University of Southern California Dissertations and Theses

SUMOylation regulates RNA polymerase III -- dependent transcripton via MAF1

(USC Thesis Other)

SUMOylation regulates RNA polymerase III -- dependent transcripton via MAF1

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
SUMOYLATION REGULATES RNA POLYMERASE III - DEPENDENT
TRANSCRIPTION VIA MAF1

By

Aarti D. Rohira

A Dissertation presented to
THE FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the degree
DOCTOR OF PHILOSOPHY
(GENETIC, CELLULAR AND MOLECULAR BIOLOGY)

May 2013

Copyright 2013 Aarti D. Rohira
ii

Dedication
To My Family
iii

Acknowledgements
I am thankful to Dr. Debbie Johnson for providing me with the opportunity to pursue my
graduate studies in her lab. She took me on as a Master’s student and worked hard to
transfer me into the PhD program. I am grateful for her constant encouragement, support
and mentorship throughout my graduate studies. I would also like to thank Dr. Zoltan
Tokes who admitted me into the Master’s program in Biochemistry and who let me
transfer into the PhD program.

I would like to thank Dr. Sandra Johnson who mentored me when I first joined the lab
and taught me all that I know about performing experiments. I would like to thank Abe
Kaslow and Justin Allen for helping me with running the experiments. I would like to
acknowledge Dr. Beth Palian who was a great support to me during my graduate studies.
She was always available to help me with experiments and to have useful discussions
with about my project. I would like to thank Shuping Zhong, Jody Fromm, Elliot Chen,
Eric Wang, Alexa Billow and Alex Guo for being wonderful and supportive labmates. I
would also like to acknowledge my past and present committee members, Drs. Michael
Stallcup, Woojin An, Zea Borok and Young Hong, for their guidance and support.

Lastly, I would like to acknowledge my family who encouraged me and supported me to
come so far from home and pursue my PhD. I would like to thank my friends who helped
me maintain my good spirits throughout my graduate studies.

iv

Table of Contents

Dedication .......................................................................................................................... ii
Acknowledgements .......................................................................................................... iii
List of Figures ................................................................................................................... vi
Abstract ........................................................................................................................... viii
Chapter 1: Introduction ................................................................................................... 1
1.1 Maf1 and transcription .............................................................................................. 3
1.2 Small Ubiquitin like Modifier (SUMO) .................................................................... 7
1.3 Consequences of covalent and non-covalent SUMO interactions .......................... 10
1.4 Sumoylation and disease ......................................................................................... 13
Chapter 2: Covalent SUMO modification regulates Maf1 function .......................... 18
2.1 SUMO negatively regulates RNA polymerase III mediated transcription ............. 18
2.2 Maf1 is covalently modified by SUMO .................................................................. 19
2.3 SUMO enhances Maf1 protein expression .............................................................. 20
2.4 Maf1 is covalently modified by SUMO at lysine 35 .............................................. 21
2.5 Maf1 phosphorylation does not affect covalent SUMO modification .................... 24
2.6 Sumoylation is required for Maf1 function ............................................................. 25
2.7 SENP1 desumoylase regulates Maf1 sumoylation .................................................. 26
2.8 SUMOylation at lysine 35 does not alter Maf1 subcellular localization or protein
stability .......................................................................................................................... 27
2.9 SUMOylation affects Maf1 interaction with RNA pol III ..................................... 29
2.10 SUMOylation affects Maf1 recruitment to RNA polymerase III- dependent gene
promoters ....................................................................................................................... 30
2.11 Summary ............................................................................................................... 31
Chapter 3: Non-Covalent SUMO interactions regulate Maf1 function ..................... 32
3.1 Maf1 non-covalently interacts with SUMO1 and SUMO2 ..................................... 32
3.2 Serine 123 on Maf1 is important for non-covalent interaction with SUMO1 and
SUMO2 ......................................................................................................................... 33
3.3 Mutation of Maf1 serine 123 to alanine enhances Maf1 function .......................... 36
v

3.4 Maf1 S123A interaction with RNA polymerase III is enhanced ............................ 38
3.5 Summary ................................................................................................................. 40
Chapter 4: Identification of a novel protein-protein interaction between Maf1 and
Mediator ........................................................................................................................... 41
Results ............................................................................................................................... 44
4.1 Maf1 interacts with the CDK8 subcomplex ............................................................ 44
4.2 CDK8 regulates Maf1 target genes ......................................................................... 44
4.3 CDK8 is recruited to gene promoters regulated by Maf1 ....................................... 45
4.4 Summary ................................................................................................................. 47
Chapter 5: Discussion ..................................................................................................... 48
5.1 Sumoylation regulates RNA polymerase III mediated transcription ...................... 48
5.2 Maf1 is covalently modified by SUMO .................................................................. 49
5.3 Covalent SUMO modification is required for Maf1 function ................................. 51
5.4 SUMO modification and Maf1 protein expression ................................................. 54
5.5 Maf1 and non-covalent SUMO interaction ............................................................. 55
5.6 Non-covalent SUMO interaction and Maf1 function .............................................. 57
5.7 Maf1 interacting proteins ........................................................................................ 59
5.8 Summary ................................................................................................................. 60
Chapter 6: Materials & Methods .................................................................................. 63
References ........................................................................................................................ 72

vi

List of Figures
Figure 1 SUMO negatively regulates RNA pol III transcribed genes 18
Figure 2 Endogenous Maf1 is sumoylated 19
Figure 3 Maf1 is covalently modified by SUMO1 and SUMO2 20
Figure 4 Ubc9 silencing inhibits Maf1 protein expression 21
Figure 5 Maf1 protein sequence 22
Figure 6 Maf1 lysine 35 is the major site of sumoylation 22
Figure 7 Maf1 is sumoylated at lysine 35 by both SUMO1 and SUMO2 23
Figure 8 Maf1 sumoylation does not affect Maf1 phosphorylation 24
Figure 9 Covalent SUMO modification is required for Maf1 function 25
Figure 10 SENP1 desumoylase brings about Maf1 desumoylation and positively
regulates RNA pol III –dependent transcription
26
Figure 11 Covalent SUMO modification does not affect Maf1 localization 28
Figure 12 Sumoylation at Maf1K35 does not alter protein stability 29
Figure 13 SUMO modification is required for Maf1 interaction with RNA pol III 30
Figure 14 Covalent SUMO modification is required for Maf1 recruitment to
tRNA
Leu
promoter
31
Figure 15 Maf1 interacts non-covalently with SUMO1 and SUMO2 32
Figure 16 Covalent SUMO modification is not required for non-covalent Maf1-
SUMO interaction
33
Figure 17
Figure 18
Maf1 protein sequence
Maf1S123A displays enhanced interaction with SUMO
34
35
Figure 19 Non-covalent SUMO interaction enhances Maf1 function 36
Figure 20 Maf1S123A subcellular localization is unchanged compared to
Maf1WT
37
vii

Figure 21 Maf1S123A displays enhanced interaction with RNA pol III 39
Figure 22 Maf1 interacts with CDK8 and MED13 44
Figure 23 Decreased CDK8 expression reduces TBP and tRNA gene expression 45
Figure 24 CDK8 and MED13 are present at TBP and tRNA
Leu
promoters 46
Figure 25 Model for covalent SUMOylation of Maf1 regulating Maf1 function 53
Figure 26 Maf1K35R no longer functions to regulate RNA pol III –dependent
transcription
54

viii

Abstract
Our studies have identified a role for sumoylation in mediating the ability of Maf1 to
repress RNA pol III-dependent transcription. As increased expression of either SUMO1
or SUMO2 represses the expression of tRNA and U6 RNA genes, we further examined
whether the transcriptional repressor, Maf1, might be regulated by sumoylation. We find
that Maf1 is covalently modified by both SUMO1 and SUMO2. To identify the specific
sumoylation site, each of the eleven lysine residues within Maf1 was mutated to arginine
revealing that Maf1 is covalently modified by SUMO at K35. Compared to wild type
Maf1, Maf1K35R is impaired in its ability to repress tRNA gene transcription and to
suppress colony growth. Covalent SUMO modification of Maf1 at K35 is required for
Maf1 association with RNA pol III and recruitment to RNA pol III –dependent gene
promoters. Additionally, Maf1 can also non-covalently interact with SUMO1 and
SUMO2. Covalent modification of Maf1 at K35 is not required for this interaction.
Further mutational analysis identified S123 as an important residue that controls non-
covalent interaction of Maf1 with SUMO, potentially, by its phosphorylation state.
Compared with wild type Maf1, Maf1S123A displays an enhanced ability to both, repress
transcription and suppress colony growth, whereas Maf1S123D exhibits a decrease in
these Maf1 functions. Furthermore, Maf1S123A shows enhanced interaction with RNA
pol III. To understand how Maf1 maybe recruited to promoters we have found that Maf1
interacts with CDK8 and MED13, members of the Mediator-CDK8 subcomplex, which
are also bound at the TBP promoter. While Maf1 represses TBP expression, CDK8
induces TBP gene expression. This supports the idea that association between Maf1 and
ix

the CDK8 subcomplex regulates gene activity. Together, our results demonstrate that
covalent and non-covalent interactions of SUMO1 and SUMO2 with Maf1 positively
regulate its capacity to repress transcription and cell growth. Additionally, Maf1
associates with the CDK8 subcomplex to repress CDK8 mediated gene activation
function.
1

Chapter 1: Introduction
Transcription is the process of RNA synthesis carried out by large enzymes called RNA
polymerases (pols). RNA polymerase I transcribes the largest rRNA genes, except 5S
rRNA while RNA polymerase II transcribes mRNA and certain U RNAs. RNA
polymerase (pol) III transcribes untranslated small RNAs such as tRNA and 5S rRNA,
which are required for protein synthesis, 7SL RNA which is involved in intracellular
protein transport, and the U6, H1 and MRP RNAs, which are involved in the processing
of RNA transcripts (Paule and White, 2000). Together, RNA pol I and RNA pol III
products account for 80% of total cellular RNA synthesized in eukaryotes (Paule and
White, 2000). These processes are tightly regulated since the growth of cells is directly
proportional to the rate of protein accumulation which requires tRNAs and rRNAs.

RNA pol I and RNA pol III transcription is deregulated in transformed and tumor cells
and accumulation of pol III transcripts is observed in cancer (Schwartz et al., 1974).
tRNA and 5S rRNA are overproduced in ovarian tumors compared to matched normal
tissue (Winter et al., 2000). 7SL RNA has been shown to accumulate in 80 tumor samples
compared with matched normal tissue derived from 19 different cancers (Chen et al.,
1997). Increased levels of pol III transcripts have also been observed in neoplastic cells
relative to the surrounding healthy tissues in breast, lung and tongue carcinomas by in
situ hybridization (Chen et al., 1997). These results indicate that enhanced RNA pol III
transcription is important for tumorigenesis. Furthermore, studies in our lab and others
have confirmed that increased RNA pol III activity is required for oncogenic
2

transformation (Johnson et al., 2008a). Indeed, increased RNA pol I and III transcription
is a hallmark of transformation and tumorigenesis.

While RNA polymerase III transcribes 5S rRNA, tRNA, U6 RNA, and other small
untranslated RNAs (White, 2004) different DNA sequence elements and transcription
factors are required to regulate their expression. Transcription from the promoters of
these genes requires recruitment of RNA polymerase III. This recruitment is brought
about by the assembly of multi-subunit protein complexes that together form the
transcription initiation complex. RNA pol III transcription from all RNA pol III –
dependent genes requires the TFIIIB complex. This complex is made up of three
subunits, which are, the TATA box binding protein (TBP), the SANT domain protein
BDP1 and either of the TFIIB-related factors, Brf1 or Brf2. While Brf1 is used by RNA
pol III dependent genes with promoter elements located within the transcribed region
such as tRNA genes, Brf2 is used by RNA pol III dependent genes that have promoters
located upstream of the initiation site such as U6 and snRNA. Additionally, while the
tRNA and 5s rRNA genes promoters require the TFIIIC transcription complex the U6
gene promoter uses the SNAPc complex for transcription initiation. All three proteins of
the TFIIIB complex are regulated by a host of oncogenes and tumor suppressors to
control RNA pol III transcription.

Many cell signaling pathways that bring about cellular transformation are known to
modify components of the RNA pol III transcription factor components that affect their
3

interactions and transcriptional activity. The extracellular signal-regulated kinase (ERK)
phosphorylates Brf1, thereby, enhancing the interaction between TFIIIB-TFIIIC (Felton-
Edkins et al., 2003). The epidermal growth factor (EGF) receptor 1 activates the Ras-
mitogen activated protein kinase (MAPK) pathway, which induces the expression of
TBP, thus increasing RNA pol III –dependent transcription (Zhong et al., 2004). The
phosphatase PTEN on the other hand controls the phosphorylation of Brf1 and Bdp1 and
hence the association between TBP and Brf1 to negatively regulate RNA pol III –
dependent transcription (Woiwode et al., 2008). PTEN does this by inhibiting the
phosphatidylinositol 3-kinase (PI 3-kinase) – AKT- mammalian target of rapamycin
(mTOR) – S6 kinase signaling pathway. The oncogenic protein c-Myc associates with
Brf1 and activates transcription by RNA pol III (Gomez-Roman et al., 2003). Conversely,
tumor suppressors such as the retinoblastoma protein (Rb) and p53 bind to Brf1 and TBP
respectively (Crighton et al., 2003; Larminie et al., 1997). This disrupts the interaction of
the TFIIIB complex with TFIIIC and RNA pol III, thereby repressing RNA pol III –
dependent transcription. Maf1 (Johnson et al., 2007) and Jnk2 (Zhong et al., 2004) are
two other proteins that repress expression of TBP which represses RNA pol III –
dependent transcription.

1.1 Maf1 and transcription
Maf1 was discovered as a key player in the repression of RNA pol III transcription in
Saccharomyces cerevisiae (Boguta et al., 1997). Yeast strains that have lost Maf1 are
unable to repress pol III transcription under conditions of nutrient limitations, DNA
4

damage and environmental stresses. Since these conditions activate three distinct
signaling pathways namely, secretory, target of rapamycin (TOR) and DNA damage
pathways, it indicates that these pathways must converge upon Maf1 to affect RNA pol
III transcription repression (Upadhya et al., 2002; Willis et al., 2004). Yeast Maf1 is a
phosphoprotein which is phosphorylated by protein kinase A (PKA) (Moir et al., 2006)
and in this phosphorylated form it is sequestered in the cytoplasm. On treatment with
rapamycin it is rapidly dephosphorylated by protein phosphatase 2A (PP2A) and shuttled
into the nucleus where it represses RNA pol III -dependent transcription (Oficjalska-
Pham et al., 2006).

Yeast Maf1 does not have any apparent DNA binding domain and is thought to repress
RNA pol III mediated transcription by two possible mechanisms (Desai et al., 2005). In
the first mechanism, Maf1 interferes with the de novo assembly of the transcription
initiation factor, TFIIIB, onto the DNA. In the second mechanism, Maf1 prevents the
recruitment of RNA pol III to the TFIIIB-DNA complexes. Consistent with this, yeast
Maf1 associates directly with RNA pol III.

The Maf1 protein is conserved from yeast to humans (Pluta et al., 2001). The eukaryotic
proteins share three conserved domains (A, B and C). Within these regions signature
sequences for this protein family can be identified (PDYDFS and WSfnYFFYNkklKR).
The functions of these motifs have not yet been identified. Our lab and others have
5

identified mammalian Maf1 as a negative regulator of RNA pol III transcription
(Goodfellow et al., 2008; Reina et al., 2006; Rollins et al., 2007).

Reina et al. (Reina et al., 2006) have shown that human Maf1 can repress RNA pol III
transcription from all three types of promoters in vitro and in vivo. Mammalian Maf1 can
interact weakly with Brf1, and the largest RNA pol III subunit (RPC1), and can associate
strongly with another RNA pol III subunit, RPAC2. Similar to yeast Maf1, human Maf1
is a phosphoprotein and treatment of cells with rapamycin or MMS leads to the loss of a
slower migrating band when compared with untreated cells. It is this dephosphorylated
form of Maf1 that interacts with RPC1.

