Functional interaction among thyroid-specific transcription factors: Pax8 regulates the activity of Hex promoter

Molecular and Cellular Endocrinology 214 (2004) 117–125

Functional interaction among thyroid-specific transcription factors: Pax8 regulates the activity of Hex promoter Cinzia Puppin a , Ivan Presta b , Angela V. D’Elia a , Gianluca Tell c , Franco Arturi b , Diego Russo b , Sebastiano Filetti d , Giuseppe Damante a,e,∗ a

b

Dipartimento di Scienze e Tecnologie Biomediche, Università di Udine, Piazzale Kolbe 1, 33100 Udine, Italy Dipartimento di Scienze Farmacologiche e Dipartimento di Medicina Sperimentale e Clinica, Università di Catanzaro, Catanzaro, Italy c Dipartimento di Biochimica, Biofisica e Chimica della Macromolecole, Università di Trieste, Trieste, Italy d Dipartimento di Scienze Cliniche, Università di Roma “la Sapienza”, Roma, Italy e M.A.T.I. Center, Italy Received 10 June 2003; accepted 28 October 2003

Abstract The transcription factor Hex is expressed in the thyroid follicular cells (TFC) and in several other cell types. In TFC, Hex contributes to the control of the tissue-specific gene expression. By means of RT-PCR assays we found a correlation between the Hex and Pax8 (a different tissue-specific transcription factor, expressed in TFC) mRNA levels in normal and neoplastic thyroid tissues. This finding suggested the presence of a functional correlation between the two transcription factors. Therefore, we tested whether Pax8 regulates the transcriptional activity of Hex promoter. Indeed, by using cotransfection experiments in non-thyroidal cells, we show that increasing doses of Pax8 expression vector elicited a dose-dependent increase of the transcriptional activity of Hex promoter. Accordingly, gel-retardation assays indicated that in the Hex promoter are present several Pax8 binding sites. The Pax8 activating effect on Hex promoter was further increased by the contemporary presence of Hex protein. In fact, cotransfection of both Hex and Pax8 expression vectors doubled the transcriptional activity of Hex promoter with respect to the condition in which the Pax8 expression vector only was transfected. In addition, we show that also the transcriptional cofactor APE/Ref-1 cooperated with Pax8 for upregulation of Hex promoter activity. These findings, together with other published data, suggest that a network of functional interactions between transcriptional regulators is present in TFC. © 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Hex; Pax8; Transcription factor; Promoter; Thyroid

1. Introduction Cell type-specific transcription factors are major control elements of cell differentiation (Kingston and Green, 1994; Struhl, 1999; Lai and Darnell, 1991; Dasen and Rosenfeld, 2001). In the thyroid follicular cell (TFC), several cell type-specific transcription factors have been identified, including TTF-1, TTF-2, Pax8 and Hex (Damante et al., 2001; Pellizzari et al., 2000). Knockout mice have revealed that each of these factors is critical for development of the thyroid gland (Kimura et al., 1997; Mansouri et al., 1998; De Felice et al., 1998; Martinez Barbera et al., ∗ Corresponding author. Tel.: +39-0432-494374; fax: +39-0432-494379. E-mail address: [email protected] (G. Damante).

2000). Accordingly, it has been shown that inactivating mutations of TTF-1, TTF-2 and Pax8 genes give rise to congenital hypothyroidism in humans (Macchia et al., 1998; Clifton-Bligh et al., 1998; Krude et al., 2002). In addition to developmental roles, TTF-1, TTF-2, Pax8 and Hex regulate cell type-specific gene expression in the adult TFC. All of these factors, for example, play a role in the control of transcriptional activity of thyroglobulin promoter (Damante et al., 2001). Therefore, in order to understand molecular mechanisms of TFC differentiation, it is critical to delineate how expression of thyroid-specific transcription factors is controlled. Some molecular mechanisms responsible for maintenance of thyroid-specific transcription factors in the adult TFC have been recently proposed. In fact, several data demonstrate the presence of positive feedback loops between thyroid-specific

