(Received for publication, August 21, 1996, and in revised form, December 5, 1996)
From the Comprehensive Cancer Center, University of
Wisconsin, Madison, Wisconsin 53792 and Departments of Pathology,
Urology and Biochemistry, University of Rochester, Rochester, New York
14642, the ¶ Department of Medical Pharmacology, Utrecht
University, 3584 CG Utrecht, The Netherlands, and
University
Heart Center, University of Arizona, Tucson, Arizona 85724
While the TR4 orphan receptor (TR4) is able to
repress the expression of its target genes via its interaction with the
direct repeat 1-hormone response element (DR1-HRE) and DR2-HRE, we now report that TR4 can also induce the transcriptional activity of the
reporter gene containing a DR4-HRE via chloramphenicol
acetyltransferase assay. Electrophoretic mobility shift assay and
Scatchard analysis reveal a strong binding affinity (dissociation
constant = 2 nM) between TR4 and DR4-HRE. The
induction mediated by TR4 was detected not only in the synthetic
DR4-HRE but also in some genes, such as rat -myosin heavy-chain and
S14 genes, containing the DR4 or DR4-like motif, which have been
suggested to be the response elements for a thyroid hormone receptor.
Our data also demonstrate this TR4-mediated gene induction is TR4 dose-
and DR4 sequence-dependent. Together, our data suggest that
DR4-HRE can be a positive regulatory element for TR4, which may be able
to induce the transcriptional activity of the genes containing such
positive HREs.
The nuclear receptor superfamily comprises a large group of ligand-dependent transcriptional factors that control the expression of target genes by binding to their cognate hormone response elements (HREs)1 (1-4). It includes the receptors for steroid hormones, thyroid hormones, and retinoids. In addition to the classical receptors, a large number of members in this superfamily have been cloned via their sequence conservation, for which ligands were not found and therefore were classified as orphan receptors (5). The human TR4 orphan receptor (TR4) was initially isolated by our laboratory from human prostate and testis cDNA libraries by degenerative polymerase chain reaction cloning (6). TR4 shows a high degree of nucleotide sequence homology with the TR2-11 orphan receptor and thus forms a unique subfamily in this superfamily (7). Recently, TR4 has also been identified from human lymphoblastoma cells, called TAK1 (8). Northern blot analysis shows that TR4 is expressed in many tissues, being predominantly located in granule cells of the hippocampus and the cerebellum (6). This implies that the TR4 may regulate the specific gene function in a cell-specific manner.
A cysteine-rich cluster that comprises two zinc finger-like structures can be easily identified within the DNA binding domain of the nuclear receptors (9-12). Three amino acids in the stem of the first zinc finger (the P box) are critical to specify the half-site sequence, while 5 amino acids in the second finger (the D box) can alter the selection pattern of the half-site spacing. On the basis of its distinct P box, TR4 can be classified into the estrogen receptor/thyroid hormone receptor subfamily which displays cross-recognition of the palindromic or direct repeat (DR) HREs. These HREs consist of a core half-site, AGGTCA, or related sequence (13-15). Our previous data suggested that TR4 preferred to bind to the HREs, which consist of two DRs of consensus half-site, rather than a palindrome.2 Furthermore, TR4 can repress the retinoic acid-induced transactivation by competitive DNA binding to the response elements of other receptors (i.e. DR5 for retinoic acid receptor (RAR) and DR1 for retinoid X receptor (RXR)).2 In addition, the TR4-mediated suppression can also be found in the gene promoter containing DR2-like response element that is located in the +55 region of the SV40 major late promoter (16).
All of the above findings suggest that TR4 may function as a repressor to block other receptor-mediated transcriptional activities through competitive DNA binding to the same HREs. In the present study, we tested the regulation of TR4 on a DR4-HRE which has been shown to be a hormone response element for thyroid hormone receptor (T3RE) by using a reporter gene assay (17-19). Through these studies, we found that the TR4 can induce gene expression through the interaction with DR4-HRE in a sequence- and dose-dependent manner. In addition to repressing its target genes, our data provide the first evidence to show that TR4 can also activate its target genes, through a DR4- or DR4-like T3RE. This finding may expand our understanding of the physiological function of the TR4 orphan receptor.
