0013-7227/02/$15.00/0
Printed in U.S.A.
Endocrinology 143(4):1538 –1544
Copyright © 2002 by The Endocrine Society
Nicotinamide Adenine Dinucleotide PhosphateDependent Cytosolic T3 Binding Protein as a
Regulator for T3-Mediated Transactivation
JUN-ICHIROU MORI, SATORU SUZUKI, MUTSUHIRO KOBAYASHI, TAKESHI INAGAKI,
AI KOMATSU, TEIJI TAKEDA, TAKAHIDE MIYAMOTO, KAZUO ICHIKAWA, AND
KIYOSHI HASHIZUME
Department of Aging Medicine and Geriatrics (J.-i.M., S.S., T.I., A.K., T.T., T.M., K.I., K.H.), Shinshu University School of
Medicine, 3-1-1, Asahi, Matsumoto, 390-8621, Japan; and Iida Municipal Hospital (M.K.), 438, Yawata, Iida, 395-0814,
Japan
Nicotinamide adenine dinucleotide phosphate (NADPH)dependent cytosolic T3 binding protein (CTBP) plays a role in
the regulation of nuclear transport of T3 in vitro. However, it
is not known whether CTBP regulates the T3 action. In this
study, we examined the effects of CTBP on cellular translocation of T3 and on transcriptional activation using established CTBP-expressing CHO or GH3 cells.
The expression of CTBP increased cellular and nuclear uptake of T3 in the CTBP-expressing cells. The efflux rate was
decreased by induction of CTBP. Efflux from nuclei also inhibited by induction of CTBP.
Expression of CTBP suppressed the T3-regulated luciferase
I
N OUR PREVIOUS studies, we demonstrated that nicotinamide adenine dinucleotide phosphate (NADPH)-dependent cytosolic T3-binding protein (CTBP) plays an important
role in the regulation of nuclear transport of T3 in vitro (1).
Activation of CTBP with NADPH inhibits the nuclear entry of
T3, whereas activation with NADP enhances the entry (2, 3).
Further, we observed that acceptor protein for NADP-activated
CTBP is present in nuclei (3, 4). These observations suggest that
CTBP plays roles not only in regulation of nuclear T3 entry but
also in regulation of T3-induced gene expression. However, it
is not yet known whether CTBP regulates the T3 action.
In this study, we examined the effects of CTBP on cellular
translocation of T3 and on transcriptional activation induced by
T3 using CTBP-expressing CHO and GH3 cells. To determine
whether CTBP affects the T3-induced transactivation, we transfected with thyroid hormone response element (TRE)-fused
reporter gene and measured the reporter activity after adding
T3 in CTBP-expressing GH3 cells. Further, we examined the
effect of CTBP on expression of rat GH mRNA which is known
as one of the T3 response genes.
activity in GH3 cells. Suppression was observed to be related to
the expression level of CTBP. T3 induction of rat GH mRNA was
lower in the cells expressing CTBP than that in CTBP-null cells.
These results suggest that CTBP regulates the T3-induced
gene expression, with which an increase in the nuclear content of the T3 is associated. Because we observed that a part
of CTBP could be transported into nuclei and that acceptor
protein for CTBP is present in nuclei as previously reported,
interaction of CTBP with certain proteins, including transcription factors or nuclear T3 receptor, may contribute to the
regulation. (Endocrinology 143: 1538 –1544, 2002)
GGC-3⬘; antisense, 5⬘-CCTCAAGCATCCATCTCAACATCAAGT-3⬘.
Human fetal brain Marathon Ready cDNA (CLONTECH Laboratories,
Inc., Palo Alto, CA) was used as a template. Amplified fragment was
sequenced and checked fidelity of the sequences (GenBank accession no.
U85772).
The amplified fragment was cloned into TA cloning site of pT7-Blue
(Novagen, Madison, WI). BamHI-SalI fragment of pT7-CTBP was ligated
into pcDNA3.1 (Invitrogen, Carlsbad, CA), which is available for the
expression in mammalian cells and in vitro transcription and translation
(pcDNA-CTBP).
