Single Strand DNA-Binding
Proteins
and Thyroid Transcription
Factor-l
Conjointly
Regulate Thyrotropin
Receptor Gene Expression
Hiroki Shimura*, Yoshie Shimurat,
Shoichiro
IkuyamaS,
and Leonard
Masayuki
D. Kohn
Ohmori,
Section
on Cell Regulation
Laboratory
of Biochemistry
and Metabolism
National
Institute
of Diabetes
and Digestive
and Kidney
National
Institutes
of Health, Bethesda,
Maryland
20892
An element,
-186 to -176 base pairs (bp), in the
minimal TSH receptor
(TSHR) promoter
binds thyroid transcription
factor-l
(TTF-1) and is important
for both constitutive
expression
and TSHIcAMPinduced
negative
autoregulation
of the TSHR in
thyroid cells. An element on the noncoding
strand
of the TSHR, contiguous
with the V-end
of the
TTF-1 element, has single strand binding activity. It
is distinct
from the TTF-1 site, as evidenced
by
competition
experiments
using gel shift assays;
but the association
of the two elements is not random. Thus, the single
strand
binding
protein
(SSBP) element also exists contiguous
to the 5’end of an upstream
TTF-1 site, -881 to -866 bp;
mutation
of two conserved
nucleotides
in each
SSBP element results in the loss of SSBP binding
and cross-competition.
Transfection
experiments
indicate that full, constitutive
TSHR gene expression in FRTL-5 thyroid cells requires the binding of
both SSBPs and TTF-1, since mutation
of either
element halves thyroid-specific
promoter
activity,
whereas
mutation of both decreases
promoter
activity to values near those of a control
vector.
Transfection
experiments
with rat liver cells support their independent
activities and show that the
SSBP site contributes
to TSHR gene expression
in
non-thyroid
tissue. The SSBPs function conjointly
with TTF-1 in thyroid-specific,
TSH/cAMP-induced
negative autoregulation
of the TSHR. Thus, TSH or
forskolin-treated
FRTL-5
cells coordinately
decrease TSHR RNA levels and TSHR DNA binding to
both the SSBPs and TTF-1; also the maximal TSH/
CAMP-induced
decrease
in gene expression
requires both elements.
The TSH-induced
effect in
each case is inhibited
by cycloheximide;
the TSHinduced
decrease
in SSBP/DNA
complex
formation requires the presence
of insulin or calf serum,
exactly as does TSH-induced
down-regulation
of
0888-8809/95/53.00/O
Molecular
Endocmology
CopyrIght
0 1995 by The Endocme
Diseases
TSHR RNA levels. In sum, full, constitutive
expression of the TSHR in thyroid cells requires TTF-1 and
the SSBPs to bind separate,
contiguous
elements
on the TSHR promoter.
TSH/cAMP
decreases
the
binding of each factor to its respective
site, thereby
decreasing
TSHR gene expression.
The role of the
SSBP and TTF-1 sites in constitutive
TSHR expression and in TSH/cAMP-induced
negative regulation
of the TSHR is, therefore,
additive and independent. (Molecular
Endocrinology
9: 527-539, 1995)
INTRODUCTION
Functional expression of the TSH receptor (TSHR) and
thyroglobulin
(TG) genes is characteristic
of the thyroid; nevertheless, the regulation and role of each is
different. The TSHR regulates thyroid growth and thyroid hormone production;
TG is important only for the
latter (l-4). TSHR initiates the CAMP signal; TG responds to it (l-4). In FRTL-5 thyroid cells, TSHR gene
expression
is increased,
then decreased
by TSH/
CAMP, as a function of time and CAMP levels (5); TG
gene expression
is only increased
(6, 7). Although
TSHR RNA is detected in the thyroid and not brain,
spleen, liver, or lung (8), it can be detected in retroorbital (9) adipose (9), or osteosarcoma
cells (10) and
lymphocytes
(11); TG RNA expression
is absolutely
thyroid-specific
(l-4, 12-l 5).
We have defined a minimal TSHR promoter, -220
to -39 bp’, which exhibits thyroid-specific
expression
and TSH/cAMP
autoregulation
(16-21). Thyroid transcription factor-l (TTF-1) binds to the minimal TSHR
promoter,
-186
to -176
bp (20, 22, 23), and
is involved in both activities (20, 23). lTF-1 is not,
’ Numbering is relative to the ATG start codon which is
defined as + 1. We have used this convention
since the TSHR
5’-flanking
region is a GC-rich
promoter
with multiple
major
start sites, as in the case of several housekeeping
genes (16).
Since it is not yet clear which start site is used physiologically,
we have adapted
this convention.
We have used the same
numbering
system in our separate
reports
(16-21,
23).
Smety
527
MOL
END0
. 1995
Vol 9 No. 5
528
however, solely responsible for TSHR gene expression
in the thyroid, since significant
expression
remains
after deletion of the TTF-1 element (16-21). Further,
the region encompassing
the TTF-1 element in the
TSHR promoter, -194 to -169 bp, expresses weak
enhancer activities in nonthyroidal,
Buffalo rat liver
(BRL) cells. Although TSH/cAMP-induced
up-regulation of the TSHR (20) gene, like the TG gene (24, 25),
appears to be associated
with protein kinase A-dependent increases in lTF-l/DNA
complex formation,
TSH/cAMP-induced
down-regulation
of the TSHR
gene is associated with decreased lTF-1 mRNA levels
and decreased TTF-l/TSHR
complex formation (20).
There is no concurrent
down-regulation
of TG gene
expression. One explanation
for this is a compensatory increase in Pax-8 interactions
with a lTF-1 recognition site in the TG promoter (20, 23, 25); Pax-8
does not interact with the l-l-F-1 site in the minimal
TSHR promoter (20, 22, 23). The possibility remains,
however, that other factors, binding to different promoter elements, are important in TSH/cAMP
downregulation, as well as full expression, of the TSHR gene
in thyroid cells. This report identifies and characterizes
one of these elements.
We identify, for the first time, an element on the noncoding strand, 5’- and contiguous with lTF-1 sites in the
TSHR, which binds single strand binding proteins
(SSBPs). We show that the SSBP element on the TSHR
is distinct from that for TTF-1, is not associated with
lTF-1 sites on the TG promoter, and is required for full
expression of the TSHR gene, functioning additively with,
but independently of, the lTF-1 site. We also show that
A
the SSBP element functions conjointly with the TTF-1
site in TSH/cAMP-induced
down-regulation
of TSHR
gene expression in the thyroid. The conjoint function of
the SSBP and lTF-1 elements in the TSHR appears,
therefore, to additionally contribute to differences in transcriptional regulation of the TSHR and TG genes. Finally,
we show that the SSBP element contributes to TSHR
gene expression in nonthyroid cells. Although the description of the SSBPs themselves remains preliminary,
characterization
of the function and role of the SSBP
element on the TSHR should allow their subsequent
cloning and comparison to other hormonally regulated,
functionally defined SSBPs.
RESULTS
TSH-Regulated
Single Strand DNA-Binding
Proteins Specifically
Recognize
a Site on the
Noncoding
Strand of TSHR Minimal Promoter
Near, but Separate
from, the TTF-1 Site
A double-stranded,
oligonucleotide
probe, - 194 to
-169 bp, can form a lTF-VTSHR
DNA complex with
nuclear extracts from FRTL-5 thyroid cells (Fig. 1A;
Ref. 20). lTF-l/TSHR
complexes are not formed with
extracts from FRT or BRL cells, because they contain
no lTF-1 (Fig. lA, lane 1 vs. 3 and 4; Ref. 20); and
treatment of FRTL-5 cells with TSH (6H) decreases
formation of the TTF-l/TSHR
DNA complex (Fig. lA,
lane 2 vs. lane 1; Ref. 20). When the noncoding
(Fig.
1 B), but not the coding strand (Fig. 1 C), of this oligo-
C
B
Probe
Non
Double-strand
Coding
coding
Extract
‘\_ ,.)
TTF-1 -
.‘I
:
1
Fig. 1. Ability of Nuclear
2
3
4
1
2
3
4
1
2
3
4
Proteins in Extracts
from FRTL-5,
FRT, or BRL Cells To Form Protein/DNA
Complexes
when DoubleStranded
or Single-Stranded
Forms of an Oligonucleotide
Spanning
- 194 to - 169 bp Were Used as Radiolabeled
Probes
Double-stranded
(A), noncoding
(B), and coding strand (C) oligonucleotides
were used as probes. They were incubated
with nuclear
extracts from FRTL-5 cells maintained
with (6H) or without (5H) TSH, from FRT cells, or from BRL cells. The position of the lTF-l/lSHR
DNA complex is indicated (20,22,23).
Solidamws in panels A and B denote protein/DNA
complexes
distinct from the TTF-1 complex
and which form with the noncoding
strand only. The dashed arrow depicts another complex
formed with the noncoding
strand but
distinct from the others based on their apparent
absence in FRT and BRL cells and by competition
data to be presented
in Fig. 2.
529
SSBPs and TSHR Gene Expression
tous factors that interact with the lTF-I sites on the
TG promoter. Conversely, the TTF-1 sites of the TG
promoter do not interact with these SSBPs.
