£%. 3..

‘114
I VII-p

. . . .52...»
.. u... . :1:
1.12:3: . . .

A...
Fairs .314.
mil

7:5...
33.3.... k d.

1,... 1.2. .. .T .
2.. t..-) . 1.2.... 1
, $3.11 I:

fix: . 3.
3.32:».
.. {in .v
‘1: i .3 A
...§R.~.¢.I.uat.lvni&
- .«~.Ln.1tp’

rlJr; 7...!

.55.. .3)
3.. . 9 52.97 .

:.\ .7
7y 8%...

_\
v feta-Lulu». 4.
.11 t I. .11.: .

2 V1

.1 .6...
.rl) . . .1. r
‘. s:

 

 

magma

300?

This is to certify that the
dissertation entitled

FUNCTIONAL CHARACTERIZATION OF Drosophila
C-TERMINAL BINDING PROTEIN

presented by

Priya Mani

has been accepted towards fulfillment
of the requirements for the

degree in Cell and Molecular Biology

 

 

MQMM

Major Professor’s Signature
Il/Ilj/Dé

Date

 

MSU is an Affimtative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

i DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p2/CIRCIDateDue.indd-p.1

 

FUNCTIONAL CHARACTERIZATION OF Drosophila
C-TERMINAL BINDING PROTEIN

By

Priya Mani

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Program in Cell and Molecular Biology

2006

ABSTRACT

FUNCTIONAL CHARACTERIZATION OF Drosophila
C-TERMINAL BINDING PROTEIN
BY

Priya Mani

Transcriptional repression results from the selective actions of
negatively acting repressors and corepressors that function by multiple
mechanisms. C-terminal Binding Protein (CtBP) is one such interesting
corepressor important for development, oncogenesis and transcriptional
regulation. Distinct forms of this protein have been uncovered in vertebrates
that are capable of performing a variety of functions, both in the nucleus and
cytoplasm. All CtBP isoforrns have a high degree of similarity to NAD-
dependent hydroxy acid dehydrogenase enzymes, with some isoforrns
carrying C-tenninal extensions (tail). Distinct CtBP proteins are produced in
Drosophila that differ in the presence or absence of the C-terminal tail. The
significance of the homology to metabolic dehydrogenase enzymes and the
function of the C-tenninal extension is still largely unknown.

Herein we demonstrate that the expression and coding information for
distinct CtBP isoforrns differing in their C-terrnini is a conserved feature of
phylogenetically divergent organisms. in many Drosophila species, the
relative levels of isoforrns are altered during development with a significant
drop in the tailed isoform following embryogenesis. This developmental shift in

protein levels can be traced back to a difference in the abundance of

individual spliceform levels and points to the tailed isoform contributing an
evolutionarily conserved role.

lntriguingly, all CtBP proteins contain dehydrogenase domains that
include an nucleotide (NAD) binding Rossman fold, a putative catalytic core
and a broad dimerization region. To investigate the significance of this
dehydrogenase homology, mutants in these critical domains were
misexpressed in the fly and analyzed for biological activity. Overexpressed
wild type isoforms and catalytic mutants resulted in aberrant phenotypes while
the NAD binding mutant and the dimerization mutants were completely
disrupted for biological activity, with low protein expression. We show that the
ability to dimerize is important for CtBP function and the biological activity of
the catalytic mutant might be a consequence of dimerization with endogenous
CtBP proteins. Our results also indicate that NAD binding is not needed for
dimerization and we speculate that it may be required for association with
cofactors.

The experiments described here provide a broader understanding of
CtBP function in the context of the whole organism, and should in turn
facilitate our mechanistic understanding of CtBP function and eukaryotic gene

regulation.

To my Dad, Mr. Chandra Shekhar Mani, my Mom, Mrs. Lakshmi Mani,
my husband, Dr. Adwait U. Telang
AND

To my mentor, Dr. David N. Amosti

iv

ACKNOWLEDGEMENTS

This thesis is by far the most significant scientific accomplishment of
my life and it would not exist without the people who supported me and
believed in me. I am indebted to my mentor Dr. David Arnosti, for his
guidance and his personal commitment towards my graduate training. In his
own unique way, he challenged and inspired me to be a better thinker, a
better orator and a better researcher. Thanks to him, science will always be
fun for me! My experience in this lab taught me much more than just science,
I learnt that the virtues of patience, dedication and tenacity can take you a

long way and these are qualities that I take with me.

I am very grateful to members of my thesis committee, Dr. R. William
Henry, Dr. Laurie S. Kaguni, Dr. Karl Olson and Dr. Donald Jump for their
insightful ideas and constructive comments that helped to shape my
manuscript and this thesis.

Many thanks to my colleagues in the Amosti laboratory, past (Meghana
Kulkarni, Paolo Struffi, Scott Keller and Montserrat Sutrias-Grau) and present
(Sandhya Payankaulam, Carlos Martinez, Walid Fakhouri, Yang Zhang, Li Li
and Martin Buckley) for their friendship and scientific expertise. I am
especially grateful to Meghana Kulkarni for getting me started in the Amosti
lab and for her guidance through my formative years. Many thanks to
Montserrat Sutrias-Grau whose work served as a starting point for this thesis.
I am also very grateful to Sandhya Payankaulam for being a close friend and

confidante, and for pushing me towards my goal.

I am also very thankful for my friends and colleagues in various
departments (Gauri Jawdekar, Dorothy Tappenden, Soumya Korrapati, Tejas
Kadakia, Chetan Sukuru, Zakir Ullah and Deborah Yoder) for being there in
good and bad times.

To my family, thank you for being my family and for being there; my
deepest gratitude to my mom, dad, sister, mother-in-law and brother-in-law
who taught me to be tough by example.

Finally, I don’t have enough words to thank my husband Adwait, who
supported my ambition in getting a Ph.D. years ago, and from whom I
received a daily dose of motivation to accomplish it. I thank him for his
acceptance, love, and understanding of my elevated stress levels, especially
in the last two years. He has my heartfelt gratitude for urging me to look at the

bigger picture in life ...always.

vi

TABLE OF CONTENTS

LIST OF FIGURES ................................................................................ ix

LIST OF TABLES .................................................................................. xi

KEY TO ABBREVIATIONS .................................................................... xii

CHAPTER I

Introduction ...................................................................................... 1-45
1.1 CtBP family of proteins ......................................................... 3
1.2 Domain structure — Implications for CtBP function ...................... 5
1.3 Regulation of CtBP proteins ................................................. 10
1.4 Mechanism of CtBP mediated transcriptional repression ............ 20
1.5 A distinct cytosolic role ascribed to CtBP ................................. 22
1.6 CtBP as a corepressor during Drosophila development .............. 24
1.7 Alternative splicing of Drosophila CtBP ................................... 34

CHAPTER II

Developmental Expression and Phylogenetic Characterization of CtBP

corepressor protein ......................................................................... 46-76
Abstract .................................................................................... 47
Introduction ............................................................................... 48
Materials and Methods ................................................................. 52
Results ..................................................................................... 57
Discussion ................................................................................. 71
References ................................................................................ 74

CHAPTER III

NAD binding modulates CtBP biological activity .................................... 77-110
Abstract .................................................................................... 77
Introduction ................................................................................ 78
Materials and Methods ................................................................. 82
Results ..................................................................................... 87
Discussion ............................................................................... 103
References .............................................................................. 108

vii

CHAPTER IV

Conclusions and Future Directions ................................................... 111-123
APPENDIX A
Rescue of CtBP function in the fly .................................................... 124-140
APPENDIX B
An RNAi approach to depleting the CtBPL isoform ............................... 141-147

viii

Figure l-1:
Figure l-2:

Figure I-3:

Figure l-4:

Figure "-1:

Figure "-2:

Figure "-3:

Figure "-4:

Figure III-1:

Figure Ill-2:

Figure III-3:

Figure III-4:

Figure Ill-5:

LIST OF FIGURES

Key domains of CtBP proteins ....................................... 9
CtBP genomic structure and splicing patterns .................. 14

Drosophila CtBP proteins are ubiquitously distributed during
development ............................................................ 28

Alignment of different CtBP isoforms generated by
alternative splicing in Drosophila ................................ 36

Developmental expression profile of CtBP isoforms in
Drosophila melanogaster ............................................ 59

Conservation of coding information for CtBPL-specific C-
terminus ............................................................. 63, 64

Conserved developmental regulation of CtBP protein
expression in Drosophila moja vensis and Drosophila
virilis ....................................................................... 67

Adult expression of CtBP proteins in four Drosophila
species, Anopheles gambiae, Apis mellifera and Tribolium
castaneum ............................................................... 69

Phenotypes induced by overexpression of wild type and
mutant CtBP proteins ................................................. 88

Relative steady state levels of recombinant CtBP proteins:
CtBPL (NAD) and CtBPL (DIM) mutants and CtBPs isoform
are expressed at low levels ......................................... 93

CtBPL (CAT) protein is capable of
dimerization .............................................................. 95

lmmunoprecipitation of endogenous CtBPL and CtBPs with

FLAG-tagged CtBP proteins indicate CtBP homo— and
heterodimerization .................................................... 98

CtBPL (NAD) and CtBPL (DIM) mutants are expressed at low
steady state levels ..................................................... 99

ix

Figure III-6:

Figure A-1:

Figure A-2:

Figure A-3:

Figure A-4:

Figure 8-1:

NAD binding mutant migrates in molecular mass consistent
with dimers but not the dimerization mutant .................. 101

Genomic organization of the Drosophila CtBP locus ....... 126

Expression of CtBP isoforms under the control of
endogenous promoter .............................................. 127

Expression of recombinant CtBPL assayed by Western
blotting .................................................................. 128

Genetic crosses for whole animal rescue of CtBP mutant
phenotype .............................................................. 131

A transgenic system to trigger RNAi and selectively eliminate
CtBPL in Drosophila ................................................ 144

Images in this dissertation are presented in color.

LIST OF TABLES

Table Ill-1: Quantitation of CtBP misexpression phenotypes .............. 90
Table A-1: CtBP promoter driven transgenes do not rescue CtBP
mutants ................................................................. 133

xi

aa

bp
cDNA:
Co-IP
CtBP
dCtBP
DNA
DTT
Eco/i
EDTA
HAT
HDAC
HEPES
HMT
hsp
Kb
KDa
NAD
NLS
PAGE
PcG
PCR
Pol II
Rb
RNA
RNAi
RT-PCR
TSA
Wt

KEY TO ABBREVIATIONS

amino acid

anterior-posterior

base pair(s)

complementary deoxyribonucleic acid
Co-immunoprecipitation

C-terrninal Binding Protein

Drosophila C-tenninal Binding Protein
Deoxyriboncleic acid

Dithiothreitol

Escherichia coli

Ethylene dinitrilo tetra acetic acid

Histone acetyltransferase

Histone deacetylase
N-2-hydroxyethylpiperazine—N’-2-ethanesulfonic acid
Histone methyltransferases

heat shock promoter

Kilo base

Kilo Dalton

Nicotinamide adenine dinucleotide
Nuclear localization signal

Poly acrylamide gel electrophoresis
Polycomb

Polymerase chain reaction

RNA Polymerase ll

Retinoblastoma tumor suppressor protein
Ribonucleic acid

RNA interference

Reverse transcriptase polymerase chain reaction
Trichostatin A

Wild type

xii

CHAPTER I

INTRODUCTION

Gene expression in eukaryotes is a highly elaborate process under
tight regulation at several junctures. An ever-growing body of evidence shows
that changes in transcriptional profiles form a vital part of the genetic basis of
the evolution of development and contribute to animal diversity (Wray, 2003).
Developmental processes are regulated by extensive protein: protein and
protein: DNA interactions during the regulation of transcription. Appropriate
transcriptional regulation is thus central to cellular homeostasis, and
misregulation of gene expression has been implicated in several disease
states. Mutations that disrupt developmental processes often map to loci
encoding transcription factors, which emphasizes the importance of
transcription in development. These regulatory factors are the key to spatio—
temporal differential gene expression that must transpire for appropriate
cellular differentiation and development to ensue (Lawrence, 1992).

The specificity of differential gene expression in eukaryotes is
controlled by the coordinated actions of sequence specific, context-dependent
activator and repressor proteins. Early investigations have led to the
identification of these coregulators and recent efforts are aimed at the precise
characterization of their molecular function. Repressor proteins are aided by
corepressor proteins that function by multiple mechanisms and may either
help quantitatively by enhancing the overall repression activity (Struffi et al.,
2004) or qualitatively, by providing a unique repression activity (Lunyak et al.,

2002). Data from several sources suggests that such corepressors link

sequence-specific repressors to the core RNA polymerase II machinery to
accomplish direct repression (Austin and Biggin, 1995; Mannervik et al.,
1999). In addition, corepressors are also aided by histone modifying activities
like deacetylases, methyl transferases and newly-identified demethylases
which act cooperatively and are documented to be important players of
several repressor assemblies (Arnosti, 2004; Chinnadurai, 2006). An
expanding array of post-translational modifications is being acknowledged as
means of regulating protein localization and abundance. The ability to use
various repression mechanisms in a combinatorial fashion may be exploited
to provide unique patterns of gene expression. Current investigations are
aimed at defining the precise mechanisms of cofactor function and regulation.

C-terminal Binding Protein (CtBP) is a versatile prototype of an
expanding family of corepressor proteins that aids in the repression of a
diverse array of DNA binding transcriptional repressors in vertebrates and
invertebrates alike. In this thesis, I describe studies that analyze the different
isoforms of the Drosophila homolog of CtBP and reveal dynamic changes
occurring in protein levels during development.

Recently, the cellular redox status has also been suggested to
influence gene transcription (lmai et al., 2000; Rutter et al., 2001). CtBP
proteins resemble NAD-dependent dehydrogenases in structure but whether
this similarity extends to function and whether such activity is linked to
transcription is unclear. This study also extends to testing the significance of
dehydrogenase-like conserved domains within CtBP in the context of a whole
organism. These investigations are fundamental to understanding the

molecular basis of CtBP’s function as a corepressor involved in development.

1.1 CtBP family of proteins

The founding member of the C-terrninal Binding Protein (CtBP) family
was discovered in 1993 as a 48KDa cellular phosphoprotein that associated
with the C-terminal region of the adenoviral 2/5 E1A oncoprotein. This
interaction was found to negatively modulate oncogenic transformation,
tumorigenesis and metastasis by E1A. Binding was found to depend on a
PLDLS (Pro-Leu-Asp-Leu-Ser) motif present near the C-terminal end of E1A
protein, conserved among E1A proteins of primate adenoviruses (Boyd et al.,
1993; Schaeper et al., 1995). The N-terminal exon of E1A is capable of
cellular transformation, but the presence of this motif was shown to restrain
exon 1 mediated transformation, suggesting that CtBP binding could influence
E1A mediated gene expression (Sollerbrant et al., 1996). Deletion of the
CtBP-binding motif abrogated the repressive activity of the C-terminal region.
These studies provided early evidence of CtBP acting to antagonize
transactivation and led to cloning of the cDNA for human CtBP1 (Schaeper et
al., 1995). Shortly thereafter, a highly related human protein termed CtBP2 of
48KDa was identified by BLAST analysis of EST databank sequences.
Coincidentally, CtBP2 was simultaneously discovered in a two-hybrid screen
using the murine BKLF (Basic Kruppel-like factor) as prey (T umer and
Crossley, 2001). Human CtBP1 and CtBP2 are present on different
chromosomes and share over 83% amino acid similarity (Katsanis and Fisher,
1998). Other vertebrates such as mice and frogs also have two CtBP
homologs, while the invertebrate genomes including those of C. elegans and

Drosophila melanogaster contain a single gene for CtBP. It is in the fly system

that an early role for CtBP in transcriptional repression was established (See
Section 1.5). Since then, cell culture based assays and vertebrate models
have conclusively shown that CtBP is an essential protein and acts as a

bonafide corepressor in these systems.

CtBP is essential for vertebrate development

In mice, both Ctbp1 and Ctbp2 transcripts are expressed widely during
development in unique and overlapping patterns (Furusawa et al., 1999). To
address the in vivo significance of multiple CtBP functions, mice carrying
mutations in both CtBP1 and 2 were generated. Ctbp1-null mice are reduced
in size but viable. In contrast Ctbp2-null mice perish by embryonic day 10.5
and show defects in several tissues. This data demonstrates that there are
spatial differences between expression patterns of the proteins and suggests
that they might have unique functions. CtBP proteins have also been ascribed
roles in regulating Golgi dynamics (See Section 1.4), however no Golgi
defects in these mutants was noted. Compound mutants for Ctbp1-I- and
Ctbp2-l- arrested at the head fold stage and a reduction in the dosage of one
gene enhanced the severity of phenotype associated with the mutation in the
other gene (Hildebrand and Soriano, 2002). These results indicate that CtBP1
and 2 genetically interact. A similar picture emerges from the recent avian
expression study of Ctbp1 and Ctbp2, wherein the proteins are expressed in

partly overlapping expression domains in the avian embryo (Van Hateren et

al., 2006). This study suggests that avian Ctbp1 and Ctbp2 might show

functional redundancy in some tissues and specific roles in others.

1.2 Domain structure - Implications for CtBP function

From a structural standpoint, CtBP proteins can be divided into three
distinct domains. The first is the substrate-binding cleft that is responsible for
specific recognition of PXDLS motifs in binding partners. Co—crystallization of
rat CtBP with NAD (H) and a PXDLS like peptide show the presence of this
peptide bound to the N-terminus of CtBP. Structural comparisons between
binary and ternary complexes reveal that binding of this short peptide is not
associated with any major conformational changes in the protein and the
peptide binding cleft has no contact with the nucleotide binding region (Nardini
et al., 2003).

The second and most conserved of all includes the central nucleotide
binding regions. The initial cloning of human CtBP1 and subsequent
homology searches revealed extensive homology with NAD-dependent D-
isomer specific 2-hydroxy acid dehydrogenases (Schaeper et al., 1995). This
homology extends over a nucleotide/NAD binding Rossman fold (GXGXXG)
and a putative catalytic histidine residue (towards the C-terminus) in the active
site of these enzymes. Dehydrogenase activities have not been previously
associated with transcriptional repressors. The structural similarities prompted
speculation that CtBP might possess an intrinsic dehydrogenase-like activity

that assists in remodeling chromatin structure (Kumar et al., 2002). More

importantly, it also pointed to a role for NAD as a modulator of gene
expression. Cellular NAD levels have been linked to transcription before when
the Sir2 histone deacetylase (HDAC) was identified as having an NAD-
dependent HDAC activity (lmai et al., 2000).

Early attempts to identify dehydrogenase-like activity associated with
CtBP were unsuccessful, however a weak activity was reported for CtBP1
(Kumar et al., 2002; Schaeper et al., 1995). Here, Kumar and colleagues
tested the dehydrogenase activity of CtBP by combining the reduction of
pyruvate to lactate with the oxidation of NADH to NAD+. Recombinant CtBP
was able to catalyze this in vitro enzymatic reaction in a dose dependent
manner, albeit inefficiently, indicating it is a functional dehydrogenase. The
crystal structure for CtBP’s minimal dehydrogenase domain revealed a CtBP
monomer that forms extensive dimer contacts with another monomer, in the
presence of NAD. The dehydrogenase-like domains provide the majority of
the contacts needed for dimerization. NAD binds within a cleft at the
confluence of two such monomers, inducing conformational changes that
allow enhanced binding to transcription factors like E1A. When tested
biochemically, mutations in key residues of the dinucleotide binding fold
(G181V,G183V, D204A) or the dimerization interface (R141A, R142A,
R163A, R171A) disrupt binding to E1A and impair repression by E1A (Kumar
et al., 2002). Also conserved in CtBP is the His/Glu/Asp triad that constitutes
the active center of dehydrogenases. These residues are conserved in CtBP
proteins from vertebrates and invertebrates (See Figure l-1). Mutations in
these residues were also shown to compromise the ability of CtBP to contact

E1A, and the protein was non-functional when tested as a corepressor.

Moreover, this putative catalytic mutant was unable to catalyze the
dehydrogenase reaction, unlike the wild type CtBP (Kumar et al., 2002).
These observations lend support to the PXDLS binding cleft making direct
contacts with the dehydrogenase-like regions.

The structural data described above agrees well with the crystal
structure of CtBP3/BARS (a truncated form of rat CtBP1, discussed in Section
1.5). Overall, evidence from both reports concur that dinucleotide binding
triggers a closed conformation, triggering closer contacts at the dimerization
interface. However, there are some minor differences noted. BARS was co-
crystallized was with a short PXDLS-like peptide and the peptide was found to
associate exclusively with the N-terminal part of CtBP and binding of the
peptide does not induce any noticeable structural alterations (Nardini et al.,
2003). This is in contrast to the observations made by Kumar and colleagues
for CtBP1, wherein the PXDLS motif is predicted to directly interact with the
dehydrogenase-like regions, on the basis of structural data and mutagenesis
experiments. Mutations in dehydrogenase-like regions (61720 in the
nucleotide binding region or H304L in the catalytic center) were found to
prevent NAD binding, but did not affect binding to transcription factor E1A.

Lastly, the least well-conserved part of the protein is the variable C-
terminal extension. This C-terminal tail harbors sites for various post-
translational modifications and may play a regulatory role (discussed in
Section1.3, See Figure l-1). Vertebrate CtBP proteins all possess C-terrninal
extensions of ~90 residues that are predicted to be intrinsically unstructured in

solution, being composed primarily of disorder promoting residues (56%)

Figure M: Key domains of CtBP proteins

CtBP proteins display striking similarity to alpha hydroxy acid
dehydrogenases. Regions of highest homology between CtBP and hydroxy-
acid dehydrogenases are shown in blue shading. The putative catalytic
residue His-315 and the central nucleotide (NAD) binding domain are
indicated (Chinnadurai, 2002). The PXDLS binding region has been mapped
to the N-terminus of the vertebrate protein (Koipally and Georgopoulos, 2000;
Nardini et al., 2003). Residues important for dimerization map within the
dehydrogenase domains. CtBP is subject to a host of post-translational
modifications, the majority of which have been identified to act on the C-
terrninal tail of vertebrate CtBP proteins. Phosphorylation sites by Pak1,
HlPK2 and c-Jun-NHZ kinase, sumoylation motif, PXDLS binding region and
the nNOS-PDZ binding regions are indicated.

 

wm<m E

 

 

 

:65 055m
« 82... «12-570 2:205 amuo E. 5 323.80 E33.
2. 83.3 . P
ow
82’. F

:20 9.6% Bax...

m 5&1
8F-oxx0x0-BF
@523 o<z
_

_i
399:0... mmmcoaocgcco

_i
co=m~tmE_voEor

(Nardini et al., 2006). Its disordered structure may provide CtBP with the
flexibility to contact other proteins that modify this region. It is also plausible
that the C-terminus affects CtBP function after recruitment to a transcription

factor bearing a PXDLS motif.

1.3 Regulation of CtBP proteins

NAD binding and dimerization

The finding that CtBP-E1A associations and dimerization are regulated
by NAD suggest that this cofactor can influence transcriptional outputs.
Structural data from CtBP proteins show that small metabolic intermediates
like NAD+ and NADH can bind and induce conformational changes that have
functional consequences relating to repression (Kumar et al., 2002). Although
the effect of dinucleotide binding on CtBP function has been investigated by
many, the differential efficacy of NAD+ (oxidized) v/s NADH (reduced) is
debatable. Goodman and colleagues speculate that since free NAD+ levels
exceed free NADH levels in a cell, changes in the redox state could be
manifested by NADH levels that are limiting. Consistently, NADH was found to
efficiently enhance vertebrate CtBP-E1A binding three orders of magnitude
better than NAD+ (Zhang et al., 2002). Another study measuring binding
affinities shows CtBP to have higher affinity for NADH than NAD+ (>100 fold)
(Fjeld et al., 2003), however other studies find NAD+/NADH to be equally

effective in enhancing CtBP binding E1A (Kumar et al., 2002). Thus, a

10

speculative model has been suggested for CtBP to play the part of a cellular
redox sensor that interprets the cellular metabolic demands and accordingly
directs transcription (Zhang et al., 2002). Binding of dinucleotide is thought to
bring about conformational changes in protein structure that stimulate
dimerization (Balasubramanian et al., 2003). This might be one mechanism by
which CtBP proteins enter the nucleus and regulate target genes. In this
manner, NAD can be considered a switch that assembles the CtBP complex.
A good example of the impact of NAD on CtBP activity is a recent
study describing CtBP to be a corepressor for neuronal genes such as those
involved in drug-resistant epilepsy. CtBP was recruited by NRSF (neuronal
restrictive silencing factor) and was found to enhance NRSF-dependent
repression. The CtBP-NRSF direct interaction was disrupted upon increasing
levels of NADH (by speeding up glycolysis) resulting in de-repression of
neuronal targets. Conversely, the glycolytic inhibitor 2DG (2-deoxy-D-glucose)
was found to enhance NRSF directed repression of target genes, that showed
a marked reduction in levels of H3-K9 acetylation (mark of activation) and an
increase in H3-K9 methylation (mark of repression). Repression was lost in
CtBP null mouse embryonic fibroblasts (MEFs) (Garriga-Canut et al., 2006).
This demonstration of small molecule regulation of energy metabolism directly
impacting chromatin structure and gene expression in disease states is novel,
but altering central processes like glycolysis is very likely to cause pleiotropic

effects.

11

Multiple CtBP proteins are encoded by multifunctional genes

In mammals, the CtBP family is encoded by two genes: Ctbp1 and
Ctbp2. Both genes generate multiple isoforms by either alternative splicing or
differential promoter utilization (See Figure l-2). Ctbp1 undergoes alternative
splicing at the N-terminus to yield two splice variants: CtBP1-L (Long et al.)
and CtBP1-S (short) that differ by the presence of 11aa additional at the N-
terminus of CtBP1-L. Differential promoter usage and gene splicing from Ctbp
2 generates CtBP2 and a retinal-specific variant, RIBEYE (Chinnadurai,
2002). RIBEYE protein is expressed from a tissue specific promoter located
within an intron in Ctbp2 that results in a large (~ 565 aa, also called the A
domain) N-terminal domain being fused to CtBP2 except for the 20 aa from
the N-terminus (Schmitz et al., 2000). CtBP2 uses an upstream promoter and
alternative splicing to eliminate the exon that encodes the N-terminal domain
of RIBEYE.

