone trimethylation performed by PRC2 to initiate genomic silencing is necessary for PRC1 to be recruited to specific genes in order to maintain th e silenced state (Ren et al., 2008) When PRC2 methylates H3K27 through EZH2, this provides a binding site for PRC1. Once PRC1 binds, it can ubiquitinate H2AK119 for genomic silencing (Lehmann et al., 2012) Additionally, research has shown that the synergistic effects of the Polycomb Repressive Complexes are involved in the regulation of stem cell pluripotency. Specifically, this stem cell regulation is controlled by the di fferent unique function s of the Cbx proteins, which maintains the balance between self renewal and differentiation (Morey et al., 2012) In general, determining the regulatory interactions and the mechanistic prope rties between lncRNA and both Polycomb Repressive Complexes will improve the understanding of the many regulatory roles of lncRNA s as well as aid in pharmaceutical design to target disease.

PAGE 26

! %& ! In addition to regulating gene expression through chromatin modif ications, characterized lncRNAs can function as transcriptional activators by acting as cofactors that help to enhance kinase activity in order to modify chromatin architecture. LncRNAs have also been shown to control the co and post translational process es including splicing and translation of mRNA as well as the subcellular localization of proteins (L. Yang et al., 2014) Another common function of lncRNAs is to act as modular scaffolds in order to initiate the assembly of multiple protein complexes and recruit these complexes to the necessary location to regulate gene expression. These regulatory complexes no t only include interactions with protein, but RNA DNA and RNA RNA interactions as well (Guttman & Rinn, 2012) For example, HOTAIR acts as a scaffold for PRC2 and a H3K4 demethylase complex, LSD1. It has been sh own that HOTAIR binds to both PRC2 and LSD1, bridging these two complexes together in order to target this complex to silence multiple HOXD genes as well as other genes on select chromosomes (Tsai et al., 2010) Ho wever, the mechanism behind how HOTAIR guides PRC2 to its target genes is yet to be understood (Guttman & Rinn, 2012) In general, the formation of these lncRNA bridges allows lncRNAs to target specific genomic loci with high specificity (L. Yang et al., 2014) Due to this identified function of lncRNAs, it is suggest ed lncRNAs resemble transcription factors when regulating chromatin states because the target gene information is highly selective and resides in the RNA (Chu et al., 2011) It is evident that determining the functi ons of identified lncRNAs is challenging but will only add to the greater understanding of regulatory mechanisms used by mammalian cells to control gene expression. Despite the progress epigenetic scientists

PAGE 27

! %' ! have made in characterizing lncRNAs, a greater a nd more in depth analysis is needed to fully understand lncRNAs and their role in human disease. The goal of this research project is to functionally and biochemically characterize a novel long non coding RNA, LINC00486, which was discovered from analysis of hypermethylation patterns in BPD patients. This will be accomplished by determining in which cell types and tissues each human transcript of LINC00486 is expressed. Additionally, identifying protein RNA interactions such as an association between LINC 00486 and PRC1 and /or PRC2, as will help to determine the function of LINC00486. As stated previously, the specific aims of this project include: Aim #1: Characterizing the expression of endogenous LINC00486 transcripts Aim #2: Biochemical characterization of LINC00486 transcripts Materials and Methods Over Expression of LINC00486 Transcripts in Mammalian Cells Gibson Assembly Cloning of LINC00486 Transcripts 1 3 and L ncRNA HOTAIR into pCAGEN 1. Expression vector of interest, pCAGEN, was linearized using 2uL of EcoR1 restriction enzyme, 2uL EcoR1 Buffer, and 2ug of circular pCAGEN. To make the final volume 20uL, dH20 was added to each sample. The digest reaction was incubated at 37¡C for 15 minutes. 2. To remove phosphate groups on pCAGEN in order to avoid plasmi d recircularization, 1uL of ant arctic phosphatase was added to the digest reaction. Samples were incubated at 37¡C for 15 minutes.

PAGE 28

! %( ! 3. The effectiveness of the digest reaction was validated by running the digested products on a 0.8% agarose gel. The circular p lasmid was added as a control. 4. After validation of successful digestion, the remaining linear plasmid was purified using the QIAquick Gel Extraction Kit from QIAGEN. a. 1 volume of isopropanol was added to remaining digest reaction. Solution was added to spi n column in a collection tube. Tube was spun at 10,000 rpm for 1 minute and the flow through was discarded. b. The wash was performed by adding 750uL of PE Buffer to the spin column. Tube was spun at 10,000 rpm for 1 minute and the flow through was discarded. c. Empty column was re spun at 10,000 rpm for 1 minute. d. The spin column was placed in a fresh collection tube. 30uL Buffer EB was added to the spin column and the tube was spun at 10,000 rpm for 1 minute to elute purified digest product. 5. The synthesized gBl ock gene fragments (IDT) were diluted with TE buffer (pH 7.5) to a concentration of 10ng/uL. Solution was homogenized by vortexing and was placed on ice for immediate use. (Eventually stored at 20¡C). 6. A 3 fold molar increase of DNA fragments and 50ng of l inear plasmid was combined for the assembly (up to a 6 fold molar increase in fragments could have been used if the number of fragments increased above 4). DNA fragments were converted into pmoles using Equation 1.

PAGE 29

! %) ! 7. The Gibson assembly was prepared on ice by combining 50ng linear vector, 3 fold molar excess fragments and 10uL Gibson Assembly Master Mix. dH20 was added to make a total volume of 20uL. 8. Samples were incubated at 50¡C for 15 minutes ( samples that contained more than 3 fragments were incubated at 50¡C for 30 minutes). After incubation, samples were immediately placed on ice or stored at 20¡C. 9. NEB 5 alpha competent E. coli cells (included in Gibson Assembly Kit) were transformed with 2uL of assembly reaction. Overnight cultures were made from colo nies on LB plates. Plasmids were purified through mini preps (GeneJET Plasmid Miniprep Kit from Thermo Scientific). 10. To c onfirm if Transcripts 1 3/HOTAIR was successfully cloned into pCAGEN, restriction digests were performed and products were run out on a 1% agarose gel. N2A and HEK293T Mammal ian Cell Transfection Using PEI Day 1: Cells were plated at 5.0x10 5 cells per well in 6 well plates in cell specific media (Appendix) Day 2: Media was not changed in 6 well plates. 2.0ug of total DNA (1800ng pCAGEN Transcript or HOTAIR + 200ng pMax GFP) was added to 100uL Opti MEM. In a separate tube, 5uL of PEI was added to 95uL Opti MEM per transfection. Tube was tapped to mix and incubated at room temperature for 5 minutes. 100uL PEI:Opti MEM was transferred to D NA:Opti MEM tubes. Tubes were tapped to mix, centrifuged at lowest speed for collection, and incubated at room temperature for 20 minutes. Mixture was added to cells drop wise. Cells were incubated at 37¡C with 5% CO 2 overnight.

PAGE 30

! %* ! Day 3: To determine if tra nsfection was successful, presence of green fluorescence from pMax GFP was observed. mESC Transfection Using PEI 1. ESCs were trypsinized and resuspended in Opti MEM. 10uL of resuspension was added to 10uL trypan blue dye. The total number of ESCs was counted using the Countess. 2. 5 x 10 5 cells were aliquotted into 2mL microcentrifuge tubes (the same number as wells to be plated). Tubes were centrifuged for 2 minutes at 1500rpm. Supernatant was carefully aspirated. 3. Cell pellets were resuspended in 300uL Opti MEM 4. 2ug of total DNA was added to the resuspended cells (200ng pMax GFP and 1800ng plasmid of interest). Cells were mixed by careful pipetting. 5. 100uL of PEI was added to the resuspended cells. Cells were mixed by pipetting. 6. Cells were incubated for 30 minu tes at room temperature. Mixture was pipetted every 10 minutes. 7. The transfection reaction was plated on freshly prepared gelatin coated 6 well plates with warm and complete ESC media. 8. Transfection was incubated overnight and visualized the next day using the fluorescence microscope. RNA Isolation RNA isolated using mirVana mirRNA Isolation Kit by Life Technologies followed kit protocol. When RNA was to be isolated using Direct zol RNA Miniprep by Zymo, 500uL TRIzol was added per well of a 6 well plate to begin lysing process (1mL

PAGE 31

! %+ ! of TRIzol was added per 10cm plate). RNA isolation followed kit protocol. All isolated RNA was quantified using Nanodrop 2000 made by ThermoScientific. cDNA Synthesis cDNA was synthesized using High Capacity Reverse Transcriptas e kit from Applied Biosystems. Kit protocol was followed and 100ng 2ug of total RNA was used to synthesize cDNA. The amount of RNA used was kept constant for each set of RNA. TA Cloning Sequences of interest were cloned into pCR2.1 vector using T4 DNA li gase. Kit protocol was followed. RT PCR To amplify sequences of interest RT PCR was performed by adding 10x Buffer, 5mM dNTPs, 50mM MgCl 2 200ng/uL Forward primer, 200ng/uL Reverse primer, 2u/uL Platinum Taq, 200ng of cDNA and water, for a final volume of 25uL. Thermocycler Program *Note: Thermocycler program was optimized by changing annealing temperature depending on size of fragment desired for amplification. 10 min 94¡C [1 min 94¡C x30 1 min 65¡C 1 min 72¡C] 7 min 72¡C hold 4¡C qPCR To quant ify gene expression, biological replicates were created for each cell or treatment type. RNA was isolated from cells; cDNA was synthesized using 100ng 2ug RNA. If 2ug RNA was used, the final concentration of synthesized cDNA was 100ng/uL.

PAGE 32

! %, ! cDNA was diluted in sterile water to 8.89ng/uL. A master mix for each probe was made by combining 2x Taqman Mix (with UNG), 20x Taqman probe, and DEPC H20. Master mix components were multiplied by total number of wells designated for each probe. After 19.38uL of master mix was added to each designated well, 5.63 uL of diluted cDNA was added to plate in triplicates. Plates were spun down to collect all liquid at bottom of plate, and then placed in StepOne qPCR Machine. Data was collected by running the Comparative C T ( ##C T ) program, which followed the cycling parameters: qPCR Program: 2 min 50¡C x 1 cycle 10 min 95¡C x 1 cycle [15 sec 95¡C 1 min 60¡C] x30 cycles RNA Immunoprecipitation Biotinylation of Oligos %/ Each primer was diluted to 100uM to make a stock solution. &/ A serial dilution was performed on the stock solution (100uM) to create 10uM and 1uM solutions. These serial dilutions were important in order to determine optimal primer concentration. '/ Biotin was diluted to 5uM (prepared fresh every time). (/ Just before use, a master mix was prepared by diluting TdT to 2U/uL in 1X TdT Reaction Buffer and ultrapure water, resulting in a total volume of 5uL. A master mix was prepared fresh every time and once made, was kept on ice and used immediately. )/ The labeling reactions wer e prepared by adding Ultrapure Water, 10X TdT Reaction Buffer, 1uM Oligo, 5uM Biotin 11 UTP, and 2U/uL diluted TdT to

PAGE 33

! %! make a final volume of 50uL (labeling reactions were prepared individually for each reaction). */ Reactions were incubated at 37 ¡ C for 30 min utes. +/ To stop the reactions, 2.5uL 0.2M 0.5M EDTA was added. ,/ In fume hood, 50uL choloroform:isoamyl alcohol was added to each reaction to extract TdT. -/ Mixtures were vortexed briefly and tubes were centrifuged for 2 minutes at 14,000 rpm. %./ In fume hood, for each reaction the aqueous layer containing biotinylated oligos was carefully removed and placed in fresh microcentrifuge tube. Biotinylated oligos were stored at 20 ¡ C. RNA Pul l Down with Biotinylated Oligos 1. To determine the optimal concentration of biotin ylated oligos, serial dilutions of the 1uM biotinylated oligos were prepared: 1:10, 1:100, and 1:1000. 2. Biotinylated oligos were blocked in 100uL streptavidin magnetic beads (New England Bio Labs) for 1 hour at room temperature while rotating. For each tra nscript RNA, 3 tubes were prepared: Tube 1: 1uL of each 1:10 diluted oligo Tube 2: 1uL of each 1:100 diluted oligo Tube 3: 1uL of each 1:1000 diluted oligo 3. After 1 hour, 2ug of RNA and water were added to each tube to make a final volume of 25uL (in addi tion to 100uL streptavidin beads).

