| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Division of Genetics, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02115
Address all correspondence and requests for reprints to: Dr. S. G. Ball, Department of Medicine, University of Newcastle, Newcastle upon Tyne, United Kingdom NE2 4HH. E-mail: nsgb%ALBA{at}newcastle.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There are three TR isoforms, encoded by two independent genes on
separate chromosomes. The TR
gene encodes a single functional
receptor, TR1. In contrast, the TRß gene encodes two receptors,
TRß1 and TRß2, generated through differential promoter choice and
alternative splicing of 5'-exons (6). The TRs have conserved
carboxyl-terminal ligand-binding domains, but unique amino-termini.
They differ in relative tissue distribution; TR1 levels are highest
in skeletal muscle and brown adipose tissue, whereas TRß1 is most
prevalent in brain, liver, and kidney. In contrast, TRß2 is
restricted largely to the pituitary (7), with low levels also expressed
in hypothalamus, the developing retina, and the auditory system
(8, 9, 10). In addition, TRs are differentially regulated by
T3 and TRH, with TRß2 messenger RNA (mRNA)
being markedly down-regulated (11, 12).
The conserved diversity, restricted tissue distribution, and differential regulation of TRs raise the possibility of isoform-specific functions. Initial data indicated that the three TRs have similar T3-binding, DNA-binding, and trans-activation properties (13). However, there are data supporting the differential regulation of defined promoters by TR isoforms. TRß1 selectively modulates the activities of the TRH and myelin basic protein promoters in heterologous transfection systems (14, 15). Furthermore, treatment of GH3 cells with sodium butyrate and TRH specifically decreases TRß2 levels, producing a coincident fall in T3-stimulated GH gene expression (12, 16). However, there are several limitations to these data. Specificity studies based on the overexpression of TRs are difficult to interpret in a physiological context. Furthermore, the use of heterologous reporter constructs may result in effects not seen in the context of natural promoters. In addition to these problems, sodium butyrate and TRH, used in previous studies to modulate specific TR expression, have indirect actions and may affect several points in the T3 response pathway.
Antisense RNAs can extinguish the expression of a target gene by stimulating mRNA degradation and/or blocking its translation (17). Such an approach has been applied to abrogate expression of the glucocorticoid receptor in vitro (18). As the TR isoforms have unique amino-termini, they are ideal targets for isoform-specific antisense RNAs directed against the isoform-specific sequences of the corresponding mRNAs. GH3 cells express all three functional TRs. The T3-stimulated GH response of this cell line is well characterized and represents a physiologically relevant phenomenon. In the present study, we have addressed the roles of individual TRs in GH gene expression using an isoform-specific antisense RNA to produce a specific TR deletion in GH3 cells. We present data on the effect on GH gene expression of TRß1 deletion by this approach.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Production of stable transfected cell lines expressing TRß1
antisense RNA and control constructs
Three GH3-derived cell lines were produced through
stable transfection. In addition to one expressing the TRß1 antisense
construct (B1 a/sense), two control cell lines were established: one
containing the expression vector alone (vector control), and a second
expressing the TRß1 sense construct (B1 sense). GH3 cells
were grown to 70% confluence in DMEM (Life Technologies, Grand Island,
NY) and 10% FBS (Flow Laboratories, Rockville, MD) and transfected
with the TRß1 antisense construct, control sense construct, or vector
alone by electroporation (220 V; 960 microfarads; GenePulsar, Bio-Rad
Laboratories, Richmond, CA). Approximately 40 µg plasmid DNA were
used per T150 flask. Cells were grown in DMEM-10% FBS for 48 h
after electroporation, and then transferred to selective medium
containing 600 µg/ml G418 (Life Technologies). After continuous
growth in selective medium for 8 weeks, stable transfected clones were
expanded and subsequently maintained under G418 selection. To avoid
potential position artifacts from dissimilar nonhomologous
recombination of the transfected vector in different clones, data are
typical of those obtained from several different clones of each
respective cell line.
Cell culture
Wild-type GH3 cells were maintained in DMEM
supplemented with 10% FBS. All other cells were maintained in similar
medium that also contained 600 µg/ml G418. All cell lines had similar
morphology and growth characteristics. For experiments in which the
effect of T3 was assessed, cells were grown in
medium supplemented with 10% hormone-stripped FBS for 48 h before
experimentation, with medium changed daily. T3
(Sigma Chemical Co., St. Louis, MO) was added to achieve a final
concentration of 5 nM. Vehicle was added to the appropriate
control incubations. Incubations with hormone or vehicle were conducted
for 24 h. All experiments were performed in cells grown to 70%
confluence.