CtBP1-L and CtBP2 have been deemed corepressors, while CtBP1-S
has been ascribed a unique cytoplasmic role in Golgi fissioning. CtBP1—L has
been found to associate with synaptic ribbons and its roles in the cytoplasm
are speculative. The RIBEYE variant of CtBP2 is found in abundance in
specialized photoreceptor ribbon synapses involved in vision, and may be
involved in regulating neurotransmitter release (Schmitz et al., 2000).
Interestingly, the exon encoding the RIBEYE specific region is absent in worm
and fly genomes, suggesting it is a vertebrate novelty. A recent study also

reported the identification of a splice variant of CtBP2 that is localized

12

Figure I-2: CtBP genomic structure and splicing patterns

A. Alternative splicing of the vertebrate CtBP1 gene generates CtBP1-L and
CtBP1-S possessing divergent functions. Exons exclusive to CtBP1-L (red)
and CtBP1-S (blue) are color coded.

B. Alternative promoter utilization and gene splicing create CtBP2 that
includes an NLS (yellow box) while RIBEYE is produced from an intronic
promoter that encodes a large A domain fused to the remainder of the CtBP2
coding sequence. RIBEYE hence lacks the NLS. CtBP2-S is a splice variant
of CtBP2 that is devoid of the NLS and thus predominantly cytosolic.

C. Drosophila melanogaster CtBP is spliced to produce CtBPL and CtBPs that
differ by the presence of the C-terrninal extension encoded by exons 6 and 7

(numbered orange boxes), exclusive to CtBPL.
Cartoons are not drawn to scale and are adapted from (Verger et al., 2006).

13

Cpr1

 

CtBP1-L LI--—: l—L ooreressorp

 

a m 7 ~ ,Golgi
Ll

CtBP1-S [I-—-—---—---~~I T---- rim-"cram“u'i if“: fission

3- 5' UTR NLS

0‘pr l3 f3 “it.

CtBP2 DWDMEmflE—U*fl§flfi- Corepressor
CtBP2-S DmfI-m—Dm-WU-E-D-flfiflfifij Unknown
my.-~--n---a-—a~clmi£:1Swap“

fibbons
C.
dCtbp ”dig/h
CtBPs E-E—:3~Q~Ej Corepressor
CtBPL E~E-flflij———~[:ID Corepressor

    

 

specifically to the cytoplasm (See Figure l-2 and Section 1.4) (Verger et al.,
2006). The Drosophila homolog of CtBP is also subject to alternative splicing
and generates two major isoforms (See Figure l-2) as discussed in detail in

Section 1.5 and Chapter 2.

PXDLS interactions: Significance and determinants of binding

Early studies with the E1A protein suggested that CtBP is recruited to
promoters through interactions with a PXDLS motif (Pro-Leu-Asp-Leu-Ser in
E1A)), where X is usually a bulky group like Leucine, lso-leucine or Valine.
This has been found true for dozens of transcriptional effectors containing
PXDLS-like motifs that bind CtBP. These include zinc-finger proteins like
lkaros (Perdomo and Crossley, 2002), BKLF (van Vliet et al., 2000) , FOG,
Snail, delta EF1 (Furusawa et al., 1999), Evi-1 (Palmer et al., 2001), bHLH
group proteins like Hairy (Poortinga et al., 1998) and viral oncoproteins like
E1A. Subsequent studies confirmed that a vast majority of DNA binding
transcription factors that need CtBP contained variants of this motif, some in
more than one copy. Mutation in this short stretch of residues has been tested
in many proteins and has been found to abolish CtBP interaction and
repression (Boyd et al., 1993; Keller et al., 2000; V0 et al., 2001). Most
notable is a study describing a naturally occurring mutation in the PXDLS in
human homeodomain protein, TGIF. As a consequence of this mutation,
TGIF-CtBP interaction is disrupted, resulting in holoprosencephaly and

craniofacial abnormalities (Melhuish and Wotton, 2000). In Xenopus, the

15

hematopoietic transcription factor FOG represses erythroid development by
the recruitment of CtBP via a PXDLS motif. Mutations in this motif are seen to
abrogate the interaction. These results suggest that CtBP plays an important
role in hematopoiesis (Deconinck et al., 2000) .

CtBPs also bind transcription factors that do not contain an identifiable
PXDLS. Many of the vertebrate effectors like HDACs (HDAC1, 2 and 5) fall
into this category. This is also true for the Drosophila HDAC de3 that does
not have an discernible dCtBP binding motif but is a part of the Knirps
repressive complex that includes CtBP, and interacts physically and
functionally with Knirps (Struffi and Arnosti, 2005). CtBP physically interacts
with several histone acetyltransferase (HAT) enzymes whose bromodomains
contain a consensus PXDLS like motif. However, deletions of the
bromodomain or mutations in the PXDLS within the bromodomain retain
interaction with CtBP, suggesting that additional non-canonical sites must
exist (Senyuk et al., 2005).

Each unit of CtBP harbors a PXDLS binding cleft, suggesting that
dimers can serve to bridge two PXDLS partners together. When fused to a
heterologous DNA binding domain, disruption of the PXDLS responsive cleft
in CtBP does not affect its repression activity. In contrast, repression is
abolished when tested in the absence of the heterologous domain, suggesting
that the cleft does not determine association with other effectors, but primarily
functions to localize CtBP to repressors bound at target promoters (Quinlan et
al., 2006).

In the case of many transcription factors, it has been suggested that

neighboring residues flanking the PXDLS motif might contribute to the folding

16

capacity of this region and consequently affect its interaction with CtBP.
Structures of synthetic peptides examined using NMR spectroscopy suggest
that peptides have different affinities for different CtBPs eg. the Drosophila
Hairy protein binds Drosophila CtBP stronger than human CtBP suggesting
alternative CtBPs have distinct binding specificities (Molloy et al., 2001). In
E1A protein, acetylation of a lysine residue close to the PXDLS motif in the C-
terminus disrupts binding to both CtBP1 and 2 when tested as a synthetic
peptide. This interference could be through structural changes, charge

neutralization or a change in the subcellular localization of CtBP (Molloy et al.,

2006).

Mechanisms dicta ting Localization and Stability of CtBP proteins

CtBP1 and 2 are highly similar proteins and possess similar
transcriptional repression activities, although new evidence suggests that their
activities can be differentially regulated, by specific control of their subcellular
location. CtBP proteins have been detected both in the nucleus and in the
cytoplasm. Much of our current knowledge of CtBP localization is derived
from studies centered on CtBP1-L, whose cellular location is actively
regulated. CtBP1-L is predominantly cytoplasmic in Cos cells but when
coexpressed with a PXDLS containing protein, such as Net or HIC1
transcription factors, it is relocated into the nucleus (Criqui-Filipe et al., 1999;
Deltour et al., 2002). CtBP1-L has been shown to interact with the PDZ

domain (protein-protein interaction domain; PSD-95, discs-large and zona

17

occludens-1) of nNOS (neuronal nitric oxide synthase) via a C-terrninal DXL
sequence, which shuttles CtBP from the nucleus to the cytoplasm. This
mechanism might also potentially serve to make CtBP available for its
cytoplasmic roles as discussed in Section 1.5.

Posttranslational modifications are also accountable for CtBP
subcellular distribution. CtBP1-L is SUMOylated at Lys 428 via P02
(Polycomb group protein) and this modification has profound effects on its
distribution into the nucleus, antagonistic to that of nNOS. Mutation of Lys 428
into Arg 428 shifts CtBP from the nucleus to the cytoplasm and consistent
with its localization, restricts CtBP from acting as a corepressor (Lin et al.,
2003). CtBP2 lacks this SUMOylation site and the PDZ binding domain and is
likely to be regulated differently.

Independent phosphorylation of CtBP by three different kinases also
plays a regulatory role. The p21-activated kinase (Pak1) has been shown to
phosphorylate CtBP1-L on Ser 158, a modification that relocates CtBP into
the cytoplasm similar to nNOS activity (Barnes et al., 2003). Homeodomain
Interacting Protein kinase 2 (HIPK2) is a serine/threonine kinase involved in
transcription and apoptosis. HIPK2 targets Ser 422 in the C—terminus of CtBP
for phosphorylation. HIPK2 has previously been shown to phosphorylate p53
in response to UV stress, thereby promoting apoptosis by induction of genes
like Bax, PERP, p21 and Noxa which are also upregulated in CtBP knockout
MEFs (Grooteclaes et al., 2003). HIPK2 modification marks CtBP for UV
triggered ubiquitination and subsequent degradation via the proteasome
providing an alternative pathway for apoptosis, by a p53 independent

mechanism (Zhang et al., 2005). c-Jun-NH2 terminal kinase is the third

18

kinase that also targets Ser 422 for phosphorylation and degradation. C-jun-
NH2 kinase plays important roles in inducing apoptosis under cellular stress
conditions like UV irradiation or cisplatin treatment (Wang et al., 2006). Under
both these conditions, CtBP levels were markedly reduced. These
observations indicate Ser 422 to be a mark for CtBP proteasomal degradation
and lead to the speculation that CtBP may respond to a wider network of
signals than those currently known.

Amongst the CtBP proteins, CtBP1-S and RIBEYE appear to have
roles in the cytosol, CtBP2 is predominantly nuclear and CtBP1-L is both
nuclear and cytoplasmic. A search for regions that serve as signals for
localization within CtBP1-L and CtBP2 identified a nuclear localization signal
(NLS) within CtBP2 (Verger et al., 2006). This ‘KRQR’ sequence is present in
the first exon of CtBP2 and is critical for its nuclear distribution. This motif is
not present in CtBP1-L. Within this unique CtBP2 NLS, mutagenesis data
indicates that the Lys 10 and Arg 11 are critical for localization. Interestingly
this region is conserved in the Dros0phila CtBP protein. This result is
consistent with reports that describe CtBP2 as a nuclear protein (Chinnadurai,
2006). When cotransfected with other CtBP proteins, the NLS of CtBP2 is
capable of shuttling the other CtBP proteins into the nucleus by virtue of
dimerization. A splice isoform of CtBP2 that lacks this NLS (CtBP2-S)
localizes to the cytoplasm. In contrast, a parallel study suggests that instead
of functioning as a classical NLS, the Lysine 10 residue might be a site for
acetylation of CtBP by p300 that directs it to the nucleus. Using a non-
acetylatable version of K10R, this study demonstrated that this acetylation is

required for nuclear retention of CtBP2 (Zhao et al., 2006). Further studies are

19

needed to distinguish whether the lysine 10 residue in CtBP2 plays a role in
nuclear import or nuclear retention. This mode of regulation of CtBP proteins
might have important ramifications on their roles in transcription, development

and cellular differentiation.

1.4 Mechanism of CtBP mediated transcriptional repression

CtBP is recruited to target promoters by DNA-binding repressors, but
details of events that follow are unknown. The precise nature of CtBP’s mode
of repression is a subject of intense investigation. Human CtBP1 associates
with several HDACs and histone deacetylation might be a major contributor to
repression (Sundqvist et al., 1998). However, CtBP mediated repression has
been reported to be insensitive to the Class I HDAC inhibitor, Trichostatin A,
signifying that HDAC-independent mechanisms might be at play (Ryu and
Arnosti, 2003). A first step was taken at identifying these cofactors with the
biochemical purification of vertebrate CtBP1 in HeLa cells. The associated
proteins were: DNA binding proteins like ZEB1, histone modifiers like histone
deacetylases and methyltransferases (HDACs 1 and 2, related HMTs like
EuHMT1 and 69a), chromodomain containing protein like HPC2 (human
polycomb), Co-REST and related proteins (corepressor for REST transcription
factor) and NPAO (nuclear polyamine oxidase). CtBP2 was also found in the
same complex, supporting the notion that CtBP proteins can heterodimerize.
This core CtBP complex could efficiently cause H3K9 and H3K27 methylation

of core histones, both of which are marks of repression. These modifications

20

were also enriched at a CtBP target, E-cadhen‘n. siRNA targeting CtBP not
only reduced occupancy of CtBP and EuHMT1 on the E-cad promoter but
also decreased methylation of H3 (Shi et al., 2003). Studies on the enzymatic
activity of NPAO led to the identification and characterization of the histone
demethylase, LSD1 (lysine specific demethylase) (Shi et al., 2005). LSD1
specifically demethylates H3K4 (a mark of active chromatin) via an amine
oxidation reaction (Shi et al., 2004). Its presence in the CtBP complex
suggests that CtBP proteins co-ordinate distinct enzymatic events that convert
active chromatin to a repressed state.

Another recently characterized mechanism of CtBP repression includes
physical interactions with HAT coactivators like p300 and CBP that contain
PXDLS motifs. CtBP was shown to specifically impair the HAT function of
these proteins by suppressing H3 acetylation (K9, K14 and K18) in an NADH
dependent manner (Kim et al., 2005; Senyuk et al., 2005). The association of
CtBP with both coactivators and corepressors of transcription suggests that
diverse CtBP complexes with distinct transcriptional properties might exist in
vivo.

In humans, CtBP also interacts with Cth (C-terminal Interacting
Protein). Cth bridges CtBP to Retinoblastoma protein, human Polycomb and
BRCA1 (breast cancer tumor suppressor) (Li et al., 1999; Meloni et al., 1999;
Sewalt et al., 1999). CtBP proteins also interact with the Polycomb protein
(Pc) in Xenopus (Sewalt et al., 1999). A functional role for this interaction has
already been discussed with regard to the localization of CtBP proteins by
SUMOylation. The Polycomb complex isolated from Drosophila does not

appear to contain CtBP and a Cth-like protein is yet to be identified (Levine et

21

al., 2002). These observations postulate that there might be key differences in
the complexes formed by human CtBP1 vs. Drosophila CtBP.

CtBP proteins may act mostly as bridging molecules to recruit the
above-described activities to the DNA binding repressor. As discussed
above, CtBP itself has an inherent dehydrogenase activity that might partake
in repression, but this remains unproven. Dehydrogenases have not yet been
documented to possess repressive activities, however an Oct-1 transcription
complex was recently found to utilize a dehydrogenase enzyme for the

activation of the histone H2B promoter (Zheng et al., 2003).

1.5 A distinct cytosolic role ascribed to CtBP

In a study of factors that regulate membrane tubulation, CtBP3 was
identified as a protein that is ribosylated in rat kidney cells upon treatment with
the fungal toxin Brefeldin A (Nardini et al., 2003; Spano et al., 1999). CtBP3
displays 97% identity to human and mouse CtBP1-L and 79% identity to
CtBP2. CtBP3 lacks the first 11 aa encoded by CtBP1; whether this protein
represents the rat version of CtBP1 or an N-terminal splice variant of CtBP1 is
not certain. The current consensus in the field is that CtBP3 represents. an N-
terminally truncated version of CtBP1. This protein is termed CtBP1-S or
BARS (Brefeldin-A ribosylated substrate). Recombinant BARS has been
shown to possess a weak acyl transferase enzymatic activity that catalyzes
acylation of lysophosphatidic acid to phosphatidic acid (a lipid involved in

membrane fissioning) (Weigert et al., 1999). This coincides with the onset of

22

mitosis that requires organelle partitioning between daughter cells. This
activity of BARS is thought to lead to accumulation of phosphatidic acid in the
Golgi membrane, enhancing the curvature of the phospholipid bilayer and
thereby causing disassembly of the Golgi network (Corda et al., 2006). To test
the role of BARS in this process, BARS activity was reduced in rat kidney
cells using dominant negative mutants or by mRNA depletion using antisense
oligos. These treatments prevented Golgi partitioning and arrested the cell
cycle at 62. Both Golgi fragmentation and the onset of cell cycle were
restored upon addition of recombinant BARS (Hidalgo Carcedo et al., 2004).
In support of this data, another study found that in mitotic cells, human
CtBP becomes associated with centrosomes during mitosis. Golgi
morphogenesis is initiated and regulated within centrosomes, implicating
CtBP in this process (Spyer and Allday, 2006). A more recent study re-
evaluated the acyl transferase activity associated with BARS and claimed it to
be a contaminant associated with the recombinant BARS protein (Gallop et
al., 2005). lntriguingly, dual Ctbp1 and Ctbp2 mouse knockouts are embryonic
lethal but no Golgi defects are observed and MEFs can be cultured from
these animals with no visible Golgi defects. Clearly, much remains to be

proven concerning the role of CtBP in Golgi fission.

23

1.6 CtBP as a corepressor during Drosophila development

The Drosophila blastoderrn embryo serves as a premier model for the
study of eukaryotic transcriptional regulation, providing an amenable
background for the study of both cis- and trans-regulatory elements. Early
embryonic development in Drosophila is controlled by a cascade of genes that
encode for transcription factors. The cascade is turned on by preformed
mRNAs and proteins that are synthesized by the mother and deposited in the
egg prior to fertilization. The maternally derived proteins set up spatially
distributed morphogenetic gradients that establish the antero-posterior (A-P)
and dorso-ventral axes. These morphogen proteins can then activate or
repress zygotic genes at specific positions along both axes and thereby
specify patterning along the embryo (St Johnston and Nusslein-Volhard,
1992).

Gap genes are the first category of zygotically transcribed genes,
expressed in broad expression domains. They were identified based on
phenotypes in a genetic screen looking for mutants lacking large sections of
abdominal segments along the A-P axis (Nusslein-Volhard and Wieschaus,
1980). Gap genes encode well-characterized repressor proteins whose
primary function is to refine the patterns of pair-rule genes, the next class of
genes in the segmentation cascade. Their repression activity is described as
short-range, acting at distances of up to 100-150 bp to inhibit the basal
promoter or upstream bound activators (Gray and Levine, 1996; Nibu et al.,
2003). This is in sharp contrast to long-range repressors that globally

suppress transcription over several kilobases (Cai et al., 1996). Short-range

24

repressors employ CtBP as a corepressor, while long-range repressors use
the Groucho corepressor. It has been long believed that Groucho and CtBP
represent two distinct pathways of repression that form the basis for this so
called ‘range' of repressor activity.

Pair-rule genes are expressed in seven transverse stripes along the
anterior-posterior axis and in turn, they control the fourteen-stripe pattern of
the subsequently expressed segment polarity genes that delineate the future
segments of the fly. Identification of these zygotic gene products involved in
body plan determination of a Drosophila embryo is the outcome of large
genetic screens that searched for embryonic lethal mutations.

CtBP plays a vital role in this hierarchical organization by contributing
to the repression abilities of short-range repressors encoded by gap genes.
The first thread of evidence linking CtBP to short-range repression came from
yeast two-hybrid assays, in which Knirps and Snail, two short-range
repressors were used as bait to screen an embryonic cDNA library (Nibu et
al., 1998b). Both proteins selected Drosophila CtBP cDNA, specifying a
putative protein of 383 residues and exhibiting high sequence similarity with
the human CtBP. The CtBP gene is expressed during early oogenensis and
the message is deposited in the egg prior to fertilization. Following zygotic
induction, CtBP proteins can be traced throughout development to be
ubiquitously and uniformly expressed in the fly (See Figure l-3).

It has been demonstrated that CtBP is essential for short-range
transcription during development by analyzing embryos lacking maternal
CtBP. dCtBP is encoded by a single gene whose locus has been mapped

cytologically to 87D8—9 on the third chromosome. P1590 is a homozygous

25

lethal P-element insertion within the CtBP coding region that results in
homozygotes dying as pharate adults (i.e. still within pupal case). Mutant
embryos containing reduced CtBP levels demonstrate a host of patterning
defects that can be attributed to impaired activity of short-range repressor
proteins like Knirps, Kriippel and Giant. In situ hybridization assays show that
the expression of these repressors is mostly normal in CtBP mutants, but the
expression of their pair-rule targets such as even-skipped and fushi-tarazu is
severely disturbed (Nibu et al., 1998a). Like vertebrate repressors, short-
range repressors in Drosophila recruit CtBP via a PXDLS motif. Mutations in
this motif compromise the activity of these proteins, indicating that CtBP-
dependent repression represents a major form of transcriptional regulation
during development. A direct transcriptional repression activity of dCtBP has
also been shown by Gal4-CtBP in a tethering assay wherein CtBP can
repress nearby transcriptional activators but not the basal promoter elements
(Nibu et al., 1998b; Sutrias-Grau and Arnosti, 2004). This tethering assay has
been exploited further to question the relevance of CtBP’s homology to
dehydrogenases. Mutations made in the putative catalytic histidine are found
not to compromise CtBP corepressor function while mutations in the
nucleotide-binding fold destroy its ability to obstruct repression on integrated
reporter genes (Sutrias-Grau and Arnosti, 2004). A protein interaction map in

Drosophila emphasizes that CtBP forms a major transcriptional regulatory

26

Figure l-3: Drosophila CtBP proteins are ubiquitously distributed during
development

Stage specific embryos were collected and imaginal discs were harvested
from third instar larvae. Samples were fixed in a heptane-formaldehyde
solution as per lab protocol and immunostained with anti-CtBP rabbit
polyclonal serum (1:500) (F) embryo stained with rabbit preimmune, (B-E)
embryos from progressive stages during development, (G-l) wing, eye and leg
imaginal discs extracted from late third instar larvae (P.Mani, unpublished).

27

 

 

 

28

network where each interactor contains a discernible P-DLS like motif (Giot et

aL,2003)

CtBP and long-range repression by Hairy

As typified by the Hairy protein in Drosophila, long-range repressors
can function over many kilobases to block transcription, in sharp contrast to
short-range repressors. The mechanisms underlying these modes of
repression are unclear (Courey and Jia, 2001 ). Possible clues have arisen
from the observations that short-range repressors mediate repression by the
recruitment of CtBP, while long-range repressors recruit a different
corepressor, Groucho. These corepressor may mediate repression by distinct
pathways (Chen and Courey, 2000). However, both CtBP and Groucho have
been proposed to utilize chromatin remodeling mechanisms through the
recruitment of histone deacetylases like de3 (Chen et al., 1999; Struffi and
Arnosti, 2005). The long-range repressor Hairy also shows a weak interaction
with CtBP and has a canonical CtBP binding motif. The removal of the weak
dCtBP interaction motif (PLSLV) does not impair Hairy-mediated repression of
its targets. Instead, removal of this motif augments Hairy function. CtBP and
Groucho binding to Hairy may be antagonistic because just nine amino acid
residues separate the CtBP binding P-SLV-K and Groucho binding WRPW
motifs. Mechanistically speaking, when dCtBP and Groucho both bind, they
might be unable to interact with additional corepressors or contact their target
proteins in the core transcription complex (Zhang and Levine, 1999). Such

antagonism is supported by genetic studies, which suggest that lowering the

29

dose of maternal dCtBP products can partially suppress the embryonic
phenotypes of hairy mutants while reducing groucho can enhance it (Phippen

et al., 2000; Poortinga et al., 1998).

The simplest interpretation of these observations is that the dCtBP and
Groucho interfere with one another when both are bound to Hairy, but the
rules governing the choice of cofactor recruited are undetermined. Another
cofactor of Hairy is the Sir2 protein that encodes an NAD+ -dependent histone
deacetylase. A global chromatin profiling study looking at genomic recruitment
of Hairy and its cofactors demonstrates that while Groucho is believed to be
the major cofactor, it is found to be associated with only subsets of Hairy
targets. Surprisingly, CtBP and Sir2 are predominantly found at Hairy targets
with their association being largely overlapping and distinct from Groucho

(Bianchi-Frias et al., 2004).

CtBP as a corepressor for signaling pathways in the fly

Numerous signaling pathways are seen to converge upon general
transcription coregulators. Decapentaplegic (dpp) encodes a TGF-fl homolog
that functions as a long-range signaling morphogen, specifying cell fates, in a
dose-dependent manner. A response to thresholds of Dpp proteins is
regulated by an opposing gradient of Brinker, which is a dpp-target and a
transcriptional repressor. Brinker’s repression domain contains binding sites
for both CtBP and Groucho corepressors and physical interactions have been

established. For Brinker to silence its endogenous targets, Groucho alone is

30

sufficient. For simpler targets such as Brinker autoregulation, both
corepressors can function interchangeably and adequately. Like Hairy, in the
absence of Groucho, Brinker’s repressive capacity is impaired and unlike
Hairy, CtBP’s absence also seems to affect Brinker output to a smaller extent
(Hasson et al., 2001). These results suggest that Brinker uses multiple
modes of repression (provided by CtBP and Groucho) that provide it with a
flexibility to silence a variety of genes in response to a gradation in Dpp
activity during development.

The Notch signaling pathway is employed in a variety of cell fate
decisions during development. The Notch signal is transduced via Su(H)
[Suppressor of Hairless] that acts as a repressor in the absence of Notch
signal and repression is critical for appropriate cell fate specification. In the
presence of Notch signaling, Su(H) is an activator. Hairless (H) acts as a
corepressor for Su(H) by antagonizing its activation via the recruitment of
CtBP and Groucho. Reduction in levels of either CtBP or Groucho enhances
Hairless loss of function phenotypes, and mutating Hairless’s CtBP binding
motif debilitates its activity, suggesting that both corepressors are vital
components of this signaling pathway (Barolo and Posakony, 2002; Barolo et
al., 2002).

Regulation of the Wnt/ ,B-catenin signaling cascade is also utilized in
cell fate determination and oncogenesis. Constitutive induction of this
pathway is observed in many cancers, especially in colorectal carcinoma. In
the absence of Wnt signal, TCFs (T-cell factor) function as repressors of Wnt
target genes, similar to the functionings of the Notch pathway. In Xenopus,

thBP is known to act as a corepressor for xTCF-3 (Brannon et al., 1999). In

31

the fly, TCF’s are thought to recruit Groucho for repression and when Wnt
signaling is on, binding of B-catenin displaces Groucho and repression is
relieved. CtBP can bind to vertebrate TCF’s, and CtBP overexpression inhibits
TCF mediated activation of Wnt targets, but no direct interaction of TCF and
CtBP has been shown. One model suggests that CtBP antagonizes Wnt
pathway by binding to APO tumor suppressor (Adenomatous polyposis coli)
that causes it to sequester B-catenin away from TCFs (Hamada and Bienz,
2004). Some support for this model comes from a study where APC and CtBP
can be traced on a Wnt target and this coincides with a loss of B-catenin from
TCFs (Sierra et al., 2006). As colorectal cancer cells all have mutations in
APC proteins, the functional interaction of APC and CtBP suggests a role for
CtBP as a tumor suppressor. In contrast, another model suggests CtBP to be
recruited by an unknown DNA binding protein and act in parallel to TCF to

repress Wnt targets (Fang et al., 2006).

CtBP- independent repression activity

Multiple repression activities may assist quantitatively or qualitatively in
regulating gene expression. Three general mechanisms of transcriptional
repression in eukaryotes include competition, where repressors directly
compete with activators for a common cis-element, and blocking the basal
transcriptional machinery or direct repression. The third is by quenching,
whereby repressors repress activators co-occupying adjacent sites.