PAGE 34

! &. ! 4. Tubes were incubated on ice for 30 minutes and then vortexed and briefly spun down. 5. A magnetic plate was used to remove any clear liquid from tubes without disturbing the beads. 6. Beads were washed with 100uL Wash 2/3 fro m mirVana mirRNA Isolation Kit (Life Technologies). Beads were vortexed quickly and spun down. The magnetic plate was used to remove clear liquid from beads. 7. Wash step was repeated 2 more times and clear liquid was removed and discarded. 8. cDNA was synthesiz ed by combining entire amount of samples (streptavidin beads + RNA + Biotinylated oligos) to cDNA master mix (Applied Biosystems). 9. Tubes were spun down to collect beads at bottom and 2uL of clear liquid on top of beads (the cDNA) was used to perform RT PC R. Transcript specific primers were used. Non pulldown cDNA was used as a negative control. 10. PCR products were observed by running out on a 1% agarose gel. EZH2 Over Exp ression and Immunoprecipitation 1. Two sets of N2A 6 well plates were made one for HA beads and one for control beads. For each set, 2ug of total DNA was transfected: 1800ng pCMV HA EZH2 + 200ng pMax GFP (negative control) 900ng pCMV HA EZH2 + 900ng pCAGEN T ranscripts 1 3 (or pCAGEN HOTAIR ) + 200ng pMax GFP

PAGE 35

! &% ! 2. Biotinylated oligos were blocked in 1 00uL streptavidin magnetic beads for 1 hour at room temperature while rotating. The previously determined optimal dilution was used for each set of oligos. 3. In cell hood, transfected cells were washed, trypisinized and pelleted. 4. Cytoplasm were isolated by resuspending the cell pellets in 400uL Buffer A in order to lyse cell (Appendix) Tubes were left on ice for 15 minutes to allow cells to swell. To each tube, 25uL of IGEPAL was added. Tubes were vortexed vigorously for 30 seconds and centrifuged at 4¡C f or 30 seconds at 14,000 rpm. The supernatant was removed and placed in fresh tube (cytoplasmic isolation). Isolated cytoplasm was stored at 80¡C. 5. Nuclei was isolated by resuspending remaining pellet in 50uL ice cold Buffer B (Appendix) Tubes were rocked vigorously for 15 minutes at 4¡C. Tubes were centrifuged at 4¡C for 5 minutes at 14,000 rpm. Supernatant was removed and place in fresh tube (nuclei isolation). Isolated nuclei were stored at 80¡C. 6. Before immunoprecipitation, the negative controls were ac quired by placing 10uL of the nuclei s upernatants into fresh tubes. 3X volume of TRIzol was added to tubes in fume hood and nuclei were lysed by pipetting up and down. Tubes were stored at 80¡C. 7. To the remaining nuclei, 200uL of Lysis/Binding Solution (m irVana RNA Isolation Kit from Life Technologi es) was added to each tube. Nuclei c pellets were resuspended by vigorous vortexing. 8. Resuspended nucleic pellets were separated into two groups: HA magnetic beads and the Control beads. To the designated HA bead s tubes, 10uL of HA magnetic

PAGE 36

! && ! beads were added to the nucleic lysate. To the Control beads tubes, 10uL of the control beads were added to the nucleic lysate. Tubes were incubated overnight at 4¡C while rotating. 9. After incubating, both the HA and Control tu bes were spun down and a magnetic rack was used to remove liquid from the beads. Liquid was discarded and beads were washed 3x with sterile H 2 0 in hood. 10. In the fume hood, 200uL of TRIzol was added to each tube. Tubes were vortexed and incubated at room te mperature for 5 minutes. Tubes were the centrifuged at 12,000 x g for 1 minute and the supernatants were transferred to fresh tubes. 11. RNA was isolated (Direct zol RNA Miniprep by Zymo) and cDNA was synthesized (Applied Biosystems). RT PCR was performed usi ng transc ript specific primers and HOTAIR primers for each sample. Cbx Protein Over Exp ression and Immunoprecipitation For the Anti Flag and Anti GFP immunoprecipitations to determine if Transcripts 1 3 were associated with Cbx proteins, the same protoco l as the HA tagged EZH2 immunoprecipitation was followed, except for a few adjustments. For the Anti Flag IP, N2A cells were transfected with each pVenus Cbx plasmid individually. Once cell pellets were resuspended, Anti Flag beads (Sigma Aldrich) were add ed for an overnight incubation. No control beads were used. Similarly for Anti GFP IP, the same protocol was followed. Except, N2A cells were transfected with pTripz MCS1 Cbx2 (ful l length) and Anti GFP beads (ChromoTek) were added to cell suspension. No control beads were used. Results

PAGE 37

! &' ! Over Expression of LINC00486 Transcripts in Mammalian Cells In order to determine the function of LINC00486 transcripts, we decided to start by overexpressing the lncRNAs in mammalian cells. Therefore, we needed to clone t he three primary predicted LINC00486 transcripts into a mammalian expression vector. The Gibson Assembly cloning technique (Gibson et al., 2009) was used to individually clone each transcript and lncRNA HOTAIR into a mammalian expression vector, pCAGEN. Nucleotide fragments, gBlocks, were synthesized to be identical to the sequences of each transcript using the PrimerQuest Design Tool provided by Integrative DNA Technologies (IDT). In addition, because of how Gibson Assembly cloning works, 15 bp correspon ding to pCAGEN were added to the ends of each synthetic DNA fragment. With the concerted action of the three necessary enzymes, an exonuclease, a polymerase, and a DNA ligase, the fragments were assembled into a single fragment, which was inserted into the linearized pCAGEN vector. At the end of the single tube isothermal reaction, a fully sealed double stranded DNA molecule was created that contained each transcript inserted into pCAGEN (Supplemental Data Figure 1) (Gibson, 2009). A restriction digest usin g 200ng of Gibson product was performed to ensure the correct insertion into pCAGEN (Figure 1A). Next, to confirm that we had successfully cloned the LINC00486 transcripts and HOTAIR we performed Sanger sequencing. Each correctly sized RT PCR product was TA Cloned into pCR2.1 vector using the TA cloning Kit (The Original TA Cloning Kit from Invitrogen). The ligations were transformed into E. coli competent cells, overnight cultures were made from the transformation colonies, and mini preps (Thermo Scientif ic) were performed to isolate cloned plasmids. Restriction digests were performed on

PAGE 38

! &( ! isolated plasmids for verification of correct band size (Figure 1B). Finally, 200ng of digest products were sent for sequencing using the DNA sequencing service provided b y Eton Biosciences to validate the correct human transcript sequence. After successful cloning of each individual transcript into the pCAGEN vector, it was imperative to determine if the plasmids were capable of successful expression in mammalian cells. U sing polyethylenimine (PEI), a linear cationic polymer as a transfection agent, each plasmid was transfected into mouse neuroblastoma (N2A) cells (Bartman, 2014). We chose N2A cells because murine LINC00486 homologues have not yet been identified, and the r efore the human PCR primers we designed would not amplify anything in mouse cDNA (No Template Control, Figure 2B). This allowed us to take advantage of the clean background in N2A cells to assess the expression of human LINC00486 transcripts. We also clone d the lncRNA HOTAIR into pCAGEN to include in our experiments as a positive control for a representative lncRNA. In addition, pMax GFP was cotransfected with each transcript into the cells to act as a positive control for successful transfection. LINC00486 Transcripts 1 3 and HOTAIR were transfected into N2A cells individually. Verification of successful transfection was exhibited through green fluorescence, indicating the nucleus of the N2A cells successfully acquired the introduced pMax GFP and transcript (Figure 2A). To verify that we were able to overexpress the LINC00486 transcripts and HOTAIR, RNA was isolated from transfected N2A cells (MirVana mirRNA Isolation Kit from Life Technologies) and cDNA was synthesized (High Capacity cDNA Reverse Transcrip tion Kit from Applied Biosystems) using 2ug of total isolated RNA. Next, RT PCR was performed to verify expression of each transcript using primers specific to each

PAGE 39

! &) ! transcript and HOTAIR Primers were designed using Primer3Plus (www.bioinformatics.nl/cgi b in/primer3plus/primer3plus.cgi) and each transcript primer pair was tested on the transfected cells to determine the primer specificity (Figure 2B). The primers that successfully showed expression of the transfected transcript indicated high primer specifi city; we were able to demonstrate that each primer pair was specific to each LINC00486 transcript (Transcript 1: 298bp; Transcript 2: 129bp; Transcript 3: 179bp). GAPDH primers were used as a control for RNA integrity (GAPDH: 150bp). RT PCR products were validated on a 1% agarose gel to confirm correct band size of transcripts. The no template controls were used as a negative control to verify that the primers were not contaminated. Many lncRNAs are found in the nucleus due to their association with trans cription factors to initiate gene silencing (L. Yang et al., 2014) Therefore, we were curious to see where in the cell LINC00486 was localized. To do so, we isolated the nuclei and cytoplasm from the cell lysate in order to focus our downstream experiment s on a specific cellular region. RT PCR revealed the presence of each transcript in both the nuclei and cytoplasm. As predicted, LINC00486 transcripts were found in the nucleus supporting the idea that LINC00486 may associate with different transcription factors Additionally, HOTAIR was localized to the nucleus as predicted. Transcript expression seen in the cytoplasm was most likely due to the over expression of each transcript As expected, the GAPDH control was only localized to the cytoplasm, validati ng successful nucleic and cytoplasmic isolation (Figure 2C).

PAGE 40

! &* ! Figure 1 Verification of Successful Cloning of LINC00486 Transcripts 1 3. (A) XhoI and NotI double restriction digest on pCAGEN LINC00486 Transcript 2. An example of ensuring LINC00486 Transcripts were correctly inserted int o pCAGEN through Gibson Assembly Cloning. (B) EcoRI restriction digest on pCR2.1 LINC00486 Transcript 2 to determine successful TA cloning. This is an example of the digests performed on each LINC00486 Transcript for cloning validation.

PAGE 41

! &+ ! Figure 2 Verification of Over Expression and L ocalization of LINC00486 Transcripts 1 3 in N2A cells. (A) After co transfecting N2A cells with individual LINC00486 transcript s and pMaxGFP, successful transfection is validated by observing green fl uorescence in cells. (B) RT PCR on over expressed LINC00486 Transcripts 1 3 throughout entire cell and (C) in nuclei and cytoplasm using transcript specific primers (Expected sizes Transcr ipt 1: 298bp; Transcript 2: 129bp; Transcript 3: 179bp; HOTAIR : 250 b p ; GAPDH: 150bp. HOTAIR primers were used as a positive control for a representative lncRNA ; GAPDH primers were used as a control for RNA integrity).

PAGE 42

! &, ! Endogenous Expression in Human Cells and Tissues To determine if LINC00486 Transcripts 1 3 and HOTAIR were endogenously expressed in human cells, RNA was isolated from untrans fected human embryonic kidney (HEK293T) cells (Life Technologies) and cDNA was synthesized using 2ug total RNA (Applied Biosystems). Using the transcript specific primers, RT PCR was performed to observe endogenous transcript expression in untransfected HE K293T cells. Only the Transcript 3 primer showed a single band at the correct size (179bp), indicating specificity to endogenous Transcript 3 expression (Figure 3A). D ue to the lack of detection of T ranscripts 1 and 2 at the correct sizes (Figure 3A), new primers of varying sizes were designed to overlap different exons in each transcript (to prevent the detection of genomic DNA). RT PCR was performed on human cDNA using the newly designed primers. RT PCR products were ran out on a 1.7% agarose gel (Figure 3B, Transcript 1 not included). The primers with the most robust expression specific to each transcript were used to look at transcript expression in different human tissues (Figures 4A C). After confirming presence of endogenous transcript expression in c ultured human cells, we were curious to see the endogenous expression pattern in human tissues. A panel of human tissue RNA (Human Total RNA Master Panel II) was obtained from Clontech Laboratories. cDNA was synthesized from 2ug total RNA for each sample, and standard RT PCR was performed. Each transcript showed specific levels of expression in different tissues. The tissues that showed no expression were chosen as controls for downstream experiments (Table 1). Next, we wanted to specifically quantify the expression of each transcript in the same human tissues by performing quantitative PCR (qPCR). However, pre made

PAGE 43

! &! Taqman assays for each transcript were not available for purchase. Therefore, we designed our own Taqman probes using the sequences of the tra nscript primers previously shown to have robust and specific expression in human tissues (Figures 4A C). Using these probes, qPCR was performed to quantify transcript expression in each tissue. The control tissue used for each transcrip t probe was the tiss ue lacking transcript expression seen in the previous RT PCR results (Table 1) This selected tissue was held consistent for all tissues while looking at the expression of each transcript. Transcript 1 revealed the highest expression in testes and salivary gland, Transcript 2 in the liver and salivary gland and Transcript 3 in testes and placenta (Figure 5A). To confirm accurate results, the qPCR products were run out on a 1.7% agarose gel. Absence of primer dimers on agarose gel validated specific transcri pt expression (Figure 5B D). After determining the tissues with high transcript expression, we noticed a trend of relatively high transcript expression in both the testes and placenta. Due to the genetic similarities between germ cells and ESCs, we were c urious to observe endogenous LINC00486 expression in human embryonic stem cells (H7 hESCs) and induced pluripotent cells (CWRU205 iPSCs). We received the hESC and iPSC RNA from Dr. Paul Tesar at Case Western Reserve University. cDNA was synthesized using 5 00ng RNA and qPCR was performed to quantify transcript expression in the different cell types. For each transcript, the same tiss ue that previously showed no transcript expres sion was used as a control. Reduced expression for each transcript was observed in H7 hESCs and CWRU205 iPSCs (Figure 5E), likely indicating no functional role for LINC00486 in human ESCs

PAGE 44

! '. ! One common capability of lncRNAs is their enhancer like functions on neighboring genes. When this occurs, the flanking genes often show similar ln cRNA expression, which is visualized using qPCR. To determine if LINC00486 had cis capabilities, we looked at endogenous flanking gene (genes TTC27 and LTBP1) expression in human cells. Probes were designed for flanking genes LTBP1 and TTC27 (IDT). We perf ormed qPCR on untransfected HEK293T cells and HEK293T cells transfected with each individual transcript to observe endogenous TTC27 and LTBP1 expression. Low TTC27 expression was observed in HEK293T cells transfected with LINC00486 Transcripts 1 3. LTBP1 e xpression remained relatively constant in all HEK293T cells compared to the untransfected sample (Figure 6A). However, in N2A cells transfected with each transcript, Transcript 3 revealed an almost 2 fold increase in LTBP1 expression compared to the other transcripts and untransfected samples (Figure 6B). This suggests there is a correlation in gene expression between the flanking genes LTPB1 and LINC0486 Transcript 3 in N2A cells These results are expected due to prior knowledge of enhancer capabilities i n other characterized lncRNAs. Since LINC00486 was identified as being hypomethylated in human lymphoblasts from patients with bipolar disorder (BPD) that had been treated with lithium, we wanted to examine whether the expression of the three LINC00486 tr anscripts was changed when cells were treated with lithium or another Gsk 3 inhibitor, SB 415,286. In order to look at endogenous transcript expression when Gsk 3 is inhibited, untransfected HEK293T cells were either treated with 10mM LiCl or 20uM SB 415,2 86 for 24 hours. HEK293T cells were chosen to act as a baseline for Gsk 3 inhibition in human cells. Results revealed more than a 2 fold increase in Transcript 1

PAGE 45

! '% ! expression in the 30uM SB 415,286 treated cells. However, transcript expression in the other c ell treatments remained relatively constant (Figure 7). After determining the baseline transcript expression in human cells treated with a Gsk 3 inhibitor, we wanted to see endogenous transcript and flanking gene expression in lymphoblast cells derived fro m the BPD individual that were treated with a Gsk 3 inhibitor. qPCR revealed there was no difference in LTBP1 or TTC27 gene expression between 3188 lymphoblast cells that were untreated or treated with 1uM LiCl for 21 days (Figure 8A). Transcript expressio n was compared between treated and untreated cells for the same affected individual, revealing a 2 fold increase in 3188 lymphoblast cells treated with 30uM SB for 2 days compared to untreated and LiCl treated cells (Figure 8B). To quantify endogenous tra nscript expression in a (BPD) patient (3188) compared to their unaffected sibling (3189), qPCR was performed to observe transcript and flanking gene expression in these individuals. When comparing the BPD patient (3188) with their unaffected sibling (3189) there was much lower LTBP1 expression in 3189 compared to 3188. TTC27 expression remained relatively constant between the siblings. Transcript 3 expression revealed a 0.5 fold decrease in 3189 compared to 3188 (Figure 8C). Transcript 3 was chosen due to having the most robust expression in N2A cells in previous results (Figure 6B).