0303-7207/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mce.2003.10.061

118

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

transcription factors themselves. It has been shown, for example, that TTF-1 binds to its own gene promoter, upregulating its transcriptional activity (Oguchi and Kimura, 1998). Moreover, we have recently shown that either TTF-1 or Hex proteins are able to regulate the transcriptional activity of Hex promoter (Puppin et al., 2003). In the present study the finding of a positive relationship between Pax8 and Hex mRNA levels in normal and pathological human thyroid tissues suggests the possibility that Pax8 is able to regulate the transcriptional activity of Hex promoter. Indeed, our data indicate that, in non-thyroidal cells, Pax8 upregulates the Hex promoter activity and that its action is enhanced by the overexpression of either the redox factor APE/Ref-1 or the Hex protein itself. Altogether, these findings suggest the presence in TFC of a network of functional interactions between transcriptional regulators.

2. Materials and methods 2.1. PAX8-HEX RT-PCR from thyroid tissues Twenty-two thyroid tumors (10 cold adenomas and 12 differentiated thyroid carcinomas, six follicular and six papillary) and eight non-nodular, normal thyroid tissues were analyzed. Informed consent was obtained by all patients. Total RNA extraction and cDNA synthesis were performed as previously described (Arturi et al., 2001) and the RT-PCR amplification was performed using 3 ␮l of cDNA as previously described (Arturi et al., 1998). The samples were subjected to 27 cycles of amplification. The PCR conditions were as follows: a first step of predenaturation at 95 ◦ C for 10 min; denaturation at 95 ◦ C for 30 s, annealing at 59 ◦ C for 30 s and extension at 72 ◦ C for 30 s. Primers oligonucleotides for the human PAX8 gene were: 5 -CAACCTCCCTATGGACAGCT-3 and 5 -CATCCGTGCGAAGGTGCTTT-3 . The amplification yielded a 259 base pair DNA product, corresponding to fragment 430-689 of the human PAX8 gene, according to the sequence reported in the GenBank accession no. L19606. Primers oligonucleotides for the human HEX gene were: 5 -TACTCTGGAGCCCCTTCTTG-3 and 5 -TTCAAGGTCTTCCTGGGAGG-3 . The amplification yielded a 370 base pair DNA product, corresponding to fragment 397–767 of the human HEX gene, according to the sequence reported in the GenBank accession no. XM018176. The reaction conditions were optimized to ensure that the amplification for both couples of mRNAs remained within the exponential range. Different ratios of primers were tested to get the same efficiency of amplification and the ratio 1:1 was used (data not shown). To confirm the data obtained by eteroduplex RT-PCR for PAX8/HEX, we also performed, on the same samples, a semiquantitative RT-PCR coupling amplification of PAX8 or HEX with a human beta-actin gene fragment (data not shown). PCR conditions to amplify

beta-actin together with PAX8 or together with HEX were: one cycle at 95 ◦ C for 10 min, 30 cycles of 95 ◦ C for 30 s, 59 ◦ C for 30 s, 72 ◦ C for 30 and one cycle at 72 ◦ C for 10 min. Three microliter of cDNA were incubated into the PCR mix with one of the oligonucleotide pair listed above, for the amplification of PAX8 or HEX, and the following pair used to amplify a 483 bp fragment of the human beta Actin gene (fragment 241–723, GenBank accession no. NM001101): 5 -CGAGGCCCAGAGCAAGAGA-3 and 5 -CACAGCTTCTCCTTA ATGTCAC-3 . Ten of 50 ␮l of the amplification products were run on 1.5 tris–borate– EDTA agarose gel containing ethidium bromide. The bands of the positive gel were scanned and the density and the width of each PCR product was measured using the NIH Image Program (Wayne Rasband, National Institutes of Health, USA). The correlation between Hex and Pax8 transcript levels was studied by linear regression analysis using the Prism program (GraphPad). A level of P < 0.05 was considered statistically significant. A RT-PCR was performed to verify the presence of Pax8 transcript in total RNA extracted from HeLa cells using the above-described primers and conditions. No specific band was visible, even increasing the number of cycles of amplification up to 40 (data not shown). 2.2. Plasmids The constructs, containing the mouse Hex promoter fragments −103/+22 and −235/+22, have been previously described (Denson et al., 2000). In these constructs, the fragments −103/+22 and −235/+22 of the Hex promoter were cloned in the plasmid pGL3B, 5 to the luciferase (LUC) gene. Mutations of the −222 and −57 Pax8 binding sites were performed in the context of the −235/+22 construct by using the Quick Change Site Directed mutagenesis kit (Stratagene) and were called −222 m and −57 m, respectively. RSV–CAT and CMV–LUC plasmids contained the Rous Sarcoma Virus (RSV) and the Cytomegalovirus (CMV) promoters linked to the chloranphenicol acetyl-transferase (CAT) and luciferase (LUC) genes, respectively. The Pax8 and APE/Ref-1 expression vectors have been previously described (Tell et al., 1998). 2.3. Cell cultures and transfections HeLa cells were cultured in DMEM medium with 10% calf serum (Gibco, Milan, Italy). The calcium phosphate co-precipitation method was used for transfections, as described elsewhere (Tell et al., 1998). HeLa cells were plated at 6 × 105 cells/100 mm culture dish 20 h prior to transfection. The plasmids were used in the following amounts: CMV–LUC, 2 ␮g; pGL3B, 8 ␮g; −103/+22 and −235/+22 Hex promoter, 8 ␮g; Pax8 expression vector, 0.5 and 0.2 ␮g; APE/Ref-1 expression vector 2 ␮g,

