(Received for publication, July 24, 1996, and in revised form, December 2, 1996)
From the Laboratory of Molecular Biology, The present study evaluated the expression and
regulation of endogenous thyroid hormone receptors (TRs) in cultured
cells. In COS-1 cells, the endogenous TR, subtype Thyroid hormone nuclear receptors
(TRs)1 are members of the steroid
hormone/retinoic acid receptor superfamily. Two TR genes, TR One important but less well studied mode of modulation of TR activity
is the regulation of TR expression at the protein level. TR Monkey kidney COS-1 cells (8.2 × 105 cells/100-mm dish) were plated. Forty-eight h later,
cells were preincubated in methionine or phosphate-free medium for 90 min and incubated with 100 µCi of [35S]methionine (ICN
Biomedicals, Inc., Costa Mesa, CA) or 1 mCi (74 MBq) of carrier-free
[32P]orthophosphate (Amersham Corp.) in the presence of
OA (100-250 nM) for 3 h. For phosphorylation and
metabolic labeling of the transfected h-TR Gel proteolysis of 35S-labeled TR Two-dimensional analysis of phosphoamino acids and
tryptic mapping was carried out according to Ref. 20.
COS-1 cells (2 × 106 cells/15-cm dish)
were cultured in the presence or absence of 100 nM OA at
37 °C for 3 h. Nuclei were isolated according to Samuels
et al. (21) and Zhao and Padmanabha (22).
[125I]T3 binding was carried out as described
by Samuels et al. (21) and Koerner et al. (23).
Briefly, cells were scraped in buffer H (20 mM Hepes, pH
7.5, 5 mM KCl, 0.5 mM MgCl2, and
0.5 mM dithiothreitol) with or without 100 nM
OA, and cells were pelleted at 800 × g for 10 min at
4 °C. Cells were lysed by passing back and forth (27 times) through
a 1-ml syringe with a 25 × 1.56-cm needle. The lysed cell
suspension was made isotonic by adding 1 ml of 0.5 M
sucrose, and the nuclei were pelleted at 1,200 × g for
10 min. The nuclei was washed two times by resuspending in 2 ml of binding buffer B (0.32 M sucrose, 3 mM
MgCl2 and 20 mM Tris, pH 7.8). The washed
nuclei were incubated with 0.5 nM
[125I]T3 in 200 µl of binding buffer B for
30 min at 37 °C. The binding was also carried out in the presence of
1 µM unlabeled T3 to determine the
nonspecific binding. At the end of the incubation, an equal volume of
1% Triton X-100 was added, and the nuclei were pelleted. The nuclei
were washed further with 1 ml of binding buffer B containing 0.5%
Triton X-100, and the radioactivity bound to nuclei was determined in a
COS-1 cells (3 × 105/60-mm dish)
were plated 1 day before transfection. The cells were transfected with
1.5 µg of a reporter gene containing two copies of the palindromic
TRE in tandem (pTK28m) or the chicken lysozyme TRE (pTKLys) upstream of
the chloramphenicol acetyltransferase gene. pBluescript plasmid
(Stratagene, La Jolla, CA) was used to adjust the total DNA to 3 µg/dish. Various concentrations of OA were added to the cells 3 h before the cells were lysed for analysis of chloramphenicol
acetyltransferase activity, which was normalized to protein
concentration.