For the transient transfection or
coupled in vitro transcription and translation of the
full-length TR4 protein, the pCMX-TR4 and pET14b-TR4 plasmids were
constructed as described, respectively (16). pCD-TR1, the expression
vector for thyroid hormone receptor
, was provided by Dr. Leslie J. Degroot. The reporter plasmids -32TK-CAT and TK-(DR4)2-CAT,
either without or with two copies of DR4
(5
-GATATAGGGGTCAAATAAGGTCAAATG-3
, Fig. 1A), were kindly provided by Dr. RC. Ralff. TK-(mDR4)-CAT containing one copy of the
mutated DR4 (5
-GATCTAGAATAAATG-3
)
inserted into the BamHI site of -32TK-CAT plasmid was
generated as a negative control. The reporter genes used pCATp reporter
genes fused with either the 5
-flanking region of rat
-myosin heavy
chain (
-MHC) (from
374 to +420) (17), the far upstream regulatory
region (FUR) (from
2952 to
2448) in the rat S14 gene (FUR1/2/3-CAT) (18), or the long terminal repeat of the human immunodeficiency virus I
(from
454 to +82, HIVLAILTR.CAT) (19), which contains the
T3RE motif in the natural promoter region.
Coupled in Vitro Transcription and Translation
Expression
plasmids pET14b-TR4 and pCD-TR1 were in vitro transcribed
and translated by the TNT system (Promega) as described previously
(16).
EMSA analysis
was performed according to the methods described by Cooney et
al. (20). In general, the double-stranded DR4 oligonucleotides
were end-labeled with a [-32P]ATP (specific
activity = 6000 Ci/mmol, DuPont NEN) by T4 polynucleotide kinase
to 2-8 × 108 cpm/µg. For competitive DNA binding
reactions, 100-fold excess of cold double-stranded DR4 or mDR4
oligonucleotides were mixed with the proteins prior to the addition of
labeling probe to the reactions. For antibody supershift analysis, 1 µl of the monoclonal anti-TR4 antibody (G232-303.4) was added to the
reaction. As a control for supershift assays, the monoclonal
anti-androgen receptor antibody (N1-15) (21) was added. Gels were
fixed in 50% ethanol and 10% acetic acid for 30 min before drying.
The radioactive gels were analyzed by either PhosphorImager (Molecular
Dynamics Inc.) or autoradiography.
The DNA-protein binding assay was performed as described previously (16). In brief, free probes (32P-DR4) and DNA-protein complexes were resolved in native gel and quantified by scintillation counting. The dissociation constant (Kd) was determined from the minus reciprocal of the slope of the line generated from the empirical data.
Transient TansfectionHepG2 cells were routinely maintained
in Dulbecco's modified Eagle's medium with 5% heat-inactivated fetal
bovine serum. HepG2 cells (3 × 105) were seeded
in 6-cm culture dishes 24 h before transfection. The medium
was changed to Dulbecco's modified Eagle's medium with 5%
charcoal dextran-treated serum at least 1 h before transfection. The cells were transfected using a modified calcium phosphate precipitation method (22). To normalize the transfection efficiency, the -galactosidase expression vector was co-transfected.
The data are presented as mean ± S.D. One-way analysis of variance was used to determine the differences among multiple groups. Student's t test was used for analyzing differences between two groups. The degree of significance (p value) was shown with asterisks. p < 0.05 was accepted as the level of statistical significance.
EMSA was used to determine binding specificity for TR4 to DR4-HRE. In vitro translated TR4 protein was incubated with 32P-labeled DR4-HRE and analyzed on a 5% polyacrylamide gel. As shown in Fig. 1B, a specific DNA-protein complex (indicated with arrowhead, lane 2) was identified that has a migration distinct from the nonspecific DNA-protein complexes in the presence of mock-translated product (lane 1). The binding specificity was further confirmed by adding excess unlabeled DR4-HRE or mutated DR4 (mutated DR4-HRE, in which the third G was replaced with C) (Fig. 1A). Competition by 100-fold excess unlabeled DR4-HRE abolished the specific DNA protein complex (lane 3). In contrast, this complex could not be abolished completely with the unlabeled mutant DR4-HRE (lane 4). This DNA-protein complex could not be affected when androgen receptor monoclonal antibody (N1-15) was added (lane 5). In addition, when the anti-TR4 monoclonal antibody (G232-303.4) was added in the reaction, a supershifted band made by DNA-receptor-antibody complex was visible (lane 6).