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

RSV–CAT, 2 ␮g. Cells were harvested 42–44 h after transfection, and cell extracts were prepared by a standard freeze and thaw procedure. CAT activity was measured by an ELISA method (Amersham). LUC activity was measured by a chemiluminescence procedure (Tell et al., 1998). 2.4. Gel-retardation assay Pax8 binding to the Hex promoter region −235/+22 was investigated using a series of overlapping oligonucleotides (shown in Fig. 4). Double-stranded oligonucleotides, labeled at the 5 end terminal with 32 P, were used as probes. The DNA binding domain of Pax8 (the Pax8 paired domain) was used as protein. In gel-retardation assays, protein and DNA (both at the final concentration of 0.1 ␮M) were incubated for 30 min at room temperature in a buffer containing 20 mM Tris–HCl (pH 7.6), 75 mM KCl, 0.25 mg/ml bovine serum albumin (BSA), 5 mM dithiotreitol (DTT), 10 ␮g/ml calf thymus DNA, and 10% glycerol. Protein-bound DNA and free DNA were separated on native 7.5% polyacrylamide gel run in 0.5× TBE for 1.5 h at 4 ◦ C. Gels were dried, exposed to a BIO-RAD GS-525 Molecular Imager and quantified by the Multianalyst software. The DNA-binding activity was obtained by calculating the protein-bound/free DNA ratio.

119

3. Results 3.1. Hex and Pax8 mRNA levels in human thyroid tissues The possibility of functional interactions between Hex and Pax8 genes was firstly evaluated by measuring their mRNA levels in normal human thyroid tissues. As shown in Fig. 1, Hex and Pax8 mRNA levels are positively correlated (P = 0.029; r2 = 0.51). Although a certain degree of variability was detectable among the pathological samples, a significant correlation between the levels of the Hex and Pax8 transcripts was found both in cold adenomas (P < 0.001; r 2 = 0.91) and in differentiated carcinomas (P < 0.001; r 2 = 0.95) (Fig. 2). The correlation was maintained even after correction of the transcripts expression levels with the beta-actin housekeeping gene (data not shown). These findings suggest the presence of a relationship between Pax8 and Hex genes. 3.2. Effects of Pax8 on Hex promoter The recent identification of mouse Hex promoter (Denson et al., 2000) allows us to verify whether Pax8 regulates Hex gene expression. We used two previously described constructs (Tanaka et al., 1999), each containing distinct deletions of the mouse Hex 5 -flanking sequence (−235/+22

Fig. 1. Correlation between Hex and Pax8 mRNA levels in normal human thyroid tissues. Hex and Pax8 mRNA levels were evaluated by multiplex RT-PCR as described in Section 2. (Top) Agarose gel stained by ethidium bromide. The Hex band is 370 bp-long; the Pax8 band is 259 bp-long. (Bottom) Correlation of values obtained by scanning densitometry of the gel. Parameters P and r2 were calculated by the Prism software (GraphPad).