Formation of the SSBPITSHR
DNA complexes was
not prevented
by the noncoding
(nc) or coding (c)
strands of oligo K (27), which contains the insulinresponsive
element in the TG promoter
and binds
thyroid transcription
factor-2 (TTF-2) (Fig. 2B, lanes 8
and 9 vs. 1). There is an upper protein/DNA
complex
formed with the noncoding
strand (Fig. lB, dashed
arrow) which migrates faster than the l-l-F-1 complex
(Fig. 1 B vs. 1 A), does not interact with oligonucleotides
from the TG promoter that bind TTF-1 (oligo A) or
lTF-1 and Pax-8 (oligo C) (Fig. 28, dashed arrows), as
evidenced by the lack of competition,
and does not
interact with double strand TSHR DNA (Fig. 1A; Fig.
2A, lanes l-5). This complex is not, however, present
in FRT or BRL cells (Fig. 1B) and its formation is not
similarly prevented, as a function of concentration,
by
the homologous
unlabeled double- or single stranded
oligonucleotides
(Fig. 2A). It is different, therefore,
from the SSBPs to be characterized
herein. Interestingly, its formation is also decreased by TSH treatment
of FRTL5 cells (Fig. 1 B, lane 2 vs. 1) and appears to be
inhibited by the noncoding
strand of oligo K (Fig. 28,
lane 9). The nature of this complex is being separately
characterized.
nucleotide is used as probe, a major and minor protein-DNA complex (Fig. 1 B, lane 1, so/id arrows) can
be identified that has the same electrophoretic
mobilities as observed with lower DNA complex(es) formed
by the double-stranded
probe (Fig. 1 B vs. IA, lane 1).
Unlike the TTF-1TTSHR
DNA complex, DNA-binding
activity of the noncoding single strand probe appears
to be more ubiquitously expressed, since it is seen in
nuclear extracts from FRT and BRL cells, albeit at
lower levels than in extracts from FRTL-5 cells (Fig. 1 B,
lanes 3 and 4 vs. 1 and 2). Like the lTF-l/TSHR
DNA
complex, the formation of these complexes
is decreased by TSH treatment of the FRTL5 cells (Fig.
lB., lane 2 vs. 1).
The DNA complexes formed by the FRTL-5 thyroid
cell extract and radiolabeled
noncoding strand probe
(Fig. 2, so/id arrows) are competed by the homologous
unlabeled
single strand (Fig. 2A, lanes 7-10 vs. 6)
significantly better than by double-strand
oligonucleotide (Fig. 2A, lanes 2-5 vs. l), but not at all by a
250-fold excess of the noncoding
(nc) or coding (c)
single strand oligonucleotides
containing
the TTF-1
binding sites on the TG promoter:
oligo A (Fig. 2B,
lanes 4 and 5 vs. 1) or oligo C (Fig. 2B, lanes 6 and 7
vs. 1). Double-stranded
oligo A or C also did not prevent complex formation with the radiolabeled,
noncoding strand probe, -194 to -169 bp, of the TSHR
(data not shown). These results indicate that the
SSBPs in the FRTL-5 thyroid cell extracts interact with
a site on the noncoding
strand distinct from l-l-F-1 on
the TSHR promoter. Additionally,
since oligo C also
interacts with Pax-8 and oligo A with a ubiquitous
factor present in FRTL-5 cell extracts (14, 26), the
SSBP complexes do not involve Pax-8 or the ubiqui-
The SSBP and lTF-1
Sequences
Are Contiguous
To define the binding site of the SSBPs relative
TTF-1, the contacts with purines on both strands
to
of
B
A
Double
strand
f
Competitor
Elements
0
1
Non coding
7
5
20 50
2
3
4
250
0
5
6
;
7
20
50
25:
8
9
10
1
23456769
Fig. 2. Competition Analyses of the Protein/DNA Complexes Formed by the Noncoding Strand of the Oligonucleotide Spanning
-194 to -169 bp and FRTL-5 Cell Nuclear Extracts
Nuclear extract from FRTL-5 cells maintained without TSH for 7 days was incubated with the radiolabeled noncoding strand
probe in the presence or absence of unlabeled oligonucleotides that were used as competitors. Solid anddashedarrows
in panels
A and B denote protein/DNA complexes that react with the noncoding strand. In panel A, increasing amounts of unlabeled,
double-stranded and unlabeled noncoding single-stranded oligonucleotides spanning -194 to -169 bp in the TSHR minimal
promoter were tested as competitors at the noted fold-excess over probe. In panel B, a 250-fold excess over probe of the same
unlabeled double strand (ds) or noncoding strand (nc) oligonucleotide is compared with a 250-fold excess of single stranded
oligonucleotides from the coding (c) or noncoding (nc) regions A, C, or K from the TG minimal promoter. The SSBP complexes
noted by the solid arrows can be distinguished from the complex denoted by the dashed arrow based on differences in
competition by double strand DNA from the region (lanes l-5 in panel A and lane 2 vs. 1 in panel B) and in sensitivity to
competition by different concentrations of the noncoding strand (lanes 6-10 in panel A).
MOL
530
END0
Vol 9 No. 5
. 1995
strand (open circles, bottom line) interfered with SSBP
binding. In contrast, lTF-l-binding
was inhibited by
methylation of nucleotides
at -183, -182, and -181
bp on the noncoding
strand (Fig. 3C, next to bottom
the oligonucleotide
spanning -194 to -169 bp were
determined
by methylation interference (Fig. 3, A and
B). As summarized
in Fig. 3C, methylation
of G
residues at -190 and -185 bp on the noncoding
:
Go
C
C
T
C
Go
T
Gx
Ax
Ax
-190
-185
_
C
T
C
Gx
**I/
-183
-182
C
Ao
-181
T
C
C
T
T
C
Gx
Gx
‘.’
‘: .e.&
\
., \
Go
Ao
C
~;*‘
<,u;
T
..i i’
\ .~ c
9
\
&w
,.
C
G o -176
Ao -177
Go -178
Ao
-179
Go
-180
T
Go
C
A
Gx
Gx
SSBP
Gx
Non coding
Ax
‘_’
.._
Gx
TTF-1
TTF-1
Coding
Non coding
C
TTF-l-+
-1T4
Coding
xx
0
00000
-1y4
-169
I
CAAGCGGAGCACTTGAGAGCCTCTCC
Non coding
TTF-l+
SSBP-,
GTTCGCCTCGTGAACTCTCGGAGAGG
x
x 000
o
0
xxx
xx
xx
Fig. 3. Methylation
Interference
Analyses
of the Complex
Formed
between
SSBPs or lTF-1
in FRTL-5 Cell Extracts
and the
TSHR Minimal Promoter
Region between
-194 and -169
bp
In panel A the 32P-labeled,
dimethyl
sulfate-modified,
noncoding
strand spanning
-194 to -169 bp and in panel B the dimethyl
sulfate-modified,
double-stranded
oligonucleotide
spanning
-194 to -169 bp 32P-labeled
in either the noncoding
(left) or coding
(right) strand were incubated
with 20 pg nuclear extract
from FRTL-5 cells maintained
for 7 days with no TSH. Protein-DNA
complexes
were resolved
on a preparative
native gel as described
in Materials and Methods. The bands corresponding
to bound
complexes
and free DNA were eluted from the gel and cleaved at the modified
residues.
The cleavage
products
were resolved
by electrophoresis
through
a 12% denaturing
gel and visualized
by autoradiography.
The sequence
of the bases is indicated
at
the right; numbering
of nucleotides
is relative to the ATG start codon which is defined as + 1. Open circles define the bases whose
modification
reduces the proportion
of DNA in the bound fraction;
x defines bases whose methylation
does not alter the protein
binding. In panel C the sequences
of both strands are summarized,
together
with those bases whose methylation
interferes
with
the binding of SSBPs or TTF-1. Open circles define the bases whose modification
reduces
the proportion
of DNA in the bound
fraction;
x defines bases whose methylation
does not alter protein binding.
SSBPs
and TSHR
Gene
Expression
531
A
-if”
-194
I OQOOOOOO
CAAG$GGA+ACTTGAGAGCCTCTCC
NSl
------A-A-G-CA-T-C-A------
NS2
----T-CT-G----------------
-179
----T-CT-GG-CA-T-C-A------
NS1+2
TTF-l--r
SSBPv
FRTL-5
E
in
BRL
i;,
d
Fig. 4. Effects of Mutations
in the SSBP- and TTF-l-Binding
Sites on the Ability of Proteins
in FRTL-5 Cell Nuclear
Extracts
To Form Complexes
with TSHR DNA or on the CAT Activity
TSHR Promoter-CAT
Chimeras
Transiently
Expressed
in FRTL-5
or BRL Cells
Mutations
are denoted
in panel A by comparison
to the sequence
of wild type promoter
and by comparison
to bases at which
methylation
interferes
with the binding of TTF-1 on both strands (open circles) or of SSBPs on the noncoding
strand (filled circles).