PAGE 46

! '& ! Figure 3 Endogenous LINC00486 Transcript Expression in HEK293T Cells. (A) RT P CR on untransfected HEK293T cells to observe endogenous transcript expression. (Expected sizes Transcr ipt 1: 298bp; Transcript 2: 129 bp; Transcript 3: 179bp ; GAPDH: 150bp. GAPDH primers were used as a control for RNA integrity). (B) Due to the inability of transcript primers to detect endogenous transcript expression (A), new transcript primers were designed to enhance detection of endogenous expression. Each primer name indicates expected band size. Asterisk (*) indicates selected primer used in downstream experiments (Transcript 1 primer that was chosen is not shown).

PAGE 47

! '' ! Figure 4 Detecting Presence of Endogenous LINC00486 Transcript Expression in Human Tissue. Primers selected from Figure 3B were used to observe (A) Transcript 1 ( 105bp), (B) Transcript 2 (84bp), and (C) Transcript 3 (106bp) in 20 human tissues. Asterisk (*) indicates tissue with robust transcript expression. ( GAPDH primers (150bp) were used as a control for RNA integrity).

PAGE 48

! '( ! Table 1 Tissues with robust transcript expression and designated tissue controls. Expressed Transcript: Robust Tissue Expression : Low E xpression Tissues as Chosen Controls: Transcript 1 Salivary Gland Brain Cerebellum Colon Transcript 2 Prostate Whole Brain Salivary Gland Spleen Transcript 3 Whole Brain Small Intestines Testes Salivary Gland Bone Marrow

PAGE 49

! ') ! Figure 5 Quantification of Endogenous LINC00486 Transcript Expression in Human Tissues and Cells. (A) Quantified LINC00486 Transcript 1 3 expression in 20 human tissues. Each transcript was normalized to designated tissue control (Table 1). (B D) RT PCR on qPCR pro ducts from (A) using transcript specific primers. Absence of primer dimers was indicative of accurate results (GAPDH primers were used as a control for RNA integrity). (E) LINC00486 Transcript 1 3 expression in hESCs and iPSCs (same tissue control used for each transcript as shown in Table 1 ).

PAGE 50

! '* ! Figure 6 Endogenous LTBP1 and TTC27 Gene Expression in HEK293T and N2A Cells. (A) Relative quantification of LTBP1 and TTC27 gene expression in untransfected and transfected HEK293T cells. (B) Relative quantification of LTBP1 in untransfected and transfected N2A cells.

PAGE 51

! '+ ! Figure 7 Effect of Gsk 3 Inhibition on Endogenous LINC00486 Transcript Expression in HEK293T Cells. HEK293T cells were treated with 10mM LiCl or 3 0uM SB 415,286 for 24 hours. LINC00486 Transcript 1 3 expression was quantified relative to untransfected HEK293T cells.

PAGE 52

! ', ! Figure 8 LINC00486 Transcript and Flanking Gene Expression in BPD Patient and Unaffected Sibling. Lymphobl ast cells were derived from BPD individual (3188) and unaffected sibling (3189). (A) Lymphoblast cells from 3188 were treated with 1uM LiCl for 21 days. Relative quantification of LTPB1 and TTC27 gene expression in cells treated with Gsk 3 inhibitor. (B) L INC00486 Transcript 3 expression in 3188 lymphoblast cells treated with 1uM LiCl for 21 days or 30uM SB 415,286 for two days. (C) LTBP1, TTC27 and LINC00486 Transcript 3 expression in 3188 and 3189 lymphoblast cells.

PAGE 53

! '! RNA Immunoprecipitation Many characterized lncRNAs have been shown to be associated with chromatin regulatory complexes in order to initiate genomic silencing. These discoveries have often occurred through the determination of RNA protein interactions. Therefore, in order for us to biochemically characterize LINC00486 we wanted to determine if it physically associated with any chromatin regulatory complexes. One approach that has been taken to discover interactions between ln cRNAs and histone proteins is ChIRP (chromatin immuno RNA precipitation) (Chu et al., 2011) We reasoned that a similar approach should detect interactions between lncRNA and its associated protein. Primers of roug hly 20 nucleotides in length were designed to span across the entirety of each transcript. These primers were designed using the Stellaris Probe Designer by Biosearch Technologies. Primers were also designed for HOTAIR as a positive control. The first step was to biotinylate each primer using the Biotin 3' End DNA Labeling Kit from Thermoscientific. Next, the biotinylated primers were added to the streptavidin beads in varying concentrations. The concentrations of the biotinylated oligos were varied to opti mize the pull downs for each transcript. In order to determine if each transcript could be affinity precipitated from the streptavidin beads, 2ug RNA (from transfected N2A cells) was added to the biotinylated oligos and streptavidin beads. After a one hour incubation, cDNA was synthesized from RNA attached to beads and RT PCR was performed. Results indicated the biotinylation of oligos allowed for Transcripts 1 3 to be affinity precipitated using streptavidin beads. The optimal concentrations for each oligo were Transcript 1 1:1000, Transcript 2 1:10, Transcript 3 1:10; these concentrations

PAGE 54

! (. ! remained constant throughout downstream experiments (Supplemental Figures 2A and 2B). After val idation of successful RNA affinity precipitation from streptavidin beads o ur goal was to determine if the LINC00486 transcripts would associate with EZH2. EZH2 is the catalytic subunit of PRC2, which many characterized lncRNA have been known to associate with in order to inhibit genomic loci. N2A cells were transfected with each transcript individually and RNA was isolated from cells (Life Technologies). This process was repeated for untransfected HEK293T cells. An immunoprecipitation was performed to determine if each transcript associated with HA tagged EZH2 protein (pCMV HA h EZH2 from the Ren lab). EZH2 was immunoprecipitated using HA coated magnetic beads to an isolated nucleic lysate from transfected N2A cells. After an overnight incubation, RNA was isolated from beads and cDNA was synthesized. Process was repeated using IgG control beads. RT PCR was performed to observe immunoprecipitation results. In N2A cells over expressing each transcript HA t agged EZH2 successfully pulled down Transcript 2. This was confirmed by the lack of Tran script 2 expression in the IgG c ontrol be ads set. Transcript 1 and 3 enriched in HA showed non specific binding (Figure 9A). This immunoprecipitation was repeated with HEK293T cells transfected with EZH2 (1800ng) and pMax GFP (200ng) to determine if endogenous LINC00486 transcripts were capable o f co precipitating HA tagged EZH2. Results showed EZH2 pulled down endogenous Transcript 2 from the nuclei and Transcript 1 from the cytoplasm (Figure 9B). In addition to the identified lncRNA association with PRC2, many lncRNAs are also physically associa ted with PRC1. PRC1 is another chromatin regulatory complex

PAGE 55

! (% ! that initiates gene silencing and is involved in maintaining stem cell pluripotency Therefore, we wanted to determine if there was any physical interaction between LINC00486 Transcripts 1 3 and t he Cbx protein family. Flag tagged Cbx proteins were tested for their capability to immunoprecipitate each transcript. N2A cells were transfected with individual transcripts (900ng), Flag tagged pVenus Cbx (900ng), and pMax GFP (200ng). Immunoprecipitation procedure was repeated for Cbx proteins: Cbx2, Cbx4, Cbx6, Cbx7, and Cbx8 using Anti Flag magnetic beads instead of HA magnetic beads (Figures 10A E) No control beads were used. Results were summarized in Table 2. Immunoprecipitation was repeated on a G FP tagged full length Cbx 2 protein using Anti GFP beads (Figure 10F). Results were consistent with the Flag tagged Cbx2 IP using Anti Flag beads (Figure 10A).

PAGE 56

! (& ! Figure 9 E ZH2 Immunoprecipitation of LINC00486 Transcripts 1 3. (A) RT PCR of nucleic isolate from N2A cells transfected with EZH2 and individual transcripts after HA tagged EZH2 immunoprecipitation. (B) RT PCR of endogenous LINC00486 transcripts in HEK293T cells af ter HA tagged EZH2 immunoprecip i tation. For A and B, transcript specific primers were used to amplify any transcripts that were pulled down with EZH2. (Expected sizes Transcript 1: 105bp; Transcript 2: 84bp; Transcrip t 3: 106bp; GAPDH: 150bp, HOTAIR : 250bp GAPDH primers were used as a control for RN A integrity and HOTAIR was used as a positive control for a representative lncRNA).

PAGE 57

! (' ! Table 2 Summarized results of Cbx protein immunopr ecipitations Transcript 1 Transcript 2 Transcript 3 pVenus Cbx2 Very Weak None Robust pVenus Cbx4 Very Weak None Very Weak pVenus Cbx6 Robust None Robust pVenus Cbx7 Very Weak None Very Weak pVenus Cbx8 None None Robust Figure 10 Cbx Protein Immunoprecipitation of LINC00486 Transcripts 1 3. N2A cells were cotransfecte d with individual Cbx proteins and LINC00486 Transcripts 1 3. A F) Transcript specific primers were used in RT PCR to detect presence of transcript pull down from Cbx immunoprecipitations. (A E) Each Flag tagged Cbx protein was individually immunoprecipita ted from nucleic isolate. (F) GFP tagged Cbx2 protein was immunoprecipitated from N2A nucleic isolate. (Expected sizes Transcript 1: 105bp; Transcript 2: 84bp; Transcrip t 3: 106bp; GAPDH: 150bp, HOTAIR : 250bp. GAPDH primers were used as a cont rol for RNA i ntegrity and HOTAIR was used as a positive control for a representative lncRNA).

PAGE 58

! (( ! CHAPTER III INVOLVE MENT OF GSK 3 IN STEM CELL PLURIPOTENCY Introduction Pluripotency in Mouse Embryonic Stem Cells Mouse embryonic stem cells (mESCs) originate from the inner cell mass of pre implantation mouse embryos and are characteristically pluripotent, meaning an indi vidual cell can give rise to all cell types (Burdon, 2002) Uniquely, mESCs are also capable of indefinite self renewing expansion when cultured (Loh et al., 2006) When necessary, env ironmental signals can induce differ entiation into specific cell types, establishing a promising model for regenerative disease (Loh et al., 2006; Welham et al., 2011) In order to establish and maintain pluripoten cy, extrinsic factors, signaling pathways, and a regulatory network of transcription factors synergistically work together. The Master' transcription fa ctors (Nanog, Oct4, Klf4, Esr rb etc. ) are essential for establishing pluripotency (Pei, 2009) These transcription factors contain similar DNA binding and transactivation domains, which are necessary to interact with important cofactors. Additionally, these genes are tightly controlled by feedback circuits, which allow them to self regulate as well as regulate each oth er. Data suggests when these genes are suppressed, pluripotent ESCs begin to differentiate (De et al., 2014) However, when Gsk 3 is inhibited there is increased expression of the Master' transcription factors all owing maintenance of ESC proliferation and pluripotency (Welham et al., 2011) A more in depth focus on these Master' transcription factors shows that Nanog is a homeodomain containing protein required for the m aintenance of pluripotency, which occurs through the modulation of Oct 4 (Chen et al., 2011; Loh et al., 2006) Oct4 is a

PAGE 59

! () ! POU transcription factor, encoded by Pou5f1. In the absence of Oct4, differentiation occurs, suggesting that Oct4 is essential for maintaining the identity of pluripotency in ESCs. Together, Nanog and Oct4 control the downstream genes required for maint aining pluripotency, such as Esr rb (estrogen related receptor ) (Loh et al., 2006) Esr rb, a direct transcriptional target of Nanog, is part of a nuclear hormone receptor superfamily (Sanchez Ripoll, 2013) It also lies downstream of Gsk 3 and is required for suppressing differentiatio n (Hanna, 2010) During Gsk 3 inhibition, Esr rb gene expression is increased (Martello et al., 2012) Finally, it is speculated that Klf4 (KrŸppel like factor 4) helps to regulate differentiation through modifying the chromatin structure of downstream pluripotent genes. This allows the genes to bind to their target s and activate the genes that are normal ly silenced in differentiated cells (Romeo, 2012) These genes work together in a tightly regulated network to maintain pluripotency in ESCs. Effect of LIF on Stem Cell Pluripotency The key factor in m ESC pluripotency and self renewal i s LIF (leukemia inhibitory factor), which activates the STAT3 (signal transducer and activator of transcription 3) signal ing pathway and c M yc. Once LIF binds to the receptor it results in the activation of extracellular regulated kinases (Erk1 and Erk2) and PI3K signaling (Welham et al., 2011) There is a necessary balance between these pathways because activation of STAT3 and PI3K promote pluripotency and self renewal through Gsk 3 inhibition, while ERK activatio n promotes differentiation. This important balance determines the fate of m ESCs (Storm et al., 2007) ESCs will naturally respond to environmental signals to change from a pluripotent state to a differentiated stat e, b ut in cultured medium, o ther extrinsic factors such as s erum or Bone Mo rphogenetic Proteins 2

PAGE 60

! (* ! or 4 (BMP 2 or BMP4 ) are supplemented to work synergistically with LIF to tip the balance in order to sustain pluripotency (Chen et al., 2011; Sanchez Ripoll, 2013) Without supplementing LIF, mouse embryo fibroblast ( feeders ) can be used to co culture with mESCs to sec rete necessary growth factors and maintain pluripotency (Tamm, Pijuan Galito, & Ann eren, 2013) Ground state pluripotency can also be created by growing ESCs in serum free media in the absence of LIF with the simultaneous inhibition of both Gsk 3 and MEK (Mitogen activated and extracellular regulated kinase kinase) through the addition of the two specific inhibitors (CHIR99021 and PD184352m respectively; 2i) (Sanchez Ripoll, 2013; Ying et al., 2008) Inhibiting the MEK signaling pathway results in the inhibition of the differentiation inducing ef fects secreted by FGF4 (fibroblast growth factor 4). However, introducing the cells to only MEK inhibitors caused the cells to grow poorly. When the Gsk 3 inhibitor was added, the cell state was drastically improved, providing insight into the requirement for Gsk 3 inhibition to stimulate cellular proliferation and metabolism. Th is is consistent with the r ole s of PI3K and Wnt in inhibition of Gsk 3 and regulating proliferation of ESCs. Additionally, Gsk 3 DKO ESCs mimic effects of Gsk 3 inhibitor s and are m ore resistant to differentiation than WT ESCs (Welham et al., 2011) Effect Gsk 3 Inhibition on Stem Cell Pluripotency Inhibiting Gsk 3 has shown to enhance self renewal of ESCs and facilitates resistance to diff erentiation (Sanchez Ripoll, 2013) Previous studies have shown that when Gsk 3 was inhibited through the PI3K pathway, expression of the Master' pluripotency transcription factors were altered at the protein level (Sanchez Ripoll, 2013) Nanog, a core element of the pluripotency network, is imperative for cell vitality