120

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

Fig. 2. Correlation between Hex and Pax8 mRNA levels in thyroid neoplastic tissues. (Left panel) correlation in cold adenomas (no. 10). (Right panel) Correlation in differentiated carcinomas (no. 6 papillary and no. 6 follicular). Parameters P and r2 were calculated by the Prism software (GraphPad).

and −103/+22). To avoid interference by the native thyroid Pax8, we tested the effect of Pax8 on Hex promoter activity in the non-thyroid HeLa cells, in which the Pax8 transcript resulted absent, as analyzed by RT-PCR (data not shown). Thus, HeLa cells were transfected with constructs bearing the −235/+22 and −103/+22 deletions of Hex promoter, with or without a Pax8-expressing plasmid. The CMV promoter was used to normalize the promoter activities. Increasing doses of Pax8 expression vector (0.2 and 0.5 ␮g) elicited a dose-dependent increase of the transcriptional activity of both −235/+22 and −103/+22 Hex promoter fragments (Fig. 3). These data indicate that Pax8 increases

the activity of Hex promoter, when expressed in non-thyroid cells. 3.3. Pax8-binding sites on mouse Hex promoter The finding that Pax8 is able to increase the activity of Hex promoter predicts the existence of Pax8-binding sites on this promoter. This possibility was tested by gel-retardation assay. Overlapping double-stranded oligonucleotides, spanning the entire mouse Hex promoter were constructed (Fig. 4). The binding of the DNA-binding domain of Pax8 (the Pax8 paired domain, see Pellizzari et al.,

Fig. 3. Effect of Pax8 expression vector on the transcriptional activity of Hex promoter in HeLa cells. Effect of cotransfection of 0.2 and 0.5 ␮g of the Pax8 expression vector on transcriptional activity of the −235/+22 and −103/+22 Hex promoter deletions. Each point indicate the mean value ± standard deviation of four independent transfections.

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

121

Fig. 4. Pax8 binding on Hex promoter. (Top) Sequence of the mouse Hex promoter; arrowed lines indicate position of oligonucleotides used in gel-retardation assays. (Bottom) Pax8 binding activity to oligonucleotides of Hex promoter. The binding activity is expressed as the fraction of the binding value obtained for the site C of thyroglobulin promoter. For all oligonucleotides, the Pax8 paired domain was used at the final concentration of 0.1 ␮M.

1999) to these oligonucleotides was then measured. In Fig. 4, the binding activity of the Pax8 paired domain to the Hex promoter oligonucleotides is expressed as the fraction of the binding value obtained for the site C of thyroglobulin promoter, a well characterized high-affinity Pax8 binding site (Pellizzari et al., 1999). Several binding sites were found, with sites −222 and −57 showing the highest Pax8 binding activity. As an example, the Pax8 binding to sites −222 and −57 is shown in Panel A of Fig. 5. A significant homology between the −222 oligonucleotide and the Pax8 consensus sequence (Damante et al., 2001) is present (Fig. 5, Panel B). A weaker homology is present in the −57 oligonucleotide. In order to show the functional relevance of the Pax8-binding sequences, either the −222 or the −57 binding site was subjected to site-directed mutagenesis to disrupt the Pax8 consensus (Fig. 5, Panel B). The effect of either mutation was tested by cotransfection assays. As shown in Panel C of Fig. 5, either mutation attenuate the Pax8 effect on Hex promoter activity. These data suggest that the Pax8 binding sites identified in the Hex promoter have an “in vivo” functional relevance. 3.4. Cooperation with other transcriptional regulators A single transcription unit is usually regulated by several different factors. For this reason we tested whether other transcriptional regulators cooperates with Pax8 in activat-