Preparation and establishment of CTBP-expressing CHO
cell line (CPC45) and GH3 cells
Materials and Methods
Cloning of human CTBP and construction of plasmids
CHO-K1 cells, which do not possess NADPH-dependent T3 binding
activity, were purchased from ATCC (Manassas, VA). The cells were
transfected with pQBI-CTBP by calcium-phosphate method. After selection with 400 ␮g/ml G418, the cells were cloned (CPC45). Because
parental CHO-K1 cell does not possess NADPH-dependent T3 binding
activity, expression of CTBP was confirmed by the assay for NADPHdependent T3 binding. pcDNA-CTBP or pQBI-25-fc2 plasmid (Quantum
Biotechnology Inc., Québec, Canada), which induces green fluorescent
protein (GFP), was transfected into GH3 cells by electroporation as
previously described (5). The clones were selected by the incubation
with 400 ␮g/ml G418. The expression of CTBP in CTBP-transfected cells
was confirmed by NADPH-dependent T3 binding and Western blotting.
A parental GH3 cell and a series of GH3-CTBP cells were cultured in
DMEM without and with 100 ␮g/ml G418, respectively.
Full length of human CTBP cDNA was amplified by PCR using the
following primers: sense, 5⬘-AGACTGAGGTTAGAAGGCACAGGT-
Preparation of polyclonal antibodies to CTBP
Abbreviations: CTBP, Cytosolic T3 binding protein; EF1, elongation
factor 1␣; GFP, green fluorescent protein; Ka, affinity constant; MBC,
maximal T3 binding capacity; NADPH, nicotinamide adenine dinucleotide phosphate; TRE, thyroid hormone response element.
Synthetic peptides containing a part of human CTBP amino acid
sequence were used in the immunization. Amino acid sequence (CNRTKENAEKFADTV) was chosen because of its high antigenicity index,
determined by Epitope Adviser (Fujitsu, Shizuoka, Japan). The peptide
was conjugated to form hapten with keyhole limpet hemocyanin. An-
1538
Mori et al. • CTBP Affects T3-Mediated Transactivation
tibodies were raised in rabbits obtained from Takara Inc. (Ohtsu, Japan).
The antibody was purified by affinity column chromatography with the
immunized peptide.
Western blotting
Cells were washed twice in ice-cold PBS and lysed by adding lysis
buffer (0.05% SDS, 1% Nonidet P-40, 150 mm NaCl, 50 mm Tris-HCl, pH
7.2, containing 1 mm phenylmethylsulfonyl fluoride). The lysate was
boiled for 5 min, and stored at ⫺80 C. Proteins were resuspended in lysis
buffer containing 2% SDS, and samples were separated in 10% SDSPAGE gels and transferred to immobilon-P membranes (Millipore
Corp., Bedford, MA) by semidry electrophoretic transfer (Bio-Rad Laboratories, Inc., Richmond, CA). The membranes were blocked with
TBS-T (100 mm NaCl, 10 mm Tris-HCl pH 7.5, and 0.1% Tween 20)
containing 1% skim milk. Detection was done by measuring the enhanced chemiluminescence using a horseradish peroxidase-coupled
mouse-antigoat IgG antibody (Amersham Pharmacia Biotech, Arlington
Heights, IL).
Studies of uptake and efflux of T3
T3 uptake and diffusion were estimated by the method as previously
described (6) with minor modification. The cells were grown in 24-well
plates at 37 C in a humid atmosphere of 5% CO2 in air with exchanging
the media every other day. After obtaining the late logarithmic phase of
growth, the cells were cultured in the fresh media containing 10%
resin-stripped FCS (7) for 24 h. After the incubation, the media were
changed to the same media without FCS. Two times exchange of the
media depleted the cells of the measurable amount of T3 (6), then 70 pm
[125I] T3 (3,300 ␮Ci/␮g, Dupont NEN, Boston, MA) was added. After
incubation for indicated times, the media were aspirated. After 1 min
incubation with ice-cold Dulbeco’s PBS, it was replaced with fresh icecold PBS. After repeating this procedure three times, 2 ml of 0.25 m
sucrose, 1 mm MgCl2, and 20 mm Tris-HCl (pH 7.4) containing 0.5%
Triton X-100 was added. They were incubated for 10 min to lyse the cells.
The suspension was divided into two 0.8-ml aliquots; one was for measurement of whole cell T3 uptake and the other was for measurement of
nuclear uptake, respectively. The nuclear uptake was determined by
measuring the radioactivity of the pellet obtained by centrifugation of
the suspension at 1,500 ⫻ g for 10 min. Before measurement, the pellet
was washed two times. Radioactive T3 uptake in the presence of 1 ␮m
unlabeled T3 (nonspecific uptake) was less than 4% of total uptake.