PAGE 61

! (+ ! and resisting differentiation (Storm et al., 2007) Gsk 3 inhibition cultured in the absence of LIF re sulted in selective upregulation in Nanog protein levels, which preceded changes in RNA levels. These results indicate that these changes are translational, but not transcriptional. However, the stability of these proteins were not altered (Sanchez Ripoll, 2013) Despite the numerous studies suggesting Gsk 3 inhibition promotes pluripotency, the exact mechanism behind this evidence remains unsolved. Another recent discovery relating to Gsk 3 inhibition is now commonly used to benefit researchers working with induced pluripotent stem cells ( iPSCs ) The addition of ascorbic acid (commonly known as Vitamin C) and a Gsk 3 inhibitor (called AGi) has been shown to be the most effective combination of treatments to reprogram mouse somatic ce lls into iPSCs Vitamin C acts to alleviate cell senescence (Bar Nur et al., 2014) During the reprogramming process, tumor suppressive mechanisms (largely through tumor protein p53) are activated, which initiates cellular senescence to act as a barrier to iPSC generation (Banito & Gil, 2010) Vitamin C relieves this barrier by reducing p53 activity to promote the generation of iPSCs (Esteban et al., 2010) To reprogram somatic cells Vitamin C and Gsk 3 inhibition work synergistically by initiating early activation of the key pluripotency genes (Bar Nur et al., 2014) Additionally, Vitamin C is a known cofactor for histone demethylases and al pha ketoglutarate dependent dioxygenases. It is suggested that Vitamin C may allow reprogramming to run more smoothly by aiding in the histone demethylation process (Esteban et al., 2010) Previously, forcing the c ellular change from somatic cells to iPSCs was only possible through tedious and time consuming genetic manipulation (Bar Nur et al., 2014) Today, t he addition of AGi improves the efficiency and speed of iPSC

PAGE 62

! (, ! form ation. It remains to be determined how these mechanisms fully work; yet AGi provides great potential into making the transition from somatic cells to iPSCs more efficient. m 6 A mRNA Modification RNA epigenetics plays an important role in stem cell pluripo tency. Over a hundred different post transcriptional modifications decorate RNA molecules, however our knowledge of their location and function are limited (Dominissini et al., 2012) N 6 methyladensoise (m 6 A) is th e most abundant posttanscriptional modification found on thousands of mRNAs and lncRNAs in somatic cells. It is suggested that m 6 A abundance is roughly 0.1% 0.4% of total adenosine residues in m RNA ( Meyer et al., 2012) The novel approach to identify and localize m 6 A sites, m 6 A seq, has exposed the RNA methylome to improve our understanding of this modification. m 6 A seq has identified the fact that methylation can be found individually or in clusters and are typically enriched near the stop codon or within the 3' untranslated region (UTR) in mRNA (Linder et al., 2015) The presence of this modification near stop codons suggests translational control (Dominissini et al., 2012) Association with 3' UTR indicates a role in RNA regulation because the 3'UTR influences RNA stability, subcellular localization and translational regulation. Additionally, there was no m 6 A found in the mRNA poly A tail (Meyer et al., 2012) Performing m 6 A seq on RNA isolated from HEK293T cells over 12,000 m 6 A sites were identified in the transcripts of more than 7,000 human genes, indicating this modification is very widespread throughou t the transcriptome (Dominissini et al., 2012) mRNA methylation is non stoichiometric, meaning only a fraction of transcripts are methylated at a specific position. In addition, some transcripts are completely voi d of

PAGE 63

! (! methylation. These observations suggest m 6 A locations are non random and most likely play a regulatory role in translation (Dominissini et al., 2012) This modification is found on genes that are involved in n umerous cellular functions including transcriptional regulation, intracellular signaling pathways, and have been linked to diseases such as cancer and obesity. Additionally, m 6 A is prominently found on transcripts that encode the major pluripotency transcr iption factors Research has sh own that this mRNA modification is directly correlated to stem cell pluripotency reduced levels of m 6 A result in the inability for ESCs to differentiate (Batista et al., 2014) Althou gh this modification is essential for cellular development and viability, the precise role of the addition of m 6 A on m RNA remains to be determined (X. Wang et al., 2014) The addition of the m 6 A modifi cation is ca talyzed by Mettl3 (M ethyl Transferase like 3) and a related, but uncharacterized protein, Mettl14 (Y. Wang et al., 2014) These proteins add methyl groups onto the C 6 position of adenosine, which does not hinder it s ability to base pair with thymidine or uracil (Guela, 2015; Meyer et al., 2012) Two members of the alpha ketoglutarate dependent dio xygenases protein family, FTO (F at and Obesity Gene) and Alkbh 5 (A lkB homolog 5) are the demethylases that oxidatively remove the methyl group from adenosine s (Zheng et al., 2013) This proces s is catalyzed by the glutarate dependent enzyme co factor Vitamin C (Gerken, 2007) FTO is a regulator of metabolism and energy utilization and SNPs with increased levels of FTO expression resulted in an elevated body mass and an increased risk for obesity (Meyer et al., 2012) FTO is present not only in the nucleus, but in the cytoplasm where posttranscriptional modifications occur, such as the addition of m 6 A (Gerken, 2007) When FTO was overexpressed in human cells, there was a decrease in the amount of m 6 A

PAGE 64

! ). ! observed. In contrast, when FTO was knocked down, increased amounts of m 6 A on mRNA was observed (Jia et al., 2011) These data validate the role of FTO as a demethylase and demonstrates RNA methyl ation is a reversible process that regulate s mRNA metabolism (X. Wang et al., 2014) Recently, it has been suggested that the addition of m 6 A directly reduces the stability of mRNA and therefore decreases translation frequency (Guela, 2015) In general, a large number of transcripts exhibiting the m 6 A modification show an inverse correlation with mRNA stability and gene expression (Y. Wang et al., 2014) The transcripts marked with m 6 A showed a reduced half life and increased rates of decay, indicating the association between m 6 A and transcript turnover (Batista et al., 2014) This is due to t he binding of the YTH domain family proteins to m 6 A sites, which recruit the transcripts to RNA decay processing bodies (P bodies) (Batista et al., 2014) Five different YTH domain family members have been identifi ed, two of which have been highly studied (Y. Yang et al., 2015) YTDHF1 and YTDHF2 play a role in mRNA translational abilities (X. Wang et al., 2014) The YTH domain family is known to bind onto the C terminal YTH domain of single stranded RNA. Once bound, each protein has a unique role to ensure the m 6 A marked mRNA has efficient gene expression and controllable production of proteins (Wang et al., 2015) The carboxyl terminal domain of YTHDF2 selectively binds the m 6 A methylated mRNA while the amino terminal domain localizes the YTHDF2 mRNA complex to mRNA P bodies. This indicates YTHDF2 has the ability to affec t translation and the lifetime of mRNA (X. Wang et al., 2014) Working synergistically with YTHDF2, YTHDF1 directly promotes the translation efficiency of methylated mRNAs by trafficking more m 6 A mRNA transcripts

PAGE 65

! )% ! to translational machinery. Additionally, YTHDF1 directly accelerates the translation initiation rate once the mRNA is bound to the ribosome (Wang et al., 2015) YTHDF1 and YTHDF2 possess very distinct functions, w hich intuitively result in two opposing mRNA fates. This suggests a more complex understanding of the overall m 6 A function. In order to surpass this issue, data has shown that both proteins have their own individual sets of target mRNAs as well as a large set of common target mRNAs. For consistent regulation of the shared targets, YTHDF1 binds to mRNA prior to YTHDF2 in order to activate the translation of the methylated mRNAs for optimal production of proteins. Once YTHDF2 binds, the lifetime of the mRNAs is limited. This dynamic gene expression regulation allows for a more precise and effective process for protein production with a decreased response time between the stimulus and the translational output. This occurs during various biological transformatio ns, one of which is promoting cellular differentiation (Wang et al., 2015) Together, Mettl3, FTO/Alkbh5 and YTHDF1/2 act as "molecular switches" that regulate the m 6 A modification on mRNA (Guela, 2015) Using epigenetic terminology, Mettl3 acts as the writer', FTO/Alkbh5 are the erasers' and YTHDF proteins are the readers' of m 6 A modifications. Despite identifying these epigenetic regulators, the precise mechanism behind how these enzymes know when to add or remove the m 6 A tags is still unknown. However, since the addition of m 6 A plays a critical role in mRNA degradation and turnover, it can be suggested that the m 6 A process is highly regulated and is capable of being finely tuned in response to varying environmental stimuli. An interesting observation from m 6 A RNA reductions is maintenance of pl uripotency in ESCs, even under conditions when differentiation should occur. Mettl3

PAGE 66

! )& ! has been shown to terminate pluripotency and promote differentiation in ESCs. When Mettl3 was knocked out, cells showing normal pluripotent morphology did not start to diff erentiate (Y. Wang et al., 2014) Add itionally, the expression of Esr rb and Nanog was not altered. This indicates that ablation of Mettl3 depletes almost all m 6 A tags on mRNA, increasing the stability of the plurip otent mRNA transcripts as well as the translational efficiency, resulting in the maintained pluripotent state (Batista et al., 2014) To take this one step farther, when the Mettl3 KO ESCs were grown in the absence of LIF, they became even more resilient to differentiation, further validating the idea that Mettl3 terminates pluripotency (Guela, 2015) This data suggests that the addition of m 6 A may play an important role in stem cell differentiation. Interestingly, the Mettl3 KO and the Gsk 3 DKO ESCs showed a very similar phenotype due to their inabi lity to differentiate. This observation inspired us to hypothesize that the addition of the m 6 A modification on mRNA is regulated by Gsk 3 activity through FTO. We hypothesize that Gsk 3 phosphorylates FTO, resulting in FTO ubiquitination and degradation. Reduced FTO levels would lead to reduced demethylase activity, which would cause an increase in m 6 A m RNA promoting the degradation of the mRNA, resulting in the differentiation of ESCs. However, in the absence of Gsk 3 activity FTO would remain active, r esulting in reduced m 6 A levels and decreased mRNA degradation, allowing the cells to maintain in the pluripoten t state. If our hypothesis is correct, this discovery of Gsk 3, a multifaceted regulator of multiple signaling pathways, as the controller of m 6 A modifications on mRNA, would provide the first mechanistic explanation into how this RNA epigenetic process is regulated. It would also provide

PAGE 67

! )' ! insight into understanding how different Gsk 3 dependent signaling pathways regulate the fine balance between d ifferentiation and pluripotency in ESCs. The purpose of this project is to examine the relationship between Gsk 3 and stem cell pluripotency by a novel mechanism: the regulation of the m 6 A modification on mRNA. We hypothesize the m 6 A tag is regulated by G sk 3 through controlling the protein levels FTO, a demethylase. Inhibition of Gsk 3 would result in increased FTO protein levels, resulting in reduced m 6 A levels and maintenance of stem cell pluripotency. Identifying the regulatory role of Gsk 3 will provi de a mechanistic understanding of stem cell differentiation as well as directly link m 6 A modifications with Gsk 3 function. As previously stated, the specific aims include: Aim #1: Determine if the effect of Gsk 3 on stem cell pluripotency is through the regulation of mRNA methylation. Aim #2: Investigate if Vitamin C and Gsk 3 inhibition work synergistically to promote stem cell pluripotency Materials and Methods RNA Immunoprecipitation and ELISA: m 6 A Immunoprecipitation Anti N6 methyladenosin e (m 6 A) (Mi llipore) was coupled to dynabeads using the Dynabead Antibody Coupling Kit from Life Technologies. 10mg dynabeads was added to 50uL m 6 A antibody (at 1ug/mL) to make a final concentration of 10mg/mL. The coupling process was performed by following the kit p rotocol. Immunoprecipitation protocol was similar to EZH2 protocol, except 40uL of Anti m6A beads were added to untransfected WT and DKO cells resuspensions. No control beads were used. After

PAGE 68

! )( ! successful immunoprecipitation of m 6 A mRNA, qPCR was performed t o quantify pluripotency gene expression. m 6 A ELISA U sing Qubit 1. Total RNA was isolated (Zymo) from untransfected N2A cells; the additional DNase restriction step was performed. 2. Per RNA sample, 50uL of Oligo d(T) magnetic beads were added 200uL Lysis/Bindi ng buffer (mirVana Kit). Mixture was vortexed briefly and then mixed on rotator for 2 minutes at room temperature. 3. 5ug of total isolated RNA was added to magnetic beads. Tubes were placed on rotator to incubate overnight at 4¡C. 4. Using a magnetic plate, the original liquid was taken off beads and saved as "dT Other". 5. Beads were washed with 100uL RNase free water and water was discarded. Wash was repeated 2 more times. 6. To elute RNA off of beads, 20uL cold 10mM Tris HCl pH 7.5 was added to RNA and beads. Mixture was incubated at 80¡C for 2 minutes. 7. After incubation, tube was immediately placed on magnet and eluted RNA was placed into a new tube on ice. 8. RNA was quantified using Qubit to determine how much poly A RNA was pulled down from Oligo dT beads. 9. De sired amount of RNA was added to 10uL, 50uL and 100uL of m 6 A beads. Negative controls with the same volume of m6A beads and no RNA were created.