Studies in our lab have demonstrated that in addition to repressing RNA pol III
transcribed genes, Maf1 can also repress transcription by RNA pols I and II (Johnson et
al., 2007). Maf1 represses RNA pol I transcribed genes through its ability to modulate
expression of the central transcription factor, TATA-box Binding Protein (TBP). Maf1
represses RNA pol II transcribed TBP gene expression by displacing Elk-1 from the TBP
promoter. Maf1 can also selectively regulate other Elk-1 regulated genes, such as egr-1
but not cfos. Hence, Maf1 can negatively regulate a subset of RNA pol II transcribed
genes. In order to repress RNA pol III transcribed genes, occupancy of Maf1 at tRNA
promoters is inversely correlated with occupancy of the TFIIIB subunits, Brf1 and Bdp1,
as well as the RNA pol III specific subunit, Rpc155. We have also shown that changes in
Maf1 protein expression brings about phenotypic changes in U87 glioblastomas which
6

include changes in the actin cytoskeleton and the ability of these cells to grow in an
anchorage-independent manner. These findings suggest that Maf1 might be able to
regulate the transformation state of cells thereby functioning as a potential tumor
suppressor.

To better understand the mechanism by which mammalian Maf1 represses RNA pol III
mediated transcription the role of phosphorylation on Maf1 function was examined.
Studies have shown that mammalian Maf1 is phosphorylated by mTOR which negatively
regulates Maf1 function. mTORC1 phosphorylates Maf1 at serines 60, 68 and 75.
Mutation of serines 60, 68 and 75 to alanines (3A) enhances Maf1’s ability to repress
RNA pol III mediated transcription (Michels et al., 2010). Furthermore,
immunoflourescence staining of cells treated with WYE-132, which represses RNA pol
III transcribed genes and dephosphorylates Maf1, displayed a significant increase in the
amount of nuclear Maf1 (Shor et al., 2010). In another study, Kantidakis et al.
(Kantidakis et al., 2010) proposed that mTORC1 associates with TFIIIC on RNA pol III
transcribed gene promoters and relieves Maf1 mediated repression.

Recent studies in our lab have identified PTEN as a positive regulator of Maf1 expression
(Palian et al., unpublished). PTEN positively regulates Maf1 expression by blocking the
PI3 kinase/Akt pathway. FoxO1 is a downstream effector that positively regulates Maf1
expression. Importantly, we identified FASN and ACC1, enzymes that regulate
lipogenesis, as new target genes that are repressed by Maf1. These lipogenenic enzymes
7

are known to be upregulated in cancer. Furthermore, decreased Maf1 expression in Huh7
cells increases the formation of lipid droplets. Since Maf1 is a negative regulator of
lipogenesis and RNA polymerase III mediated transcription, reduced expression of Maf1
could contribute to a transformed phenotype, supporting the idea that Maf1 might be
functioning as a novel tumor suppressor. Consistent with this idea, we have found a
marked decrease in Maf1 nuclear staining in human liver and prostate cancer tissue
samples compared to matched normal tissue. It is therefore important to determine how
Maf1 is regulated to understand how it is deregulated in cancer. So far we know that
Maf1 function is regulated by phosphorylation and since Maf1 has no known DNA
binding domain it is most likely recruited to promoters through protein-protein
interactions. While trying to understand how Maf1 expression and function might be
regulated, we found that a genome wide yeast two hybrid analysis had annotated an
interaction between Maf1 and SUMO2 (Stelzl et al., 2005).

1.2 Small Ubiquitin like Modifier (SUMO)
Initially discovered in the yeast Saccharomyces cerevesiae, SUMO (Small Ubiquitin like
Modifier) is a family of proteins that uses an ubiquitin like modification system to
covalently attach to the lysine residue of target proteins. In humans, four SUMO isoforms
have been identified. They are termed SUMO1, SUMO2, SUMO3 (also known as
sentrin, sentrin3 and sentrin2 respectively) and SUMO4. While SUMO-1, SUMO-2 and
SUMO-3 are ubiquitously expressed SUMO-4 is mainly expressed in the kidney, lymph
node and spleen. SUMO-1 shares 18% homology with ubiquitin and about 50%
8

homology with SUMO-2/3, which are almost identical (Muller, Ledl and Schmidt 2004).
The covalent addition of SUMO to target proteins is termed sumoylation and involves the
formation of an isopeptide bond between the C-terminal glycine residue of SUMO and a
Є-amino group of a lysine residue in the target protein. This lysine residue is most often
embedded in a consensus sequence composed of a characteristic ψKxE/D motif, where ψ
is a large hydrophobic residue and x is any amino acid. In some cases, SUMO
conjugation can occur at lysine residues that are not embedded within this consensus
motif such as the Daxx protein (Lin et al., 2006) and poly (A) polymerase (Vethantham et
al., 2008). It should also be noted that not all proteins that have the consensus ψKxE/D
sequence are modified by SUMO.

Sumoylation is a multi-step reversible pathway that is analogous to ubiquitination but
uses a SUMO specific enzymatic machinery. SUMO is produced as an immature
precursor protein which carries a C-terminal stretch of amino acids of variable length that
must first be removed by SUMO-specific proteases (SENPs/SUSPs) (Geiss-Friedlander
and Melchior, 2007). This processing is a prerequisite for SUMO conjugation. The first
step in the sumoylation pathway is the ATP-dependent activation of the mature SUMO
protein at its C terminus by the SUMO-specific E1 activating enzyme heterodimer AOS1-
UBA2 (SAE1/SAE2). In the next step, SUMO is transferred from UBA2 to the only E2
conjugating enzyme, Ubc9. In the final step, Ubc9 transfers SUMO to a target protein by
forming an isopeptide bond between the C-terminal glycine residue of SUMO and a
lysine side chain of the target. Ubc9 itself is able to recognize and bind the consensus
9

motif in the target proteins. However, this process is very inefficient. The efficiency of
this process is increased by SUMO E3 ligases which catalyze the transfer of SUMO from
Ubc9 to the target protein. In mice and humans, the PIAS family of proteins contain the
SP-RING domain and function as SUMO E3 ligases in addition to having SUMO
independent functions (Johnson and Gupta, 2001; Takahashi et al., 2001a; Takahashi et
al., 2001b). More recently the RanBP2 and polycomb group protein 2 (Pc2) have also
been identified as SUMO E3 ligases but they are unrelated to the ubiquitin E3 ligases
(Kagey et al., 2003; Pichler et al., 2002).

Sumoylation is a highly dynamic process and the SUMO-conjugated proteins are rapidly
deconjugated by the SUMO isopeptidases (SENPs). There are 6 mammalian SENPs
(Sentrin-specific Protease) that can deconjugate sumoylated proteins. While SENP1 and
SENP2 can desumoylate proteins modified by any SUMO isoform, the remaining 4
SENPs deconjugate proteins modified by SUMO2 and SUMO3 (Bawa-Khalfe and Yeh,
2010). The 6 mammalian SENPs are divided into 3 subfamilies based on their subcellular
distribution. For Family 1, SENP1 and SENP2 can shuttle between the nucleus and
cytoplasm, for Family 2, SENP3 and SENP5 are retained in the nucleolus and for Family
3, SENP6 and SENP7 are distributed in the nucleoplasm.

In addition to covalent sumoylation, recent studies have shown that SUMO isoforms can
also mediate distinct protein-protein interactions in vivo by promoting non-covalent
binding to other proteins that contain specific motifs which recognize SUMO paralogues
10

(Hecker et al., 2006). Minty et al defined h-h-X-[S]-[X]-[S/T]-a-a-a, where h is any
hydrophobic amino acid and a is any acidic amino acid, as a SUMO-interacting motif
(SIM) (Minty et al., 2000). Recently two other SIMs have been identified which are
[V/I]-X-[V/I]-[V/I] and K-X
3-5
-[I/V]-[I/L]-[I/L]-X
3
-[D/E/Q/N]-[D/E]-[D/E] (Hannich et
al., 2005; Song et al., 2004). Identification of these SUMO-interacting motifs has opened
up new possibilities in how cellular proteins might interact with each other and regulate
various cellular activities.

1.3 Consequences of covalent and non-covalent SUMO interactions
Sumoylation of proteins is a highly dynamic process and its outcomes are very diverse.
These include changes in localization, altered activity, signal transduction pathways and,
in some cases, stability of the modified protein (Park et al., 2007). It is important to note
that only a few proteins are quantitatively sumoylated while a majority of proteins appear
to be modified to a small percentage in the steady state (Geiss-Friedlander and Melchoir
2007). Non-covalent SUMO interactions play a role in the assembly of large multi-
protein complexes that also affect a variety of cellular processes such as transcription
regulation, genomic integrity and cellular signaling. Importantly, both covalent and non-
covalent SUMO – protein interactions work together to maintain proper cellular function.

Sumoylation and protein stability
It is has been shown that lysine residues can also act as acceptors for ubiquitin, ubiquitin
like proteins, acetyl and methyl groups (Hay 2005). It is therefore possible that SUMO-
11

modification can block other lysine dependent modifications such as ubiquitination.
Sumoylation and ubiquitination can compete for the same lysine residue in a target
protein. If the protein gets ubiquitinated it will be targeted for proteosome-dependent
degradation. However, if the protein is sumoylated, thereby blocking ubiquitination, then
the stability of the protein is increased.

Sumoylation and Transcription
SUMO modification is capable of influencing the assembly of transcription factors,
recruitment of chromatin remodeling enzymes, and can itself act as a component of the
histone code. In most cases reported to date, SUMO modification is associated with
transcriptional repression. There are three probable models for SUMO mediated
repression of transcription. In the first model, it is thought that sumoylation of
transcription factors can lead to the recruitment of repressive factors with chromatin
remodeling activity. In the second model, sumoylation of the transcription factor initiates
recruitment of a repressive complex. This repressive complex includes proteins that
contain a SUMO interacting motif (SIM) that are recruited to sumoylated transcription
factors and serve as a platform for the assembly of a repressor complex. In the third
model, the repressive effect of SUMO on transcription is a result of the recruitment of
SUMO-modified proteins into the repressive environment of particular subnuclear
domains, the best known of which is the PML nuclear body (Hay 2005). Hence
sumoylation brings about transcriptional repression by a number of different
mechanisms.
12

Sumoylation and Subcellular Transport
The SUMO E1 and E2 enzymes are predominantly nuclear but new evidence has shown
that they are also associated with filaments of the nuclear pore complex projecting into
both the nucleus and the cytoplasm (Rodriguez et al., 2001; Zhang et al., 2002). Hence
cytoplasmic proteins are also capable of being modified by SUMO. It is known that
unmodified RanGAP1 is cytoplasmic and on sumoylation at lysine 526 it gets directed
into the nucleus (Joseph et al., 2002). Alternatively, the transcriptional corepressor CtBP
is normally nuclear and a mutation of its single SUMO modification site results in its
cytoplasmic localization and failure to repress transcription (Lin et al., 2003).

SIM and PML nuclear body assembly and function
Promyelocytic nuclear bodies (PML) are dynamic nuclear structures without membranes
that are regulated under conditions of stress (Borden, 2002). These PML bodies are
transcriptionally repressive domains that contain proteins such as Daxx, PML, and Ubc9
among others. The PML proteins act as scaffolds for these structures and have important
tumor suppressive functions. A number of proteins that are found in these PML bodies
contain SIMs and interact non-covalently with SUMO. The PML proteins are themselves
heavily sumoylated and the combination of SUMO and SIMs allow the PML proteins to
nucleate and establish the PML bodies (Shen et al., 2006). Furthermore, sumoylated PML
also interacts non-covalently with other proteins through their SIMs to recruit them to the
PML bodies.

13

In addition to the above mentioned consequences, covalent and non-covalent SUMO
interactions are also known to play a role in the functioning of various signaling
pathways such as steroid hormone signaling, Wnt signaling and cytokine signaling (Hay
2005, Geiss-Friedlander and Melchoir 2007). In addition, SUMO plays important roles in
regulating genome integrity, base excision repair, DNA replication and homologous
recombination.

1.4 Sumoylation and disease
The balance between SUMO conjugation and deconjugation is essential for the normal
function of proteins in cells. An imbalance in the SUMO machinery has been shown to
disrupt normal cellular physiology. Loss of the SUMO conjugating enzyme, Ubc9 in
mice is embryonic lethal just after development into blastocysts (Nacerddine et al., 2005).
The Ubc9-deficient cells display chromosomal defects and abnormal nuclear
morphology. Knockout of SENP1 or SENP2 in mice is also embryonic lethal suggesting
that excessive sumoylation of cellular proteins is also harmful (Cheng et al., 2007; Chiu
et al., 2008; Kang et al., 2010). Indeed, several studies have suggested that the SUMO
pathway plays a role in a number of human diseases such as cancer, neurodegenration
and heart disease.
SUMO and cancer
Several studies have identified a role for the SUMO modification pathway in
tumorigenesis. Gene-expression studies have found changes in the levels of various
14

SUMO conjugating and deconjugating enzymes in various cancers, suggesting an
imbalance in the SUMO system (Jacques et al., 2005; Lee and Thorgeirsson, 2004;
McDoniels-Silvers et al., 2002). While SENP1 levels are elevated in thyroid oncocytic
adenocarcinoma (Jacques et al., 2005) and human prostate cancer (Cheng et al., 2006),
SENP3 is elevated in a number of carcinomas, including ovarian, lung, rectum, colon,
and prostate (Han et al., 2010). Conversely, SENP6 mRNA is downregulated breast
tumor tissue compared to matched normal tissue (Mooney et al., 2010).

In addition to the overexpression of the SENPs in many cancers, studies have also
identified increases in expression of the SUMO conjugating machinery in a number of
cancers. Ubc9 expression is upregulated in certain lung and ovarian cancers (McDoniels-
Silvers et al., 2002). Injection of MCF7 breast cancer cells with attenuated Ubc9
expression in nude mice prevents tumor formation suggesting that Ubc9 may be a
potential therapeutic target (Mo et al., 2005). Additionally, increased expression of
SUMO2 and SAE1 in hepatocellular carcinoma patients is associated with poor prognosis
(Lee and Thorgeirsson, 2004). The SUMO E3 ligase PIAS3 has also been found to be
upregulated in a variety of human tumors including brain, colorectal, breast and prostate
cancers (Wang and Banerjee, 2004). Together, these studies suggest that both SUMO
conjugation and deconjugation contribute to cancer progression.
SUMO and neurodegenerative diseases
A number of proteins that play a role in a variety of human neurodegenerative diseases
such as Huntington’s disease, Parkinson’s disease, spinocerebellar ataxia type I (SCA1),
15

amyotrophic lateral sclerosis (SOD1) and Alzheimer’s disease are known to be
sumoylated. The huntingtin protein exhibits an expansion of the polyglutamine tract in
huntington’s disease. Sumoylation has been observed on lysine residues in the N-
terminus of the mutant huntingtin protein that has the extended polyglutamine tract
(Steffan et al., 2004). This sumoylation has been suggested to increase the stability of the
mutant protein leading to the accumulation of toxic intermediate poly-Q oligomers.
Hence, sumoylation of huntingtin promotes the neurodegenerative process in
Huntington’s disease. In SCA1 disease, ataxin-1 is the protein that has an abnormal
polyglutamine expansion. In this disease sumoylation of the mutant protein is decreased
compared to the wildtype protein, however the effect of this loss of sumoylation on
protein function is not known (Riley et al., 2005). In Parkinson’s and Alzheimer’s disease
the tau protein is expressed at high levels and is sumoylated at lysine 340 (Dorval and
Fraser, 2006). It is thought that this sumoylation of the tau protein competes with
ubiquitination and increases the stability of the protein. Another protein, PARK7 (DJ-1),
which functions as a transcriptional co-activator and molecular chaperone, and accounts
for 1-2% of early onset mutations in Parkinson’s disease is sumoylated at lysine 130
(Shinbo et al., 2006; Wang et al., 2006). This sumoylation is required for DJ-1-mediated
transforming and cell growth-promoting activities. In Alzheimer’s disease, the amyloid-b
(AB) protein, produced by the APP processing through the amyloidogenic proteolytic
pathway, is believed to be the causative factor. The APP protein is sumoylated at two
lysine residues (Zhang and Sarge, 2008a). Loss of sumoylation increases AB protein
aggregation suggesting that treatments that alter APP sumoylation could have therapeutic
16

potential for Alzheimer’s patients (Sarge and Park-Sarge, 2009). Hence, SUMO-mediated
regulation of a vast range of proteins plays an important role in neurodegeneration.

SUMO and heart disease
The sumoylation pathway has also been found to play a role in certain cardiomyopathies.
Lamin A is a protein that is important in maintaining nuclear structure and function.
Mutations in this protein have been identified in patients with familial dilated
cardiomyopathy and conduction system disease (Malhotra and Mason, 2009). In both
diseases, the lamin A protein is mutated at glutamic acid residue 203 (Zhang and Sarge,
2008b). Interestingly, this glutamic acid residue lies within the SUMO consensus motif.
Consistent with this the mutated lamin A protein from fibroblast cells of patients shows
decreased sumoylation which in turn leads to abnormal lamin A localization and nuclear
morphology.

