| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Instituto de Investigaciones Biomédicas (J.A.V., A.A., J.M.-P., P.S.), Consejo Superior de Investigaciones Científicas, Arturo Duperier 4, 28029 Madrid, Spain; Dipartimento di Biologia e Patologia Cellulare e Molecolare (M.Z.), Universitá degli Studi di Napoli Federico II, 80131 Naples, Italy; and Stazione Zoologica Anton Dohrn (R.D.L.), Villa Comunale, 80121 Naples, Italy
Address all correspondence and requests for reprints to: Dr. Pilar Santisteban, Instituto de Investigaciones Biomédicas, Consejo Superior de Investigaciones Científicas, Arturo Duperier 4, 28029 Madrid, Spain. E-mail: PSantisteban{at}biomed.iib.uam.es
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Loss of differentiation as a result of cellular transformation is a well characterized phenomenon (12). In many thyroid tumors, expression of molecular markers such as Tg and TPO and their corresponding transcription factors TTF-1 and Pax-8 is lost, particularly in late stages of thyroid carcinogenesis (13). Oncogenic transformation of thyroid cells in culture mimics this loss of differentiation, and experiments with FRTL-5 thyroid cells demonstrate that Ha- and Ki-ras transformation also abolish Tg and TPO expression completely. However, while in Ki-ras transformed cells TTF-1 is undetectable, Ha-ras-FRTL-5 cells display normal levels of TTF-1 messenger RNA (mRNA) and protein (14).
Previous results from our laboratory showed that TTF-1, encoded by the titf-1 locus, is a phosphoprotein required for specific expression of Tg and TPO (15). Precise mapping of the TTF-1 phosphorylation sites demonstrated that this transcription factor is phosphorylated on a maximun of seven serine residues distributed in three different phosphotryptic peptides (15). Therefore, as described for many other transcription factors, phosphorylation appears to be a key regulatory mechanism by which TTF-1/DNA binding and/or transactivation could be modulated. Several observations indicate that phosphorylation of TTF-1 may be relevant only in specific cell types. Previous works in thyroid cells proposed that phosphorylation is a critical step for TTF-1 binding to the Tg promoter (16), whereas studies in TTF-1-transfected HeLa cells indicated that phosphorylation does not play an important role in the regulation of TTF-1 DNA binding activity (15). Therefore, the biological significance of TTF-1 phosphorylation remains to be elucidated.
In this study, we have searched for regulatory mechanisms affecting TTF-1 activity that could be altered upon Ha-ras transformation. Using a construct containing a specific TTF-1 responsive element within the Tg promoter (17), we demonstrate that the ability of TTF-1 to promote transcription is lost in cells transformed with the Ha-ras oncogene. This ability is partially reestablished when the catalytic subunit of the protein kinase A (PKA) is expressed in these cells. In vivo cell labeling and immunoprecipitation of TTF-1 from Ha-ras-FRTL-5 cells show a clear decrease in the overall extent of phosphorylation, along with a heterogeneity in the distribution of the phosphorylated serines. This heterogeneous phosphorylation is more prominent within the amino terminal region of the protein.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TTF-1 protein analysis
FRTL-5 or transformed Ha-ras-FRTL-5 cells were
labeled in methionine-free medium containing 0.5 mCi of
[35S] methionine/ml for 2 h. Cells were lysed in a
buffer containing 10 mM sodium phosphate (pH 7.4), 0.1
M NaCl, 1% Triton X-100, 0.1% SDS, 0.5% sodium
deoxycholate, 2 mM phenylmethylsulfonil fluoride, 1 µg/ml
pestatin, 1 µg/ml leupeptin, 1 µg/ml aprotinin, 50 mM
sodium fluoride, 10 mM sodium pyrophosphate, and 5
mM sodium vanadate. Extracts were incubated with an
specific TTF-1 antibody (15) coupled to protein A sepharose beads
(Pharmacia Biotech Inc., Piscataway, NJ). Immune complexes were washed
extensively with lysis buffer, boiled in SDS-Laemmli sample buffer, and
resolved by SDS 8%-PAGE. For Western blot, nuclear extracts were
obtained as described (20) and subjected to SDS-PAGE. Proteins
were then transferred to nitrocellulose membranes and probed with the
same TTF-1 specific antibody (15). After incubation with a secondary
antibody, immunoreactive bands were visualized by enhanced
chemiluminescence ECL (Amersham Corp., Arlington Heights, IL).