Diffusion of T3 was studied by using the cells in which 70 pm [125I]
T3 was incubated for 24 h as described for the uptake study. After
washing two times with warmed resin-stripped media, the cells were
incubated with fresh resin-stripped media. The contents of radioactivity
in whole cells and in nuclear pellet were measured by the same procedures for the uptake study.
T3 binding assay
T3 binding activity of CTBP was measured as previously described
(8). CTBP containing fractions were incubated with TED buffer (10 mm
Tris-HCl, pH 7.4, containing 0.5 mm EGTA, and 1 mm DTT) to make a
final volume of 200 ␮l. Incubation was performed in the absence or
presence of 100 ␮m NADPH. After appropriate incubation times, bound
and free hormones were separated by dextran-coated charcoal. The
dissociation constant and the maximal binding capacity were estimated
by the method of Scatchard (9). The concentration of the protein and
DNA were measured by the methods of Bradford (10) and of Burton (11),
respectively.
Nuclear T3 binding was measured as previously described (12).
The characteristics of T3 binding was determined by Scatchard analysis. Radioactive T3 binding to nuclei prepared from GH3 cells was
displaced by unlabeled T3 or its analogues. Triiodo-l-thyroacetic acid
was the most potent to displace, which was followed by T3, triiodod-thyronine, and l-T4.
Luciferase assay
GH3 cells and cloned cells were treated with trypsin for 48 h before
transfection and were plated into 10-cm dishes. Twenty-four hours
Endocrinology, April 2002, 143(4):1538 –1544 1539
before transfection, the medium was changed to DMEM with 10% serum
pretreated with resin. The cells were electroporated with 10 ␮g 2xPALTK-Luc (13), and 1.0 ␮g pSV-␤-galactosidase vector (Promega Corp.,
Madison, WI) as previously described. The cells were distributed into 24
wells with DMEM containing serum pretreated with 10% resin. Twelve
to 16 h after incubation, the medium was changed to fresh DMEM
containing 10% serum pretreated with resin and various concentrations
of T3. After additional 24 h incubation, the cells were harvested. Luciferase activity was determined by the Promega Corp. Luciferase Assay
System according to the protocol using Berthold Lumat (E.G. & G.,
Berthold, Evry, France). ␤-Galactosidase activity was measured by the
method previously described (5), and all luciferase data were corrected
for ␤-galactosidase activity to account for variations in transfection
efficacy.
RNA analysis
Total RNA corresponding to 106 cells was extracted and eluted by
using Midi kit (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol. The cDNA fragments of rat GH and elongation factor
1␣ (EF1) (14) were amplified by reverse transcriptase PCR using GH3
cell-derived mRNA as template. The primers used were 5⬘-ATGGCTGCAGACTCTCAG-3⬘ and 5⬘-GAAAGCACAGCTGCTTTC-3⬘ for rat
GH, and 5⬘-TCCCAGTGGTCATCACCATG-3⬘ and 5⬘-ATGGACAATTTGGCACCT-3⬘ for EF1. After cloning to pGEM-T easy vector (Promega
Corp.), fidelity of each fragment was confirmed by sequencing. The
probes were obtained after digestion with EcoRI. Autoradiographic signals were quantitated by a bio-imaging analyzer BAS-1500 (Fuji Photo
Film Co., Ltd., Kanagawa, Japan).
Results
Effects of expression of CTBP on cellular and nuclear
T3 uptakes
Three different clones (clones 3, 7, and 8) that can express
CTBP were obtained. As shown in Fig. 1, high, intermediate,
and low expressions were observed in clones 8, 7, and 3,
respectively. The maximal T3 binding capacity (MBC) and
affinity constant (Ka) for T3 were estimated in each cell line.
As shown in Table 1, MBC, which was determined in the
presence of NADPH, was related to the levels of CTBP expressed. In contrast to the MBC, Ka was not significantly
different among each cell line.
Effect of CTBP expression on T3 uptake was examined in
GH3 cells, which are abundant in nuclear T3 receptors (15)
but do not express CTBP. The uptake was estimated by
measuring the radioactivity after exposure of 70 pm 125I T3 in
parental and series of CTBP-expressing cells. The maximal
FIG. 1. Expression of CTBP in established cell lines derived from
GH3 cells. GH3 cells were transfected with cDNAs coding for CTBP.
Twenty-four hours after incubation, media were changed which contained 400 ␮g/ml of G418 to select the CTBP-expressing clones. Cell
extracts from the obtained clones were submitted to Western blot
analysis with an antibody (CTBP168). The intensity of the band was
measured by the densitometer. The value obtained from the lane of
GH3 was subtracted from that of each cell line. The relative expression levels were calculated from the data in which the expression level
in clone C3 was defined as 1.