Total RNA was isolated from COS-1 cells
cultured in the presence or absence of 100 nM OA (RNeasy,
Qiagen, Chatsworth, CA). Fifteen µg of RNA was electrophoresed on a
1% formaldehyde gel and blotted onto a nylon membrane, which was
prehybridized at 68 °C with QuikHyb solution (Stratagene) for 3 h, followed by hybridization at 65 °C for 6 h with a
radioactive rat TR 2
µg of total RNA was reverse transcribed using Moloney murine leukemia
virus reverse transcriptase (Clontech, Palo Alto, CA) after treatment
with RNase-free DNase I (Boehringer Mannheim). 5-20 µl of the
resultant 100-µl cDNA reaction was used for PCR to ensure
linearity of response. Primers identical to sequences common to human
and mouse TR COS-1
cells were transfected with h-TR When COS-1 cells
were treated with 250 nM OA, we found that a protein band
with a molecular weight identical to that of transfected h-TR
To support further the conclusion that the protein induced by OA was
indeed TR mAb C4 recognizes both TR
We further examined the effect of the concentration of OA, ranging from
25 to 250 nM, on the induction of TR Previously we have
shown that transfected h-TR
Whether the TR
Comparison of [125I]T3 binding to isolated nuclei of
COS-1 cells cultured in medium with or without OA
Molecular and Cellular Endocrinology
Branch,
1 (TR
1), but
not subtype
2 or
1, was induced to express by okadaic acid (OA) in a concentration-dependent manner. The induced TR
1 had
immunoreactivity and partial V8 proteolytic maps similar to those of
the transfected and in vitro translated human TR
1
(h-TR
1). The OA-induced expression of endogenous TR
1 was,
however, not observed in a variety of other cultured cell lines tested,
indicating that the induction was cell type-dependent.
TR
1 induced by OA was a multisite phosphorylated protein, in which
serine and threonine in a ratio of 10:1 were phosphorylated. The
induced TR
1 was functional as it could mediate the thyroid
hormone-dependent transcriptional activity via several thyroid hormone response elements. The induction of endogenous TR
1
expression by OA was not accompanied by an increase in mRNA levels
but was the result of an increase in the stability of the TR
1
protein. This is the first report to indicate that one of the
mechanisms by which the TR isoforms are differentially expressed is via
the tissue-specific stabilization of the TR isoform proteins. Furthermore, this selective stability of TR
1 could be conferred by
phosphorylation.
and
TR
, have been identified which are located on chromosome 17 and
chromosome 3, respectively. Four isoforms,
1,
2,
1, and
2,
are generated from each of two TR genes,
and
, by alternative splicing (1, 2). They are transcription factors that regulate the
expression of target genes by interacting with specific DNA sequences
known as thyroid hormone response elements (TREs) in the promoter
region of target genes (1, 2). The transcriptional activity of TRs is
not only dependent on thyroid hormone
3,3
,5-triiodo-L-thyronine (T3) but also on the
type of TRE. Recent studies have indicated that the transcriptional
activity of TRs is further modulated via interaction with four groups
of cellular proteins: (a) members of the nuclear receptor
superfamily, notably the retinoid X receptors (1, 2); (b)
corepressors including p270/N-CoR (3), SMRT (4), TRUP (5), SHP (6), and
TRACs (7); (c) coactivator SRC-1 (8); and (d) the
tumor suppressor p53 (9). It is clear that regulation of the
transcriptional activity of TRs is more complex than envisioned
previously.
1 and
TR
1 are differentially expressed at different developmental stages
and at different levels in various tissues (10-12). Moreover, using
the specific T3 binding activity and immunoreactivity as measures of the expressed receptor proteins, it was found that there
are marked variations in the TR protein:mRNA ratios in different tissues (11, 12), suggesting isoform- and tissue-dependent posttranslational modification of TRs. However, it is unknown how TRs
could be modulated at the protein level. We studied this problem by
first using the cultured rat growth hormone-producing cell line GH3,
which has long been used as a model cell line for studying thyroid
hormone action (13). Unfortunately, preliminary experiments indicate
that the anti-TR antibodies currently available are not sensitive
enough to detect the endogenous TRs in GH3 cells either by
immunoprecipitation or Western blotting. However, we subsequently
discovered that in COS-1 cells, endogenous TR
1 could be induced to
express by okadaic acid (OA), a potent and specific inhibitor of
serine/threonine phosphoprotein phosphatases 1 and 2A (14). COS-1 cells
are known to be functionally deficient in TRs and have been used widely
as recipient cells for gene transfer to study TR action. We therefore
characterized the process of induction of TR
1 by OA as we hoped to
use COS-1 cells as a model cell line to understand the regulation of
expression of endogenous TRs. Thus, in the present paper, we studied
the molecular basis of the induction of endogenous TR
1 expression
and found that the OA-induced TR
1 was a phosphoprotein possessing
the same characteristics and functions as transfected human (h)-TR
1
in mediating thyroid hormone-dependent transactivation.