Mutations
are noted as coding strand changes;
appropriate
noncoding
strand mutations
which complemented
these were also
made. In panels B and C, the double-stranded
and noncoding
strand of the oligonucleotide
spanning
-194 to -169 bp, or their
mutated
counterparts,
were, respectively,
used as the radiolabeled
probe. Each was incubated
with 1 pg nuclear extract
from
FRTL-5 cells maintained
for 7 days without
TSH. In panels D and E, FRTL-5 and BRL cells, respectively,
were transfected
with
pTRCAT5’-199,
the pTRCAT5’-199
mutants
as indicated,
pTRCAT5’-177,
or a p8CAT control. CAT activity is presented
relative
to that in the p8CAT control whose activity is arbitrarily
set at unity. All cells were cotransfected
with pSVGH and conversion
rates
were normalized
to GH levels. In the right portion
of panel E, BRL cells were also cotransfected
with or without an expression
vector containing
lTF-1,
Rc/CMV-THA,
or its control plasmid,
pRc/CMV,
In this panel, CAT activities
are presented
as the ratio
of activity
in the presence
or absence
of TTF-1 cotransfection.
Values are presented
as the mean + SEfrom three separate
experiments.
In panel D, one star @) denotes
a statistically
significant
decrease
(f < 0.05 or better) relative to the activity of
pTRCAT5’-199
but a significant
level (P c 0.05 or better) of activity
by comparison
to pTRCATS’-177;
two stars k) denotes
a
statistically
significant
decrease
(P < 0.05 or better) relative to the activity of pTRCATS’-199
but no significant
difference
in activity
by comparison
to pTRCAT5’-177.
In panel E one star (+) denotes
a statistically
significant
decrease
(P < 0.05 or better) relative
to the activity
of pTRCAT5’-199.
line, open circles) and by methylation
of nucleotides
at
-186
bp and between
-180
through
-176
bp on the
coding
strand
(Fig. 3C, top line, open circles). Thus, on
the noncoding
strand,
bases
involved
in the formation
of complexes
with SSBPs
and lTF-I are contiguous
and not overlapped.
On the coding
strand,
the lTF-1
site, the G residue
at -186
bp is located
between
the
residues
interfering
with the binding
of the SSBPs
on
the noncoding
strand,
- 190 and -185
bp; this overlap
does not, however,
appear
to be functionally
relevant,
MOL
532
END0
1995
since the SSBPs are noncoding strand-specific
interaction and are not competed for by TTF-I
site oligonucleotides
from the TG promoter.
Vol 9 No. 5
in their
binding
The SSBP and TTF-1 Elements Are Both
Required
for Full Expression
of the TSHR
To determine the functional contribution
and relationship of the SSBP and TTF-1 elements, we mutated
both in the pTRCAT5’-199
chimeric promoter construct and in oligonucleotides
used for electrophoretic
mobility shift assays (EMSA) (Fig. 4A). Nonsense mutation 1 (NSl) is a mutation identical to the pTRCAT5’199NS construct in our previous report (20). The mutation involves nucleotides reacting only with TTF-1 by
methylation analysis (Fig. 4A, open circles) and eliminates TTF-1 activity, as well as protein complex formation (20). Nonsense mutation 2 (NS2) changes both
residues of the SSBP-binding
site which were identified as contact points on the noncoding
strand by
methylation interference (Fig. 4A, dark circles vs. Fig.
3). NS2 also makes two changes at - 188 and - 187
bp which would be 5’ and contiguous
with residues
identified as contact points on the coding strand portion of the double-stranded
TTF-I site (Fig. 3) and in
the consensus TTF-1 -specific site defined experimentally and by sequence comparisons
(20, 22, 23). Nonsense mutation 1+2 (NS1+2)
is composed
of NSl
plus NS2 mutations.
The TTF-1ITSHR
DNA complex formed by the radiolabeled
double-stranded,
wild type oligonucleotide, -194 to -169 bp, and FRTL-5 cell extracts (Fig.
4B, first lane), is no longer evident when radiolabeled,
double-stranded
NSl is probe (Fig. 4B, second lane).
The TTF-l/TSHR
DNA complex, confirmed by inhibition with oligo A of the TG promoter (data not shown)
is, in contrast,
formed with radiolabeled
doublestranded NS2 (Fig. 4B, third lane). The lower level of
complex formation with radiolabeled
double-stranded
NS2 may be accounted for by the mutations on both
strands at -185 bp, which would be next to the 5’terminus of the TTF-1 binding region (Fig. 4A). In contrast, radiolabeled
NS2 (Fig. 4C, lane 3), not NSl (Fig.
4C, lane 2), does not form the SSBP/TSHR
complex
when compared with the radiolabeled
wild type noncoding strand probe (Fig. 4C, first lane). The NS1+2
mutation abolished
formation
of both protein/DNA
complexes
(Fig. 4, B and C, last lanes). These data,
together with the data in Figs. 2 and 3, support the
conclusion
that the SSBPs and TTF-1 interact with
contiguous,
but distinct sites within -194 to -169 bp
of the minimal TSHR promoter.
In FRTL-5 cells, the CAT activity expressed
by
pTRCAT5’-199NS1,
which no longer binds TTF-1 but
still binds SSBPs (Fig. 4, B and C), is halved by comparison to the activity of wild type pTRCAT5’-199
(Fig.
4D, second vs. first bar). Activity was not, however,
reduced to the level of pTRCATS’-177
(Fig. 4D, fifth
bar), which is missing both TTF-1 and SSBP binding
sites; nor was it reduced to the activity level of the
p8CAT control (Fig. 4D, last bar). The activity of
pTRCAT5’-199NS2,
which does not bind SSBPs but
does retain binding activity to lTF-1, was also halved
by comparison
to the activity exhibited by wild type
pTRCAT5’-199
(Fig. 4D, third vs. first bar), but was not,
again, reduced to the level of pTRCAT5’-177
(Fig. 4D,
fifth bar) nor to the level of the p8CAT control (Fig. 4D,
last bar). In contrast, the CAT activity of pTRCAT5’199NS1+2,
missing both the lTF-1 and SSBP binding
sites, was identical to pTRCAT5’-177
(Fig. 4D), which
deletes both sites. These data (Fig. 4) support the
conclusion
that the SSBP and TTF-1 elements are
both required for full, constitutive
expression
of the
TSHR promoter in FRTL-5 thyroid cells. Each contributes to about half the activity necessary for maximal
TSHR gene expression; their activity is additive, and
their actions are independent.
In BRL cells that express no TTF-1 (20) the CAT
activity of pTRCAT5’-199
is 2-fold
higher than
pTRCAT5’-177
when cotransfected
with TTF-1 (Fig.
4E, right panel, first vs. fifth bar). The activity of
pTRCAT5’-199NS2,
which still binds TTF-1 but not
SSBPs, is also 2-fold higher than pTRCAT5’-177
when
cotransfected
with TTF-1 (Fig. 4E, rightpanel,
third vs.
fifth bar). In contrast,
the activities of pTRCATS’199NS1, which does not bind TTF-1 but still binds
SSBPs, is unchanged
by cotransfection
with TTF-1
(Fig. 4E, right pane/, second vs. fifth bar). These data
establish that the SSBP binding site is not involved in
TTF-1 -dependent
TSHR gene expression, confirming
its independent
mode of action in FRTL-5 thyroid cells.
In BRL cells, the activity of pTRCAT5’-199
is slightly
but significantly greater (P < 0.05) than the activity of
pTRCAT5’-177
in the absence of TTF-1 (Fig. 4E, left
pane/, first vs. fifth bar). The activity of pTRCAT5’199NS1, which still binds the SSBPs but not lTF-1, is
also significantly greater (P < 0.05) than the activity of
pTRCAT5’-177
in the absence of TTF-1 (Fig. 4E, leff
pane/, second vs. fifth bar). In contrast, pTRCATS’199NS2 and pTRCAT5’-199NS1+2,
neither of which
can bind the SSBPs, expressed the same activities as
pTRCAT5’-177
(Fig. 4E, left pane/, third and fourth
bars, respectively, vs. fifth bar). These last results indicate that the SSBP element plays a role in TSHR
gene expression
in nonthyroid cells and can explain
the positive enhancer activity of this region of the
TSHR promoter in cells with no TTF-1 (20).
The weak enhancer
activity associated
with the
SSBP element in BRL, by comparison
to FRTL-5 cells,
is consistent with the lesser amounts of SSBP/TSHR
DNA complex formation in the BRL vs. FRTL-5 cell
extracts (Fig. 1 B). Thus, levels of complex formation in
BRL cell extracts are less than 20% those in FRTL-5
cell extracts (Fig. 1 B, lanes 4 vs. 1). The small SSBPdependent
functional effect, comparing the mean activities of pTRCAT5’-199
and pTRCAT5’-199NSl
vs.
pTRCAT5’-177,
is in that range when compared with
FRTL-5 cell activity (Fig. 4E vs. 4D).
SSBPs
and TSHR
Gene
Expression
533
The SSBPs and lTF-1 Sites are Involved
.cAMP Negative Regulation
of the TSHR
in TSH/
SSBP/TSHR
DNA complex formation was decreased
in FRTL-5 cells treated with TSH for 6 days (Fig. lA,
lane 2 vs. 1). The TSH effect is evident 8 and 24 h after
treatment (Fig. 5, lane 1 vs. 2; Table 1) and is CAMP
mediated, since it is duplicated
by forskolin (Fig. 5,
lanes 3 vs. 4), as well as cholera toxin or (Bu),cAMP
(data not shown). Forskolin treatment of FRT and BRL
cells did not cause a decrease in the SSBP/DNA complex formation (Fig. 5, lanes 5 to 8), i.e. the TSH/cAMPinduced decrease is FRTL-5 cell specific, as is the
decrease
in TSHR RNA or receptor levels (5) and
TTF-VTSHR
DNA complex formation (20).