All these data clearly demonstrated that the TR4 can bind specifically
to DR4-HRE. We also calculated the binding affinity between TR4 and
DR4-HRE by the Scatchard analysis. Fixed amounts of in vitro
translated TR4 proteins were incubated with an increasing amount of
32P-labeled DR4 (0.1-12.8 ng) and resolved in EMSA. As
shown in Fig. 2, Scatchard analysis displays a single
binding component with a dissociation constant (Kd)
of 2 nM and Bmax of 0.075 nM. These data fit the Kd range for
steroid receptors and their HREs.
TR4 Mediates Dose-dependent TK-(DR4)2-CAT Transcriptional Activation
As shown above, TR4 can bind to a
DR4-HRE with high specificity and affinity. We were then interested in
knowing the consequences of this binding. The reporter gene
TK-(DR4)2-CAT and the TR4 expression (pCMX-TR4) plasmid
were co-transfected into HepG2 cells. Surprisingly, we found that TR4
itself was able to activate CAT activity (Fig. 3A, lane 3). This transactivation
could not be observed when pCMX-TR4 was co-transfected with the
reporter plasmids without DR4-HRE (-32TK-CAT, lane 1), or
with the mutated DR4-HRE (TK-(mDR4)-CAT, lane 2). Moreover,
the CAT activity was increased as increasing amounts (from 0.01 to 1 µg) of pCMX-TR4 were co-transfected (Fig. 3B, lanes
1-6) Transcriptional activity reached a plateau when 2 µg of
pCMX-TR4 was co-transfected (lane 7). In summary, these data
clearly demonstrated that TR4 induced transcriptional activation is TR4
dose- and DR4-HRE sequence-dependent.
The Correlation between TR4 and TR
All of the data above suggest that
DR4-HRE is a positive novel response element for the TR4. Previous
reports suggest that the DR4-HRE can also serve as the response element
for the T3RE. Therefore we investigated how these two
receptors interacted with each other on the same target sequence by
designing the following experiments. First, a fixed amount (2 µg) of
thyroid hormone receptor expression vector, pCD-TR1, was transfected
singly or co-transfected with increasing amounts of pCMX-TR4 (from
0.125 to 1 µg) in the absence of T3. As shown in Fig.
4A, CAT activities were activated only when
pCMX-TR4 was transfected (lane 1 versus 2). No activation was detected when pCD-TR
1 was transfected alone (lane 1 versus 3). The induction was significant increasing (p < 0.005) by increasing amounts of pCMX-TR4 (lanes 3-7). The
induced CAT activities mediated by TR4 were slightly reduced when the
unliganded pCD-TR
1 (2 µg) was co-transfected (lane 2 versus
7). However, the reduction of TR4-mediated CAT activation by
unliganded TR
1 was not very significant (p < 0.1)
by Student's t test. This phenomenon was also observed by
co-transfection of pCMX-TR4 (1 µg) with increasing amounts of
unliganded pCD-TR
1 (Fig. 4C, lanes 3-7). The
activities induced by TR4 (1 µg) was reduced 50% when 2.0 µg of
unliganded TR
1 was co-transfected. (Fig. 4C, lane 3 versus 7). Again, the dose of unliganded TR
1 does not
significantly (p < 0.1) repress the activation of CAT
activities by TR4. This may be due to the fact that TR4 is more potent
than liganded TR
1 in the activation of TK-(DR4)2-CAT reporter gene. The tendency for repression of TR4-mediated
transcriptional activity by the unliganded TR
1 is consistent with
previous reports (23-25) in which suggested that unliganded TR
1 may
act as a gene silencer.
In contrast, in the presence of T3 (107
M), both pCD-TR
1 and pCMX-TR4 can activate CAT
activities up to 15- and 22-fold alone, respectively (Fig. 4,
B and D, lanes 1-3). The
transcriptional activity slightly increased when we co-transfected 2 µg of pCD-TR
1 and increasing amounts of pCMX-TR4 (0.125 to 1 µg)
(Fig. 4B, lanes 3-7). Furthermore, when a fixed
amount of pCMX-TR4 (1 µg) was transfected together with increasing
amounts of pCD-TR
1 (from 0.5 to 2 µg) (Fig. 4D,
lanes 3-7), the results showed that CAT activity induced by
TR4 did not increase significantly by adding additional TR
1. This
may suggest that transcriptional activity could be saturated by
expression of 1 µg of TR4 alone.
