Download The Human Skeletal a-Actin Promoter Is Regulated by Thyroid

Vol.
3, 31-42,
January
1992
Cell Growth
The Human
Skeletal
a-Actin
Promoter
Is Regulated
Thyroid
Hormone:
Identification
of a Thyroid
Hormone
Response
Element’
Elaina
S. R. Collie
University
of Queensland,
Biotechnology,
Ritchie
Queensland,
and George
Centre
Research
Biology
St Lucia,
and
4072,
Australia
Introduction
Abstract
Skeletal a-actin mRNA increases in the adult heart
during cardiac hypertrophy
after the imposition
of
hemodynamic
overload/aortic
restriction.
3,3’,5Triiodo-L-thyronine
(13) elicits a cardiac response
similar to the effect of prolonged
exercise and was
recently shown to cause a rapid increase in the amount
of skeletal a-actin mRNA in hearts from normal and
hypophysectomized
animals. We used transient
transfection
analysis to show that 13 induces the
expression
of the native skeletal a-actin promoter
between nucleotide
positions -2000
and +239 linked
to the chloramphenicol
acetyltransferase
reporter
gene
in COS-1 fibroblasts
and myogenic
C2C12 cells. This T3
(10-100
nM)-induced
transcriptional
activation
is
dependent
on the expression
of the thyroid hormone
receptors from transfected
a1 and fl c-erbA
complementary
DNA expression
vectors.
Electrophoretic
mobility shift assays were used to
identify a thyroid hormone
response element (IRE) in
the human skeletal a-actin gene. This IRE is located
between nucleotide
positions -1 73 and -149 with
respect to the start of transcription
at +1 (5’
TGGTCAACGCAGGGGACCCGGGCGG
3’).
Electrophoretic
mobility shift assay experiments
showed that the putative skeletal a-actin TRE and
defined rodent growth hormone TREs (that bind thyroid
hormone receptors in vitro and in vivo) interacted
with
an identical
nuclear factor in vitro in muscle cells that
was developmentally
regulated
during myogenesis.
Transient transfection
analysis utilizing 5’
unidirectional
deletions of the skeletal a-actin
promoter
indicated
that cis-acting sequences between
nucleotide
positions -432 and -153,
which
encompassed
the TRE, were required
for T3/thyroid
hormone receptor-dependent
trans-activation
in vivo.
Furthermore,
we demonstrated
that the skeletal a-actin
TRE is juxtaposed
next to SRF and SpI binding sites, at
its 5’ and 3’ flanks, respectively.
It is also surrounded
by sequences densely populated
by other Spi, SRF, and
OF binding sites. In conclusion,
these results indicate
that T3-induced
increases in a-actin mRNA in animals
Thyroid
hormones
exert profound
effects on the growth,
development,
and homeostasis
of vertebrate
organisms.
These effects are primarily
mediated
by nuclear receptor
proteins
that act to increase the rates of transcription
of
target genes. These receptors
are the cellular
homologues of the v-erbA
protooncogene
and are members
of
the steroid
hormone
receptor
superfamily
of ligand-responsive transcriptional
factors. Cloning of v-erbA-related
cDNA3 sequences
has revealed
the existence,
in mammals, of at least two distinct
but closely
related
genes
that encode
TRs. These genes, which reside on separate
chromosomes,
have been termed
the c-erbA-a
and cerbA-fl
genes. The c-erbA genes are alternatively
spliced,
resulting in heterogeneous
mRNAs and protein products.
This alternative
splicing
results in non-hormone-binding
and tissue-specific
TR forms which
may fine tune the
hormonal
response
(1, 2).
The DNA-binding
domains
of the TR-a and -13 proteins
are highly related
(90-97%),
as are the ligand-binding
domains
(88-98%),
whereas
the extreme
amino termini
are completely
unrelated.
The TRs show distinct
temporal and spatial patterns
of expression.
The TR-a form
exerts important
functions
in central
nervous
system,
kidney, cardiac, and skeletal muscle (1, 2).
The thyroid
hormone
receptor
is thought
to act by
binding
to specific
DNA sequences
in genes responsive
to T3. These sequences
are known
as TREs and are
generally
recognized
by their ability to confer T3 responsiveness to heterologous
promoters
and reporter
genes.
These TREs are purine-rich
elements
that contain
a consensus core T3 receptor-binding
motif G/A GGT/A cA/s
(1,
3, 4), which
may be part of a larger palindromic
sequence,
AGGTCA_.TGACCT
(1-4).
Thyroid
hormones
exert marked
effects
on cardiac
and skeletal
muscle.
Thyroid
hormone
elicits a cardiac
response similar, in many ways, to the effect of prolonged
exercise.
Hyperthyroidism
results in an increased
heart
rate, cardiac output,
and synthesis of a-MHC
mRNA (Vi
protein
isoform).
This in part reflects the increased
metabolic
demands
imposed
by augmented
oxygen
consumption
in peripheral
tissue. In addition,
increases
in 13-
The abbreviations
used are: cDNA, complementary
amphenicol
acetyltransferase;
TRE, thyroid
hormone
TR, thyroid hormone
receptor,
T3, 3,3’,S-triiodo-i-thyronine;
3
binding
Received
1This
work
9/8/91.
was
supported
by
are mediated
by a direct transcriptional
mechanism
that may involve interactions
with ubiquitous
proteins.
E. 0. Muscat2
for Molecular
Laboratories,
& Differentiation
by the
Australian
Health and Medical
Research Council,
Special Projects Grant.
2 To whom
requests for reprints should
Research
and a University
be addressed.
Council,
National
of Queensland
factor;
SRF,
serum
response
factor;
CTF/NF-l,
transcription
factor; EMSA, electrophoretic
mobility
myosin heavy chain; rGH, rodent growth hormone;
modified
Eagle’s medium;
FCS, fetal calf serum;
dioleoyloxy)propyl]-N,N,N-trimethylammonium-methyl
N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic
DNA; CAT, chlorresponse
element,
CBF, CArGCCAAT-binding
shift assay; MHC,
DMEM,
Dulbecco’s
DOTAP,
N-[1-(2,3sulfate;
acid.
HEPES,
31
32
Characterization
of a-Actin
TREs
adrenergic
receptor
number
lead to increased
sensitivity
to catecholamines.
Thyroid
hormone
is a limiting/essential factor during postnatal
skeletal muscle development
and maturation
(5, 6). Hyperthyroidism
induces
the precocious
development
of adult fast contractile
protein
genes (7).
An examination
of muscle-specific
genes has allowed
the identification
of additional
direct targets of thyroid
hormone
action.
These include
several members
of the
myosin heavy chain gene family, atrial natriuretic
factor,
sarcolemmal
calcium
ATPase, and sarcoplasmic
calcium
ATPase (Ref. 8 and references
therein).
Furthermore,
the
effects of T3 on the expression
of a particular
gene are
regulated
in a highly tissue-specific
manner (1, 5, 6).
The actin monomers
assemble
into long cables that
are twisted
into a double
helix to form the thin filament
of the muscle
sarcomere.
In small mammals
such as
rodents,
cardiac
a-actin
predominates
(>95%)
in the
adult heart, although
both isoforms
are present
neonatally. However,
in larger mammals
such as the porcine,
bovine,
and human species, the skeletal isotype of actin
encodes
up to 20-30%
of the net a-actin
in the adult
heart (9-16). The skeletal isotype
is under pathophysiological regulation
(1 7, 18). The amount
of skeletal a-actin
increases
in the adult
heart during
hypertrophy
after
imposition
of a hemodynamic
overload/aortic
restriction
(1 7), in cultured
neonatal
cardiomyocytes
by administra(ion of a1- or /3-adrenergic
agonists (19), and in cultured
neonatal
card iomyocytes
stimulated
by serum antigens
(discussed
in Ref. 18). These results correlated
with the
high levels of skeletal a-actin
found in diseased
human
heart (11).
In ventricular
muscle,
T3 stimulates
the expression
of
the rodent a-MHC
gene (Vi), while inhibiting
the expression of the 13-MHC gene (V3) (20). Very recently,
it has
been demonstrated
that the amount
of skeletal
a-actin
mRNA
in hearts from normal
and hypophysectomized
animals
is also rapidly
induced
under these conditions.
The presence
of a-MHC
in the thyrotoxic
heart is in part
responsible
for the increase
in maximum
velocity
of
shortening
of the unloaded
muscle and for a contraction
that is less efficient
(5, 6). The a-MHC
gene is also
expressed
in the atria, but T3 exerts minimal
influence
on its expression
in this tissue (21). Functional
analysis
carried out in vivo and in vitro has identified
the regulatory sequences
required
for hormonal
regulation
of the
MHC genes. Thyroid-sensitive
elements
have been identified in the rodent
a-MHC
gene between
nucleotide
positions
-599/-576
and -144/-132
(22, 23) and in the
human a-MHC
gene between
nucleotide
positions
-151
and -138
(24). These cis-acting
sequences
have been
shown to interact
with protein
factors in vitro (25). The
latter site in the rodent
species
has been definitively
shown
to interact
with the thyroid
hormone
receptor,
TR-a1 (23). The sequences
that mediate
hormonal
regulation are 5’ GGAGGTGACAGGA
3’ between
nucleotide positions
-144/-i32
and -151/-138
in the rodent
and human
species,
respectively.