Northern blot analysis of TR and GH mRNAs
GH3 cells were harvested in 4 M
guanidinium thiocyanate, 0.5% sodium n-lauryl sarcosine, 25
mM sodium citrate, and 0.1 M mercaptoethanol.
RNA was prepared by centrifugation through 5.7 M cesium
chloride. The pellet was washed twice with 0.3 M sodium
acetate, ethanol precipitated on each occasion, and finally resuspended
in ribonuclease-free distilled water. The purity and concentration of
RNA were assessed by optical density
(OD260/OD280) and integrity of 18S and 28S
ribosomal RNA after electrophoresis.
GH3 cell RNA (5 µg) was subjected to electrophoresis through 1% agarose containing 1.8% (vol/vol) formaldehyde. The running buffer was 1 x MOPS (3-N[morpholino]propanesulfonic acid; Sigma). After electrophoresis, RNA was transferred to Duralon membranes (DuPont, Wilmington, DE). Hybridization with riboprobes was performed in 50% formamide, 5 x SSC, 5 x Denhardt’s solution [2% (wt/vol) polyvinyl pyridoline, 2% (wt/vol) BSA, and 2% (wt/vol) Ficoll 400], 5% SDS, and 500 µg/ml yeast RNA at 65 C. Membranes were washed in 0.2 x SSC (standard saline citrate)-0.1% SDS, twice at room temperature and twice at 65 C. For incubations with cDNA probes, membranes were prehybridized in 50% formamide, 5 x SSC, 10 x Denhardt’s solution, 50 mM sodium phosphate (pH 6.7), 1% SDS, and 1 µg/ml salmon sperm DNA (Sigma). Subsequent hybridization was performed in 50% formamide, 5 x SSC, 5% dextran sulfate, 20 mM sodium phosphate (pH 6.7), 1 x Denhardt’s solution, 0.5% SDS, and 20 µg/ml salmon sperm DNA at 45 C. Membranes were washed successively at room temperature in 1% SDS-1 x SSC, 0.5% SDS-0.5 x SSC, 0.1% SDS-0.1 x SSC twice, and finally at 50 C in 0.1% SDS-0.1 x SSC.
32P-Labeled riboprobes for rat GH and rat TRß2 were
prepared by standard procedures from the respective cDNAs subcloned in
pBS-KS (7, 19). The proximal 300-bp isoform-specific coding region of
the rat TRß1 gene was PCR amplified from the full-length TRß1
sequence contained in the plasmid CDM (Dr. R. Koenig, University of
Michigan, Ann Arbor, MI) and directionally subcloned into pBS-KS. This
plasmid, rTRß1sp, was linearized and used to generate a complementary
RNA riboprobe. A [32P]cDNA probe for rat TR1 was made
by the random hexamer method, using the full-length TR1 cDNA
subcloned in pCDNA1/Amp as a template (20). The specificities of the TR
constructs used have been demonstrated previously (7, 11). A
32P-labeled cDNA cyclophilin probe was made by the random
primer method, using a 900-bp cyclophilin sequence in pBS (21).
GH and TR mRNA expression were quantified by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) and corrected for loading and transfer by normalizing to a nonregulated cyclophilin mRNA signal. Statistical analysis was performed by Student’s t test.
Preparation of GH3 cell nuclear
extracts
GH3 cell nuclear extracts were prepared
as previously described (22). Nuclear extracts from individual cell
lines were prepared from four T-150 culture flasks of cells, grown to
70% confluence. Briefly, cells were lysed in 3 vol buffer containing
20 mM HEPES (pH 7.8), 1.5 mM MgCl2,
40 mM KCl, 0.5 mM dithiothreitol (DTT), and 0.1
mM phenylmethylsulfonylfluoride (PMSF). Ten percent
(vol/vol) Nonidet P-40 was added to a final concentration of 0.5%. The
suspension was incubated on ice for 10 min. Nuclei were then isolated
by centrifugation (15,000 x g, 30 sec, 4 C). The
nuclear pellet was raised in an equal volume of buffer containing 20
mM HEPES (pH 7.8), 0.6 M KCl, 0.02
mM ZnCl2, 0.2 mM EDTA, 0.5
mM DTT, 0.1 mM PMSF, and 1 µg/µl pepstatin
and incubated on ice for 30 min. After additional centrifugation
(15,000 x g, 20 min, 4 C), the resultant supernatant
was dialyzed for 4 h against buffer containing 20 mM
HEPES (pH 7.8), 5 mM mercaptoethanol, 10% (vol/vol)
glycerol, 50 mM NaCl, 2 mM EDTA, and 0.1
mM PMSF. The protein concentration was determined by
standard methods, and aliquots were rapidly frozen and stored at -80
C. The extraction procedure generated nuclear extract protein yields of
14–18 µg/µl, which were similar for all cell lines. Nuclear
extracts were used within 2 weeks of preparation.