In another study of a patient with severe developmental defects including cleft lip and
cardiac malformation, a deletion of SAE1, the SUMO activating enzyme, was identified
which implicates a role for this pathway in normal cardiac development (Leal et al.,
2009).

The above examples illustrate the importance of maintaining the SUMO balance in a cell
for proper functioning of the cell. Furthermore, loss of this SUMO balance affects a wide
range of proteins and the outcomes of this deregulation depends on the proteins that are
17

affected. Sumoylation of a protein provides the protein with a region to interact with a
host of other proteins and can help in the assembly of protein complexes which then
affects transcriptional regulation.

SUMO interactions are most often associated with transcriptional repression. The
annotated yeast two hybrid interaction between Maf1 and SUMO2 indicates a potential
role of SUMO in regulating Maf1 function. In this study, we determined that Maf1
interacts with SUMO both covalently and non-covalently. We further examined how this
interaction affects Maf1 mediated transcriptional repression. This study highlights a new
mechanism by which Maf1 works to repress RNA pol III mediated transcription.

18

Chapter 2: Covalent SUMO modification regulates Maf1 function
2.1 SUMO negatively regulates RNA polymerase III mediated transcription
Previous studies in our lab demonstrated that enhanced expression of RNA polymerase
III transcribed genes drives cellular transformation and tumorigenesis (Johnson et al.,
2008b). To further determine how RNA pol III transcribed genes are regulated we
investigated the role of sumoylation. Myc- tagged SUMO1 or SUMO2 were
overexpressed in U87 glioblastoma cells to increase the sumoylation status of cellular
proteins. Increased expression of SUMO1 and SUMO2 resulted in a decrease in
expression of RNA pol III-dependent pre-tRNA
Leu
, pre-tRNA
i
met
and 7SL RNA
transcripts (Fig. 1).

FIGURE 1. SUMO negatively regulates RNA pol III
transcribed genes. U87 cells were transfected with empty
vector, myc-SUMO1ρ or myc-SUMO2ρ as indicated. 48 hours
post transfection total RNA and protein were isolated. RNA was
converted to cDNA which was then subjected to qRT-PCR
analysis using primers specific for tRNA
Leu
, tRNA
i
met
, 7SL and
GAPDH. Gene expression was quantified using GAPDH as an
internal control. Protein lysates were subjected to immunoblot
analysis using antibodies specific for myc and actin.

19

2.2 Maf1 is covalently modified by SUMO
Since Maf1 is a negative regulator of RNA pol III- dependent genes we investigated
whether Maf1 is sumoylated. To determine whether Maf1 is sumoylated, 293T cells were
transfected with a myc-tagged SUMO1 expression vector. Cell lysates were
immunoprecipitated with anti-Maf1 antibodies and immunoblotted with antibodies to
Maf1 and myc. Immunoblot analysis with Maf1 antibodies displayed a higher molecular
weight band in the myc-SUMO1 transfected cells, in addition to the unmodified Maf1
bands (Fig 2B). Immunblot analysis with myc antibodies revealed that the higher
molecular weight band represented sumoylated Maf1 (Fig 2C).

Figure 2. Endogenous Maf1 is sumoylated. (A, B & C) 293T cells were transfected with a myc-
SUMO2ρ expression vector. Cell lysates were immunoprecipitated with anti-Maf1 antibodies and
immunoblotted with anti-Maf1 or anti-myc antibodies.

To further confirm that Maf1 is covalently modified by SUMO, an HA- tagged Maf1
expression vector was coexpressed with myc- tagged SUMO1 or SUMO2 in COS7 cells.
Immunoblot analysis revealed a marked increase in Maf1 expression upon
A
.

B.
.

C.
.

20

overexpression of SUMO1 and SUMO2 (Fig. 3). Furthermore, a new higher molecular
weight band appeared at around 50kDa in addition to the normally observed Maf1
polypeptides. Immunoprecipitation with HA antibodies followed by immunoblot analysis
with myc antibodies confirmed that this shifted band was sumoylated Maf1. Together, the
results confirm that Maf1 is covalently modified by SUMO1 and SUMO2.

FIGURE 3. Maf1 is covalently modified by SUMO1 and SUMO2. COS-7 cells were
transfected with wild type Maf1-HA and myc-SUMO1ρ or myc-SUMO2ρ expression vectors
were co-transfected, where indicated. Cell extracts were collected 48 h post-transfection and
analyzed by immunoblot with anti-HA and anti-actin antibodies. Cell lysates from above were
immunoprecipitated with anti-HA antibody and immunoblotted with anti-myc antibody.

2.3 SUMO enhances Maf1 protein expression
Sumoylation is known to affect protein stability (Hay, 2005) and since we observed an
increase in Maf1 protein expression upon increased SUMO1 or -2 expression we
investigated whether sumoylation affects endogenous Maf1 protein expression. We used
siRNA to downregulate Ubc9, the only SUMO conjugating enzyme, in COS7 cells.
Immunblot analysis revealed decreased expression of endogenous Maf1 in cells
transfected with Ubc9 siRNA compared to control siRNA (Fig. 4A). QPCR analysis
21

determined no change in Maf1 mRNA level upon Ubc9 downregulation suggesting that
sumoylation may regulate Maf1 protein expression (Fig. 4B). Consistent with decreased
Maf1 protein expression, RNA pol III transcribed genes, namely pre-tRNA
Leu
, 7SL and
U6 RNA were upregulated upon Ubc9 downregulation (Fig 4B). Hence, sumoylation
regulates RNA pol III –dependent transcription.

FIGURE 4. Ubc9 silencing inhibits
Maf1 protein expression. (A) COS7 cells
were transfected with either mmRNA or
Ubc9 siRNA. Cell lysates were collected
48 hours later and analyzed by
immunoblot with anti-Ubc9, anti-Maf1
and anti-actin antibodies respectively.
Densitometry was used to quantify the
bands using β-actin as a control. (B) Total
RNA was isolated from cells transfected
same as above. qRT-PCR analysis was
carried out using primers specific for
Maf1, pre-tRNA
Leu
, 7SL, U6 and GAPDH.
Gene expression was quantified relative to
GAPDH as an internal control.

2.4 Maf1 is covalently modified by SUMO at lysine 35
SUMO is most often conjugated to a lysine residue that lies within a SUMO consensus
motif which is ΨKxE/D. However, there are proteins that are sumoylated at lysine
residues that do not lie within a SUMO consensus motif such as the Daxx protein (Lin et
al., 2006). Analysis of the Maf1 protein sequence (Fig. 5) revealed no SUMO consensus
22

motif. Hence, we mutated each of the eleven lysine residues in Maf1 to arginines by site
directed mutagenesis and performed an in vivo sumoylation assay with myc- tagged
SUMO2 in COS7 cells.
MKLLENSSFEAINSQLTVETGDAHIIGRIESYSCKMAGDDKHMFKQFCQEGQPHV
LEALSPPQTSGLSPSRLSKSQGGEEEGPLSDKCSRKTLFYLIATLNESFRPDYDFST
ARSHEFSREPSLSWVVNAVNCSLFSAVREDFKDLKPQLWNAVDEEICLAECDIYS
YNPDLDSDPFGEDGSLWSFNYFFYNKRLKRIVFFSCRSISGSTYTPSEAGNELDME
LGEEEVEEESRSRGSGAEETSTMEEDRVPVICI
Figure 5. Maf1 protein sequence with mutated lysine residues highlighted in red.
All Maf1 mutants, except Maf1K2R, were expressed at comparable levels. Upon
expression of myc- tagged SUMO2 all mutant proteins were capable of forming the
higher molecular weight sumoylated polypeptide with the exception of Maf1K35R (Fig
6). The Maf1K196R mutant showed a decrease in the amount of sumoylated band as
compared to Maf1WT. Since, the K196 residue lies in a putative nuclear localization
motif it is possible that this interferes with Maf1 sumoylation.

Figure 6. Maf1 lysine 35 is the major site of sumoylation. COS-7 cells were transfected with
wild type Maf1-HA or Maf1HA mutants and myc-SUMO2ρ expression vectors were co-
transfected. Cell extracts were collected 48 h post-transfection and analyzed by
immunoblot with anti-HA and anti-actin antibodies.
23

Next, to confirm that Maf1 lysine 35 is the major site of sumoylation by both SUMO1
and SUMO2 we coexpressed myc- tagged SUMO1 or SUMO2 with HA- tagged Maf1
wildtype (WT) or Maf1K35R. Immunoblot analysis revealed that upon expression of
SUMO1 or SUMO2, both Maf1WT and Maf1K35R protein expression was increased.
However, Maf1K35R showed a significant decrease in the higher molecular weight
sumoylated polypeptides compared to Maf1WT (Fig. 7). Furthermore,
immunoprecipitation of lysates with HA antibodies followed by immunoblot analysis
with myc antibodies demonstrated a significant loss of sumoylated Maf1 in lysates
transfected with Maf1K35R as compared to Maf1WT. This confirmed that lysine 35 on
Maf1 is the major site of sumoylation by both SUMO1 and SUMO2. Since some residual
sumoylation was observed with the Maf1K35R mutant there might be some other lysine
residue on Maf that is also sumoylated and hence Maf1K35R is the major site of
sumoylation.

Figure 7. Maf1 is sumoylated at lysine 35 by both SUMO1 and -2. COS7 cells were
transfected with either wildtype Maf1HA or Maf1HAK35R and myc-SUMO1 or SUMO2
expression vectors. Cell extracts were collected 48h post-transfection and analyzed by
immunoblot with anti-HA and anti-actin antibodies. Cell lysates were then immunoprecipitated
with anti-HA antibody and immunoblotted with anti-myc antibody.
24

2.5 Maf1 phosphorylation does not affect covalent SUMO modification
Previous studies have identified serine 75 as an important residue on Maf1 that is
phosphorylated by mTOR kinase which negatively regulates Maf1 function. Studies have
also shown that mutation of this serine residue to alanine improves Maf1 mediated
repression of RNA pol III dependent promoters. To determine whether there was
interplay between Maf1 phosphorylation and sumoylation we examined the ability of
Maf1 S75 mutants to be sumoylated. Our cell based sumoylation assays with Maf1S75A
and Maf1S75D determined that both forms could be covalently modified by SUMO1 and
SUMO2 (Fig. 8B). Additionally, protein expression of both the phospho mutant and
phosphor mimic was increased by SUMO. This indicated that phosphorylation at serine
75 does not affect sumoylation of Maf1. Next, to determine whether sumoylation at
lysine 35 affected phosphorylation of Maf1 we examined changes in the banding pattern
of Maf1WT and Maf1K35R in the absence and presence of SUMO on a phostag gel, that

Figure 8. Maf1 sumoylation does not affect Maf1 phosphorylation. (A) COS7 cells were
transfected with either Maf1WT or Maf1K35R and myc-SUMO1ρ expression vectors. Cell
extracts were collected 48 h post-transfection and analyzed by immunoblot on a phos-tag gel with
anti-HA antibodies. (B) COS-7 cells were co-transfected with plasmids expressing myc-SUMO1ρ
and Maf1-HAWT or Maf1HA mutants as indicated. Cell lysates were immunoblotted with anti-
HA and anti-β-actin antibodies.
25

seperates proteins based on the number of phosphor residues.. Our results demonstrated
that Maf1K35R had the same banding pattern as Maf1WT and this banding pattern was
unchanged in the presence of SUMO (Fig. 8A). Hence sumoylation and phosphorylation
at S75 have independent effects on Maf1.

2.6 Sumoylation is required for Maf1 function
Sumoylation is known to affect the function of the protein. We next determined the
functional consequences of Maf1 sumoylation at K35. We examined Maf1 -mediated
transcription repression of pre-tRNA
Leu
gene expression in U87 cells. QRT-PCR analysis
revealed that mutation at lysine 35 reduced the ability of Maf1 to repress RNA pol III –
dependent gene activity (Fig. 9A). To further assess Maf1 function, a colony suppression
assay was used to determine how Maf1 affects the cell growth. While expression of

Figure 9. Covalent SUMO modification is required for Maf1 function. (A) U87 cells were
transiently transfected with Maf1WT and Maf1K35R expression vectors. 48 hours post
transfection total protein and RNA were isolated from the cells. The protein was subjected to
immunoblot analysis with anti-HA and anti-actin antibodies. The RNA was converted to cDNA
and subjected to qRT-PCR using primers specific for TBP and GAPDH. Gene expression was
quantified using GAPDH as an internal control. (B) U87 cells were transiently transfected with
Maf1WT, Maf1K35R and puromycin resistant expression vectors. 48 hours post transfection
puromycin resistant colonies were selected. 3 weeks post selection colonies were stained with
crystal violet and counted.
26

Maf1WT resulted in a decrease in the number of colonies that formed, colony
suppression was abrogated with the expression of Maf1K35R (Fig. 9B). Taken together
our results demonstrate that covalent modification of Maf1 on lysine 35 is required for
Maf1 to repress transcription and suppress cellular growth.

2.7 SENP1 desumoylase regulates Maf1 sumoylation
Sumoylated proteins are rapidly deconjugated by a family of proteins called the SUMO
isopeptidases (SENPs). Since Maf1 is present in both the nucleus and the cytoplasm we

Figure 10. SENP1 desumoylase brings about Maf1 desumoylation and positively regulates
RNA pol III dependent transcription. (A) U87 cells were transfected with Maf1HA and myc-
SUMO1 expression vectors. mmRNA or SENP1 siRNA were transfected as indicated. Cells
transfected with mmRNA were used as a negative control. Protein lysates were immunoblotted
with anti-HA and anti-actin antibodies. (B) U87 cells were transfected with either mmRNA or
SENP1 siRNA. Total RNA was isolated and converted to cDNA. QPCR analysis was carried out
using primers specific for pre-tRNAimet and U6 RNA. Gene expression was calculated using
GAPDH as an internal control.

determined whether SENP1 desumoylates Maf1. SENP1 was downregulated in U87 cells
and the fraction of sumoylated Maf1WT was determined by immunoblot analysis using
27

HA antibodies. U87 cells transfected with mmRNA was used as a negative control. We
found that compared to mmRNA, knockdown of SENP1 displayed an increase in the
sumoylated Maf1 fraction (Fig 10A). Furthermore, knockdown of SENP1 repressed
transcription of tRNA
i
met
and U6 RNA (Fig. 10B). This is consistent with our previous
data that covalent SUMO modification of Maf1 is required for Maf1-mediated repression
of RNA pol III –dependent transcription.

2.8 SUMOylation at lysine 35 does not alter Maf1 subcellular localization or protein
stability
To further investigate the mechanism by which covalent modification of Maf1 at lysine
35 affects Maf1 function we determined whether SUMO regulates Maf1 localization.
Immunofluorescence staining of U87 cells transfected with either HA-tagged Maf1WT or
Maf1K35R revealed that Maf1 was predominantly nuclear with a small fraction in the
cytoplasm (Fig 11). These results indicate that covalent attachment of SUMO at lysine 35
does not affect the subcellular distribution of Maf1.
28

Figure 11. Covalent SUMO modification does not affect Maf1 localization. U87 cells were
transiently transfected with either Maf1WT or Maf1K35R expression vectors. 24 hours post
transfection immunofluorescence staining was carried out with HA antibody and FITC-
conjugated secondary antibody. The nuclei were counterstained with propidium iodide. Confocal
microscopy was used to visualize the cells.

Since we observed an increase in Maf1 protein expression upon SUMO overexpression
we determined whether covalent SUMO modification of Maf1 on lysine 35 is required to
increase Maf1 protein stability. COS7 cells transiently transfected with either Maf1WT or
Maf1K35R expression vectors were treated with cyclohexamide for 0, 2, 4, 6 and 8
hours. Protein lysates were then subjected to immunoblot analysis for HA and actin. Both
Maf1WT and Maf1K35R showed similar protein decay rates (Fig. 12). Hence,
29

sumoylation at lysine 35 does not affect Maf1 protein stability. Taken together, our data
demonstrate that the loss of function of Maf1K35R is not due to a change in subcellular
localization or protein stability.

Figure 12. Sumoylation at Maf1K35 does not alter protein stability. COS7 cells were
transfected with either Maf1WT or Maf1K35R expression vectors. 48 hours post-transfection
cells were treated with cyclohexamide and protein was isolated at the indicated time points.
Immunoblot analysis was carried out using anti-HA and anti-actin antibodies. The graph
represents the change in total protein content at different times after cyclohexamide treatment
compared to actin.