1540
Endocrinology, April 2002, 143(4):1538 –1544
Mori et al. • CTBP Affects T3-Mediated Transactivation
TABLE 1. MBC and Ka of NADPH-dependent T3 binding in the CTBP-expressing cell lines and in various rat tissues
CTBP-expressed GH3 cell lines
Relative expression level estimated
by Western blotting
MBC (fmol/␮g DNA)
Ka (⫻ 109 M⫺1)
Rat tissues
C3
C7
C8
1
7.84
27.6
5 ⫾ 1.6
1.31 ⫾ 0.3
43 ⫾ 14
1.48 ⫾ 0.4
171 ⫾ 54
1.41 ⫾ 0.5
Cerebrum
Heart
Kidney
72 ⫾ 21
1.13 ⫾ 0.25
121 ⫾ 29
1.13 ⫾ 0.21
162 ⫾ 37
1.53 ⫾ 0.35
Preparation of cytosolic fraction from rat tissues was performed as previously described (8). MBC and Ka were obtained by estimation with
Scatchard analyses. There are no significant (P ⬎ 0.1, determined by an analysis of variance) differences in Kas between GH3 cell lines and
rat tissues. Each value indicates mean ⫾ SD of triplicate determinations.
FIG. 2. Effects of CTBP expression on cellular and nuclear T3 uptakes. A, Whole cell uptake of T3. CTBP-expressing clones (clones 3, 7, and
8) were obtained as described in Fig. 1. The cells were preincubated in serum free media for 24 h to deplete T3 endogenously presents in the
cells, then 70 pM [125I] T3 were added to each incubation. After appropriate incubation times, radioactivities incorporated into the cells were
measured. The CTBP-dependent uptakes which were obtained by subtraction of the radioactivity incorporated into GH3 cells (CTBP-independent uptake) are shown. Inset shows the total incorporation of radioactive T3 in each cells. Abscissa shows the time after adding radioactive
T3. Each datum indicates the mean of triplicate determinations. B, Nuclear uptake of T3. After incubation of the cells with 70 pM [125I] T3 as
described in A, the nuclear fractions were prepared as described in Materials and Methods. The CTBP-dependent nuclear uptake was obtained
by subtraction of CTBP-independent nuclear uptake, which was obtained in CTBP nonexpressing GH cells, from the uptake obtained in each
CTBP-expressing cells as mentioned in (A). Inset shows the total nuclear uptake into each cells. Each datum indicates the mean of triplicate
determinations, which did not vary by more than 5%.
uptake in whole cell was larger in CTBP-expressing cells than
that in CTBP-null cells. Subtraction of CTBP-independent T3
uptake revealed that the level of expression of CTBP well
correlated to the level of T3 uptake (Fig. 2, left panel). Not only
whole cell uptake but also nuclear uptake was enhanced by
expressing CTBP. The increase in nuclear uptake was also
dependent on the levels of CTBP expression (Fig. 2, right
panel). The MBCs of CTBP positively related to the uptakes
in whole cells and nuclei of established cell lines (Fig. 3).
To clarify the effect of CTBP expression on T3 uptake in
other cell line, we established CTBP-expressing CHO cells
line. In contrast to the GH3 cells, which are derived from rat
pituitary gland and are abundant in nuclear T3 receptors (15)
but are quite low in NADPH-dependent T3 binding activity,
CHO cells possess neither nuclear T3 binding nor NADPHdependent T3 binding. As shown in Table 2, the specific
whole cell uptake was larger in CTBP-expressing CHO cells
than that in CTBP-null CHO cells. Thus, it is possible that
FIG. 3. Relation between MBC and T3 uptake in CTBP-expressing
cell lines. MBC and T3 uptake demonstrated in Tables 1 and 2, respectively, were plotted. Closed squares and circles indicated the
mean values of whole cellular and nuclear uptake, respectively. The
bars represented SD. Dotted lines were constructed by linear regression analysis.