Binding of
[125I]T3 to nuclear
extracts
T3 binding was performed as previously
described (23). Briefly, 40 µg nuclear extract were incubated for
18 h at 4 C with 2 nM
[125I]T3 (2200 Ci/mmol;
DuPont-New England Nuclear Research Products, Boston, MA) in a total
reaction volume of 150 µl KMTD [0.3 M KCl, 1
mM MgCl2, 10 mM Tris-HCl (pH 8.0),
and 1 mM DTT]. Dowex AG-1x8 was used to separate bound and
free ligand, and nonspecific binding was measured in the presence of a
500-fold molar excess of unlabeled ligand. Statistical analysis was
performed using Student’s t test.
Electrophoretic mobility shift assay (EMSA) for identification of
individual TR isoform expression at a protein level
Inverted palindrome TRE (F2) DNA fragments were end labeled with
T4 polynucleotide kinase in the presence of [
-32P]ATP.
Gel-purified probes were incubated with 2.5 µg nuclear extract at
room temperature for 30 min in the presence of 100 mM
Tris-HCl (pH 7.5), 50 mM NaCl, 5% (vol/vol) glycerol, 1
mM EDTA, and 1 mM DTT (reaction volume, 20
µl). The TR isoforms in the nuclear extract binding the
oligonucleotide probe were identified by supershifting the protein-DNA
complexes with TR isoform-specific antisera. One to 2 µl
isoform-specific antibody (24) were added per 20 µl mixture, and the
reactions were incubated for an additional 2 h at 4 C. Control
preimmune serum was added in parallel incubations. Reaction mixtures
were subjected to electrophoresis through a 5% polyacrylamide gel
(29:1, acrylamide-bisacrylamide) in 0.5 x Tris-borate-EDTA buffer
for 90 min at 4 C. Gels were dried and subjected to autoradiography.
Quantitative differences in antibody supershifting between nuclear
extracts were assessed by PhosphorImager. Statistical analysis was
performed using Student’s t test.
Transient cotransfection of reporter constructs in control and
stable TRß1-deleted cells
Control TRß1 sense and TRß1 antisense RNA-expressing cells
were grown in hormone-free medium for 24 h. Cells were then
transfected by electroporation with one of the following luciferase
reporter constructs; rGH250-Luc, containing the proximal 250 bp of the
rat GH promoter; DR4-Luc, containing two idealized direct repeat TRE
half-sites separated by 4 bp; or F2-Luc, containing two idealized TRE
half-sites arranged as an inverted palindrome. The reporter plasmid
rGH250-Luc was constructed by insertion of the PCR-generated fragment
(-250 to -1) of the rat GH promoter into the
BglII-SacI sites of the luciferase reporter
plasmid PXPI (23). The reporter plasmids DR4-Luc and F2-Luc were
constructed by insertion of the respective synthetic oligonucleotides
upstream of the herpes simplex virus thymidine kinase promoter in the
luciferase reporter plasmid PT109 (25). Electroporation parameters were
previously described. The reporter plasmid concentration was 3
µg/well. Each well also received 1 µg RSV-ß-galactosidase
expression vector. Control wells were transfected with vector alone
(PXP). The total amount of DNA transfected was kept constant. Cells
were placed in six-well plates and grown in the presence of 0.1
µM T3 for 48 h before
harvesting. Luciferase activity was assayed by standard techniques.
Data were corrected for transfection efficiency by relative
ß-galactosidase expression. In addition, data were corrected for
luciferase expression in control (PXPI or PT109 alone) wells.
Statistical analysis was performed using Student’s t
test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. TRß1
mRNA levels were 92% lower in the line expressing the antisense RNA
than in wild-type or control cells. In contrast, basal TRß2 mRNA was
increased 3.5-fold. Expression of TR1 mRNA was similar in all cell
lines. TRß1 and TR1 mRNAs were not markedly affected by
T3 treatment, whereas TRß2 mRNA was markedly
down-regulated in all cell lines. These changes are in keeping with the
results of previous studies, which have shown either a modest (11) or
no (12) change in TR1 and TRß1 levels in response to
T3 and marked down-regulation of TRß2.
|
, lanes
2, 4, and 6). However, no such supershift band was visible with
extracts from cells expressing the TRß1 antisense construct (lane 8).