These sequences
are
similar
to the many
variant
TREs that are scattered
throughout
the rodent growth
hormone
promoter,
have
recently
been found in the sarcoplasmic
calcium
ATPase
promoter
(8), and are known
to function
in vivo and in
vitro (3, 4). Winegrad
et a!. (18), who showed
that thyroid
hormone
induced
the accumulation
of skeletal
a-actin
mRNA, did not find the TR-binding
sites in the a-actin
gene and could
and/or
not discriminate
transcriptional/posttranscri
between
direct/indirect
ptional
mechanisms.
This was probably
attributable
to the fact that they were
scanning
the sequence
for consensus
dyad repeat sites,
using the new TR “half-site”
core-binding
motif described
by Glass and Holloway
(1) and Norman
et a!. (3). We
have identified
some probable
candidate
TRE sequences
for the interaction
ofthe receptor
with this promoter
that
are accommodated
by the core consensus
sequence.
This sequence
contains
a purmne-rich
element,
G/
GGT/A
cA/s
that is highly conserved.
These putative
TREs may
account
for the induction
of skeletal
a-actin
during
T3
treatment,
and they are located
between
nucleotide
positions
-273/-249
and -173/-149.
In the current
study, we showed
that T3 induced
the
transcription
of the native skeletal a-actin promoter
between
nucleotide
positions
-2000
and +239 which
is
linked to the CAT reporter
gene in COS-i fibroblasts
and
myogenic
C2C12
cells. This T3 (10-50
nM) -induced
transcriptional
activation
is dependent
on the expression
of the TR a1 and fl isoforms.
EMSA assays were used to
identify
a putative
TRE in the human
skeletal
a-actin
gene. This TRE is located
between
nucleotide
positions
-1 73 and -149 with respect to the start of transcription
at +1 (5’ TGGTCMCGCAGGGGACCCGGGCGG
3’).
This
sequence
fits
the
six-base
pair
core-binding
motif
for the TR G/ GGT/
cA/
and is similar to TREs defined
in the rodent
growth
hormone
gene (3). Transfection
experiments
indicated
that this TRE was functional
in
vivo. The putative
skeletal a-actin TRE and defined
rodent
growth
hormone
TREs (which
bind TRs in vitro and in
vivo) interacted
with an identical
nuclear
factor in vitro
in muscle cells that was developmentally
regulated
during myogenesis.
This TRE is flanked on its boundaries
by
the transcription
factors, SpI (26-28),
CTF (29-31),
and
SRF (32-34).
Results
Transcriptional
Activation
of the a-Actin Promoter
by T3
and Thyroid Hormone
Receptors in Nonmuscle
and Myogenic Cells
Transcriptional
Activation
of the Skeletal
a-Actin
Promoter by the a and 13 c-erbA Products in COS-1 Fibroblasts. Transfection
studies
were carried
out to investigate whether
the Cis-acting
sequence
between
nucleotides -2000
and +239
in the native
skeletal
a-actin
promoter
were responsive
to T3 via the thyroid
hormone
receptor
in vivo. For this purpose,
we used the plasmid
pHSA2000CAT,
which
contains
the native
skeletal
aactin promoter
linked to the CAT gene.
COS-1 cells, which are a fibroblast
cell line deficient
in TRs, were grown
in thyroid
hormone-deficient
medium for 24 h prior to transfection.
In transient
expression
assays, equal amounts
of either pUC18
or a mixture
of
the rodent
c-erbA-a
and c-erbA-f3
cDNAs
cloned
into
CDM8 (a eukaryotic
expression
vector) (35) were cotransfected into this cell line with the skeletal a-actin promoter
pHSA2000CAT.
The protein
products
of the c-erbA-a
and c-erbA-13
cDNAs
are the a and 13 isoforms
of the
thyroid
hormone
receptor.
The cells transiently
expressing the a and 13TRs were then grown in the presence
T3
(10 nM) for 48-72
h from the day after transfection.
The
cells that were transfected
with pUCi8
were grown
in
T3-deficient
medium
over the same time period.
The
Cell
2
1
3
4
5
6
2
1
3
4
Growth
5
& Differentiation
6
33
7
...
..
......
1.
pCAT+pUC18
2. pCAT
+
c-erbA.a
+
c-erbA-
3.
pHSA2000CAT
+
pUCI
4.
pHSA2000CAT
+
c-erbA-cx
5.
pHSA2000CAT
+
pUCI8
6.
pHSA2000CAT
+
c-erbA-ct
+
T3
....#{149}S#{149}
8
+
c-erbA-
+
T3
+
c-erbA-
+
T3
Fig. 1. The skeletal
a-actin
promoter
is trans-activated
by c-erbA
and T3
in COS-1
cells.
CAT assays
demonstrating
the effect
of T3/TRs
on the
human
skeletal
a-actin
promoter
sequences
between
-2000
and +239
in COS-1
cells,
in the presence
of c-erbA
TRs and T3 (10 nM).
The
transfections
and CAT assays were
performed
as described
in “Materials
and Methods.”
CAT activity
in the
presence
from
pHSA2000CAT
of T3 and
4 and 6), as opposed
TRs,
was increased
respectively
early
linked
c-erbA-cx
+
c-erbA-3
5.
pHSA2000CAT
+
pUCI8
1 , Lanes
6.
pHSA2000CAT
+
c-erbA-a
+
T3
with
7.
pHSA2000CAT
+
c-erbA-3
+
T3
to CAT)
in the
and a 3-4-
T3-dependent
trans-activation
they are cotransfected
with
of
the
rodent
a and 3 forms
of the TR. The T3/TR-independent
expression
of pCAT demonstrates
that the T3-dependent
induction
mediated
by the a-actin
TRE is sequence
specific.
The a- and i9-c-erbA Genes Mediate
Similar Functions
in the T3/TR Transfection
Assay. The requirement
for cerbA gene products
activation
indicated
thyroid
hormone
+
pCAT
These transient
cotransfection
experiments
have demonstrated
that the sequences
between
nucleotide
positions -2000
and +239
in the skeletal
a-actin
gene are
capable
of mediating
CAT expression
when
pUC18
4.
presence
ofT3 (Fig. 1, Lanes 1 and 2). The pCAT
plasmid
expresses
at high levels
in COS-1
cells, because
of very
efficient
replication
from
the SV4O origin
of replication
posttransfection.
These
transfections
were
performed
in
triplicate
with different
plasmid
preparations,
fold T3/TR induction
was seen in each case.
+
pCAT+pUC18
to that in cells not transfected
SV4O) promoter
2. pEMSV-CAT
3-4-fold
(Fig.
pEMSV-CAT
3.
TRs and grown
in the absence
of T3 (Fig. 1, Lanes 3 and
5). This level of induction
with T3 and TRs is similar
to
that observed
by Zilz et a!. (35) in their
work
on the
hepatic
S14 gene.
In addition,
this transcriptional
induc(ion was not observed
when c-erbA-cx
and c-erbA-f3
were
cotransfected
into COS-1
cells with the pCAT vector
(an
enhancerless
1.
during
T3-dependent
transcriptional
that the process
is mediated
through
receptors.
We investigated
the func-
+
c-erbA-a
+
c-erbA-3
+
T3
+
T3
Fig. 2.
The skeletal
cx-actin
promoter
is trans-activated
by T3 and either
c-erbA-a
or -/3 in COS-1
cells. CAT assays demonstrating
the functional
activity
of the c-erbA-a
or -(. TRs with
respect
to trans-activation
of the
human
skeletal
actin
promoter
in the presence
of T, (10 nM) in COS-1
cells. From the pEMSV-CAT,
pCAT,
and pHSA2000CAT
transfections
in
this experiment,
we used 5, 5, and 25 M1 of extract
for the CAT assays,
respectively,
because
the skeletal
a-actin
promoter
did
not express
efficiently
in fibroblasts
relative
to pEMSV-CAT.
tional
differences
between
the a and 3 receptors
with
respect
to the trans-activation
of the cs-actin
TRE in the
presence
of T3. The a-actin
promoter
pHSA2000CAT
construct
was cotransfected
into COS-1 cells with either
pUC18,
the a isoform,
or the /3 isoform
of the thyroid
hormone
receptor.
The CAT activity
was approximately
3-fold
and 5-fold
higher
in the presence
of a-c-erbA
or
13-c-erbA,
respectively,
after T3 treatment
(Fig. 2, Lanes 57). As demonstrated
by other investigators,
no significant
difference
in functional
activity
was found
between
the
a- and
13-c-erbA
products
with
respect
to the extent
of
trans-activation
of the skeletal
a-actin
promoter
after 1,
treatment.