Chromatin modifications that reduce accessibility of transcription factors to the

32

DNA template may involve covalent modifications of histones or DNA (Strahl
and Allis, 2000). In addition, cis—acting elements named silencers or boundary
elements may prevent transcription of a gene by preventing crosstalk when
located between a promoter and enhancer (Gerasimova and Corces, 2001 ).

Embryo repression assays performed with short-range repressors
Knirps, Giant and Kriippel all provide basis for CtBP-dependent and CtBP-
independent modes of repression. Knirps has a repression domain towards
the C-terminus that encodes a PXDLS like motif important to bind CtBP. In
addition, an N-tenninal region has also been mapped that is active for
repression and does not contain any canonical CtBP binding motifs. This
activity is evident in experiments in which the CtBP corepressor is absent. In
transgenic embryos containing integrated reporters and lacking CtBP, Knirps
can still repress its endogenous target, the eve stripe 3 enhancer (Keller et al.,
2000)

In case of Krilppel, when its binding sites do not overlap but are
adjacent to activator binding sites, repression is dependent on CtBP by
quenching. Quenching might involve hindering activator-basal machinery
crosstalk or it might invoke the intrinsic dehydrogenase activity ascribed to
CtBP to remodel chromatin. However, Krinpel can repress by competition
without any help from CtBP, when the activator sites overlap with Kriippel
repressor binding sites, indicating a CtBP-independent mode of repression
(Nibu et al., 2003). Giant is also able to effectively regulate hunchback while
its other targets like eve stripe 2 are derepressed in a CtBP mutant
background proposing that Giant too can function via CtBP-independent

pathways (Strunk et al., 2001). The relevance of these multiple activities was

33

put to test in the case of Knirps. It was found that supplying CtBP-independent
activity at higher than regular levels could repress a target that normally
requires CtBP, suggesting that CtBP contributes in a quantitative fashion to

boost repressor output (Struffi et al., 2004).

1.7 Alternative splicing of Drosophila CtBP

In the fly embryo, multiple CtBP transcripts are expressed ubiquitously
throughout development that encode polypeptides of 383, 386, 476 and 479
amino acids (See Figure M). The first 376 amino acids required for
interaction with short-range repressors are common to all four isoforms
(essentially composed of only the dehydrogenase-like domain). Any
distinctions in the function of individual isoforms is yet unidentified. These
distinct isoforms are produced from alternative splicing that use alternative
donor / acceptor sites and differ largely in the presence or absence of a C-
ten'ninal tail of ~90 amino acids of unidentified significance. The 476 and 479
aa isoforms both include the C-terminal tail and differ in the alternate use of
the splice acceptor that results in the presence/absence of an ‘LNGGYYT’
motif at the start of exon 6 and a ‘VSSQS' motif at the start of exon 7 (Also
see Chapter II). Both the 383 and 386 aa isoforms lack the C-terminal
extension and differ in the inclusion of a ’VFQ’ motif by resorting to different
splice acceptors at the start of exon 5. It is very likely that these different

isoforms might function in a tissue or stage specific manner with

34

Figure l-4: Alignment of different CtBP Isoforrns generated by alternative
splicing in Drosophila

The Drosophila CtBP gene encodes differentially spliced products that are
divergent at their COOH-terminal regions (conserved residues are shaded in
yellow). Two isoforms [AY060646 (479aa) and NM_001014617 (476 3a)]
include an unstructured C-terminal extension of ~90 aa of unknown function
and differ in the alternative use of a splice donor that results in the inclusion of
‘LNGGYYT’ motif in the 479 aa isoform and an inclusion of a ‘VSSQS’ motif in
the 476 aa isoform. The two shorter isoforms [AY069170 and ABO11840] are
essentially composed of just the dehydrogenase-like domains and do not
include the C-terrninal tail. These also differ in the use of a splice donor that
results in the inclusion of a ‘VFQ' motif in AY069170. The red arrow indicates
a conserved splice donor that results in the generation of the longer isoforms.
This splice donor represented by a conserved glycine codon within exon 5
that is not subject to silent mutations in several organisms, supporting the
notion that isoforms with tails are functionally important.

The experiments in Chapter II, III and Appendix A were performed using the
479 and 383 aa isoforms obtained from Yutaka Nibu, Weill Medical College of
Cornell University, New York.

35

 

 

EEHEEESH§ESH§3§=§>§ a . I . N. :30
gaugHflnEngfi .. .. .. .. .. H3A§§Efl§>§ o coouohd

H2BgagmfisgfigmH6>HSHE§HHH6¢2§BH§ 3:32

HoneHggafiggfissaugfigmH6>H§HHHH¢§HHH662§aBHHz 2.2.82
«engage—H6835; ........ asgggfiasgfigaH6>H6~Huagfla662§a85z 7...». £38
aggufiofimfiugflgasfigggmggfigH6>§HHH§HE662§EHE .3322

H

9356-- -zuggggfiggzsa ans—2H HHHLE. :32? Egg 6.8662 .53 3.324
goém Haggis—Hm H:eo§§am 85 8:633 26.. ..HHBG 5223:2263 262 92H .82 :E 5322 3:23 2.22:2
226959.. :2 22H :8 326299662? :H6HBBH H:quHBEwmz.HHb= 6:2 as: 3:5 .HSBHon 6:3 1...: .836
22462.55 655366 :23 agaifig :H6aaaH u§:§§ :2 :0 flgfia H33. 23.. .58 333.2

2HE§>66630>§¢§0§2HE§§Hiya—«562553663366 3.22:
2Engages6326285603226265655HHoEQEnBBHEEHSuogHEo 2.22:2
22225236HHSoSoHSeHHfiuegHu—Sfifignggg HHoEQBEEHEZHSHogHzgo 7...: £30
EH§§§66H¢3§6§¢H§§2§§§§HgguogiHofieaBHfi—Bme $335

HgHHxOHgHgan—MHBHHH HHH§§§ ”Mam §>B§H3Hu§>a gagngdagn—l ONE—Hon“
HgHH “OH gH Haifa: :5 EH” BOO‘ flethGEAHiH :UGMOn—dqd .323 AOENBOEH Haggai a;

HgHHu Hon—“Hg :gggnHHggg §:‘ :EHgHmoanna4 :mmdg mOahmegHumfimI-aanl : h. ‘30
HgHH :OH gH :flgggaH H§>§ HHMHOO‘BH‘EEAHSHaUn—MOQAA‘» :3 aOEHIOEH “nag ovoouo:

36

the splicing machinery favoring certain splice sites. CtBPL is generated from
splicing originating within exon5 that joins core CtBP sequences (exons 1-5)
to that of exon 6 and 7.

In this thesis, we study the expression and activities of two of these
isoforms that we designate as CtBPL (479 aa) and CtBPs (383 aa)
respectively. These two splice variants are found to mediate effective
repression on reporter genes when assayed as Gal4 fusion proteins both in
the cultured cells and in the embryo (Ryu and Arnosti, 2003; Sutrias-Grau and
Amosti, 2004). This prompts an understanding of whether CtBP isoforms in
Drosophila are functionally equivalent or have distinct roles in vivo. Chordate
genomes like those of human, mouse and Ciona are seen to contain only
CtBPL-Iike isoforms (not shown). CtBPs isoforms have only been described
thus far in Drosophila, raising significant questions concerning the function of

forms of this protein that do not contain the unstructured C-terminal extension.

37

REFERENCES

Arnosti, D. N. (2004). Multiple Mechanisms of Transcriptional Repression in
Eukaryotes. In Handbook of Experimental Pharmacology, vol. 166 (ed. K. J. a.
T. S. J. Gossen M.): Springer(Berlin).

Austin, R. J. and Biggin, M. D. (1995). A domain of the even-skipped protein
represses transcription by preventing TFIID binding to a promoter: repression
by cooperative blocking. Mol Cell Biol 15, 4683-93.

Balasubramanian, P., Zhao, L. J. and Chinnadurai, G. (2003). Nicotinamide
adenine dinucleotide stimulates oligomerization, interaction with adenovirus
E1A and an intrinsic dehydrogenase activity of CtBP. FEBS Lett 537, 157-60.

Barnes, C. J., Vadlamudl, R. K., Mishra, S. K., Jacobson, R. H., Li, F. and
Kumar, R. (2003). Functional inactivation of a transcriptional corepressor by a
signaling kinase. Nat Struct Biol 10, 622-8.

Barolo, S. and Posakony, J. W. (2002). Three habits of highly effective
signaling pathways: principles of transcriptional control by developmental cell
signaling. Genes Dev 16, 1167-81.

Barolo, 8., Stone, T., Bang, A. G. and Posakony, J. W. (2002). Default
repression and Notch signaling: Hairless acts as an adaptor to recruit the
corepressors Groucho and dCtBP to Suppressor of Hairless. Genes Dev 16,
1964-76.

Bianchl-Frias, D., Orian, A., Delrow, J. J., Vazquez, J., Rosales-Nieves, A.
E. and Parkhurst, S. M. (2004). Hairy transcriptional repression targets and
cofactor recruitment in Drosophila. PLoS Biol 2, E178.

Boyd, J. M., Subramanian, T., Schaeper, U., La Regina, M., Bayley, S. and
Chinnadurai, G. (1993). A region in the C-terminus of adenovirus 2/5 E1a
protein is required for association with a cellular phosphoprotein and
important for the negative modulation of T24-ras mediated transformation,
tumorigenesis and metastasis. Embo J 12, 469-78.

Brannon, M., Brown, J. D., Bates, R., Kimelman, D. and Moon, R. T.
(1999). XCtBP is a Xch-3 oo-repressor with roles throughout Xenopus
development. Development 126, 3159-70.

Cal, H. N., Arnosti, D. N. and Levine, M. (1996). Long-range repression in
the Drosophila embryo. Proc Natl Acad Sci U S A 93, 9309-14.

Chen, G. and Courey, A. J. (2000). Groucho/1' LE family proteins and
transcriptional repression. Gene 249, 1-16.

38

Chen, G., Fernandez, J., Mische, S. and Courey, A. J. (1999). A functional
interaction between the histone deacetylase de3 and the corepressor
groucho in DrOSOphila development. Genes Dev 13, 2218-30.

Chinnadurai, G. (2002). CtBP, an unconventional transcriptional corepressor
in development and oncogenesis. Mol Cell 9, 213-24.

Chinnadurai, G. (2006). CtBP family proteinszunique transcriptional
regulators in the nucleus with diverse cytosolic functions: Landes Bioscience.

Corda, D., Colanzl, A. and Luinl, A. (2006). The multiple activities of
CtBP/BARS proteins: the Golgi view. Trends Cell Biol 16, 167-73.

Courey, A. J. and Jia, S. (2001). Transcriptional repression: the long and the
short of it. Genes Dev 15, 2786-96.

Criqui-Filipe, P., Ducret, C., Maira, S. M. and Wasylyk, B. (1999). Net, a
negative Ras-switchable TCF, contains a second inhibition domain, the CID,
that mediates repression through interactions with CtBP and de-acetylation.
Embo J 18, 3392-403.

Deconinck, A. E., Mead, P. E., Tevosian, S. G., Crlsplno, J. D., Katz, S. G.,
Zon, L. I. and Orkln, S. H. (2000). FOG acts as a repressor of red blood cell
development in Xenopus. Development 127, 2031 -40.

Deltour, S., Pinte, S., Guerardel, G., Wasylyk, B. and Leprince, D. (2002).
The human candidate tumor suppressor gene HIC1 recruits CtBP through a
degenerate GLDLSKK motif. Mol Cell Biol 22, 4890-901.

Fang, M., Li, J., Blauwkamp, T., Bhambhani, C., Campbell, N. and
Cadlgan, K. M. (2006). C-terrninal-binding protein directly activates and
represses Wnt transcriptional targets in Drosophila. Embo J 25, 2735-45.

Field, C. C., Birdsong, W. T. and Goodman, R. H. (2003). Differential
binding of NAD+ and NADH allows the transcriptional corepressor carboxyl-
terminal binding protein to serve as a metabolic sensor. Proc Natl Acad Sci U
S A 100, 9202-7.

Furusawa, T., Moribe, H., Kondoh, H. and Higashl, Y. (1999). Identification
of CtBP1 and CtBP2 as corepressors of zinc finger-homeodomain factor
deltaEF1. Mol Cell Biol 19, 8581-90.

Gallop, J. L., Butler, P. J. and McMahon, H. T. (2005). Endophilin and
CtBP/BARS are not acyl transferases in endocytosis or Golgi fission. Nature
438, 675-8.

Garrlga-Canut, M., Schoenlke, B., Qazl, R., Bergendahl, K., Daley, T. J.,
Pfender, R. M., Morrison, J. F., Ockuly, J., Stafstrom, C., Sutula, T. et al.
(2006). 2-Deoxy-D-glucose reduces epilepsy progression by NRSF-CtBP-
dependent metabolic regulation of chromatin structure. Nat Neurosci.

39

Gerasimova, T. I. and Corces, V. G. (2001). Chromatin insulators and
boundaries: effects on transcription and nuclear organization. Annu Rev
Genet 35, 193-208.

Glot, L., Bader, J. S., Brouwer, C., Chaudhurl, A., Kuang, 8., Li, Y., Hao,
Y. L., Ool, C. E., Godwin, B., Vltols, E. et al. (2003). A protein interaction
map of Drosophila melanogaster. Science 302, 1727-36.

Gray, S. and Levine, M. (1996). Transcriptional repression in development.
Curr Opin Cell Biol 8, 358-64.

Grooteclaes, M., Deveraux, G., Hildebrand, J., Zhang, 0., Goodman, R. H.
and Frlsch, S. M. (2003). C-terminal-binding protein corepresses epithelial
and proapoptotic gene expression programs. Proc Natl Acad Sci U S A 100,

4568-73.

Hamada, F. and Bienz, M. (2004). The APC tumor suppressor binds to C-
terrninal binding protein to divert nuclear beta-catenin from TCF. Dev Cell 7,

677-85.

Hasson, P., Muller, B., Basler, K. and Paroush, Z. (2001). Brinker requires
two corepressors for maximal and versatile repression in Dpp signalling.
Embo J 20. 5725-36.

Hldalgo Carcedo, C., Bonazzi, M., Spano, S., Turacchlo, G., Colanzl, A.,
Luinl, A. and Corda, D. (2004). Mitotic Golgi partitioning is driven by the
membrane-fissioning protein CtBP3/BARS. Science 305, 93-6.

Hildebrand, J. D. and Soriano, P. (2002). Overlapping and unique roles for
C-terrninal binding protein 1 (CtBP1) and CtBP2 during mouse development.
Mol Cell Biol 22, 5296-307.

Imal, 8., Armstrong, C. M., Kaeberlein, M. and Guarente, L. (2000).
Transcriptional silencing and longevity protein Sir2 is an NAD-dependent
histone deacetylase. Nature 403, 795-800.

Katsanls, N. and Fisher, E. M. (1998). A novel C-terrninal binding protein
(CTBP2) is closely related to CTBP1, an adenovirus E1A-binding protein, and
maps to human chromosome 21q21.3. Genomics 47, 294-9.

Keller, S. A., Mao, Y., Struffi, P., Margulies, C., Yurk, C. E., Anderson, A.
R., Amey, R. L., Moore, 8., Ebels, J. M., Foley, K. et al. (2000). dCtBP-
dependent and -independent repression activities of the Drosophila Knirps
protein. Mol Cell Biol 20, 7247-58.

Kim, J. H., Cho, E. J., Kim, S. T. and Youn, H. D. (2005). CtBP represses
p300-mediated transcriptional activation by direct association with its
bromodomain. Nat Struct Mol Biol 12, 423-8.

Koipally, J. and Georgopoulos, K. (2000). lkaros interactions with CtBP
reveal a repression mechanism that is independent of histone deacetylase
activity. J Biol Chem 275, 19594-602.

40

Kumar, V., Carlson, J. E., Ohgl, K. A., Edwards, T. A., Rose, D. W.,
Escalante, C. R., Rosenfeld, M. G. and Aggarwal, A. K. (2002).
Transcription corepressor CtBP is an NAD(+)-regulated dehydrogenase. Mol
Cell 10, 857-69.

Lawrence, P. A. (1992). The Making of a Fly: Blackwell Sci. Pub. (Oxford).

Levine, S. 8., Weiss, A., Erdjument-Bromage, H., Shao, Z., Tempst, P.
and Kingston, R. E. (2002). The core of the polycomb repressive complex is
compositionally and functionally conserved in flies and humans. Mol Cell Biol
22, 6070-8.

Li, S., Chen, P. L., Subramanian, T., Chinnadurai, G., Tomlinson, G.,
Osborne, C. K., Sharp, 2. D. and Lee, W. H. (1999). Binding of Cth to the
BRCT repeats of BRCA1 involved in the transcription regulation of p21 is
disrupted upon DNA damage. J Biol Chem 274, 11334-8.

Lin, X., Sun, 8., Liang, M., Liang, Y. Y., Gast, A., Hlldebrand, J.,
Brunlcardi, F. C., Melchior, F. and Feng, X. H. (2003). Opposed regulation
of corepressor CtBP by SUMOylation and PDZ binding. Mol Cal/11, 1389-96.

Long, J., Zuo, D. and Park, M. (2005). Pc2-mediated sumoylation of Smad-
interacting protein 1 attenuates transcriptional repression of E-cadherin. J Biol
Chem 280, 35477-89.

Lunyak, V. V., Burgess, R., Prefontaine, G. G., Nelson, C., Sze, S. H.,
Chenoweth, J., Schwartz, P., Pevzner, P. A., Glass, C., Mandel, G. et al.
(2002). Corepressor-dependent silencing of chromosomal regions encoding
neuronal genes. Science 298, 1747-52.

Mannervik, M., Nibu, Y., Zhang, H. and Levine, M. (1999). Transcriptional
coregulators in development. Science 284, 606-9.

Melhulsh, T. A. and Wotton, D. (2000). The interaction of the carboxyl
terminus-binding protein with the Smad corepressor TGIF is disrupted by a
holoprosencephaly mutation in TGIF. J Biol Chem 275, 39762-6.

Meloni, A. R., Smith, E. J. and Nevins, J. R. (1999). A mechanism for
Rb/p130-mediated transcription repression involving recruitment of the CtBP
corepressor. Proc Natl Acad Sci U S A 96, 9574-9.

Molloy, D., Mapp, K. L., Webster, R., Galllmore, P. H. and Grand, R. J.
(2006). Acetylation at a lysine residue adjacent to the CtBP binding motif
within adenovirus 12 E1A causes structural disruption and limited reduction of
CtBP binding. Virology.

Molloy, D. P., Barral, P. M., Bremner, K. H., Galllmore, P. H. and Grand, R.
J. (2001). Structural determinants outside the PXDLS sequence affect the
interaction of adenovirus E1A, C-terrninal interacting protein and Drosophila
repressors with C-terminal binding protein. Biochim Biophys Acta 1546, 55-

70.

41

Nardini, M., Spano, S., Cericola, C., Pesce, A., Massaro, A., Millo, E.,
Luinl, A., Corda, D. and Bolognesl, M. (2003). CtBP/BARS: a dual-function
protein involved in transcription co-repression and Golgi membrane fission.
Embo J 22, 3122-30.

Nardini, M., Svergun, D., Konarev, P. V., Spano, S., Fasano, M., Bracco,
C., Pesce, A., Donadini, A., Cericola, C., Secundo, F. et al. (2006). The C-
terrninal domain of the transcriptional corepressor CtBP is intrinsically
unstructured. Protein Sci 15, 1042-50. '

Nibu, Y., Senger, K. and Levine, M. (2003). CtBP-independent repression in
the Drosophila embryo. Mol Cell Biol 23, 3990-9.

Nibu, Y., Zhang, H., Bajor, E., Barolo, 8., Small, S. and Levine, M. (19983).
dCtBP mediates transcriptional repression by Knirps, Kruppel and Snail in the
Drosophila embryo. Embo J 17, 7009-20.

Nibu, Y., Zhang, H. and Levine, M. (1998b). Interaction of short-range
repressors with Drosophila CtBP in the embryo. Science 280, 101-4.

Nusslein-Volhard, C. and Wieschaus, E. (1980). Mutations affecting
segment number and polarity in Drosophila. Nature 287, 795-801.

Palmer, 5., Broulllet, J. P., Kilbey, A., Fulton, R., Walker, M., Crossley, M.
and Bartholomew, C. (2001). Evi-1 transforming and repressor activities are
mediated by CtBP co-repressor proteins. J Biol Chem 276, 25834-40.

Perdomo, J. and Crossley, M. (2002). The lkaros family protein Eos
associates with C-terminal-binding protein corepressors. Eur J Biochem 269,

5885-92.

Phippen, T. M., Swelgart, A. L., Moniwa, M., Krumm, A., Davie, J. R. and
Parkhurst, S. M. (2000). Drosophila C-terminal binding protein functions as a
context-dependent transcriptional co-factor and interferes with both mad and
groucho transcriptional repression. J Biol Chem 275, 37628-37.

Poortinga, G., Watanabe, M. and Parkhurst, S. M. (1998). Drosophila CtBP:
a Hairy-interacting protein required for embryonic segmentation and hairy-
mediated transcriptional repression. Embo J 17, 2067-78.

Quinlan, K. G., Verger, A., Kwok, A., Lee, S. H., Perdomo, J., Nardini, M.,
Bolognesl, M. and Crossley, M. (2006). The role of the C-terminal binding
protein PXDLS motif binding cleft in protein interactions and transcriptional
repression. Mol Cell Biol.

Rutter, J., Relck, M., Wu, L. C. and McKnight, S. L. (2001). Regulation of
clock and NPASZ DNA binding by the redox state of NAD cofactors. Science
293, 510-4.

Ryu, J. R. and Arnosti, D. N. (2003). Functional similarity of Knirps CtBP-
dependent and CtBP-independent transcriptional repressor activities. Nucleic
Acids Res 31, 4654-62.

42

Schaeper, U., Boyd, J. M., Verma, S., Uhlmann, E., Subramanian, T. and
Chinnadurai, G. (1995). Molecular cloning and characterization of a cellular
phosphoprotein that interacts with a conserved C-terminal domain of
adenovirus E1A involved in negative modulation of oncogenic transformation.
Proc Natl Acad Sci U S A 92, 10467-71.

Schmitz, F., Konlgstorfer, A. and Sudhof, T. C. (2000). RIBEYE, a
component of synaptic ribbons: a protein's journey through evolution provides
insight into synaptic ribbon function. Neuron 28, 857-72.

Senyuk, V., Sinha, K. K. and Nucifora, G. (2005). Corepressor CtBP1
interacts with and specifically inhibits CBP activity. Arch Biochem Biophys
441, 168-73.

Sewalt, R. G., Gunster, M. J., van der Vlag, J., Satijn, D. P. and Otte, A. P.
(1999). C-Terminal binding protein is a transcriptional repressor that interacts
with a specific class of vertebrate Polycomb proteins. Mol Cell Biol 19, 777-

87.

Shl, Y., Lan, F., Matson, C., Mulligan, P., Whetstlne, J. R., Cole, P. A.,
Casero, R. A. and Shl, Y. (2004). Histone demethylation mediated by the
nuclear amine oxidase homolog LSD1. Cal/119, 941-53.

Shl, Y., Sawada, J., Sui, G., Affar el, 8., Whetstlne, J. R., Lan, F., Ogawa,
H., Luke, M. P., Nakatani, Y. and Shl, Y. (2003). Coordinated histone
modifications mediated by a CtBP co-repressor complex. Nature 422, 735-8.

Shl, Y. J., Matson, C., Lan, F., lwase, S., Baba, T. and Shl, Y. (2005).
Regulation of LSD1 histone demethylase activity by its associated factors. Mol
Cell 19, 857-64.

Sierra, J., Yoshida, T., Joazeiro, C. A. and Jones, K. A. (2006). The APC
tumor suppressor counteracts beta-catenin activation and H3K4 methylation
at Wnt target genes. Genes Dev 20. 586-600.

Sollerbrant, K., Chinnadurai, G. and Svensson, C. (1996). The CtBP
binding domain in the adenovirus E1A protein controls CR1-dependent
transactivation. Nucleic Acids Res 24, 2578-84.

Spano, S., Silletta, M. G., Colanzl, A., Alberti, S., Fiuccl, G., Valente, C.,
Fusella, A., Salmona, M., Mironov, A., Luinl, A. et al. (1999). Molecular
cloning and functional characterization of brefeldin A-ADP-ribosylated
substrate. A novel protein involved in the maintenance of the Golgi structure.
J Biol Chem 274, 17705-10.

Spyer, M. and Allday, M. J. (2006). The transcriptional co-repressor C-
terrninal binding protein (CtBP) associates with centrosomes during mitosis.
Cell Cycle 5, 530-7.

St Johnston, D. and Nussleln-Volhard, C. (1992). The origin of pattern and
polarity in the Drosophila embryo. Cell 68, 201-19.

43

Strahl, B. D. and Allis, C. D. (2000). The language of covalent histone
modifications. Nature 403, 41-5.

Struffl, P. and Arnosti, D. N. (2005). Functional interaction between the
Drosophila knirps short range transcriptional repressor and RPD3 histone
deacetylase. J Biol Chem 280, 40757-65.

Struffl, P., Corado, M., Kulkarni, M. and Arnosti, D. N. (2004). Quantitative
contributions of CtBP-dependent and -independent repression activities of
Knirps. Development 131, 2419-29.

Strunk, B., Struffl, P., Wright, K., Pabst, B., Thomas, J., an, L. and
Arnosti, D. N. (2001). Role of CtBP in transcriptional repression by the
Drosophila giant protein. Dev Biol 239, 229-40.

Sundqvist, A., Sollerbrant, K. and Svensson, C. (1998). The carboxy-
terrninal region of adenovirus E1A activates transcription through targeting of
a C-terrninal binding protein-histone deacetylase complex. FEBS Lett 429,

183-8.

Sutrias-Grau, M. and Arnosti, D. N. (2004). CtBP contributes quantitatively
to Knirps repression activity in an NAD binding-dependent manner. Mol Cell
Biol 24, 5953-66.

Turner, J. and Crossley, M. (2001). The CtBP family: enigmatic and
enzymatic transcriptional co-repressors. Bioessays 23, 683-90.

Van Hateren, N., Shenton, T. and Borycki, A. G. (2006). Expression of
avian C-terminal binding proteins (Ctbp1 and Ctbp2) during embryonic
development. Dev Dyn 235, 490-5.

van Vliet, J., Turner, J. and Crossley, M. (2000). Human Kruppel-like factor
8: a CACCC-box binding protein that associates with CtBP and represses
transcription. Nucleic Acids Res 28, 1955-62.