As described above, TR4 can
recognize synthetic DR4-HRE and activate TK-(DR4)2-CAT
reporter gene in a dose- and DNA sequence-dependent manners. We then explored whether this TR4-mediated CAT induction could
also be applied to the DR4-T3RE-containing or nonclassical T3RE natural promoters. The sequences corresponding to
T3RE were shown in Fig. 5A with
an arrow or underline. The first gene used was
from the 5-flanking region of rat
-MHC gene (from
347 to +420),
which contains a DR4-HRE that serves as a T3RE (17). The
expected results were shown in Fig. 5B, panel a.
The transcriptional activity of this gene can be induced up to 9.5-fold
with co-transfection of pCMX-TR4. The second gene tested was the FUR of
the rat S14 gene (from
2952 to
2448) which contains multiple
T3REs that synergize with each other in their
responsiveness to T3 (18, 26). As shown in Fig.
5B, panel b, the transcriptional activity of rS14
can be induced up to 6.3-fold upon co-transfection with pCMX-TR4. The
third native gene we tested was the HIV-LTR (from
454 to +82), which
contains a nonclassical T3RE, GGCGGG, and can be activated by
unliganded TR
1 (27). A 3.9-fold induction mediated by pCMX-TR4 can
also be observed (Fig. 5B, panel c). The
differences of CAT activity obtained between TR4-transfected and
control in these three native T3RE responsive genes were
considered significant (p < 0.05) based by Student's
t test. All these data suggest that the DR4-T3RE
as well as some nonclassical T3REs can serve as potential
target sequences for both T3R and TR4.
To further define the correlation between TR4 and TR1 in these
natural promoter constructions, we then tested the overexpression of
both pCMX-TR4 and pCD-TR
1 in the reporter gene assay. Fig. 6A showed that CAT activities were
significantly (p < 0.05) induced by the transfection
of pCMX-TR4 alone in all reporter plasmids we had tested, but induction
could not be observed in the transfection of unliganded pCD-TR
1. In
the absence of T3, the activation induced by TR4 was
slightly reduced by unliganded TR
1 in synthetic DR4,
-MHC, and
S14 reporter gene constructs but not in HIV-LTR reporter gene. These
results agree with a previous report in which HIV-LTR was shown to be
activated by unliganded TR
1, and the addition of T3
could reverse this effect (27). On the other hand, in the presence of
T3 (Fig. 6B), cotransfection of pCMX-TR4 with pCD-TR
1 slightly increased the activation as compared with the single receptor transfection but the increasing effect is not very
significant. Together, our data provides the first evidence showing
that the TR4 is capable of activation of CAT reporter genes containing
a DR4-T3RE. This is true for both synthetic and native
T3REs as well as other nonclassical T3REs.
Like the TR2 orphan receptor, TR4 can also serve as a repressor in
RAR/RXR-mediated gene induction and transcription from the major late
promoter of the SV40+55 gene (16, 22, 28). In the present study, we
showed TR4 can also induced the transcriptional activity of genes
containing a DR4 or DR4-like sequence. These data suggest that the
selectivity of the receptor response is dependent on the core motif,
including the spacing, orientation, and precise sequence composition of
the adjacent core motifs. In Fig. 1, when we replaced the third G with
C in DR4-HRE (mDR4), an excessive amount of unlabeled mDR4 could not
abolish with the specific DNA-receptor complex completely. This
suggests that the precise sequence of the core motif plays an essential
role in TR4-mediated gene regulation, a result in agreement with our
previous studies as the TR2 orphan receptor (22, 28). In addition, the
spacing of core binding motifs may also be a key factor in dictating
specificity of transcriptional response to a transcriptional factor.
TR4 can either induce or suppress transcription through the binding to
HREs with same sequence of the core binding motif but different
spacing. Activation or repression may be dependent on different spacing
of the motifs (DR1/DR5 for the repression of RAR/RXR-mediated gene
induction and DR2 in suppression of SV40 gene versus DR4 in
induction of -MHC and S14 genes). Similar results are also observed
in the case of the thyroid responsive pathway, as transactivation was
observed only on the unspaced palindromic element (PAL), a native rat
growth hormone T3RE but not on the 3-base pair-spaced
vitellogenin A2 estrogen response element, PAL(+3) (29). Interestingly,
when we tested the TR4-mediated induction on different native
T3REs, we found that the levels of induction are different
among these three T3REs. The order for the TR4-mediated
induction among these three native gene promoters is
-MHC > rS14 > HIV-LTR. This implies that TR4 had a better induced effect
on the DR4-T3REs than TR
1 on such nonclassical T3REs. This suggests that the orientation, and the sequence
composition of these adjacent putative core motifs are also very
important in the determination of gene regulation. Overall, the
combination of core motif spacing, orientation the sequence composition
of the core motif and adjacent area could determine the pattern of responsiveness exhibited by genes that are regulated by TR4.