Moreover, the induction of endogenous TR
1 protein expression was
cell type-dependent in that it did not occur in GH3 (13),
neuro-2
(16), human choriocarcinoma JEG-3, human epidermal carcinoma
A431, mouse fibroblast NIH3T3, and the human cervical carcinoma HeLa
cells. The increase in TR
1 protein expression by OA in COS-1 cells
was not accompanied by increases in mRNA level or in the rate of
synthesis of TR
1 protein. Rather, our results showed that the
increase in TR
1 protein expression by OA was the result of increased
stability of the TR
1 protein most probably via phosphorylation and
defined a novel role of phosphorylation in regulating the levels of
TR
1 in different cell types.
Immunoprecipitation
1, cells were transfected
with 8 µg of h-TR
1 expression vector pCLC51 (16) by using
LipofectAMINE according to the manufacturer's instructions (Life
Technologies, Inc.). After labeling with either
[35S]methionine or [32P]orthophosphate,
labeled TR
1 were immunoprecipitated with 5 µg of monoclonal
antibody (mAb) C4 (17) or J52 (18) as described by Lin et
al. (16). Identical conditions were used for detection of
endogenous TR
1 in GH3, neuro-2
, JEG-3, A431, NIH3T3, and HeLa
cell lines.
1 was
performed according to Cleveland et al. (19). Briefly, the
bands containing labeled TR
1 were excised and loaded into the sample
wells of a 15% acrylamide gel. Three µg of Staphylococcus
aureus V8 protease (Boehringer Mannheim) in soaking solution was
overlaid. Electrophoresis was performed, and the dried gel was
autoradiographed.
-counter. The nuclear protein concentrations were determined by
Coomassie Plus Protein Assay Reagent (Pierce Chemical Co.) using
aliquots of nuclei suspension.
1 on
Transcription
1 cDNA probe prepared from a rat-TR
1
expression plasmid (gift of Dr. R. Koenig, University of Michigan
Medical Center). A human
-actin cDNA probe was also prepared as
an internal control. The membrane was washed under stringent conditions
with 0.2 × SSPE at 60 °C for 10 min. The labeled filters were
analyzed by exposure to x-ray film, stripped, and reprobed with the
-actin cDNA.
1 were used to amplify a 378-base pair cDNA
fragment. h-TR
1 was also used as a template for PCR to serve as a
control. PCR products were assessed by Southern analysis using the
above rat TR
1 cDNA probe.
1 Protein
1 expression vector (pCLC51) and
incubated in methionine-free medium in the presence or absence of OA.
Cells were pulse labeled with 50 µCi of [35S]methionine
for 30 min and chased for various time periods. Cell lysates were
prepared, and immunoprecipitation was performed as described above.
Cell Type- and TR Isoform-dependent Induction of
Expression of Endogenous TR1 Protein by OA
1 was
immunoprecipitated with monoclonal anti-h-TR
1 J52 (mAb J52; 19)
(lane 8 versus lane 9 of Fig. 1A).
This band was not present when the cells were not treated with OA as
shown in lane 1 (Fig. 1A). As a control for
specific expression of transfected h-TR
1, lane 2 shows
that when an identical expression plasmid lacking the coding sequence
of h-TR
1 was used, no expression of h-TR
1 was detected.