Thus, TRß1 was greatly reduced in cells expressing the TRß1
antisense construct. This result is consistent with the mRNA data and
confirms the depletion of the isoform at the protein level in this cell
line. TRß2 antibody also produced a clear supershift band with
wild-type and control cell extracts (Fig. 3, lanes 2, 4,
and 6). However, in contrast to the TRß1 data, there was a 4.2-fold
greater supershift of the antisense cell nuclear extract (lane 8). This
confirms an increase in TRß2 levels in this line. The supershift
pattern produced with the TR1 antibody was similar in all nuclear
extracts, indicating that expression of TR1 protein was similar in
all cell lines (data not shown). Addition of additional TRß1 antibody
produced no additional supershift in extracts from wild-type cells,
control cells, or cells expressing the antisense construct. Similarly,
additional TRß2 antibody resulted in no additional supershift in the
wild-type, control, or antisense-transfected cell extracts (data not
shown). The decreased TRß1 and increased TRß2 supershifts in the
nuclear extracts of anti-sense-transfected cells were, thus, not a
result of suboptimal amounts of antibody limiting the formation of
supershift complexes.
|
|
). These results
indicate that in the antisense line in which TRß1 is depleted, total
TR expression is maintained by a reciprocal increase in TRß2
levels.
|
. Basal levels of GH mRNA were increased 10.5-fold in
TRß1-depleted cells relative to those in wild-type or control cells.
Addition of T3 produced a marked increase in GH
mRNA in all cell lines. However, T3-stimulated GH
mRNA levels were 5.0-fold higher in TRß1-deleted cells than in
wild-type or control cells. These data indicate that depletion of
TRß1 coupled with a concurrent increase in TRß2 leads to both
enhanced basal and T3-stimulated GH gene
expression.
|
. There were no
differences between control and TRß1-depleted cells in either basal
or T3-stimulated expression of luciferase
reporters containing F2-TRE, DR4-TRE, or GH250-TRE. The increase in GH
gene expression in TRß1-depleted cells is thus not a manifestation of
a nonspecific increase in either basal or
T3-stimulated gene expression in this cell line.
These data are further supported by the lack of any difference in the
expression of cyclophilin mRNA between any of the cell lines (Fig. 5A
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the present study, antisense RNA was used to abrogate TRß1 expression in GH3 cells. The depletion was confirmed at both the mRNA and protein levels. Total TR expression was maintained in TRß1-depleted cells by a reciprocal increase in TRß2 levels. This alteration in the population of TRs resulted in an absolute increase in basal and T3-stimulated GH gene expression. To test that these observed increases were not the result of a generalized increase in basal or T3-stimulated gene expression in GH3 cells lacking TRß1, we performed a series of transient cotransfection studies. Basal and T3-stimulated expression of DR4-Luc, F2-Luc, and rGH250-Luc, were similar in TRß1-depleted and control cells. Similarly, there were no differences in the expression of the cyclophilin housekeeping gene among wild-type, control, and TRß1-depleted cells. The observed increase in GH gene expression was thus not the result of a generalized effect on basal or T3-stimulated gene expression and appears to reflect a selective effect on the GH gene. As the results were typical of those obtained from several, independent clones of stable transfected cells, the observed differences were not the result of random variation in the phenotype of parent cells or positional effects resulting from differences in the insertion of the transfected constructs into the host cell genome.
The mechanism underlying the increase in GH gene expression in
TRß1-depleted cells must involve either decreased repressor-silencer
activity or, alternatively, enhanced activator activity at the GH
promoter. In the absence of T3, apo-TRs repress
T3-responsive gene expression, including that of
the GH gene, by mechanisms that may involve direct interaction with
both key general transcription factors and newly characterized
corepressors (4, 5, 29, 30, 31). In the absence of
T3, TRß1 may be a more powerful repressor of
the GH gene than TRß2. Abrogation of TRß1 expression together with
a reciprocal increase in TRß2 levels may thus lead to effective
derepression of the GH locus. Alternatively, TRß2 may be a more
powerful activator of the GH gene than is TRß1, and thus, the
reciprocal increase in TRß2 levels associated with TRß1 deletion
may lead to enhanced GH gene expression. As the levels of GH mRNA are
10.5-fold higher in TRß1-depleted cells relative to those in
wild-type and control cells, even in the absence of
T3, such selective properties would have to
extend to ligand-independent activation. TR1 and TRß1 have been
shown to trans-activate several promoters in a
ligand-independent manner (28, 32, 33, 34). Recent studies have shown that
TRß2 is capable of ligand-independent trans-activation of
certain TREs, including the rat GH promoter, through an
isoform-specific, transferable amino-terminal domain (28). This domain
also confers selective ligand-dependent trans-activation
properties that are TRE half-site dependent, making TRß2 a more
powerful ligand-dependent trans-activator of some natural
TREs than is TRß1.