ing Hex promoter. We focused on two factors: APE/Ref-1 and Hex protein. It has been demonstrated that the multifunctional protein APE/Ref-1 upregulates the transcriptional effects exerted by Pax8 on the thyroglobulin promoter (Tell et al., 1998). By cotransfection experiments, we verified whether a similar effect occurred in the context of the Hex promoter. Results are shown in Panel A of Fig. 6. Cotransfection of an APE-Ref-1 expression vector doubles the −235/+22 and −103/+22 Hex promoter activities in presence of different doses of the Pax8 expression vector. APE/Ref-1 can be considered, therefore, as a regulator of Hex promoter. The protein Hex, depending on the transcription units to which is bound, can exert either inhibitory or activatory effects (Pellizzari et al., 2000; Tanaka et al., 1999; Sekiguchi et al., 2001). We have recently demonstrated that the Hex protein activates its own promoter and that its action is additive to that exerted by TTF-1 (Puppin et al., 2003). Panel B of Fig. 6 shows the effect exerted by Hex protein on its own gene promoter, either in the absence or presence of Pax8. As previously shown (Puppin et al., 2003), Hex protein is able to activate its own gene promoter. The activating effect is detectable also in the presence of 0.2 and 0.5 ␮g of Pax8 expression vector, either on −235/+22 or −103/+22 Hex promoter fragments. On both promoters the expression of Hex protein doubles the activating effects exerted by Pax8.

122

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

Fig. 5. Major Pax8-binding sites on Hex promoter. (Panel A) Representative gel retardation assays by using oligonucleotides C, −222 and −57 (see Fig. 4). B, protein-bound DNA; F, free DNA.The Pax8 paired domain was used at concentrations of 0.03, 0.1 and 0.3 ␮M. (Panel B) Sequence homology of −222 and −57 oligonucleotides to the Pax8 consensus sequence. Boxes indicates nucleotides that have been changed to disrupt the Pax8 binding site in the −222 and −56 sequences (corresponding mutants are called −222 m and −57 m). (Panel C) Pax8 transcriptional activity on mutant Hex promoters. 0.5 ␮g of the Pax8 expression vector was used. Each bar indicate the mean value ± standard deviation of four independent transfections.

4. Discussion The finding of a positive correlation between Pax8 and Hex mRNA levels both in normal and pathological thyroid tissues suggests two, not mutually exclusive, possibilities. The first possibility consists in the presence of an upstream control mechanism that modulates the expression of both Pax8 and Hex genes in the same direction, without any form of functional interaction between the two factors. In our opinion, however, this latter mechanism is unlikely: the find-

ing that Pax8 expression, but not that of Hex, is suppressed in most undifferentiated thyroid carcinomas (Fabbro et al., 1994; D’Elia et al., 2002) suggests that the mechanisms of control for the expression of these two transcription factors are different. The second possibility is that one transcription factor controls the expression of the other one. Indeed, this second explanation is corroborated by our data. In fact, we show that Pax8 upregulates the Hex promoter activity. In addition, Hex cooperates with Pax8 in activating its own promoter. We have recently shown that also TTF-1 is able

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

123

Fig. 6. Effects of APE/Ref-1 and Hex on the Pax8-activating effect on Hex promoter. HeLa cells were transfected as described in Section 2. Each point represent the mean value of three independent transfections. For each value, standard deviation was not over ±10%.

to upregulate Hex promoter activity (Puppin et al., 2003). Moreover, it has been demonstrated that TTF-1 is able to activate its own gene promoter (Oguchi and Kimura, 1998). Altogether, these findings demonstrate that thyroid-specific transcription factors establish networks of auto/cross regulatory circuits by interaction between protein products and gene promoters. Similar regulatory networks have been found in other systems. Positive cis-regulatory elements of the Pdx-1 gene (which plays a role in pancreas development and beta-cell function, see McKinnon and Docherty, 2001) are bound by several transcription factors including: HNF-3␤, HNF-1␣, SP1/3, and Pdx-1 itself (Melloul et al., 2002). Also, the expression of Krox20, a transcription fac-

tor expressed in the hindbrain and in neural crest cells migrating toward the third branchial arch, is regulated by an enhancer element bound by Krox20 itself ad Sox10, a crest specific HMG box protein (Ghislain et al., 2003). The biological meaning of these cross/auto regulatory circuits can be easily proposed. In fact, positive feed-back loops reinforce expression of a given factor in a given cell type and contribute to make a sharp boundary between expressing and not-expressing cell types (Jiang et al., 1991). The Hex gene is extremely important for thyroid development. In the Hex −/− mouse, development of the thyroid is arrested at the thyroid bud stage at 9.5 d.p.c. (Martinez Barbera et al., 2000). Thus, positive feed-back loops between