Transfection of this TR
1-less vector had no effect on the induction
of TR
1 in COS-1 cells by OA, as the intensity of the TR
1 band
seen was identical to that in its absence (lane 8 versus lane
7). The intensity of the TR
1 band in the presence of OA
(lane 9) was slightly higher than that in the absence of OA
(lane 3), reflecting the combined intensities of the
endogenous and the transfected TR
1. The minor lower bands in
lanes 3 and 9 probably represented the
degradation products. The epitope for mAb J52 is in the middle of the
A/B domain (18). To be certain that the band detected in lane
7 was TR
1, we also used another monoclonal antibody, C4, whose
epitope is located at the COOH-terminal
Glu457-Val-Phe-Glu-Asp461 (17). As shown in
lane 10, a protein band with the same electrophoretic mobility as for transfected h-TR
1 (lanes 12 and
6) was detected which was not seen in the absence of OA
(lane 4). Lanes 13-15 are controls using an
irrelevant mAb that did not immunoprecipitate TR
1 to confirm the
specificity of TR
1 bands seen in lanes 7 and
10.
Fig. 1.
Panel A, effect of OA on the expression
of endogenous TR1 in COS-1 cells. Untransfected COS-1 cells (2 × 106 cells/100-mm dish) (lanes 1,
4, 7, 10, and 13) and cells
containing transfected h-TR
1 expression vector (lanes 3,
6, 9, 12, 15) or empty
vector (lanes 2, 5, 8, 11,
and 14) were labeled with 100 µCi
[35S]methionine in the absence of OA (lanes
1-6) or in the presence of 250 nM OA (lanes
7-15). After lysis of cells, immunoprecipitation was carried out
with mAb 52 (lanes 1-3, 7-9), mAb C4
(lanes 4-6, 10-12), a control antibody MOPC
(lanes 13-15). Panel B, partial peptide maps of
TR
1. The gel slices containing immunoprecipitated endogenous TR
1
(lane 3), transfected h-TR
1 (lane 2), in
vitro translated h-TR
1 (lane 4), and the cytosolic thyroid
hormone-binding protein, p58 protein (lane 1) were digested
with S. aureus V8 protease according to "Experimental
Procedures."
[View Larger Version of this Image (26K GIF file)]
1, we carried out one-dimensional partial V8 protease
digestion. The peptide map of the immunoprecipitated [35S]methionine-labeled TR
1 derived from OA-treated
cells (Fig. 1B, lane 3) was compared with the
maps obtained from either transfected h-TR
1 (lane 2) or
in vitro translated h-TR
1 (lane 4). An
identical pattern of four peptides, a, b, c and d, with molecular
masses of 14, 12.5, 11, and 9 kDa, respectively, was seen in
lanes 2-4. The weaker band (peptide a) and the stronger
band (peptide d) seen in lane 3 compared with those in
lanes 2 and 4 probably were due to more extensive
proteolysis of peptide a to peptide d in TR
1 derived from COS-1
cells. The differences in the protease susceptibility of peptide a
could reflect the differences in local aggregated charge or structure
surrounding the V8 cleavage sites due to minor differences in the
primary sequences of human and monkey TR
1. However, a completely
different protein (p58, a cytosolic thyroid hormone-binding protein;
24) yielded a completely different peptide map, as shown in lane
1. Therefore, based on the immunoreactivity and peptide mapping,
we have clearly established that in COS-1 cells, OA induced the
expression of TR
1.
2 and TR
1 because of the common
conserved epitope at the COOH terminus. However, examination of lanes 10 and 11 of Fig. 1A further
shows that no protein band larger or smaller than TR
1 (molecular
mass ~55,000) was detected, indicating that neither TR
2 (molecular
mass ~58,000) nor TR
1 (molecular mass ~47,000) was induced to
express by OA. To assess whether OA could also affect the expression of
TR
1 in other cell types, we similarly treated the cells derived from
different tissues with OA. Fig. 2A shows that
no OA-induced expression of TR
1 was detected in GH3, neuro-2
,
JEG-3, A431, NIH3T3, and HeLa cells. These results indicate that the
induction of the expression of TR
1 was TR isoform- and cell
type-dependent.
Fig. 2.
Panel A, effect of OA on the expression
of endogenous TR1 in different cell types. Cells were labeled in the
absence (odd numbered lanes) or presence of 100 nM OA (even numbered lanes), and the lysates
were immunoprecipitated using mAb C4. Panel B, dose
dependence of OA-induced TR
1 expression. COS-1 cells were labeled as
described in Fig. 1A either in the absence of OA (lane 1) or in the presence of 25, 50, 100, and 250 nM OA
for lanes 2, 3, 4, and 5,
respectively. Immunoprecipitation using mAb C4 was carried out as
described in Fig. 1A.
