Thyroid hormone induces the expression of
4-1BB
and activation
of caspases in a thyroid hormone receptor-dependent manner
Toshiko Yamada-Okabe
1
, Yasuo Satoh
2,
* and Hisafumi Yamada-Okabe
1,3
1
Department of Hygiene,
2
Department of Surgery, School of Medicine, Yokohama City University, Fukuura, Kanazawa,
Yokohama;
3
Pharmaceutical Research Department 4, Kamakura Research Laboratories, Chugai Pharmaceutical Co. Ltd,
Kajiwara, Kamakura, Kanagawa, Japan
Thyroid hormone has various effects on cell proliferation,
growth and apoptosis. To gain more insight into the
molecular dynamics caused by thyroid hormone, gene
expression in HeLaTR cells that constitutively over-
expressed the thyroid hormone receptor (TR) was analyzed.
Gene expression profiling of the HeLaTR cells with an
oligonucleotide microarray yielded 229 genes whose
expression was significantly altered by T3. Among these
genes, the expression of 4-1BB, which is known to initiate a
signal cascade activating NF-jB, was significantly up-regu-
lated by T3. Although treatment of the HeLaTR cells with
T3 did not induce expression of NF-jB reporter luciferase,
even in the presence of the 4-1BB-Ligand, it increased the

genes and are expressed as TRa,TRb1orTRb2 [4–6]. Each
TR forms either a homo- or a hetero-dimer with a
retinoid X receptor and binds to the thyroid hormone
response element (TRE) that is present in the promoter/
enhancer of certain genes [7]. In the absence of T3, a
corepressor such as N-CoR or SMRT is associated with the
homo- or heterodimer of TR and represses the transcription
mediated by TR. Upon binding of T3 to TR, the
corepressor is displaced by a coactivator such as SRC1,
TIF2 (GRIP1) or TRAM-1 (p/CIP, AIB1, ACTR, RAC3),
making TR active in transcription [8,9].
Recently, a number of genes that were regulated by
thyroid hormone were identified by cDNA microarray with
hepatic RNA prepared from hypothyroid mice treated with
T3 [10,11], and the expression of approximately 2.5% of the
examined genes (55 of 2225 genes) appeared to be modulated
by thyroid hormone treatment. Genes under the control of
T3 included those involved in hepatocyte proliferation, cell
survival (apoptosis), gluconeogenesis, lipogenesis, and insu-
lin signaling [10]. As thyroid hormone causes growtharrest of
thyrotropic tumors, genes involved in thyrotropic tumor
growth were also explored by cDNA microarray. Among
1176 genes examined, 40 genes were down-regulated and
seven were up-regulated by thyroid hormone [12]. However,
because these studies were carried out with cells or tissues
expressing only low levels of TR, cells might only weakly
respond to T3, and therefore, cellular responses to T3 might
not be clearly detected by DNA microarray.
In this study, we examined the effects of T3 on gene
expression in the HeLa cells that overexpressed TR by

activate caspases to induce apoptosis.
Materials and methods
Generation of cell lines that constitutively overexpress
TR
The entire open reading frame (ORF) of the human thyroid
hormone receptor (hTRa1) cDNA was amplified by PCR
using primers containing the sequences of the 5¢-and
3¢-coding regions of TR and ligated at the HindIII cleavage
site of pCMV-Tag4 (Stratagene). The HeLa S3 cells were
transfected with either the resulting plasmid or pCMV-Tag4
by lipofectamine reagent (Invitrogen) and selected with
1mgÆmL
)1
of G418 for 2 weeks. G418-resistant clones were
isolated and tested for expression levels of hTRa1. Cells
from a single clone that expressed a high level of TR were
designated HeLaTR and used for the study. HeLa S3 cells
were cultured in Dulbecco’s modified Eagle’s medium
(DME) supplemented with 10% fetal bovine serum
(FBS). Exposure of the cells to T3 was carried out in
DME supplemented with 10% FBS that had been treated
with charcoal and dextran (HyClone). The nucleotide
sequences of the primers used for amplifying the hTRa1
cDNA were: 5¢-CCCGGGAAGCTTCGGACCATGG
AACAGAAGCCAAGCAAGGTG-3¢ and 5¢-CCCGGG
GTCGACGACTTCCTGATCCTCAAAGACCTC-3¢.In
order to overexpress 4-1BB and TRAF1 in the HeLaTR
cells, the entire coding regions of the 4-1BB and TRAF1
cDNAs were amplified by RT-PCR and ligated at the
HindIII site of pRC/CMV eukaryotic expression plasmid.