Verger, A., Quinlan, K. G., Crofts, L. A., Spano, S., Corda, D., Kable, E. P.,
Braet, F. and Crossley, M. (2006). Mechanisms directing the nuclear
localization of the CtBP family proteins. Mol Cell Biol 26, 4882-94.

Vo, N., Field, C. and Goodman, R. H. (2001). Acetylation of nuclear
hormone receptor-interacting protein RlP140 regulates binding of the
transcriptional corepressor CtBP. Mol Cell Biol 21, 6181-8.

Wang, S. Y., Iordanov, M. and Zhang, Q. (2006). c—Jun NH2-terminal kinase
promotes apoptosis by down-regulating the transcriptional co-repressor CtBP.
J Biol Chem.

Welgert, R., Silletta, M. G., Spano, S., Turacchlo, G., Cericola, C., Colanzl,
A., Senatore, S., Mancini, R., Polishchuk, E. V., Salmona, M. et al. (1999).
CtBP/BARS induces fission of Golgi membranes by acylating
lysophosphatidic acid. Nature 402, 429-33.

44

Wray, G. A. (2003). Transcriptional regulation and the evolution of
development. Int J Dev Biol 47, 675-84.

Zhang, H. and Levine, M. (1999). Groucho and dCtBP mediate separate
pathways of transcriptional repression in the Drosophila embryo. Proc Natl
Acad Sci U S A 96, 535-40.

Zhang, G., Nottke, A. and Goodman, R. H. (2005). Homeodomain-
interacting protein kinase-2 mediates CtBP phosphorylation and degradation
in UV-triggered apoptosis. Proc Natl Acad Sci U S A 102, 2802-7.

Zhang, Q., Piston, D. W. and Goodman, R. H. (2002). Regulation of
corepressor function by nuclear NADH. Science 295, 1895-7.

Zhao, L. J., Subramanian, T., Zhou, Y. and Chinnadurai, G. (2006).
Acetylation by p300 regulates nuclear localization and function of the
transcriptional corepressor CtBP2. J Biol Chem 281, 4183-9.

Zheng, L., Roeder, R. G. and Luo, Y. (2003). S phase activation of the
histone H28 promoter by OCA-S, a coactivator complex that contains GAPDH
as a key component. Cell 114, 255-66.

45

CHAPTER II

Developmental Expression and Phylogenetic Conservation of
Alternatively Spliced Forms of the
C-termlnal Binding Protein Corepressor

Mani-Telang, P. and Amosti D. N. (2006). Development, Genes and Evolution

In Press

46

ABSTRACT

The C-terrninal binding protein (CtBP) is an evolutionarily conserved
transcriptional corepressor found in multicellular eukaryotes. Multiple forms of
the protein are typically found in animal cells, produced from separate genes
and by alternative splicing. CtBP isoforms have also been implicated in
cytoplasmic functions, including Golgi fission and vesicular trafficking. All
forms of CtBP contain a conserved core domain that is homologous to a-
hydroxyacid dehydrogenases, and a subset of isoforms (CtBPL) contain
extensions at the C-terminus. Despite distinct developmental profiles and
knockout phenotypes in the mouse, the properties of different isoforms of the
protein are found to be similar in many transcriptional assays. We have
investigated the expression and conservation of distinct isoforms of the CtBP
protein in insects, and found that the expression of multiple, developmentally
regulated isoforms is widely conserved. In a variety of Drosophila species,
the relative abundance of CtBPL to CtBPs drops sharply after embryogenesis.
revealing a conserved developmental shift. Despite the overall lower levels of
this isoform, bioinforrnatic analysis reveals that exons encoding the C-terminal
extension in CtBPL are conserved from Diptera to Coleoptera, suggesting that

the CtBPL isoform contributes an important, evolutionarily conserved function.

47

Introduction

The C-terminal binding protein (CtBP) is an evolutionarily conserved
factor that has been implicated in a variety of cellular processes, including
transcriptional repression, Golgi function, and vertebrate retinal synapse
activity (Chinnadurai, 2003). Originally identified by its ability to interact with
the C-terminus of the adenovirus E1A protein, CtBP has been shown to
directly bind to a variety of transcription factors in vertebrate cells and in
Drosophila, and recruit chromatin-modifying factors including histone
deacetylases and histone demethylases (reviewed in Turner and Crossley,
2001; Chinnadurai, 2005). CtBP proteins share a high degree of similarity
within a central domain comprised of an NAD-binding domain and a
substrate-binding fold. The proteins form dimers, and demonstrate extensive
structural similarity to NAD-dependent dehydrogenases (Kumar et al., 2002;
Nardini et al., 2003). CtBP proteins also possess C-terminal sequences of
variable lengths that are likely to be unstructured (Nardini et al., 2006).

CtBP proteins exhibit a weak NAD-dependent catalytic activity in vitro,
however the physiological relevance of this activity is unknown (Kumar et al.,
2002; Balasubramanian et al., 2003; Barnes et al., 2003). NAD binding has
also been suggested to regulate CtBP allostery, permitting the interaction of
the protein with binding partners. In vitro, association of CtBP with E1A
proteins is stimulated by NAD and NADH, suggesting a possible molecular
switch that might regulate CtBP activity (Kumar et al., 2002; Barnes et al.,
2003). Differential affinity of the protein for NADH relative to NAD has been

suggested to endow CtBP with the potential to respond to differing

48

intracellular levels of these cofactors, possibly linking gene regulation to
cellular redox states (Zhang et al., 2002). A possible lysophosphatidic acid
acyl transferase activity relevant to membrane trafficking has also been
ascribed to one form of CtBP (CtBP3/CtBP1-S/BARS), however this result
has been disputed (Weigert et al., 1999; Gallop et al., 2005).

Distinct CtBP isoforms are expressed as a result of alternative splicing,
alternative promoter usage, and different genes. In vertebrates, the ctbp1 and
ctbp2 genes are expressed in overlapping patterns during development and
exhibit distinct functions. ctbp1 knockout mice are viable, but are smaller and
show increased postnatal mortality, while the ctbp2 mutation is embryonic
lethal (Hildebrand and Soriano, 2002). Similar to the expression pattern in
mice, avian orthologs of Ctbp1 and Ctbp2 are expressed in overlapping
regions and distinct domains suggesting that Ctbp1 and Ctbp2 might have
unique roles in certain tissues (Van Hateren et al., 2006). The rat CtBP1
isoform termed CtBP1-S/CtBP3/BARS lacks a short region at the N terminus;
this protein has been implicated in membrane fission events that are required
for Golgi trafficking and Golgi fragmentation during mitosis (reviewed in Corda
et al., 2006). In vertebrates, the RIBEYE splice form of CtBP2 produces a
protein containing CtBP residues fused to a unique N terminus; this protein is
localized to synaptic vesicles of the retina (Schmitz et al., 2000). , It is
unknown whether the distinct developmental phenotypes seen with vertebrate
CtBP mutants are largely a reflection of the genes' unique promoter activities,
or whether differences in the proteins themselves play a large role.

Posttranslational modifications and association with other binding

proteins have been shown to regulate the stability, activity and localization of

49

 

CtBP proteins in vertebrates. Some of these modifications target the central
conserved region of the protein; the Pak1 kinase phosphorylates CtBP1 at
Ser158, stimulating nuclear to cytoplasmic translocation and downregulating
CtBP repression activity (Barnes et al., 2003). Other modifications are
targeted to the C-terminal nonconserved portion of the protein;
phosphorylation of CtBP1 ser422 by the HIPK2 kinase promotes degradation
of the protein, whereas SUMOylation of the C-terminus is required for nuclear
localization of CtBP1 (Kagey et al., 2003; Lin et al., 2003; Zhang et al., 2005).
In addition to being covalently modified, the C-terminus can also serve as the
binding target for a PDZ-domain containing protein, neuronal nitric oxide
synthase, that drives cytoplasmic localization of the CtBP1 (Riefler and
Firestein, 2001).

Drosophila CtBP is vital for embryonic development and embryos
lacking dCtBP function exhibit grave defects in segmentation. In contrast to
vertebrates, distinct Drosophila CtBP proteins are produced from a single
gene. Two major isoforms, termed CtBPL and CtBPs, differ by the presence
or absence of a ~90 amino acid extension at the C-terminus, which, although
of similar size and amino acid composition, is not homologous to C-terminal
extensions found in vertebrate CtBP proteins (Poortinga et al., 1998; Nibu et
al., 1998a). In light of the fact that the unstructured C-terminus can play a
regulatory role in vertebrates, it seems possible that Drosophila CtBPL may be
subject to similar covalent modifications as those found in vertebrates, but
currently there is little understanding of the biological importance of the
different isoforms. In vitro, both CtBPL and CtBPs are able to bind to short-

range transcriptional repressors such as Knirps and Kriippel, and in

50

transcriptional assays, both isoforms exhibit similar activities (Sutrias-Grau
and Arnosti, 2004; Fang et al., 2006). Therefore, we have utilized
biochemical and phylogenetic analysis to study expression of the protein in
disparate orders to gain more insight into the significance of distinct isoforms
of this widely conserved protein. Biochemical and phylogenetic evidence
indicates that the alternatively spliced CtBPL isoform represents a conserved,
developmentally regulated form of the protein, suggesting a specific functional

role for this protein.

51

MATERIALS AND METHODS
Insect stocks and Iysate preparation

The fly stocks used in this study were: Drosophila melanogaster yW" (Amosti
lab), Drosophila sechellia, Drosophila moja vensis (Tucson Drosophila Stock
Center), Drosophila virilis (Dr. Scott Pitnick). Tn'bolium castaneum was a gift
from Dr. Susan Brown (Kansas State University), Anopheles gambiae from
Dr. Ned Walker (Michigan State University) and Apis mellifera from Dr.
Zachary Huang (Michigan State University). All files were maintained on
standard cornmeal/molasses food and embryos collected at 25°C on apple
juice/agar.

For developmental expression analysis, staged embryos were collected,
dechorionated and resuspended in lysis buffer (150mM NaCl, 50mM Hepes,
pH 7.9, 10% glycerol, 0.1mM EDTA with Complete mini-EDTA free protease
inhibitor cocktail tablet, Roche) and sonicated using a Branson-250 sonifier.
Larvae, pupae and adults were homogenized in lysis buffer with a steel pestle
and then sonicated under the same conditions. Lysates were cleared by
centrifugation and total protein concentration of the supernatant was
measured by a Bradford assay with bovine serum albumin as the standard.
The supernatant was mixed with Laemmli sample buffer for sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SOS-PAGE) analysis. For
identification of CtBP isoforms in bee, beetle, and mosquito and flies, whole
adult animals were homogenized in Laemmli sample buffer for SOS-PAGE

analysis.

52

Western Blot Analysis

Immunoblotting was performed using 10% SOS-PAGE gels in a tank transfer
system (Bio-Rad Mini Trans-Blot® Cell) and proteins were transferred to
lmmuno—BlotTM PVDF membrane (Bio-Rad). Antibody incubation was
performed in TBST (20mM Tris-HCI, pH 7.5, 120mM NaCl, 0.1% Tween-20)
supplemented with 5% non-fat dry milk as a blocking agent. Rabbit polyclonal
antibodies used to detect CtBP (1:10.000) and monoclonal mouse antibody
for tubulin (1:6000, Iowa Hybridoma Bank) were visualized using HRP-
conjugated secondary antibodies (Pierce) and SuperSignal® West Pico
chemiluminiscent substrate (Pierce). Western blots shown are representative

of at least three biological replicates for each experiment.

Antibodies and recombinant CtBP proteins

Polyclonal anti-CtBP antibodies were generated as described in Struffi et al.
(2005). For the production of recombinant proteins, the cDNAs for CtBPL and
CtBPs bearing two Flag epitope tags at the C-terminal end was cloned into
the pET15b expression vector and used to transform E.coli BL-21 cells.
Expression of bacterial proteins was induced by treating log-phase cultures
with 0.4mM IPTG. The cells were then sonicated in lysis buffer, centrifuged

and supernatant was dissolved in Laemmli sample buffer for western analysis.

53

RT-PCR measurements of splice form abundances In embryos and

adufls.

Total RNA of D. melanogaster embryos (stage 0-12) and adults was isolated
by tissue homogenization in Trizol reagent (Sigma) according to the
manufacturer's protocol. RNA was treated with Rnase-free DNase (RQ1,
Promega) to remove contaminating genomic DNA. RT-PCR was performed
using AccessQuickTM RT—PCR System from Promega. Transcripts for CtBPL
were amplified using primer pairs DA1 147 — 5'
CCCCACAGTACAACCAACCT 3' and DA1 148 - 5'
TCCGTTTTTATGCTCGATGA 3', CtBPs using primer pairs DA 1146 — 5'
CTCAACGAGCACAACCATCA'ITI’AATC 3' and DA 1150 — 5'
CTCTACTTTTCTTGATTTGATATCATTTGTAG 3' and total CtBP was
amplified using primer pairs DA 1 146 - 5'
CTCAACGAGCACAACCATCATTI’AATC 3' and DA 1151 -
5'GCACGTCTGGAATATTGCCGAC 3’. The transcript for tubulin was
amplified using primers BWH 763 - 5' CCGCCACCTTCATCGGCAAC 3' and
BWH 764 - 5' TTAGTTCTCGTCGACCACAG 3'. All primer pairs spanned an
intron such that amplification of contaminating genomic DNA could be
distinguished from the RT-PCR amplified products. The RT step. was
performed at 45°C for 45 minutes followed by 30 cycles of PCR in a 25 pl
reaction mix for 94°C for 1min, 56°C for 1 min and 72°C for 1 min for CtBP
transcripts and 25 cycles for the tubulin transcript. PCR products were
visualized by agarose gel electrophoresis and quantitated using BioRad

Quantity One software Version 4.4.1. . Three separate RNA preparations for

54

embryos and three for adults were quantitated and a scatter plot was
generated in Excel. Linear fits to the data were used to determine relative
amounts of CtBPL and CtBPs mRNA for each RNA preparation, and these
amounts were compared to levels of total CtBP and to tubulin. Relative levels
of CtBPs mRNA increase from embryo to adult by 3-8 fold, when normalized
to either total CtBP or tubulin controls, whereas normalized CtBPL mRNA
levels do not change significantly. The data shown in Figure “-10 is a
representative result of RT-PCR analysis of biological triplicates that were
each analyzed at least two times. In this particular dataset, CtBPs was
upregulated 6 fold in adults when normalized to both total CtBP and tubulin

controls.
Sequence alignments

To determine the conservation of CtBP exons in diverse insect genomes we
searched the Flybase database (Release 4.2) using FLYBASE BLAST for the
assembled genomes of Drosophila melanogaster, D. sechellia, D. persimilis,
D. mojavensis, D. virilis, D. grimshawi, Anopheles gambiae, Aedes aegypti,
Apis mellifera and Tribolium castaneum. Matches to conserved exons 1-4 of
CtBP were obtained for D. sechellia (AAKOO1000254.1), D. persimilis
(AAIZ01000471), D. mojavensis (contig_8705), D. virilis (contig_15233), D.
gn'mshawi (contig_21987), A.gambiae (AAABO1008805), Aedes aegypti
(supercontig_1.155), A.mellifera (AADGOSOO6060) and 7'. castaneum
(CM000284.1). These automated alignments generally did not identify exons
6 and 7 however, therefore sequences 3’ to the conserved exons were

searched in all three reading frames for conserved coding information and

55

aligned using Clustal W. Predicted gene sequences for A. mellifera
(XM_392682) and T. castaneum (XP_972241) were included in these
alignments. Relative similarities were calculated based on the pairwise
alignment with the D. melanogaster peptide sequence for exons 6+7 as a

reference.

56

RESULTS

Expression of CtBP Isoforrns In Drosophila

Four major CtBP transcripts are detected ubiquitously during
development and are predicted to produce proteins of 383, 386, 476 and 479
amino acids (Poortinga et al., 1998; Nibu et al., 1998b; Sutrias-Grau and
Arnosti, 2004). To analyze endogenous CtBP proteins in Drosophila, we
generated polyclonal rabbit antibodies against CtBPL protein (aa1-479)
expressed in E.coli. The antibodies are specific and detect proteins of the
expected sizes in embryonic extracts, approximately 42 (CtBPs) and 50 KDa
(CtBPL ) (Figure lI-1A). lmmunostaining revealed that CtBP proteins are
ubiquitously present in the nuclei of pre- and post-blastoderrn embryos and
imaginal discs from third Instar larvae (data not shown).

To analyze the developmental expression profile of CtBP isoforms, we
analyzed soluble extracts from different developmental stages of the fly
(Figure ll-1B). Both CtBPL and CtBPs isoforms are detected throughout the
first 15 hours of embryogenesis. with relatively higher levels of CtBPs than
CtBPL (antibody recognition of CtBPL is expected to be equal or better than
that of CtBPs because the two proteins are virtually identical in the central
domain, and the antibody was raised against CtBPL). The relative levels of
CtBPL to CtBPs drop further after embryogenesis. showing weak expression
of CtBPL in the larva, pupa, and adult (Figure II-1B). The lower abundance of
CtBPL in postembryonic stages is not simply due to sequestration of the
protein in an insoluble form, because similar low levels of CtBPL were

observed in whole animal extracts prepared in boiling SDS (discussed below).

57

Figure "-1: Developmental expression profile of CtBP isoforms In
Drosophila melanogaster.

A. Specificity of anti-CtBP antibody tested in Western blot with Drosophila
melanogaster embryonic extract (lanes 1,2) or bacterial extracts containing
recombinant CtBPL (lane 3) or CtBPs (lane 4). Preimmune serum did not
cross react with any proteins in embryo extract, while anti-CtBP recognized
two isoforms of approximately 42 and 50 kDa in embryonic extracts.
Recombinant proteins migrate slower than endogenous counterparts due the
presence of an N-terminal hexahistidine tag and a C-terminal Flag tag.
Markers (kDa) are indicated to the left.

B. Expression of CtBP isoforms in embryos, larvae, pupae, and adults. 50 pg
of total soluble protein was loaded on 10% SOS-PAGE and analyzed by
immunoblotting with anti-CtBP. Relative CtBPL and CtBPs levels were
unchanged during embryogenesis. A marked reduction in the relative level of
CtBPL was observed from the larval through adult stages. CtBPs levels
remained relatively unchanged throughout the developmental time course.
The bottom panel shows B-tubulin as a loading control.

C. Steady-state levels of CtBP mRNAs measured by RT-PCR analysis. Total
mRNA from embryos and adults was reverse transcribed and PCR amplified
using primers specific to CtBPL exons, CtBPs regions, or a region common to
both isoforms as indicated. Tubulin mRNA was measured as a control.
Reverse transcription reactions were primed with 60, 30, or 15 ng of total
RNA, as indicated by triangular symbol. The —RT control reactions were
primed with 60 ng of RNA. In this experiment, CtBPs levels increased from
embryo to adult six-fold relative to either total CtBP or tubulin, while
normalized CtBPL levels were unchanged.

58

 

 

 

 

 

 

A 8” 83V 83
\{9 KO k(.-
0 0° 90 ,bo
3“ e \o \o
6‘ 0° 6‘8 69
0‘ <9 <9
<26” <28“ 4“ Q9
5°“ "' . o
36 -
4 ’5 0 A
B.
<9 0 \09
(<3, «3’ 69 $89 989 e“ vs”
v“ >9 '0 s s 3*
6° 65" '9 V26 Q°Q v8 v8
<10th
. <10th
IB :anti-CtBP

 

 

 

 

 

55i - - - _ - - ‘ IB :anti-tubulin
1 2 3 4 5 6 7

 

 

l Embryos

 

Adults

 

 

59

We measured the relative levels of specific CtBP mRNA splice forms in
embryonic and adult stages to determine if this developmental switch reflects
a change in alternative mRNA isoform abundance. Primer pairs specific to
the CtBPs, CtBPL, and to a region of the gene common to both isoforms were
used in RT-PCR reactions. The absolute amounts of CtBPs and CtBPL RT-
PCR products are not directly comparable because different primer sets were
used, however the relative ratios in different stages of development are
informative. The levels of CtBPs transcripts relative to total CtBP or tubulin
mRNAs undergo a marked shift between these two stages, with CtBPs
increasing 3-8 fold in the adults, while little change is seen in CtBPL levels
(Figure ll-1C). This change suggests that the protein profile favoring CtBPs
later in development may be dictated by changes in the abundance of distinct
splice forms of the mRNA. Additional post-transcriptional effects may also

contribute to the decreased CtBPL protein levels observed.

Identification of conserved CtBPL-specific coding information

We examined genomic sequences of 10 different insects representing
>300 million years of evolutionary divergence - the fruit flies D. melanogaster,
D. sechellia, D. persimilis, D. moja vensis, D. grimshawi and D. virilis, the
mosquitoes Anopheles gambiae and Aedes aegypti (Diptera), the honey bee
Apis mellifera (Hymenoptera) and the red flour beetle Tribolium castaneum
(Coleoptera) - to determine if these organisms might also express diverse

isoforms of CtBP. Analysis of putative open reading frames 3' of core

60

conserved CtBP sequences identified regions homologous to D.
melanogaster exons 6 and 7, which encode the C-terminal extension of CtBPL
(Figure ll-2A). In Drosophila species, the sequences of exon 6 appear to be
separated from an upstream exon by a ~3 kbp intron, while the intron is of
smaller size in mosquito and beetle. In the honey bee, this intron appears to
have been entirely eliminated. The overall similarity among putative C-
terminal coding regions is clearly lower than that observed for the core CtBP
sequences (>92%), suggesting a lower level of evolutionary constraint (Figure
Il-2B). However, some motifs are conserved; the similarities include several
distinctive motifs involving less abundant amino acids, not simply tracts of
repeating residues that would show similarities by chance. Splice signals
following the terminal codons for exon 5 (YPEG), are conserved in all
Drosophila, as well as lower Diptera and T. castaneum, suggesting that the
downstream coding information is likely to be incorporated into mRNAs
(Figure lI-ZC). Splice acceptor sites are present immediately 5' of the
conserved LNGGYYT coding region of exon 6 in Drosophila species. A
conserved splice acceptor sequence is not found directly 5' of INNGGY
coding sequences present in T. castaneum and A. gambiae, raising the
possibility that acceptor sites in alternative locations may be used (Figure ll-
20). In the bee, the open reading frame for the C-terminal extension is fused

to the core sequences,

61

Figure "-2: Conservation of coding infonnatlon for CtBPL-speclflc C-
terminus.

A. Peptide coding information present in dipterans, bee, and beetle genomic
sequences homologous to alternatively spliced exon 6 and 7 in Drosophila
melanogaster encoding CtBP “tail” region. Conceptual translations of
genomic sequences are shown below sequence of CtBPL, in which YPEG
represents the end of the exon 5 coding sequence for the CtBPL isoform.
Predicted intron size in nucleotides is indicated between exons. lntrons in
Apis mellifera have apparently been eliminated. Dark gray shading indicates
widely conserved sequences; light gray shading partially conserved
sequences. Possible sumoylation sites (IN K X E) are indicated by gray bars
above exon 7 residues. An alternative splice acceptor site 5' of the junction
shown for exon 7 would produce an mRNA encoding an additional VSSQS
motif at the beginning of exon 7 (not shown); this sequence is not conserved
outside of Drosophila, unlike the residues 5' of exon 6 shown in ZC.

B. Phylogenetic tree of species analyzed, displaying relative similarities of the
C-terrninal ‘tail’ in different species compared to Drosophila melanogaster.
Pairwise comparisons were performed using D. melanogaster exons 6 and 7
(103 amino acids) as a reference. Percentage indicates identical amino acids
or a conserved substitution. Residues present in expansions in more
divergent species such as D. mojavensis and D. virilis were not scored. C.
The cDNA sequences reported for D. melanogaster CtBPL contain alternative
splice acceptor sites for the 5' end of exon 6; a sequence isolated from adult
head uses a downstream acceptor site (NP_001014617), while a different
sequence isolated from embryo uses a more upstream acceptor (Sutrias-Grau
and Arnosti, 2004) incorporating the residues LNGGYYT. This portion of the
protein is evolutionarily conserved and contains appropriate splice acceptor
sequences both 5' and 3' of this motif, thus alternative splicing may be a
conserved feature here as well.

62

 

 

 

 

 

 

         

 

 

 

 

      

 

 

 

mm B a $4 _ E32330 h
m m ”Ewan—mama --o ..... .. p-406Em---<m.E - 8-- 80:39: <
.....m, E m :2..:_ .mH:::-H ................................. 3884
mm ”1” w -537; a ..... SBEwa<
I I , I I .
mm m B --«3--- w.
I I I I
mu m H --«3---
I I I I
mm m H 137.

, l u .-
mm m B :3...
mm ... m meme... ......... 9%... ...... 3.22....
Pal m E :3. , . .......... m m. ...filmfifimg. ..... 8.mauoca.cE.o

n.2nsam c.4n3xm

 

 

]
J

 

 

 

. . . . . 55¢:ng H
. . 2229: .<
m a 3.68 .<
528% .o

maid
.... “2%.“—
sfifioad
..ofiouocSoE d

 

 

 

 

u. ............ I - - mznmmmo, .gmmo.

  

 

"III-II-II-Ii-ll-Il-Ii-ld .1.--
IIIIIIIIIIIIIIIII-IIII'IIIIIY

mnIm-n-n-IaI-mI-nl

 

.<

63

 

 

mum umm umm umdemoumuumuommmmmommmummmmuommooooumumm ........... an em mmm any on» Eamcmummo .._.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40mm Owwwwz>wdmm> 6mm?
moo muu mmm um_muoumouuwuucuommommummmumummuommfiuuommmmmmummo ..... um _m mom mo... om» 3Q>mwm .<
oawmeMMOUZHU‘NHAxKQ Ummw
umu muo mom om Tmaumumuumuommommummmuummme ....................... um F mom 000 any hmuwmmocflme .D
m a d U .H. w m. U U z A O m m M.
o zOXw m ZOxm .0
Ezccflmmog. .x. hm
93:38.1 fix. on
aq>mmm.< .X. ..v
mmEEmm< o\a :4
_>>mcmE_._m.o .x. no
m___.__> D o\o mm ,|.
m_wco>m_oE d in
m___E_9_8 o Slum
9.350090 3. v0 , .m

chmmmocsoE d

 

 

o\o oow

supporting the notion that these are indeed coding sequences. Similar to the
case with vertebrate CtBP proteins, the predicted C-terminal extensions of
these CtBP isoforms are probably unstructured in solution. The sequences
are rich in disorder promoting amino acids (ala, gly, pro, ser) and are
predicted to not assume a globular structure by the GlobPlot program (not

shown) (Linding et al., 2003).