[View Larger Version of this Image (27K GIF file)]
1 in COS-1 cells. Lane 2 of Fig. 2B shows that at 25 nM, no TR
1 was induced. At 50 nM (lane
3), a low level of TR
1 was expressed, and a maximal level of
TR
1 was expressed at 100 nM of OA (lane 4). A
further increase in the OA concentration (250 nM) led to a
fall in the expression of TR
1.
1 Induced by OA Was a Phosphoprotein
1 is a phosphoprotein and that
phosphorylation is enhanced by OA (16, 25). Therefore we asked the
question of whether the induced TR
1 was also a phosphoprotein. Fig.
3 shows that endogenous TR
1 induced by OA was a
phosphoprotein. The extent of phosphorylation was similar at 100 nM (lane 2 of Fig. 3) and 250 nM of
OA (lane 3 of Fig. 3). Lane 4 is a negative
control to indicate that when an irrelevant antibody was used, no
phosphorylated TR
1 was immunoprecipitated. However, it is important
to point out that the TR
1 induced at 250 nM OA was about
1/25 of that at 100 nM (see Fig. 2 B).
Therefore, the extent of phosphorylation of TR
1 at 250 nM would be expected to be higher than that seen for TR
1
induced at 100 nM OA (lane 3 versus lane 2 of
Fig. 3). These results are consistent with earlier observations that
phosphorylation of transfected h-TR
1 was enhanced when the OA
concentration was increased from 100 to 250 nM (16).
Similar to the transfected h-TR
1 (16), phosphorylated serine and
threonine in a ratio of 10:1 were detected for the induced endogenous
TR
1 (data not shown). Two-dimensional tryptic mapping indicates one
major phosphopeptide, corresponding to 60% of total counts, and five
other minor phosphopeptides (data not shown), suggesting that multiple
phosphorylation sites were present in phosphorylated TR
1 induced by
OA.
Fig. 3.
Effect of OA on the phosphorylation of
endogenous TR1 in COS-1 cells. COS-1 cells were labeled with 1 mCi of [32P]orthophosphoric acid in the absence of OA
(lane 1) or in the presence of 100 nM
(lane 2) and 250 nM (lanes 3 and
4) OA. After lysis of cells, immunoprecipitation was carried
out using mAb C4 (lanes 1-3) or control mAb MOPC
(lane 4).
[View Larger Version of this Image (18K GIF file)]
1-mediated Transcription
1 induced by OA
was functional was first assessed by T3 binding. Nuclei
were isolated from COS-1 cells cultured in the presence or absence of
OA and incubated with [125I]T3. As shown in
Table I, the nuclei isolated from cells treated with OA
had nearly 2-fold more T3 binding than those not treated with OA, indicating that OA-induced TR
1 was functional as a
T3 binder.
-counter, and the proteins were determined as described under
"Experimental Procedures." The nonspecific binding was obtained by
incubating nuclei with 0.5 nM [125I]T3 in
the presence of 1 µM unlabeled T3. The
nonspecific binding was ~15% of total binding. Data are averages of
three independent experiments each with duplicates (mean ± S.D.,
n = 3).
OA
Specific
[125I]T3 bound
fmol/50 µg nuclear
proteins
0.34 ± 0.06
+
0.58
± 0.07
We further examined the transcriptional activity of OA-induced
endogenous TR1. Two reporter genes, one containing palindromic TRE
and the other containing chicken lysozyme TRE, were transfected into
cells. As shown in Fig. 4, the
T3-dependent transcriptional activity of TR
1
was OA-dependent. An approximately 1.8- and 2.5-fold increase in the transcriptional activity on chicken lysozyme TRE was
observed at 100 and 250 nM OA, respectively. On palindromic TRE, the increase was slightly lower (~1.5- and 1.8-fold increase at
100 and 250 nM, respectively). These findings are similar
to that observed previously of a 2-fold increase in the transcriptional activity of the transfected TR
1 by OA (17). Thus, the induced TR
1
was functional, and its transcriptional activity was enhanced by
phosphorylation.