2.9 SUMOylation affects Maf1 interaction with RNA pol III
To elucidate the mechanism by which covalent SUMO modification affects Maf1
function we examined how sumoylation at this site would affect the interaction between
Maf1K35R and RNA pol III. Protein lysates from 293T cells expressing either HA-
tagged Maf1WT or Maf1K35R were immunoprecipitated with antibodies against the
RNA pol III subunit, RPC39. Immunoblot analysis with HA antibodies revealed that
Maf1K35R had a reduced ability to interact with RNA pol III as compared to Maf1WT
(Fig. 13) and this interaction was impaired even in the presence of SUMO.
30

Figure 13. SUMO modification is required for Maf1 interaction with RNA polymerase III.
293T cells were cotransfected with Maf1WT or Maf1K35R and myc-SUMO1 expression vectors.
48 hours post transfection total protein was isolated form cells and immunoprecipitated with anti-
RNA pol III antibody followed by immunoblot analysis with anti-HA antibody. Cells transfected
with empty vector were used as a negative control.

2.10 SUMOylation affects Maf1 recruitment to RNA polymerase III- dependent gene
promoters
Since the Maf1K35R mutant protein could no longer associate with RNA pol III, we next
determined whether this prevented Maf1 recruitment to the RNA pol III -dependent gene
promoters. We performed chromatin immunoprecipitation (ChIP) assays to determine
whether the Maf1K35R mutant is recruited to the tRNA
Leu
promoter. Chromatin from
U87 cells transfected with either HA-tagged Maf1WT or Maf1K35R were pulled down
with HA antibodies. QPCR analysis determined enrichment of Maf1WT but not
Maf1K35R at the tRNA
Leu
promoter (Fig. 14). Consistent with this RNA pol III was
displaced from the tRNA
Leu
promoter in cells transfected with Maf1WT but not
Maf1K35R. Hence, sumoylation of Maf1 at lysine 35 is required for recruitment of Maf1
to RNA pol III transcribed promoters.
31

Figure 14. Covalent SUMO modification is required for Maf1 recruitment to tRNA
Leu

promoter. U87 cells were transfected with either HA-tagged Maf1WT or Maf1K35R expression
vectors. Cells transfected with empty vector were used as a negative control. 48 hours post
transfection chromatin was isolated and immunoprecipitated with antibodies to HA, RPC39 and
IgG. DNA was then isolated and qPCR analysis was carried out using primers to the tRNA
Leu

gene promoter. After normalizing to IgG the vector control was set to 1 and fold change was
calculated.

2.11 Summary
In this chapter, we have identified sumoylation as a new pathway by which RNA
polymerase III mediated gene transcription is regulated. Sumoylation negatively regulates
RNA pol III-mediated transcription. A major mechanism by which this occurs is through
the covalent modification of Maf1 by SUMO1 and SUMO2. Maf1 is modified by SUMO
at lysine 35. This covalent modification of Maf1 by SUMO is required for Maf1’s ability
to repress RNA pol III mediated transcription and to suppress colony growth.
Furthermore, sumoylation of Maf1 at lysine 35 is required for Maf1 to associate with
RNA pol III and to be recruited to RNA pol III -dependent promoters.
32

Chapter 3: Non-Covalent SUMO interactions regulate Maf1 function
3.1 Maf1 non-covalently interacts with SUMO1 and SUMO2
A genome-wide yeast two hybrid screen annotated an interaction between Maf1 and
SUMO2. To determine whether Maf1 non-covalently interacts with SUMO1 and
SUMO2, COS7 cells were cotransfected with HA- tagged Maf1WT and myc-SUMO1 or
-2. Immunoprecipitation with SUMO1 or SUMO2 antibodies was carried out to pull
down all sumoylated proteins. Immunoblot analysis with HA antibody confirmed the
presence of the sumoylated Maf1HA band at 50kDa. Additionally, a band was observed
at 37kDa which represents non-sumoylated Maf1HA and this indicates that Maf1 can
non-covalently interact with SUMO1 and SUMO2 (Fig. 15).

Figure 15. Maf1 interacts non-covalently with SUMO1 and SUMO2. COS-7 cells were
transfected with wild type Maf1-HA and myc-SUMO1ρ or myc-SUMO2ρ expression vectors
were co-transfected, where indicated. Cell extracts were collected 48 h post-transfection and
analyzed by immunoblot with anti-HA. Cell lysates were immunoprecipitated with anti-SUMO1
or anti-SUMO2 and immunoblotted with anti-HA antibodies.

Next, we asked whether covalent modification of Maf1 at lysine 35 by SUMO was
required for its non-covalent interaction with SUMO. Immunoprecipitation with anti-
SUMO antibodies followed by immunoblot analysis did not detect a 50kDa band for the
33

Maf1K35R non-SUMO modified mutant compared to the Maf1WT. However,
Maf1K35R retained non-covalent interaction with SUMO1 as observed by the presence
of the 37kDa Maf1 bands (Fig. 16). Hence, covalent modification by SUMO at K35 is
not required for non-covalent interaction of Maf1 with SUMO.
Figure 16. Covalent SUMO modification is not
required for non-covalent Maf1-SUMO
interaction. COS7 cells were transfected with
either wildtype Maf1HA or Maf1HAK35R and
myc-SUMO1 or SUMO2 expression vectors. Cell
extracts were collected 48h post-transfection and
analyzed by immunoblot with anti-HA antibodies.
Cell lysates were immunoprecipitated with anti-
SUMO1 or anti-SUMO2 and immunoblotted with
anti-HA antibodies.

3.2 Serine 123 on Maf1 is important for non-covalent interaction with SUMO1 and
SUMO2
Non-covalent interaction of a protein with SUMO is mediated through a region on the
protein that is termed the SUMO Interacting Motif (SIM). So far, three SIM motifs have
been defined, which are, h-h-X-[S]-[X]-[S/T]-a-a-a (Minty et al., 2000),where h is any
hydrophobic amino acid and a is any acidic amino acid, [V/I]-X-[V/I]-[V/I] and K-X
3-5
-
[I/V]-[I/L]-[I/L]-X
3
-[D/E/Q/N]-[D/E]-[D/E] (Hannich et al., 2005; Song et al., 2004). To
identify the SIM motif on Maf1 we carried out mutagenesis within sequences that
resemble the S-X-S motif (Fig. 17). In these putative SIM sequences, we mutated each
serine to alanine and performed cell based sumoylation assays. Our results demonstrated
that all four Maf1 mutants could be covalently modified by SUMO1 (Fig. 18A) and
34

SUMO2 (data not shown). With the exception of Maf1S123A, which was expressed at
lower amounts relative to Maf1WT, the other mutations did affect Maf1 expression.
MKLLENSSFEAINSQLTVETGDAHIIGRIESYSCKMAGDDKHMFKQFCQEGQPHV
LEALSPPQTSGLSPSRLSKSQGGEEEGPLSDKCSRKTLFYLIATLNESFRPDYDFST
ARSHEFSREPSLSWVVNAVNCSLFSAVREDFKDLKPQLWNAVDEEICLAECDIYS
YNPDLDSDPFGEDGSLWSFNYFFYNKRLKRIVFFSCRSISGSTYTPSEAGNELDME
LGEEEVEEESRSRGSGAEETSTMEEDRVPVICI
Figure 17. Maf1 protein sequence with mutated serines highlighted in blue.
Next, to determine the ability of these mutants to non-covalently interact with SUMO, we
immunoprecipitated total protein lysates with anti-SUMO1 and immunoblotted with anti-
HA. We found that while S31A, S75A and S233A Maf1 mutants maintained non-
covalent interaction with SUMO1, the Maf1S123A mutant displayed an enhanced ability
to interact with SUMO1 as compared to Maf1WT (Fig 18B).

Serine residues on proteins can be modified which can affect protein function. Mutation
of serine to alanine prevents phosphorylation of the protein. Since Maf1S123A interacts
better than Maf1WT with SUMO1, we asked whether the charge at the serine residue
introduced by phosphorylation prevents SUMO1 interaction. We mutated serine 123 to
aspartic acid which acts as a phospho-mimic and performed cell based sumoylation and
immunoprecipitation assays. We found that Maf1S123D could be covalently modified by
SUMO1 and SUMO2 (Fig. 18C). However, mutation of serine 123 to aspartic acid
reduced non-covalent interaction with SUMO1 compared to Maf1S123A (Fig. 18C). We
35

concluded that the charge at Maf1S123 likely due to phosphorylations inhibits non-
covalent interaction of Maf1 with certain sumoylated proteins.
A.

B.

C.

FIGURE 18. Maf1S123A displays enhanced interaction with SUMO. (A & C) COS-7 cells
were co-transfected with plasmids expressing myc-SUMO1ρ and Maf1-HA wild type or Maf1HA
mutants as indicated. Cell lysates were immunoblotted with anti-HA and anti-β-actin antibodies.
(B & C) These cell lysates were immunoprecipitated with anti-SUMO1 and immunoblotted with
anti-HA antibodies.
36

3.3 Mutation of Maf1 serine 123 to alanine enhances Maf1 function
Next, we determined how non-covalent interaction of Maf1 with SUMO affects Maf1
function. We examined the ability of Maf1S123 mutants to repress RNA polymerase III -
dependent transcription in U87 cells. QPCR analysis determined that mutation of serine
123 to alanine enhanced the ability of Maf1 to repress pre-tRNA
Leu
gene expression
compared to Maf1WT (Fig. 19A). Conversely, Maf1S123D demonstrated a reduced
ability to repress pre-tRNA
Leu
gene expression compared to Maf1S123A suggesting that
the charge at Maf1S123 negatively affects Maf1 function.
A. B.
A.

FIGURE 19. Non-covalent SUMO interaction enhances Maf1 function. (A) U87 cells were
transiently transfected with Maf1WT, Maf1S123A and Maf1S123D expression vectors. 48 hours
post transfection total RNA was isolated from the cells, converted to cDNA and subjected to
qRT-PCR using primers specific for tRNA
Leu
and GAPDH. Gene expression was quantified using
GAPDH as an internal control. (B) U87 cells were transiently transfected with Maf1WT,
Maf1S123A, Maf1S123D and puromycin resistant expression vectors. 48 hours post transfection
puromycin resistant colonies were selected. 3 weeks post selection colonies were stained with
crystal violet and counted.

To further determine the growth suppressive properties of the Maf1S123 mutants, a
colony suppression assay was performed in U87 cells. While the Maf1S123A mutant
37

demonstrated an enhanced ability to suppress the number of colonies, Maf1S123D had a
reduced ability to suppress colony growth (Fig 19B). Together, our results demonstrate
that enhanced non-covalent interaction of Maf1 with some sumoylated protein(s)
enhances its ability to repress transcription and to suppress colony growth.

Figure 20. Maf1S123A subcellular localization is unchanged compared to Maf1WT. U87
cells were transiently transfected with either Maf1WT or Maf1S123A expression vectors. 24
hours post transfection immunofluorescence staining was carried out with HA antibody and
FITC-conjugated secondary antibody. The nuclei were counterstained with propidium iodide.
Confocal microscopy was used to visualize the cells.

38

To investigate the mechanism by which non-covalent SUMO interaction affects Maf1
function we determined the subcellular localization of the Maf1 mutants.
Immunofluorescence staining of U87 cells transfected with either HA-tagged Maf1WT or
Maf1S123A revealed that, in either case, Maf1 was predominantly nuclear with a small
fraction in the cytoplasm (Fig. 20). These results indicate that non-covalent SUMO –
Maf1 interactions do not affect the subcellular distribution of Maf1.

3.4 Maf1 S123A interaction with RNA polymerase III is enhanced
Previous studies have shown that Maf1 is recruited to RNA polymerase III transcribed
gene promoters and Maf1 occupancy is inversely correlated with occupancy of RNA pol
III and the TFIIIB subunits. To elucidate the mechanism by which non-covalent
interaction of Maf1 with SUMO enhances Maf1-mediated repression we determined the
ability of Maf1S123A to interact with RNA polymerase III. Protein lysates from 293T
cells transfected with either HA-tagged Maf1WT or Maf1S123A were
immunoprecipitated with antibodies to the RNA pol III subunit, RPC39. Immunoblot
analysis with anti-HA revealed that two fold more Maf1S123A immunoprecipitated with
RNA pol III compared to Maf1WT (Fig. 21A).

Additionally, in the presence of SUMO1, the interaction between Maf1S123A and RNA
pol III was further enhanced compared to Maf1WT (Fig. 21B). Our results indicate that
the charge at Maf1S123 prevents the interaction with RNA polymerase III. Abolishing
39

the charge at Maf1S123 by mutating serine to alanine enhances the interaction of Maf1
with RNA pol III in the presence of SUMO.

A.

B.

FIGURE 21. Maf1S123A displays enhanced interaction with RNA polymerase III. (A & B)
293T cells were co-transfected with Maf1WT or Maf1S123A and myc-SUMO1ρ. Cells lysates
were immunoblotted with HA and β-actin antibodies. Cells lysates were then immunoprecipitated
with RNA pol III antibodies and probed with HA antibody.

40

3.5 Summary
In this chapter, we showed that Maf1 interactions with certain cellular proteins are
mediated through a non-covalent interaction between Maf1 and SUMO. Covalent
modification of Maf1 was not required for this non-covalent SUMO interaction. We
identified serine 123 as an important residue on Maf1 that affects non-covalent
interaction with SUMO. Additionally, we provide evidence that serine 123 on Maf1 is
phosphorylated and that mutation of this serine to alanine enhances Maf1 non-covalent
interaction with SUMO. Conversely, mutation of serine to glutamic acid does not
enhance Maf1 non-covalent interaction with SUMO compared to Maf1S123A. This
increased interaction with SUMO increases Maf1 mediated repression of RNA pol III
transcribed genes. The Maf1S123A also displays an enhanced ability to suppress colony
growth compared to Maf1WT and Maf1S123D. The Maf1S123A mutant also exhibits
increased interaction with RNA polymerase III and this interaction is further enhanced in
the presence of SUMO. These results support the idea that Maf1 associates with cellular
proteins and these interactions require sumoylation of the Maf1-interacting proteins.
While the various Maf1-associated proteins remain to be identified, our results support
the idea that RNA pol III may be sumoylated and that it’s association with Maf1 requires
sumoylation. Hence, non-covalent interaction between Maf1 and some sumoylated
protein enhances Maf1 mediated repression of RNA pol III transcribed genes.

41

Chapter 4: Identification of a novel protein-protein interaction between
Maf1 and Mediator
Previous work in our lab has shown that Maf1 is recruited to TBP and other select Elk-1
regulated genes. Increased binding of Maf1 results in the displacement of Elk-1 from the
promoter (Johnson et al., 2007). Maf1 is also recruited to RNA pol III gene promoters
and its occupancy is inversely correlated with the occupancy of TFIIIB and RNA pol III
subunits. Examination of the Maf1 sequence, however, did not highlight any putative
DNA binding sequence suggesting that Maf1 is probably recruited to promoters through
specific protein-protein interactions. In order to understand how Maf1 is recruited to
promoters, we performed coimmunoprecipitation and mass spectrophotometric analyses
to identify Maf1 interacting proteins (Sandra Johnson, unpublished data). Our analyses
identified Mediator 13 as one of several Maf1-interacting proteins.

Mediator is a large multi-subunit protein complex that was first discovered in
Saccharomyces cerevisiae. The core mediator contains more than 20 subunits that bind
DNA and are required for activator-dependent transcription (Myers and Kornberg, 2000).
Mediator (MED) 13 is part of a smaller sub-complex that can associate with the large
Mediator complex to affect transcription. The sub-complex is made up of 4 proteins,
namely, MED 12, MED13, Cyclin Dependent Kinase (CDK) 8, and cyclin C. This
complex is known as the CDK8 sub-module. The CDK8 sub-module can function to
positively or negatively regulate transcription.

42

The CDK8 sub-module can negatively regulate transcription in both, a kinase-dependent
and kinase-independent manner. In a kinase-dependent manner, the CDK8 sub-module
phosphorylates and inactivates CDK7 (TFIIH) and the general transcription machinery
(Taatjes, 2010). The CDK8 sub-module can also stearically hinder the association
between RNA pol II and the large mediator complex and repress transcription in a kinase
independent manner (Elmlund et al., 2006; Knuesel et al., 2009). In yet another kinase-
independent mechanism of transcription repression by the CDK8 sub-module, the
MED12 subunit can recruit G9a, the H3K9 methyltransferase, to the promoters of genes
regulated by the RE1 silencing transcription factor (Ding et al., 2008).