Mori et al. • CTBP Affects T3-Mediated Transactivation
Endocrinology, April 2002, 143(4):1538 –1544 1541
TABLE 2. Total uptakes of T3 in CTBP-CHO cells and a series of CTBP-expressing GH3 cells
Cell line
CHO-K1
CTBP-CHO
GH3
C3
C7
C8
Whole cell uptake (cpm/1 ⫻ 106 cells)
701 ⫾ 55
1538 ⫾ 115
2031 ⫾ 154
2438 ⫾ 188
3010 ⫾ 224
4033 ⫾ 301
ND
ND
1649 ⫾ 132
1845 ⫾ 164
2059 ⫾ 188
2661 ⫾ 223
Nuclear uptake (cpm/1 ⫻ 106 cells)
Cells were grown in resin-stripped media for 48 h. Twenty-four hours after incubation with 70 pM [125I] T3, cells were harvested and divided
into two aliquots. The radioactivity of one aliquot was directly measured for the determination of whole cell [125I] T3 uptake. Another aliquot
was centrifuged at 1500 ⫻ g for 10 min at 4 C. After removing the supernatant, the pellet was washed twice with SMT containing 0.5% Triton
X-100, followed by a single wash with SMT. The radioactivity in the nuclear pellet was counted. ND, Not done.
FIG. 4. Effects of CTBP expression on
cellular and nuclear T3 efflux. MBC and
half-life demonstrated in Tables 1 and 3,
respectively were plotted. Open squares
and circles indicated the mean values of
half-life obtained from the whole cells
and the nuclei, respectively. The abscissa represents semilogarithmic plot
of MBC. The ordinate indicates half-life.
Inset, Cells were incubated with 70 pM
[125I]T3 for 48 h in resin-stripped media.
After further incubation for appropriate
times in the absence of radiolabeled T3
in resin-stripped media, cells were harvested, and radioactivities in cell homogenate (left panel) and in nuclear pellet (right panel) were measured. Closed
square, circle, triangle, and cross represent the radioactivities in GH3, C3, C7,
and C8, respectively. The abscissa indicates the time after depletion of radiolabeled T3, and the ordinate represents
semilogarithmic plot of radioactivity.
Each datum indicates the mean of triplicate determinations, which did not
vary by more than 5%.
CTBP plays a role in the retention of whole cell T3 uptake not
only in GH3 but also CHO cells.
Effects of CTBP expression on T3 effluxes
Export of T3 from whole cell to media and from nuclei to
cytoplasm was examined in GH3 and CTBP-expressing
cloned GH3 cells. The cells were incubated in the presence
of radiolabeled T3 for 48 h, and the export was estimated by
measuring contents of radiolabeled T3 in whole cells (Fig. 4
inset, left panel) and in nuclei (Fig. 4 inset, right panel) after
deprivation of T3 from media. The first-order dissociation
kinetics was calculated from data obtained within 7.5 h from
the beginning of incubation. The rate of whole cellular efflux
was significantly decreased by expressing CTBP. The rate of
nuclear efflux was also decreased by expressing CTBP. Calculated half life of nuclear T3 content was longer in the cells
expressing CTBP than that observed in CTBP-null cells (Table 3). The prolongation of the half-life was correlated to the
expression level of CTBP (Fig. 4).
Effect of CTBP on T3-mediated gene expression
Above results suggested that CTBP expression influences
the gene expression regulated by T3 because the T3 content
in nuclei could be augmented by CTBP expression. We estimated the effect of CTBP on T3-mediated transcription in
CTBP-expressing and nonexpressing GH3 cell lines. We
TABLE 3. Half-life of [125I] T3 contents in whole cells and nuclei,
after deprivation of T3 from the media
Cell line
GH3
Clone 3
Clone 7
Clone 8
Half-life (hours)
Whole cell
Nuclei
4.89 ⫾ 0.48
NS
5.35 ⫾ 0.55
a
6.43 ⫾ 0.68
7.03 ⫾ 0.73b
]
5.19 ⫾ 0.50
NS
6.17 ⫾ 0.68
a
7.72 ⫾ 0.70
8.92 ⫾ 0.87b
]
The data were calculated from three separate experiments. Halflife was determined by the method as previously described (17). Statistical significance was determined by an analysis of variance followed by the Bonferroni multiple comparison test. a, P ⬍ 0.05, GH3
vs. Clone 7. b, P ⬍ 0.01, GH3 vs. Clone 8. NS, No significance.
transfected TRE containing luciferase reporter gene (TREx2TK-Luc) into GH3 cells and cloned cells (C3, C7, and C8). The
levels of nuclear T3 receptor were not different among these
cell lines (Table 4). T3 increased luciferase activity in CTBP
nonexpressing GH3 cells. The luciferase activity induced by
T3 was significantly lower in CTBP-expressing cells than that
in nonexpressing cells. The diminishment of luciferase activity correlated to the expression level of CTBP (Fig. 5).