PAGE 69

! )) ! 10. Experimental and control tubes were rotated for 2 hours at 4¡C. 11. After incubation, tubes were placed on a magne tic plate and original liquid was removed and saved as "m 6 A Other". 12. Beads were washed with 100uL RNase free water and water was discarded. Wash was repeated 2 more times. 13. To each tube, 300uL Elution Buffer (Appendix) was added. Samples were incubated at 50¡C for 1.5 hours to elute RNA off of beads. 14. After incubation, tubes were placed on magnetic plate and eluted RNA was placed into a new tube on ice and saved as "After m 6 A". 15. RNA was quantified using Qubit to determine how much m 6 A RNA was immunoprecipita ted from Anti m6A beads. 16. To purify RNA, 1 volume of isoamyl:chloroform was added to 1 volume of Elution Buffer (300uL). Tubes were vortexed vigorously for 15 seconds. Tubes were then left to homogenize at room temperature for 3 minutes. 17. Tubes were centrifu ged at 12,000 xg for 15 minutes at 4¡C. Upper aqueous phase was transferred to a fresh tube. Volume of this phase was determined. 18. 0.1 volume of 3M Sodium Acetate pH 5.2 and 2 volumes of ice cold Ethanol was added to each tube. Tubes were mixed. 19. Ethanolic solution was incubated from 1 hour to 18 hours (overnight) at 20¡C to allow RNA to precipitate. 20. After incubation, RNA was recovered by centrifugation at 13,000 x g at 4¡C. 21. All traces of supernatant were removed without disturbing the invisible pellet.

PAGE 70

! )* ! 22. The pellet was washed with 500uL ice cold 70% Ethanol. Tubes were centrifuged at maximum speed for 10 minutes at 4¡C. Steps 21 and 22 were repeated. 23. Tubes were left open at room temperature for all traces of Ethanol to evaporate. However, the RNA pellet wa s not left to dry completely. 24. After evaporation, RNA pellet was dissolved in 30uL DEPC water. The walls of the tube were rinsed with the water to obtain as much RNA as possible. To aid solubilization, RNA pellet was incubated in resuspensions solution fo r 5 minutes at 65¡C with intermittent gentle vortexing. 25. Tubes were stored as "Final RNA" and quantified using the Qubit. All RNA was stored at 80¡C. m 6 A ELISA (Based Off Epigentek P rotocol) 1. Total RNA was isolated (Zymo) from untransfected N2A cells; the additional DNase restriction step was performed. 2. In a 96 well plate, 80uL DNA/RNA Binding Buffer (Zymo) was added to each well. 3. 5ug of total isolated RNA was added to designated wells. Solution was mixed gently by tilting plate side to side, ensuring th e solution coated the bottom of each well evenly. 4. The plate was covered with a lid and incubated at 37¡C for 90 minutes. 5. After incubation, DNA/RNA Binding Buffer was removed from each well. Each well was washed with adding 150uL TBST. TBST was then removed and discarded. Wash was repeated 2 more times.

PAGE 71

! )+ ! 6. m6A Antibody was diluted 1:1000 in TBST. 50uL of diluted m 6 A Antibody was added to each well. Plate was covered with lid and incubated at room temperature for 60 minutes. 7. After incubation, m 6 A Antibody was re moved from each well and each well was washed with 150uL TBST for a total of three times. 8. Anti Rabbit antibody was diluted 1:5000 in TBST. 50uL of Anti Rabbit Antibody was added to each well. Plate was covered and incubated at room temperature for 60 minut es. 9. After incubation, Anti Rabbit antibody was removed from each well and discarded. Wells were washed with 150uL TBST for a total of four times. 10. PNPP was equilibrated to room temperature and was mixed end over end. 100uL of PNPP was added to each well. 11. Plate was gently agitated to mix thoroughly and then incubated at room temperature until color developed. 12. Reaction was stopped by the addition of 100uL 2M Sulfuric Acid. Plate was agitated gently to mix. 13. Plate was placed in illuminometer and the absorbanc e was measured at 450nm. Process was repeated with AP linked Antibody and TMB to determine optimal protocol. Observing Protein Expression: Western Blotting

PAGE 72

! ), ! For FTO westerns, membrane was blocked in 5% milk/TBST for 1 hour at room temperature, while rota ting. Primary FTO antibody (PhosphoSolutions Reference#: 597 FTO) was added to the membrane at 1:1000 in 10mL of 1% milk/TBST. Membrane was incubated in primary antibody overnight at 4¡C while rotating. Membrane was then washed with TBST (1x for 15 min and 2x for 5 min) and then Anti Mouse secondary antibody was added at 1:20,000 in 10mL of 1% milk/TBST for 30 minutes at room temperature while rotating. Membrane was washed with TSBST (3x for 5 minutes) and then was ECL was added to membrane for 5 minutes. M embrane was blotted dry and then imaged. For Alpha Tubulin westerns, membrane was blocked in 4% BSA/TBST for 1 hour at room temperature, while rotating. Primary Alpha Tubulin antibody (Cell Signaling Reference#: A11126) was added at 1:1000 in 10mL of 4% B SA/TBST. For secondary antibody, Anti Mouse was added at 1:20,000 in 10mL of 4% BSA/TBST. Similar procedure was followed as used in FTO westerns. For Gsk 3 "/! westerns, membrane was blocked in 5% BSA/TBST for 1 hour at room temperature, while rotating. Primary Gsk 3 "/! antibody (CalBioChem Reference #: 368662) was added to membrane at 1:1000 in 10mL 5% BSA/TBST. For secondary antibody, Anti Mouse wa s added at 1:20,000 in 10mL 5% BSA/TBST. Similar procedure was followed as used in FTO westerns. Results Inhibition of Gsk 3 has been shown to maintain stem cell pluripotency. However, the mechanism behind how this works is yet to be established. Due to t he similar phenotypes between METTL3 KO and Gsk 3 DKO ESCs, we reasoned these genes could

PAGE 73

! )! be working within the same pathway. Therefore, we hypothesized this resistance to differentiation in Gsk 3 DKO ESCs may be due to the addition of m 6 A through the regu lation of Gsk 3. m 6 A levels on mRNA directly correlate with pluripotency; ESCs that maintain pluripotency have reduced m 6 A mRNA levels. As these methylation levels increase, stem cells begin to differentiate. Due to the common knowledge that Gsk 3 is a neg ative regulator, we began by exploring the possibility that Gsk 3 is regulating mRNA methylation by the phosphorylation and suppression of the demethylase, FTO possibly by targeting for degradation We hypothesize that when Gsk 3 is inhibited, there is an increase in FTO protein levels, resulting in decreased m 6 A levels, which leads to enhanced pluripotency. Fortunately, we had the necessary reagents and tools to begin dissecting these questions. Pluripotency Gene Expression in Immunoprecipitated m 6 A RNA To determine the relative levels of m 6 A modifications found on specific mRNAs, we immunoprecipitated m 6 A RNA from untransfected WT and Gsk 3 DKO cells grown in complete ESC media with the addition of LIF. After lysing and adding anti m 6 A magnetic beads, fo llowed by incubation, all liquid was removed from beads (Input) and m 6 A RNA was then isolated off of the beads. RNA was isolated from the Input liquid and cDNA was synthesized from both Input and m 6 A mRNA. qPCR was performed to observe GAPDH, Nanog, and Kl f4 gene expression (Figure 11). Results revealed relatively no change in expression levels, which did not support our hypothesis. Next, we wanted to determine the half life of the m 6 A mRNA. This was accomplished by treating WT and Gsk 3 DKO cells with amanatin, a potent inhibitor of RNA polymerase II. By blocking RNA polymerase II, the cell is unable to enter the next round of synthesis,

PAGE 74

! *. ! thereby halting transcription (Brueckner & Cramer, 2008) This allows for the determination of the stability of mRNA through qPCR. To determine the half life of Nanog in m 6 A mRNA, the m 6 A immunoprecipitation was repeated in WT and Gsk 3 DKO ESCs grown in complete ESC media with the addition of LIF. These cells were treated with 1ug/mL amanatin for 0 or 2 hrs. qPCR revealed a higher CT mean for m 6 A IP of WT at the 2 hour treatment and th e m 6 A IP of DKO at the 0 hour, indicating reduced Nanog expression at these time points. The CT mean for the input of both WT and DKO at all time points remained relatively constant (Figure 11D). These results did not match up with our hypothesis. Next, w e were curious about FTO protein levels in ESCs. In normal ESC culturing conditions with the addition of LIF, FTO protein expression is high in both WT and Gsk 3 DKO ESCs (" tubulin western served as a loading control) (Figure 12) This was anticipated bec ause the addition of LIF maintains pluripotency in the WT ESCs. Additionally, Gsk 3 DKO ESCs remain pluripotent due to the ablation of Gsk 3, even in the absence of LIF. As previously mentioned, pluripotency is correlated with m 6 A levels pluripotent stem cells have high levels of m 6 A due to the suppressed FTO demethylase function. Unfortunately, the immunoprecipitation and western blot results did not support our hypothesis. We concluded that we needed to find a way to separate the FTO mechanism from plur ipotency because cells grown in the presence of LIF maintain pluripotency and mask the mechanistic effects of FTO. Therefore, we decided to manipulate the cells by removing LIF to see the mechanistic changes on FTO protein levels.

PAGE 75

! *% ! Figure 11 m 6 A mRNA Gene Expression in WT and Gsk 3 DKO ESCs. m 6 A mRNA was immunoprecipitated from total RNA isolated from WT and Gsk 3 DKO ESCs gr own in complete ESC media in the presence of LIF qPCR was performed to look at (A) GAPDH, (B) Nanog and (C) Klf4 gene expression in the m 6 A IP compared to Input total RNA. (D) Prior to m 6 A IP, WT and Gsk 3 DKO ESCs were treated with amanatin and total R NA was isolated at zero or two hour time points. m 6 A mRNA was immunoprecipitated from RNA and qPCR was performed to look at Nanog gene expression. Figure 12 FTO Protein Expression in WT and Gsk 3 DKO ESCs Cultured in Complete E SC Media i n the Presence of LIF. Protein band sizes are indicated on left side. tubulin western served as a loading control

PAGE 76

! *& ! Effect of LIF on mESC Protein and Gene Expression A paper from the Samuel Lunenfeld Research Institute demonstrated that Gsk 3 DKO ESCs remained pluripotent even after removing LIF for 14 days (Doble et al., 2007) Therefore, we decided to use the same experimental regimen We grew WT and Gsk 3 DKO ESCs in normal ESC media in the absence of LIF for 14 days. In addition to these cell types, we decided to include p110* ESCs. These cells have the myristolated form of the PI3K subun it, p110", which promotes a constitutively active insulin pathway, resulting in the constant inhibition of Gsk 3. We were curious to see how these cells would maintain pluripotency in comparison to Gsk 3 DKO ESCs when grown without LIF. On day 14, protein was isolated and quantified using Qubit. Next, we determined protein expression by performing FTO, tubulin (Figure 13A) and Gsk 3 "/! western blots (Figures 13 A B). FTO expression in Gsk 3 DKO ESCs was much more robust than in WT or p110* ESCs (Figures 13 A B). The tubulin western served as a loading control and the Gsk 3 "/! western was performed to validate sufficient knockout of both isoforms in the Gsk 3 DKO cells. In addition to protein, RNA was isolated from the embryonic stem cells (Zymo) and cDNA was synthesized (Applied Biosystems). We then assessed gene expression for Nanog, Oct4, and Esrrb (the Master' pluripotency genes). Gene expression was quantified through qP CR in order to observe levels of pluripotency in cells when LIF is absent (Figure 14). qPCR results revealed almost a 6 fold increase in Nanog expression in Gsk 3 DKO without LIF for 14 days compared to WT without LIF. The Nanog expression in WT cells treated without LIF for 14 days was almost identical to both the WT and Gsk 3 DKO cell s with LIF for 14 days (Figure 14A). Esrrb expression in Gsk 3

PAGE 77

! *' ! DKO cells treated without LIF for 14 days increased 10 fold compared to WT without LIF and WT and Gsk 3 DKO plus LIF (Figure 13B). Due to the consistencies in the high expression of pluripotenc y genes in Gsk 3 DKO ESCs when grown in the absence of LIF, we wanted to further investigate exactly how much higher the gene expression would be in other pluripotency genes. When qPCR results for Nanog, Oct4 and Esrrb expre ssion in WT and Gsk 3 DKO ESCs w ithout LIF for 14 days were combined, Esrrb expression in Gsk 3 DKO ESCs without LIF had the greatest increase in expression with an eleven fold increase compared to WT without LIF for all three genes (Figure 14C). Next, we were curious about FTO gene expr ession in WT and Gsk 3 DKO ESCs when grown without LIF for 14 days. Results indicate a slight decrease in FTO expression in the Gsk 3 DKO cells, which is consistent with our assumption that regulation by FTO occurs at the translational level, not at the tr anscriptional level (Figure 14D). After seeing successful results that supported our hypothesis, we wanted to see if we could alter the cellular medium conditions to speed up the time frame. First, we repeated the no LIF experiment in complete ESC media. However, this time we isolated protein on various days to compare to protein isolated on day 14. WT and Gsk 3 DKO ESC protein isolated on day two did not show robust FTO expression in Gsk 3 DKO ESCs compared to day 14 (Figure 15A). Therefore, we concluded that optimal FTO expression is seen in cells grown for 14 days in the absence of LIF in normal ESC media. tubulin was used as a protein loading control. To take this optimization process one step further, we wanted to determine if we could alter the me dium conditions and receive the same, if not better, results. WT and Gsk 3 DKO ESCs were grown in N2B27 media supplemented with 10% FBS

PAGE 78

! *( ! (Appendix). These cells were grown in the absence of LIF for a total of four days. Each day, protein wa s isolated from a subset of cells. At the end of the four day experiment, western blots were performed to determine which day had the optimal FTO expression in W T and Gsk 3 DKO ESCs (Figure 15B ). The tubulin western was used as a protein loading control and the Gsk 3 "/! western was performed to validate sufficient knockout of both isoforms in the Gsk 3 DKO cell. Unfortunately, FTO expression in Gsk 3 DKO ESCs was low in comparison to the Gsk 3 DKO ESCs grown in complete ESC media in the absence of LIF for 14 days. This allowed us to conclude that the most optimal medium condition for robust FTO expression is growing cells in complete ESC media in the absence of LIF for 14 days. By optimizing th ese cellular conditions, it allowed us to move forward and proceed on downstream experiments.

PAGE 79

! *) ! Figure 13 FTO Expression in ESCs Cultured in Complete ESC Medi a in the Absence of LIF for 14 D ays. FTO expression was observed in (A and B) WT, Gsk 3 DK O and (B) p110* ESCs that were grown in the absence of LIF for 14 days. (B) A subset of WT, Gsk 3 DKO and p110* ESCs were grown with the addition of 50ug/mL Vitamin C. Protein band sizes are indicated on the left. The tubulin western was used as a protei n loading control and the Gsk 3 "/! western was performed to validate sufficient knockout of both isoforms in the Gsk 3 DKO cell.