The TSH/cAMP-induced
decrease
in SSBPIDNA
complex formation in FRTL-5 cells mimics the TSH/
CAMP-induced
decrease in TSHR gene expression (5) in
two important additional aspects. First, the decrease is
dependent on the presence of insulin and/or serum in the
medium (Fig. 6A). Second, it is prevented by cycloheximide (Fig. 6B), as is TTF-l/TSHR
DNA complex formation (20). Finally, TSH and forskolin decrease the promoter activity of pTRCAT5’-199NS1,
which binds the
SSBPs but not lTF-1, and pTRCAT5’-NS2,
which binds
TTF-1 but not the SSBPs (Table 2). In each case, the
decrease is to the level of the TSH/forskolin
effect on
pTRCAT!Y-199,
pTRCAT5’199NS1+2,
and pTRCAT5’177 (Table 2). Thus, the SSBP, as well as the TTF-I
element, appears to contribute to the TSWcAMP-induced decrease in activity measurable in the region between -199 and -177 bp.
In sum, the TSH/cAMP-induced
decrease in SSBPI
TSHR DNA complex formation
coincides
with the
FRTL-5
FRT
BRL
-nn
z z 1 z 1 z 1 z
1
2
3
4
5
6
7
8
Fig. 5. Effect of TSH or 10 PM Forskolin
on SSBP/TSHR
DNA Complex
Formation
in FRTL-5,
FRT, or BRL Cells
FRTL-5
cells were maintained
7 days with no TSH, then
rechallenged
for 24 h with 1 X lo-”
M TSH or 10 PM forskolin. FRT and BRL cells were cultured
in their appropriate
media, then incubated
with or without
forskolin
for 24 h.
Nuclear extracts
were prepared
from each of the cell populations and evaluated
for their ability to form SSBP/TSHR
DNA complexes
using a radiolabeled,
single-stranded
oligonucleotide
identical to the noncoding
strand of - 194 to - 169
bp of the TSHR gene as probe. The data represent
a typical
EMSA experiment
that was repeated
with the same results
three times, with different
extracts,
on different
days.
TABLE
1. Effect of 1 X 10-l’
M TSH on TTF-l/TSHR
DNA
Complex
Formation
or SSBPKSHR
DNA Complex
Formation
as a Function
of Time
lTF-I/TSHR
DNA
SSBP/TSHR DNA
Time after TSH
Complex (% of zero
Complex (% of zero
addition (h)
time control)
time control)
0
2
8
24
100
119 2 5a
81 -t-6’
86 t 8c
100
93 k 5
86 IT 6’
60 + 4”
Cells were maintained
7 days with no TSH in the medium then
rechallenged
with 1 x lo-”
M TSH for the noted
times.
Nuclear
extracts
were prepared
from the cells and equal
amounts
were incubated,
in the absence
or presence
of a
250-fold
excess of unlabeled
oligonucleotide,
with the radiolabeled
double-stranded
oligonucleotide
probe
spanning
-194
to -169
bp to measure
TTF-l/TSHR
DNA complex
formation
(see Fig. 4B) or the radiolabeled
noncoding
strand
probe spanning
-194
to -169
bp to measure
SSBP/TSHR
DNA complex
formation
(see Fig. 4C). Quantification
of complex formation
was determined
by cutting the labeled bands
from dried gels and counting
the radioactivity;
an area of the
gel with no band served as a control value, which was subtracted
in each case. The amount
of specific complex
at 0
time, i.e. in 5H medium
just after TSH addition,
was set at
100%; the amounts
at other times were calculated
based on
this value. Values are expressed
as a percent of that control
and are the mean + SE of three separate
experiments
with
different
batches of cells.
a Statistically
significant
increase,
P < 0.05.
b Statistically
significant
decrease,
P < 0.05.
c Statistically
significant
decrease,
P < 0.01.
TSH/cAMP-induced
decrease
in TTF-l/TSHR
DNA
binding (Table 1) and has properties in common with
the TSH/cAMP-induced
decrease in TSHR (Figs. 5
and 6; Ref. 5). Since full expression of the TSHR requires
lTF-1 and the SSBPs to bind to their respective
elements on the TSHR promoter, it is reasonable to
conclude that decreases in both these interactions, as
a result of TSH/cAMP
treatment and as evidenced by
decreased
complex formation,
result in decreased
TSHR gene expression (Table 2).
Unlike TTF-l/TSHR
DNA complex formation, there
is no increase in SSBP/TSHR
DNA complex formation
2 h after TSH treatment of FRTL-5 cells (Table 1).
Further, protein kinase A (PKA) treatment of the extracts does not increase SSBP/TSHR
DNA complex
formation (data not shown), as it does lTF-l/TSHR
(20) or TG DNA complex formation (24, 25) an effect
associated with increased TSHR or TG gene expression (20, 24, 25). These data would suggest that TSH/
CAMP modulation of the SSBP/lSHR
DNA interaction
is only involved in TSH/cAMP-induced
negative and
not positive autoregulation
of the TSHR gene.
The Same SSBP Site Exists Contiguous
5’ to an Upstream lTF-1 Site
with
and
An upstream lTF-1 element, -881 to -866 bp, of the
TSHR 5’-flanking
region (23) acts as an enhancer of
MOL END0 .1995
Vol 9 No. 5
534
Insulin + Serum
-
-
+
+
Cycloheximide
-
-
+
+
TSH
-
+
-
+
Forskolin
-
+
-
+
1 2 3 4
1
2
3
4
Effect of Insulin and Serum (A) or 10 PM Cycloheximide (B) on TSH- or Forskolin-Induced Decreases in SSBP/TSHR DNA
Complex Formation
A radiolabeled, synthetic single-stranded oligonucleotide identical to -194 to -169 bp of the noncoding strand of the TSHR
gene was incubated with nuclear extracts from FRTL-5 cells. In panel A, nuclear extracts were prepared from cells cultured for
7 days without TSH and without insulin and in 0.2% serum (-) or with 10 pg/ml insulin and in 5% serum (+). Cells were then
treated with (+) or without (-) 1 x IO-” M TSH for 24 h as noted. In panel B, FRTL-5 cells were maintained in 5H medium with
no TSH for 7 days and then challenged with 10 ~,LM forskolin which duplicates the ability of TSH to decrease the SSBP/TSHR DNA
complex formation (see Fig. 5). In addition, 10 PM cycloheximide were added alone or with the forskolin. After 24 h, the nuclear
extracts were prepared from each population of cells and were incubated with the probe as noted in Materials and Methods.
Cycloheximide at these concentrations and under these conditions inhibits protein synthesis more than 95% in FRTL-5 cells and
prevents down-regulation of TSHR RNA levels by TSH or its CAMP signal (5). The data represent a typical experiment that was
repeated with the same results three times, with different batches of cells, on different days.
Fig.
6.
2. Effect of 1 x lo-” M TSH or 10 PM Forskolin
Treatment of FRTL-5 Cells on CAT Activity of the NSl,
NS2, and NSl + 2 Promoter-CAT Mutants (Fig. 4) by
Comoarison to oTRCAT5’-199 or oTRCATS’-177
TABLE
CAT activity
Promoter
construct
pTRCATS’-199
NSI mutation
NS2 mutation
NSl + 2
mutation
oTRCATS’-177
No addition
14.1 2 1.0
relative
to p8CAT
+TSH
(1 x 10-10
M)
control
+Forskolin
(10 IL@
2.1
1.5
1.6
1.5
Ifr 0.7
t 0.3
? 0.4
t 0.3
2.6 t 0.7
7.0 k 0.8
6.5 L 1.2
2.8 I? 0.5
2.9 ‘-’ 0.3
1.6 ? 0.3
1.8 t 0.3
1.8 2 0.5
1.4 t 0.4
2.1 ‘-’ 0.7
FRTL-5 cells were grown to 80% confluency in 6H medium,
shifted to 5H medium (no TSH) for 4 days, then returned to 6H
medium 1 day before transfection with the noted chimeras
and for 16 h thereafter. At that time fresh 5H medium was
added alone (No addition), plus 1 x 109” M TSH, or plus 10
PM forskolin. Cells were maintained
in culture for 60 additional hours before CAT activity was measured. CAT activity
is presented relative to that in the p8CAT control whose
activity is arbitrarily at unity (see Fig. 4D). All cells were
cotransfected with pSVGH and conversion rates were normalized to GH levels.
inhibited
by the noncoding
strand containing
the
downstream
SSBP bind element (Fig. 7A).
Comparing
sequences
and data from methylation
interference analyses, the same two nucleotides identified as important SSBP contact points for binding
and activity are conserved 5’ and contiguous
to the
TTF-1 sites in both loci (Fig. 7B, dark circles). We
mutated both residues on the noncoding
strand of the
upstream and downstream
oligonucleotides
(Fig. 8A,
underlined
or bold). As exemplified
using the mutant
upstream oligonucleotide
as radiolabeled
probe (Fig.
8B), complex formation was markedly reduced. Competition experiments
were confirmatory;
thus, as exemplified using radiolabeled,
wild type, upstream oligonucleotide
as probe (Fig. 8C), competition
by the
mutated downstream
sequence was significantly decreased (Fig. 8C, lanes 5-8) by comparison
to competition by an oligonucleotide
with the downstream
SSBP site intact (Fig. 8C, lanes l-4).