saline and 40 lCi of a mixture of [
35
S]methionine and
[
35
S]cysteine (1000 Ci mmol
)1
;pro-mix
L
-[
35
S] in vitro cell
labelling mix, Amersham) for 2 h. Cell extracts were
prepared by lysing the cells with RIPA buffer: [150 m
M
NaCl, 1% NP4O, 0.5% deoxycholate, 0.1% SDS, 50 m
M
Tris-HCl (pH 8.0)]
4
,andthehTRa1 protein was immuno-
precipitated from the cell extracts as described in a previous
paper but with 2 lgÆmL
)1
of the anti-(TR polyclonal Ig)
(clone FL-408, Santa Cruz) [14]. The hTRa1proteinwas
separated by 10% SDS/PAGE and visualized with the Fuji
BAS 2000 system.
Luciferase reporter assay
To monitor the TR-mediated transcription, the reporter
plasmid was constructed with the pGL2-promoter plasmid

AA-3¢ and 5¢-GATCTTGTCCTTGTCCTCCC-3¢.
Electrophoretic mobility shift assay
Nuclear extracts were prepared as described by Dignam
et al. [15]. Two microliters of the oligonucleotide harboring
Ó FEBS 2003 T3 induces 4-1BB in a TR-dependent manner (Eur. J. Biochem. 270) 3065
the NF-jB consensus sequence (Promega) that had been
end-labelled with [c-
32
P]ATP (6000 CiÆmmol
)1
; Amersham
Pharmacia) was incubated in a reaction mixture (10-lL final
volume) containing 4% (w/v) Ficoll, 20 m
M
Hepes/KOH
(pH 7.9), 50 m
M
KCl, 1 m
M
EDTA, 1 m
M
dithiothreitol,
2.5 lg of poly(dI-dC) (Roche Diagnostics), 6 m
M
MgCl
2
and 10-lg protein amounts of nuclear extract at room
temperature for 10 min. Thereafter, 2 lL aliquot was
electrophoresed on a 6% polyacrylamide gel and radio-
labelled oligonucleotides were visualized with an image

ization was carried out in 200 lL of a buffer containing
0.1
M
Mes (pH 6.7), 1
M
NaCl, 0.01% (v/v) Triton X-100,
20 lg herring sperm DNA, 100 lg acetylated bovine serum
albumin, 10 lg of the fragmented cRNA and biotinylated-
control oligonucleotides at 45 °C for 12 h. After the chips
were washed with a buffer containing 0.01
M
Mes (pH 6.7),
0.1
M
NaCl and 0.001% (v/v) Triton X-100, they were
further incubated with biotinylated anti-streptavidin Ig and
stained with streptavidin R-Phycoerythrin (Molecular
Probes) to amplify intensities. Each pixel level was collected
by laser scanner (Affymetrix) and the levels of the expression
of each cDNA and of the reliability (Present/Absent call)
were calculated with software (
AFFYMETRIX MICROARRAY
SUITE
ver.4.0). For comparisons of the expression level
of each gene, average differences below 20 were considered
to be 20.
Determination of the caspase activities
One million cells of HeLa, HeLaTR, HeLaTR/4-1BB,
HeLaTR/TRAF1 and FRTL were plated onto 10-cm
diameter