While in vitro studies point to a repressive role for the CDK8 sub-module in transcription
regulation, conflicting evidence in vivo indicate a positive role for CDK8 in regulating
transcription. CDK8 positively regulates several serum response genes through
recruitment of P-TEFb and BRD4 to the elongating RNA pol II (Donner et al., 2010).
CDK8 also activates several p53 target genes such as p21 and Hdm2, where the CDK8
sub-module occupancy correlates positively with p53-dependent gene activation (Donner
et al., 2007). The large Mediator complex is also required for thyroid receptor (TR) and
thyroid hormone (T3) dependent activation of transcription (Fondell et al., 1996).
Furthermore, mediator complexes containing the CDK8 sub-module are recruited to the
human type I deiodinase (DioI) gene in a T3- and TR- dependent manner and knockdown
of CDK8 decreases RNA pol II occupancy at the DioI gene promoter (Belakavadi and
Fondell, 2010).
43

CDK8 can also function to positively regulate transcription of β-catenin-driven genes
(Firestein et al., 2008). CDK8 is recruited to the myc promoter where it positively
regulates myc transcription. Suppression of CDK8 reduced b-catenin occupancy at the
myc promoter suggesting that the CDK8 sub-module functions as a direct regulator of b-
catenin-driven transcription. The canonical Wnt/b-catenin signaling pathway is aberrantly
activated in almost all colorectal cancers. Additionally, the CDK8 locus is amplified in a
significant number of colon cancers. Furthermore, knockdown of CDK8 in colon cancer
cells, characterized by CDK8 and β-catenin hyperactivity, inhibits proliferation. CDK8
kinase activity is required for β-catenin-driven transformation.

Our preliminary studies identify Maf1 as a potential tumor suppressor while other studies
have established CDK8 as a potent oncogene. In this study, we confirm the interaction
between Maf1 and the CDK8 sub-module and begin to examine how the interaction
between Maf1 and the CDK8 sub-module may regulate Maf1 function.

44

Results
4.1 Maf1 interacts with the CDK8 subcomplex
To confirm the interaction between Maf1 and the CDK8 subcomplex,
immunoprecipitation of protein lysates expressing HA-tagged Maf1 was carried out with
either CDK8 or MED13 antibodies. Immunoprecipitation with anti-MED8, a member of
the large Mediator complex was also carried out. Immunoprecipitation with anti-IgG was
performed as negative controls. Immunoblot analysis with HA antibodies detected the
presence of Maf1HA in lysates immunoprecipitated with CDK8 and MED13 but not with
MED8 and IgG (Fig. 22). This revealed that Maf1 interacts with and the CDK8
subcomplex, but not the Mediator complex.

Figure 22. Maf1 interacts with CDK8
and MED13. 293T cells were
transiently transfected with either
Maf1HA overexpression plasmid or
vector alone. 48 hours post transfection
cell extracts were collected and
subjected to immunoprecipitation with
anti-MED8, anti-CDK8, anti-MED13 or
IgG antibodies and imuunoblotted with
anti-HA antibody. Whole cell lysates
were analyzed by immunoblot using
anti-HA antibody.

4.2 CDK8 regulates Maf1 target genes
After identifying the interaction between Maf1 and the CDK8 sub-module we asked
whether CDK8 regulates Maf1 targets, namely TBP and tRNA
Leu
genes. We used siRNA
to downregulate CDK8 expression in U87 cells and performed qPCR to look at the
45

expression of CDK8, Maf1, TBP and pre-tRNA
Leu
. While Maf1 expression was
unchanged, TBP expression was decreased upon CDK8 knockdown (Fig. 23).
Interestingly, CDK8 downregulation also led to a decrease in pre-tRNA
Leu
gene
expression. This is the first demonstration of RNA pol III dependent regulation by
CDK8. Our results suggest that CDK8 positively regulates Maf1 target genes.

Figure 23. Decreased CDK8 expression reduces TBP and tRNA gene expression. U87 cells
were transiently transfected with either a mismatch RNA or a CDK8-specific siRNA. Total RNA
was isolated, reverse transcribed and subjected to qRT-PCR analysis using primers specific for
CDK8, TBP, Maf1, pre-tRNA
Leu
and GAPDH. Fold changes in expression were quantified using
GAPDH as an internal control.

4.3 CDK8 is recruited to gene promoters regulated by Maf1
The CDK8 sub-module is known to occupy promoters of target genes where it positively
regulates expression. Since CDK8 positively regulates expression of Maf1 target genes
we determined whether this effect is through the direct recruitment of CDK8 to Maf1
target gene promoters. ChIP analysis determined that CDK8 and MED13 are recruited to
the TBP promoter near the transcription start site compared to an upstream control region
(Fig 24). In addition to TBP, we also found Maf1 enrichment on the tRNA
Leu
gene
46

promoter as expected. However, the enrichment signal of CDK8 and MED13 while
above IgG was very low compared to Maf1. Hence, more experiments are required to
confirm the occupancy of the CDK8 sub-module on RNA pol III –transcribed gene
promoters.

Figure 24. CDK8 and MED13 are present at TBP and tRNA
Leu
promoters. U87 cells were
used to prepare chromatin extracts that were then subjected to immunoprecipitations with Elk1,
Maf1, CDK8 or MED13 antibodies. qPCR analysis was performed using specific primers
designed to amplify sequences of TBP or tRNA gene promoters. (A) Analysis of transcription
factors bound to sequences adjacent to the TBP gene. Primers used to amplify sequences relative
to the transcription start site of the TBP gene are designated in schematic (top). (B) Analysis of
transcription factors bound to the tRNA
Leu
gene promoter. Primers that span the coding sequence
of the tRNA
Leu
gene were used. Bars represent three independent determinations from two
separate chromatin preparations.
47

4.4 Summary
In this chapter, we identified new Maf1 interacting proteins. Maf1 interacts with atleast
two members of the CDK8 sub-module, which consists of MED12, MED13, cyclin C and
CDK8. We also found that CDK8 positively regulates Maf1 target genes, namely TBP
and tRNA
Leu
. This is the first indication of regulation of RNA pol III dependent
transcription by CDK8. Additionally, CDK8 and MED13 co-occupy the TBP and
tRNA
Leu
gene promoters with Maf1. These results highlight the possibility that CDK8
and Maf1 function in opposing fashion to co-regulate common target genes.
48

Chapter 5: Discussion
5.1 Sumoylation regulates RNA polymerase III mediated transcription
Here we present evidence to show that RNA pol III mediated transcription is negatively
regulated by sumoylation. Increased RNA pol III dependent transcription is a hallmark of
tumorigenesis and this process is tightly regulated by the cell in a number of ways. The
TFIIIB transcription initiation complex is regulated by a number of tumor suppressors
and oncogenes either at the level of expression or function (White, 2004). Brf1 and Bdp1
are phosphorylated by a number of cell signaling pathways which alters their function. In
this study, we identified sumoylation as a new regulatory mechanism that controls RNA
pol III mediated transcription. We have shown that by either increasing or decreasing the
cellular sumoylation concentration, RNA pol III -mediated transcription is affected.
While it is not clear how loss of sumoylation or lack of desumoylation contributes to
tumorigenesis, different components of the SUMO machinery have been shown to be
deregulated in cancer. Interestingly, Ubc9 was found to be overexpressed in breast cancer
tumor samples (Wu et al., 2009). Alternatively, SENP1 expression was overexpressed in
prostate cancer samples (Wang et al., 2012) and stable overexpression of SENP1 in
mouse prostate tissue leads to the formation of PIN lesions followed by progression to
prostate cancer (Bawa-Khalfe et al., 2010). These studies indicate that while increased
sumoylation may contribute to breast cancer, decreased sumoylation is important in
prostate cancer. This indicates that it might be the specific SUMO targets that may be
differentially regulated in different cancers. In our study, decreased expression of Ubc9
enhanced RNA pol III dependent transcription and decreased expression of SENP1
49

repressed RNA pol III dependent transcription. Since SENP1 positively regulates RNA
pol III dependent transcription, it is possible that increases in SENP1 expression in
prostate cancer may contribute to increased RNA pol III dependent transcription.

Our data also showed that increased expression of SUMO represses transcription of
tRNALeu and U6 RNA. Both these genes represent two different classes of RNA pol III
–dependent genes that require different components of the general transcription
machinery. While sumoylation could regulate RNA pol III dependent transcription
through modification of a number of factors of the transcription machinery such as
TFIIIB, TFIIIC, or RNA polymerase itself, we specifically examined the Maf1, which is
the common negative regulator of RNA pol III –dependent transcription from all classes
of RNA pol III-dependent genes. Furthermore, preliminary studies in our lab have found
that Maf1 is lost in PTEN negative prostate and liver tumors and functions as a tumor
suppressor. In our study, we identified Maf1 as one potential target through which
sumoylation negatively regulates RNA pol III mediated transcription.

5.2 Maf1 is covalently modified by SUMO
Maf1 is an important repressor of RNA pol III mediated transcription and was first
identified in yeast. So far, phosphorylation of Maf1 has been identified as a key
posttranslational modification that regulates Maf1 function. In yeast, phosphorylated
Maf1 is sequestered in the cytoplasm and upon dephosphorylation it is shuttled into the
nucleus where it acts to repress RNA pol III transcription. However, many of these
50

phosphosites are not conserved in mammalian Maf1 suggesting that other mechanisms
function to regulate mammalian Maf1 function. In our study, we have shown that
mammalian Maf1 is covalently modified by both SUMO1 and SUMO2. Using site-
directed mutagenesis we identified lysine 35 as the site of covalent SUMO modification
on Maf1. While lysine 35 does not lie within a SUMO consensus motif, it is known that
sumoylation can also occur at lysine residues that are not within SUMO consensus
motifs.

Recent studies have shown that mammalian Maf1 is phosphorylated at serine 64, 68 and
75 by mTOR kinase (Michels et al., 2010). This phosphorylation inhibits Maf1 mediated
repression of RNA pol III transcription. Certain proteins have been identified that require
phosphorylation at a serine residue to stimulate sumoylation and the modified lysine
residue lies within a region that is termed the phosphorylation dependent sumoylation
motif (PDSM) (Hietakangas et al., 2006). While Maf1 lysine 35 does not lie within a
SUMO consensus motif it was possible that these two post translational modifications
could work to regulate modification of each other in order to regulate Maf1 function.
However, our results demonstrate that sumoylation did not affect the phosphorylation
state of Maf1 as indicated by the unchanged phophorylation profile of Maf1 on a phostag
gel in the absence and presence of SUMO. Furthermore, the Maf1K35R mutant had the
same phosphorylation profile as that of Maf1WT. Additionally, phosphorylation at serine
75 did not affect sumoylation at lysine 35 on Maf1 since both Maf1S75A and Maf1S75D
51

were sumoylated by both SUMO1 and SUMO2. This indicated that Maf1 function is
separately regulated by these two post translational modifications.

5.3 Covalent SUMO modification is required for Maf1 function
Sumoylation of a protein is known to affect the protein in a number of ways. It can
change protein stability, subcellular localization and function. In our study, we showed
that sumoylation of Maf1 at lysine 35 did not affect subcellular localization or protein
stability. However, covalent SUMO modification at lysine 35 was required for Maf1
mediated repression of RNA pol III dependent genes. The Maf1K35R mutant failed to
repress pre-tRNA
Leu
transcription compared to Maf1WT. This lysine 35 residue in Maf1
is conserved from yeast to humans. In yeast, Maf1 lysine 35 is an important residue as a
mutation from lysine to glutamic acid renders yeast Maf1 non-functional (Gajda et al.,
2010). The lysine 35 residue in Maf1 resides within the A box and mutation of this
residue blocks intra-molecular interaction between the A and BC box of yeast Maf1. As
there is one SUMO isoform in yeast, it is possible that yeast Maf1 is sumoylated and this
sumoylation might be important in the interaction between the A box and BC box in
Maf1, thereby maintaining Maf1 conformation and function.

SENP1 was identified as one of the desumoylases that reverses covalent SUMO
modification of Maf1. Consistent with this, SENP1 downregulation represses RNA pol
III dependent transcription. Interestingly, while SENP1 is overexpressed in prostate
cancer we have observed a decrease in Maf1 expression in prostate tumor samples. It is
52

possible that the low amounts of Maf1that are expressed in these tumors are rendered
non-functional by increased SENP1 expression and its ability to deconjugate sumoylated
Maf1.

Sumoylation of a target protein is most often associated with transcription repression. It is
thought that sumoylation of a target protein helps it to recruit chromatin remodeling
enzymes and other transcriptional corepressors and helps to assemble a repressive
complex on the target promoter. Previous reports have shown that Maf1 represses
transcription by getting recruited to promoters and displacing RNA pol III from the RNA
pol III dependent gene promoters. Our data demonstrated that sumoylation of Maf1 at
lysine 35 was required for Maf1 interaction with RNA pol III. Consistent with this, ChIP
analysis determined that Maf1K35R was no longer able to be recruited to the tRNA
Leu

gene promoter.

Furthermore, the Maf1K35R mutant also failed to suppress colony growth. Based on our
data, we propose a model wherein, under conditions of increased cellular sumoylation,
Maf1 is covalently modified by SUMO at lysine 35 (Fig. 25). This sumoylation of Maf1
then facilitates Maf1 association with RNA pol III and recruitment to RNA pol III
dependent gene promoters, where it displaces RNA pol III and represses transcription.
Hence, covalent modification of Maf1 by SUMO represses transcription and suppresses
colony growth.

53

Figure 25. Model for covalent SUMOylation of Maf1 regulating Maf1 function. Under
increased cellular sumoylation conditions Maf1 is modified at lysine 35. Sumoylated Maf1
associates with RNA pol III and displaces it from the promoter. Sumoylated Maf1 can therefore
repress RNA pol III dependent transcription and suppress growth.

Alternatively, when the cellular SUMO concentration is low, Maf1 is not sumoylated at
lysine 35, and as a result, it can no longer associate with RNA pol III (Fig. 26). This
prevents Maf1 recruitment to RNA pol III transcribed gene promoters and, therefore
Maf1 can no longer repress transcription or suppress growth. Therefore, sumoylation of
Maf1 at lysine 35 is required for Maf1 mediated repression of RNA pol III dependent
transcription and Maf1-mediated suppression of colony growth.

54

Figure 26. Maf1K35R no longer functions to regulate RNA pol III –dependent
transcription. Under decreased cellular sumoylation conditions Maf1 is not modified at lysine 35
and fails to interact with RNA pol III to displace RNA pol III from the promoter. Maf1K35R can
therefore no longer repress RNA pol III dependent transcription and suppress growth.

5.4 SUMO modification and Maf1 protein expression
Lysine residues can be modified by ubiquitnation, sumoylation, acetylation and
methylation. Sumoylation of a protein at a lysine residue can block ubiquitination of the
protein, thereby increasing the stability of the protein. In our study, Maf1 protein
expression was increased when the cellular sumoylation concentration was increased and
55

Maf1 protein expression was decreased when Ubc9 was downregulated. However, Maf1
mRNA expression remained unchanged indicating that SUMO regulates Maf1 protein
expression. Additionally, the expression of the Maf1K35R mutant is unchanged
compared to Maf1WT. This suggests that covalent SUMO modification at lysine 35 does
not regulate Maf1 expression or protein stability. Proteins are also known to non-
covalently interact with other sumoylated proteins and, in the case of Oct4, this non-
covalent interaction is known to increase Oct4 protein stability (Wei et al., 2007). An
interaction between Maf1 and SUMO2 was identified by Stelzl et al. (Stelzl et al., 2005)
in a genome-wide yeast two hybrid screen. Our data confirmed that Maf1 can non-
covalently interact with some as yet unidentified sumoylated protein(s). By increasing or
decreasing cellular sumoylation concentrations, all proteins regulated by SUMO would
be affected, and it is possible that non-covalent interaction between Maf1 and some
sumoylated protein(s) regulates Maf1 protein expression.