Developmental expression of alternative Isoforrns in D. mojavensis and

D. virilis

The presence of the regions correlating to D. melanogaster exons 6
and 7 does not in itself reveal whether distinct CtBP isoforms are produced,
therefore we measured CtBP protein levels in embryos, larvae, pupae and
adults from D. mojavensis and D. virilis, which are estimated to have shared
the last common ancestor with D. melanogaster about 40-60 million years
ago. Western blot analysis revealed that two major bands of sizes similar to
CtBPs and CtBPL were present in these species (Figure "-3). The relative
abundance of the CtBPL isoform decreases in larval and pupal stages, staying
low in D. mojavensis in the adult, but increasing again in adult D. virilis. While
differing in details, these changes suggest that developmental changes in
relative abundances of CtBP isoforms are a conserved feature in

Drosophilids.

65

Figure "-3: Conserved developmental regulation of CtBP protein
expression in D. mojavensis and D. virilis.

Expression of CtBP isoforms in embryos, larvae, pupae, and adults of D.
mojavensis (A.) and D. virilis (B.). As in D. melanogaster, two predominant
species were observed in both species, but the CtBPL isoform has a lower
mobility (~60kDa vs. 50kDa in D. melanogaster). The relative levels of CtBPL
to CtBPs in the embryo was greater in these species than in D. melanogaster,
but just as in that species there is a pronounced decrease in relative levels of
CtBPL in the larva and pupa. Adult levels of CtBPL remain low in D.
mojavensis, but recover in D. virilis. 50 pg of total soluble protein was loaded
on 10% SOS-PAGE and analyzed by immunoblotting with anti-CtBP. The
bottom panels shows B-tubulin as a loading control.

66

 

 

50-
36-

- -- , ‘ <ICtBPL

W’ ' “CtBPs
- —- IB :anti-CtBP

 

55-

 

 

50-

36—

55-

W IB :anti-tubulin

 

 

Q1 0 e e «50%
(go <8“ 6" 8'8 .98 «9° e“
95°" 8 5° 64} Q8 0“ 0“
o’ ‘1’ Q" v8 ‘9

 

w3~ ---- «tapi-

W“ ~ a “ ” QCtBPS
. - IB :anti-CtBP

 

 

 

w IB :anti-tubulin

 

 

67

 

Expression of CtBP isoforms in diverse orders

To determine whether expression of CtBPs and CtBPL-like isoforms is
generally conserved in insects, we measured expression of CtBP proteins in
organisms whose sequenced genomes had been examined for CtBPL-specific
coding information (Figure ll-2A). Crude extracts from adults were analyzed
by Western blotting, including three Drosophila species of increasing
phylogenetic distance from D. melanogaster, Anopheles gambiae (lower
Dipteran), Apis mellifera (Hymenoptera), and Tn'bolium castaneum
(Coleoptera) (Figure "-4). Relative to D. melanogaster, the closely related D.
sechellia, (diverged ~3 Mya) expresses CtBPs and CtBPL isoforms of the
same size. The extracts from the more distantly related species D.
mojavensis and D. virilis (diverged 40-60 ~Mya) contained proteins of similar
size to CtBPs (~42 kDa) and an additional, lower mobility form (~60 kDa) that
migrated slower than D. melanogaster CtBPL. Two proteins were also evident
in the mosquito, both of somewhat faster mobility than the Drosophila
counterparts. Three cross-reacting species were found in the honey bee, all
of similar abundance, including one protein of ~25 kDa that migrates
considerably faster than CtBPs, similar to a minor species noted in D.
mojavensis extracts. Only one major isoform of ~50 kDa was detected in
extracts from T. castaneum, similar in mobility to D. melanogaster CtBPL,
although upon overexposure, weak bands of faster mobility could be seen. In
this figure, the relative levels of CtBPL and CtBPs in Drosophila appear to be
similar, but this is only because the gel was exposed for a long time to bring

out

68

 

 

 

 

50-

36- w ”fl”

 

 

 

Figure "-4: Adult expression of CtBP proteins in four Drosophila
species, Anopheles gambiae, Apis mellifera, and Tribolium castaneum.

Soluble extracts from adults were analyzed by Western blotting using anti-
CtBP. Cross-reacting species similar in size to CtBPs were noted in all
Dipterans. Slower mobility proteins consistent with CtBPL-like species were
present in all extracts; multiple bands were detected in extracts from all
species except T. castaneum. The relative abundance of CtBPL and CtBPs is
masked by the long exposure of the gel; lower panel shows a separate
Western blot (lanes 8-11) that was exposed for a shorter time to demonstrate
the lower abundance of CtBPL to CtBPs in D. melanogaster, D. sechellia, and
D. mojavensis adults.

69

the weaker A. mellifera bands. A Western blot of the Drosophila extracts in
which the exposure was shorter reveals that the ratio of CtBPL to CtBPs in
adults was low in all Drosophila species except D. virilis (Figure "-4, lanes 8-
11), which is consistent with the developmental profiles for D. virilis and D.

mojavensis shown in Figure "-3.

7O

DISCUSSION

In the mouse, the CtBP1 and CtBP2 genes have been found to provide
overlapping but functionally distinct activities in development (Hildebrand and
Soriano, 2002). These different activities might be transcriptionally based, a
situation in which homologous genes encode functionally interchangeable
products, but the distinct timing and levels of transcriptional activity of the
promoters are unique, as has been described for the Drosophila prd, gsb, and
gsbn genes (Li & Noll 1994). However, this model cannot be applied to cover
all vertebrate CtBP proteins, because the RIBEYE spliceform of CtBP2 and
CtBP1-S/BARS splice variant of CtBP1 encode distinct polypeptides, and
appear to have acquired unique roles in retinal function and membrane
trafficking, respectively (Corda et al., 2006). With respect to the
transcriptional regulatory forms of CtBP1 and 2, biochemical studies have
identified molecular modifications that may distinguish the two isoforms
functionally. CtBP1 is phosphorylated at serine158 by the Pak1 kinase, a
modification that induces nuclear to cytoplasmic translocation (Barnes et al.,
2003). CtBP2 has a unique N-terminus that is acetylated, which facilitates
nuclear retention of the protein (Zhao et al., 2006). Whether these differences
play a role directly in transcription is unclear; both proteins may function
similarly when recruited to promoters.

In Drosophila, less is known about distinctions among isoforms.
Previous work from our and other laboratories has indicated that multiple
CtBP isoforms are expressed in Drosophila, but no functional distinctions

have been drawn between CtBPs and CtBPL isoforms until now. Our study

71

 

provides evidence that the presence of these isoforms is not simply “noise”,
for example, aberrant splicing that is tolerated by the system. The
evolutionary conservation of multiple isoforms and developmental regulation
strongly points to functional differentiation between these proteins. It is
striking that all the organisms surveyed express proteins whose size
corresponds to the D. melanogaster CtBPL isoform. In addition, all contain
conserved coding sequences in their genomes for the unstructured C-terminal
extension of the protein, which in the case of mammals is the subject of
sumoylation, phosphorylation, and binding of regulatory proteins. Our
measurements of relative levels of CtBP mRNAs in embryos and adults
indicate that an increase in the amount of the CtBPs spliceform may account
for the shift toward this form of the protein later in the adult. However,
differences in mRNA translation efficiency or stability of the two forms of the
protein may also play a role. Phosphorylation of vertebrate CtBP by HIPK2
and c-Jun amino-terminal kinases at ser 422 in the unstructured C-terminal
portion of the protein triggers degradation of the protein via the proteasome
(Zhang et al., 2005; Wang et al., 2006). It is possible that a similar
modification may account for the drop in Drosophila CtBPL levels.

We note that putative sumoylation signals are conserved in Dipteran

sequences (Figure ll-2A), suggesting that insect CtBP proteins may similarly

be modified by SUMO. All vertebrate CtBP proteins possess some form of C-

terminal extension, however the presence of CtBPs isoforms in insects may

indicate that potential regulation by modification of the C-terminus may not be

required, at least in some stages or roles. Additional biochemical and genetic

72

studies will be required to identify possible functional distinctions between

these isoforms.

ACKNOWLEDGMENTS

We thank Scott Pitnick, Susan Brown, Zachary Huang, Ned Walker and
Tucson Drosophila Stock Center for files and insects used in this study and
Dr. Casey Bergman for advice on alignments, and two anonymous reviewers
for their helpful suggestions. We thank Paolo Struffi for generating the anti-
CtBP antibody, as described in (Struffi et al., 2005), and Montserrat Sutrias-
Grau for pointing out the potential SUMOylation sequences in CtBPL. This

project was supported by NIH GM56976 to DNA.

73

 

REFERENCES

Balasubramanian P, Zhao LJ, Chlnnadural G (2003) Nicotinamide adenine
dinucleotide stimulates oligomerization, interaction with adenovirus E1A and
an intrinsic dehydrogenase activity of CtBP. FEBS Lett 537: 157-160

Barnes CJ, Vadlamudi RK, Mishra SK, Jacobson RH, Li F, Kumar R
(2003) Functional inactivation of a transcriptional corepressor by a signaling
kinase. Nat Struct Biol 10: 622-628

Chlnnadural G (2003) CtBP family proteins: more than transcriptional
corepressors. BioEssays 1: 9-12

Chlnnadural G (2005) CtBP Family Proteins: Unique Transcriptional
Regulators in the Nucleus with Diverse Cytosolic Functions. In: G.
Chlnnadural (ed) CtBP Family Proteins. Eurekah.com, Landes Bioscience.
Georgetown, Texas, pp 1-17

Corda D, Colanzl A, Luini A (2006) The multiple activities of CtBP/BARS
proteins: the Golgi view. Trends Cell Biol 16: 167-173

Fang M, Li J, Blauwkamp T, Bhambhani C, Campbell N, Cadlgan KM
(2006) C-terminaI-binding protein directly activates and represses Wnt
transcriptional targets in Drosophila. EMBO J 25: 2735-2745

Gallop JL, Butler PJ, McMahon HT (2005) Endophilin and CtBP/BARS are
not acyl transferases in endocytosis or Golgi fission. Nature 438: 675-678

Hildebrand JD, Soriano P (2002) Overlapping and unique roles for C-
tenninal binding protein 1 (CtBP1) and CtBP2 during mouse development.
Mol Cell Biol 15: 5296-5307

Kagey MH, Melhuish TA Wotton D (2003) The polycomb protein P02 is a
SUMO E3. Cell 1: 127-137

Kumar V, Carlson JE, Ohgl KA, Edwards TA, Rose DW, Escalante CR,
Rosenfeld MG Aggamal AK (2002) Transcription corepressor CtBP. is an
NAD(+)-regulated dehydrogenase. Mol Cell 4: 857-869

LI X, Noll M. (1994) Compatibility between enhancers and promoters
determines the transcriptional specificity of gooseberry and gooseberry neuro
in the Drosophila embryo. EMBO J 13: 400-406

Lin X, Sun B, Llang M, Llang YY, Gast A, Hildebrand J, Brunlcardl FC,
Melchior F Feng XH (2003) Opposed regulation of corepressor CtBP by
SUMOylation and PDZ binding. Mol Cell 5: 1389-1396

Llnding R, Russell RB, Neduva V, Gibson TJ (2003) GlobPlot: Exploring

74

protein sequences for globularity and disorder. Nucleic Acids Res 31: 3701-
3708

Nardinl M, Spano S, Cericola C, Pesce A, Massaro A, Mlllo E, Luini A,
Corda D, Bolognesl M (2003) CtBP/BARS: a dual-function protein involved

in transcription co-repression and Golgi membrane fission. EMBO J 22: 3122
— 31 3O

Nardini M, Svergun D, Konarev PV, Spano S, Fasano M, Bracco C, Pesce A,
Donadini A, Cericola C, Secundo F, Luini A, Corda D Bolognesl M (2006)

The C-terrninal domain of the transcriptional corepressor CtBP is intrinsically
unstructured. Protein Sci 15: 1042-1050

Nibu Y, Zhang H, Bajor E, Barolo S, Small S, Levine M (1998a) dCtBP

mediates transcriptional repression by Knirps, Kruppel and Snail in the
Drosophila embryo. EMBO J 23: 7009-7020

Nibu Y, Zhang H, Levine M (1998b) Interaction of short-range repressors
with Drosophila CtBP in the embryo. Science 5360: 101-104

Poortinga G, Watanabe M, Parkhurst SM (1998) Drosophila CtBP: a Hairy-

interacting protein required for embryonic segmentation and hairy-mediated
transcriptional repression. EMBO J 7: 2067-2078

Riefler GM, Firestein BL (2001) Binding of neuronal nitric-oxide synthase
(nNOS) to carboxyl-terminal- binding protein (CtBP) changes the localization

of CtBP from the nucleus to the cytosol: a novel function for targeting by the
PDZ domain of nNOS. J Biol Chem 51: 48262-48268

Schmitz F, Konlgstorfer A, Sudhof TC (2000) RIBEYE, a component of

synaptic ribbons: a protein's journey through evolution provides insight into
synaptic ribbon function. Neuron 28: 857-872

Struffl P, Arnosti DN (2005) Functional interaction between the Drosophila

knirps short range transcriptional repressor and RPD3 histone deacetylase. J
Biol Chem 280: 40757-40765

Sutrias-Grau M, Arnostl DN (2004) CtBP contributes quantitatively to Knirps

repression activity in an NAD binding-dependent manner. Mol Cell Biol 24:
5953-5966

Turner J, Crossley M (2001) The CtBP family: enigmatic and enzymatic
transcriptional co-repressors. BioEssays 8: 683-690

Van Hateren N, Shenton T, Borycki AG (2006) Expression of avian C-

lerminal binding proteins (Ctbp1 and Ctbp2) during embryonic development.
Dev Dyn 235: 490-495

Wang SY, lordanov M, Zhang Q (2006) c-Jun-NH2-terminal kinase promotes
apoptosis by down-regulating the transcriptional corepressor CtBP. J Biol

75

 

Chem [in press, Epub ahead of print]

Weigert R, Silletta MG, Spano S, Turacchlo G, Cericola C, Colanzl A,
Senators S, Mancini R, Polishchuk EV, Salmona M, Facchlano F, Burger

KN, Mironov A, Luini A, Corda D (1999) CtBP/BARS induces fission of Golgi
membranes by acylating lysophosphatidic acid. Nature 6760: 429-433

Zhang Q, Nottke A, Goodman RH (2005) Homeodomain-interacting protein

kinase-2 mediates CtBP phosphorylation and degradation in UV-triggered
apoptosis. Proc Natl Acad Sci U S A 102: 2802-2807

Zhang Q, Piston DW, Goodman RH (2002) Regulation of corepressor
function by nuclear NADH. Science 5561: 1895-1897

Zhao LJ, Subramanlan T, Zhou Y, Chlnnadural G (2006) Acetylation by

p300 regulates nuclear localization and function of the transcriptional
corepressor CtBP2. J Biol Chem 281: 4183-4189

76

 

 

CHAPTER III

FUNCTIONAL ANALYSIS OF CtBP DEHYDROGENASE-LIKE DOMAIN

ABSTRACT

The transcriptional corepressor CtBP lies at the center of several
protein networks involved in gene expression, development and oncogenesis.
In this study, we have sought to uncover the functional relevance of the
dehydrogenase-like domain for CtBP biological function in a whole organism
context. Our observations show that Drosophila CtBP requires NAD binding
for biological activity in vivo. NAD binding by CtBP has been postulated to
allow formation of higher order complexes by CtBP. Our results show that
dimerization of CtBP proteins is not dependent upon NAD. However, the
binding of NAD might be critical for inducing allosteric changes in CtBP that
perhaps allow for subsequent recruitment of cofactors needed for

transcriptional repression.

77

Introduction

A hallmark of Drosophila development is the regulation of gene
expression by transcriptional repression (Gray and Levine, 1996). Multiple
mechanisms effecting repression have been identified that have increased our
understanding of gene regulation, but the characterization of how and when
these mechanisms are utilized is still in its rudimentary stages. A general
functional distinction has been made categorizing the range of repressor
action in Drosophila (Gray and Levine, 1996). Short-range repressors are
locally-acting proteins that are fundamental regulators of gene expression and
segmentation in Drosophila. Repression by these proteins is possible when
they are located within a short distance (~100-150 bp) of either activators or
the transcription initiation site, and their short range of action means that
individual enhancers can be repressed without interfering with the functioning
of other nearby enhancers.

All short-range repressors studied in Drosophila employ the C-terminal
Binding Protein (CtBP) as an effector to mediate repression. CtBP was first
identified in human cells through its interaction with the adenoviral E1A
oncoprotein where it was found to repress E1A oncogenicity (Boyd et al.,
1993). Since then, numerous transcription factors have been found to employ
this conserved corepressor (Chinnadurai, 2002). Several vertebrate and
invertebrate DNA-binding proteins are known to employ CtBP for their
biological functions (Criqui-Filipe et al., 1999; Furusawa et al., 1999; Kegel et
al., 2002; Koipally and Georgopoulos, 2000). Mice mutant for CtBP1 and

CtBP2 genes display grave developmental defects that may be ascribed to

78

the compromised abilities of numerous regulators of gene expression,
emphasizing the importance of CtBP in development (Hildebrand and
Soriano, 2002). The Drosophila homolog of CtBP contributes quantitatively to
total repressor activity to generate appropriate levels of repression by short-
range repressors like Knirps and Krilppel (Keller et al., 2000; Struffi et al.,
2004; Sutrias-Grau and Arnosti, 2004). CtBP also interacts with the long-
range repressor Hairy; in this case, however, CtBP appears to antagonize
rather than facilitate repression by Hairy’s primary cofactor Groucho, through
a competitive binding mechanism (Zhang and Levine, 1999).

Multiple mechanisms may be used by transcriptional repressors to
enhance the overall repression output or to provide unique activities that
regulate transcription negatively at a given promoter (Arnosti, 2004). Recent
work has established that the repression activity of CtBP proteins relies on the
formation of a multiprotein assembly containing an assortment of enzymes.
This might represent one mechanism by which CtBP acts to effect
transcriptional repression, namely as an adaptor that links chromatin-
modifying activities to target promoters. The CtBP1 corepressor complex
purified from human cells includes chromatin remodeling enzymes that
mediate histone modifications by deacetylation and methylation of H3-K9, and
the demethylation of H3-K4 (Shi et al., 2003). An alternate mechanism posits
that CtBP directly binds bromodomains of histone acetyltransferases such as
p300 and CBP, and blocks them from contacting and modifying histone tails
(Kim et al., 2005; Senyuk et al., 2005).

Growing evidence indicates that CtBP proteins are multifunctional

performing roles in addition to controlling gene expression. Multiple CtBP

79

isoforms have been implicated in cytosolic functions such as Golgi
maintenance and synaptic ribbon formation in retinal cells (Chinnadurai,
2006b; Corda et al., 2006; Schmitz et al., 2000).

Despite the multiplicity of CtBP isoforms, all CtBP proteins exhibit a
highly conserved signature sequence and structural similarity to alpha
hydroxy acid dehydrogenases (Chinnadurai, 2002; Kumar et al., 2002). This
similarity extends over a centrally located NAD+/NADH dinucleotide-binding
domain and a triad of residues known to be important for mediating catalysis
in related dehydrogenase enzymes. The discovery of an NAD-binding fold
within CtBP has provoked several studies that attempt to put CtBP at the
focus of redox-regulated pathways. Regulation of gene expression by redox
regulators represents an emerging concept in transcription (Imai et al., 2000;
Rutter et al., 2001).

In addition to possibly serving in a catalytic mode, NAD is also
suggested to play a role in modulation of CtBP structure. NAD binding to
CtBP has been shown to promote oligomerization of the protein in vitro
(Balasubramanian et al., 2003). Possibly as a consequence of this, mutations
that disrupt NAD binding affect CtBP activity. Mutations that impede NAD
binding (G181A, G183A and 0204A) reduce CtBP1 binding to an E1A-CST
fusion protein and prevent CtBP mediated repression of a Gal4-E1A activated
reporter in cell culture (Kumar et al., 2002). An NAD binding mutant of CtBP2
(G189A) de-represses a heterologous SV40 enhancer-activated reporter in
C33 cells (Thio et al., 2004). In one case, a CtBP protein mutated in the NAD
binding cleft was found to retain activity (Grooteclaes et al., 2003). One

problem with these assays is that the transiently transfected reporters do not

80

represent a physiological chromatin environment. Thus, it is not certain
whether relevant CtBP chromatin modifying activities are being assayed. In
contrast, whole animal assays using integrated reporter genes test gene
activity in a native chromatin context. We have assayed CtBP proteins in such
a setting and found that mutations in key glycines (G181A, G183A) implicated
in NAD binding abolish repression by GaI4-CtBP fusion proteins (Sutrias-Grau
and Arnosti, 2004). This result suggests that NAD binding is crucial in this
context.

o-hydroxy acid dehydrogenases contain a well-conserved triad of
active site residues. This catalytic center of related dehydrogenases is well
conserved in vertebrate and invertebrate CtBP family members. Triple
mutations that disrupt the putative catalytic ability (H3150, 6295, R266) were
seen to diminish CtBP1 interaction with an E1A-GST chimera, and reduced
repression of a GaI4-E1A activated reporter in cell culture (Kumar et al.,
2002). These mutations were also seen to abrogate a weak dehydrogenase
activity seen with the wild type vertebrate CtBP1 protein. In contrast, a single
mutation disrupting the putative active site (H3150) did not obstruct
Interaction with E1 A and this mutant was found to retain repressive potential
comparable to the wild type CtBP protein (Grooteclaes et al., 2003). In whole
animal assays using integrated reporters, mutations in the putative active site
of CtBP abolished repression (Sutrias-Grau and Arnosti, 2004). This result
suggests that the presumptive catalytic function is not required for CtBP

mediated transcriptional repression.

81

MATERIALS AND METHODS

Construction of P-element vectors

cDNA fragments for CtBP isoforms (Accession numbers: AY060646 for
dCtBPL and ABO11840 for dCtBPs) carrying a C-terrninal double FLAG tag
(FF) were introduced downstream of 5X UAS sites in the pUAST vector using
unique Kpnl-Xbal sites in the polylinker (Brand and Perrimon, 1993). An.
optimal Kozak sequence and start codon was provided by the insertion of an
adaptor upstream of the cDNA sequence using Bglll and Kpnl restrictions
sites (DA 683 5’-GATCACCCGGGACCAAAATGGGTAC-3’ and DA 684 5’-
CCAT'ITI'GGTCCCGGGT-B’). dCtBPL NAD binding and catalytic mutants
have been described elsewhere and contain mutations in G181A, G183A
disrupting NAD binding (NAD) and H3150 abolishing the putative active site
(CAT) (Sutrias-Grau and Arnosti, 2004). For the dimerization mutant, point
mutations were introduced in dCtBPL-FF at four key arginines (R141A,
R142A, R163A and R171A). All mutants described were made in the context
of the full length CtBPL. The constructs expressing CtBP ubiquitously were
generated in the pUAST vector containing CtBP wild type and mutant cDNA
fragements. The vector has a basal hsp70 promoter and 5X UAS sites that
were removed by restriction digestion with Sphl and Kpnl and replaced with
an adaptor that containing Sphl, a new Ascl and Kpnl restriction sites and an
optimal Kozak sequence (DA 885 5’-
CACCGGCGCGCCACCAAAATGGGTAC-3’; DA 886 5’-
CCAT‘ITTGGTGGCGCGCCGGTGCATG-3’). The promoter region of CtBP

82

(~8054 bp) was amplified from embryo genomic DNA in two pieces: the first
4000 bp fragment containing one of the predicted start sites was amplified as
an Sphl-Ascl fragment (068841302-8845092) with DA 926 (5’-
GTGCATGCGAAATGG‘I'I’AGCCAGCGTGGTG-3’) and DA 927 (5’-
CGGGCGCGCCTTGAAATCGAGAATCCTGCAATGG-3’) and inserted
immediately upstream of the cDNA in frame. The second fragment (~4054 bp)
with the other putative transcriptional start site was amplified as an Sphl-Sphl
fragment (CG8837323-8841301) with primer sets DA 924 (5’-
CTGCATGCATACCATAATTCTTGCAGT‘I'I’GCC-3’) and DA 925 (5'-
CGGCATGCAGCTTI'CTGT'I'I'CATGCATATGCAC -3’) and introduced into

the Sphl site, upstream of the first fragment (See Appendix A, Figure A-1).

lmmunoprecipitations and Western Blot Analysis

0-12 hour embryos were collected from transgenic lines expressing FLAG-
tagged CtBP proteins, resuspended in lysis buffer (150mM NaCI, 50 mM
HEPES, pH 7.9, 10% glycerol, 0.1mM EDTA with Complete mini-EDTA free
protease inhibitor cocktail, Roche) and sonicated using a Branson-250
sonifier. Soluble lysates (~20 mgs total protein) were immunoprecipitated
overnight at 4°C with 25 pl of anti-M2 (Sigma) crosslinked to protein G beads,
by incubation with the chemical crosslinker dimethyl pimelimldate
dihydrochloride to a final concentration of 20mM. lmmunoprecipitation with
mouse monoclonal lgG and anti-M2 with non-transgenic embryo extracts was
performed in parallel as negative controls. Beads were collected and washed

thrice in lysis buffer and eluted in 40 pl of Laemmli sample buffer.

83

lmmunoprecipitates were separated on SOS-PAGE gels and analyzed by
western blotting as described below using rabbit polyclonal anti-CtBP
(1:10.000) (Struffi and Arnosti, 2005).

Total extracts from adult flies were immunoblotted using 10% SOS-PAGE gels
in a tank transfer system (Bio-Rad Mini Trans-Blot® Cell) and proteins were
transferred to lmmuno-BlotTM PVDF membrane (Bio-Rad). Antibody
incubation was performed in TBST (20mM Tris-HCI, pH 7.5, 120mM NaCl,
0.1% Tween-20) supplemented with 5% non-fat dry milk as a blocking agent.
Mouse monoclonal antibodies include anti-M2 used to detect FLAG epitope-
tagged CtBP (1:10.000) and tubulin (1:6000, Iowa Hybridoma Bank). These
were visualized using HRP-conjugated secondary antibodies (Pierce) and

SuperSignal® West Pico chemiluminiscent substrate (Pierce).

P-element transformation and antibody staining of Imaginal discs

All the P-element transformation vectors described were introduced into
Drosophila germ line by injection of yw67 embryos as described previously.