Electrophoretic mobility shift assays
DNA binding assays were performed as described (21) using
32P-labeled double-stranded oligonucleotide C
(5'-CACTGCCCAGTCAAGTGTTCTTGA-3') derived from the -92 to -68 region
of the Tg promoter and shown to recognize the transcription factor
TTF-1 (6). Nuclear proteins (3 µg), prepared as described (20), were
preincubated 15 min on ice in a binding reaction mixture containing 40
mM HEPES, pH 7.9, 200 mM KCl, 0.5
mM DTT, 0.2 mM EDTA, 5% Ficoll, and 3 µg
poly(dI-dC). In competition and supershift experiments, 100-fold excess
of unlabeled competitor or 1 µg of TTF-1 antibody were added,
respectively. Then 50 pg of labeled C probe were added to the mixture
and incubated for 30 min at room temperature. The resulting DNA-protein
complexes were separated from free DNA on a 5% polyacrilamide gel.
Plasmids and transfections
For detection of TTF-1 transcriptional activity, the constructs
C5E1b-CAT and E1b (17), referred in this work as 5C-CAT and TATA-CAT,
respectively, were transiently transfected by the calcium phosphate
precipitation technique (22) into FRTL-5 and Ha-ras-FRTL-5
cells. The 5C-CAT construct contains five tandem repeats of the C
binding site from the Tg promoter, and it is exclusively dependent on
TTF-1 for transactivation (17). TATA-CAT was used as a negative
control. The plasmid CMV-Luc (17) was used to correct for transfection
efficiency. Forty eight hours after transfection, cell extracts were
prepared and CAT (22) and luciferase activities (23) determined. For
PKA overexpression assays, vectors containing either a wild-type or a
mutated PKA catalytic subunit (24) were transfected in
Ha-ras-FRTL-5 cells. TTF-1 mutants carrying Ser/Ala
substitution were generated by PCR as previously described (15).
Expression vectors for wild-type and mutant TTF-1 complementary DNAs
were transfected into HeLa cells 36 h before in vivo
labeling.
In vivo labeling and fingerprinting analysis of phosphopeptides
Exponentially growing FRTL-5, their Ha-ras
derivatives and transfected HeLa cells were maintained in
phosphate-free medium for 30 min, then labeled with 0.5 mCi of
[32P] orthophosphate/ml for 3 h. Cell extracts and
TTF-1 immunoprecipitation were performed as described above.
Immunoprecipitated, in vivo phosphorylated TTF-1 and the
different mutant proteins were separated in SDS 8% polyacrilamide gel
electrophoresis and electroblotted to nitrocellulose. After
autoradiography of the membranes, the bands were excised and digested
with trypsin-TPCK (Worthington) as described (15). The resulting
solubilized peptides were collected, oxidized, and loaded onto TLC
plates (25). First dimension separation was performed by
electrophoresis at 1,000 V in 0.13 M
(NH4)2CO3, pH 8.9, for 30 min. TLC
plates were then subjected to chromatography in isobutiric: pyridine:
acetic acid: n-butanol:water (62.5: 4.8: 2.9: 1.9: 27.9). When
necessary, phosphopeptides were recovered from the TLC plate and
subjected to chemical cleavage with cyanogen bromide (CNBr) in the
conditions described (26).
TTF-1 mutational analysis
A 200-bp DNA fragment containing the N-terminal region of the
TTF-1 protein was amplified from FRTL-5 and Ha-ras-FRTL-5
genomic DNA by PCR, using the primers 5'-AGTAGCGAGGCTTCGCCTTC-3' and
5'-TGTCTGTAAGCTGCGAGCGG-3' for forward and reverse annealing,
respectively. Amplification was performed following standard protocols
using Taq polymerase (Perkin Elmer Co., Foster City, CA).
Amplified DNA fragments were subcloned in the pCRII vector (Invitrogen,
Carlsbad, CA), and a representative number of positive clones were
sequenced (Applied Biosystems, Foster City, CA). Sequences were aligned
using the Pileup program (GCG package, University of Wisconsin DNA
software) and compared with the previously published TTF-1 sequence
(27).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The levels of TTF-1 phosphorylation in thyroid FRTL-5 cells and their
Ha-ras transformed counterpart were analyzed by in
vivo 32P labeling and immunoprecipitation. Equal
number of exponentially growing cells were incubated with radiolabeled
orthophosphate. After protein extraction, immunoprecipitated TTF-1 was
separated by electrophoresis (Fig. 1A
). A
marked decrease in TTF-1 phosphorylation was observed in
Ha-ras transformed FRTL-5 cells, as determined by the amount
of 32P incorporated into the transcription factor.