One thousand cells of HeLaTR and FRTL were cultured in
100 lL of the medium in the presence or absence of T3 and
Z-VAD-FMK at 37 °C for 48 h. At the end of the culture,
20 lL of CellTiter 96 AQ
ueous
One Solution Reagent
(Promega) containing 5-(3-carboxymethoxyphenyl)-2-(4,5-
dimenthylthiazoly)-3-(4-sulfophenyl) tetrazolium salt (MTS)
was added to the medium, and the cells were incubated
at 37 °C for 1–4 h. Absorbance at 490 nm, which represents
a viable cell, was measured with a microplate reader.
Detection of DNA fragmentation
Both floating and adherent cells of HeLaTR and FRTL
that were cultured in the presence or absence of T3 were
collected and lysed with a lysis buffer containing 10 m
M
Tris/HCl (pH 8.0), 10 m
M
EDTA and 0.5% (v/v) Triton
X-100. Cell lysates were treated with RNase A
(0.1 mgÆmL
)1
)at37°C for 1 h and then with proteinase K
(1 mgÆmL
)1
)at50°C for 2 h. DNA was extracted with
phenol/chloroform mixture and precipitated by isopropa-
nol. Five lg of DNA was fractionated by agarose gel
electrophoresis.
Results

trations of T3 to induce the luciferase activity in the
HeLaTR cells ranged from 25 to 100 ngÆmL
)1
,whichis
consistent with a previous report by Fondell et al.[16].
Furthermore, the increase in luciferase activities correlated
with increase in the luciferase mRNA (not shown). No
increase in the luciferase activity by T3 was observed in the
vector-transfected HeLa cells. These results demonstrate
that hTRa1wasactivatedbyT3intheHeLaTRcellsand
stimulated the transcription of genes in a TRE-dependent
manner and that the parental HeLa cells were TR-negative
and thereby not responsive to T3.
Genes whose expression is altered by T3
To identify the genes whose expression is regulated by T3,
we performed oligonucleotide microarray for approxi-
mately 11 000 human genes. Gene expression was com-
pared between the HeLaTR cells and those treated with
T3; genes whose mRNA levels showed a change of more
than threefold between the two were selected. Among the
229 genes that matched this criterion, 113 genes were
up-regulated and 116 genes were down-regulated by T3
(Fig. 2). The expression of these genes was further verified
by RT-PCR. As conventional RT-PCR did not amplify
detectable levels of the cDNA fragments from the
mRNA, whose average difference in the DNA microarray
was less than 400, we focused on the genes with average
differences greater than 400 either in the HeLaTR cells
or those treated with T3. RT-PCR confirmed that T3
up-regulated the expression of 4-1BB, pregnancy specific

35
S]methionine/cys-
teine, and the TRa1 protein was immunopre-
cipitated from the cell extracts with an anti-
human TRa Ig. The human TRa1proteinwas
fractionated on a polyacrylamide gel and
visualized by an image analyzer. The arrow
indicates the band of human TRa1protein.
(C) Reporter luciferase activities induced by
T3. HeLaTR and vector-transfected HeLa
cells were transfected with plasmid carrying
DR0orDR4thatwaslinkedtotheluciferase
gene and cultured in the presence or absence of
T3 for 48 h. Cell extracts were prepared from
the cells, and the luciferase activities in the cell
extracts were determined. One relative luci-
ferase unit (RLU) represents 1.66 light units
examined by recombinant luciferase.
Ó FEBS 2003 T3 induces 4-1BB in a TR-dependent manner (Eur. J. Biochem. 270) 3067
As shown in Fig. 3B, RT-PCR at different time points
after the addition of T3 revealed that 4-1BB expression was
induced as early as 1 h after the T3 treatment and its
mRNA level continued to increase at least up to 48 h. The
PSG7 mRNA reached a detectable level within 2 h and
remained at a low level until 24 h. Its level, however, was
elevated strongly between 24 and 48 h. Whereas the
HeLaTR cells expressed a certain level of the fmfc mRNA
even in the absence of T3, T3 further increased its level at
2–4 h. The BMP6 mRNA level started declining between
2 and 4 h and continued to decline up to 48 h after the