Effect of CTBP on T3 response gene expression
Above results suggested that CTBP possesses suppressive
effect on T3-induced mRNA expression. To evaluate the ef-
1542
Endocrinology, April 2002, 143(4):1538 –1544
Mori et al. • CTBP Affects T3-Mediated Transactivation
TABLE 4. MBC and Ka of nuclear T3 receptors in the parental GH3 and CTBP-expressing cell lines
Cell line
GH3
C3
C7
C8
MBC (fmol/␮g DNA)
Ka (⫻ 1010 M⫺1)
2.41 ⫾ 0.3
1.11 ⫾ 0.2
2.52 ⫾ 0.4
1.02 ⫾ 0.2
2.48 ⫾ 0.3
1.04 ⫾ 0.2
2.62 ⫾ 0.4
1.08 ⫾ 0.2
Preparation of nuclei was performed as previously described (12). MBC and Ka were obtained by estimation with Scatchard analyses. There
are no significant (P ⬎ 0.1, determined by an ANOVA) differences in MBC and Kas between GH3 cell lines and other cell lines. Each value
indicates mean ⫾ SD of five separate determinations.
FIG. 5. Effect of CTBP expression on trans-activation induced by T3
in GH3 cells. 2xPAL-TK-Luc reporter gene and 1.0 ␮g pSV-␤-galactosidase vector were cotransfected into GH3 cells and cloned cells (C3,
C7, and C8), and incubated with various concentrations of T3 for 24 h.
After cell harvest, luciferase and ␤-galactosidase activities were measured. All luciferase activities were corrected for ␤-galactosidase activities. Data were expressed as fold induction by T3. Each value
represents the mean ⫾ SD of five separate determinations. Asterisk
indicates statistical significance (P ⬍ 0.05). N.S., No significance.
fect of CTBP on the TR activation-induced transcription, we
measured the amount of rat GH mRNA induced by T3 in
GH3 and the cloned cells. Twenty-four hours after culture of
these cells with 10 nm T3, rat GH mRNA levels were estimated by Northern blot analysis. As shown in Fig. 6A, T3
increased the level of rat GH mRNA in a dose-dependent
manner. The T3-induced increase in GH mRNA expression
was inhibited by expressing CTBP (Fig. 6, A and B). To
exclude the possibility that overexpressed protein may influence the T3-induced mRNA expression (nonspecific expression), we developed GFP-expressing GH3 cells, and the
changes in rat GH mRNA expression levels were examined.
As shown in Fig. 6, C and D, overexpression of GFP did not
affect the expression of rat GH mRNA.
Discussion
We have previously suggested that CTBP plays roles not only
in the cytoplasmic T3 reservoir but also in the regulation of T3
nuclear transport or in shuttling T3 between cytoplasm and
nucleus (3). The levels of expression of CTBP correlated to the
MBC of T3 in CTBP-expressing GH3 cells prepared in this study.
The Kas for T3 in CTBP expressed in these cells were similar to
those obtained in CTBP prepared from rat tissues, indicating
that the characteristics of T3-binding in CTBP expressed in GH3
cell are similar to those of native CTBP.
Although we have demonstrated the T3 binding of purified CTBP in the presence of NADPH or NADP⫹ in vitro (3),
it was not determined whether the expression of CTBP molecule alters the T3 content in living cells. The positive corelation between the expression of CTBP and T3 contents in
established cell lines indicates that CTBP holds T3 in living
cells. These data imply that physiologically, CTBP affects the
tissue content of T3 in vivo.
We observed that maximal level of T3 uptake or the efflux
from nucleus to cytoplasm was correlated to the expression
level of CTBP in GH3 cells. These results suggest that CTBP
plays a role not only in cytoplasmic reservoir but also in
nuclear retention of T3. In our previous study, we observed
that CTBP acceptor sites present in nuclei (4), and NADPactivated form of CTBP can accelerate the nuclear import of
T3 (2). Thus, the observation that increase in nuclear uptake
induced by expression of CTBP may reflect the acceleration
of nuclear import induced by NADP-dependent activation of
CTBP. A delayed efflux from nucleus, induced by expression
of CTBP, may reflect the presence of CTBP-acceptor interaction in nucleus. We found that T3 was passively or actively
transferred into nuclei of CTBP-null GH3 cells, indicating
that free T3 or T3-bounded other proteins may also be transferred into nuclei. Thus, not only the expression of CTBP, but
also other mechanisms may affect the nuclear content of T3
in the CTBP-expressing cells.