This set of transfections
also demonstrated
that a plasmid
ney
sarcoma
CAT
gene
presence
containing
virus
long
(pEMSV-CAT)
of T3 and
a strong
viral
terminal
(36)
cotransfected
promoter
repeat)
was
not
TRs (Fig.
linked
induced
2, Lanes
(Mobto
the
in the
1 and
34
Characterization
1
of (,-Actin
2
TREs
3
4
5
6
.
(A)
.
HUMAN
SKELETAL
ALPHA
ACTIN
TREs
:
-23
-24
5’
GGGCAACTGGGTCGGGTCAGGAGOG
3’
:
-:3
-149
5’
TGGTCAACGCAGGGGACCCGGGCGG
3’
(B)
RODENT
HORMONE
GROWTH
5’
:T;;;
-1
r
-16j/-14E)
rH
-7/-46
rH
-27/-6
S
TREs
AAGGTAAGATCAGGACTGACCG
3’
CGCAGGAGAGCAGT1GGGACCG
3’
w
1
I
fig.
‘
AAAAAGGCAGGAGCCTTGGGTC
3’
5’
3’
AAAAAGGGCATGCAAGGAC9b,
4.
Thyroid
hormone
response
elements.
A, the putative
TREs in the
skeletal
a-actin
promoter.
Sequences
of one strand
of doublestranded
oligonucleotide
probes
that
were
synthesized
to represent
putative
thyroid
hormone
receptor
binding
sites in the skeletal
a-actin
promoter.
The direction
and location
of the four putative
HSA TREs are
indicated
by arrows.
These
sequences
match
the
six-base
pair core
sequence-binding
motif
for the thyroid
hormone
receptor:
G1 GGT/A CA,
human
1.
pCAT+pUC18
2. pCAT
+
c-erbA-cc
+
c-erbA-
3.
pHSA2000CAT
+
pUC18
4.
pHSA2000CAT
+
c-erbA-a
5.
pHSA2000CAT
+
pUC18
6.
pHSA2000CAT+
c-erbA-a
+
T3
C. The
+
+
c-erbA-
c-erbA-
#{247}
T3
+
2). The T3/TR-independent
expression
of the Moloney
sarcoma
viral promoter
and pCAT
demonstrates
that the
T3/TR-dependent
trans-activation
mediated
by the aactin promoter
is sequence
specific.
TranscriptionalActivation
of the Skeletal a-Actin Promoter by the a- and 13-c-erb A Products
in Myogenic
C2C12 Cells. Similar experiments
to those outlined
above
carried
line. Initially,
out
either
in C2C12
cells,
a mouse
pUC1 8 or the c-erbA-a
myogenic
and
to the transcripgrowth
hormone
DNA sequences
have been
shown
binding
of thyroid
nucleotide
positions
at +1.
to confer
T3 responsiveness
to the rGH promoter
via
hormone
receptors
to these
cis-acting
regions.
The
are indicated
with respect
to the start of transcription
were
in C2C12
T3
Fig. 3.
The skeletal
a-actin
promoter
is trans-activated
by c-erbA
and T3
in C2C12
ells. CAT assays
demonstrating
the effect
of T3/TRs
on the
human
skeletal
actin
Promoter
sequences
between
-2000
and +239
in
C2C12
cells, in the presence
of c-erbA
TRs and T, (10 nM(.
were
nucleotide
positions
are indicated
with
respect
tional
start site at +1 . B, the defined
TREs in the rodent
promoter
(rGH).
Boxed region,
the four rGH TREs. These
cell
-13expres-
sion vectors
were
cotransfected
into this cell line with
the skeletal
a-actin
promoter.
The cells were then grown
in thyroid
hormone-deficient
medium
in the absence
or
presence
(10 nM) of T3, respectively.
In cells transfected
with the TRs and grown
in the presence
of T3, a 2-4-fold
increase
in
CAT
activity
was
observed
with
pHSA2000CAT
(Fig. 3, Lanes 4 and 6) when
compared
to cells transfected
with pUC1 8 and grown
in T3-deficient
medium
(Fig. 3, Lanes 3 and 5). In addition,
no induction
was observed
in the presence
of T3 in cells cotransfected
with
the receptors
and the pCAT
vector
DNA
(Fig. 3,
Lanes 1 and 2). These
transfection
studies
were
carried
out in duplicate.
The pCAT plasmid
expressed
at the
expected
low levels in these cells, because
the SV4O
origin of replication
does not function
in rodent cells.
Additional
experiments
were then performed
to determine
whether
exogenous
thyroid
hormone
receptors
required
the skeletal
a-actin
cells
to elicit
promoter.
a T3 response
For this
purpose,
from
C2C12
cells transfected
with the skeletal
a-actin
pHSA2000CAT
plasmid
were grown
either
in the presence
(10-50
nM)
or absence
of T3. Expectedly,
little
or irreproducible
increases
in CAT activity
were seen in C2 myogenic
cells
transfected
with the skeletal
a-actin
promoter
(data not
shown).
This has been observed
in other
cell lines derived from
tissues,
including
liver and muscle
that normally express
large amounts
of endogenous
T3 receptors
and has also been seen in primary hepatocytes
(35), and
cardiocytes
(8). One possible explanation
for these findings,
put
forward
by ZiIz
the endogenous
from
fied
et a!. (35)
T3 receptors
among
others,
are limiting
the transfected
constructs.
This
by our data, which
demonstrated
is that
or sequestered
hypothesis
that the
is yenskeletal
a-actin sequences
and +239 mediate
between
nucleotide
positions
a response
to T3 dependent
-2000
on co-
transfected
hormone
cells.
thyroid
Characterization
sponse Elements
Identification
core
consensus
CI was
putative
promoter
receptors
in C2C12
of the Putative Thyroid
Hormone
Rein the Human Skeletal a-Actin Gene
of Putative Skeletal a-Actin
TREs. The
T3 receptor-binding
motif
used to scan the human
skeletal
TREs (3). Putative
TREs were
region
of this gene
tions -273/-249
and -173/-149,
transcriptional
start site at +1
regions
are indicated
as follows:
GGGCAACTGGGTCGGGTCAGGAGG
between
.
[G/A
GGT/A
a-actin
identified
nucleotide
cA/
gene for
in the
posi-
with
respect
to the
The sequences
of these
(a) HSA -273/-249
5’
3’ and (b) HSA
Cell
A
B
CArG
Probe
.J.
I-
Q_o
sayed as described
and Methods.”
-
1 73/-i
5‘
49
I-F-
TGGTCAACGCAGGGGACCGGGCGG
of possible
Oligonuclewere synputative
cis-
-173/-149.
Nuclear
Extracts from Differentiated
Cells Form a Complex
with a Putative
regulated
of differentiation
to assay
stages
regulation
extracts
myoblasts,
days after
tracts were
ically
and
confluent
myoblasts,
and myotubes
(3 and 4
mitogen
withdrawal,
respectively).
These
exassayed
for levels of CBF, which
is biochemimmunologically
indistinguishable
from
SRF
of Oct-i
constitutively
(data
not
expressed
proteins
from
shown)
the
a
in different
devel-
and
are
used
extracts
of expression
(Fig.
58).
of a nuclear
This
to
demonstrated
factor
known
in these
of these
experiments,
cells and verified
the
extracts.
the
HSA
-273/-249
and
a-actin
teins
are
sequences
binding
The
protein
sequence
-173/-149
factor.
This
differentiated
demonstrates
to these
that
regions
that
of the
interacted
with
is a constitutively
the protein
different
skeletal
the
expressed
that
bound
proa-actin
HSA
factor
to the HSA
sequence
is a developmentally
regulated
protein
factor
was expressed
only in highly
muscle
cells (myotubes,
day 4) and could
be detected
in myogenic
cells
at an earlier
stage
of
development
(Fig.
The Skeletal a-Actin
TRE/Protein
Complex Is Specifically Competed
by Defined
Rodent
Growth
Hormone
TREs. Previously
defined
thyroid
hormone
response
ele5D).
(38, 39); these
factors
EMSA
two
not
in muscle.
proliferating
standardize
the amount
of nuclear
extracts
used in the
experiments
(Fig. 5A). The amount of MEF-2 protein (40),
which is a differentially
expressed
factor, was also measnuclear
-
HSA -1 73/-149
oligonucleotides
specifically
interacted
with a factor(s)
in mouse
myogenic
cells. The differential
regulation
of these factors
is depicted
in Fig. 5, C and D.
The expression
patterns
of the factors
that bound
the
-273/-249
opmental
Specifically,
of TRE-bound
were
isolated
In the
stage
(Fig. 5C). In contrast,
mobility
shift assays (37) were
used to determine
the putative
skeletal
a-actin
TREs interacted
with
nuclear
factor in vitro. Nuclear
extracts
were
prepared
from
mouse
myogenic
C2C12
myoblasts
and myotubes
in the
*
i
promoter.