mRNA Level of TR
The induction of
expression of TR1 protein by OA may occur at either the level of
transcription or translation. To understand the molecular basis of the
induction, we first determined the TR
1 mRNA levels in cells
treated with or without OA by Northern analysis. Although it was
possible to detect TR
1 mRNA from total RNA (Fig.
5A, lane 1) and mRNA
(lane 2) prepared from control cells (GH3), no TR
1
mRNA was detectable from total RNA (lanes 3 and
4) and mRNA (lanes 5 and 6)
prepared from COS-1 cells in the presence (lanes 4 and
6) or absence (lanes 3 and 5) of OA.
-Actin mRNA was detectable in GH3 cells (lane 2 of
Fig. 5B) and also in COS-1 cells either from total RNA
(lanes 3 and 4 of Fig. 5B) or at a
much greater abundance using the mRNA (lanes 5 and 6 of Fig. 5B), confirming the presence of intact
RNA. Comparing the intensities of the
-actin bands in lanes
3 and 4 or 5 and 6, it can be
seen that OA had no effect on the expression of
-actin mRNA.
In light of the inability to detect TR1 mRNA in COS-1 cells by
Northern analysis, we performed reverse transcriptase-PCR, and the
resultant cDNA was amplified by PCR and identified by Southern
analysis. In preliminary experiments, we had first titrated the amounts
of RNA so that RNA used in reverse transcriptase-PCR was in the linear
range, and the resultant cDNA can be compared quantitatively.
Lane 1 of Fig. 5C shows that at 25 nM
OA, TR
1 mRNA was similar to that without OA treatment. Further
increases in OA concentration led to a dose-dependent
reduction of mRNA up to 100 nM, which then leveled off
at 250 nM. Lanes 6-10 are controls to indicate
that when no reverse transcription was carried out, no band was
detected. These results indicate that OA affected the expression of
endogenous TR
1 at the mRNA level.
Because the
expression level of endogenous TR1 was virtually nondetectable under
the present experimental conditions (Fig. 1A and lane
1 of Fig. 2A), it was not possible to compare the effect of OA on the expression of endogenous TR
1 at the translation level. However, since we have shown that the characteristics and functions of endogenous TR
1 were indistinguishable from those of
transfected TR
1 (see above), we used transfected TR
1 to study the
effect of OA on the stability of TR
1. The stability of TR
1 was
examined by a brief pulse with [35S]methionine followed
by a chase for different lengths of time in the presence (Fig.
6A, lanes 8-14) or absence of 100 nM OA (lanes 1-7). Comparison of the
intensities of labeled TR
1 at each time point indicates that TR
1
from cultures in the presence of OA was more stable than without OA
treatment. Quantitation analyses indicate that
t1/2 of TR
1 was ~1.88 ± 0.07 and
0.96 ± 0.02 h (Fig. 6B) in the presence of OA or
absence of OA, respectively, indicating that treatment of cells with OA
led to a 2-fold increase in the stability of TR
1.
We further examined the effect of OA on the rate of synthesis of
transfected TR1 by labeling cells with [35S]methionine
for various lengths of time in the presence (Fig. 6C,
lanes 7-12) or absence (lanes 1-6) of OA. The
intensities of bands were quantified, and the rates of synthesis were
calculated to be 3,546 ± 607 and 3,402 ± 250 counts/min for
TR
1 synthesized in the absence or presence of OA, respectively.
These data indicate that OA had no effect on the rate of synthesis of
TR
1, suggesting that the increased stability of the TR
1 protein
was the major factor responsible for the OA induction of the TR
1
protein.