PAGE 80

! ** ! Figure 14 Pluripotency Gene Expression in ESCs Grown in the Absence of LIF for 14 Days. Comparison of (A) Nanog and (B) Esrrb gene expression in WT and Gsk 3 DKO cells grown in the presence and absence of LIF for 14 days. (C) Quantification of Nanog, Oct4 and Esrrb and (D) FTO gene expression in WT and Gsk 3 DKO cells grown in the absence of LIF for 14 days.

PAGE 81

! *+ ! Figure 15 Optimizing ESC Media Conditions for Detection of Robust FTO Protein Expression. (A) Testing time frame of WT and Gsk 3 DKO ESCs cultured complete ESC media in the absence of LIF. Protein isolated on day two of n o LIF experiment was compared to cells on day 14. (B) WT and Gsk 3 DKO ESCs were grown in N2B27 media with 10% FBS. FTO protein levels were observed in cells that were isolated each day during four day experiment. The tubulin western was used as a protei n loading control and the Gsk 3 "/! western was performed to validate sufficient knockout of both isoforms in the Gsk 3 DKO cell.

PAGE 82

! *, ! Effect of Gsk 3 Inhibition on FTO Protein Expression Since the addition of a Gsk 3 inhibitor to ESC media helps to promote pluripotency in t he absence of LIF, we wanted to ask whether Gsk 3 inhibitors could increase FTO levels in WT ESCs similar to the FTO levels in Gsk 3 DKO ESCs. To determine the effects of Gsk 3 inhibition on FTO levels, WT cells that were grown without LIF for 14 days were then treated with differing concentrations of SB 415,286 or LiCl on day s 15 and 16. On day 16, protein was isolated from cells and FTO and tubulin western blots were performed. Results indicated that increasing amounts of SB 415,286 increased FTO expression, with the most robust expression seen in cells treated with 15uM SB (Figure 16 A). The LiCl treated cells showed a similar trend in increasing FTO expression with increasing LiCl concentrations. The cells treated with 10mM LiCl showed the most robust FTO expression (Figure 16 B). These data show that FTO protein levels in WT ESCs can be rapidly increased by Gsk 3 inhibition in a dose dependent fashion.

PAGE 83

! *! Figure 16 Effect of Gsk 3 Inhibition on FTO Protein Expression in WT ESCs. WT cells were cultured i n complete ESC media in the absence of LIF for 16 days. On days 14 and 15, cells were treated with varying concentrations of (A) SB 415,286 or (B) LiCl. Protein was isolated on day 16 to observe the effect of Gsk 3 inhibition on FTO expression in a dose d ependent ma nner. The tubulin westerns were used as a protein loading control.

PAGE 84

! +. ! Effec t of Vitamin C on Pluripotency Vitamin C play s a role in promoting pluripotency in the generation of iPSCs from somatic cells (Bar Nur et al., 2014) While Vitamin C is best known as an anti oxidant, it has also been shown to serve as a co factor for alpha ketoglutarate dependent dioxygenases, such as FTO. Therefore, we wondered if Vitamin C is promoting pluripotency b y enhancing FTO mRNA demethylase activity. To answer this question, WT, Gsk 3 DKO, and p110* ESCs were grown in the absence of LIF for 14 days. Every day during the 14 day exper iment, a subset of the cells was treated with a relatively low dose of Vitamin C ( 50ug/mL ) Over the 14 day experiment, the cellular mo rphology changed greatly in WT plus Vitamin C compared to without Vitamin C (Figure 17A). The morphology of Gsk 3 DKO and p110* ESCs stayed relatively consistent over the 14 day experiment (Figures 17 B C). On day 14, protein was isolated and western blots were performed to observe FTO expression (Figure 13B). Results indicated there was no change in FTO protein levels between the WT and Gsk 3 DKO ESCs +/ Vitamin C. H owever, p110* cells treated with V itamin C revealed robust FTO expression compared to without Vitamin C. The robust expression was almost identical to Gsk 3 DKO ESCs +/ Vitamin C. The tubulin western served as a loading control and the Gsk 3 "/! western was performed to validate sufficie nt knockout of both isoforms in the Gsk 3 DKO cells. On day 14, RNA was also isolated from WT, Gsk 3 DKO, and p110* ESCs +/ Vitamin C. cDNA was synthesized and qPCR was per formed to quantify Nanog Oct4, and Esr rb expression levels. For the DKO cells that were treated with 50ug/mL Vitamin C for 14 days in the absence of LIF, there was no change between DKO plus and minus

PAGE 85

! +% ! Vitamin C in Nanog, Esrrb, or Oct 4 gene expression (Figure 18A). This suggests that Vitamin C cannot further enhance the pluripote ncy se en in Gsk 3 DKO ESCs. A ddition ally, we were curious to observe FTO expression in Gsk 3 DKO ESCs cultured in the absence of LIF. Quantitative PCR revealed no change in FTO expression (Figure 18B). However, WT and p110* ESCs treated with Vitamin C revealed a n increase in Nanog, Esrrb, and Oct4 expression compared to cells lacking the Vitamin C treatment (Figure 19). Interestingly Nanog expression in p110* ESCs treated with Vitamin C revealed a 40 fold increase compared to WT ESCs treated with Vitamin C. This qPCR data shows that Vitamin C alone can enhance expression of pluripotency factors, yet this enhancement is not seen in Gsk 3 DKO ESCs.

PAGE 86

! +& ! Figure 17 Morphology of WT, Gsk 3 DKO and p110* ESC s Grown in the Presence and A bsence of Vitamin C. ( A) WT, (B) Gsk 3 DKO and (C) p110* were grown in complete ESC media without LIF for 14 days. A subset of cells was treated with 50ug/mL Vitamin C everyday during 14 day experiment.

PAGE 87

! +' ! Figure 18 Quantified Gen e Expression in Gsk 3 DKO ESCs Cultured in the Absence of LIF with the A ddition of Vitamin C. Gsk 3 DKO ESCs were grown in complete ESC media in the absence of LIF for 14 days. A subset of cells was treated with 50ug/mL Vitamin C everyday during 14 day period. (A) Esrrb, Nanog and Oct4 expression in Gsk 3 DKO ESCs +/ Vitamin C. (B) F TO expression in Gsk 3 DKO ESCs +/ Vitamin C.

PAGE 88

! +( ! Figure 19 Quantified Gene Expression in WT and p110* ESCs Cultured in the Absence of LIF with the Addition of Vitamin C. WT and p110* ESCs were grown in complete ESC media in the a bsence of LIF for 14 days. A subset of cells was treated with 50ug/mL Vitamin C everyday during 14 day period. Esrrb, Nanog, and Oct4 expression in WT and p110* ESCs +/ Vitamin C.

PAGE 89

! +) ! CHAPTER V DISCUSSION Conclusions Characterizing a lncRNA, LINC00486 Functional characterization of LINC00486 Successful cloning into the mammalian expression vector, pCAGEN, allowed for numerous downstream characterizing experiments. RT PCR with transcript specific primers revealed the ability to over express each functi onal transcript in both N2A and HEK293T cells. LINC00486 transcript expression was shown to be present in the nucleus, which was predicted due to the commonly known association between lncRNAs and transcription factors to initiate gene silencing. Presence of LINC00486 in the cytoplasm may have been due to the overexpression of each transcript, which caused an abundant amount of transcript to be found in both parts of the cell. However, i dentification of this lncRNA in the nucleus allowed us to narrow our fu ture exp eriments on the nuclear isolate. Afte r validation of successful over expression of LINC00486 Transcripts 1 3 in mammalian cells, we were curious if endogenous transcript expression could be detected. RT PCR revealed our transcript specific primers could only detect endogenous Transcript 3 expression in HEK293T cells suggesting the presence of this transcript in human embryonic kidney tissue. These results led us to question in what other tissues these LINC00486 transcripts were expressed. However, newly designed primers were required to detect endogenous expressi on of Transcripts 1 and 2. C onfir mation of successful detection confirmed our prediction that three transcripts exist, each unique in sequence. A

PAGE 90

! +* ! panel of human tissue RNA allowed us to run multiple RT PCR experiments to determine the location of each expressed transcript. Robust Transcript 1 expression was observed in salivary gland and colon tissue; Transcript 2 expression was observed in prostate, salivary gland and spleen tissues; Transcr ipt 3 expression was see in whole brain, testes, salivary gland and bone marrow. Each transcript was expressed in different types of tissues, with the exception of the salivary gland, once again supporting the idea that three transcripts of LINC00486 exist Next, we want ed to quantify the expression of each transcript in each human tissue. Using a tissue that revealed no expression during RT PCR as th e control for each transcript, quantitative PCR revealed a much deeper understanding of transcript expressio n Transcript 1 and 3 had the greatest amount of expression in the testes while Tr anscript 2 had the most in the l iver. Transcript 3 was also high in the placenta and the salivary gland. Transcript 1 revealed high expression in the salivary gland, the live r and the spinal cord, and Transcript 2 was high in the salivary gland testes and prostate. By determining what tissues show the g reatest amount of expression of each transcript, we could now focus on specific tissues to continue functionally characterizin g this lncRNA. Due to the large amount of expression of all three tran scripts in the testes and high T ranscript 3 expression in the placenta, we were curious to see if there was any correlation between LINC00486 in germ cells and stem cells Germ cells an d human embryonic stem cells (hESCs) have many geneti c similarities required for developmental process es Therefore, we looked at transcript expression in hESCs and iPSCs. Results revealed low transcript expression in both hESCs and iPSCs compared to the d esignated tissue controls for each transcript, indicating no correlation in LINC00486 transcript

PAGE 91

! ++ ! expression between hESCs, iPSCs and human tissues involved in reproduction and development. Many lncRNAs have shown to have enhancer like capabilities whic h allows them to regulate neighboring genes. We were curious to see if LINC00486 Transcripts 1 3 had the ability to regulate flanking genes, TTC27 and LTBP1. Quantitative PCR revealed modest changes in LTBP1 and TTC27 endogenous expression in HEK293T cells transfected with each transcript. However, when each transcript was overexpressed in N2A cells, LTBP1 expression was almost 2 folds higher in cells transfected with Transcript 3. This result is suggestive that Transcript 3 has the capability to regulate i ts neighbor, LTBP1. Since the LINC00486 transcripts were originally identified in bipolar patients through lithium treatment we were interested in determining transcript expression in other human cell s treated with Gsk 3 inhibitors: LiCl and SB 415,286 HEK2 93T cells treated with either 1 mM LiCl or 3 0uM SB 415,286 for 24 hours revealed relatively no change in Transcripts 2 and 3 in comparison the untreated cells. However, Transcript 1 revealed a slight increase by half a fold in cells treated with LiCl an d almost 2.5 fold increase in cells treated with SB 415,286. These results suggest an association between LINC00486 Transcript 1 and the Gsk 3 pathway which was predicted due to means by which LINC00486 was discovered After establishing a baseline of tr anscript expression in mammalian cells treated with Gsk 3 inhibitors, we were interested in comparing these results to the original lymphoblast cells from the BPD patient (3188) and their unaffected sibling (3189) that were treated with either LiCl or SB 4 15,286 qPCR results indicated a 2 fold increase in

PAGE 92

! +, ! Transcript 3 expression in 3188 (BPD) lymphoblast cells treated with 30uM SB 415,286 for 24 hours. Results also indicated no change in LTBP1 and TTC27 expression in 3188 (BPD) cells treated with 1uM LiCl Finally, we wanted to take a step backwards and look at the endogenous expression pattern in untreated lymphoblast cells from 3188 (BPD) and 3189 (unaffected) LTBP1, TTC27, and Transcript expression were all reduced in 3189 (unaffected) compared to 3188 (BPD) In conclusion, this transcriptional data allowed us to further characterize LINC00486 by gaining insight into the functional capabilities of Transcripts 1 3 as well as its regulation in Gsk 3 signaling pathways. Biochemica l characterization of LINC0 0486 In order to biochemically characterize LINC00486, it was imperative to determine what proteins are associated with this lncRNA. Research has shown many lncRNAs have an association with the PRC2 complex to regulate transcriptional inhibition. Therefore we wanted to investigate if this novel lncRNA followed the same trend. Immunoprecipitation of HA tagged EZH2 in transfected N2A cells revealed the succe ssful pull down of Transcri pt 2. These results indicate a physical association between EZH2 and Transc ript 2 Identical results were seen in the EZH2 IP in HEK293T cells which revealed e ndogenous Transcript 2 immunoprecipitation from HA tagged EZH2. Since EZH2 is a protein within PRC2 which binds to RNA with high affinity t hese results imply LINC00486 T ranscript 2 is physically associated with PRC2 and therefore may have a functional role in genomic silencing. Further investigation is required to determine the mech anisms behind this interacti on and transcriptional regulatio n. H owever, one possibility is the ability for LINC00486 to tether PRC2 to the site of transcription. Once PRC2 reaches the site, the chromatin structure is remodeled to

PAGE 93

! +! induce genomic inhibition. A more in depth examination of this process is required to fully understand this mechanism Many other characterized lncRNAs have also been shown to be associated with the PRC1 complex which maintains genomic silencing through the binding of PRC2. Additionally, the synergistic activity of PRC1 and PRC2 has been shown to play a role in regulat ing pluripotency in mouse ESCs specifically through the individual Cbx proteins PRC1 is comprised of different Cbx subunits each of which has a unique function. We were interested in determining which transcript associates with each Cbx subunit in hope of determining the involvement of LINC00486 Transcripts 1 3 in the unique Cbx functions. RT PCR on Flag tagged Cbx immunoprecipitation s of Transcripts 1 3 revealed varying amounts of Transcript s 1 and 3 that were immunoprecipitated using different Cbx prot eins These results suggest T ranscripts 1 3 are physically associated with PRC1 through different Cbx proteins. Transcript 1 had the most robust expression after the Cbx6 IP and weaker expression for Cbx2, Cbx4 and Cbx7. Additionally, Transcript 3 had the most expression after the Cbx2, Cbx6 and Cbx8 IPs. Cbx7 is considered the main Cbx subunit that maintains pluripotency. Cbx7 accomplishes this through repression of Cbx2, Cbx4 and Cbx8, which prevents premature differentiation. Recent findings have suggest ed that Cbx6 works alone and does not interact with any other PRC1 subunit. It is believed to be recruited to genomic targets by different ncRNAs for an independent function. Therefore, LINC00486 Transcript 1 may recruit Cbx6 to target genes. The function of Cbx6 remains to be established therefore a further investigation is required to reveal the purpose of the association between Transcript 1 and Cbx6. However, d ue to tight association between Transcript 3 and subunits Cbx2 and Cbx8, it can be suggested

PAGE 94

! ,. ! t hat Transcript 3 may have a role in cellular differentiation through PRC1 dependent mechanisms, which remains to be determined. After determining the physical associations between each LINC00486 transcript and the Polycomb Repressive Complexes, it is impo rtant to return back to the BPD patient and their unaffected sibling to determine if LINC00486 is actually directly related to BPD. Involvement of Gsk 3 in Stem Cell Pluripotency Our data suggests Gsk 3 plays an important role in stem cell differentiatio n through the regulation of the m 6 A modification on mRNA Due to the similar morphological phenotypes between Mettl3 KO and Gsk 3 DKO ESCs, we hypothesized that Gsk 3 activity regulates the mRNA m 6 A modification process through the phosphorylation of FTO, the m 6 A demethylase. Phosphorylation of FTO would cause inhibition, resulting in an increased presence of m 6 A on mRNA, resulting in the tagged pluripotency mRNAs to be degraded. This leads to the differentiation of ESCs. However when Gsk 3 is i nhibited, FT O levels are raised, resulting in reduced m 6 A levels and a maintained pluripotent state in ESCs. When WT ESCs are cultured in LIF, pluripotency is promoted, but FTO levels are also kept high. Therefore, i n order to determine if our hypothe s is was correct, we needed to optimize the cell culturing conditions in order to separate pluripotency from the mechanistic regulation by FTO. ESCs cultured in complete ESC media in the absence of LIF for 14 days resulted in the best observation of FTO protein expression. When WT, Gsk 3 DKO, and p110* ESCs were grown in these conditions, FTO expression in WT cells was greatly reduced compared to the robust expression seen in Gsk 3 DKO cells.