5.5 Maf1 and non-covalent SUMO interaction
Certain transcription repressor proteins that are sumoylated can also non-covalently
interact with other proteins that possess a SUMO interacting motif (SIM). This SIM helps
them to recruit other sumoylated proteins to assemble transcription repressor complexes
on target gene promoters. Since Maf1 does not have any known DNA binding domain it
is possible that Maf1 might be recruited to target promoters in a similar manner. Our data
reveals that Maf1 non-covalently interacts with SUMO and covalent SUMO modification
at lysine 35 is not required for this interaction.
56

We mapped serine 123 on Maf1 as an important residue that affects non-covalent SUMO
interaction. While mutation of Maf1 serine 123 to alanine enhanced non-covalent
interaction with SUMO, mutation of serine 123 to aspartic acid did not enhance non-
covalent SUMO interaction compared to Maf1WT. Serine residues on proteins are
subject to modification by phosphorylation. Muller et. al. (Stehmeier and Muller, 2009)
identified phospho-regulated SIM modules in PML, PMSCL1 and PIAS1, wherein, CK2
phosphorylates serine residues that lie near the hydrophobic core V/I – V/I – X - V/I SIM
motif. This phosphorylation is required for non-covalent interaction of PML, PMSCL1
and PIAS1 with other sumoylated proteins. Since mutation of serine 123 to alanine and
not aspartic acid in Maf1 enhances non-covalent SUMO interactions, it is possible that
serine 123 on Maf1 is phosphorylated. Rather than enhance interaction, phosphorylation
of S123 may block non-covalent interaction between Maf1 and some sumoylated protein.
Consistent with this idea, the Maf1S123D mutant does not display an enhanced non-
covalent interaction with some SUMO compared to Maf1S123A. This would indicate
that Maf1S123 lies within the putative SIM and that phosphorylation blocks interaction
between the Maf1 SIM and SUMO. However, since we observed that mutation of Maf1
at serine 123 to alanine results in a gain of functional interaction between Maf1 and
SUMO instead of a loss of interaction, it is also possible that Maf1S123 does not lie
within the putative Maf1 SIM. In this case, phosphorylation at S123 blocks the
interaction between a Maf1 SIM at another site and SUMO.

57

Bioinformatic predictions place Maf1 serine 123 within a putative phosphorylation
consensus motif (Liu et al., 2012) for Polo-like kinase 1. Interestingly, a recent report
determined that Polo-like kinase 1 stimulates RNA pol III mediated transcription during
interphase (Fairley et al., 2012). It is directly recruited to tRNA and 5S rRNA gene
promoters where it phosphorylates Brf1. It is possible that Polo-like kinase I
phosphorylates Maf1 at serine 123 which causes a decreased association between Maf1
and some sumoylated protein(s). It will be interesting to determine whether Polo-like
kinase I directly phosphorylates Maf1 at serine 123 using in vitro kinase assays and
whether this phosphorylation blocks non-covalent interaction between Maf1 and SUMO.

5.6 Non-covalent SUMO interaction and Maf1 function
Non-covalent interactions between sumoylated proteins and proteins that have SIM’s
facilitate the assembly of transcription complexes on gene promoters. The ability to non-
covalently interact with specific sumoylated transcription factors would offer a
mechanism by which Maf1 could interact with a variety of proteins to get recruited to
target gene promoters. We found that increased interaction of Maf1 with some
sumoylated protein(s) enhanced Maf1 mediated repression of RNA pol III dependent
transcription. This increased interaction also enhanced Maf1’s ability to suppress colony
growth indicating that non-covalent interaction between Maf1 and some sumoylated
protein(s) is required for Maf1 mediated transcription repression and growth suppression.

58

Our results demonstrated that, (1) Maf1S123A expression is lower compared to Maf1
WT, (2) Maf1S123A shows an enhanced ability to interact with some sumoylated
protein(s) and (3) Maf1S123A shows enhanced interaction with RNA pol III. In addition
to this, the Maf1K35R mutant which is no longer covalently modified by SUMO, fails to
associate with RNA pol III, but can still non-covalently interact with some sumoylated
protein(s). These results suggest the following for how Maf1 function and expression is
regulated through sumoylation. There are at least two SUMO driven non-covalent
interactions that regulate Maf1 function and expression. In the first scenario, Maf1
interaction with RNA pol III requires a sumoylation event apart from Maf1 covalent
sumoylation. This is because Maf1S123A shows enhanced interaction with RNA pol III
which is further augmented in the presence of SUMO. It is possible that RNA pol III,
itself is sumoylated. Evidence of yeast RNA polymerase III sumoylation has been shown
in in vitro studies (Panse et al., 2004). Yeast RNA pol III subunits Rpc 160, Rpc 128, Rpc
82, Rpc 40 and Rpc 37 were found to be sumoylated in a proteome-wide approach using
affinity chromatography. The human homologues of these subunits were also found to be
potentially sumoylated in mammalian cells using a proteomic approach (Zhao et al.,
2004). Interestingly, Maf1 has been shown to interact with the human homologue of Rpc
160. Furthermore, crystal structure analyses have shown that yeast Maf1 associates with
Rpc 82, Rpc 34 and Rpc 31 which together form the RNA pol III clamp structure
(Vannini et al., 2010). The association with Maf1 at this structure rearranges the clamp
and prevents binding of RNA pol III to the TBP-Brf1 complex to inhibit transcription.
Hence, we propose that Maf1 associates with sumoylated RNA pol III to inhibit RNA pol
59

III –dependent transcription. This supports the idea that both sumoylation of Maf1 and
sumoylation of RNA pol III are required for their interactions.

In the second scenario, our results demonstrated that SUMO affects expression of Maf1
and covalent SUMO attachment does not impair Maf1 expression. This suggests that the
non-covalent interaction between Maf1 and some sumoylated protein(s) affects Maf1
expression and stability. It is possible that SUMO affects the expression of some protein
that non-covalently interacts with Maf1 and stabilizes Maf1 expression. Alternatively, it
is possible that sumoylation of a Maf1-associated protein is required for its non-covalent
interaction with Maf1 which leads to increased Maf1 protein expression or stability.
Hence, there are distinct non-covalent interactions between Maf1 and some sumoylated
protein(s) that are affected by covalent or non-covalent Maf1 modification by SUMO.

5.7 Maf1 interacting proteins
In addition to repressing RNA pol III transcribed genes our lab demonstrated that Maf1
can also repress certain RNA pol II transcribed genes. While previous studies have shown
that Maf1 can associate with RNA pol III to repress RNA pol III transcription, how Maf1
is recruited to RNA pol II transcribed gene promoters is not known. Furthermore, the
subunits in RNA pol III that Maf1 interacts with are not conserved to RNA pol II
suggesting that the mechanism by which Maf1 represses RNA pol III –mediated
transcription is probably not conserved for RNA pol II –mediated transcriptional
repression by Maf1. Additionally, Maf1 has no known DNA binding domain and is
60

thought to be recruited to gene promoters through protein-protein interactions. Our results
identified a novel interaction between Maf1 and the CDK8 sub-module. In vitro studies
first identified the CDK8 sub-module as a negative regulator of RNA pol II dependent
transcription through its interaction with the mediator. However, in vivo studies have
shown that CDK8 positively regulates transcription. Our results demonstrate that CDK8
positively regulates the transcription of TBP, which is a Maf1 target gene. Additionally,
this is also the first demonstration that CDK8, which is part of the Mediator complex,
positively regulates RNA pol III dependent gene transcription. Consistent with this,
CDK8 and MED13 were recruited to the TBP and tRNA
Leu
gene promoters near the
transcription start site, and they co-occupy these promoters with Maf1. Studies have
shown that CDK8 functions as a potent oncogene and is overexpressed in colon cancer.
This would be consistent with its positive effect on transcription of TBP and RNA pol III
dependent genes whose induction are both known to promote transformation and
tumorigenesis. Based on our preliminary results, we propose that Maf1 is recruited to
these promoters through interaction with the CDK8 sub-module where it might then
displace the CDK8 sub-module from the promoter and repress transcription.

5.8 Summary
In our study, we identified sumoylation as a novel pathway that negatively regulates
RNA polymerase III mediated transcription via Maf1. Maf1 is covalently modified by
SUMO1 and SUMO2 at lysine 35 and this sumoylation is required for Maf1 mediated
repression of RNA pol III dependent transcription and suppression of colony growth.
61

This is the first indication of a post translational modification other than phosphorylation
regulating mammalian Maf1 function. Furthermore, loss of this covalent modification led
to a loss of association between Maf1 and RNA pol III and resulted in the inability of
Maf1 to be recruited to RNA pol III transcribed gene promoters. In addition to covalent
modification, Maf1 also non-covalently interacts with some sumoylated protein(s). Serine
123 on Maf1 was mapped as an important residue that regulated non-covalent Maf1
interaction with some SUMO modified protein. Mutation of Maf1 serine 123 to alanine
enhanced Maf1 interaction with RNA polymerase III. This Maf1S123A mutant also
displayed an enhanced ability to repress RNA pol III mediated transcription and suppress
colony growth.

While this study mainly focused on covalent and non-covalent SUMO interactions that
regulate Maf1 mediated repression of RNA pol III dependent genes, mammalian Maf1
can also repress certain RNA pol II dependent genes like TBP and egr-1. In this study, we
also identified Maf1 interacting proteins CDK8 and MED13 that co-localize with Maf1
on the TBP promoter. This opens up new avenues to understand how Maf1 functions to
repress RNA pol II dependent genes. It will be interesting to study whether Maf1 is
recruited to RNA pol II dependent promoters by the CDK8 sub-module. It will also be
interesting to understand whether covalent and non-covalent SUMO interactions with
Maf1 play a role in regulating Maf1 mediated repression of RNA pol II dependent genes.
It is possible that one or more components of the CDK8 sub-module might be sumoylated
or might possess a SIM through which Maf1 might be recruited to promoters to repress
62

RNA pol II dependent transcription. This study is also the first indication of the CDK8
sub-module regulating RNA pol III dependent transcription. Since CDK8 is known to
function as a potent oncogene and Maf1 functions as a potential tumor suppressor it will
be important to understand the relationship between Maf1 and CDK8 to understand how
they control transformation and tumorigenesis.

63

Chapter 6: Materials & Methods
Cell lines and plasmids
U87 glioblastoma cells were purchased from the American Type Culture Collection
(ATCC). The COS7 and 293T fibroblast cell lines were kindly provided by Michael
Stallcup (University of Southern California). All three cell lines were cultured in high
glucose Dulbecco’s Modified Eagle’s Media (DMEM) (Mediatech) supplemented with
10% Fetal Bovine Serum (FBS), Penicillin (500 units), Streptomycin (500 ug) and
Glutamine (500 units).

The pcDNA3 – Maf1HA plasmid is previously described (Johnson et al., 2007). The
myc-SUMO1ρ and myc-SUMO2ρ plasmids were kindly provided by David Ann (City of
Hope, Duarte, California).

Site-directed mutagenesis
Maf1 mutants were generated using the Quikchange Lightening site-directed mutagenesis
kit (Stratagene) as per the manufacturer’s protocol. Primers for mutageneis were designed
using the Stratagene Quikchange Primer design software. After transformation plasmids
were isolated using the Promega Mini-prep kit (Promega). Mutants were screened by
performing an EcoRI digestion overnight. Mutagenesis was confirmed by sequencing at
the USC/Norris Microchemical Core.

64

Transient Transfections
COS7 and 293T cells were transiently transfected using the lipid based F1 transfection
reagent (Targeting Systems). 5 x 10
5
cells were seeded on 100 mm
2
plates in 10 ml of
10% FBS/DMEM. The cells were transfected 24 hours later with plasmids, as indicated
in the figure legends, and pSK for a total of 10 ug of DNA per plate. 2ul of F1
transfection reagent for every 1ug of DNA was mixed with plain DMEM and incubated
at 37°C for 25 mins. 1ml of transfection complex was added to each plate. 48 hours post
transfection cells were harvested for total protein or RNA and QPCR or immunoblot
analyses were carried out.

3 x 10
5
U87 cells were seeded in 60mm
2
dishes in 10% FBS/DMEM without antibiotics
and transfected using Lipofectamine 2000 (Invitrogen). 2ul of Lipofectamine 2000 for
every 1ug of DNA was mixed in plain DMEM and incubated at room temperature for 5
mins. The lipofectamine 2000/DMEM mix was then added to the plasmid DNA/DMEM
mix and incubated at room temperature for 20 mins. The transfection complexes were
added to the cells for 4 hours at 37°C after which the transfection mix was removed and
DMEM with 10% FBS was added to the plates. 24 hours later cells were rinsed with PBS
(Mediatech) and serum starved with 0.5% FBS in DMEM media. 16 hours post serum
starvation RNA was isolated from the cells.

65

Transient transfection of siRNA
COS7 cells were transfected with mmRNA or Ubc9 siRNA using the Targfect siRNA kit
(Targeting Systems). The siRNA was added to plain DMEM and mixed followed by
addition of Solution A and then Solution B in a ratio of 1:2. The transfection complexes
were incubated at 37°C for 25 mins. The cells were washed twice with plain DMEM
before addition of the transfection complexes. 2 hours after addition DMEM with 10%
FBS was added to the plates. The next morning, the media was removed and fresh
DMEM with 10% FBS was added and RNA and protein were isolated 24 hours later. The
siRNAs were obtained from the USC/Norris Cancer Center Microchemical Core facility.
The sequences of the siRNAs are as follows:
mmRNA Sense strand – 5’ – GCUUCCCCCUCGUAAUUACUU – 3’
mmRNA Anti-sense strand – 5’ – GUAAUUACGAGGGGGAAGCUU – 3’
Ubc 9 siRNA Sense strand – 5’- CAA AAA AUC CCG AUG GCA CTT – 3’
Ubc 9 siRNA Anti-sense strand – 5’ - GUG CCA UCG GGA UUU UUU GTT – 3’

U87 cells were transfected with either CDK8 or control siRNA using Lipofectamine 2000
(Invitrogen) same as above. The sequences of the CDK8 siRNAs are as follows:
CDK8 siRNA Sense strand – 5'- AUAUAAUAGUGACUUCACCAUUCCCTT – 3’
CDK8 siRNA Anti-sense strand – 5'- GGGAAUGGUGAAGUCACUAUUAUAUTT - 3'

66

RT-PCR Analysis
Total RNA was isolated from cells using RNA-STAT 60 (Teltest) according to the
manufacturer’s protocol. 10ug of RNA was subjected to DNase digestion using the
DNase Turbo kit (Ambion) according to the protocol provided. 0.5ug – 1ug of RNA was
converted to cDNA using the Superscript First Strand Synthesis kit (Invitrogen). Briefly,
the RNA was incubated with 100ng random hexamers and 10mM dNTPs at 65°C for 5
mins in a total volume of 10ul. A master mix, containing 1X RT Buffer, 50mM MgCl
2
,
0.1M DTT, RNase OUT and SuperScript III, was added to a final volume of 20ul and
incubated at 25°C for 10 mins followed by incubation at 50°C and then 85°C for 15 mins.
The cDNA was then diluted 1:100 and used for quantitative RT-PCR (QPCR). Real time
PCR was performed on the Mx3000P QPCR System (Brilliant SYBR Green QPCR
Master mix, Stratagene) with primers sets for Maf1, pre-tRNA
Leu
, pre-tRNA
i
met
, 7SL,
TBP, CDK8 and GAPDH. Relative amounts of transcripts were quantified by the
comparative threshold cycle method (ΔΔCt) with GAPDH as the endogenous reference
control. The primer sequences are as follows:
Maf1 Forward 5’ – GTG GAG ACT GGA GAT GCC CA – 3’
Reverse 5’ – CTG GGT TAT AGC TGT AGA TGT CAC A – 3’
Pre-tRNA
Leu
Forward 5’ –
GTC
AGG
ATG
GCC
GAG
TGG
TCT
AAG
-‐3'

Reverse 5’ –
CCA
CGC
CTC
CAT
ACG
GAG
AAC
CAG
AAG
ACC
C
-‐3'

Pre-tRNA
i
met
Forward 5’ – CTG GGC CCA TAA CCC AGA G – 3’
Reverse 5’ – TGG TAG CAG AGG ATG GTT TC – 3’
67

7SL Forward 5’ – GTG
TCC
GCA
CTA
AGT
TCG
G
-‐
3'

Reverse 5’ –
TAT
TCA
CAG
GCG
CGA
TCC
-‐3’
GAPDH Forward 5’ – GAT
GGG
TGT
GAA
CCA
CGA
GAA-‐3'

Reverse 5’ –
GGG
CCA
TCC
ACA
GTC
TTC
TG-‐3'

CDK8 Forward 5’ - GGG ATC TCT ATG TCG GCA TGT AG – 3’
Reverse 5’ - AAA TGA CGT TTG GAT GCT TAA GC – 3’

Immunoprecipitation and Immunoblot Analysis
Total protein was isolated from cells in 200ul – 300ul of cell lysis buffer
[20 mM Tris
(pH 7.5), 150 mM NaCl, 1mM EDTA, 1mM EGTA, 0.1% Triton X-100, 2.5 mM sodium
pyrophosphate, 1mM β-glycerol phosphate, 1 mM Na
3
VO
4
, Halt phosphatase inhibitor
(Pierce), and Protease Inhibitor Cocktail Set III (Calbiochem)].
Cells were incubated in
cell lysis buffer on ice for 15 mins after which they were scrapped and sonicated for 15
seconds. After sonication they were centrifuged for 15 mins at 10,000 g. The supernatant
was then collected and the protein concentration was determined by the Bradford method
using the Bio-Rad Protein Assay Reagent. 100ug – 200ug of protein was then subjected
to sodium dodecyl sulphate – polyacrylamide gel electrophoresis (SDS-PAGE). To
separate out phosphorylated Maf1 a phostag acrylamide (Wako Chemicals) was used to
run a SDS-PAGE gel. Proteins were then transferred onto a Hybond nitrocellulose
membrane (GE-Healthcare, Amersham). The membrane was then blocked with 5% milk
in TBS for 45 mins followed by incubation in primary antibody overnight. After
68

overnight incubation, the bound primary antibody was visualized with either biotinylated
IgG secondary antibody and standard Vectastatin ABC kit (Vector Laboratories) or
horseradish peroxidase-conjugated secondary antibody (IgG) followed by enhanced
chemiluminescence (ECL) (Perkin Elmer) or SuperSignal West Pico Chemiluminescent
Substrate (Pierce). The membranes were exposed to Denville film.
The resultant films
were scanned and quantified using UN-Scan-It software (Silk Scientific).