Skeletal Muscle
TRE. Electropho-
retic
that
pattern
.
I
be differentially
thesized
to enable characterization
of these
acting motifs.
In the following
study, these oligonucleotides will be referred
to as HSA -273/-249
and HSA
ured
.#{149}.#{149}.
in “Materials
Each of these regions
contained
a number
TREs, which
have been outlined
in Fig. 4A.
otides
corresponding
to the above
sequences
the
0
0
developmental
(32, 34), and
49
m
mm(Y)
1.1
3’.
are both
HSA -173/-i
1
mmc’)’4
F-I--
.
Fig. 5.
Detection
of a factor that
interacts
with the skeletal
a-actin
TRE that
is differentially
regulated
during
myogenesis.
Electrophoretic
mobility
shift analysis
of C2C12
myogenesis
using the
following
DNA
probes:
(A)
CArG; (B) MEF-2;
(C) HSA -273/
-249;
and (0) HSA -173/-149.
Proliferating
myoblasts
)PMB(,
confluent
myoblasts
(CMB),
and
myotubes
after 3 and 4 days of
mitogen
withdrawal
(MT3
and
MT4,
respectively).
The
DNA
probes
were
incubated
with
510 g of nuclear
extract
and as-
35
D
HSA -273/-249
II
cv)
& Differentiation
C
MEF-2
I
Developmental
Stage
Growth
to
ments
from
the
rat
[rGH TREs -190/-167,
-6 (Fig. 48)], which
purified
TRs
(3) and
growth
hormone
promoter
region
-168/-146,
-701-46,
and -27/
have been shown to interact
with
function
in vivo
(41,
42),
were
used
in EMSA experiments
with the myogenic
C2C12
cell
nuclear
extracts.
These defined
TREs interacted
with a
specific factor that was differentially
regulated
in mouse
myogenic
cells and more abundantly
expressed
in differentiated
myotubes
(data
not
shown).
This
finding
was
36
Characterization
of o-Actin
TREs
Probe
HSA
-1 731-1 49
Competitor
HSA
Molar Excess
C 40 80 160 C 40 80 1 60 C
rGH -1681-146
-173/.149
I
rGH
#{149}
#{149}
#{149}
1
#{149}
1 #{149}
#{149}
.
.
1
-701-46
l
,
2
40 80 160
3
4
5
.
pUCI8
pHSA432C:T
+
pUC18
pHSA432CAT
+
c-erbA-a
pHSA432CAT
+
c-erbA.cc
pHSA432C;T
+
c-erbA-)3
pHSA432CAT
S
#{149}
K
9
10
pHSA153CAT
+
pHSAIS3CAT
+
pUCI8
pHSA153CAT
+
c-erbA-cz
pHSAI53CAT
+
c.erbA-a
+
c-erbA-3
11
#{149}
pHSA1S3CAT
12
#{149}
pHSAI53CAT
The skeletal
o-actin
TRE is functional
in vivo. CAT
the effect
of T3 (100 nM( on two deleted
human
actin
promoters,
pHSA432CAT
and pHSA153CAT,
in COS-1
presence
of c-erbA-o
and -.
The transfections
and CAT
performed
as described
in “Materials
and Methods.”
quences
Probe
LISA - 1 7 1/- 1 49 in C2 myotube
nuclear
of each
DNA
( ompetitor
is indicated.
C. the
the absence
of any unlabeled
competitor.
analogous
to the EMSA
-173/-149
oligonucleotide
We tested
the ability
hormone
TREs, rGH
extra
ontrot
results
obtained
(Fig. SD).
of the defined
-168/-146
and rGH
ts. The molar
excess
binding
reaction
in
with
the
rodent
HSA
growth
-70/-46
(Fig.
48), to compete
specifically
for binding
to HSA -273/
-249
and -1731-149.
The complex
formed
by HSA
-273/-249
was not competed
by the rGH TREs (160fold) or HSA -173/-149
(data not shown).
In contrast,
the HSA -173/-149
binding
site bound
to a nuclear
factor
that was specifically
competed
by both rGH TREs
at a 40-fold
molar excess
(Fig. 6). This DNA-protein
interaction
was also competed
by a 20-fold
molar excess
(data not shown).
This competition
by sequences
that
interacted
with purified
TR provided
evidence
vitro interaction
of a IRE-associated
factor with
ified
DNA
sequence
in the skeletal
a-actin
for the in
the specpromoter
between
nucleotide
positions
T3
+
T3
+
+
c-erbA-3
+
T3
pUCI8
T3
+
+
T3
T3
assays dem-
Fig. 7.
onstrating
[‘B. .
The skeletal
a-actin
TRE-associated
factor
is specifically
competed
by the defined
rodent
growth
hormone
TREs. The effect
of competition,
by various
DNA
fragments,
on the complex
formed
in vitro
with
the
+
c-erbA.3
+
#{149}
1
#{149}
#{149}
#{149}
#{149}
7
,
+
#{149}
#{149}
6
I
pHSA432C:\T
skeletal
acells, in the
assays
were
-432/+239
and
-153/+239,
respectively
(43). These regions were linked
to the gene coding for CAT. The plasmids
pHSA432CAT
and pHSA1S3CAT
were cotransfected
into COS-i
cells
with equal
amounts
of either
pUC18
or c-erbA-a
and cerbA-(3 cDNAs
cloned
into the CDM8
expression
vector.
After
iO
M T3 treatment,
only pHSA432CAT
exhibited
a T3-dependent
ence of either
trans-activation
(3.2-4.4-fold)
in the pres-
c-erbA-a
or c-erbA-13 (Fig. 7, Lanes 3 and
5 versus Lanes 4 and 6). This construct
contained
the TRE
sequences
defined
in vitro between
nucleotide
positions
-173
and -149.
However,
the plasmid
pHSA153CAT,
which does not contain the TRE defined
in vitro, was not
inducible
(Fig. 7,
expenibetween
nucleotide positions
-432 and -153 were required
for the T3/
TR-dependent
trans-activation.
This indicated
that the
IRE (-173/-i49)
was functional
in vivo. The presence
of
cotransfected
TR-a or -I in the absence
of ligand led to
a decrease
in pHSA432CAT
(Fig. 7, Lanes 1 and 2 versus
Lanes 3 and 5), indicating
that TRs display
a repressorlike function
in the absence of hormone,
which has been
noted in other systems (8).
Lanes
ments
by T3 in the presence
of TR-a
9 and 11 versus Lanes 10 and
showed
that the cis sequences
or TR-fl
12). These
region.
The cis-acting
Region
between
Nucleotide
Positions
-432 and -153 in the Skeletal a-Actin Gene Contains a
Functional
TRE
We then examined
the ability of this putative
TRE to
mediate
a T,/TR-dependent
transactivation
in vivo. We
used
two
5’
unidirectional
deletion
mutants,
pHSA432CAT
and
pHSA1S3CAT,
which
contained
se-
The Skeletal a-Actin
1/CTF Binding Sites
TRE Is Flanked
by SpI, SRF, and NF-
We decided
to determine
whether
other transcription
factors interacted
with the skeletal a-actin
promoter,
5’
and 3’ of the putative
TRE between
nucleotide
positions
-173 and -149.
We
have
interference
previously
analysis,
shown
and DNase
by
EMSA,
1 studies
methylation
(31) that CBF
Cell Growth
Fig.
8.
Schematic
37
#{149}4-SS
representa-
tion of the proteins
interacting
with
the skeletal
a-actin
promoter, showing
the relative
position of the TRE versus characterized SRF, SpI, and NF-1/CTF
-700
-600
-500
-300
-400
-200
-100
#{149}o
sites.
Spi
TRE
SRF
gene that
(A)
SpI
Consensus
5’G
G
G
G
TA
C
A
(B)
HSA
& Differentiation
1-724
C G G
AATAAT
Alpha
67 9 I - 6 5 5 TGTGCGTGGGGGAAGGGGTCGACG
6 3 5 I - 6 1 3 CAGGGAGCTCGGGGGTGGGAAGAGA
HSA
-379/-355
HSA
-2
HSA
-
HSA
-129/-109
bers represent
3’
-
SpI
1 4 1 GGGGACCCGGGCGGGGGCCCC
GGCCGAGGGAGGGGGCTCGAG
SpI
lettering.
binding
binding
sites
a-actin
gene. A, the SpI
is displayed
in bold
Nucleotides
appearing
at lower
frequencies
in desites are displayed
in plain
text. B, sequences
of one
described
in the
skeletal
by Evans et al. (48)
strand of double-stranded
oligonucleotide
probes that were synthesized
to represent
putative
SpI binding
sites in the skeletal a-actin
promoter.
The nucleotides
in these probes that were accommodated
by the consensus are underlined.
interacts
with the skeletal a-actin
promoter
at positions
-98/-89,
-1 79/-i
70, and -225/-2i6.