The mechanism by which depletion of TRß1 produces increased expression of TRß2 is unclear. It is possible that TRß1 represses TRß2 expression in the absence of T3, and that deletion of TRß1 relieves this repression. However, in contrast to GH, TRß2 is negatively regulated by T3. In keeping with other negatively regulated genes, nonliganded TR would thus be expected to function as an activator at the TRß2 promoter (14, 35). It may be that TRß2 is a more powerful ligand-independent trans-activator of the TRß2 promoter than is TRß1, and that increases in TRß2 levels result in positive feedback on the TRß2 promoter in the absence of T3. In this context it is of interest to note that regulation of TRß2 mRNA levels by T3 remains intact in TRß1-depleted cells, indicating that, as with the GH promoter, TRß2 is sufficient to maintain T3-dependent regulation. The mouse TRß2 promoter has recently been characterized (36). It is a complex promoter, containing multiple TRE half-sites in addition to the response elements of other transcription factors. The mechanisms underlying regulation of the TRß1 and TRß2 promoters and the differential expression of the two TRs remain unknown.
The rat GH promoter contains three TRE half-sites between positions -188 and -165 (37). Chimeric reporter constructs containing these promoter sequences exhibit T3 sensitivity, basal repression by unliganded TRs, and ligand-independent trans-activation in transient cotransfection experiments (28, 29, 38). In addition, they mediate active silencing by TRs in a reconstituted in vitro transcription system, composed of individual components of the basal transcriptional machinery (31). The lack of any difference in expression of the rGH250-Luc reporter between TRß1-depleted and control cells indicates that the proximal 250 bp of the rat GH promoter are not sufficient to mediate the selective, TR isoform-specific modulation of GH gene expression demonstrated in the present study when using the endogenous GH gene. It seems likely that additional portions of the GH promoter, distal to -250, may be required to demonstrate such selective effects, at least at the levels of TRß1 and TRß2 found in cells not transiently overexpressing TRs. Higher order DNA effects, not apparent with episomally expressed, transiently transfected reporters, may be involved. Alternatively, the increased basal and T3-stimulated GH mRNA levels in TRß1-deleted cells could reflect posttranscriptional effects. T3 increases the stability of GH mRNA (39) while it decreases that of the TSH ß-subunit mRNA (40). The precise role of the TRs in this aspect of T3 action remains unclear. The differentiation between these phenomena and the identification of the additional cis-elements or cellular components involved will be major steps in understanding the molecular basis for the diversity of thyroid hormone action.
We have shown that the expression of TRß1 can be greatly reduced in GH3 cells by coexpression of a TRß1 antisense RNA. This is associated with an increase in TRß2 expression and a concurrent increase in basal and T3-stimulated GH gene expression. This is the first study to demonstrate the depletion of a specific TR isoform in this cell type. As depletion of TRß1 is associated with an increase in basal and T3-stimulated GH gene expression, we conclude that TRß1 is not necessary for T3 induction of the GH gene in GH3 cells. Furthermore, TRß1 is responsible for substantial basal repression at this locus, whereas TRß2 is sufficient for T3-dependent activation. Although our data show that TRß2 is sufficient for T3 induction of the GH gene in the absence of TRß1, the reciprocal increases in TRß2 levels associated with deletion of TRß1 do not result in equivalent basal repression. These data highlight the potential for promoter- and isoform-specific dissociation of the repression and activation properties of TRs.
Received December 20, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and N-terminal variant ß receptors. Development 114:39–47[Abstract]
and ß
thyroid hormone receptor (TR) gene expression during auditory
neurogenesis: evidence for TR isoform-specific transcriptional
regulation in vivo. Proc Natl Acad Sci USA 91:439–443-related protein which
binds deoxyribonucleic acid but does not bind thyroid hormone. Mol
Endocrinol 2:893–901This article has been cited by other articles:
 |
|
 |
|
  J. Manzano, B. Morte, T. S. Scanlan, and J. Bernal Differential Effects of Triiodothyronine and the Thyroid Hormone Receptor {beta}-Specific Agonist GC-1 on Thyroid Hormone Target Genes in the Brain Endocrinology, December 1, 2003; 144(12): 5480 - 5487. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Yen Physiological and Molecular Basis of Thyroid Hormone Action Physiol Rev, July 1, 2001; 81(3): 1097 - 1142. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