The association
of the SSBP and TTF-1 binding
elements in both loci of the TSHR promoter is not likely
to be random. This strengthens the concept that they
are functionally associated.
DISCUSSION
the TTF-1 element in the minimal TSHR promoter and
functions conjointly with it in thyroid-specific
expression and TSH/cAMP
autoregulation.
Using the noncoding strand of the -194 to - 169 bp region containing the downstream
(DS) SSBP and l-l-F-1 region as
radiolabeled
probe, we found the unlabeled noncoding strand of the upstream (US) region, -886 to -858
bp, prevented formation of the SSBP/TSHR
DNA complex (Fig. 7A). When the noncoding
strand of the upstream region, -886 to -858 bp, was used as probe,
it formed a complex that was self-inhibited
and was
We have identified a minimal TSHR promoter, -220 to
-39 bp, exhibiting thyroid-specific
expression of the
gene (16-21). A double-stranded
sequence,
- 186 to
- 176 bp, interacts with lTF-1 and has been identified
as a positive regulatory element that contributes
to
“thyroid-specific”
expression
of the
constitutive,
TSHR in FRTL-5 thyroid cells (20, 22, 23). In the
present study, we show that the noncoding
strand of
this region of the minimal TSHR promoter
interacts
with nuclear proteins other than lTF-1, two of which
SSBPs
and TSHR
Gene
A
535
Expression
A.
non coding strand
Probe
Competitor
T-=-l(-1
.
MUT lJ?‘ST NC
us
IX
US
(-)
US
UPST NC
DS
CTTGTTACACGCTGAATTCACGAGAGAAG
-Eat3
CTTGTTGCACGGTGAATTCACGAGAGAAG
DNST NC
MUT DNST NC
1
2
3
4
5
-858
GTTCG-CCTCG-TGAACTCTCGGAGAGG
-194
-169
GTTCACCTCCTGAACTCTCGGAGAGG
6
B.
C.
Competition
by Downstream
SSBP Site
-194
0
00 oooooOm
-169
CyG$GAGCACTTGAGAGCCTCTCC
IIIII
II
II
US-‘ITF-1
site GAACqACGTGCCACTTAAGTG;CTCTCTTC
-858
-886
Fig. 7. Competition
and Methylation
Interference
Analyses
Indicate
that SSBPs
Bind the Comparable
Element
5’ and
Contiguous
to the Upstream
TTF-1 Site at -881 to -866
bp
of the 5’-Flanking
Region of the TSHR
In panel A, a radiolabeled,
synthetic
single-stranded
oligonucleotide
identical
to the noncoding
strand spanning
-194
to -169
(lanes l-3) or -886 to -858
bp (lanes 4-6) of the
TSHR gene was used as probe and incubated
with nuclear
extracts
from FRTL-5 cells. These contain,
respectively,
the
downstream
(DS) and upstream
(US) TTF-1 sites of the 5’flanking
region of the TSHR (20, 22, 23). Nuclear
extracts
were prepared
from FRTL-5 cells cultured
in 5H medium
for
6 days. In lanes 2 and 6, a 250-fold
excess of the unlabeled
noncoding
strand of the DS oligonucleotide
was present
in
the incubation
mixture;
in lanes 3 and 5, a 250-fold
excess of
the unlabeled
noncoding
strand of the US oligonucleotide
was present.
The arrows denote the SSBP complexes
(Figs.
1 and 2). In panel B the sequences
of the coding strands
of
the DS and US oligonucleotides
are noted; identities
in sequence are noted by solid lines. Closed circles define bases
whose methylation
reduces
SSBP binding,
determined
by
methylation
interference
analyses
as described
in Fig. 3.
Open circles define the bases in the US oligonucleotide
whose methylation
reduces the proportion
of TTF-I reactive
DNA in the bound fraction.
DS-TIT-1
VW Type
5 2050
0
B
0
Mutant
52050
site
appear to be ubiquitously
expressed,
albeit at much
higher levels in FRTL-5 thyroid cells. These proteins
bind to an element on the noncoding
strand which is
contiguous
with, and 5’ to, the TTF-1 element. The
association of the lTF-1 and SSBP binding elements
does not appear to be random. This is evidenced by
data showing that an SSBP site is 5’ and contiguous to
an upstream, as well as the downstream,
TTF-1 site;
the upstream SSBP site binds the same SSBPs, as
evidenced by mutual competition
of the downstream
and upstream SSBP regions and the loss of crosscompetition
after mutation of only two conserved
residues.
In FRTL-5 thyroid cells, the SSBP and lTF-1 function independently
and additively to achieve full, constitutive expression of the TSHR gene. Thus, deletion
of each element results in about half of the maximal
TSHR gene expression;
deletion of both results in a
B lj
P =
Probe = Upstream SSBP Site
Fig. 8. Effects of Mutations
in the Conserved
Residues
of
the SSBP-binding
Sites Contiguous
with the Upstream
and
Downstream
TTF-1 Sites (A) on the Ability of Proteins
in
FRTL-5 Cell Nuclear Extracts To Form Complexes
with TSHR
DNA
Mutations
of the noncoding
(NC) strand are denoted
as
underlined
or bold residues in panel A, by comparison
to the
sequence
of wild type promoter..
In panel B the noncoding
strand
of the wild type oligonucleotide
spanning
-886
to
-858
bp, and its mutant
counterpart,
were used as the
radiolabeled
probes;
In panel C only the wild type oligonucleotide spanning
-886 to -858 bp is probe. In panels Band
C, probe
was incubated
with 1 pg nuclear
extract
from
FRTL-5 cells maintained
for 7 days without TSH. In panel C,
the noted fold-excess
of wild type or mutant
downstream
-194
to -169
bp, is included
as a
noncoding
strand,
competitor.
decrease to levels close to the control p8CAT chimera.
Unlike the TTF-1 element, the SSBP element is involved in TSHR gene expression in nonthyroid cells.
Thus, mutation of the SSBP-binding
element in TSHR
promoter-CAT
chimeras and transfection
into BRL
cells results in the loss of promoter
activity that is
independent
of l-l-F-1. The low level of SSBP-related
gene expression
in BRL cells is consistent with the
lesser amounts of SSBP complex formation in these
cells. The SSBPs may, therefore, contribute
to low
levels of TSHR expression
in tissues without TTF-1,
together with the house-keeping
feature of the TSHR
promoter, i.e. being GC-rich with multiple transcription
start sites (16). This may be important in retroorbital
tissues where TSHR RNA has been identified (9),
where functional expression
of the TSHR has been
MOL
536
END0
1995
implicated in exophthalmos
(28, 29) and where TTF-I
has not been demonstrated
(30).
In FRTL-5 thyroid cells, TSH/cAMP
autoregulates
TSHR gene expression; within the first few hours there
is up-regulation;
however, down-regulation
is apparent by 8 h and maximal within 24 (5). Understanding
the mechanism of biphasic modulation
of TSHR RNA
via the CAMP signal is important, because it is a feature of gonadotropin
as well as TSHR regulation and is
suggested to be important in the mechanism by which
a single receptor of this family modulates both growth
and differentiation
(31-36). In our previous reports (20,
23) we suggested that the TTF-1 site was involved in
TSH/cAMP-induced
negative regulation of the TSHR.
Thus, TSH/cAMP
decreased lTF-1 mRNA levels and
TTF-l/TSHR
complex formation coincidentally
as a
function of time and TSH concentration;
further, the
decreases
were cycloheximide-sensitive,
as was
TSHR gene expression. We showed (20, 23) that one
explanation
for the absence of coincident
negative
regulation
of the TG gene was a compensatory
increase in Pax-8 complex formation with TG, but not
TSHR, TTF-1 sites.
In the present report we show that TSH decreases
the association of SSBPs, as well as TTF-1, with the
minimal TSHR promoter, as evidenced by decreased
complex formation. TSH decreases TSHR RNA levels,
TTF-l/TSHR
DNA, and SSBP/TSHR
DNA complex
formation similarly with respect to time (Refs. 5, 20,
and 23; this report); the decrease in each case is
mimicked by forskolin and is prevented by cycloheximide (Refs, $20, and 23; this report). LikeTSHR gene
expression (5), the decrease in SSBP/TSHR DNA complex formation requires the presence of insulin, serum,
or insulin-like growth factor-l in the medium.
Since the binding of both TTF-I and SSBPs to their
respective elements is associated with full, constitutive TSHR gene expression
in the absence of TSH/
CAMP, we suggest that TSH-induced
decreases in the
association
of each contributes
to the TSH-induced
decrease in activity measurable in the region between
~ 199 and - 177 bp. An additive, positive effect of the
SSBP and TTF-1 elements for full, constitutive expression of the TSHR does not contradict their additive role
in TSH/cAMP-induced
negative regulation; this result
is not discrepant, but rather complementary.
Thus, the
binding of both factors, SSBPs and TTF-1, to their
respective elements is necessary for full expression; a
decrease in binding of both factors, induced by TSH/
CAMP, would result in a decrease in TSHR gene expression from the maximal and would be expected to
be additive and independent,
as evidenced in Table 2.