These regulatory
regions contain
the CCArGG
(CC A-rich GG) sequence
motif,
CC (A+T-nich)6GG
(32). CBF has been shown to
be necessary
for efficient
myogenic
specific
expression
(32, 33, 43-47),
and it is biochemically
and immunologically mndistinguisdhable
from SRF (33, 34). Interestingly,
the SRF site at -i79/-i70
is adjacent
and 5’ to the
skeletal a-actin TRE (Fig. 8).
We decided
to scan the promoter
sequence
for putative SpI consensus
sequence
motifs.
The trans-acting
factor SpI has been shown to interact with the degenerate
binding
sites, conforming
to the following
consensus
as
described
by Evans et a!. (48) (Fig. 9A) (numbers
are
percentage
of each base at that position):
GM G,9 G1
121
occurrences
Sites
ACATGGTTGGGGAGGCCTTTGGGAC
sequence
sequence
(num-
out of seven):
G7 G7 G7 Cs A5 G7 G7 G7 G5 G4
I -27 3 CTCGCCCCACCCCATCCCCTCCGGC
Putative
uppercase
fined
SpI
C
GCCCGCGCCTCCTCCCCGGCCGTCCGCCCTCGCCTCCCCCCGCACGT
-
consensus
G
Actin
-
16 1I
consensus
TA
HSA
Fig. 9.
G
CCG
HSA
91
have the following
Sequence
T
Skeletal
-771
CTF
A11
C1
A4
G1
C82 C96 C,9 G,, G,,, C58
A,4 A,,
T
A21 A1,
14
Cli
C7 G14
17
Gustafson
and
ized seven
SpI binding
Kedes
(49) have defined
sites in the human
121
A7
and charactercardiac a-actin
Cl
C1
C C2
11
T
TI
1
In the case of the human
skeletal
a-actin
gene, we
found
at least seven putative
SpI binding
sites in the
promoter
between
nucleotide
positions
-77i
and -i08
(Fig. 98). We synthesized
oligonucleotide
probes
for
each of these sites and incubated
them with purified
HeLa cell SpI protein
(26-28)
(kindly
provided
by J.
Kadonaga,
S. Jackson, and R. Tjian). All seven oligonucleotides, at nucleotide
positions
-77i/-724,
-679/-655,
-635/-6i3,
-379/-355,
-297/-273,
-i61/-i4i,
and
-129/-i08,
interacted
with SpI in gel retardation
assays
but with variable
degrees
of affinity
(Fig. iO). Interestingly, an SpI site was found
to be located
3’ of the
skeletal a-actin TRE, between
nucleotide
positions
-173
and -149
(Fig. 8). In addition,
these seven oligonucleotides interacted
with SpI present
in crude nuclear
extracts, and these interactions
were competed
with cardiac a-actin
SpI binding
sites (data not shown).
The SpI
binding
activity
in these extracts
was reduced
during
myogenic
differentiation
(data not shown), as previously
reported
by Gustafson
and Kedes (49).
Additional
scrutiny
of the nucleotide
sequences
of the
skeletal a-actin
promoter
identified
a putative
NF-i/CTF
binding
site between
the second and third CBF binding
sites (Fig. 8). This
sequence
at HSA
-i99/-i86,
TGGGACCGGGCCAA,
is compatible
with the consensus
sequence
proposed
by Jones
et a!. (30):
TGG(A/
c)NNNNNGCCAA.
In addition,
the actin
gene
site
matches,
at 1 i of i4 bases (underlined),
a human
13globmn high affinity CTF site (29): TGGTATGGGGCCAA.
An oligonucleotide
probe,
HSA -2i0/-i78,
was shown
to interact with a factor present in crude nuclear extract
and was specifically
competed
by itself and by a proven
CTF/NF-i
site from ras [-319/-295
(29)] (Fig. hA).
The
identity ofthe oligonucleotide
as a CTF/NF-i
binding site
was convincingly
demonstrated
by its interaction
with
purified
CTF/NF-i
from HeLa cells (kindly
provided
by
S. Jackson and R. Tjian), as shown in Fig. i 1 B.
In summary,
we have demonstrated
that the skeletal
a-actin TRE is juxtaposed
to SRF and SpI binding sites, at
its 5’ and 3’ flanks, respectively.
Furthermore,
it is sunrounded
by
CTF binding
a region
carpeted
sites (summarized
by
other
in Fig. 8).
SpI,
SRF,
and
38
(harac
t(’ri/.itIOri
ot
,ir\c
Cr)
Lt)
‘-
(9
(0
it)
N-
(Y)
(9
(0
I
I ig.
tin TREs
IC)
Lt)
C’)
C’)
NC’J
0)
NC’)
NcY)
(‘4
,
,
-
“zt
-
0
‘-
-
(0
0)
1
T-
(\j
NN-
I
(‘4
<
<
<
<
<
(I)
I
(I)
I
(I)
I
(I)
I
(I)
(I)
I
I
1 (1.
SpI interacts
with
niultiple
sites in the skeletal
o-actin
gene.
Spl Protein
troll)
t-lela
cells int(’racts
svith a number
ol Grich
sites in the skeletal
o-,u tin I)r111tr.
The (iligonucleoti(le
pU)bes denoted
in Fig. 9B sver(’ inc ubated
with
puriti(’(l
Spt, and the l)ound
complexes
ssc’r(’ ObstrV(’(l
l)V gel ‘k’ trophoresis
nlol)ility
shift analysis
as descriI’d
in ‘Materials
,uil
Methods,”
Purltiv(l
Discussion
In the
human
present
skeletal
investigation,
we have
shown
that the
a-actin
promoter
is capable
of mediating
a T1/TR-dependent
trans-activation
of CAT
expression.
Our data indicate
that the Trinfluenced
induction
of aactin
mRNA
in animals
(18) is a direct
transcriptional
effect.
The putative
skeletal
a-actin
TRE [5’ TGGTCAACGCAGGGGACCCGGGCGG
3’ (-1 73/- 1 49)] contains sequences
that fit the consensus
motif,
C1
GG /A/
(, A/(,
(3) and show
homology
to TREs defined
by meth-
ylation
interference
analysis
in the growth
hormone
gene
(GGGACC
and GGGACG)
(3). T3/TR-dependent
transactivation
was shown
to require
sequences
between
nucleotide
positions
-432
and -153
that encompassed
the consensus
sequences.
The skeletal
a-actin
TRE interacts with a developmentally
regulated
protein
from muscle cells. This DNA/protein
complex
is specifically
com-
peted
by defined
rGH TREs that have been
shown
to
interact
with
purified
thyroid
hormone
receptors
and
confer
T1 regulation
(3, 41, 42).
The ability
of T to activate
the skeletal
a-actin
promoter
was strictly
dependent
on the expression
of either
cotransfected
aor fl-c-erbA
genes in fibroblastic
and
muscle
cells. These
results
were expected
in the COS-1
cells because
these
cells have been
shown
to contain
insufficient
levels of TRs for trans-activation
of genes after
15 treatment.
In contrast,
muscle
is one of the tissues
targeted
by thyroid
hormone,
and hence
it was unexpected
that C2C12
cells should
require
cotransfected
aor [3-c-erbfl
in order to mediate
a significant
T response.
However,
a number
of explanations
may account
for
these observations.
Zilz et a!. (35), among
others
(8), have
suggested
that the level of endogenous
receptors
is insufficient
to bind to a large number
of transfected
DNA
molecules
introduced
into cells. Alternatively,
the endogenous
receptor
may be sequestered
in a form not freely
available
to new sites in the time frame of the experiment.
For example,
the T1 receptor
may be bound
to another
protein
or to a chromosomal
site. Consistent
with this
latter possibility,
it is known
that the receptor
is localized
as a chromosomably
bound
protein
in either the presence
or absence
of the ligand.
Thus, the TR might
actually
be
bound
to its cognate
TRE sequence
in the absence
of
hormone
and not readily
reequibibrated
with
the transfected
DNA
molecules.
These
data are consistent
with
other
tissue
culture
model
systems
(including
primary
hepatocytes
and cardiocytes)
(8, 35) that express
endogenous
TRs, where
it has been observed
that transfected
IRs are required
to mediate
T3 regulation
of a transfected
gene.
Furthermore,
other
immortalized
myogenic
cell
lines (e.g. L6E69) have failed to support
the induction
of
transcription
after T3 treatment,
although
the identical
cis-acting
sequence
has responded
to hormonal
treatment in primary
cardiocytes
(23). However,
it should
also
be noted
that the sarcoplasmic
calcium
ATPase
TRE only
supports
T1-dependent
trans-activation
in the presence
of transfected
TRs in primary
cardiocytes
(8). These
myogenic
cells may have adapted
to continuous
cell culture
in vitro in media
containing
TH by expressing
the nonTi-binding
forms
ofthe
a-TRs (a2).