Comparison with the parental FRTL-5 cells and quantitation of the
signal by densitometric scanning indicated up to a 5-fold reduction in
the amount of radiolabeled 32P in TTF-1 immunoprecipitated
from Ha-ras-FRTL-5 cells. We ruled out the possibility that
differences in radioactive labeling were due to the presence of
different amount of TTF-1 protein in both extract preparations. For
that purpose, steady-state levels and de novo protein
synthesis were determined by Western blot and immunoprecipitation of
35S-methionine-labeled TTF-1 (Fig. 1B), respectively. In
both cases, comparable levels of immunoreactive TTF-1 signal was
observed, unambiguously indicating that differences in 32P
labeling were exclusively due to a decrease in the amount of tracer
incorporation.
|
shows the results of the
electrophoretic mobility shift assay performed to detect the presence
of protein-DNA complexes in crude nuclear extracts from FRTL-5 and
Ha-ras-FRTL-5 cells. Incubation of equal amounts of nuclear
proteins with a 32P-labeled C oligonucleotide and
separation by nondenaturing polyacrylamide gel electrophoresis yielded
the same TTF-1/DNA complexes in both cell types (lanes 2 and 3).
Specificity of the complex in Ha-ras-FRTL-5 was assayed
using a 100-fold excess of an unrelated oligonucleotide (lane 4)
or unlabeled oligonucleotide C (lane 5). The identity of the complex in
FRTL-5 and Ha-ras-FRTL-5 was further confirmed by supershift
assays using an anti-TTF-1 specific antibody (lanes 6 and 7).
|
). However, when this
construct was transfected in Ha-ras-FRTL-5 cells, CAT
activity values were comparable with those obtained in the negative
control, indicating that TTF-1 is unable to promote transcription in
Ha-ras-transformed FRTL-5 cells.
Changes in TTF-1 phosphorylation pattern in Ha-ras transformed
FRTL-5 cells
The results described above provide a direct correlation between
the absence of the thyroid differentiation marker Tg and the inability
of TTF-1 to promote transcription in Ha-ras-FRTL-5 cells.
The fact that this functional alteration is associated with a decrease
in 32P-radiolabeled TTF-1 in these transformed cells
prompted us to analyze alterations of TTF-1 phosphorylation in more
detail. We investigated whether the reduction of phosphorylation either
targets to any particular serine residue/s or affects the overall
phosphorylation extent of the TTF-1 molecule. To address this question,
cultures from FRTL-5 and Ha-ras-FRTL-5 cells were in
vivo labeled with 32P-orthophosphate.
Immunoprecipitated TTF-1 was separated by electrophoresis and
transferred to nitrocellulose. Membrane-bound TTF-1 was digested with
tripsin-TPCK and resulting peptides were resolved by two dimensional
electrophoresis-chromatography, as described in Materials and
Methods. Representative experiments, where the same number of
counts were loaded, are shown in Fig. 3.
TTF-1 fingerprinting analysis from FRTL-5 cells shown the migration of
three phosphotryptic peptides (designated 1, 2, and 3), consistent with
the previously described phosphorylation pattern (15). Analysis from
Ha-ras transformed cells revealed a similar pattern with
tryptic phosphopeptides 1 and 2 migrating at the bottom of the TLC
place. However, a heterogeneous mobility of phosphopeptide 3 was
observed. This peptide, clearly defined in FRTL-5 cells, migrates as a
doublet in the transformed cell line, and two phosphopeptides were
identified (3a and 3b). Mixing experiments of both preparations
indicated that spot 3 of FRTL-5 cells corresponds to spot 3a in
Ha-ras-FRTL-5 cells (not shown), and, therefore, 3a and 3b
are different phosphorylation stages of the same peptide. The faster
mobility of peptide 3b indicates a higher amount of 32P
incorporation into this phosphopeptide.
|
). As for the endogenous TTF-1, three
major phosphopeptides were identified in HeLa cells transfected with
wild-type TTF-1 (panel A). Construct S1–24, bearing four alanines
replacing serines in the N-terminal region produced a peptide map in
which peptide 3 completely disappeared (panel B). Tryptic map of
32P labeled S64 mutant where the two serines in the
C-terminal has been replaced with alanines shows peptide 3 migration at
the upper right corner of the chromatography plate, whereas
peptide 1 is missing at the bottom. Double mutant S61 (panel
C) indicates that peptide 2 derives from the TTF-1 central region, and
confirms the identity of peptide 3 with the N-terminal.
|
). These results allowed us to conclude
that the three radiolabeled peptides correspond to the same TTF-1
protein fragment and also demonstrated that no other residue is
phosphorylated in peptide 3 in Ha-ras-FRTL-5 cells.
|
) and corresponds to that
obtained in FRTL-5 cells and to the previously published TTF-1 gene
(27). These results ruled out the possibility that differences in
phosphorylation could be the consequence of mutations affecting
residues at the N-terminal TTF-1 region, supporting the notion that as
a consequence of the Ha-ras transformation, specific kinase
pathways are altered and lead to modifications of TTF-1
phosphorylation.