In our previous report (20), we showed that TSH/
CAMP-increased,
PKA-dependent
TTF-l/DNA
complex formation was involved in positive regulation of
TSHR gene expression,
as is the case for increased
TG gene expression (24,25). TSH treatment of FRTL-5
cells does not similarly increase SSBP/TSHR
DNA
complex formation nor does PKA treatment
of the
extracts. Thus, SSBPs do not appear to contribute to
Vol 9 No. 5
TSH/cAMP-induced
positive TSHR gene regulation.
This strengthens
the hypothesis
that TSH/cAMP-induced positive and negative regulation of TSHR gene
expression
involve different mechanisms,
albeit the
same factors, and correlates with measurements
of
complex formation (5, 20, 23).
Three TTF-1 binding sites within 170 bp of the rat TG
gene transcriptional
start site are important for TG
promoter
activity (12-15, 26). The upstream TTF-1
binding site (A region) interacts with ubiquitous factor
(UFA) and the downstream
(C region) with Pax-8; both
are necessary for full TG promoter activity (14, 26).
Since SSBP/TSHR DNA complex formation is not prevented by oligo A or oligo C from these TG regions, the
SSBPs are distinct from UFA or Pax-8 and are not
associated
with TTF-1 sites on the TG promoter.
SSBP binding to SSBP elements on the TSHR promoter, therefore, contributes
to the difference
between the TSHR and TG gene expression
and the
different roles that TTF-1 plays in each gene. The
existence of SSBP elements on the noncoding strand,
contiguous
and 5’ to the upstream and downstream
TTF-1 sites in the TSHR promoter, strengthens
the
likely importance
of their role in TSHR, but not TG,
expression.
The SSBP site appears to be a discrete element
between the TTF-1 and insulin response element (IRE)
of the TSHR. The TSHR IRE is upstream of the TTF-1
site and dependent on it (21). Thus, the NSl mutation,
in constructs with both sites, -220 to -169 bp, eliminates insulin activity as well as TTF-1 binding and
function; there is a concurrent loss of the TTF-1 footprint from -186 to -176 bp and the IRE footprint,
from -217 to -210 bp (21). DNAase I digestion reveals, however, that a protected region remains, - 197
to -186 bp, between the TTF-1 and IR elements (21).
The NSl mutation in this construct does not eliminate
its ability to bind SSBPs as evidenced
in EMSA (Y.
Shimura, H. Shimura, M. Ohmori, M. Ohta, and L. D.
Kohn, unpublished
observations).
Further, the oligionucleotide
containing
the TSHR IRE, -220 to -188
bp, can bind SSBPs (21); SSBP binding remains despite mutations that eliminate
IRE binding and the
insulin response (21); and a single strand form of this
mutated oligonucleotide,
-220 to - 188 bp, can compete for the binding of the SSBPs interacting with the
noncoding
strand of the downstream
oligonucleotide,
-194 to -169 bp (Y. Shimura, H. Shimura, M. Ohmori,
M. Ohta, and L. D. Kohn, unpublished
observations).
These data are consistent with the existence of a site
between the TTF-1 and IRE elements of the TSHR
which binds SSBPs and can be detected in footprints
of the NSl mutant containing both sites. It is of interest
that a GXXXXG motif exists at - 196 to - 191 bp of the
coding strand, in addition to the GXXXXG
motif at
-190 to -185 bp linked to the SSBP site on the
noncoding strand. Of additional interest, it is the coding strand of the IRE-containing
oligonucleotide
from
-220 to -188 bp that inhibits the SSBP binding to the
noncoding strand of the oligonucleotide
from - 194 to
SSBPs
and TSHR
Gene
Expression
- 169 bp. The relationship
of the two GXXXXG
sites to
SSPB
binding
is under
investigation
as is the relation
of the function
of the SSBP
site and the IRE.
There
is increasing
evidence
that sequence-specific, single strand
DNA binding
proteins
are important
in gene expression.
Thus, similar
factors
have recently
been
reported
to be important
in the transcriptional
regulation
of c-myc (37), 3-hydroxy-3-methylglutatyl
coenzyme
A reductase
(38), tyrosine
aminotransferase
(39), actin (40, 41), adipsin
(42), p-casein
(43), plateletderived
growth
factor
A-chain
(44), and the GH (45)
genes.
In addition,
it has been
reported
that the estrogen
receptor
preferentially
binds
to the coding
strand
of its response
element
(46). Some
sequencespecific
single strand
DNA binding
proteins
can exhibit
enhancer
(39) and others
repressor
(38, 43, 45) activities. Some
also recognize
double-stranded
DNA (37,
40, 45), whereas
others,
like the SSBPs
herein,
bind
weakly
to double
strand
DNA. In the present
report,
we
describe
a role for SSBPs
in the TSHR
promoter
and
show,
as in the case of the p-casein
gene promoter,
that the activity
of the SSBPs
is hormonally
regulated.
The relationship
of the SSBPs
binding
to the TSHR
to
the SSBPs
interacting
with the above
genes
is unknown
at this time; definitive
comparisons
must await
cloning
of the SSBPs
interacting
with the SSBP
element on the TSHR which we have characterized
in this
report.
Similarly,
the relationship
of SSBPs
in FRTL-5
thyroid
and nonthyroid
cells, as well as the role of the
SSBPs
in expression
of the TSHR
in nonthyroid
cells,
is incompletely
defined
and
requires
further
investigation.
MATERIALS
AND METHODS
Materials
Bovine TSH (NIDDK-bTSH-l-l,
30 U/mg), the TTF-1 expression vector,
RcCMV-THA
(14) (the gift of Dr. Roberto
Di
Lauro,
Stazione
Zoologica
A. Dohrn,
Villa Communale,
Naples, Italy, 80121), and all other materials
were the same
as previously
detailed
(16-21,
23).
Cells
BRL cells (BRL 3A, ATCC No. CRL 1442) were grown
in
Ham’s F-12 supplemented
with 5% fetal calf serum (Biofluids, Rockville,
MD). FRTL-5 (ATCC No. CRL 8305) (47) and
FRT (48) rat thyroid
cells (16-21,
23) were grown in Coon’s
modified
Ham’s F-12 containing
5% calf serum (GIBCOIBRL
Life Technology,
Gaithersburg,
MD),
1 mM nonessential
amino acids (Microbiological
Associates,
Bethesda,
MD),
and, in the case of FRTL-5 cells, a six-hormone
mixture
(6H)
as follows:
bovine TSH (1 X lo-”
M), insulin
(10 Fg/ml),
cortisol
(0.4 rig/ml), transferrin
(5 Fg/ml),
glycyl-L-histidyl-Llysine acetate (10 rig/ml), and somatostatin
(10 rig/ml) (47).
537
until near confluency;
the 5H extracts
were from cells maintained in medium
depleted
of TSH (5H) for 7 days after they
were grown to near confluency
in 6H medium.
Extracts
used
to measure
the response
to TSH or forskolin
were from
FRTL-5
cells maintained
for 7 days in 5H medium
or in
medium with no TSH, no insulin, and only 0.2% serum before
each effector
was added to the culture medium
at the concentration
and time indicated.
FRT and BRL cells were grown
until near confluency
in their appropriate
media or to 60%
confluency
before being challenged
with forskolin.
Gel Mobility
Gel mobility shift analyses
(16-21,
23) used synthesized
single- or double-stranded
oligonucleotides
that were end-labeled with [Y-~‘P]ATP
and T4 polynucleotide
kinase, then
purified on an 8% native polyacrylamide
gel. One microgram
of nuclear extract was incubated
in a 30 ~1 reaction
volume
for 20 min at room temperature,
with or without
unlabeled
competitor
oligonucleotides,
and in the following
buffer: 10
mM Tris HCI, pH 7.6, 50 mM KCI, 5 mM MgCI,,
1 mM dithiothreitol, 1 mM EDTA, 12.5% glycerol,
0.1% Triton X-100, and
1 fig poly(dl-dC).
Labeled probe, 50,000 cpm (-0.5
ng DNA),
was added
and incubated
an additional
20 min at room
temperature.
DNA-protein
complexes
were separated
on 5%
native polyactylamide
gels.
Methylation
Interference
Assays
Methylation
interference
assays
were
performed
as described (15). Single-stranded
oligonucleotide
spanning
- 194
to -169 bp was end-labeled
with [y3’P]ATP
and T4 polynucleotide
kinase and purified on an 8% native polyacrylamide
gel. To obtain double-stranded
probes with coding or noncoding strand labeled, labeled single-stranded
oligonucleotides were annealed
with their cold complementary
strand,
then purified on an 8% native polyactylamide
gel. The singleor double-stranded
probes were modified
with dimethyl
sulfate (49) for 20 min on ice. For the preparative
mobility shift,
5 x 1 O5 cpm of modified
oligonucleotides
were incubated
with 12 pg poly(dl-dC)
and 20 pg nuclear extract. The undried
gel was exposed
to autoradiography
film for 1 h, and the
regions corresponding
to the protein-DNA
complex
and unbound probe in the gel were excised,
eluted, and then precipitated.
Base elimination
and strand scission
reactions
at
adenines
and guanines
(G>A)
were performed
(49) before
samples
were lyophilized,
resuspended
in water,
relyophilized (three times), and analyzed
on a 12% sequencing
gel.