The primary antibodies used are as follows: goat polyclonal CDK8 (Santa Cruz
Biotechnology), rabbit polyclonal Mediator 13 (Bethyl Laboratories), Mediator 8 (Santa
Cruz Biotechnology), rabbit IgG (Bethyl Laboratories), mouse polyclonal myc (Santa
Cruz Biotechnology), mouse polyclonal actin (Millipore), rat monoclonal HA (Roche),
rabbit polyclonal Maf1 (Abcam), RNA polymerase III C39 (Santa Cruz Biotechnology)
and rabbit polyclonal Ubc9 (Cell Signaling).

For immunoprecipitation reactions 500ug – 1000ug of total protein was incubated
overnight at 4°C with 2ug – 4ug of antibody as indicated. 25ul – 30ul of Protein A/G
Agarose Beads (Santa Cruz Biotechnology) were then added and incubated for an
additional 3 hours at 4°C. The beads were then washed 3 times with cold 1:1 TBS: Cell
lysis buffer. The proteins were eluted from the beads with 40 ul of 2X Sample Buffer
followed by boiling for 10 mins and subjected to SDS-PAGE. Immunoblot analysis was
performed as described before.

69

Colony Suppression Assay
U87 cells were transfected with HA –tagged Maf1WT or Maf1K35R expression vectors
and a puromycin resistant vector. Cells transfected with puromycin resistant vector and
no Maf1 were used as controls. 48 hours post-transfection colonies were selected by
treatment with 0.6ug/ml – 0.8ug/ml puromycin. Media with fresh puromycin was
changed every 48 hours. Three weeks post selection colonies were fixed and stained with
crystal violet. The colonies were counted and converted to % Survival and graphed by
setting the vector alone to 100%.

Immunofluorescence
U87 cells were plated on coverslips coated with poly-L-Lysine and collagen. The cells
were then transfected with HA –tagged Maf1WT or Maf1K35R using Lipofectamine
2000. 24 hours post-transfection cells were fixed in 4% paraformaldehyde, blocked in 1%
BSA in PBS for 1 hour at room temperature and incubated with primary HA antibody
(Covance) at a dilution of 1:1000 for 30 mins. Following incubation with primary
antibody cells were washed with 1X PBS and then incubated with FITC –conjugated
secondary antibody (Sigma) at a dilution of 1:256 for 30 mins. Cells were then washed
with 1X PBS and mounted in mounting media with propidium iodide (Vector Labs).
They were then allowed to dry overnight followed by imaging on a Zeiss LSM 510
confocal laser scanning microscope.

Chromatin Immunoprecipitation
70

Chromatin was isolated from cells plated on 145mm2 plates at 90% confluency. Cross-
linking was carried out by incubating with 1% formaldehyde at room temperature for 10
mins. The cross-linking was stopped by incubating with 0.125M glycine for 5 mins. The
cells were then washed twice with cold PBS, scrapped and collected by centrifugation at
4000 rpm for 5 mins. The cell pellet was resuspended in hypotonic lysis buffer (10 mM
Hepes-KOH pH 7.8, 10 mM KCl, 1.5 mM MgCl
2
)
supplemented with Protease Inhibitor
Cocktail III (Calbiochem). After incubation on ice for 10 mins and centrifugation at 5000
rpm for 5 mins, the nuclear pellet was resuspended in nuclei lysis buffer (50 mM Tris-Cl
pH 8.1, 10 mM EDTA, and 1% SDS) with Protease Inhibitor Cocktail III. The suspension
was incubated on ice for 10 mins followed by sonication to fragment the chromatin and
centrifugation at 13000 rpm for 10 mins. The supernatant chromatin was then collected.
20 ul of chromatin was stored separately as input.

For antibody pull down of the cross-linked chromatin, 900 ul of dilution buffer, 25 ul of
Protein A/G Plus beads and 10 ul of Protease Inhibitor Cocktail III was added to 100 ul
of chromatin and pre-cleared for 1 hour on a rotating platform at 4C. After centrifugation
at 14000rpm for 10 mins the supernatant was collected and incubated with either HA,
TBP, CDK8, MED13, RPC39, Maf1 or IgG antibody respectively overnight. Next
morning, 40 ul of Protein A/G Plus beads were added and incubated for 2 hours at 4C.
The beads were centrifuged at 1000 rpm for 1 min and then washed with Wash Buffer 1,
Wash Buffer 2 and Wash Buffer 3 once. The beads were then washed twice with TE
Buffer. 150 ul of Elution Buffer was added to the beads and vortexed at setting 3 for 15
71

mins followed by centrifugation at 14000 rpm for 5 mins. The supernatant was
transferred to a fresh tube and the elution was then repeated once more. Protein-DNA
cross links were then reversed by incubating with 19ul 5M sodium chloride at a final
concentration of 0.125M and 1ul RNAse A overnight at 65C. Next morning, the proteins
were digested by treating with Proteinase K for 2 hours at 42C. The DNA was then
isolated using phenol: chloroform extraction. The DNA was then amplified by QPCR
using primers for the tRNA
Leu
gene promoter.

72

References
Bawa-Khalfe, T., Cheng, J., Lin, S.H., Ittmann, M.M., and Yeh, E.T. (2010). SENP1
induces prostatic intraepithelial neoplasia through multiple mechanisms. J Biol Chem
285, 25859-25866.
Bawa-Khalfe, T., and Yeh, E.T. (2010). SUMO Losing Balance: SUMO Proteases
Disrupt SUMO Homeostasis to Facilitate Cancer Development and Progression. Genes
Cancer 1, 748-752.
Belakavadi, M., and Fondell, J.D. (2010). Cyclin-dependent kinase 8 positively
cooperates with Mediator to promote thyroid hormone receptor-dependent transcriptional
activation. Mol Cell Biol 30, 2437-2448.
Boguta, M., Czerska, K., and Zoladek, T. (1997). Mutation in a new gene MAF1 affects
tRNA suppressor efficiency in Saccharomyces cerevisiae. Gene 185, 291-296.
Borden, K.L. (2002). Pondering the promyelocytic leukemia protein (PML) puzzle:
possible functions for PML nuclear bodies. Mol Cell Biol 22, 5259-5269.
Chen, W., Böcker, W., Brosius, J., and Tiedge, H. (1997). Expression of neural BC200
RNA in human tumours. J Pathol 183, 345-351.
Cheng, J., Bawa, T., Lee, P., Gong, L., and Yeh, E. (2006). Role of desumoylation in the
development of prostate cancer. Neoplasia 8, 667-676.
Cheng, J., Kang, X., Zhang, S., and Yeh, E.T. (2007). SUMO-specific protease 1 is
essential for stabilization of HIF1alpha during hypoxia. Cell 131, 584-595.
Chiu, S.Y., Asai, N., Costantini, F., and Hsu, W. (2008). SUMO-specific protease 2 is
essential for modulating p53-Mdm2 in development of trophoblast stem cell niches and
lineages. PLoS Biol 6, e310.
Crighton, D., Woiwode, A., Zhang, C., Mandavia, N., Morton, J.P., Warnock, L.J.,
Milner, J., White, R.J., and Johnson, D.L. (2003). p53 represses RNA polymerase III
transcription by targeting TBP and inhibiting promoter occupancy by TFIIIB. EMBO J
22, 2810-2820.
73

Desai, N., Lee, J., Upadhya, R., Chu, Y., Moir, R., and Willis, I. (2005). Two steps in
Maf1-dependent repression of transcription by RNA polymerase III. J Biol Chem 280,
6455-6462.
Ding, N., Zhou, H., Esteve, P.O., Chin, H.G., Kim, S., Xu, X., Joseph, S.M., Friez, M.J.,
Schwartz, C.E., Pradhan, S., et al. (2008). Mediator links epigenetic silencing of neuronal
gene expression with x-linked mental retardation. Mol Cell 31, 347-359.
Donner, A.J., Ebmeier, C.C., Taatjes, D.J., and Espinosa, J.M. (2010). CDK8 is a positive
regulator of transcriptional elongation within the serum response network. Nat Struct Mol
Biol 17, 194-201.
Donner, A.J., Szostek, S., Hoover, J.M., and Espinosa, J.M. (2007). CDK8 is a stimulus-
specific positive coregulator of p53 target genes. Mol Cell 27, 121-133.
Dorval, V., and Fraser, P.E. (2006). Small ubiquitin-like modifier (SUMO) modification
of natively unfolded proteins tau and alpha-synuclein. J Biol Chem 281, 9919-9924.
Elmlund, H., Baraznenok, V., Lindahl, M., Samuelsen, C.O., Koeck, P.J., Holmberg, S.,
Hebert, H., and Gustafsson, C.M. (2006). The cyclin-dependent kinase 8 module
sterically blocks Mediator interactions with RNA polymerase II. Proc Natl Acad Sci U S
A 103, 15788-15793.
Fairley, J.A., Mitchell, L.E., Berg, T., Kenneth, N.S., von Schubert, C., Sillje, H.H.,
Medema, R.H., Nigg, E.A., and White, R.J. (2012). Direct regulation of tRNA and 5S
rRNA gene transcription by Polo-like kinase 1. Mol Cell 45, 541-552.
Felton-Edkins, Z.A., Fairley, J.A., Graham, E.L., Johnston, I.M., White, R.J., and Scott,
P.H. (2003). The mitogen-activated protein (MAP) kinase ERK induces tRNA synthesis
by phosphorylating TFIIIB. EMBO J 22, 2422-2432.
Firestein, R., Bass, A.J., Kim, S.Y., Dunn, I.F., Silver, S.J., Guney, I., Freed, E., Ligon,
A.H., Vena, N., Ogino, S., et al. (2008). CDK8 is a colorectal cancer oncogene that
regulates beta-catenin activity. Nature 455, 547-551.
74

Fondell, J.D., Ge, H., and Roeder, R.G. (1996). Ligand induction of a transcriptionally
active thyroid hormone receptor coactivator complex. Proc Natl Acad Sci U S A 93,
8329-8333.
Gajda, A., Towpik, J., Steuerwald, U., Muller, C.W., Lefebvre, O., and Boguta, M.
(2010). Full repression of RNA polymerase III transcription requires interaction between
two domains of its negative regulator Maf1. J Biol Chem 285, 35719-35727.
Geiss-Friedlander, R., and Melchior, F. (2007). Concepts in sumoylation: a decade on.
Nat Rev Mol Cell Biol 8, 947-956.
Gomez-Roman, N., Grandori, C., Eisenman, R.N., and White, R.J. (2003). Direct
activation of RNA polymerase III transcription by c-Myc. Nature 421, 290-294.
Goodfellow, S., Graham, E., Kantidakis, T., Marshall, L., Coppins, B., Oficjalska-Pham,
D., Gérard, M., Lefebvre, O., and White, R. (2008). Regulation of RNA polymerase III
transcription by Maf1 in mammalian cells. J Mol Biol 378, 481-491.
Han, Y., Huang, C., Sun, X., Xiang, B., Wang, M., Yeh, E.T., Chen, Y., Li, H., Shi, G.,
Cang, H., et al. (2010). SENP3-mediated de-conjugation of SUMO2/3 from
promyelocytic leukemia is correlated with accelerated cell proliferation under mild
oxidative stress. J Biol Chem 285, 12906-12915.
Hannich, J., Lewis, A., Kroetz, M., Li, S., Heide, H., Emili, A., and Hochstrasser, M.
(2005). Defining the SUMO-modified proteome by multiple approaches in
Saccharomyces cerevisiae. J Biol Chem 280, 4102-4110.
Hay, R. (2005). SUMO: a history of modification. Mol Cell 18, 1-12.
Hecker, C., Rabiller, M., Haglund, K., Bayer, P., and Dikic, I. (2006). Specification of
SUMO1- and SUMO2-interacting motifs. J Biol Chem 281, 16117-16127.
Hietakangas, V., Anckar, J., Blomster, H., Fujimoto, M., Palvimo, J., Nakai, A., and
Sistonen, L. (2006). PDSM, a motif for phosphorylation-dependent SUMO modification.
Proc Natl Acad Sci U S A 103, 45-50.
75

Jacques, C., Baris, O., Prunier-Mirebeau, D., Savagner, F., Rodien, P., Rohmer, V.,
Franc, B., Guyetant, S., Malthiery, Y., and Reynier, P. (2005). Two-step differential
expression analysis reveals a new set of genes involved in thyroid oncocytic tumors. J
Clin Endocrinol Metab 90, 2314-2320.
Johnson, E.S., and Gupta, A.A. (2001). An E3-like factor that promotes SUMO
conjugation to the yeast septins. Cell 106, 735-744.
Johnson, S., Dubeau, L., and Johnson, D. (2008a). Enhanced RNA polymerase III-
dependent transcription is required for oncogenic transformation. J Biol Chem 283,
19184-19191.
Johnson, S.A., Dubeau, L., and Johnson, D.L. (2008b). Enhanced RNA polymerase III-
dependent transcription is required for oncogenic transformation. J Biol Chem 283,
19184-19191.
Johnson, S.S., Zhang, C., Fromm, J., Willis, I.M., and Johnson, D.L. (2007). Mammalian
Maf1 is a negative regulator of transcription by all three nuclear RNA polymerases. Mol
Cell 26, 367-379.
Joseph, J., Tan, S., Karpova, T., McNally, J., and Dasso, M. (2002). SUMO-1 targets
RanGAP1 to kinetochores and mitotic spindles. J Cell Biol 156, 595-602.
Kagey, M.H., Melhuish, T.A., and Wotton, D. (2003). The polycomb protein Pc2 is a
SUMO E3. Cell 113, 127-137.
Kang, X., Qi, Y., Zuo, Y., Wang, Q., Zou, Y., Schwartz, R.J., Cheng, J., and Yeh, E.T.
(2010). SUMO-specific protease 2 is essential for suppression of polycomb group
protein-mediated gene silencing during embryonic development. Mol Cell 38, 191-201.
Kantidakis, T., Ramsbottom, B.A., Birch, J.L., Dowding, S.N., and White, R.J. (2010).
mTOR associates with TFIIIC, is found at tRNA and 5S rRNA genes, and targets their
repressor Maf1. Proc Natl Acad Sci U S A 107, 11823-11828.
Knuesel, M., Meyer, K., Bernecky, C., and Taatjes, D. (2009). The human CDK8
subcomplex is a molecular switch that controls Mediator coactivator function. Genes Dev
23, 439-451.
76