Using the cotransfection
assay, both the a and 13forms
of the thyroid
hormone
receptor
were found
to activate
the skeletal
a-actin
IRE in COS-1
and C2C12
cells. a-cerbA
mRNA
is much
more
abundant
in skeletal
and
cardiac
muscle
than the 13-TR, and hence
the a-TR
is
more likely to mediate
the response
of the endogenous
skeletal
a-actin
gene
in these
muscles.
However,
no
functional
differences
were seen between
the a- and 13IRs with respect
to trans-activation
of the skeletal
a-actin
promoter.
Our
data support
1990 (18), who found
in the hearts of normal
the observations
of Winegrad
that skeletal
a-actin
and hypophysectomized
2-24 h ofT3 treatment.
This induction
during
a time of preferential
a-MHC
bates the expression
of this isoform
studies
did not distinguish
a direct
thyroid
hormone
on the gene or an
ondary
to changes
in hemodynamics.
a cis-acting
motif
that accounts
for
effect
of 13 on the skeletal
a-actin
was
et a!.,
induced
rats after
ofa-actin
occurred
synthesis
(13 stimuof the MHC).
Their
regulatory
effect
of
indirect
effect
secWe have identified
a direct
regulatory
gene.
Furthermore,
Cell
CTF/NF-
I Conscnsus
PROBE
HSA
-210/-178
COMPETITOR
Human
CCCGCGTI’ACCTGGGACCGGGCCAACCCGCTCC
Harvey
ras-1
AATFCCGAATGGCGCGCAGCCAATGGTAGGCA
B
NUCLEAR
EXTRACF
PURIFIED
CFF/NF-
I
I
1
::
N
.a
COMPETITOR
Fig. 1 1. NF-l/CTF
o-actin
promoter.
was
shown
interacts
The HSA
to interact
with
and C2 nuclear
extracts
mobility
shift analysis.
0
,,
0
0
<
,
.<
0
B
C’,
X
0
<
r
.o
‘
X
I
with
the skeletal
-210/-186
probe
purified
NF-1/CTF
by gel electrophoresis
this could
account
for the quick
response
of skeletal
aactin to hormonal
or hemodynamic
changes.
The skeletal
a-actin
IRE, like the rodent
and human
a-MHC
TREs, fit into the consensus
core
13 receptorbinding
motif
G/
GGT/A CA/c
which
is a “half site” of
the larger
TGACCT.
& Differentiation
TGG(A/C)NNNNNGCCAA
A
CRUDE
Growth
palindromic/dyad
Retinoic
acid
repeat sequence
AGGTCA...
and vitamin
D receptors
that are
members
of the steroid/thyroid
superfamily
of ligandmodulated
receptors
are also known
to bind this palmdromic
sequence.
Very
recently,
palindromic
half sites
(encoding
estrogen
and glucocorticoid
response
elements)
have
been
implicated
in positive
and negative
responses
to estrogen
and glucocorticoids
via interactions
with
their
respective
receptors
and transcription
factors
such as los and jun (AP-i)
with
Oct-i
(50-54).
The function
of the skeletal
a-actin
TRE during
myogenic
ontogeny
may be controlled
by similar
regulatory
mechanisms
involving
ubiquitous
transcription
factors
such as
SpI, SRF, and CTF, which
were
found
to interact
with
regions
flanking
the a-actin
TRE. These transcription
factors have also been
shown
to be involved
in proteinprotein
interactions
involving
other
factors
(55).
The
ubiquitous
proteins
such as SpI, SRF, and CTF may act
like or in concert
with the TR auxiliary
protein
(56, 57),
and other
tissue-specific
factors
(58,
59),
which
have
been shown
to enhance
binding
of the receptor
to the
TRE in vivo via multicomponent
protein
complexes.
Materials
and Methods
Cell Culture
and Transfection.
Mouse
myogenic
C2 cells
supplemented
with 20%
previously
(43). This cell
biochemically
and mormyotubes
by mitogen
with 2% FCS in 10%
was essentially
complete
within 72to isoform
switching
in the actin mul-
(60, 61) were grown
in DMEM
FCS in 10% CO2 as described
line was induced
to differentiate
phologically
into
multinucleate
withdrawal
(DMEM
supplemented
C02).
Differentiation
96 h with
respect
39
40
Characterization
of a-Actin
tigene
(8). However,
family
differentiate
TREs
at a very
these
high
cells will spontaneously
confluence
(100%)
in
the
presence
of mitogens.
COS-i
fibroblasts
were grown in
DMEM supplemented
with i0%
FCS in iO% CO2.
Each 60-mm
dish of cells was transiently
transfected
with
10
of
reporter
plasmid
mixed with an additional
vectors
or pUCi8
DNA.
each transfection
by the addition
DNA
expressing
CAT,
amount
of the TR expression
The total amount
of DNA in
experiment
(1 1 jzg) was kept constant
of pUCi8
DNA.
Prior to transfection,
the
cells were cultured
for 24 h in thyroid-deficient
medium
containing
iO%
charcoal-stripped
FCS in DMEM.
The
DNA mixtures
were cotransfected
into C2 myoblasts
and
COS-i
fibroblasts
by the liposome-mediated
procedure.
We used the cationic
lipid DOTAP.
Unilamellar
vesicles
were
formed
by mixing-the
appropriate
DNAs
with
30-
40 I of DOTAP and 1 X HEPES-buffered
saline to a total
volume
of 200 jab. After a 10-mm
incubation
at room
temperature,
this mixture
was added
to 6 ml of fresh
culture
(thyroid-deficient)
medium
and added
to the
cells, which were between
50 and 70% confluent.
After
a period
of 20-24
h, fresh medium
with or without
T3
(10 nM) was added to the cells. The cells were harvested
for
the
assay
of CAT
enzyme
activity
60-72
h after
the
transfection
period.
Each transfection
experiment
was
performed
three times using at least two different
plasmid
preparations
in order to overcome
the variability
inherent
in transfections.
CAT Assays. The cells were harvested,
and the CAT
activity
was measured
as previously
described
(62). Aliquots of the cell extracts
were incubated
at 37#{176}C,
with
0.1-0.4
cCi of [14C]chloramphenicol
of S mM acetyl CoA
(Amersham)
in the
presence
and 0.25 M Tnis-HCI,
pH
7.8. After a 2-4-h
incubation
period,
the reaction
was
stopped
by the addition
of 1 ml ethyl acetate, which was
used to extract
the chloramphenicol
and its acetylated
forms.
The
extracted
gel thin-layer
ously
(61).
scintillation
Plasmids.
materials
were
chromatography
Quantitation
plates
of CAT
assays
analyzed
on
as described
was
performed
counting
of the chromatograms.
The plasmid
pEMSV-CAT
was described
Davis et a!. (36). The
cerless
SV4O promoter
plasmid
linked
silica
previ-
cells
were
using
10%
Nonidet
serum
albumin,
We thank
Dr. Howard
Towle
for generously
providing
and -13 expression
vectors
in CDM8
and Drs. S. Jackson,
R. Tjian for their generous
gift of SpI and CTF. We thank
the SpI and CTF oligonucleotides.
following
incubation
in 10 mi HEPES (pH 7.9), 10 mt’i
KCI, 0.1 mM EDTA, 0.1 mri ethyleneglycol
bis(f3-aminoethyl ether)-N,N,N’,N’-tetraacetic
acid, 1 mivi dithiothreitol, 0.5 mM phenylmethylsulfonyl
fluoride,
and 2 pg/mI
3-5
g
of
the rat c-erbA-a
N. Tanese,
and
Larry Kedes
for
References
1 . Glass,
C. K., and
the thyroid
hormone
Holloway,
receptor.
1. M. Regulation
Biochim.
Biophys.
of
gene expression
by
Ada,
1032: 157-176,
1990.
2.
Forman,
nuclear
B. M.,
and
hormone
nol., 1293-1301,
3. Norman,
Samuels,
receptors:
H.
H. Interactions
the regulatory
zipper
among
a subfamily
model,
Mol.
F.,
growth
hormone
Biol. Chem., 264:
Lavin,
gene
T. N.,
contains
12063-12073,
Baxter,
1. D.,
multiple
and
thyroid
West,
B. L. The
response
elements.
5. Swynghedauw,
tractile
proteins
771, 1986.
A. M., Mercadier,
J. J., and Schwartz,
during
cardiac
growth.
nt. Rev. Cytol.,
263:
Chem.,
8.
Rohrer,
myosin
Chem.,
chain
mRNA
6370-6374,
D.
hormone
hormone
receptor
1. Biol.
sequence.
B. Developmental
and functional
adaptation
in cardiac
and skeletal
muscle,
Physiol.
Rev.,
heavy
K.,
and
rat
J.
1989.
4. Lavin, T. N., Baxter,
I. D., and Horita, S. The thyroid
binds to multiple
domains
of the rat growth
5’ flanking
Chem.,
263: 9418-9426,
1988.
fast myosin
of
Endocri-
1990.