presence or absence of T3 for 48 h were analyzed by oligonucleotide microarray. Genes whose mRNA levels differed by more than fivefold between
the T3-untreated and T3-treated HeLaTR cells were selected with
AFFYMETRIX MICROARRAY SUITE
ver.4.0 software and are shown. The number on
each bar represents the -fold change of the mRNA level between the two. Genes up-regulated by T3 are shown in the upper half, and those down-
regulated by T3 are listed in the lower half. Each gene is annotated with its GenBank accession number. Dotted bars, genes with an average
difference greater than 400 after up-regulation or before down-regulation by T3. Filled bars, genes with an average difference less than 400 even after
up-regulation or before down-regulation by T3.
3068 T. Yamada-Okabe et al. (Eur. J. Biochem. 270) Ó FEBS 2003
very low level of luciferase activity was observed in the cells
transfected with plasmid carrying the luciferase gene with-
out the NF-jB enhancer sequences (Fig. 4). Because 4-1BB
has to be dimerized by binding to its ligand to initiate signal
cascade, we also treated the HeLaTR cells with T3 and
4-1BB-L. However, even in the presence of 100 ngÆmL
)1
4-1BB-L, a concentration 10 times higher than that of the
ED
50
for IL-8 production in human peripheral blood
mononuclear cells, the NF-jB reporter luciferase activity
was not induced by T3 (Fig. 4). This was a rather
unexpected result, so we next explored other possible effects
of T3. As T3 is known to induce apoptosis in rat pancreatic
beta-cell and in amphibian metamorphosis [2,23], we
examined the effects of T3 on the caspase activities in the
HeLaTR cells. As shown in Fig. 5, T3 increased the caspase
activities in the HeLaTR cells to a level similar to that of
TNFa (Fig. 5A). The activity detected in this assay was
indeed caspase activity because a pan-caspase inhibitor,

HeLaTR/TRAF1 cells was almost the same as that in the
parental HeLaTR cells, indicating that overexpression of
TRAF1 did not significantly influence the caspase activities
(Fig. 5A).
We also asked whether the up-regulation of 4-1BB
expression and activation of caspases by T3 occurred in
other cells. Rat FRTL cells that originated from normal
thyroid epithelial cells were used because they were not
genetically manipulated and T3 bound to the FRTL cells
[24]. RT-PCR demonstrated that the FRTL cells indeed
expressed a certain level of TRa (Fig. 6A). As in the
HeLaTR cells, T3 treatment of the FRTL cells significantly
increased the level of the 4-1BB mRNA and decreased the
level of the BMP6 mRNA. Furthermore, a marked increase
in the Z-VAD-FMK-sensitive caspase activities was also
observed in the FRTL cells that were cultured in the
presence of T3 (Fig. 6B). Caspase activities in the FRTL cell
extracts prepared from cells treated with T3 were signifi-
cantly inhibited by caspase 3 inhibitor DEVD-FMK and
also, to lesser extents, by caspase 1 inhibitor YVAD-CMK
and by caspase 8 inhibitor IETD-FMK (not shown),
indicating that several caspases were activated by T3. In
addition, T3 impaired the proliferation of the cells and
significantly increased the degree of DNA fragmentation in
FRTL cells (Fig. 6C,D). Thus, it appears that induction of
the expression of 4-1BB and subsequent activation of
caspases by T3 leads to apoptosis in rat FRTL cells.
Discussion
Using oligonucleotide microarray and subsequent
RT-PCR, we identified the genes whose expression is

sion levels of these genes cause profound effects on the
cells.
Among genes whose expression is modulated by T3,
4-1BB is responsible for a protein that can provide
costimulatory signals to stimulate T-cell proliferation [25].
Expression of 4-1BB has been thought to be rather
restricted, in the body, to activated T-cells
9
; it is induced by
the activation of T-cells from peptide-MHC complexes
and mitotic stimulants such as phorbol 12-myristate
13-acetate (PMA)
10
[26]. Our results show that T3 induced
the transcription of 4-1BB in the HeLa cells harboring the
functional TR and also in rat FRTL cells, which suggests
the involvement of 4-1BB not only in T-cell-mediated
immune responses but also in other systems where
functional TR is expressed.
As well as treatment of the HeLaTR cells with T3,
overexpression of 4-1BB in the HeLaTR cells led to the
activation of caspases. Activation of caspases by the
overexpression of 4-1BB was not an artefact of over-
expressing any gene because caspase activities were not
elevated in the HeLaTR cells overexpressing TRAF1.These
results demonstrate that T3 induces the expression of
4-1BB, which in turn activates caspases. Activation of
caspases in the HeLaTR cells overexpressing 4-1BB
Fig. 4. Effects of T3 on the activity of NF-jB. (A) Reporter luciferase
activities induced by T3. The cells of HeLaTR were transfected with