Imaginal discs dissected from third instar larvae were fixed in antibody
staining buffer (10 mM potassium phosphate, pH 6.8, 15 mM NaCl, 45 mM
KCI and 2 mM MgCl2) and 4% formaldehyde on ice for 45 minutes. For
detection of FLAG-tagged proteins, anti-M2 was used at a concentration of
1:1500. This antibody was detected and developed using a universal
secondary antibody conjugated to HRP (1:250, Vectastain ABC Kit). Discs
were mounted in 80% glycerol and the expression pattern was visualized

using an Elite PK-62000 Universal Vectastain ABC kit (Vector Laboratories,

84

lnc., Manufacturer’s protocol).

Recombinant CtBP proteins and Gel filtration chromatography

The cDNAs for wild type CtBP (both CtBPL and CtBPs) and the three mutants
(NAD binding mutant, catalytic mutant and dimerization defective mutant of
CtBP) were cloned into a modified pET-15b expression vector using Kpn l and
Not I restriction sites (created by D.N. Amosti). This vector provides an N-
terminally fused hexahistidine tag. Except for the catalytic mutant cDNA that
contains 2 FLAG-tags at the C-terminus, all other cDNAs were designed to
carry a single tag. These vectors were then used to transform heat shock
competent E.coli strain BL21 (DE3). 25 ml of log phase bacterial cultures
were induced with 0.8mM isopropyl thiogalactose (IPTG) overnight at 16°C.
The cells were collected, resuspended in buffer containing 50mM HEPES pH
7.9, 150mM NaCI, 10% glycerol, 10 mM imidazole with a Roche cocktail
protease inhibitor tablet and 10 mM B-mercaptoethanol, sonicated and
centrifuged to separate the soluble supernatant fraction. 0.1% Triton X was
added to supernatant before incubation with Ni2°-NTA beads (Sigma) at 4°C
for 3 hours. The beads were washed thrice in the same buffer with 20 mM
imidazole and eluted with 250 mM imidazole.

~10 ug of purified protein was fractionated on a Superdex 200 10/30 size
exclusion column (Amersham) equilibrated with buffer containing 150 mM
NaCI, 50mM HEPES pH 7.9, 0.1 mM EDTA and 10% glycerol. Following
sample injection, 0.5 ml fractions were collected at a flow rate of 0.5 ml/min.

Pools of three fractions (4 ul each fraction) were resolved on SDS-PAGE gels

85

and analyzed by western blotting using anti-M2 to detect FLAG-tagged
recombinant proteins as described before (1:10.000). For the CtBPL-NAD
mutant a pool of three fractions (10 ul each fraction) was analyzed by western
blots. Size markers (MW-GF-1000, Sigma) were separated under similar

conditions.

86

RESULTS

Assaylng the biological activity of wild type and mutant CtBP proteins in

vivo

We have previously observed that an NAD binding mutant of CtBP
(G181A, G183A) is impaired for repression. whereas a putative catalytic site
mutant (H3150) retains its repression activity when tested as Gal4 tethered
proteins in embryo repression assays, suggesting that NAD binding is
required for CtBP repression activity (Sutrias-Grau and Arnosti, 2004).
However, these assays are limited in that they only test the protein’s activity in
a repression context, and as a DNA binding fusion protein. To investigate the
functional relevance of residues within the dehydrogenase-like domain of
CtBP in a more general context, we generated Drosophila transgenic lines
expressing proteins that carry mutations in conserved residues. The changes
introduced are predicted to disrupt NAD binding (NAD), disrupt the putative
catalytic site (CAT) or abrogate dimerization (DIM) based on the effects of
these mutations in the highly similar vertebrate CtBP1 (Balasubramanian et
al., 2003; Kumar et al., 2002; Nardini et al., 2003). NAD binding by CtBP has
been suggested to promote oligomerization that might lead to recruitment of
cofactors effecting repression on a CtBP-dependent target (Balasubramanian
etaL.2003)

We overexpressed the wild type CtBPL and the three mutant proteins
using the UAS-Gal4 system (For UAS-Gal4 assay see Appendix B - Figure B-

1). Ectopic expression of the wild type protein and a putative catalytic mutant

87

 

wing blister

Figure Ill-1: Phenotypes induced by overexpression of wild type and
mutant CtBP proteins

Misexpression of CtBP proteins in eye (A) induces rough eye phenotype. and
in wing (C) induces vein abnormalities and blistering. Normal eye and wing
appearance shown in B and D. Phenotypes were observed with
misexpressed wild type CtBPL and CtBPL (CAT). Misexpression of CtBPL
(NAD) and CtBPL (DIM) mutants and the short splice variant of CtBP did not
yield any abnormal phenotype. Expression was achieved with eye specific
GMR-Gal4 driver or the wing-specific en-Gal4 driver.

For a schematic of UAS-Gal4 assay. see Appendix B-Figure B-1.

A minimum of three independent lines per transgene were examined for
phenotypic outcomes. SEM images are courtesy of Geoffrey Williams, Brown
University.

88

Table Ill-1: Quantitation of CtBP misexpression phenotypes

Flies with phenotypes generated by wild-type CtBP isoforms and mutant
forms that affect the CtBPL (CAT), CtBPL (NAD), or CtBPL (DIM) residues were
scored. CtBPL and CtBPL (CAT) mutants showed similar effects; no wing or
eye phenotypes were observed with the CtBPL (NAD and CtBPL (DIM)
mutants or the CtBPs isoform. A minimum of three independent lines per
transgene were counted for phenotypes.

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ems - o SN - o mco:
Eb m o NR m o is $90
one v o mg m o -25 .590
aux. m O can m o .-o<z 590
8m 3. $8 E m 5.3 -Eoimzo
82. e 8.» m2 4 $2 is 590
n: 87.3 8:: u Anxcvmsooemc as; u: 86$ 8:: u §vwe>o cmaom mcmmmcmc.
mmctozmzm wwmtbzmzc
02:5 m>m

 

 

 

90

of CtBP (H3150 mutation) induced wing blisters and rough eyes when
misexpressed in the wing using an en-Gal4 driver and in the eye using the
GMR-Gal4 driver. In contrast, misexpression of the NAD binding mutant,
dimerization mutant, or the short isoform of CtBP induced no obvious
phenotype in this assay (Figure Ill-1, Table Ill-1).

To test if the differences in induced phenotypes reflected intrinsic
differences in protein activity or rather in levels of protein expression, adult
flies were assayed for CtBP protein levels by immunoblotting and larvae were
assayed by immunostaining of eye imaginal discs. The CtBP proteins carry a
double FLAG-tag at the C-terminus that permits detection of the recombinant
protein by an anti-M2 antibody. As shown in Figure III-2, the steady state
protein levels CtBPL (NAD), CtBPL (DIM) and CtBPs were significantly lower
than the wild type CtBPL or CtBPL (CAT). The low expression of CtBPL (DIM)
suggests that lack of dimerization may destabilize the protein, which raises
the possibility that the CtBPL (NAD) mutant protein may also be poorly
expressed due to lack of dimerization. The expression of CtBPL (CAT) protein
was consistently observed to be comparable to the wild type protein for

several independent lines tested.

91

Figure III-2: Relative steady state levels of recombinant CtBP proteins:
CtBPL (NAD) and CtBPL (DIM) mutants and CtBPs isoform are expressed
at low levels.

A. Single adult flies misexpressing the transgene were solubilized in Laemmli
sample buffer and extracts were probed with anti-M2 antibody to detect the
recombinant FLAG-tagged proteins. The CtBPL (NAD) (lanes 5.6), or CtBPL
(DIM) (lanes 7.8) were detected at low levels compared to CtBPL (lanes 1,2)
and CtBPL (CAT) (lanes 9,10). CtBPs isoform was practically undetectable,
although upon darker exposure a faint band could be observed (lanes 3. 4;
data not shown). The lower panel shows the same blot stripped and reprobed
with B-tubulin antibody as a loading control. B-tubulin was reduced in lanes 1-
3, indicating these lanes may be underloaded. Two independent lines per
transgene are shown.

B. Third instar wing imaginal discs were dissected from third instar larval
progeny misexpressing the transgenes described above. Discs were
immunostained with anti-M2 (121500). The eye driver GMR-Gal4 expresses
CtBP in the posterior half of the imaginal disc. CtBPL and CtBPL (CAT)
showed comparable expression (Panels 1, 2) while the CtBPL (NAD) (Panel
3). CtBPL (DIM) (Panel 4) and CtBPs (Panel 5) were expressed at low levels.
Discs from GMR-Gal4 crossed to non-transgenic flies served as negative
control (Panel 6).

92

 

GMR (eye driver) crosses

 

 

 

 

 

 

 

 

 

 

IB :anti-M2
v-r *
50- g. lB:anti-tubulin
1 2 3 4 5 6 7 8 9 10 11
B.
CtBPL CtBPL-CAT CtBPL-NAD CtBPL-DIM CtBPs yw

    

‘1

      

 

 

 

 

93

 

DroSOphiIa CtBPL presumptive catalytic mutant is competent for

dimerization

Consistent with previous cell-based and embryo repression assays,
mutation of the well-conserved histidine residue does not measurably affect
CtBP biological activity. The mutated histidine is important for catalysis in
dehydrogenase enzymes and reported to be critical for the weak in vitro
dehydrogenase activity of CtBPL (Balasubramanian et al.. 2003; Kumar et al.,
2002). It is possible that this presumptive catalytic residue in CtBP might
function in a different biological context not tested in our assay. In addition,
the mutant protein may form heterodimers, recruiting one wild-type CtBP that
may provide enough function to mask the effect of the mutation (Figure Ill-3).

To test this possibility, co-immunoprecipitations were performed on
embryo extracts expressing recombinant, epitope-tagged mutants
ubiquitously. lmmunoprecipitation using the anti-M2 FLAG antibody recovered
the epitope tagged CtBPL (CAT) protein. as well as endogenous CtBPL and
CtBPs proteins (Figure Ill-3). The association of CtBPL (CAT) with
endogenous CtBP proteins sheds new light on our previous embryo
repression assays in which Gal4-CtBPL (CAT) was assayed (Sutrias-Grau and
Arnosti, 2004). This result suggests that the repression activity of the catalytic
mutant might be attributable to its dimerization with wild-type endogenous
proteins. To better characterize the impact of this specific mutation. a CtBP
null background is desired where possible heterodimerization can be ruled out

(See Appendix A).

94

Input cg as , Q/
Q
o8 e“ QK {‘3

 

*
CtBPL-FLAG
- - fiCtBPL
"' .- <ICtBPs
IB :anti-CtBP

 

 

 

Figure III-3: CtBPL (CAT) protein is capable of dimerization

Embryo extracts prepared from transgenic lines that express recombinant
CtBPL (CAT) (FLAG-tagged) from CtBP promoter. were immunoprecipitated
with anti-M2 (lane 5). Endogenous CtBP isoforms specifically co-precipitated
with the recombinant protein. and no non-specific associations were observed
with either beads alone (lane 6) or anti-M2 IP of non transgenic embryo
extracts (lane 4). A sample of non-transgenic embryo extract (yw EE) was
loaded as a size marker to judge migration of CtBP proteins (lane 7). A non-
specific band observed is marked with an asterisk.

95

Drosophila CtBPL and CtBPs can homo- and heterodimerlze

The ability of CtBP to dimerize is predicted to be central to its role as a
bridging factor, allowing one monomeric unit to contact a DNA bound factor
and the other to interact with chromatin modifying factors. Vertebrate CtBP
proteins are known to dimerize, and dimerization may allow the formation of a
CtBP repressor assembly. All CtBP isoforms identified in vertebrates possess
unstructured C-tenninal extensions (‘tails'), in contrast to Drosophila that
expresses a tailed (CtBPL) in addition to a unique untailed isoform (CtBPs).
The association of CtBPL (NAD) with endogenous CtBPL and CtBPs suggests
that hetero and homodimers between tailed and untailed proteins are able to
form. To test if wild type CtBPL and CtBPs proteins can dimerize, we
performed co-immunoprecipitations to test the ability of recombinant, flag-
tagged CtBPL expressed in a ubiquitous fashion, to dimerize with endogenous
CtBPL and CtBPs. The recombinant wild type CtBPL effectively and
specifically immunoprecipitated endogenous long and short proteins,
indicating that the CtBPL can homo- and hetero-dimerize (Figure Ill-4A).
Coimmunoprecipitations with recombinant CtBPs also co-precipitated both
endogenous isoforms comparable to recombinant CtBPL, indicating that both
tailed and untailed forms of CtBP are capable of forming dimers. indicating
that dimer formation is characteristic of all CtBP proteins (Figure Ill-4B). Thus,
the low expression levels of CtBP proteins do not appear to be reflective of a

lack of dimerization potential.

96

Figure Ill-4: lmmunoprecipitation of endogenous CtBPL and CtBPs with
FLAG-tagged CtBP proteins Indicate CtBP homo- and
heterodimerization

A. Embryo extracts prepared from transgenic lines that express recombinant
CtBPL (FLAG-tagged) from the CtBP native promoter were
immunoprecipitated with anti-M2 (lane 5) and anti-IgG as a negative control
(lane 7). Endogenous CtBP isoforms specifically co-precipitated with the
recombinant protein, and no non-specific associations were observed with
either beads alone (lane 6) or lgG (lane 7). A sample of non-transgenic
embryo extract (yw EE) was loaded as a size marker to judge migration of
CtBP proteins (lane 4).

B. Embryo extracts prepared from transgenic lines that express recombinant
CtBPs (FLAG-tagged) from CtBP native promoter were immunoprecipitated
with anti-M2 (lane 5). Endogenous CtBP isoforms specifically co-precipitated
with the recombinant protein, and no non-specific associations were observed
with either beads alone (lane 6) or anti-M2 IP of a non-transgenic embryo
extract (lane 4). A sample of non-transgenic embryo extract (yw EE) was
loaded as a size marker to judge migration of CtBP proteins (lane 7).

97

 

<I CtBPL-FLAG
<1 CtBPL

<1 CtBPs
IB :anti-CtBP

 

 

 

 

 

<1 CtBPL
<1 CtBPs-F LAG
4 CtBPs

 

 

IB :anti-CtBP

 

 

98

 

€59 09’ 0““,
QV Q" Q" Q" 9‘;
v8? 8? 8’ 8? *8?
0’0 0’0 0’0 0’0 0'0
Q Q Q Q Q
- . -.
50-
36 - u IB :anti-M2

 

 

 

50 ' ~ - - C- IB :anti-tubulin
1 2 3 4 5

 

 

Figure III-5: CtBPL (NAD) and CtBPL (DIM) mutants are expressed at low
steady state levels

Total extracts from single flies were prepared from transgenic lines that
express recombinant wild type or putative dehydrogenase mutants from CtBP
promoter (PC-CtBP). Extracts were analyzed by immunoblotting with anti-M2.
Consistent with low levels of CtBPL (NAD) or CtBPL (DIM) observed in
misexpression experiments, Figure Ill-2, we observed low expression of these
mutants when driven from the native CtBP promoter as well (lanes 3 and 5).
In comparison to low levels of protein observed in misexpression experiments,
CtBPs was clearly well expressed in this context (lane 2). The bottom panel
shows the same blot reprobed with B-tubulin as a loading control.

99

CtBPL (NAD) binding mutant Is not defective for dimer formation

NAD binds to a central region of CtBP that has been postulated to
mediate an allosteric modulation of the protein allowing dimer formation and in
vitro, proteolytic accessibility of the protein changes in the presence of NAD
(Kumar et al., 2002; Zhang et al., 2002). Data from the crystal structure of
vertebrate CtBP1 suggests that residues important for dimerization largely
overlap with the central NAD binding region (Kumar et al., 2002). In light of
our results with low expression levels of CtBPL (NAD) or CtBPL (DIM), we
thought it likely that both these mutants are affected similarly, namely loss of
NAD binding might prevent dimerization through an allosteric mechanism,
resulting in the loss of dimerization, just as mutating key residues in the dimer
interface might block physical association. In both these circumstances, we
hypothesized that the inability to form dimers might ultimately lead to
instability of proteins that might account for low steady state protein levels
observed.

To test this hypothesis, we sought to study the mutant Drosophila
proteins. In contrast to the presumptive catalytic mutant, the NAD binding and
dimerization mutants were expressed at low steady-state levels both when
overexpressed using the Gal4-driver as well as when expressed using the
8Kb CtBP regulatory region. This low expression level limits our ability to
physically characterize the protein produced in transgenic flies (Figure Ill-5).
Therefore, to determine if these mutants are compromised for dimer
formation, we expressed the wild type and mutants as His- and Flag-tagged

proteins in bacteria (~64 KDa), purified them over a Ni2°-NTA matrix and

100

200 K03 150 KDa 66 KDa 34 KDa

¢ L 8 i 10 l11

 

 

 

1 2 3 4 5 6 7 12
A M -—- CtBPL
B .“I-c— -—- —-—- _. .— CtBPL (CAT)

 

 

CtBPs

 

--- c—fi."

 

 

D j 8” H’ ..... j CtBPL(NAD)
E a. -w CtBPL (DIM)

 

 

 

 

Figure III-6: NAD binding mutant migrates In molecular mass consistent
with dimers but not the dimerization mutant

Peak fractions from Superdex 200 size exclusion experiments with wild type
CtBP and mutants were resolved on SOS-PAGE gels and western blotted
with anti-M2 antibody (to detect FLAG-tagged purified proteins). A pool of 3
fractions was loaded per lane. The CtBPL (CAT), CtBPs and CtBPL were found
to migrate as dimers (~150 KDa) consistent with their ability to associate with
partner CtBPs (Also see coimmunoprecipitations with CtBPs and CtBPL in
Figure Ill-3). The CtBPL (DIM) (his-tagged and FLAG-tagged) was seen to
peak in fractions in accordance with monomers. Significantly, the CtBPL
(NAD) mutant was observed in fractions corresponding to dimers, suggesting
that this mutant is able to associate with other CtBP proteins. Size markers
were resolved under the same conditions and are indicated in the figure.

101

analyzed them by gel filtration chromatography. When fractions were
analyzed by immunoblotting. CtBPL was observed to migrate in fractions
compatible with the size of a dimer, having a molecular mass of about 150
KDa. consistent with size exclusion results of recombinant vertebrate CtBP1
by others (Balasubramanian et al., 2003; Nardini et al., 2003). The
presumptive catalytic mutant that has activity comparable to the wild type
CtBPL, had a similar elution profile as CtBPL. CtBPs (~50 KDa) eluted in
slightly later fractions, but with a migration pattern consistent with dimer
formation, in accordance with our coimmunoprecipitation results. Unlike the
others, the CtBPL (DIM) mutant predominantly migrated in fractions consistent
with monomers. Interestingly, the CtBPL (NAD) mutant was observed to elute
in fractions of higher molecular mass similar to the profile of wild type CtBP,
suggesting that mutation of residues critical for NAD binding does not disrupt

dimerization.

102

DISCUSSION

The role of CtBP as an essential corepressor protein has been well
established but the significance of its dehydrogenase-like domain is still
unclear (Chinnadurai, 2006a). Previous studies aiming at similar questions
have relied upon cell culture techniques where contrasting results have been
obtained. Here, we sought to tease apart the contributions of its NAD binding
abilities, dimerization and its putative active site in the context of the whole

organism.

Dimerization is essential for CtBP functions

In our hands, misexpression of wild type full-length CtBP results in
developmental abnormalities. but interestingly only in specific settings. The
constitutive expression of CtBP proteins from both the tubulin promoter and
the CtBP promoter did not disrupt development, suggesting that such
outcomes are dependent on certain thresholds of misexpression that were
achieved using strong tissue specific drivers. The phenotypic consequences
of excessive corepressor activity may be due to direct repression of
endogenous targets by CtBP in eye/wing tissues, or indirect effects resulting
from the titration of cofactors required for regulation of other genes in these
fissues.

Repression mediated via CtBP might be established by associated
cofactors, with CtBP functioning as a tether to connect these repressive
activities to target promoters. Alternatively, CtBP may employ its putative

catalytic activity to mediate repression. A weak dehydrogenase activity has

103

been associated with vertebrate CtBP1, but the significance of this enzymatic
activity warrants further investigation into physiological substrates and
function (Balasubramanian et al., 2003; Kumar et al., 2002). Mutations
affecting the presumptive catalytic histidine analogous to those used to
inactivate dehydrogenases, did not impair the biological activity of CtBP as
shown here or its repression ability seen earlier (Sutrias-Grau and Arnosti,
2004). Analysis of the presumptive catalytic mutant shows that it can
associate effectively with other CtBP proteins; this attribute of dimerization
might not only likely mask the outcome of the catalysis defect but also
explains our misexpression phenotypes obtained with this mutant. However,
this putative catalytic residue is strictly conserved in metazoan CtBP proteins
and strongly speaks for conservation in function. Thus, the enzymatic function
if any, might pertain to either transcriptional repression or other cellular
functions. I propose that a role for this conserved residue might be best
investigated in a CtBP null context.

Dimerization amongst CtBP proteins appears to be a conserved
concept. Wild type isoforms of Drosophila CtBP can homo- and
heterodimerize in vivo. This dimer formation might form the mechanistic basis
for CtBP to nucleate of a repressive complex. with one monomer attaching to
the DNA bound repressor and the other monomer allowing its PXDLS site to
be contacted by repressive cofactors. CtBP may thus use its dimerization

ability to tether transcriptional repression activities to target promoters.

104

NAD binding and function

Substitutions of key glycines in the central NAD binding domain in
CtBP have been shown to inhibit NAD binding (Kumar et al., 2002; Nardini et
al., 2003). Our misexpression analyses with a CtBPL (NAD) mutant revealed
low steady state levels of this protein. Previous reports analyzing this mutant
fused to a Gal4 tether found a complete loss of transcriptional repression and
stable protein levels in vivo (Sutrias-Grau and Arnosti, 2004). The presence of
a heterologous DNA binding domain might help stabilize this protein by
directly tethering it to the promoter. The CtBPL (DIM) mutant also exhibited
low steady state levels. In light of prior studies that suggest that NAD binding
by CtBP allows oligomerization (Balasubramanian et al.. 2003), we
speculated that the reason for low steady state levels of the CtBPL (NAD) and
CtBPL (DIM) mutants might be a consequence of inability to dimerize. Our
size exclusion results with recombinant versions of these mutants raised in
bacteria showed that the NAD mutant was capable of migrating in fractions
consistent with dimers, even in the absence of NAD. This profile is in sharp
contrast to the DIM mutant that elutes predominantly in fractions consistent
with monomers. The CtBPL (NAD) elutlon profile is also somewhat different
from that observed for CtBPL. suggesting that the protein might adapt a
different conformation perhaps due to lack of NAD binding. Clearly, this
dimerization is not enough to rescue the transcriptional repression function of
this mutant when this mutant was tested for repression of an integrated
reporter. These results also rule out the earlier speculation that an inability to

dimerize might be responsible for low steady state levels of protein observed

105

in vivo. We speculate that in spite of dimer formation. the absence of NAD
binding prevents these dimers from undergoing a conformational change that
excludes binding to cofactors. This might be one plausible explanation for the

low steady state levels and loss of repression observed.

Future Considerations

Relevant to this idea that NAD induced conformational change is
needed to contact transcriptional cofactors, I propose to perform
coimmunoprecipitations to test for the association of candidate cofactors like
LSD1, de3 and a DNA binding repressor like Kriippel, with wild type vs. the
NAD binding and DIM mutants. A complementary approach will be to
compare the migration properties of epitope tagged CtBP wild type and
mutants generated in Drosophila by gel filtration chromatography. Here, the
ability of CtBP to migrate with cofactors will be tested. If the low steady state
levels of the CtBPL (NAD) mutant are indeed a result of lack of association
with cofactors, this mutant will be seen to migrate in fractions of lower
molecular weight and the CtBPL (DIM) mutant will adequately serve as a
control.

It has been documented that in the presence of NAD+/NADH, wild type
CtBPL migration on a sizing column is seen to shift to higher order oligomers .
Experiments with recombinant CtBPL demonstrate that both forms of the
nucleotide were effective in stimulating oligomerization of CtBP, which was

seen to migrate in fractions of higher molecular weight (Balasubramanian et

106

al., 2003). This observation prompted us to test the migration characteristics
of the recombinant CtBPL (NAD) mutant relative to the wild type CtBPL, in the
presence and absence of nucleotides. However. in our hands the elution
profile of recombinant wild type CtBPL protein in the absence vs. presence of
NAD+ remained unchanged (P.Mani, data not shown). Further experiments
are warranted to test if the CtBPL (NAD) mutant is unresponsive to the effects
of NAD and migrates in similar fractions even in the absence of NAD. If NAD
induces a conformational change in the protein that enhances oligomerization,
then the CtBPL (NAD) mutant should be completely unaffected, and its

observed migration as dimers eluting at ~150 KDa should remain unchanged.

107

REFERENCES

Arnosti, D. N. (2004). Multiple Mechanisms of Transcriptional Repression in
Eukaryotes. In Handbook of Experimental Pharmacology, vol. 166 (ed.
Gossen M. Kaufmann J. and Triezenberg S.J.): Springer(Berlin).

Balasubramanian. P., Zhao, L. J. and Chlnnadural, G. (2003). Nicotinamide
adenine dinucleotide stimulates oligomerization, interaction with adenovirus
E1A and an intrinsic dehydrogenase activity of CtBP. FEBS Lett 537, 157-60.

Boyd, J. M., Subramanian, T., Schaeper, U., La Regina, M., Bayley, S. and
Chinnadurai, G. (1993). A region in the C-terminus of adenovirus 2/5 E1a
protein is required for association with a cellular phosphoprotein and
important for the negative modulation of T24-ras mediated transformation,
tumorigenesis and metastasis. Embo J 12, 469-78.

Brand. A. H. and Perrimon, N. (1993). Targeted gene expression as a
means of altering cell fates and generating dominant phenotypes.
Development 1 1 8, 401 -1 5.

Chlnnadural. G. (2002). CtBP, an unconventional transcriptional corepressor
in development and oncogenesis. Mol Cell 9, 213-24.

Chlnnadural, G. (2006a). CtBP family proteinszunique transcriptional
regulators in the nucleus with diverse cytosolic functions: Landes Bioscience.

Chinnadurai, G. (2006b). Cth. a candidate tumor susceptibility gene is a
team player with luminaries. Biochim Biophys Acta 1765, 67-73.

Corda, D., Colanzl, A. and Luinl, A. (2006). The multiple activities of
CtBP/BARS proteins: the Golgi view. Trends Cell Biol 16, 167-73.

Criqul-Flllpe. P., Ducret, C., Malra, S. M. and Wasylyk, B. (1999). Net, a
negative Ras-switchable TCF, contains a second inhibition domain, the CID,
that mediates repression through interactions with CtBP and de-acetylation.
Embo J 18, 3392-403.