M.
by
P-40
and
as a nonspecific
The assays were
mm and electropolyacrylamide:
and 2 mi EDTA.
acid, dried, and
Acknowledgments
7. Russel,
S. D., Cambon,
N., Nadal-Ginard,
Thyroid
hormone
induces
a nerve independent
pCAT promoter
(an enhanto CAT in a pUCi9
back-
lysed
6 cg of bovine
polydeoxyinosinic-deoxycytidylic
acid
competitor
in Dignam
buffer C (64).
incubated
at room temperature
for 20
phoresed
through
a
6%
(20:1
bisacrylamide)
gel in 80 mi Tnis borate
Gels were briefly soaked in 10% acetic
autoradiographed.
6. Lompre,
expression
.
The
of protein,
by
bone) was purchased
from Promega.
The plasmids
expressing the rodent
a- and fl-c-erbA
genes in the eukaryotic
expression
vector
CDM8,
containing
the
cytomegalovirus
promoter,
and SV4O origin of replication
were
described
by Zilz
et a!. (35). The
plasmids
pHSA2000CAT,
pHSA432CAT,
and pHSA153CAT
were
described
by Muscat et a!. (32, 43).
Oligonucleotides.
The sequences
of the oligonucleotide probes used in the EMSA experiments
are as follows:
CArG
5’ GAAGGGGACCAAATAAGGCAAGGTGG
3’;
MEF-2
5 ‘ ATCTGAAAGGCATAGCCCCATATATCAGTGATATAAATAGAACCTGCAG
3 ‘ The underlined
sequences
indicate
the binding
region for the associated
factor.
The TRE, SpI, and CTF oligonucleotides
were
described
in the text.
Nuclear
Extracts and Gel Mobility
Shift Assays. Nuclear extracts were prepared
by the method
of Schreiber
et a!. (63).
of leupeptin
and aprotinin
(Boehninger
Mannheim).
Nuclear proteins
were extracted
with 0.4 NI NaCI, 20 mi
HEPES (pH 7.9), 1 mM EDTA, 1 mi ethyleneglycol
bis(13aminoethyl
ether)-N,N,N’,N’-tetraacetic
acid, i m’i dithiothreitol,
1 mM phenylmethylsulfonyl
fluoride,
and 2
tg/ml aprotinin
and leupeptin.
Each binding
mixture
(25 tI) contained
i -2 ng of a T4
polynucleotide
kinase-labeled
DNA fragment,
5-10
zg
K. Changes
124: 137-186,
B., and
skeletal
in gene
1991.
Whalen,
precocious
in rat hindlimb
of con66: 710-
R. G.
expression
of
J.
muscle.
Biol.
1988.
Hartong,
retinoic
R., and
acid
Dillman,
on slow
W.
H.
sarcoplasmic
heavy
chain
a-gene
expression
266: 8638-8646,
1991.
in
Influence
of thyroid
calcium
cardiac
ATPase and
myocytes.
J. Biol.
9, Bains, W., Ponte, P., Blau, H., and Kedes, L. Cardiac actin is the major
actin gene product
in skeletal muscle differentiation
in vitro.
Mol. Cell.
Biol.,
10.
4: 1449-1453,
Gunning,
1984.
P., Hardeman,
and Kedes, L. Differential
human myogenesis.
Mol.
11. Gunning,
alpha-cardiac
P., Ponte,
actin genes
cle and heart.
Mol.
Cell.
E., Wade,
patterns
Cell. Biol.,
R., Ponte,
P., Bains,
of transcript
7: 4100-4114,
P., Blau,
H., and
are co-expressed
Biol., 3: 1985-1995,
Kedes,
in adult
W.,
Blau,
accumulation
1987.
H.,
during
L. Alpha-skeletal
human
skeletal
and
mus-
1983.
12. Hayward,
L. j., and Schwartz,
R. J. Sequential
expression
of chicken
actin genes during
myogenesis.
J. Cell Biol., 2: 1044-1051,
1982.
13. Mayer,
Y., Czosnek,
H., Zeelon,
P., Yaffe,
sion of the genes coding
for the skeletal
and
Nucleic
Acids
Res., 12: 1087-1 100, 1984.
14.
Minty,
skeletal
actin
A.
I.,
actins
Alonso, S., Caravatti,
M., and Buckingham,
actin mRNA in the mouse and its identity
Cell, 30: 185-192,
1982.
muscle
mRNA.
D., and Nudel,
cardiac
U. Expresin the
heart.
M. E. A fast
with cardiac
15. Vandekerckhove,
expression
of skeletal
muscle
tissues
and
16. Vandekerckhove,
evolution
of muscle
j., Bugaisky, G., and Buckingham,
M. Simultaneous
muscle and heart actin protein
in various striated
cells. J. Biol. Chem.,
261: 1838-1843,
1986.
J., de Couet,
H-G.,
specific actins: a protein
and Weber,
K. Molecular
chemical
analysis. In: C. G.
Cell Growth
dos
Remedios,
Muscle
1983.
and
and
j. A. Barden
Non-muscle
Cells,
Actin:
Structure
pp. 241-248.
(eds.),
Sydney:
and
1 7. Schwartz,
K., Dela Bastie, D., Bouvaret,
P., Olieviero,
and Buckingham,
M. a-Skeletal
actin mRNA’s
accumulate
phied adult rat hearts. Circ. Res., 59: 551-556,
1986.
18. Winegrad,
S., Wisnewsky,
C., and
hormone
on the accumulation
of mRNA
hearts from normal and hypophysectomized
USA, 87: 2456-2460,
1990.
Function
Academic
in
Press,
P., Alonso, S.,
in hypertro-
Schwartz,
K. Effect of thyroid
for skeletal and cardiac actin in
rats. Proc. NatI. Acad. Sci.
19. Bishopric,
N., and Kedes, L. Adrenergic
regulation
of the skeletal aactin gene promoter
during
myocardial
cell hypertrophy.
Proc. NatI.
Acad. Sci. USA, 88: 2132-2136,
1991.
20.
Lompre,
cardiac
A-M.,
ventricular
Nadal-Ginard,
B., and Mahdavi,
V. Expression
a- and fi-MHC
genes
is developmentally
Chem.,
259: 6437-6446,
1984.
and
of the
hormon-
J. Biol.
ally regulated.
21. Isumo, S., Nadal-Ginard,
B., and Mahdavi,
V. All members
MHC multi-gene
family
respond
to thyroid
hormone
in a highly
specific manner.
Science (Washington
DC), 231: 597-600,
1986.
of the
tissue
22. Gustafson,
T. A., Markham,
B. E., BahI, J., and Morkin,
E. Thyroid
hormone
regulates
expression
of a transfected
a-MHC
fusion gene in
fetal heart cells. Proc. NatI. Acad. Sci. USA, 84: 3122-3126,
1987.
23. Izumo,
S., and Mahdavi,
V. Thyroid
hormone
receptor
generated
by alternative
splicing differentially
activate
MHC
cription.
Nature
(Lond.), 334: 539-543,
1988.
a-isoforms
gene trans-
24. Tsika, R. W., BahI, j., Leinwand,
L. A., and Morkin,
E. Thyroid
hormone
regulates
expression
of a transfected
human
a-MHC
fusion
gene in fetal rat heart cells. Proc. NatI. Acad. Sci. USA, 87: 379-383,
39. Sturm, R. A., Das, G., and Herr, W. The ubiquitous
protein
Oct-i
contains
a POU domain
with a homeo
Genes & Dev., 2: 1582-1599,
1988.
9: 5022-5033,
1985.
expression.
44.
Mol.
Miwa,
and mutually
cardiac actin
45,
negative
skeletal
47,
Walsh,
K., and
N. C., Rigby, P. W., and Ziff, E. B. Trans-acting
protein factors
of eucaryotic
transcription.
Genes & Dev., 2: 2657-
2681,
1988.
32. Muscat,
G. E. 0., Gustafson,
T. A., and Kedes, L. A common
factor
regulates
skeletal and cardiac a-actin transcription
in muscle. Mol. Cell.
Biol.,
8:4120-4133,
1988.
33. Tuil, D., Clergue,
N., Montarass,
D., Pinset, C., Kahn, A., and PhanDinh Tuy, F. CCArGG
boxes, cis-acting
elements
with dual specificity:
muscle specific
activation
and serum responsiveness.
J. Mol. Biol., 213:
677-686,
34.
Boxer,
actin
35.
L. M., Prywes,
CArG-binding
sponse
thyroid
1990.
factor.
ZiIz,
factor
Mol.
514 promoter.
36. Davis,
transfected
1987.
9: 515-522,
M. B., and Towle,
response
J. Biol.
G., and Kedes,
is indistinguishable
Cell. Biol.,
N. D., Murray,
hormone
R., Roeder,
elements
located
Chem., 265: 8136-8143,
R., Weintraub,
cDNA converts
H., and
fibroblasts
from
L. The sarcomeric
c-los serum
re-
the
1989.
H. Identification
far
upstream
of multiple
from
the
rat
1990.