and secreted from HeLaTR cells, because the addition of
4-1BB-L to the culture media of the HeLaTR/4-1BB cells
further elevated the caspase activities. Further study is
necessary to clarify the involvement of 4-1BB-L in the
activation of caspases by T3. The biological significance of
the modulation of PSG7, fmfc and BMP-6 expression by T3
also awaits additional studies.
Up-regulation of 4-1BB and the subsequent activation of
the caspases occurred not only in HeLaTR cells but also in
rat FRTL cells that expressed only authentic TR, indicating
that the up-regulation of 4-1BB and the subsequent
activation of the caspases were not specific to the HeLaTR
cells but rather common cellular responses to T3. Further-
more, T3 induced DNA fragmentation and impaired the
proliferation of rat FRTL cells. This implies that T3 induces
apoptosis, which is mediated by 4-1BB, and subsequent
activation of caspases in the tissues and organs in the body.
Impairment of cell proliferation by T3 was more profound
in FRTL cells than in HeLaTR cells. As HeLaTR cells are
derived from tumor cells, it is possible that HeLaTR cells
express anti-apoptotic factors thereby becoming more
resistant to apoptotic stimuli. Nevertheless, both proapop-
totic effects and anti-apoptotic effects by 4-1BB have been
demonstrated; activation of 4-1BB abrogated the granulo-
cyte-macrophage colony-stimulating factor-mediated anti-
apoptosis in neutrophils and eosinophils [28,29], whereas
it induced Bcl-x
L
and Bfl-1 expression and promoted the
survival of CD8+ T lymphocytes [30]. Thus, proapoptotic

of TRAF1 in HEK293T cells and TRAF1-deficient mice
[31,32]. On the other hand, in HeLa cells, it was reported
that the expression of the full length TRAF1 potentiated
the activation of NF-jB induced by TNFa or IL-1, but
that of the C-terminal part of TRAF1 blocked it [33]. In
this study, T3 did not activate NF-jB of the HeLaTR
cells despite that the expression of both 4-1BB and
TRAF1 were strongly induced by T3 and that of TRAF2
remained at a high level, being unaffected by T3. Because
TRAF2A, a splice variant of TRAF2, was shown to inhibit
the NF-jB activation [34], one possibility is that HeLaTR
cells expressed TRAF2A. Nevertheless, the mechanism of
how 4-1BB induces apoptosis remains elusive. Cannons
et al. [35] reported that in T cell activation, 4-1BB recruits
TRAF2 and activates apoptosis signal-regulating kinase-1
(ASK1), which leads to downstream activation of JNK/
SAPK activation [35]. Therefore, it is possible that 4-1BB
activates ASK1 to induce apoptosis.
Acknowledgements
We thank T. Aono for technical assistance, K. Okumura for anti-
4-1BB-L monoclonal Ig TKS-1, and F. Ford for proofreading the
manuscript. This study is supported in part by a grant from the
Ministry of Education, Culture, Sports, Science and Technology, Japan
to T. Y O.
References
1. Spangenberg, D.B. (1971) Thyroxine induced metamorphosis in
Aurelia. J. Exp. Zool. 178, 183–194.
2. Shi, Y B. & Ishizuya-Oka, A. (2001) Thyroid hormone regulation
of apoptotic tissue remodeling: implications from molecular ana-
lysis of amphibian metamorphosis. Prog. Nucl. Acid Res. Mol.