PKA overexpression partially reestablishes TTF-1 transcriptional
activity in Ha-ras-FRTL-5 cells
To gain further insights in the mechanisms responsible for the
differential phosphorylation of TTF-1 in Ha-ras-FRTL-5
cells, we explored the participation of signal transduction pathways
that could be involved in the regulation of TTF-1 transcriptional
activity. Because previous works have correlated Ha-ras
transformation with alteration of different kinase pathways (28) and
considering a recent report describing and characterizing a functional
PKA phosphorylation site within the N-terminal motif of TTF-1 (29), we
analyzed whether exogenous expression of PKA in
Ha-ras-FRTL-5 cells modulates transcriptional activity of
TTF-1. Ha-ras-FRTL-5 cells were cotransfected with different
concentrations (0.5, 1, 2, 4, and 6 µg) of an expression vector
containing either the wild-type or the mutated PKA catalytic subunit
(24), along with the 5C-CAT or the TATA-CAT constructs. CAT and
luciferase activities were determined in cell extracts (Fig. 6).
Overexpression of PKA in Ha-ras-FRTL-5 cells yielded a
moderate increase (2-fold) in 5C-CAT activity, whereas no increase was
found with the TATA-CAT construct. The maximum activity was obtained
when 2 µg of the wild-type PKA vector was cotransfected, whereas no
increase in CAT activity was found when the mutated PKA was used. These
data indicate that PKA is able to partially reestablish transcriptional
activity of TTF-1. Nevertheless, CAT activity values in
Ha-ras-FRTL-5 cells in the presence of PKA were
significantly lower than those obtained in normal FRTL-5 (Fig. 2B),
indicating that other kinase pathways, presumably altered in
Ha-ras-FRTL-5 cells, may be responsible for the final
phosphorylation stage of TTF-1.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One of the most common mechanisms affecting transcription factor activity is phosphorylation, so we focused our attention in changes in orthophosphate incorporation into TTF-1. Because this factor is inactive in Ha-ras transformed FRTL-5 cells, we also explored whether the transcription factor was able to bind its DNA recognition site and/or whether its transcriptional activity was impaired. These two possibilities, along with changes in the subcellular distribution, are the most common ways by which phosphorylation alters the activity of a transcription factor (33, 34). The detection of equal amount of TTF-1 protein in nuclear extracts from both normal and transformed cells ruled out the latter possibility. In addition, our results indicate that hypophosphorylated TTF-1 in Ha-ras-FRTL-5 cells is able to bind DNA, although its ability to transactivate is lost. Modification of phosphorylation upon oncogenic transformation of some transcription factors have been described in other cell systems (35, 36). This modification normally affects DNA binding activity of the targeted transcription factor. Interesting for our studies is the fact that ras affects phosphorylation of c-jun transcriptional activation domain (37). Activation of protein kinase C by ras transformation also decreases phosphorylation of c-jun at sites that inhibit its DNA binding capacity (38). Due to the pleiotropic effect of ras, there is a possibility that it could also modify TTF-1 transactivation activity.
In a previous work, transfection in HeLa cells of TTF-1 mutant derivatives carrying Ser/Ala substitutions determined that phosphorylation is not involved in controlling TTF-1/DNA binding (15). However, other studies indicate that phosphorylation of this transcription factor is essential for DNA binding in thyroid cells (16), suggesting that regulation of TTF-1 function is likely to be specific in the thyroid context. Our results of phosphopeptide mapping of TTF-1 in Ha-ras-FRTL-5 cells, along with a transfection approach with mutant derivatives, indicate that the reduction of TTF-1 phosphorylation is accompanied by an heterogeneous pattern of the phosphopeptides affecting the N-terminal domain of the protein. The higher mobility of the spot 3b, only present in Ha-ras transformed cells, indicates that this peptide corresponds to a highly phosphorylated form of peptide 3a. We interpret these results considering that, although the equilibrium between phosphorylated and nonphosphorylated forms of the TTF-1 protein is displaced to the latter form, some population of highly phosphorylated TTF-1 molecules are present in those cells transformed by Ha-ras. This observation may indicate that a precise equilibrium between kinases and phosphatases responsible for TTF-1 phosphorylation is necessary for the correct function of this transcription factor.