However, the molecular structure of the receptor also contributes to the specificity of gene regulation. For example, the two helical regions in the DNA binding domain of the receptor may also contribute to the specificity of base-specific contacts (30, 31). The third helix formed between the two helices can also serve as a gap to separate two core binding motif (32, 33). When the DNA binding domain of the receptor binds to HREs, the receptor may change its conformation and allow access to different factor(s). As a result, TR4 may bind to different co-factors and contribute to both positive or negative gene regulation through different protein-protein interactions (34). Previous reports have suggested that other nuclear receptors, such as RXR, can serve as a common heterodimerization partner for several nuclear receptors, including the thyroid hormone receptors and retinoic acid receptors (35). The intrinsic binding properties of RXR are masked in such T3R-RXR and RAR-RXR heterodimers. In contrast, RXR is active as a non-DNA-binding co-factor with the NGFI-B/nurr1 orphan receptors (35). Although we were unable to demonstrate the interaction between RXR and TR4, it is still possible that TR4 may need other co-factors for proper function. Using a yeast-two hybrid system, we are in the process of isolating several TR4-associated proteins.3 Recently, COUP-TF has been shown to inhibit the transcriptional activities of T3R, 1,25-dihydroxyvitamin D3 receptor (36), RAR, RXR (37), the peroxisome proliferator-activated receptor (38, 39), and hepatocyte nuclear factor 4 (40). The molecular mechanism of COUP-TF-mediated transcriptional repression may also be involved in some other activator(s) (41). Together, these findings suggest that interactions among proteins may contribute to receptor-mediated gene induction or suppression.
Ligands have always played a central role in the activation of steroid
hormone receptors. For example, unliganded T3R is
associated with cellular co-repressor p270/N-CoR which serves as a gene
silencer (24). Upon ligand binding, T3R dissociates with
the co-repressor and actively turns on its target genes (23, 42).
However, a similar effect was also observed in this study. Unliganded
TR1 has a potential suppressive effect on the TR4-mediated
transactivation in both synthetic DR4 or native T3RE-CAT
reporter genes, although the effect is not very significant
(p < 0.1). One of the potential explanation is that
TR4 has higher binding affinity than TR
1 to T3REs.
Therefore, TR4 may compete stronger than TR
1 on binding to the same
HRE to activate its target gene. On the other hand, the physiological
function of orphan receptors still remains obscure because information
regarding ligand is lacking. Nevertheless, the induced CAT activity
shown in this study may still provide us a useful assay system for
searching for potential ligands or activators for the TR4 orphan
receptor.
What then is the relationship between TR4 and TR1, since they both
recognize the same HRE? Transient transfection studies show that TR4
can activate CAT activity in the presence or absence of T3
suggesting that T3 is not the ligand or activator for TR4. In contrast, T3R can only activate transcriptional activity
when the ligand, T3, is present. These results suggest that
two different receptors may use different ligands to induce similar or
different sets of target genes. Another possibility is that although
two different receptors can recognize the same HRE, they may function differently and independently in gene regulation. For example, T3R can also bind with high affinity to estrogen response
element and serves to antagonize estrogen-dependent
transcriptional activation (15). In our present study, we found that
the T3-induced transactivation of TR
1 could be
overwhelmed by TR4. In addition, the repression effects from unliganded
TR
1 could be overcome by adding more TR4. The dominant effect of TR4
may be due to its higher binding affinity to DR4-HRE as compared with
that of TR
1 (data not shown). Together, our data suggested TR4 may
play an important role in TR
-mediated gene regulation.
In summary, our data suggest that the ability of TR4 to induce genes with T3RE-DR4-HRE may provide us with a model system for future studies in determining the mechanisms to determine how the TR4 activates and represses its target genes. Moreover, this TR4-mediated induction system may provide a potential assay for finding a possible ligand and/or activators for TR4.
We thank Dr. Leslie J. Degroot for thyroid
hormone receptor expression vector, pCD-TR1. We also thank Dr. H. C. Towle for kindly providing the rat S14 gene promoter, Dr. J. A. A. Ladias for HIV-LTR reporter gene, and Dr. Alan Saltzman for his helpful comments and editing on the manuscript.