Furusawa, T., Moribe, H., Kondoh, H. and ngashl. Y. (1999). Identification
of CtBP1 and CtBP2 as corepressors of zinc finger-homeodomain factor
deltaEF 1. Mol Cell Biol 19, 8581-90.

Gray, S. and Levine, M. (1996). Transcriptional repression in development.
Curr Opin Cell Biol 8, 358-64.

Grooteclaes, M., Deveraux, Q., Hlldebrand, J., Zhang, Q., Goodman, R. H.
and Frlsch, S. M. (2003). C-terminal-binding protein corepresses epithelial
and proapoptotic gene expression programs. Proc Natl Acad Sci U S A 100,

4568-73.

108

Hildebrand. J. D. and Soriano, P. (2002). Overlapping and unique roles for
C-terrninal binding protein 1 (CtBP1) and CtBP2 during mouse development.
Mol Cell Biol 22. 5296-307.

Imai, 8., Armstrong, C. M., Kaeberlein, M. and Guarente, L. (2000).
Transcriptional silencing and longevity protein Sir2 is an NAD-dependent
histone deacetylase. Nature 403, 795—800.

Kegel, K. B., Meloni, A. R., YI, Y., Kim, Y. J., Doyle, E., Culffo. B. G., Sapp,
E., Wang, Y., an, Z. H., Chen. J. D. et al. (2002). Huntingtin is present in the
nucleus, interacts with the transcriptional corepressor C-terminal binding
protein, and represses transcription. J Biol Chem 277, 7466-76.

Keller, S. A., Mao, Y., Struffi, P., Margulles. C., Yurk, C. E., Anderson, A.
R., Amey, R. L., Moore. 8., Ebels, J. M., Foley, K. et al. (2000). dCtBP-
dependent and -independent repression activities of the Drosophila Knirps
protein. Mol Cell Biol 20, 7247-58.

Kim. J. H., Cho, E. J., Kim, S. T. and Youn, H. D. (2005). CtBP represses
p300-mediated transcriptional activation by direct association with its
bromodomain. Nat Struct Mol Biol 12, 423-8.

Koipally, J. and Georgopoulos, K. (2000). lkaros interactions with CtBP
reveal a repression mechanism that is independent of histone deacetylase
activity. J Biol Chem 275. 19594-602.

Kumar, V., Carlson, J. E., Ohgl. K. A., Edwards, T. A., Rose, D. W.,
Escalante, C. R., Rosenfeld, M. G. and Aggarwal, A. K. (2002).
Transcription corepressor CtBP is an NAD(+)-regulated dehydrogenase. Moi
Cell10, 857-69.

Nardini, M., Spano. S., Cericola, C., Pesce, A., Massaro, A., Millo, E.,
Luini, A., Corda, D. and Bolognesl, M. (2003). CtBP/BARS: a dual-function
protein involved in transcription co-repression and Golgi membrane fission.
Embo J 22, 3122-30.

Rutter, J., Relck. M., Wu. L. C. and McKnight, S. L. (2001). Regulation of
clock and NPASZ DNA binding by the redox state of NAD cofactors. Science
293, 510-4.

Schmitz, F., Konlgstorfer, A. and Sudhof, T. C. (2000). RIBEYE, a
component of synaptic ribbons: a protein's journey through evolution provides
insight into synaptic ribbon function. Neuron 28. 857-72.

Senyuk, V., Sinha, K. K. and Nuclfora. G. (2005). Corepressor CtBP1
interacts with and specifically inhibits CBP activity. Arch Biochem Biophys
441, 168-73.

Shl, Y., Sawada, J., Sui. G., Affar el, 8., Whetstlne, J. R., Lan. F., Ogawa,
H., Luke. M. P., Nakatanl, Y. and Shl, Y. (2003). Coordinated histone
modifications mediated by a CtBP co-repressor complex. Nature 422, 735-8.

109

Struffi, P. and Arnosti, D. N. (2005). Functional interaction between the
Drosophila knirps short range transcriptional repressor and RPD3 histone
deacetylase. J Biol Chem 280, 40757-65.

Struffi. P., Corado, M., Kulkarni, M. and Arnosti, D. N. (2004). Quantitative
contributions of CtBP-dependent and -independent repression activities of
Knirps. Development 131, 2419-29.

Sutrias-Grau, M. and Arnosti, D. N. (2004). CtBP contributes quantitatively
to Knirps repression activity in an NAD binding-dependent manner. Mol Cell
Biol 24, 5953-66.

Thio, S. S., Bonventre, J. V. and Hsu, S. I. (2004). The CtBP2 co-repressor
is regulated by NADH-dependent dimerization and possesses a novel N-
terminal repression domain. Nucleic Acids Res 32, 1836-47.

Zhang, H. and Levine, M. (1999). Groucho and dCtBP mediate separate
pathways of transcriptional repression in the Drosophila embryo. Proc Natl
Acad Sci U S A 96, 535-40.

Zhang. 0., Piston, D. W. and Goodman, R. H. (2002). Regulation of
corepressor function by nuclear NADH. Science 295. 1895-7.

110

CHAPTER IV

CONCLUSIONS AND FUTURE DIRECTIONS

Drosophila serves as an exemplary model to study the spatial control
of gene activity in development, during which boundaries of gene expression
are resolved by spatially restricted activity or localization of transcriptional
repressors (Mannervik et al., 1999). Studies on transcriptional repression in
the Drosophila embryo have identified two basic modes of repression: long-
range and short-range repression (Gray and Levine, 1996). Short-range
repression represents a more flexible, locally acting kind of gene regulation
where repressors are documented to primarily act by ‘quenching’ of activator
proteins or directly repress basal machinery components (Arnosti, 2004;
Hewitt et al., 1999). In contrast, long-range repressors function more globally
blocking multiple enhancers in a vicinity of several kilobases. As discussed in
Chapter I, short-range repressors work with the evolutionarily conserved
corepressor protein CtBP through a consensus sequence while long-range
repressors mediate their effect through the corepressor Groucho. The
different activities of of each repressor class are believed to reflect distinct
mechanisms at play; however the molecular events after corepressor
recruitment are not clearly understood. Some long-range repressors are also
seen to interact with CtBP and the role of CtBP in this context is ambiguous.
For instance. CtBP binding to Hairy has been shown to antagonize repression
while in the context of Runt long-range repressor, CtBP is apparently needed
for maintaining repression of Runt targets in the early embryo (Phippen et al.,

2000; Wheeler et al., 2002).

111

CtBP is a major contributor to the activity of short-range repressors
during development and has been demonstrated to modulate the activities of
oncogenes, signaling pathways and apoptosis (Chinnadurai, 2006). The
precise mechanism of action of CtBP has been under intense scrutiny both in

vertebrate and in invertebrate model organisms.

CtBPL and CtBPs - developmentally regulated roles

In vertebrates, multiple genes for CtBP are seen to encode for splice
variants that execute divergent functions (Discussed in Chapter I). The fly
CtBP gene also encodes for two major isoforms (CtBPL and CtBPs) of
unidentified functions. Both forms of the proteins show comparable repression
potential (Sutrias-Grau and Amosti. 2004). CtBPL differs from CtBPs by
possessing an unstructured C-terrninal tail that is a target for several post-
translational modification events. My analysis of isoform expression reveals a
conserved developmental shift in the levels of CtBPL isoform at later stages
(Mani-Telang and Arnosti, 2006). This is a first demonstration highlighting
temporal expression differences between these splice variants and provokes
the key questiOn regarding isoform specific needs for specific functions during
development. The observation that genomes of related Dipterans and other
organisms from Hymenoptera and Coleoptera, all contain CtBPL coding
information and show the presence of multiple isoforms, suggests that the
existence of multiple isoforms is a widely conserved trait, and may have

important functional ramifications. My data is in accordance with in vivo mice

112

CtBP knockouts (Hildebrand and Soriano, 2002). Mice null for CtBP1 and
CtBP2 suggest unique and overlapping phenotypes, different isoforms are
regulated through different post-translational modifications; lending support to

the hypothesis of different isoforms having specific functions.

Mechanisms behind the observed down-regulation of CtBPL

As described earlier, developmental study of isoform expression
reveals a conserved shift in the levels of CtBPL isoform at later stages. The
evident drop in CtBPL protein levels can be traced back to transcript
abundances. My investigations of relative transcript levels by RT-PCR
measurements indicate that CtBPs mRNA is upregulated relative to CtBPL at
later developmental times, which is mirrored in the protein profile with CtBPs
protein being more predominant in adults.

In addition to splicing preferences, the decrease in CtBPL levels may
also be a consequence of regulatory post-translational modification events
that target the unique C-terrninus of the protein. Numerous such modifications
have been identified that act on individual vertebrate isoforms and serve to
regulate isoform localization (availability) or abundance (discussed in Chapter
1-Section 1.3). SUMOylation of vertebrate CtBP1 is required for its nuclear
translocation and disrupting this modification has a profound effect on CtBP’s
localization and corepressor function (Lin et al., 2003). My bioinfonnatic
analysis comparing sequence information for CtBPL tail among highly

divergent organisms illustrates that SUMOylation motifs are preserved (See

113

Chapter II - Figure "-2). Previous demonstration of SUMOylation affecting
vertebrate CtBP function and the conservation of SUMO motifs in divergent
organisms suggests SUMO regulation of CtBP must be operative in
Drosophila as well. It will be interesting to determine the role of SUMOylation
on CtBP function in a developmental setting. Since SUMOylation is known to
alter the localization of CtBP1, the subcellular localization of tagged protein
forms can also be tested by transfections in Drosophila cells.

CtBP proteins are also phosphorylated; one such phosphorylation
event by HIPK2 kinase at Ser-422 in the C-terminus results in ubiquitination of
vertebrate CtBP followed by proteasomal clearance, which can be prevented
by proteasomal inhibitor treatment (Zhang et al., 2005; Zhang et al., 2003).
Preliminary evidence with the proteasomal inhibitor MG-132 shows that in
Drosophila 82 cells. CtBPL levels can be stabilized in comparison to CtBPs
levels, that stay constant (Y. Zhang, unpublished results). The C-terrnini of
vertebrate vs. fly CtBP proteins differ in sequence, but are conserved in
overall length and are abundant in disorder promoting residues (P.Mani, data
not shown). In this region of Drosophila CtBP, several serine residues exist
that could serve as potential phosphorylation sites. Analysis of truncated
derivatives of the tail region that eliminate SUMOylation and phosphorylation
sites will help shed light on the importance of these modifications. CtBPs, lacks
this stretch of the C-terminus and is the predominantly expressed protein at all
developmental times; in light of our developmental expression data, it will
prove to be interesting to test if truncated CtBPL derivatives are converted to
more stable proteins. Taking these studies one step further, the localization of

these truncated derivatives can be studied. It is likely that a deletion disrupting

114

the conserved SUMO motifs from Drosophila CtBPL might result in its
translocation to the cytoplasm. Along the same lines, it will be interesting to
study the repression activity of these deletion mutants as translocation of

CtBP into the cytoplasm will interfere with its corepressor functions.

Functional equivalency of CtBP isoforms In Drosophila

Understanding the biological functions of CtBP isoforms has been a
major challenge. In previous studies, comparing CtBPL and CtBPs activity, no
differences have been uncovered (Fang et al., 2006; Sutrias-Grau and
Amosti. 2004). The observed decline in Drosophila CtBPL levels are thus very
interesting and could be interpreted as CtBPs furnishing CtBPL like functions
in later stages. It is possible that CtBPL function is no longer required during
these developmental periods and is hence down regulated, in which case
genes targeted specifically by CtBPL should be derepressed. One motive for
the misexpression strategy was to compare the biological activity of CtBPL
and CtBPs. However, due to low expression levels of the short isoform the
results were inconclusive. Further analysis using coimmunoprecipitations and
size exclusion chromatography demonstrated that CtBPs proteins and CtBPL
proteins can homo- and heterodimerize. This observation tells us that an
appropriate assessment of individual isoform function can only be provided in
a CtBP null background, where dimerization is not feasible.

To determine if these isoforms are functionally redundant or exist for

unique function, a whole animal assay of CtBP activity is proposed. As

115

mentioned earlier, the motivation behind the experiment is to be able to assay
individual isoform activity in a CtBP null backdrop. In Drosophila,
homozygous mutations in CtBP are lethal. I identified the endogenous
regulatory region required for expression of CtBP and created transgenic lines
expressing CtBPL and CtBPs from this endogenous CtBP promoter. My initial
attempts to rescue this lethality using transgenes for CtBPL alone, CtBPs
alone or both did not suffice. Several reasons for this lack of replacement are
discussed in detail in the following section (Appendix A). Instead of a
transgene based rescue, a more sureshot way of assaying for function is to
attempt a genomic rescue. In this experiment. the entire genomic locus
including all the regulatory information for CtBP will be introduced into flies,
and assessed for its ability to rescue CtBP nulls. A successful rescue will
permit us to make modifications in this genomic locus (for e.g. block splice
site that generates CtBPL) and answer the original question regarding whether

the fly can survive on the basis of any one isoform.

116

Structure- function studies

CtBP proteins resemble dehydrogenases in having a conserved
Rossman fold required for NAD binding and a catalytic triad (His/Glu/Arg)
conserved in all enzymes where the histidine forms the active site center
(Chinnadurai, 2002). Consistent with the presence of the Rossman fold,
vertebrate CtBP1 and 2 have been documented to bind with NAD+ and
NADH, with contrasting views on which metabolite has a higher affinity for
CtBP (Kumar et al., 2002; Zhang et al., 2002). Based on these views, it has
been suggested that CtBP may regulate transcription in a redox-responsive
manner (Zhang et al.. 2002). In addition, a weak enzymatic activity has been
noted with CtBP1, which adds a layer of complexity to solving the mechanism
of CtBP-mediated repression. Previous work has suggested that NAD binding
promotes multimerization of the protein in vitro (Balasubramanian et al., 2003)
and possibly allows interaction with PXDLS-containing cofactors by inducing a
conformational alteration in the protein (Kumar et al., 2002). However, the
temporal sequence of molecular events has not been tested. It is not known if
dimerization is required for cofactor association, if the NAD induced
conformational change is mandatory for cofactor binding or if individual CtBP
monomers are already bound to cofactors and are recruited through
dimerization with another CtBP monomer. associated promoter bound
repressor.

Since these structural features are intact in the fly CtBP proteins, we
initiated studies on the relevance of these features by making mutations in

conserved residues and assaying them in a whole animal context. Our NAD

117

binding mutant was ineffective as a corepressor while a catalytic site deficient
mutant retained full repressive potential, giving us the first indication that NAD
binding is essential for repression (Sutrias-Grau and Arnosti, 2004). My
misexpression studies with the wild type and mutant proteins also painted a
consistent picture with the catalytic mutant showing comparable activity as
wild type CtBPL; however, the NAD binding mutant was expressed poorly.

When tested for dimerization, CtBPL, CtBPs and the catalytic mutant
were able to dimerize. Since these proteins were also seen to retain
repressive abilities, it is likely that dimerization allows for cofactors to be
recruited. A stepwise recruitment model can be envisioned wherein a DNA
binding repressor may recruit CtBP as a corepressor. CtBP may then
dimerize (or bind as a dimer), and then use its dimer partner to recruit
repressive cofactors. Towards this aim, I generated a dimerization mutant
that also yielded the same results as the NAD binding mutant and was poorly
expressed. Since the residues for dimerization map along the NAD binding
region (Kumar et al., 2002), it was my hypothesis that probably the reason
these proteins are expressed poorly is because of defects in CtBP
dimerization. l speculated that the NAD binding mutant cannot bind NAD and
hence cannot undergo the conformational change that allows for dimerization.
while the dimerization mutant can bind NAD but is restricted by mutations in
critical residues needed to contact its dimer partner.

Since the NAD binding and the dimerization mutant were both poorly
expressed in flies (with several promoter constructs), l generated purified
recombinant proteins in bacterial and fractionated them by size exclusion

chromatography. The wild type CtBPL and CtBPs pure proteins resolved in

118

fractions of about 150KDa. indicative of dimers. I observed the catalytic
mutant to migrate in similar size fractions and coimmunoprecipitate with the
wild type proteins, suggesting that this mutant could associate with another
CtBP monomer. This result explains my misexpression data and repression
data from our previously published study, wherein the repressive activity of
this mutant could perhaps be a result of association with endogenous CtBP
proteins. This result emphasizes that in order to be able to ascribe a role for
the dehydrogenase activity associated with CtBP, one needs to evaluate it in
a CtBP null background. In accordance with my hypothesis. the dimerization
mutant peaked in monomeric fractions. This is a first demonstration of the
inability of this mutant to dimerize. However, this mutant has been shown to
be unable to contact a PXDLS partner, E1A in cell culture based studies
suggesting that the protein will be ineffective as a repressor (Kumar et al.,
2002). It will be interesting to test if this mutant is active for repression in
Drosophila. If the hypothesis that dimerization preceded cofactor recruitment
is valid, then this mutant is predicted to fail as a repressor.

Surprisingly. our hypothesis with the NAD binding mutant proved to be
incorrect. This mutant was seen to migrate in fractions consistent with dimers
of CtBP, suggesting that NAD binding is not mandatory for associating with a
dimer partner, as has been speculated. This leaves open the question of why
the NAD binding mutant is expressed poorly and why does it lack repressive
function. My speculation is that perhaps dimers formed in the absence of NAD
are not functional. NAD acts as a switch to bring about a change in
conformation of CtBP; it is possible that monomers of CtBP can stick to each

other in the absence of NAD, but their PXLDS responsive clefts are not

119

exposed/available. This might prohibit interaction with cofactors and is a likely
explanation for the lack of repression observed.

I will test this hypothesis by performing coimmunoprecipitations, similar
to those described in Chapter III, to probe for a lack of cofactor associations
with the NAD binding and the dimerization mutant. Recombinant CtBPL
proteins expressed in files are seen to associate with a histone demethylase
LSD1 and BAF53 in coimmunoprecipitation studies. Using extracts from files
expressing the NAD binding and dimerization mutants, I will investigate
whether interaction with these cofactors is abolished. Another important step
leading to repression is the recruitment of CtBP corepressor by DNA binding
proteins to target promoters. I shall also determine if the mutants can
associate with a DNA binding repressor like Knirps or Kruppel (short-range
repressors in the fly). As a complementary approach, fly extracts expressing
wild type CtBP versus the NAD binding and dimerization mutants will be
fractionated on a sizing column; here I will be assaying the ability of these
proteins to traverse in high molecular weight fractions indicative of a CtBP
complex, containing its suite of repressive cofactors. In light of previous
studies, the dimerization mutant should serve as a control for non-specific
DNA mediated interactions, by migrating as monomeric units, if the
hypothesis is accurate. These complementary approaches focusing on the
relevance of NAD binding and dimerization will throw light on CtBP-mediated
molecular mechanisms of repression.

In conclusion, the Drosophila CtBP protein is regulated in more ways
that currently known. The studies described herein provide insights into two

very crucial aspects of CtBP mediated repression during development -

120

1)

2)

Its conserved developmental regulation strongly suggests distinct
functions during development.
NAD binding is not a pre-requisite for dimerization of CtBP proteins,

but may be required for cofactor association.

121

REFERENCES

Arnosti, D. N. (2004). Multiple Mechanisms of Transcriptional Repression in
Eukaryotes. In Handbook of Experimental Pharmacology. vol. 166 (ed. K. J. a.
T. S. J. Gossen M.): Springer(Berlin).

Balasubramanian, P., Zhao, L. J. and Chlnnadural. G. (2003). Nicotinamide
adenine dinucleotide stimulates oligomerization, interaction with adenovirus
E1A and an intrinsic dehydrogenase activity of CtBP. FEBS Lett 537. 157-60.

Chlnnadural, G. (2002). CtBP, an unconventional transcriptional corepressor
in development and oncogenesis. Mol Cell 9. 213-24.

Chlnnadural, G. (2006). CtBP family proteins:unique transcriptional
regulators in the nucleus with diverse cytosolic functions: Landes Bioscience.

Fang, M., Li, J., Blauwkamp, T., Bhambhani, C., Campbell. N. and
Cadlgan, K. M. (2006). C-terrninaI-binding protein directly activates and
represses Wnt transcriptional targets in Drosophila. Embo J 25, 2735-45.

Gray. S. and Levine, M. (1996). Transcriptional repression in development.
Curr Opin Cell Biol 8, 358-64.

Hewitt, G. F., Strunk, B. 8., Margulles. C., Prlputln. T., Wang, X. D., Amey,
R., Pabst, B. A., Kosman. D., Relnltz, J. and Arnosti, D. N. (1999).
Transcriptional repression by the Drosophila giant protein: cis element
positioning provides an alternative means of interpreting an effector gradient.
Development 1 26, 1 201 -1 0.

Hildebrand, J. D. and Soriano, P. (2002). Overlapping and unique roles for
C-terrninal binding protein 1 (CtBP1) and CtBP2 during mouse development.
Mol Cell Biol 22, 5296-307.

Kumar, V., Carlson, J. E., Ohgl. K. A., Edwards, T. A., Rose, D. W.,
Escalante, C. R., Rosenfeld, M. G. and Aggarwal, A. K. (2002).
Transcription corepressor CtBP is an NAD(+)-regulated dehydrogenase. Mol
Cell 10. 857-69.

Lin, X., Sun, 8., Liang, M., Llang. Y. Y., Gast, A., Hildebrand. J.,
Brunlcardl, F. C., Melchior, F. and Feng, X. H. (2003). Opposed regulation
of corepressor CtBP by SUMOylation and PDZ binding. Mol Cell 11, 1389-96.

Manl-Telang, P. and Arnosti, D. N. (2006). Developmental expression and
phylogenetic conservation of alternatively spliced forms of the C-terrninal
Binding Protein corepressor. Development. Genes and Evolution In Press.
Mannervik. M., Nibu, Y., Zhang, H. and Levine, M. (1999). Transcriptional
coregulators in development. Science 284, 606-9.

122

Phippen, T. M., Swelgart, A. L., Moniwa, M., Krumm, A., Davie, J. R. and
Parkhurst, S. M. (2000). Drosophila C-terminal binding protein functions as a
context-dependent transcriptional co-factor and interferes with both mad and
groucho transcriptional repression. J Biol Chem 275, 37628-37.

Sutrias-Grau, M. and Arnosti, D. N. (2004). CtBP contributes quantitatively
to Knirps repression activity in an NAD binding-dependent manner. Mol Cell
Biol 24, 5953-66.

Wheeler. J. C., VanderZwan, C., Xu, X., Swantek. D., Tracey. W. D. and
Gergen, J. P. (2002). Distinct in vivo requirements for establishment versus
maintenance of transcriptional repression. Nat Genet 32, 206-10.

Zhang, Q., Nottke, A. and Goodman, R. H. (2005). Homeodomain-
interacting protein kinase-2 mediates CtBP phosphorylation and degradation
in UV-triggered apoptosis. Proc Natl Acad Sci U S A 102, 2802-7.

Zhang, 0., Piston, D. W. and Goodman, R. H. (2002). Regulation of
corepressor function by nuclear NADH. Science 295, 1895-7.

Zhang, 0., Yoshimatsu, Y., Hildebrand, J., Frlsch, S. M. and Goodman, R.

H. (2003). Homeodomain interacting protein kinase 2 promotes apoptosis by
downregulating the transcriptional corepressor CtBP. Cell 115, 177-86.

123

 

APPENDIX A

RESCUE OF CtBP FUNCTION

CtBP function is crucial for the regulation of early gene expression in
Drosophila. Loss-of-function mutations in the gene are homozygous lethal
with lethality occurring in during the late larval to early pupal stages (Poortinga
et al., 1998) . Studies on vertebrate CtBP proteins have provided much data
on the potential roles of CtBP in cell culture, but these studies have not shed
light on the role of CtBP during development. Drosophila serves as an
excellent model to perform whole organism genetic manipulations with ease.

The CtBP gene encodes for two major splice forms, which are both
detected throughout all developing stages of the fly (Chapter I. Figure l-3)
(Sutrias-Grau and Arnosti, 2004). However, there is a marked reduction in
levels of CtBPL isoform from larval through adult stages. unlike CtBPs levels
that stay relatively constant. I found that the changing abundance of distinct
mRNA spliceforrns might be the determinants of the developmental changes
in protein profile. In addition. post-translational modifications might regulate
overall levels of this protein (Mani-Telang and Arnosti, 2006). It is unknown if
different isoforms provide distinct functions during development. The splice
variants differ by a C-terminal extension present in CtBPL and in its place,
CtBPs has a short stretch of 10 aa. Alternative splicing might provide
specificity of CtBP functions during fly development. Because CtBPL is
virtually identical to all CtBP proteins except for the C-terminal residues
(similar in length, but vary in sequence) it is possible that this form of the

protein can provide all functions of CtBP. On the other hand, a developmental

124

profile shows CtBPs to be the prevalent protein during the life cycle of the fly
(Chaper ll-Figure Il-1C). In order to test potential redundancies between
isoforms, I tested the ability of CtBPL and CtBPs to rescue a null mutant. This
assay serves as a rigorous test of individual isoform function in light of our
previously described analyses (Chapter II and Ill).

Two null mutations have been described for CtBP. the first is a P-
element insertion in a 5’UTR of CtBP (#P-1590) (FigureA-1) and the second is
a deletion around the same segment (#1663), both of which do not
complement each other (Grumbling et al., 2006). I introduced a cDNA
fragment for the long isoform into pCasper—tubulin vector, allowing expression
under the control of a constitutively active tubulin promoter in files (Basler and
Struhl. 1994). This promoter is active early during Drosophila embryogenesis
and protein expression can be detected ubiquitously by immunostaining, in a
manner that mimics endogenous CtBP expression (data not shown).
However, unlike the results obtained from tissue specific misexpression
discussed in Chapter 3, the ubiquitous ectopic expression of the CtBPL
isoform did not induce any phenotypes. Furthermore, I was unable to recover
any survivors in a CtBP-l- mutant background. The failure to rescue might be
attributed to the lack of compensation of the CtBPs isoform, insufficient levels
or patterns of induced protein, or interference from the C-terminal flag-tag.