Lassar, A. Expression
of a single
to myoblasts.
Cell, 51: 987-1000,
box
domains
have
R. J. A combination
of closely
elements
regulates
Biol., 10: 528-538,
Mol. Cell.
Schimmel,
P. DNA-binding
is important
site
for expression
positive
of the
human
associated
transcription
1990.
different
promoter
Cell. Biol., 9: 2 191for two
in muscle
skeletal
cells.
actin
Mol.
Cell.
1988.
48. Evans, T., DeChiara,
T., and Efstratiadis,
A. A promoter
of the
insulin-like
growth factor II gene consists of minimal
control
elements.
Mol. Biol., 199: 61-81,
1988.
49. Gustafson,
which
interact
50.
Activation
T. A., and Kedes,
L. H. Identification
with
functional
regions
of the
Cell. Biol., 8: 3269-3283,
1989.
Mol.
Gaub,
M.
P., Bellard,
M.,
Scheuer,
of multiple
human
cardiac
I., Chambon,
P., and
of the ovalbumin
gene by the estrogen
complex.
Cell, 63: 1267-1276,
1990.
1990.
Jones,
promoter
1987.
CArG
promoter
gene.
8: 1800-1802,
29. Santoro,
C., Mermod,
N., Andrews,
P. C., and Tjian, R. A family of
human CCAAT-box-binding
proteins activate transcription
and DNA replication:
cloning and expression
of multiple
cDNAs.
Nature (Lond.), 334:
218-224,
1988.
31.
R. J., and Moore,
hormone
regulatory
roles in expression
Cell. Biol., 7: 2803-2813,
1987.
cis-acting
a-actin
factors
51. Diamond,
cription
factor
and the regulation
growth
46. Walsh, K. Cross binding of factors to functionally
elements
in the c-los and skeletal actin genes. Mol.
2201, 1989.
los-jun
P., Kelly, T., and Tjian, R. A
transcription
and DNA repli-
rat
7: 4089-4099,
L. Duplicated
K. L., and Schwartz,
and
ofthe
Biol.,
Kedes,
dependent
gene. Mol.
Chow,
positive
Cell.
T., and
28. Jones, K. A., Kadonaga,
J. T., Luciw, P. A., and Tjian, R. Activation
of
the AIDS retrovirus
promoter
by the cellular
transcription
factor, Spl.
Science (Washington
DC), 232: 755-759,
1986.
30. Jones, K. A., Kadonaga,
J. T., Rosenfeld,
cellular
DNA binding
protein
that activates
cation. Cell, 48: 79-81,
1987.
J. W., Koenig,
43. Muscat, G. E. 0., and Kedes, 1. Multiple
5’ flanking
regions of the
human a-skeletal
actin gene synergistically
modulate
muscle specific gene
promoter.
activation
Sci., 11: 20-
Larsen,
P. R., Harney,
characterization
of the
elements
required
for induction
by thyroid
hormone
with and without
a
cotransfected
a-type thyroid
hormone
receptor.
J. Biol. Chem., 264: 178182, 1989.
Biol.,
27. Kadonaga,
J. T., Jones, K. A., and Tjian, R. Promoter-specific
of RNA polymerase
II transcription
by Spl. Trends Biochem.
23, 1986.
1989.
42. Brent,
G. A.,
D. D. Functional
25. Markham,
B., BahI, j., Gustafson,
T. A., and Morkin,
E. Interaction
of
a protein factor within a thyroid
hormone
sensitive
region of rat a-MHC
gene. j. Biol. Chem., 262: 12856-12862,
1987.
511-517,
octamer
binding
box subdomain.
41 . Forman, B. M., Yang, C-R., Stanley, F., Casanova,
J., and Samuels, H.
H. c-erbA
protooncogenes
mediate
thyroid
hormone-dependent
and
independent
regulation
of rodent growth hormone
and prolactin
genes.
Mol. Endocrinol.,
2: 902-911,
1988.
promoter
P., Kadonaga,
j. T., Barrera-Saldana,
H., Takahashi,
K.,
P., and Tjian, R. Bi-directional
Sv 40 transcription
mediated
Sp 1 binding
interactions.
Science
(Washington
DC), 230:
41
40. Gosset, L. A., Kelvin,
D. J., Sternberg,
E. A., and Olson,
E. A new
myocyte
specific
enhancer
binding
factor that recognises
a conserved
element
associated
with multiple
muscle specific genes. Mol. Cell. Biol.,
1990.
26. Gidoni,
Chambon,
by tandem
& Differentiation
M., Miner,
interactions:
from a single
DNA
J., Yoshinaga,
selectors
element.
of
Science
S. K., and
positive
(Washington
receptor
rat
J.
proteins
a-actin
Corsi,
P. 5.
involves
the
Yammamoto,
or negative
DC), 249:
K. Transregulation
1266-1271,
52.
Yang-Yen,
H-F., Chambard,
J-C., Sun, Y-L., Smeal, T., Schmidt,
T. J.,
J., and Karin, M. Transcriptional
interference
between
c-jun and
the glucocorticoid
receptor:
mutual
inhibition
of DNA binding
due to
Drouin,
direct
protein-protein
53.
Schule,
N., Verma,
tein c-jun
interaction.
R., Rangarajan,
Cell,
I., and Evans, R. M.
and the glucocorticoid
54. Bruggemeier,
U.,
Ubiquitous
transcription
promoter
through
Kalff,
1991.
55.
R. E., Shaw,
Herrara,
1205-1215,
M.,
factor
Franke,
Oct-i
W.
Darling,
heterodimeters
with
hormone
Nordheim,
58.
Murray,
59.
M.
68-72,
with
the T3 receptor.
B., and
Towle,
Mol.
M. A., and
37. Fried, M., and Crothers,
D. M. Equilibrium
and kinetics of lac repressor-operator
interactions
by polyacrylamide
gel electrophoresis.
Nucleic
Acids Res., 9: 6505-6525,
1981.
distinct
nuclear
thyroid
hormone
38. Baumruker,
T., Sturm, R. A., and Herr, W. OBP100
binds remarkably
degenerate
octamer
motifs through
specific
interaction
with flanking
sequence.
Genes & Dev., 2: 1400-1413,
1988.
60.
Cell,
1989.
H. C. Identification
Berrodin,
protein
T. J. Thyroid
dependent
response
element.
E. R., and Chin,
binds
DNA
and
Endocrinol.,
5: 73-84,
Darling,
D. S., and
Koenig,
R.
that interfere
with
transcriptional
interaction
with a nuclear
protein.
of nuclear
binding of the thyroid
hormone
receptor
element.
Mol. Endocrinol.,
3: 1434-1442,
Lazar,
receptors.
A. Occupation
of the ccomplex
is unaltered
by
J.
57, O’Donnell,
A. L., Rosen,
E. D.,
Thyroid
hormone
receptor
mutations
activation
also interfere
with receptor
Mol. Endocrinol.,
5: 94-99,
1991.
enhance
response
J., Yang,
oncopro1990.
S., Burnside,
J., Winslow,
receptor-auxiliary
protein
(TRAP)
D. S., Beebe,
Triiodothyronine
forms
1991.
between
1217-1226,
S., Scheidereit,
C., and Beato,
M.
mediates
induction
of the MMTV
fos serum response
in vivo by a multiprotein
growth factor induction.
Nature (Lond.), 340:
56.
L., Bolado,
antagonism
Cell, 62:
interaction
P. E., and
1990.
S., Ransome,
Functional
receptor.
synergistic
64: 565-572,
62:
P., Kliewer,
and
factors
to a thyroid
J.
that
hormone
1989.
hormone
receptors
form
independent
Mol.
complexes
with
Endocrinol.,
4: 1627-1635,
a
1990.
Yaffe,
ogenic
725-727,
cells
D., and Saxel,
isolated
1977.
from
0.
Serial
dystrophic
passaging
mouse
and differentiation
muscle.
Nature
(Lond.),
of my270:
42
Characterization
of a-Actin
TREs
61. Yaffe, D., and Saxel, 0.
requirements
for differentiation.
A myogenic
cell
Differentiation,
line with altered
serum
7: 1 59-166,
1977.
62.
Gorman,
C. M., Moffat,
L. F., and Howard,
B. H. Recombinant
which express chloramphenicol
acetyltransferase
in mammalian
cells. Mol. Cell. Biol., 2: 1044-1051,
1982.
genomes
63.
Schreiber,
E., Mathias,
P., Muller,
M.,
and
Schaffner,
W.
Rapid
detection
of octamer
from a small number
binding proteins
of cells. Nucleic
with mininuclear
extracts, prepared
Acids Res., 17: 6419, 1989.
64. Dignam, J. D., Lebovitz,
R. M., and Roeder, R. Accurate
transcription
initiation
by RNA polymerase
II in a soluble extract from isolated
mammalian nuclei. Nucleic Acids Res., 1 1: 1475-1489,
1983.