The fact that the N-terminal stretch of the TTF-1 protein is involved in phosphorylation alterations in Ha-ras-FRTL-5 cells may be of special interest because previous results have designated this N-terminal region as one of the TTF-1 transactivation domain (17). In the N-terminal site of the TTF-1 protein, several consensus sequences for protein kinases have been observed. The sites that we have identified by computer analysis show homology to casein kinase II (CKII), protein kinase C (PKC), and mitogen-actived protein kinase (MAPK) (15), and we have previously demonstrated that PKC is able to in vitro phosphorylate TTF-1 (15). A recent report also localizes and functionally characterizes a PKA phosphorylation site within this N-terminal motif of TTF-1 in lung cells (29). Our results cotransfecting the catalytic subunit of the PKA into Ha-ras-FRTL-5 cells show a moderate increase in a reporter construct depending on TTF-1, suggesting that although PKA may participate in the regulation of TTF-1 phosphorylation, it is not sufficient to reestablish basal levels of TTF-1 transcriptional activity in Ha-ras transformed cells. Taken together, these observations indicate that phosphorylation of TTF-1 at the N-terminal site is a rather complex mechanism in which different kinases are involved. We have to consider that Ha-ras induces cellular transformation by activating several signal transduction pathways including the Raf-MEK-MAPK and the PKC cascades (28), and also a modest reduction in PKA activity has been reported in Ha-ras transformed FRTL-5 cells (39). How these pathways integrate to precisely modulate TTF-1 function remains to be elucidated.
|
Footnotes |
|---|
2 Both authors contributed equally to this work and should be
considered first authors.
Received November 19, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Y. Maeda, V. Dave, and J. A. Whitsett Transcriptional Control of Lung Morphogenesis Physiol Rev, January 1, 2007; 87(1): 219 - 244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Garcia-Jimenez, M. A. Zaballos, and P. Santisteban DARPP-32 (Dopamine and 3',5'-Cyclic Adenosine Monophosphate-Regulated Neuronal Phosphoprotein) Is Essential for the Maintenance of Thyroid Differentiation Mol. Endocrinol., December 1, 2005; 19(12): 3060 - 3072. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. De Vita, L. Bauer, V. M. C. da Costa, M. De Felice, M. G. Baratta, M. De Menna, and R. Di Lauro Dose-Dependent Inhibition of Thyroid Differentiation by RAS Oncogenes Mol. Endocrinol., January 1, 2005; 19(1): 76 - 89. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. deFelice, D. Silberschmidt, R. DiLauro, Y. Xu, S. E. Wert, T. E. Weaver, C. J. Bachurski, J. C. Clark, and J. A. Whitsett TTF-1 Phosphorylation Is Required for Peripheral Lung Morphogenesis, Perinatal Survival, and Tissue-specific Gene Expression J. Biol. Chem., September 12, 2003; 278(37): 35574 - 35583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M-C.W. Yang, B. Wang, J.C. Weissler, L.R. Margraf, and Y-S. Yang BR22, a 26 kDa thyroid transcription factor-1 associated protein (TAP26), is expressed in human lung cells Eur. Respir. J., July 1, 2003; 22(1): 28 - 34. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Medina, M. Rivas, P. Cruz, I. Barroso, J. Regadera, and P. Santisteban RhoA Activation Promotes Transformation and Loss of Thyroid Cell Differentiation Interfering with Thyroid Transcription Factor-1 Activity Mol. Endocrinol., January 1, 2002; 16(1): 33 - 44. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Lonigro, D. Donnini, E. Zappia, G. Damante, M. E. Bianchi, and S. Guazzi Nestin Is a Neuroepithelial Target Gene of Thyroid Transcription Factor-1, a Homeoprotein Required for Forebrain Organogenesis J. Biol. Chem., December 14, 2001; 276(51): 47807 - 47813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-S. Yang, M.-C. W. Yang, B. Wang, and J. C. Weissler BR22, a Novel Protein, Interacts with Thyroid Transcription Factor-1 and Activates the Human Surfactant Protein B Promoter Am. J. Respir. Cell Mol. Biol., January 1, 2001; 24(1): 30 - 37. [Abstract] [Full Text]
|
|
|
|
|
|
L. Perrone, G. Tell, and R. Di Lauro Calreticulin Enhances the Transcriptional Activity of Thyroid Transcription Factor-1 by Binding to Its Homeodomain J. Biol. Chem., February 19, 1999; 274(8): 4640 - 4645. [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 |