The mRNAs of the TR isoforms and their encoded receptor
proteins are known to be expressed differentially in a tissue-specific manner. The ratios of the mRNA versus the protein levels
vary widely from tissue to tissue (11, 12). Moreover, the timing of the
appearance of the various TR isoforms and the ratios of these isoforms
are regulated tightly during development (10, 11). However, the
mechanisms underlying this tissue-specific and developmental regulation
of TRs are largely unknown. In the present study we discovered that the
expression of endogenous TR1 protein, but not TR
2 or TR
1,
could be induced by OA in COS-1 cells, which are believed to be
functionally deficient in response to T3. However, this
induction is unique among the cell lines examined including
T3 functionally competent GH3 cells (13) and neuro-2
cells (15) and T3 functionally deficient cells (JEG-3,
NIH3T3, HeLa, and A431 cells). Therefore, our findings highlight a
potentially novel mechanism explaining the tissue-specific expression
of TR isoforms.
Our data indicate that the molecular mechanisms of the induction of
TR1 expression by OA were complex. The induction was mediated at
both the transcriptional and translational levels. The amounts of
mRNA of TR
1 were reduced as the concentration of OA was
increased. However, paradoxically, although OA treatment led to a
reduction of mRNA, the TR
1 proteins were increased by OA except
at a high concentration of 250 nM. The sharp increase in
the level of TR
1 protein from 50 to 100 nM (Fig.
2B) was not proportional to the corresponding decrease in
mRNA (Fig. 5C), reminiscent of what has been observed in
the in vivo studies reported by Schwartz et al.
(12) in that there are large variations in the mRNA:TR protein
ratios, suggesting additional regulation at the posttranslational
level. Indeed, we found that although OA had no effect on the synthesis
of TR
1, OA increased its stability by nearly 2-fold. Therefore, the
increase in the stability of TR
1 plays a major role in the function
of TRs. This is the first clear report demonstrating that one of the
mechanisms of the regulation of TR expression is at the
posttranslational level.
The reasons for the increased stability of TR1 protein by OA are not
entirely clear. OA is an inhibitor of phosphoprotein phosphatase 1 and
2A and has been used widely as a tool to study the functions of
phosphorylation in cellular processes (14). We have demonstrated that
OA not only induced TR
1 expression but also increased dramatically
the extent of phosphorylation. The induced TR
1 was a phosphoprotein
that was phosphorylated at multiple Ser and Thr sites. Although the
phosphorylation sites are unknown, it is reasonable to assume that
phosphorylation of Ser or/and Thr residues by the treatment of cells
with OA could change the conformation of TR
1, thereby decreasing its
susceptibility to proteolytic degradation. Phosphorylation-induced
structural modification is not without precedent. Structural changes in
glycogen phosphorylase have been studied by x-ray crystallography (26). As a result of phosphorylation of Ser-14, the interaction between subunits is strengthened, and the binding sites for allosteric effectors and substrates are altered (26). However, conformational changes are not obligatory for enhanced susceptibility of TR
1 to
proteolytic degradation. The increased phosphorylation of Ser or/and
Thr sites could alter the interaction of TR
1 with cellular proteases, thus altering the kinetics of degradation. Alteration of the
stability of proteins by phosphorylation has been demonstrated directly
for human I
B-
(27), nicotinic acetylcholine receptors (28), and
the Drosophila yan protein (29). Similar to the present
studies of TR
1, the changes in the stability of these proteins led
to functional consequences (27-29).
Phosphorylation of TRs is not only limited to the 1 subtype. Chicken
TR
1 (30, 31) and rat TR
2 (32) were also found to be
phosphorylated. Therefore the different extents of phosphorylation at
different sites in TR isoforms could affect their stability differently. The present study thus raises the interesting possibility that variations in the protein stability of TR
1 may be one mechanism by which the tissue-specific and development-specific expression of the
TR isoforms are regulated. Although it remains to be established that
the stabilities of other TR isoforms are regulated by phosphorylation, what is clear from our study is that differential phosphorylation of TR
isoforms in different cell types adds yet another level of fine tuning
to the already multifaceted and complex mechanisms of regulation of
TR-dependent gene transcription.