Plasmids
Chimeras,
pTRCAT5’-199
and pTRCAT5’-177,
containing
199 and 177 bp of 5’-flanking
region of the rat TSH receptor
gene, respectively,
have been described
(16, 17). To generate
chimeras
containing
mutations
in their sequences,
objective
promoter
segments
with mutations
were generated
by polymerase chain reaction (50) using 1) forward
primers that had
the mutated
sequence
with BamHl site on the 5’-end and 2)
the ON-8L
primer described
previously
as a reverse
primer
(16). Amplified
fragments
were ligated p8CAT and sequenced
(16-21,
23, 51) to ensure nucleotide
fidelity. The preparation
of pSVGH,
used to evaluate
transfection
efficiency,
has been
described
(16-21).
All plasmid
preparations
were twice purified by CsCl gradient
centrifugation
(52).
Transient
Nuclear
Shift Analyses
Expression
Analysis
Extracts
Nuclear
extracts
were prepared
exactly
as previously
described (16-21,
23). The 6H FRTL-5 cell extracts
were from
cells grown in the presence
of complete
6H medium
(+TSH)
BRL and FRTL-5
cells were transfected
by electroporation
(Gene Pulser, Bio-Rad,
Richmond,
CA) (16-21,30).
BRL cells
were used directly
after being grown
to 80% confluency.
FRTL-5 cells were grown to 80% confluency
in 6H medium,
MOL
538
END0
1995
Vol 9 No. 5
shifted to 5H medium (no TSH) for 4 days, then returned
to 6H
medium
1 day before
transfection.
Cells were harvested,
washed,
and suspended
at 1.5 x lo7 cells/ml
in 0.8 ml
Dulbecco’s
phosphate
buffered
saline, pH 7.4, without
Mg”
or Ca’ ‘. Either 35 wg pTRCAT5’-199
or equivalent
molar
amounts
of the deletion
mutants
or p8CAT (negative
control)
were added
with 2 pg pSVGH.
For experiments
involving
cotransfection
with TTF-1,
10 pg of the TTF-1 expression
vector,
RcCMV-THA,
were added
with the pTRCAT
and
pSVGH
plasmids,
as previously
described
(20). pRc/CMV
plasmid,
with no TTF-1 insert, was used as a control
in
cotransfection
experiments
with TTF-1 (20). The amounts
of
all plasmids
used in these experiments
were determined
by
preliminary
experiments
that evaluated
transfection
conditions as a function
of plasmid concentration:
they are essentially the same as used previously
(16-21,
23).
Cells were pulsed (300 V, capacitance
960 pfarads),
plated
(-6 x lo6 cell per dish) and cultured
for 72 h in the case of
FRTL-5 cells or for 48 h in the case of BRL cells. Medium
was
taken for human GH RIA (Nichols Institute,
San Juan Capistrano, CA) and cells were harvested
for CAT assays (53) which
were performed
exactly
as described
(16-21,
23).
Other
5.
6.
7.
8.
Procedures
Protein
concentration
was measured
recrystallized
BSA was the standard.
using
a Bio-Rad
kit;
Acknowledgment
Received
August
19, 1994. Re-revision
received
and accepted February
24, 1995.
Address
requests
for reprints
to: Leonard
Kohn, Department of Biochemistry/Metabolism,
National
Institute
of Diabetes and Digestive
and Kidney Diseases,
National
Institutes
of Health, Bethesda,
Maryland
20892.
*Current
address:
Third Department
of Internal Medicine,
University
of Yamanashi
Medical
School, 1110 Tamaho-cho,
Nakakomagun,
Yamanashi-ken,
409-38
Japan.
TCurrent
address:
Deparment
of Pediatrics,
University
of
Yamanashi
Medical
School, 1110 Tamaho-cho,
Nakakomagun, Yamanashi-ken,
409-38
Japan.
*Current
address:
Third Department
of Internal Medicine,
Faculty of Medicine,
Kyushu Universtiy,
3-l -1 Maidashi,
Higashi-ku,
Fukuoka
812, Japan.
9.
10.
11.
12.
13.
14.
15.
REFERENCES
16.
Kohn LD, Saji M, Akamizu
T, lkuyama
S, lsozaki 0, Kohn
AD, Santisteban
P, Chan JYC, Bellur S, Rotella CM,
Alvarez FV, Aloj SM 1990 Receptors
of the thyroid gland:
the thyrotropin
receptor
is only the first violinist
of a
symphony
orchestra.
In: Ekholm R, Kohn LD, Wollman
S
(eds) Control
of the Thyroid:
Regulation
of its Normal
Growth
and Function.
Plenum Press, New York, pp 151210
Dumont
JE, Lefort A, Libert F, Parmentier
M, Raspe E,
Reuse F, Maenhaut
C, Roger P, Corvilain
B, Laurent
E,
Mockel
J, Lamy F, Van Sande J, Vassart G 1990 Transducing systems
in the control of human thyroid cell function, proliferation,
and differentiation.
In: Ekholm R, Kohn
LD, Wollman S (eds) Control of the Thyroid:
Regulation
of
its Normal
Growth
and Function.
Plenum
Press, New
York, pp 357-372
Vassar? G, Dumont JE 1992 The thyrotropin
receptor
and
regulation
of thyrocyte
function and growth.
Endocr Rev
13:596-611
Kohn LD, Saji M, Kosugi M, Ban T, Giuliani C, Hidaka A,
Shimura
H, Shimura
Y, Okajima
F 1993 The synthesis
17.
18.
19.
20.
and secretion
of thyroid hormones:
regulation
by multiple
hormones
and signals which can be subverted
by autoantibodies
to the thyrotropin
receptor.
in: Troncone
L,
Shapiro
B, Satta MA, Monaco
F (eds) Thyroid
Diseases:
Basic Science,
Pathology,
Clinical and Laboratory
Diagnosis. CRC Press, Boca Raton, pp 59-l 18
Saji M, Akamizu
T, Sanchez
M, Obici S, Avvedimento
E,
Gottesman
ME, Kohn LD 1992 Regulation
of thyrotropin
receptor
gene expression
in rat FRTL-5
thyroid
cells.
Endocrinology
130:520-533
Santisteban
P, Kohn LD, Di Lauro R 1987 Thyroglobulin
gene expression
is regulated
by insulin and insulin-like
growth factor I, as well as thyrotropin,
in FRTL-5 thyroid
cells. J Biol Chem 262:4048-4052
lsozaki 0, Kohn LD, Kozak CA, Kimura S 1989 Thyroid
peroxidase:
rat cDNA sequence,
chromosomal
localization in mouse,
and regulation
of gene expression
by
comparison
to thyroglobulin
in rat FRTL-5 cells. Mol Endocrinol
3:1681-l
692
Akamizu
T, lkuyama
S, Saji M, Kosugi
S, Kozak
C,
McBride
OW, Kohn LD 1990 Cloning,
chromosomal
assignment,
and regulation
of the rat thyrotropin
receptor:
expression
of the gene is regulated
by thyrotropin,
agents that increase
CAMP levels, and thyroid
autoantibodies. Proc Natl Acad Sci USA 87:5677-5681
Endo T, Ohno M, Kotani S, Gunji K, Onaya T 1993 Thyrotropin
receptor
in non-thyroid
tissues.
Biochem
Biophys Res Commun
190:774-779
lnoue M, Kotani S, Endo T, Tawata
M, Onaya
T 1993
Thyrotropin
receptor
exists on clonal osteoblast-like
osteosarcoma
cells. Thyroid
(Suppl l):T-74
Francis T, Burch HB, Cai W-Y, Lukes Y, Peele M, Carr FE,
Wartofsky
L, Burman
KD 1991 Lymphocytes
express
thyrotropin
receptor-specific
mRNA as detected
by the
PCR technique.
-Thyroid
1:223-227
Musti AM. Ursini VM. Avvedimento
EV. Zimarino
V. Di
Lauro R 1987 A cell type specific factor recognizes.the
rat thyroglobulin
promoter.
Nucleic Acids Res 15:81498166
Civitareale
D, Lonigro R, Sinclair AJ, Di Lauro R 1989 A
thyroid-specific
nuclear protein essential
for tissue-specific expression
of the thyroglobulin
promoter.
EMBO J
8:2537-2542
Sinclair AJ, Lonigro R, Civitareale
D, Ghibelli L, Di Lauro
R 1990 The tissue-specific
expression
of the thyroglobulin gene requires
interaction
between
thyroid-specific
and ubiquitous
factors.
Eur J Biochem
193:31 l-31 8
Guazzi S, Price M, De Felice M, Damante
G, Mattei M-G,
Di Lauro R 1990 Thyroid nuclear factor 1 (ITF-1) contains
a homeodomain
and displays a novel DNA binding specificity. EMBO J 9:3631-3639
lkuyama
S, Niller HH, Shimura
H, Akamizu
T, Kohn LD
1992 Characterization
of the 5’-flanking
region of the rat
thyrotropin
receptor
gene. Mol Endocrinol
6:793-804
lkuyama
S, Shimura
H, Hoeffler JP, Kohn LD 1992 Role
of the cyclic adenosine
3’, 5’-monophosphate
response
element
in efficient
expression
of the rat thyrotropin
receotor oromoter.