induced involution of a thyrotrope tumor. Endocrinology 143,
347–359.
13. Yamada, H., Omata-Yamada, T., Wakabayashi-Ito, N., Carter,
S.G. & Lengyel, P. (1990) Isolation of recessive (mediator
–
)rev-
ertants from NIH 3T3 cells transformed with a c-H-ras oncogene.
Mol. Cell. Biol. 10, 1822–1837.
14. Yamada-Okabe, T., Doi, R. & Yamada-Okabe, H. (1996) Normal
and transforming ras are differently regulated for posttranslational
modifications. J. Cell. Biochem. 61, 172–181.
Fig. 6. Effect of T3 on 4-1 B B expression, caspase activities and prolif-
erationinratFRTLcells.(A) Expression of the mRNA for TRa,
4-1BB, BMP-6 and GAPDH. Total RNA was extracted from the
subconfluent cultures of FRTL cells that were cultured in the presence
or absence of T3 for 48 h. RT-PCR was performed with the specific
primers for the indicated genes and the total RNA as the template. The
PCR products collected after 30 cycles of PCR were analyzed by
agarose gel electrophoresis. (B) Caspase activities in the cell extracts
were determined by colorimetric assay with acetyl-DEVD-p-nitroani-
lide as a substrate. The cell extracts were prepared from FRTL cells
cultured in the presence or absence of T3 or Z-VAD-FMK for 48 h.
Absorbance at 405 nm, which represents the caspase activities, is
shown. (C) Effects of T3 on cell proliferation. Five hundred cells were
plated on 96-well microplates and cultured in the presence or absence
of T3 or Z-VAD-FMK. At 48 h, 20 lL of CellTiter 96 AQ
ueous
One
Solution Reagent was added to the cells, and the cells were further
incubated for 2 h. Absorbance

HGF, inhibits spreading and invasion of human pancreatic cancer
cell. Br. J. Cancer 84, 864–873.
22. Rickard, D.J., Hofbauer, L.C., Bonde, S.K., Gori, F., Spelsberg,
T.C. & Rigg, B.L. (1998) Bone morphogenetic protein-6 produc-
tion in human osteoblastic cell lines. J. Clin. Invest. 101, 413–422.
23. Jorns, A., Tiedge, M. & Lenzen, S. (2002) Throxine induces
pancreatic beta-cell apoptosis in rats. Diabetologia 45, 851–855.
24. Erkenbrack, D.E. & Rosenberg, L.L. (1986) Binding of thyroid
hormones to nuclear extracts of thyroid cells. Endocrinology 119,
311–317.
25. Shuford, W.W., Klussman, K., Tritchler, D.D., Loo, D.T.,
Chalupny, J., Siadak, A.W., Brown, T.J., Emswiler, J., Raecho,
H.,Larsen,C.P.,Pearson,T.C.,Ledbetter,J.A.,Aruffo,A.&
Mittler, R.S. (1997) 4-1BB costimulatory signals preferentially
induce CD8
+
T cell proliferation and lead to the amplification in
vivo of cytotoxic T cell responses. J. Exp. Med. 186, 47–55.
26. YeZ., Hellstrom, I., Hayden-Ledbetter, M., Dahlin, A., Ledbetter,
J.A. & Hellsgrom, K.E. (2002) Gene therapy for cancer using
single –chain Fv fragments specific for 4-1BB. Nat. Med. 8,
343–348.
27. Futagawa, T., Akiba, H., Kodama, T., Takeda, K., Hosoda, Y.,
Yagita, H. & Okumura, K. (2002) Expression and function of
4-1BB and 4-1BB ligand on murine dendritic cells. Int. Immunol.
14, 275–286.
28. Heinisch, I.V.W.M., Bizer, C., Volgger, W. & Simon, H U. (2001)
Functional CD137 receptors are expressed by eosinphils from
patients with IgE-mediated allergic responses but not by eosino-
phils from patients with non-IgE-mediated eosinophilic disorders.

(1999) Role of the stress kinase pathway in signaling via the T cell
costimulatory receptor 4-1BB. J. Immunol. 163, 2990–2998.
Ó FEBS 2003 T3 induces 4-1BB in a TR-dependent manner (Eur. J. Biochem. 270) 3073