To faithfully mimic the endogenous CtBP expression levels and
patterns in vivo, the genomic locus of CtBP was searched for regulatory
information. The cis-regulatory elements governing CtBP transcription have

not been previously explored. The entire locus is contained within a ~15 Kb

125

. 8 Kb promoter element ‘ CtBPs 3 Kb intron
+——+

V

CG8031” C612360

I” I" CtBPL
I I

A

P-element insertion
# P-1590

 

 

Figure A-1: Genomic organization of the Drosophila CtBP locus

CtBP gene locus extends over a ~15 Kb region and encodes two splice
variants. Two predicted transcriptional start sites have been located, depicted
as red boxes in the cartoon. A P-element insertion described for CtBP has
been mapped to lie upstream of the coding sequence (Grumbling et al., 2006;
Poortinga et al., 1998). An 8Kb genomic fragment 5’ of the gene was used to
drive expression of CtBP transgenes. Exons (black boxes) coding for CtBPL
and CtBPs are indicated by dotted lines and occupy about 7 Kb of genomic
sequence. The two additional exons that constitute the C-terminal tail of
CtBPL are separated from the core CtBP sequences (exons 1-5) by a large
intron of ~3 Kb that seems to be conserved in Drosophilids. CtBPs has a short
3’ UTR of ~60 bp (light blue triangle) while CtBPL has a long ~700 bp UTR
(green triangle).

126

4 CtBPL-FLAG

  

' <1 CtBPs-FLAG
IB :anti-M2

 

'_ I.s‘~’"“‘v‘uflrlv
ss-P—m-

1 2 3

 

 

IB :anti-tubulin

 

Figure A-2: Expression of CtBP Isoforrns under the control of
endogenous promoter

An 8Kb upstream region directly upstream of the 5’ most protein coding exon
for CtBP was used to drive expression of FLAG-epitope tagged forms of
CtBPL and CtBPs. Aline containing both of these transgenes was obtained by
recombination (See Materials and Methods). Single adults from transgenic fly
lines expressing these isoforms were solubilized in Laemmli sample buffer,
loaded on 10% SDS-PAGE gels and analyzed by anti-M2 antibody to detect
FLAG-tagged CtBP. Native promoter driven CtBPs is expressed at levels
lower than CtBPL. detected both by western blots and immunostaining of
stage-specific embryos. which is in sharp contrast to the endogenous protein
profile. The bottom panel shows B—tubulin as a loading control.

127

0

Q 0 Q6)
50 ’05],

’b
06 r560

 

50 l q m “ n J IB :anti-M2

 

--—- -- - <1 CtBPL-FLAG
50- - — —— W‘“ dCtBPL

* i U"
- I— a - ‘- <CtBPS
36 IB :anti-CtBP

55- “3| IB :anti-tubulin

1 2 3 4

 

 

 

 

 

Figure A-3: Expression of recombinant CtBPL assayed by Western
blotting

To determine if levels of recombinant FLAG-tagged CtBPL decrease at later
developmental times as seen with endogenous CtBPL, western blots were
performed on stage-specific soluble extracts from transgenic fly lines
expressing CtBPL under the CtBP promoter. Western blotting with anti-M2
(first panel) to detect FLAG-tagged CtBPL, did not show any appreciable
decrease in protein levels at later developmental stages, suggesting that the
protein is not expressed at lower levels in larva and adult. However, when the
same blot was stripped and reprobed with anti-CtBP antibody, a clear
decrease in CtBPL-FLAG similar to endogenous CtBPL was observed.
suggesting that expression patterns (lmmunostaining, data not shown) and
expression levels were comparable to native protein. The discrepancy
between the two antisera could be on account of the CtBP antisera’s limited
cross-reactivity to modified forms of CtBP. The band marked with an asterisk
in 12-20 hr. extract is unknown. The blot was stripped and reprobed again to
show B-tubulin as a loading control.

128

region and the genomic organization seems to be conserved in Drosophilids
(data not shown). Exons coding for CtBP are contained within ~6.8 Kb region
with two predicted transcriptional start sites located -2 Kb and -8 Kb upstream.
of the translational initiation codon. In an effort to drive expression of CtBP
using its endogenous regulatory elements, cDNA fragment specific to each
isoform was expressed under the control of either -4 Kb or -8 Kb of genomic
sequence (Figure A-1). The shorter 4 Kb 5’ region did not induce detectable
CtBP protein expression (data not shown). In contrast, the entire upstream
region of ~8Kb yielded several transgenic lines which expressed CtBP.
suggesting that ectopic expression from this promoter does not disrupt normal
development (Figure A-2).

The rescue assay was devised to test for complementation in two
steps: the first step tests rescue of the loss of zygotic CtBP (rescue of a
homozygous null). If successful, the next stage tests it the rescued progeny
can overcome the lack of both zygotic and maternally deposited CtBP. To test
if CtBPL could compensate for endogenous CtBP functions, suitable
transgenic lines (P-element insertion lines mapped to the second
chromosome by standard genetic crosses) expressing CtBP promoter::CtBP._-
FLAG transgene under control of the native promoter was crossed into a
CtBP mutant background (See Rescue scheme, Figure A-4). Such promoter
driven transgene fusions have been effectively used for phenotypic rescue

(Donaldson et al., 2004; Moran and Jimenez, 2006; Wagner et al., 2002).

129

Figure A-4: Genetic crosses for whole animal rescue of CtBP mutant
phenotype

#P-1590 is a P-element insertion that is mapped to an upstream 5' UTR in
CtBP and #1663 is a deletion mutant in the same region and the two alleles
do not complement each other (yellow boxes) (Grumbling et al., 2006). CtBPL.
CtBPs and a recombinant chromosome containing both the transgenes were
assessed individually for rescue by crossing into different CtBP mutant
backgrounds as shown in the first step of the scheme. Three chromosomes of
Drosophila are drawn; with the CtBP mutations indicated the third
chromosome (small black squares) and the P-element on the second. where
‘P’ refers to the P-element (CtBPL, CtBPs and recombinant CtBP) driven by
the 8Kb promoter region (see bottom of the figure). All chromosomes were
marked with balancers to prevent recombination and distinct markers for
identification of balancers. In the second step, P-element strains expressing
CtBPL or CtBPs or both, were introduced into null backgrounds. The two
different null alleles of CtBP were tested together to avoid possible lethality by
additional accumulated lethals on the CtBP mutant chromosomes (shown as
red X). In the third step, transheterozygous CtBP P-1590/1663 individuals
were sought as indicated by the loss of the SMZCyo balancer, signifying
rescue from either one copy or two copies of the P-element.

130

Itanl

'R in
teles
[BPLI
were
Lianl
es of
third
here
In by
were
5 for
Islng

two
way
In as
luals
W9

nent

 

 

 

 

 

W

Rescue with one copy of transgene

 

 

 

 

SM2 Cyo

 

 

SM2 Cyo

8Kb promoter

Vr—z

PC-CtBP

 

 

TM3 Sb

CtBP-
1663

 

 

TM3 Sb

CtBP-

1663
—e—

'-9<E--4|-

CtBP-
P-1590

:3:

P-element construct

«0

OR

131

 

 

 

 

 

CtBP-

 

 

 

 

 

IEIIEI

 

 

 

+ 1663/P4590
SM2 Cyo TM3 Ser
P CtBP-
P-1590
SM2 Cyo TM3 Sb
P CtBP-
P-1590
—x—+-
—|-
P CtBP-
1663

 

Rescue with two capies of transgene

It-

 

Table A-1: CtBP promoter driven transgenes do not rescue CtBP
mutants

Either one or two copies of transgenes were assessed for rescue in CtBP
mutant backgrounds. Progeny were scored by genotype and categorized. All
flies recovered from the final cross (Figure A-4) carried a third chromosome
balancer indicting that neither CtBPL, CtBPs nor both together were able to
rescue lethality. Progeny carrying second chromosome P-element can be
homozygous or heterozygous (with a balancer). These progeny with balancer
chromosomes are generally sick and were consistently recovered at lower
ratios.

132

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o o as no $903.58
69.830
0 c on F mm mamuoéeEemno
o o E mm $998388
:28. Scene:
2.... 889:6 ago
o\oo Q o Q om o\ o wDOwwm 02 u:
o o o m touoecxm
a9 _ 89¢
in .nm6 90 gm am 9: a 8 9.: so Sm
T Itlxl l
ITIXI T >C®DOCQ
8%.? . aw . . . _
dew . -..eo ._ fig. . | ..

 

 

 

am m5 96 Sm Q 8 ms: :6 Sm

.. 89¢
.35

X

, as. ,.
_....n.m8.

133

A heteroallelic CtBP mutation combination was tested to avoid mortality
resulting from other uncharacterized recessive mutations that may be present
on these chromosomes. All chromosomes contained identifiable markers to
keep track of transgenes. Neither one nor two copies of CtBPL were able to
rescue the lethality of homozygous CtBP mutations. Similar results were
obtained when complementation was attempted with the short isoform alone,
suggesting that maybe both forms of the protein are needed to sustain
development. To test this possibility, CtBPL and CtBPs were recombined on
the second chromosome and checked for rescue. However, lethality could not
be rescued by simultaneous expression of both isoforms (See Table A-1).

The developmental profile of endogenous CtBP proteins shows CtBPs
to be present in relatively higher amounts than the long isoform (Mani-Telang
and Arnosti, 2006). However, it was noted that the expression from promoter
driven transgenic lines was the opposite, with higher CtBPL levels relative to
CtBPs (Figure A-2). The relative abundances of the two isoforms may be
critical for proper function. A developmental profile for transgenics expressing
the recombinant protein showed a detectable decrease in levels of CtBPL-
FLAG, mirroring the decrease in endogenous CtBPL levels, suggesting that
the flag-tagged transgene is subject to the same regulation as the native
proteins (Figure A-3). This decrease was more apparent with anti-CtBP
antiserum (second panel) while the anti-M2 FLAG antibody did not reveal a
detectable decrease; this decrease might reflect modified forms of CtBPL that
do not cross-react well with the CtBP antiserum. CtBPL and CtBPs transcripts
also differ in coding sequence and have different 3’UTRs. The 3’UTR is

known to be important for regulating levels of protein expression either

134

temporally and/or spatially (Fujioka et al.. 1999; Kosman and Small, 1997). It
is also conceivable that we failed to include other genomic regions that carry
necessary transcriptional information. for example this construct lacks a large
~3 kb intron located between exons 5 and 6 (Figure A-1) that is conserved in
Drosophilids (data not shown) which might contain regulatory regions that
determine expression levels of the transgene. Finally, there are at least four
CtBP transcripts have been detected by northern analysis using an embryonic
cDNA library that might be functionally relevant to the fly (Poortinga et al.,
1998). Correspondingly. four cDNAs have been recovered (Chapter l-Figure
l-4). Although the differences appear to be minor, expression of only one or
two of the isoforms might not suffice in this setting to provide complete
biological function.

These initial rescue experiments were attempted with cDNAs because
of the ease of expressing different CtBP isoforms and mutants. An alternate
approach will be to attempt rescue with the entire genomic locus contained
within ~15 Kb. Such an endeavor has been successfully demonstrated for
even-skipped, where a 16 Kb genomic region was sufficient and capable for
rescuing lethality of an a null mutant (Fujioka et al., 1999). Alternative forms of
the protein can then be tested in this context. A successful rescue with the full
length CtBPL wbuld set the stage for a more definitive analysis of
dehydrogenase mutants and C-terrninal extensions. This whole animal rescue
approach has been challenging and has not been demonstrated in vertebrate
models, thus these studies can add to our general understanding of CtBP in

development.

135

MATERIALS AND METHODS
Construction of transgenes

For rescue tests, cDNA fragments for CtBPL (1.5Kb) and CtBPs (1.2Kb) were
double flag-tagged at the C-terrnini and were inserted into the unique Kpnl
and Xbal sites of pCaspeR-tubulin vector. This vector additionally provides a
hexahistidine fusion at the N-terminus of CtBP. For rescue using CtBP
endogenous promoter (PC-CtBP), cDNA fragments for both isoforms were
introduced into the Kpnl and Xbal sites in the polylinker of the pUAST vector
(Brand and Perrimon, 1993). The original vector has a basal hsp70 promoter
and 5X UAS sites that were removed by restriction digestion with Sphl and
Kpnl and replaced with an adaptor that containing Sphl, a new Ascl and Kpnl
restriction sites and an optimal Kozak sequence (DA 885 5’-
CACCGGCGCGCCACCAAAATGGGTAC-3’; DA 886 5’-
CCATTI'TGGTGGCGCGCCGGTGCATG-3’). The promoter region of CtBP
(~8054 bp) was amplified from embryo genomic DNA in two halves: the first
4000 bp fragment containing one of the predicted start sites was amplified as
an Sphl-Ascl fragment (CG8841302-8845092) with DA 926 (5’-
GTGCATGCGAAATGG‘I'I'AGCCAGCGTGGTG-3’) and DA 927 (5’-
CGGGCGCGCCTTGAAATCGAGAATCCTGCAATGG-3’) and inserted
immediately upstream of the cDNA. The second fragment (~4054 bp) with the
other putative transcriptional start site was amplified as an Sphl-Sphl
fragment (CG8837323-8841301) with primer sets DA 924 (5’-
CTGCATGCATACCATAAT‘I’C‘ITGCAGT‘ITGCC-B’) and DA 925 (5’-

CGGCATGCAGCTI'TCTGTTTCATGCATATGCAC -3') and introduced into

136

the Sphl site, upstream of the first fragment. Another version of this promoter
driven construct (PH-CtBP) was built to retain the basal hsp70 promoter and
the 5X UAS sites, in order to achieve drive higher levels of transgene using a
combination of the CtBP regulatory region and the inducible pUAS promoter
but was not used in rescue. In this case, the Sphl site was used to insert an
Sphl-AscI-Sphl adaptor (DA 964 5’-
CACCTCAGGCGCGCCACCTGCGCATG-3’; DA 965 5’-
CGCAGGTGGCGCGCCTGAGGTGCATG-3’). The 8Kb promoter sequence

was then cloned in two steps like that described for PC-CtBP.

Western Blot Analysis

Total extracts from adult flies were immunoblotted using 10% SOS-PAGE gels
in a tank transfer system (Bio-Rad Mini Trans-Blot® Cell) and proteins were
transferred to lmmuno-BlotTM PVDF membrane (Bio-Rad). Antibody
incubation was performed in TBST (20mM Tris-HCI. pH 7.5, 120mM NaCl,
0.1% Tween-20) supplemented with 5% non-fat dry milk as a blocking agent.
Mouse monoclonal antibodies include anti-M2 used to detect FLAG epitope-
tagged CtBP (1:10.000), anti-CtBP (polyclonal, 1:10.000) (Struffi and Arnosti,
2005) and B-tubulin (1:6000, Iowa Hybridoma Bank). These were visualized
using HRP-conjugated secondary antibodies (Pierce) and SuperSignal® West

Pico chemiluminiscent substrate (Pierce).

137

CtBP fly stocks

Flies carrying CtBP promoter driven transgenes were introduced into files by
injection procedures described previously (Rubin and Spradling, 1982).
Transgenic lines were checked for protein expression and flies homozygous
for the transgene on the second chromosome were used for rescue. To test if
rescue relied on both transgenes, flies expressing PC-CtBPL and PC-CtBPs
carried over balancers were crossed and allowed to recombine on the second
chromosome. Progeny were screened by performing an anti-M2 western to
detect both FLAG-tagged forms of CtBPL and CtBPs from single flies still
carrying balancer, which is indicative of a recombined chromosome. Flies
carrying a P-element insertion in CtBP (Stock #P-1590, described in
(Poortinga et al., 1998) or a deletion (Stock #1663) were obtained from
Bloomington Stock Center. Mutant chromosomes for CtBP were maintained
over balancer chromosomes i.e. TM3. Sb. All genetic crosses were

maintained at 25 °C on standard cornmeal/molasses medium.

138

REFERENCES

Basler. K. and Struhl. G. (1994). Compartment boundaries and the control of
Drosophila limb pattern by hedgehog protein. Nature 368, 208-14.

Brand, A. H. and Perrimon, N. (1993). Targeted gene expression as a
means of altering cell fates and generating dominant phenotypes.
Development 118, 401-15.

Donaldson, T. D., Noureddine. M. A., Reynolds, P. J., Bradford, W. and
Duronlo, R. J. (2004). Targeted disruption of Drosophila Roc1b reveals
functional differences in the Roc subunit of Cullin-dependent E3 ubiquitin
ligases. Mol Biol Cal/15, 4892-903.

Fujioka, M., EmI-Sarker, Y., Yuslbova, G. L., Goto, T. and Jaynes, J. B.
(1999). Analysis of an even-skipped rescue transgene reveals both composite
and discrete neuronal and early blastoderm enhancers, and multi-stripe
positioning by gap gene repressor gradients. Development 126, 2527-38.

Crumbling. G., Strelets, V. and Consortium, T. F. (2006). FlyBase:
anatomical data, images and queries. Nucleic Acids Research 34 D484-D488;

doi:10.1 093/nar/gkj068. http://flmaseorg/

Kosman, D. and Small, S. (1997). Concentration-dependent patterning by an
ectopic expression domain of the Drosophila gap gene knirps. Development
124. 1343-54.

Manl-Telang, P. and Arnosti, D. N. (2006). Developmental expression and
phylogenetic conservation of alternatively spliced forms of the C-terrninal
Binding Protein corepressor. Development. Genes and Evolution In Press.

Moran, E. and Jimenez, G. (2006). The tailless nuclear receptor acts as a
dedicated repressor in the early Drosophila embryo. Mol Cell Biol 26, 3446-
54.

Poortinga, G., Watanabe, M. and Parkhurst, S. M. (1998). Drosophila CtBP:
a Hairy-interacting protein required for embryonic segmentation and hairy-
mediated transcriptional repression. Embo J 17, 2067-78.

Rubin, G. M. and Spradling, A. C. (1982). Genetic transformation of
Drosophila with transposable element vectors. Science 218, 348-53.

Struffi, P. and Arnosti, D. N. (2005). Functional interaction between the
Drosophila knirps short range transcriptional repressor and RPDB histone
deacetylase. J Biol Chem 280, 40757-65.

Sutrias-Grau, M. and Arnosti, D. N. (2004). CtBP contributes quantitatively
to Knirps repression activity in an NAD binding-dependent manner. Mol Cell
Biol 24, 5953-66.

139

Wagner, C. R., Mahowald, A. P. and Miller, K. G. (2002). One of the two
cytoplasmic actin isoforms in Drosophila is essential. Proc Natl Acad Sci U S
A 99, 8037-42.

140

APPENDIX B

AN RNAi APPROACH TO DEPLETING THE CtBPL ISOFORM

The CtBP gene encodes two major splice variants (CtBPL and CtBPs),
both of which can be detected through all developing stages of the fly
(Chapter I - Figure I-4). CtBP plays vital roles during fly embryogenesis,
emphasized by mutations in the gene which are homozygous lethal. It is not
known whether the existence of multiple isoforms imparts functional specificity
during development. however expression of single isoforms is not sufficient to
rescue lethality. In other systems, there are several examples of distinctions
between CtBP isoforms. Human CtBP1, but not CtBP2, is modified by
sumoylation that results in its nuclear translocation. CtBP2 but not CtBP1
contains an NLS that directs it to the nucleus, suggesting that CtBP1 and
CtBP2 are regulated differently. Mouse knockouts show distinct phenotypes,
lending support to the hypothesis of different isoforms having specific roles.

Initial attempts for a transgene-based rescue experiment do not
provide any clues regarding the functional equivalency of the two isoforms in
Drosophila (Appendix A). Both isoforms when tethered at the promoter have
been documented to display comparable repression activities (Sutrias-Grau
and Arnosti, 2004). In vivo studies assaying biological activity of CtBPs have
been of limited value due to the low expression levels of this protein.
However, we observe a decline in CtBPL protein levels from the larval to adult
stages (Mani-Telang and Arnosti, 2006). Whether this translates to CtBPs
supplying CtBPL like functions or whether CtBPL is no longer required during

these developmental stages and is hence down regulated, remains to be

141

determined. This downregulation might transpire by post-translational
modifications that occur within the C-tenninal tail (Chinnadurai, 2006) (Also
see Chapter 1).

As an independent yet complementary approach, we tried to selectively
deplete levels of CtBPL isoform to gain insights into isoform-specific roles. To
inhibit CtBPL levels. we used RNA interference (RNAi) in combination with the
inducible UAS-Gal4 system in a fly model. RNAi by double stranded RNA was
first described in worms and the mechanism has since been found to be
conserved in vertebrates, invertebrates and plants (Reichhart et al., 2002). In
Drosophila, the UAS-Gal4 system has been successfully applied to study
gain-of-function and more recently, loss-of-function phenotypes; the latter
resulting from gene specific perturbations by RNAi (Duffy, 2002; Roman,
2004). In Drosophila, The C-terrninal region of CtBPL is absent in CtBPs
(includes exon 6+7 and a ~700 bp 3’UTR) and represents a unique sequence
by which this particular isoform can be targeted. A responder line consisting of
tandem inverted repeats to the C-tenninal tail region of CtBPL (See Materials
and Methods) were cloned downstream of the UAS sites in pUAST and were
expressed ubiquitously using a constitutive ActinSC-Gal4 driver line (See
Figure B-1). This cross tests the efficacy of the hairpins to knockdown CtBPL
by inducing the expression of hairpins in a ubiquitous fashion early in
development. If the function of CtBPL is mandatory during these times, then
lethality among progeny is expected. If lethality were to be observed, it would
validate that interference with CtBP function is possible, following which the
next step would be test the specificity of knockdown. RNAi has been

demonstrated to be gene specific, but it is uncertain if the inhibitory effects

142

can spread to related family members (like splice variants). It is possible that
the inverted repeats will also cause a depletion of CtBPs levels. This
possibility will be tested by crossing the responder line to a tissue specific
driver. and testing CtBP levels in this tissue by a western blot. These
experiments are currently in progress.

If specific knockdown of CtBPL levels are observed, I shall analyze
outcomes arising from depletion of this isoform at various stages and in
different tissues during development. Another possibility is that partial
knockdown of CtBPL is observed, in which case it would still be interesting to

analyze phenotypes from temporal and spatial reductions of the long isoform.

143

 

 

 

Actin SC promoter 5X UAS CtBP IRs

   

d.s. hairpins

crap. I—I—I—I~EE
stem mica-{aw

3' UTR

Figure B-1: A transgenic system to trigger RNAi and selectively
eliminate CtBPL in Drosophila

A transgenic driver line expressing Gal4 activator under the control of a
ubiquitous driver (Actin SC) is crossed with a transgenic responder line
expressing tandem inverted repeats against CtBPL, downstream of UAS-
regulatory sites. The responder line was made homozygous for the P-element
carrying the inverted repeats and the driver line is maintained over a balanced
chromosome. The results of CtBPL knockdown will be judged by screening
progeny lacking the balancer-marker resulting from this cross.

 

144

MATERIALS AND METHODS
Construction of transgenes

Tandem inverted repeats were cloned into the pUAST vector that allows
expression of dsRNA hairpins in vivo, facilitating inducible knockdown of
CtBPL (Brand and Perrimon, 1993). Two types of double stranded hairpin
constructs were generated. The first construct was designed to target coding
sequence for exons 6+7 in CtBPL (221 bp+81 bp at the C-termini). Inverted
repeats (le) of this ~300 bp sequence were amplified using CtBPL cDNA as
a Bglll-Notl fragment using primer pairs DA 1219 (5'-
CTGGATAGATCTGCACTGCATCACCGGGCACAC-3’) and DA1220 (5'-
ATAAGATGCGGCCGCCGGCGCCTCCG‘I‘I’GACTC-3’) for the forward
repeat .The reverse repeat was amplified from as a Kpnl-Xhol fragment using
DA 1221 ( 5’-GATGGTACCGCACTGCATCACCGGGCACAC-3’) and DA
1222 (5’-GATCTCGAGCGGCGCCTCCGTTGACTCGG-3’) from CtBPL
cDNA. The second construct was designed to target unique 3’ exon regions
encoding CtBPL ‘tail’, along with ~700 bp of 3’ UTR sequence specific to this
isoform. Inverted repeat sequences (~1.3 Kb) were amplified as a Bglll-Notl
fragment using primer pairs DA 1219 (see above) and DA 1223 (5’-
ATAAGATCAGCGGCCGCG'ITI’CTCGTAATTAAAATTTI’CCAAC-S') for the
forward repeat and primer pairs DA 1261 (5’-
TATTCTAGAGCACTGCATCACCGGGCAC-3') and DA 1262 (5’-
TCAGGTACCGTTTCTCGTAATTAAAATT'I'I’CCAAC-3’) for the reverse
repeat amplified as a Xbal-Kpnl insert. In both constructs. the inverted repeats

were separated from each other using an ~800 bp region containing the

145

second intron of the mub transcript (CG7437) amplified with primers DA 753
(5’-

ATAAGATCAGCGGCCGCCAGGACGTCCAATCAAAGTGGTCAAACCCG-B’)
and DA 754 (5’-
CTGGATCTCGAGGGCTGGAG'I'TCAATAAATATACCATCGCTCTTTGGC-B’)

(Used in Reichhart et al., 2002).

Fly Stocks and Expression System

Transgenic fly lines carrying the UAS-RNAi vectors were obtained using
standard techniques of P-element mediated germline transformation. Multiple
independent lines for each construct were made homozygous and all crosses
were performed at 25°C. Knockdown efficacy was tested by crossing to a
strong ubiquitous actin50-Gal4 driver line (Bloomington Stock Center, Stock #
4414) that is carried over a marked balancer chromosome. The progeny of
this cross will first be screened for any flies that carry non-marked
chromosomes and if such progeny emerge, will be validated for CtBP protein
levels by western blotting with an anti-CtBP antisera (as described in Chapter

ll- Materials and Methods).

146

REFERENCES

Brand, A. H. and Perrimon, N. (1993). Targeted gene expression as a
means of altering cell fates and generating dominant phenotypes.
Development 11 8, 401-15.

Chlnnadural, G. (2006). CtBP family proteins:unique transcriptional
regulators in the nucleus with diverse cytosolic functions: Landes Bioscience.
Duffy, J. B. (2002). GAL4 system in Drosophila: a fly geneticist's Swiss army
knife. Genesis 34, 1-15.

Mani-Telang, P. and Arnosti, D. N. (2006). Developmental expression and
phylogenetic conservation of alternatively spliced forms of the C-terminal
Binding Protein corepressor. Development, Genes and Evolution In Press.

Reichhart. J. M., ngoxygakls, P., Naltza, S., Woerfel, G., Imler, J. L. and
Gubb, D. (2002). Splice-activated UAS hairpin vector gives complete RNAi
knockout of single or double target transcripts in Drosophila melanogaster.
Genesis 34, 160-4.

Roman, G. (2004). The genetics of Drosophila transgenics. Bioessays 26,
1243-53.

Sutrias-Grau, M. and Arnosti, D. N. (2004). CtBP contributes quantitatively

to Knirps repression activity in an NAD binding-dependent manner. Mol Cell
Biol 24, 5953-66.

147

llllllllllliilllllll