Mol Endocrinol
6:1701-l
715
Ikuyama
S, Shimura
H, lguchi H, Kono A, Haji M, Kohn
LD, Nawata
H 1993 Thyroid
autoantigen
p70Ku
is a
transcriptional
activator
of the TSH receptor
gene. In:
Nagataki
S, Mori T, Torizuka
K (eds) 80 Years of Hashimoto’s Disease. Elsevier Science Publishers
BV, Amsterdam, pp 457-462
Shimura
H, lkuyama
S, Shimura
Y, Kohn LD 1993 The
CAMP response
element
in the rat thyrotropin
receptor
promoter:
regulation
by each decanucleotide
of a flanking tandem
repeat
uses different,
additive,
and novel
mechanisms.
J Biol Chem 268:24125-24137
Shimura
H, Okajima
F, lkuyama
S, Shimura
Y, Kimura S,
Saji M, Kohn LD 1994 Thyroid-specific
expression
and
CAMP autoregulation
of the thyrotropin
receptor
gene
involves thyroid transcription
factor-l
(TTF-1). Mol Endo-
SSBPs
and TSHR
Gene
539
Expression
crinol 8:1049-l
069
21. Shimura
Y, Shimura
H, Ohmori
M, lkuyama
S, Kohn LD
1994 Identification
of a novel insulin-response
element in
the rat thyrotropin
receptor
promoter.
J Biol Chem 269:
31908-31914
22. Civitareale
D, Castelli
MP, Falasca
P, Saiardi A 1993
Thyroid
transcription
factor 1 activates
the promoter
of
the thyrotropin
receptor
gene. Mol Endocrinol
7:15891595
23. Ohmori
M, Shimura
H, Shimura
Y, lkuyama
S, Kohn LD
1995 Characterization
of an upstream
thyroid
transcription factor-l
binding site in the rat thyrotropin
receptor
promoter.
Endocrinology
136:269-282
24. Avvedimento
EV, Musti AM, Ueffing M, Obici S, Gallo A,
Sanchez
M, DeBrasi D, Gottesman
ME 1991 Reversible
inhibition
of a thyroid-specific
trans-acting
factor by Ras.
Genes Dev 5:22-28
25. Gallo A, Benusiglio
E, Bonapace
IM, Feliciello
A, Cassano S, Garbi C, Musti AM, Gottesman
ME, Avvedimento
EV 1992 v-Ras and protein kinase C dedifferentiate
thyroid cells by down-regulating
nuclear CAMP-dependent
protein kinase A. Genes Dev 6:1621-l
630
26. Zannini M, Francis-Lang
H, Plachov
D, Di Lauro R 1992
Pax-8,
a paired domain-containing
protein,
binds to a
sequence
overlapping
the recognition
site of a homeodomain and activates
transcription
from two thyroid-specific promoters.
Mol Cell Biol 12:4230-4241
27. Santisteban
P, Acebron
A, Polycarpou-Schwarz
M, Di
Lauro
R 1992 Insulin and insulin-like
growth
factor
I
regulate
a thyroid-specific
nuclear protein that binds to
the thyroglobulin
promoter.
Mol Endocrinol
6:131 O-l 317
28. Winand
RJ. DeWolf M. Consialio
E. Kohn LD 1979 Autoimmunity;
the thyrotropin
receptor,
and exophthalmos.
In: Klein E, Horster
FA (eds) Autoimmunity
in Thyroid
Diseases.
FK Schattauer
Verlag, Stuttgart-New
York, pp
125-l 47
29 Rotella CM, Zonefrate
R, Toccafondi
R, Valente
WA,
Kohn LD 1986 Ability of monoclonal
antibodies
to the
TSH receptor
to increase
collagen
synthesis
in human
fibroblasts,
an assay which appears
to measure
exophthalmogenic
immunoglobulins
in Graves’ sera. J Clin Endocrinol
Metab 62:357-367
30 Lazzaro
D, Price M, De Felice M, Di Lauro R 1991 The
transcription
factor TTF-1 is expressed
at the onset of
thyroid and lung morphogenesis
and in restricted
regions
of the foetal brain. Development
113:1093-l
104
31 LaPolt PS, Oikawa
M, Jia X-C, Dargan
C, Hsueh AJW
1990 Gonadotropin-induced
up- and down-regulation
of
rat ovarian
LH receptor
message
levels during follicular
growth,
ovulation
and luteinization.
Endocrinology
126:
3277-3279
32. Segaloff
DL, Wang H, Richards
JS 1990 Hormonal
regulation of luteinizing
hormone/chorionic
gonadotropin
receptor mRNA in rat ovarian
cells during follicular
development
and luteinization.
Mol Endocrinol
4:1856-1865
33. Wang H, Segaloff DL, Ascoli M 1991 Lutropin/choriogonadotropin
down-regulates
its receptor
by both receptor-mediated
endocytosis
and a CAMP-dependent
reduction in receptor
mRNA. J Biol Chem 266:780-785
34. Hoffman
YM, Peegel H, Sprock MJE, Zhang Q-Y, Menon
KMJ 1991 Evidence
that human chorionic
gonadotropin/
luteinizing
hormone
receptor
down-regulation
involves
decreased
levels of receptor
messenger
ribonucleic
acid.
Endocrinology
128:388-393
35. LaPort PS, Tilly JL, Aihara T, Nishimori
K, Hsueh AJW
1992 Gonadotropin-induced
up- and down-regulation
of
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
ovarian follicle-stimulating
hormone
(FSH) receptor
gene
expression
in immature
rats: effects of pregnant
mare’s
serum
gonadotropin,
human
chorionic
gonadotropin,
and recombinant
FSH. Endocrinology
130:1289-l
295
Kohn LD, Shimura
H, Shimura
Y, Hidaka A, Giuliani C,
Napolitano
G, Ohmori
M, Laglia G, Saji M 1995 The
thyrotropin
receptor.
Vitam Horm 50:287-384
Kolluri R, Torrey TA, Kinniburgh
AJ 1992 A CT promoter
element
binding
protein:
definition
of a double-strand
and a novel single-strand
DNA binding
motif. Nucleic
Acids Res 2O:ll l-l 16
Rajavashisth
TB, Taylor AL, Andalibi
A, Svenson
KL,
Lusis AJ 1989 Identification
of a zinc finger protein that
binds to the sterol regulatory
element.
Science 245:640643
Jansen-Durr
P, Boshart
M, Lupp B, Bosserhoff
A, Frank
RW, Shutz G 1992 The rat poly pyrimidine
tract binding
protein (PTB) interacts
with a single-stranded
DNA motif
in a liver-specific
enhancer.
Nucleic Acids Res 20:12431249
Santoro IM, Yi T-M, Walsh K 1991 Identification
of singlestranded-DNA-binding
proteins that interact with muscle
gene elements.
Mol Cell Biol 11:1944-1953
Kamada
S, Miwa T 1992 A protein binding to CArG box
motifs and to single-stranded
DNA functions
as a transcriptional
repressor.
Gene 119:229-236
Wilkison
WO, Min HY, Claffey
KP, Satterberg
BL,
Spiegelman
BM 1990 Control
of the adipsin
gene in
adipocyte
differentiation.
J Biol Chem 265:477-482
Altiok S, Groner
B 1993 Interaction
of two sequencespecific
single-stranded
DNA-binding
proteins
with an
essential
region of the p-casein
gene promoter
is regulated by lactogenic
hormones.
Mol Cell Biol 13:73037310
Wang Z-Y, Lin X-H, Nobuyoshi
M, Deuel TF 1993 Identification
of a single-stranded
DNA-binding
protein that
interacts
with an Sl nuclease-sensitive
region
in the
platelet-derived
growth factor A-chain
gene promoter.
J
Biol Chem 268:10681-l
0685
Pan WT, Liu Q, Bancroft
C 1990 Identification
of a growth
hormone
gene promoter
repressor
element
and its cognate doubleand single-stranded
DNA-binding
proteins.
J Biol Chem 265:7022-7028
Lannigan
DA, Notides AC 1989 Estrogen
receptor
selectively binds the “coding
strand”
of an estrogen
responsive element.
Proc Natl Acad Sci USA 86:863-867
Ambesi-lmpiombato
FS 1986 Fast-growing
thyroid
cell
strain. US Patent no 4,608,341
Ambesi-lmpiombato
FS, Coon HG 1979 Thyroid cells in
culture. Int Rev Cytol [Suppl] 10:163-l
71
Maxam
AM, Gilbert
W 1980 Sequencing
end-labeled
DNA with base-specific
cleavages,
Methods
Enzymol
65:499-560
Saiki RK, Gelfand
DH, Stoffel S, Scharf SJ, Higuchi
R,
Horn GT. Mullis KB. Erlich HA 1988 Primer-directed
enzymatic
amplification
of DNA with a thermostable
DNA
polymerase.
Science
239:487-491
Sanger F, Nicklen S, Coulson AR 1977 DNA sequencing
with chain-terminating
inhibitors.
Proc Natl Acad Sci
USA 7415463-5467
Davis LG, Dibner MD, Battey JF 1986 Basic Methods
in
Molecular
Biology.
Elsevier,
New York, pp 93-98
Gorman
CM, Moffat LF, Howard
BH 1982 Recombinant
genomes
which express
chloramphenicol
acetyltransferase in mammalian
cells. Mol Cell Biol 2:1044-1051