PAGE 95

! ,% ! These res ults support our hypothesis, suggesting that when Gsk 3 is knocked out t he FTO protein is accumulated resulting in reduced m 6 A levels and maintained pluripotency in ESCs However, when Gsk 3 levels are normal ( as seen in WT ESCs), FTO levels are low due to Gsk 3 phosphorylating FTO and tagging FTO for degradation The locatio n of this phosphorylation on FTO remains to be determined. Additionally, p110* ESCs revealed low FTO expression levels, which were almost identical to the levels in the WT ESCs indicating this mechanism might be occurring above PI3K. Further investigation is required. To confirm the sustained pluripotency in G sk 3 DKO ESCs, we compared the gene expression levels of "Master" pluripotency transcription factors in WT and Gsk 3 DKO ESCs. We observed increased expression of Nanog, Esrrb, and Oct 4 in Gsk 3 DKO ESCs, which validates the pluripotent state in these cells In addition, FTO expression in Gsk 3 DKO ESCs grown without LIF for 14 days was reduced compared to WT. This supports our hypothesis suggesting this effect occurs at the post translationally rath er than transcriptionally hence why we saw robust FTO protein expression in the w estern blot results. Additionally, treatment of WT cells with a Gsk 3 inhibitor rescued FTO protein expression, furthering the notion that Gsk 3 activity regulates FTO levels post translationally Optimal concentrations for FTO rescue were the treatment of 10mM LiCl or 15uM SB for two days. To confirm inhibition of Gsk 3 results in increased translational levels of FTO and therefore reduced m 6 A modifications on mRNA, total RN A was isolated from WT and Gsk 3 DKO ESCs cultured in complete ESC media in the absence of LIF for 14 days. mRNA was immunoprecipitated from the total RNA using oligo dT beads.

PAGE 96

! ,& ! Collected mRNA was digested to retrieve single nucleotides in order to perform Liquid Chromatography Tandem Mass Spectrometry (LC MS/MS). LC MS/MS data revealed 50% reduction in Gsk 3 DKO ESCs compared to WT, validating our hypothesis that inhibited forms of Gsk 3 would result in reduced mRNA methylation through the regulation of FTO ( data not shown; K Faulds, C Phiel; Personal Communications ). Throughout the 14 day experiment, we observed the phenotypes of the subset of ESCs that were treated with Vitamin C. There was an obvious morphological change in WT ESCs treated with Vitamin C compared to those grown in the absence of Vitamin C, indicating Vitamin C is causing a pluripotent effect on these cells. However, this pluripotent morphology did not affec t FTO expression at the transcriptional or translational level in Gsk 3 DKO ESCs, which was indicated by no change in FTO protein or gene expression between the Vitamin C treatments in Gsk 3 DKO ESCs. We originally hypothesized Vitamin C, a cofactor of FTO, would enhance pluripotency in Gsk 3 DKO ESCs by assisting the enzymatic reactio n of FTO, resulting in high FTO protein expression. However, these results suggest this antioxidant supplement does not have an effect on Gsk 3 DKO ESCs. Interestingly, FTO protein expression in p110* ESCs treated with Vitamin C was greatly increased compa red to the cells lacking Vitamin C. This robust FTO expression was almost identical to the amount of FTO expression observed in Gsk 3 DKO ESCs without LIF for 14 days. The effect of Vitamin C on p110* cells through FTO protein expression is blatantly obvio us, yet the mechanism behind why this is observed is unknown. Additionally, when we quantified pluripotent gene expression (Nanog, Oct4 and Esrrb) in p110* and WT ESCs treated with and without Vitamin C (no LIF), there was increased gene expression in Nano g, Oct 4 and

PAGE 97

! ,' ! Esrrb in the WT and p110* ESCs treated with Vitamin C. Furthermore, t here was no change in Oct 4 and Esrrb expression levels between WT and p110* ESCs with Vitamin C. However, there was a significant increase in Nanog expression in p110* ESCs with Vitamin C compared to WT ESCs with Vitamin C. These results indicate Vitamin C alone can enhance expression of pluripotency factors, yet this enhancement is not seen in Gsk 3 DKO ESCs. Final Words The large involvement of Gsk 3 regulatory activity in numerous signaling pathways indicates how important this kinase is in modulating gene expression, DNA methylation and m 6 A mRNA modifications When Gsk 3 expression is altered, it has been shown to play a role in numerous human diseases such as diabetes an d bipolar disorder. Therefore, understanding how this kinase works in these systems is extremely important for understanding disease and creating pharmaceutical therapeutics. In this thesis, w e have uncovered another layer of complexity for the regulatory role of Gsk 3 in mRNA methylation to sustain stem cell pluripotency However, the mechanistic understanding behind these modifications is still relatively unknown and will remain our main focus as we move forward.

PAGE 98

! ,( ! CHAPTER VI FUTURE DIRECTIONS Charact erizing a long non coding RNA, LINC00486 We propose to perform gain of function experiments by over expressing LINC00486 transcripts 1 3 in human and mouse cells. A microarray assay will follow over expression to observe the effect on downstream genes. Th is will help to determine which genes are associated with LINC00486 as well as what cellular pathway LINC00486 is involved in. Additionally, we propose to perform loss of function experiments to determine what the downstream effect is when the entire LINC0 0486 gene is knocked down, as well as each individual transcript. siRNAs or CRISPR will b e used to perform the knockdown/knockouts After each manipulation a microarray assay will be performed to observe effect on downstream genes. In addition to these gain and loss of function exper iments, we will determine if there is a mouse homolog of LINC00486. We will accomplish this by designing mouse primers to determine if we can detect endogenous LINC00486 in mouse cells through RT PCR. We also propose to dete rmine exactly where on each Cbx protein LINC00486 Transcripts 1 3 binds by using Cbx proteins containing different mutations through out its domain. By determining exactly where each transcript binds onto the Cbx proteins, it will hopefully give us insight into the function of that transcript and LINC00486 in general. Role of Gsk 3 in Stem Cell Pluripotency So far, we have characterized the relationship between m 6 A, FTO and pluripotency. To further investigate our findings of Gsk 3 regulating stem cell pl uripotency, we propose to determine the Gsk 3 phosphorylation site on FTO. This will

PAGE 99

! ,) ! be accomplished by using mass spectrometry on WT and Gsk 3 DKO ESCs to compare phosphorylation sites and identify those that are Gsk 3 dependent. In addition, we will pe rform mRNA half life studies by treating WT and Gsk 3 DKO ESCs with amanatin, followed by qPCR. This experiment needs to be repeated using the 14 day no LIF paradigm that I have developed. This will allow us to compare the rates of mRNA degradation betwe en WT and Gsk 3 DKO ESCs and provide insight into the overall mechanism of Gsk 3 regulated m 6 A modifications. Eventually, we will perform m 6 A seq to identify the numerous other mRNAs whose m 6 A modifications are affected by deletion or inhibition of Gsk 3 Finally, we will look at m 6 A levels in different human cells to determine if the Gsk 3 regulation of m 6 A modifications is relevant in disease.

PAGE 100

! ,* ! REFERENCES Banito, A., & Gil, J. (2010). Induced pluripotent stem cells and sen escence: learning the biology to improve the technology. EMBO Rep, 11 (5), 353 359. doi: 10.1038/embor.2010.47 Bar Nur, O., Brumbaugh, J., Verheul, C., Apostolou, E., Pruteanu Malinici, I., Walsh, R. M., . Hochedlinger, K. (2014). Small molecules facil itate rapid and synchronous iPSC generation. Nat Methods, 11 (11), 1170 1176. doi: 10.1038/nmeth.3142 Batista, P. J., Molinie, B., Wang, J., Qu, K., Zhang, J., Li, L., . Chang, H. Y. (2014). m(6)A RNA modification controls cell fate transition in mamma lian embryonic stem cells. Cell Stem Cell, 15 (6), 707 719. doi: 10.1016/j.stem.2014.09.019 Brueckner, F., & Cramer, P. (2008). Structural basis of transcription inhibition by [alpha] amanitin and implications for RNA polymerase II translocation. Nat Struc t Mol Biol, 15 (8), 811 818. Burdon, T., Smith, A., Savatier, P. (2002). Signalling, cell cycle and pluripotency in embroynic stem cells. TRENDS in Cell Biology, 12 (9). Caricasole, A., Copani, A., Caruso, A., Caraci, F., Iacovelli, L., Sortino, M. A., . Nicoletti, F. (2003). The Wnt pathway, cell cycle activation and amyloid: novel therapeutic strategies in Alzheimer's disease? Trends Pharmacol Sci, 24 (5), 233 238. doi: 10.1016/s0165 6147(03)00100 7 Chen, J., Liu, J., Chen, Y., Yang, J., Chen, J., Liu, H., . Pei, D. (2011). Rational optimization of reprogramming culture conditions for the generation of induced pluripotent stem cells with ultra high efficiency and fast kinetics. Cell Res, 21 (6), 884 894. doi: 10.1038/cr.2011.51 Chu, C., Qu, K., Zhong, F. L., Artandi, S. E., & Chang, H. Y. (2011). Genomic maps of long noncoding RNA occupancy reveal principles of RNA chromatin interactions. Mol Cell, 44 (4), 667 678. doi: 10.1016/j.molcel.2011.08.027 Cifuentes Rojas, C., Hernandez, A. J., Sarma, K ., & Lee, J. T. (2014). Regulatory interactions between RNA and polycomb repressive complex 2. Mol Cell, 55 (2), 171 185. doi: 10.1016/j.molcel.2014.05.009 Coghalin, M. P., Culbert, A., Cross, D., Corcoran, S., Yates, J., Pearce, N., Raush, O., Murphy, G., Carter, P., Cox, L., Mills, D., Brown, M., Haigh, D., Ward, R., Smith, D., Murray, K., Reith, A., Holder, J. (2000). Selective small molecule inhibitors of glycogen synthatse kinase 3 modulate glycogen metabolism and gene trranscription. Chemistry and Bio logy, 7 (10).

PAGE 101

! ,+ ! Davidovich, C., Zheng, L., Goodrich, K. J., & Cech, T. R. (2013). Promiscuous RNA binding by Polycomb repressive complex 2. Nat Struct Mol Biol, 20 (11), 1250 1257. doi: 10.1038/nsmb.2679 De, D., Jeong, M. H., Leem, Y. E., Svergun, D. I., We mmer, D. E., Kang, J. S., . Kim, S. H. (2014). Inhibition of master transcription factors in pluripotent cells induces early stage differentiation. Proc Natl Acad Sci U S A, 111 (5), 1778 1783. doi: 10.1073/pnas.1323386111 Doble, B. W., Patel, S., Wood G. A., Kockeritz, L. K., & Woodgett, J. R. (2007). Functional redundancy of GSK 3alpha and GSK 3beta in Wnt/beta catenin signaling shown by using an allelic series of embryonic stem cell lines. Dev Cell, 12 (6), 957 971. doi: 10.1016/j.devcel.2007.04.001 Dominissini, D., Moshitch Moshkovitz, S., Schwartz, S., Salmon Divon, M., Ungar, L., Osenberg, S., . Rechavi, G. (2012). Topology of the human and mouse m6A RNA methylomes revealed by m6A seq. Nature, 485 (7397), 201 206. doi: 10.1038/nature11112 Este ban, M. A., Wang, T., Qin, B., Yang, J., Qin, D., Cai, J., . Pei, D. (2010). Vitamin C enhances the generation of mouse and human induced pluripotent stem cells. Cell Stem Cell, 6 (1), 71 79. doi: 10.1016/j.stem.2009.12.001 Fang, X., Yu, S. X., Lu, Y., Bast, R. C., Jr., Woodgett, J. R., & Mills, G. B. (2000). Phosphorylation and inactivation of glycogen synthase kinase 3 by protein kinase A. Proc Natl Acad Sci U S A, 97 (22), 11960 11965. doi: 10.1073/pnas.220413597 Fiol, C., Mahrenholz, AM., Wang, Y., Roeske, RW., Roach, PJ. (1987). Formation of protein kinase recognition sites by covalent modification of the substrate. Molecular mechanism for the synergistic action of casein kinase II and glycogen synthase kinase 3. J Biol Chem, 262 (29), 14042 14048. Frame, S., Cohen, P. (2001). Gsk3 takes centre stage more than 20 years after its discovery. BIOchem J, 359 1 16. Gerken, T., Girard, C., Tung, Y., Webby, C., Saudek, V., Hewitson, K., Yeo, G., McDonough, M., Cunliffe, S., McNeill, L., Galvanovskis, J. Rorsman, P., Robins, P., Prieur, X., Coll, A., Ma, M., Jovanovic, Z., Farooqi, S., Sedgwick, B., Barroso, I., Lindahl, T., Ponting, C., Ashcroft, F., O'Rahilly, S., Schofield, C. (2007). The Obesity Associated FTO Gene Encodes a 2 Oxyglutarate Dependent Nucleic Acid Demethylase. Science, 318 Gibson, D. G., Young, L., Chuang, R. Y., Venter, J. C., Hutchison, C. A., 3rd, & Smith, H. O. (2009). Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods, 6 (5), 343 345. doi: 10.1038/nm eth.1318