These observations suggest that nuclear events induced by
thyroid hormone may be also influenced by expression of
CTBP. As is shown, the T3-induced gene expression, evaluated by estimation of reporter gene expression, was suppressed by expression of CTBP in vitro. Further, level of rat
GH mRNA, which is one of the T3-responsive genes, was
decreased in GH3 cells expressing CTBP. Although we did
not examine other thyroid hormone response genes, these
results suggest that CTBP strongly influences the T3 action
mediated through T3 receptor-TRE interaction.
Because we artificially expressed CTBP in CTBP-null GH3
cells, it was possible that the transcriptional activity was
affected by the artificial expression of CTBP, namely squelching effect. T3 response, however, was not suppressed in the
GFP-expressed GH3 cell line as a negative control, and physiological amount of CTBP was present in a series of CTBPexpressed cell lines, suggesting the possibilities of artificial
modification may be low.
Based on the findings in this study, it is possible that
nuclear content of T3 is high in nucleus of the cells expressing
Mori et al. • CTBP Affects T3-Mediated Transactivation
Endocrinology, April 2002, 143(4):1538 –1544 1543
FIG. 6. Effect of CTBP expression on GH mRNA levels in
GH3 cells. GH3 cells and cloned cells (C3, C7, and C8) were
prepared as described. Twenty-four hours after incubation
with DMEM with 10% resin-stripped serum, various concentrations of T3 (A) or 10 nM T3 (B) were added. After
incubation for 24 h (A) or appropriate time (B), total RNAs
were prepared. Eight micrograms of total RNA were analyzed by Northern blotting. Autoradiography of each blot
hybridized with 32P-labeled cDNA probes of rat GH or EF1
was carried out by BAS-1500. Similar results were obtained
in more than three experiments. C, pQBI-25-fc2 plasmid
was transfected into GH3 cells by electroporation. The
clones were selected by the incubation with 400 ␮g/ml
G418. Cell extracts from the parental GH3 cells and the
obtained clone were applied to Western blot analysis with
anti-GFP antibody. D, Eight micrograms of total RNA, extracted from GH3 and GFP-GH3 cells after 24 h incubation
with or without 10 nM T3, were analyzed by Northern blotting with rat GH and EF1 as probes. Similar results were
obtained in more than three experiments.
CTBP. Nevertheless, the T3 action was inhibited in these cells.
Precise mechanism of this phenomenon could not be solved
in this study. There are several possibilities to explain molecular mechanisms of the data. 1) CTBP may increase T3
catabolism in CTBP-expressing cells. 2) CTBP may increase
turnover of rat GH mRNA in CTBP-expressing cells. 3) CTBP
may compete for T3 binding with classic nuclear receptors.
However, a noncompetitive inhibition is seen in Fig. 6, suggesting that the third possibility is less likely.
The fourth possibility is that CTBP may directly enter into
nucleus because we could isolate the CTBP acceptor in nuclei
and we observed the entry of CTBP when the protein is
activated by NADP (2). We observed that CTBP molecule can
make homodimer in vitro, which indicated that the protein
may interact with other proteins even in nucleus (16). These
considerations lead us to a hypothesis that CTBP entered into
nucleus may regulate the redundant functions, including the
transcriptional functions of nuclear receptor or cofactor(s) for
the transactivation. However, the precise mechanism of the
regulation induced by CTBP is remained to be elucidated.
Acknowledgments
Received September 11, 2001. Accepted December 14, 2001.
Address all correspondence and requests for reprints to: Satoru Suzuki, M.D., Ph.D., Department of Aging Medicine and Geriatrics, Shinshu University School of Medicine, 3-1-1, Asahi, Matsumoto, Nagano,
390-8621, Japan. E-mail: [email protected].