Larminie, C.G., Cairns, C.A., Mital, R., Martin, K., Kouzarides, T., Jackson, S.P., and
White, R.J. (1997). Mechanistic analysis of RNA polymerase III regulation by the
retinoblastoma protein. EMBO J 16, 2061-2071.
Leal, T., Andrieux, J., Duban-Bedu, B., Bouquillon, S., Breviere, G.M., and Delobel, B.
(2009). Array-CGH detection of a de novo 0.8Mb deletion in 19q13.32 associated with
mental retardation, cardiac malformation, cleft lip and palate, hearing loss and multiple
dysmorphic features. Eur J Med Genet 52, 62-66.
Lee, J.S., and Thorgeirsson, S.S. (2004). Genome-scale profiling of gene expression in
hepatocellular carcinoma: classification, survival prediction, and identification of
therapeutic targets. Gastroenterology 127, S51-55.
Lin, D., Huang, Y., Jeng, J., Kuo, H., Chang, C., Chao, T., Ho, C., Chen, Y., Lin, T.,
Fang, H., et al. (2006). Role of SUMO-interacting motif in Daxx SUMO modification,
subnuclear localization, and repression of sumoylated transcription factors. Mol Cell 24,
341-354.
Lin, X., Sun, B., Liang, M., Liang, Y., Gast, A., Hildebrand, J., Brunicardi, F., Melchior,
F., and Feng, X. (2003). Opposed regulation of corepressor CtBP by SUMOylation and
PDZ binding. Mol Cell 11, 1389-1396.
Liu, Z., Ren, J., Cao, J., He, J., Yao, X., Jin, C., and Xue, Y. (2012). Systematic analysis
of the Plk-mediated phosphoregulation in eukaryotes. Brief Bioinform.
Malhotra, R., and Mason, P.K. (2009). Lamin A/C deficiency as a cause of familial
dilated cardiomyopathy. Curr Opin Cardiol 24, 203-208.
McDoniels-Silvers, A.L., Nimri, C.F., Stoner, G.D., Lubet, R.A., and You, M. (2002).
Differential gene expression in human lung adenocarcinomas and squamous cell
carcinomas. Clin Cancer Res 8, 1127-1138.
Michels, A.A., Robitaille, A.M., Buczynski-Ruchonnet, D., Hodroj, W., Reina, J.H.,
Hall, M.N., and Hernandez, N. (2010). mTORC1 directly phosphorylates and regulates
human MAF1. Mol Cell Biol 30, 3749-3757.
77

Minty, A., Dumont, X., Kaghad, M., and Caput, D. (2000). Covalent modification of
p73alpha by SUMO-1. Two-hybrid screening with p73 identifies novel SUMO-1-
interacting proteins and a SUMO-1 interaction motif. J Biol Chem 275, 36316-36323.
Mo, Y.Y., Yu, Y., Theodosiou, E., Ee, P.L., and Beck, W.T. (2005). A role for Ubc9 in
tumorigenesis. Oncogene 24, 2677-2683.
Moir, R.D., Lee, J., Haeusler, R.A., Desai, N., Engelke, D.R., and Willis, I.M. (2006).
Protein kinase A regulates RNA polymerase III transcription through the nuclear
localization of Maf1. Proc Natl Acad Sci U S A 103, 15044-15049.
Mooney, S.M., Grande, J.P., Salisbury, J.L., and Janknecht, R. (2010). Sumoylation of
p68 and p72 RNA helicases affects protein stability and transactivation potential.
Biochemistry 49, 1-10.
Myers, L.C., and Kornberg, R.D. (2000). Mediator of transcriptional regulation. Annu
Rev Biochem 69, 729-749.
Nacerddine, K., Lehembre, F., Bhaumik, M., Artus, J., Cohen-Tannoudji, M., Babinet,
C., Pandolfi, P.P., and Dejean, A. (2005). The SUMO pathway is essential for nuclear
integrity and chromosome segregation in mice. Dev Cell 9, 769-779.
Oficjalska-Pham, D., Harismendy, O., Smagowicz, W., Gonzalez de Peredo, A., Boguta,
M., Sentenac, A., and Lefebvre, O. (2006). General repression of RNA polymerase III
transcription is triggered by protein phosphatase type 2A-mediated dephosphorylation of
Maf1. Mol Cell 22, 623-632.
Panse, V.G., Hardeland, U., Werner, T., Kuster, B., and Hurt, E. (2004). A proteome-
wide approach identifies sumoylated substrate proteins in yeast. J Biol Chem 279, 41346-
41351.
Park, J., Kim, K., Lee, E., Seo, Y., Lim, S., Park, K., Rho, S., Lee, S., and Lee, J. (2007).
Elevated level of SUMOylated IRF-1 in tumor cells interferes with IRF-1-mediated
apoptosis. Proc Natl Acad Sci U S A 104, 17028-17033.
Paule, M.R., and White, R.J. (2000). Survey and summary: transcription by RNA
polymerases I and III. Nucleic Acids Res 28, 1283-1298.
78

Pichler, A., Gast, A., Seeler, J.S., Dejean, A., and Melchior, F. (2002). The nucleoporin
RanBP2 has SUMO1 E3 ligase activity. Cell 108, 109-120.
Reina, J., Azzouz, T., and Hernandez, N. (2006). Maf1, a new player in the regulation of
human RNA polymerase III transcription. PLoS ONE 1, e134.
Riley, B.E., Zoghbi, H.Y., and Orr, H.T. (2005). SUMOylation of the polyglutamine
repeat protein, ataxin-1, is dependent on a functional nuclear localization signal. J Biol
Chem 280, 21942-21948.
Rodriguez, M., Dargemont, C., and Hay, R. (2001). SUMO-1 conjugation in vivo
requires both a consensus modification motif and nuclear targeting. J Biol Chem 276,
12654-12659.
Rollins, J., Veras, I., Cabarcas, S., Willis, I., and Schramm, L. (2007). Human Maf1
negatively regulates RNA polymerase III transcription via the TFIIB family members
Brf1 and Brf2. Int J Biol Sci 3, 292-302.
Sarge, K.D., and Park-Sarge, O.K. (2009). Sumoylation and human disease pathogenesis.
Trends Biochem Sci 34, 200-205.
Schwartz, L.B., Sklar, V.E., Jaehning, J.A., Weinmann, R., and Roeder, R.G. (1974).
Isolation and partial characterization of the multiple forms of deoxyribonucleic acid-
dependent ribonucleic acid polymerase in the mouse myeloma, MOPC 315. J Biol Chem
249, 5889-5897.
Shen, T.H., Lin, H.K., Scaglioni, P.P., Yung, T.M., and Pandolfi, P.P. (2006). The
mechanisms of PML-nuclear body formation. Mol Cell 24, 331-339.
Shinbo, Y., Niki, T., Taira, T., Ooe, H., Takahashi-Niki, K., Maita, C., Seino, C., Iguchi-
Ariga, S.M., and Ariga, H. (2006). Proper SUMO-1 conjugation is essential to DJ-1 to
exert its full activities. Cell Death Differ 13, 96-108.
Shor, B., Wu, J., Shakey, Q., Toral-Barza, L., Shi, C., Follettie, M., and Yu, K. (2010).
Requirement of the mTOR kinase for the regulation of Maf1 phosphorylation and control
of RNA polymerase III-dependent transcription in cancer cells. J Biol Chem 285, 15380-
15392.
79

Song, J., Durrin, L., Wilkinson, T., Krontiris, T., and Chen, Y. (2004). Identification of a
SUMO-binding motif that recognizes SUMO-modified proteins. Proc Natl Acad Sci U S
A 101, 14373-14378.
Steffan, J.S., Agrawal, N., Pallos, J., Rockabrand, E., Trotman, L.C., Slepko, N., Illes, K.,
Lukacsovich, T., Zhu, Y.Z., Cattaneo, E., et al. (2004). SUMO modification of
Huntingtin and Huntington's disease pathology. Science 304, 100-104.
Stehmeier, P., and Muller, S. (2009). Phospho-regulated SUMO interaction modules
connect the SUMO system to CK2 signaling. Mol Cell 33, 400-409.
Stelzl, U., Worm, U., Lalowski, M., Haenig, C., Brembeck, F., Goehler, H., Stroedicke,
M., Zenkner, M., Schoenherr, A., Koeppen, S., et al. (2005). A human protein-protein
interaction network: a resource for annotating the proteome. Cell 122, 957-968.
Taatjes, D.J. (2010). The human Mediator complex: a versatile, genome-wide regulator
of transcription. Trends Biochem Sci 35, 315-322.
Takahashi, Y., Kahyo, T., Toh, E.A., Yasuda, H., and Kikuchi, Y. (2001a). Yeast
Ull1/Siz1 is a novel SUMO1/Smt3 ligase for septin components and functions as an
adaptor between conjugating enzyme and substrates. J Biol Chem 276, 48973-48977.
Takahashi, Y., Toh-e, A., and Kikuchi, Y. (2001b). A novel factor required for the
SUMO1/Smt3 conjugation of yeast septins. Gene 275, 223-231.
Upadhya, R., Lee, J., and Willis, I. (2002). Maf1 is an essential mediator of diverse
signals that repress RNA polymerase III transcription. Mol Cell 10, 1489-1494.
Vannini, A., Ringel, R., Kusser, A.G., Berninghausen, O., Kassavetis, G.A., and Cramer,
P. (2010). Molecular basis of RNA polymerase III transcription repression by Maf1. Cell
143, 59-70.
Vethantham, V., Rao, N., and Manley, J.L. (2008). Sumoylation regulates multiple
aspects of mammalian poly(A) polymerase function. Genes Dev 22, 499-511.
80

Wang, L., and Banerjee, S. (2004). Differential PIAS3 expression in human malignancy.
Oncol Rep 11, 1319-1324.
Wang, Q., Xia, N., Li, T., Xu, Y., Zou, Y., Zuo, Y., Fan, Q., Bawa-Khalfe, T., Yeh, E.T.,
and Cheng, J. (2012). SUMO-specific protease 1 promotes prostate cancer progression
and metastasis. Oncogene.
Wang, Z.Q., Zhou, H.Y., and Chen, S.D. (2006). The role of DJ-1 in the pathogenesis of
Parkinson's disease. Neurosci Bull 22, 232-234.
Wei, F., Scholer, H.R., and Atchison, M.L. (2007). Sumoylation of Oct4 enhances its
stability, DNA binding, and transactivation. J Biol Chem 282, 21551-21560.
White, R.J. (2004). RNA polymerase III transcription and cancer. Oncogene 23, 3208-
3216.
Willis, I., Desai, N., and Upadhya, R. (2004). Signaling repression of transcription by
RNA polymerase III in yeast. Prog Nucleic Acid Res Mol Biol 77, 323-353.
Winter, A., Sourvinos, G., Allison, S., Tosh, K., Scott, P., Spandidos, D., and White, R.
(2000). RNA polymerase III transcription factor TFIIIC2 is overexpressed in ovarian
tumors. Proc Natl Acad Sci U S A 97, 12619-12624.
Woiwode, A., Johnson, S.A., Zhong, S., Zhang, C., Roeder, R.G., Teichmann, M., and
Johnson, D.L. (2008). PTEN represses RNA polymerase III-dependent transcription by
targeting the TFIIIB complex. Mol Cell Biol 28, 4204-4214.
Wu, F., Zhu, S., Ding, Y., Beck, W.T., and Mo, Y.Y. (2009). MicroRNA-mediated
regulation of Ubc9 expression in cancer cells. Clin Cancer Res 15, 1550-1557.
Zhang, H., Saitoh, H., and Matunis, M. (2002). Enzymes of the SUMO modification
pathway localize to filaments of the nuclear pore complex. Mol Cell Biol 22, 6498-6508.
Zhang, Y.Q., and Sarge, K.D. (2008a). Sumoylation of amyloid precursor protein
negatively regulates Abeta aggregate levels. Biochem Biophys Res Commun 374, 673-
678.
81

Zhang, Y.Q., and Sarge, K.D. (2008b). Sumoylation regulates lamin A function and is
lost in lamin A mutants associated with familial cardiomyopathies. J Cell Biol 182, 35-
39.
Zhao, Y., Kwon, S.W., Anselmo, A., Kaur, K., and White, M.A. (2004). Broad spectrum
identification of cellular small ubiquitin-related modifier (SUMO) substrate proteins. J
Biol Chem 279, 20999-21002.
Zhong, S., Zhang, C., and Johnson, D.L. (2004). Epidermal growth factor enhances
cellular TATA binding protein levels and induces RNA polymerase I- and III-dependent
gene activity. Mol Cell Biol 24, 5119-5129.

Abstract (if available)

Abstract Our studies have identified a role for sumoylation in mediating the ability of Maf1 to repress RNA pol III-dependent transcription. As increased expression of either SUMO1 or SUMO2 represses the expression of tRNA and U6 RNA genes, we further examined whether the transcriptional repressor, Maf1, might be regulated by sumoylation. We find that Maf1 is covalently modified by both SUMO1 and SUMO2. To identify the specific sumoylation site, each of the eleven lysine residues within Maf1 was mutated to arginine revealing that Maf1 is covalently modified by SUMO at K35. Compared to wild type Maf1, Maf1K35R is impaired in its ability to repress tRNA gene transcription and to suppress colony growth. Covalent SUMO modification of Maf1 at K35 is required for Maf1 association with RNA pol III and recruitment to RNA pol III –dependent gene promoters. Additionally, Maf1 can also non-covalently interact with SUMO1 and SUMO2. Covalent modification of Maf1 at K35 is not required for this interaction. Further mutational analysis identified S123 as an important residue that controls non-covalent interaction of Maf1 with SUMO, potentially, by its phosphorylation state. Compared with wild type Maf1, Maf1S123A displays an enhanced ability to both, repress transcription and suppress colony growth, whereas Maf1S123D exhibits a decrease in these Maf1 functions. Furthermore, Maf1S123A shows enhanced interaction with RNA pol III. To understand how Maf1 maybe recruited to promoters we have found that Maf1 interacts with CDK8 and MED13, members of the Mediator-CDK8 subcomplex, which are also bound at the TBP promoter. While Maf1 represses TBP expression, CDK8 induces TBP gene expression. This supports the idea that association between Maf1 and the CDK8 subcomplex regulates gene activity. Together, our results demonstrate that covalent and non-covalent interactions of SUMO1 and SUMO2 with Maf1 positively regulate its capacity to repress transcription and cell growth. Additionally, Maf1 associates with the CDK8 subcomplex to repress CDK8 mediated gene activation function.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis

PDF

Maf1 is a novel target of the tumor suppressor PTEN and a negative regulator of lipid metabolism

PDF

Mechanism of repression of RNA polymerase III-dependent transcription under hypoxia

PDF

Positive regulation of RNA polymerase III-mediated transcription of tRNA genes by the Mediator kinase submodule

PDF

Characterizing the physiological roles and regulatory mechanisms of Maf1

PDF

RNA polymerase III-dependent transcription is repressed under prolonged hypoxic conditions

PDF

Repression of RNA polymerase III-dependent transcription by the tumor suppressors p53 and PTEN

PDF

Do ZFX and ZNF711 regulate the same genes in HEK293T cells?

PDF

Mechanisms for regulating the expression of the TATA-binding protein

PDF

The histone H2A deubiquitinase MYSM1 regulates CCR9 expression on CD4⁺ T cells and thymocytes

PDF

Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics

PDF

CpG poor promoter SULT1C2 regulated by DNA methylation and is induced by cigarette smoke condensate in lung cell lines

PDF

Multiple functions of the PR-Set7 histone methyltransferase: from transcription to the cell cycle

PDF

Regulation of Aurora kinase B and its effect on phosphorylation of G9a/GLP

PDF

Brain-derived neurotrophic factor (BDNF)-dependent activation of calpain: involvement of ERK pathway and regulation of AMPA receptor function and stability

PDF

The mechanisms of somatic cell reprogramming

PDF

Using genomics to understand the gene selectivity of steroid hormone receptors

PDF

Creating a multiple micrornia expression vector to target GRP78, an ER chaperone and signaling regulator in cancer

PDF

Kaposi's sarcoma associated herpes-virus induces cellular proxi expression to modulate host gene expression that benefits viral infection and oncogenesis

PDF

The role of endoplasmic reticulum chaperones in regulating hematopoietic stem cells and hematological malignancies

Asset Metadata

Creator Rohira, Aarti D. (author)

Core Title SUMOylation regulates RNA polymerase III -- dependent transcripton via MAF1

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 04/23/2013

Defense Date 10/19/2012

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

Tag CDK8,Maf1,OAI-PMH Harvest,RNA polymerase III,SUMOylation,transcriptional regulation

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Johnson, Deborah L. (committee chair), An, Woojin (committee member), Stallcup, Michael R. (committee member)

Creator Email aartirohira13@gmail.com,rohira@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-240896

Unique identifier UC11292801

Identifier etd-RohiraAart-1581.pdf (filename),usctheses-c3-240896 (legacy record id)

Legacy Identifier etd-RohiraAart-1581.pdf

Dmrecord 240896

Document Type Dissertation

Format application/pdf (imt)

Rights Rohira, Aarti D.

Type texts

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

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

Repository Name University of Southern California Digital Library

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