PAGE 102

! ,, ! Grimes, C., Jope, R. (2001). The multifaceted roles of glycogen synthase kinase 3B in cellular signaling. Prog Neurobiol, 65 391 426. Guela, S., Moshitch Moshkovitz, S., Dominissini, D., Mansour, A., Kol, N., Salmon Divon, M., Hershkovitz, V., Peer, E., Mor, N., Manor, Y., Ben Haim, M., Eyal, E., Yunger, S., Pinto, Y., Jaitin, D., Viukov,S., Rais, Y., Krupalnik, V., Chomsky, E., Zerbib, M., Maza, I., Rechavi, Y., Massarawa, R., Hanna, S., Amit, I., Levanon, E., Amariglio, N., Stern Ginossar, N., Novershtern, N., Rechavi, G., Hanna, J. (2015). m6A mRNA methylation facilitates resolution of naive pluripotency toward differentiation. Science, 347 (6225). Guttman, M., Amit, I., Garber, M., French, C., Lin, M. F., Feldser, D., . Lander, E. S. ( 2009). Chromatin signature reveals over a thousand highly conserved large non coding RNAs in mammals. Nature, 458 (7235), 223 227. doi: 10.1038/nature07672 Guttman, M., & Rinn, J. L. (2012). Modular regulatory principles of large non coding RNAs. Nature, 4 82 (7385), 339 346. doi: 10.1038/nature10887 Hanna, J., Cheng, A., Saha, K., Kim, J., Lengner, C., Soldner, F., Cassady, J., Muffat, J., Carey, B., Jaenish, R. (2010). Human embryonic stem cells with biological and epigenetic characteristics similar to tho se of mouse ESCs. PNAs, 107 (20). Hedgepeth, C. M., Conrad, L. J., Zhang, J., Huang, H. C., Lee, V. M. Y., & Klein, P. S. (1997). Activation of the Wnt Signaling Pathway: A Molecular Mechanism for Lithium Action. Developmental Biology, 185 (1), 82 91. Ji a, G., Fu, Y., Zhao, X., Dai, Q., Zheng, G., Yang, Y., . He, C. (2011). N6 Methyladenosine in nuclear RNA is a major substrate of the obesity associated FTO. Nat Chem Biol, 7 (12), 885 887. Jope, R. S. (2003). Lithium and GSK 3: one inhibitor, two inh ibitory actions, multiple outcomes. Trends Pharmacol Sci, 24 (9), 441 443. doi: 10.1016/S0165 6147(03)00206 2 Katoh, M., & Katoh, M. (2007). WNT signaling pathway and stem cell signaling network. Clin Cancer Res, 13 (14), 4042 4045. doi: 10.1158/1078 0432.C CR 06 2316 Klein, P. S., Melton, D. A. (1996). A molecular mechanism for the effect of lithium on development. Dev Biol, 93 8455 8459. Lehmann, L., Ferrari, R., Vashisht, A. A., Wohlschlegel, J. A., Kurdistani, S. K., & Carey, M. (2012). Polycomb repre ssive complex 1 (PRC1) disassembles RNA polymerase II preinitiation complexes. J Biol Chem, 287 (43), 35784 35794. doi: 10.1074/jbc.M112.397430

PAGE 103

! ,! Linder, B., Grozhik, A. V., Olarerin George, A. O., Meydan, C., Mason, C. E., & Jaffrey, S. R. (2015). Single nuc leotide resolution mapping of m6A and m6Am throughout the transcriptome. Nat Meth, 12 (8), 767 772. doi: 10.1038/nmeth.3453 Loh, Y. H., Wu, Q., Chew, J. L., Vega, V. B., Zhang, W., Chen, X., . Ng, H. H. (2006). The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet, 38 (4), 431 440. doi: 10.1038/ng1760 Margueron, R., & Reinberg, D. (2011). The Polycomb complex PRC2 and its mark in life. Nature, 469 (7330), 343 349. doi: 10.1038/nature09784 Martello, G., Sugimoto, T., Diamanti, E., Joshi, A., Hannah, R., Ohtsuka, S., . Smith, A. (2012). Esrrb is a pivotal target of the Gsk3/Tcf3 axis regulating embryonic stem cell self renewal. Cell Stem Cell, 11 (4), 491 504. doi: 10.1016/j.stem.2012.06.008 Meredith, G. D., D'Ippolito, A., Dudas, M., Zeidner, L. C., Hostetter, L., Faulds, K., . Phiel, C. J. (2015). Glycogen synthase kinase 3 (Gsk 3) plays a fundamental role in maintaining DNA methylation at imprinted loci in mouse embryonic stem cells. Mol Biol Cel l, 26 (11), 2139 2150. doi: 10.1091/mbc.E15 01 0013 Meyer, K. D., Saletore, Y., Zumbo, P., Elemento, O., Mason, C. E., & Jaffrey, S. R. (2012). Comprehensive analysis of mRNA methylation reveals enrichment in 3' UTRs and near stop codons. Cell, 149 (7), 163 5 1646. doi: 10.1016/j.cell.2012.05.003 Morey, L., Pascual, G., Cozzuto, L., Roma, G., Wutz, A., Benitah, S. A., & Di Croce, L. (2012). Nonoverlapping functions of the Polycomb group Cbx family of proteins in embryonic stem cells. Cell Stem Cell, 10 (1), 4 7 62. doi: 10.1016/j.stem.2011.12.006 Ng, S. S., Mahmoudi, T., Danenberg, E., Bejaoui, I., de Lau, W., Korswagen, H. C., . Clevers, H. (2009). Phosphatidylinositol 3 kinase signaling does not activate the wnt cascade. J Biol Chem, 284 (51), 35308 35313 doi: 10.1074/jbc.M109.078261 Ng, S. Y., Johnson, R., & Stanton, L. W. (2012). Human long non coding RNAs promote pluripotency and neuronal differentiation by association with chromatin modifiers and transcription factors. EMBO J, 31 (3), 522 533. doi: 10 .1038/emboj.2011.459 Osaki, M., Oshimura, M., Ito, H. (2004). PI3K Akt pathway: Its functions an alterations in human cancer. Apoptosis, 9 667 676. Pei, D. (2009). Regulation of pluripotency and reprogramming by transcription factors. J Biol Chem, 284 ( 6), 3365 3369. doi: 10.1074/jbc.R800063200

PAGE 104

! -. ! Phiel, C. J., Klein, P.S. (2001). Molecular targets of lithium action. Annu Rev Pharmacol Toxicol, 41 789 813. Popkie, A. P., Zeidner, L. C., Albrecht, A. M., D'Ippolito, A., Eckardt, S., Newsom, D. E., . Phiel, C. J. (2010). Phosphatidylinositol 3 kinase (PI3K) signaling via glycogen synthase kinase 3 (Gsk 3) regulates DNA methylation of imprinted loci. J Biol Chem, 285 (53), 41337 41347. doi: 10.1074/jbc.M110.170704 Ren, X., Vincenz, C., & Kerppola, T. K. (2008). Changes in the distributions and dynamics of polycomb repressive complexes during embryonic stem cell differentiation. Mol Cell Biol, 28 (9), 2884 2895. doi: 10.1128/MCB.00949 07 Romeo, F., Costanzo, F., Agostini, M. (2012). Embryonic stem cells a nd inducible pluripotent stem cells: two faces of the same coin? AGING, 4 (12). Sanchez Ripoll, Y., Bone, H., Owen, T., Guedes, A., Abranches, E., Kumpfmueller, B., Spriggs, R., Henrique, D., Welham, M. (2013). Glycogen Synthase Kinase 3 Inhibition Enhanc es Translation of Pluripotency Associated Transcription Factors to Contribute to Maintenance of Mouse Embryonic Stem Cell Self Renewal. PLOS ONE, 8 (4). doi: 10.1371/ Storm, M. P., Bone, H. K., Beck, C. G., Bourillot, P. Y., Schreiber, V., Damiano, T., . Welham, M. J. (2007). Regulation of Nanog expression by phosphoinositide 3 kinase dependent signaling in murine embryonic stem cells. J Biol Chem, 282 (9), 6265 6273. doi: 10.1074/jbc.M610906200 Tamm, C., Pijuan Galito, S., & Anneren, C. (2013). A compa rative study of protocols for mouse embryonic stem cell culturing. PLOS ONE, 8 (12), e81156. doi: 10.1371/journal.pone.0081156 Tsai, M. C., Manor, O., Wan, Y., Mosammaparast, N., Wang, J. K., Lan, F., . Chang, H. Y. (2010). Long noncoding RNA as modula r scaffold of histone modification complexes. Science, 329 (5992), 689 693. doi: 10.1126/science.1192002 Wang, X., Lu, Z., Gomez, A., Hon, G. C., Yue, Y., Han, D., . He, C. (2014). N6 methyladenosine dependent regulation of messenger RNA stability. Nat ure, 505 (7481), 117 120. doi: 10.1038/nature12730 Wang, X., Zhao, B. S., Roundtree, I. A., Lu, Z., Han, D., Ma, H., . He, C. (2015). N(6) methyladenosine Modulates Messenger RNA Translation Efficiency. Cell, 161 (6), 1388 1399. doi: 10.1016/j.cell.2015 .05.014 Wang, Y., Li, Y., Toth, J. I., Petroski, M. D., Zhang, Z., & Zhao, J. C. (2014). N6 methyladenosine modification destabilizes developmental regulators in embryonic stem cells. Nat Cell Biol, 16 (2), 191 198. doi: 10.1038/ncb2902

PAGE 105

! -% ! Welham, M. J., Kin gham, E., Sanchez Ripoll, Y., Kumpfmueller, B., Storm, M., & Bone, H. (2011). Controlling embryonic stem cell proliferation and pluripotency: the role of PI3K and GSK 3 dependent signalling. Biochem Soc Trans, 39 (2), 674 678. doi: 10.1042/BST0390674 Wray J., Kalkan, T., Gomez Lopez, S., Eckardt, D., Cook, A., Kemler, R., & Smith, A. (2011). Inhibition of glycogen synthase kinase 3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation. Nat C ell Biol, 13 (7), 838 845. Wu, P., Zuo, X., Deng, H., Liu, X., Liu, L., & Ji, A. (2013). Roles of long noncoding RNAs in brain development, functional diversification and neurodegenerative diseases. Brain Res Bull, 97 69 80. doi: 10.1016/j.brainresbull.2 013.06.001 Yang, L., Froberg, J. E., & Lee, J. T. (2014). Long noncoding RNAs: fresh perspectives into the RNA world. Trends in Biochemical Sciences, 39 (1), 35 43. Yang, Y., Sun, B. F., Xiao, W., Yang, X., Sun, H. Y., Zhao, Y. L., & Yang, Y. G. (2015). Dynamic m6A modification and its emerging regulatory role in mRNA splicing. Science Bulletin, 60 (1), 21 32. doi: 10.1007/s11434 014 0695 6 Ying, Q. L., Wray, J., Nichols, J., Batlle Morera, L., Doble, B., Woodgett, J., . Smith, A. (2008). The ground s tate of embryonic stem cell self renewal. Nature, 453 (7194), 519 523. doi: http://www.nature.com/nature/journal/v453/n7194/suppinfo/nature06968_S1.html Zheng, G., Dahl, J. A., Niu, Y., Fedorcsak, P., Huang, C. M., Li, C. J., . He, C. (2013). ALKBH5 is a mammalian RNA demethylase that impacts RNA metabolism and mouse fertility. Mol Cell, 49 (1), 18 29. doi: 10.1016/j.molcel.2012.10.015

PAGE 106

! -& ! APPENDIX Equations Equation 1: pmoles= (mass of DNA in ng) x (1000 pg/ng)_______ (# of sense strand bp) x (650 Daltons Buffers and Reagents Buffer A consists of 10mM HEPES (pH 7.9), 10mM KCl, 0.1mM EDTA, 0.1EGTA, 1:100 protease inhibitor cocktail (added fresh). Buffer B consists of 20mM HEPES (pH 7.9), 500mM KCl, 1mM EDTA, 1mM EGTA, 1:100 protease inhibitor cocktail (added fresh). Elution Buffer: 5mM Tris HCl pH 7.5, 1mM EDTA pH 8.0, 0.05% SDS, 4.2uL Proteinase K (20mg/mL) Linear PEI was made by dissolving 1g polyethylenimine (Polysciences, Inc., CAT #23966) in 1L 1X HEPES. PE I was sterilized through a 0.2uM filter and stored at 4¡C for short term use or 20¡C for long term storage (Bartman, 2014). Cellular Mediums N2A media consists of 50% DMEM (10% BGS, 1% Penicillin Streptomycin, 1% L Glutamine), 50% Opti MEM, 0.5% Pen stre p, and 0.5% L Glutamine. HEK293T media consists of DMEM, 10% BGS, 1% Penicillin Streptomycin, and 1% L Glutamine. ESC Media consists of DMEM, 15% FBS, 1% Pen Strep, 1% L Glutamine, 1% Sodium Pyruvate, 1% Non essential Amino Acids. 2 mercaptoethanol (! me) and LIF were added fresh each time (1uL of 55mM me per 1mL of media and 1uL ESGRO LIF per 1mL media). N3B27 Me dia consists of 1 volume F 12 Media, 1 volume Neurobasal Media, 1x N2, 1x B27, 0.0125% Monothioglycerol, 50mg/mL BSA, 2mM Glutamine (Sanchez Ripoll, 2013)

PAGE 107

! -' ! Supplemental Data Supplemental Data Figure 1 Gibso n Assembly Overview (NEB) Supplemental Data Figure 2 Detection of Optimal Biotinylated Oligo Dilutions for Streptavidin Bead Immunoprecipitation of LINC00486 Transcripts 1 3. N2A cells transfected with each individ ual transcript were immunoprecipitated using streptavidin beads. Different dilutions of biotinylated oligo were used for optimal immunoprecipitation.