References
1. Hashizume K, Miyamoto T, Ichikawa K, Yamauchi K, Kobayashi M, Sakurai
A, Ohtsuka H, Nishii Y, Yamada T 1989 Purification and characterization of
NADPH-dependent cytosolic 3,5,3⬘-triiodo-l-thyronine binding protein in rat
kidney. J Biol Chem 264: 4857– 4863
2. Hashizume K, Miyamoto T, Yamauchi K, Ichikawa K, Kobayashi M, Ohtsuka H, Sakurai A, Suzuki S, Yamada T 1989 Counterregulation of nuclear
3,5,3⬘-triiodo-l-thyronine (T3) binding by oxidized and reduced-nicotinamide
adenine dinucleotide phosphates in the presence of cytosolic T3-binding protein in vitro. Endocrinology 124:1678 –1683
3. Hashizume K, Miyamoto T, Ichikawa K, Yamauchi K, Sakurai A, Ohtsuka
H, Kobayashi M, Nishii Y, Yamada T 1989 Evidence for the presence of two
active forms of cytosolic 3,5,3⬘-triiodo-l-thyronine(T3)-binding protein (CTBP)
in rat kidney. Specialized functions of two CTBPs in intracellular T3 translocation. J Biol Chem 264:4864 – 4871
4. Hashizume K, Miyamoto T, Kobayashi M, Suzuki S, Ichikawa K, Yamauchi
K, Ohtsuka H, Takeda T 1989 Cytosolic 3,5,3⬘-triiodo-l-thyronine (T3)-binding
protein (CTBP) regulation of nuclear T3 binding: evidence for the presence of
T3-CTBP complex-binding sites in nuclei. Endocrinology 124:2851–2856
5. Suzuki S, Miyamoto T, Opsahl A, Sakurai A, DeGroot LJ 1994 Two thyroid
hormone response elements are present in the promoter of human thyroid
hormone receptor b1. Mol Endocrinol 8:305–314
6. Ichikawa K, Hashizume K, Kobayashi M, Nishii Y, Ohtsuka H, Suzuki S,
Takeda T, Yamada T 1992 Heat shock decreases nuclear transport of 3,5,3⬘triiodo-l-thyronine in clone 9 cells. Endocrinology 130:2317–2324
7. Samuels HH, Stanley F, Casanova J 1979 Depletion of l-3,5,3⬘-triiodothyronine and l-thyroxine in euthyroid calf serum for use in cell culture studies of
the action of thyroid hormone. Endocrinology 105:80 – 85
8. Suzuki S, Hashizume K, Ichikawa K, Takeda T 1991 Ontogenesis of the high
affinity NADPH-dependent cytosolic 3,5,3⬘-triiodo-l-thyronine-binding protein in rat. Endocrinology 129:2571–2574
9. Scatchard G 1949 The attractions of proteins for small molecules and ions. Ann
NY Acad Sci 51:660 – 672
10. Bradford MM 1976 A rapid and sensitive method for the quantitation of
micogram quantities of protein utilizing the principle of protein-dye binding.
Anal Biochem 72:248 –254
1544
Endocrinology, April 2002, 143(4):1538 –1544
11. Burton K 1956 Study of the condition and mechanism of the diphenylamine
reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem J
62:315–319
12. Suzuki S, Kobayashi H, Sekine R, Kumagai M, Mikoshiba M, Mori J, Hara
M, Ichikawa K, Hashizume K 1997 3,5,3⬘-Triiodothyronine potentiates alltrans-retinoic acid-induced apoptosis during differentiation of the promyeloleukemic cell HL-60. Endocrinology 138:805– 809
13. Kakizawa T, Miyamoto T, Ichikawa K, Kaneko A, Suzuki S, Hara M, Nagasawa T, Takeda T, Mori J, Kumagai M, Hashizume K 1999 Functional
characterization between oct-1 and retinoid X receptor. J Biol Chem 274:19103–
19108
14. Ann DK, Lin HH, Lee S, Tu Z-J, Wang E 1992 Characterization of the statin-
Mori et al. • CTBP Affects T3-Mediated Transactivation
like S1 and rat elongation factor 1 ␣ as two distinctly expressed messages in
rat. J Biol Chem 267:699 –702
15. Seo H, Vassart G, Brocas H, Refetoff S 1977 Triiodothyronine stimulates
specifically growth hormone mRNA in rat pituitary tumor cells. Proc Natl
Acad Sci USA 74:2054 –2058
16. Kobayashi M, Hashizume K, Suzuki S, Ichikawa K, Takeda T 1991 A novel
NADPH-dependent cytosolic 3,5,3⬘-triiodo-l-thyronine-binding protein
(CTBP;5.1S) in rat liver: a comparison with 4.7S NADPH-dependent CTBP.
Endocrinology 129:1701–1708
17. Ichikawa K, Brtko J, DeGroot LJ, Hashizume K, Yamada T 1989 Stabilization,
accurate determination, and purification of rat liver nuclear thyroid hormone
receptor. J Endocrinol 120:237–243