124

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125

transcriptional regulators converging on Hex promoter would increase robustness of Hex expression, providing a potential compensation for changes in gene dosage and, therefore, to reduce effects of loss-of-function mutations. It must be noted that the binding between protein products and gene promoters are not the only functional interactions established by transcription factors of the TFC. These proteins, in fact, may share the target genes. The thyroglobulin promoter, as example, is regulated by TTF-1, TTF-2, Pax8 and Hex (Damante et al., 2001; Pellizzari et al., 2000). A similar phenomenon occurs in the context of thyroperoxidase promoter (Damante and Di Lauro, 1994). In addition, thyroid-specific transcription factors establish protein–protein interactions, as recently demonstrated for TTF-1 and Pax8 (Di Palma et al., 2003). Deregulated functioning of transcription factors is crucially involved in the tumorigenic process and reduction or loss of tissue specific transcription factors expression may be considered the key alteration, leading to the disappearance of differentiated properties of a transformed cell. Accordingly, in transformed TFCs, the loss of TPO and thyroglobulin gene expression proceed in parallel with the reduction of TTF-1 and Pax8, as assessed in both human thyroid tissues and thyroid tumoral cells in culture (Fabbro et al., 1994; Chun et al., 1998). Our present data demonstrate that, as well as a positive relationship between Hex and TTF-1 previously reported (Puppin et al., 2003), a correlation between Pax8 and Hex transcripts is maintained in differentiated thyroid tumors (both benign cold adenomas and papillary and follicular carcinomas). This indicates that, in a first step of TFC transformation, the complex network of tissue specific transcription factors interaction is still working, assuring the maintenance of tissue specific target gene expression (such as thyroglobulin). During the progression toward less differentiated and more malignant phenotypes, additional alterations in the transcriptional machinery regulation determine an interruption of such positive feedback loops. Indeed, we have recently shown that Hex expression is retained in most human thyroid undifferentiated carcinomas (D’Elia et al., 2002), although Pax8 expression is abolished (Fabbro et al., 1994). In these tumors, therefore, Hex promoter activity seems to be maintained by factors other than Pax8, which is consistent with our observation of significant basal Hex promoter activity in the absence of Pax8 in non-thyroid cells (see Fig. 3). Moreover, an important role in the modulation of transcription factor function is played by the reversible post-translational modifications induced by co-factors able to act on DNA binding, transactivation/transrepression, dimerization, subcellular localization or protein stability (Calkhoven and Ab, 1996). One of these has been recently identified in the apurinic/apyrimidinic endonuclease/redox effector factor Ape/Ref-1 (Evans et al., 2000). In previous reports we have shown that this co-factor is able to regulate, through its redox activity, the DNA-binding function of Pax8 (Tell et al., 1998). Furthermore, an altered subcellular localization of APE/Ref-1 has been detected in

both thyroid tumor tissues and cell lines (Tell et al., 2000; Russo et al., 2001). The present data provide an additional role for APE/Ref-1 co-factor, independent of its modulation of Pax8 function, by directly regulating Hex promoter activity. Elucidation of the reciprocal interactions among the thyroid tissue-specific transcriptional regulatory molecules may provide new information to clarify the complex signaling networks, culminating in oncogenic transformation. Therefore, it may suggest new tools of intervention to restore the differentiated properties in the transformed cells in view of a more effective therapeutical approach.

Acknowledgements This work is funded by grants from the Consiglio Nazionale delle Ricerche (CNR, Target project on Biotechnology) to GD, from Regione Friuli Venezia-Giulia to GD, and from MIUR-COFIN to GD and SF.

References Arturi, F., Russo, D., Bidart, J.M., Scarpelli, D., Schlumberger, M., Filetti, S., 2001. Expression pattern of the pendrin and sodium/iodide symporter genes in human thyroid carcinoma cell lines and human thyroid tumors. Eur. J. Endocrinol. 145, 129–135. Arturi, F., Russo, D., Schlumberger, M., du Villard, J.A., Caillou, B., Vigneri, P., Wicker, R., Chiefari, E., Suarez, H.G., Filetti, S., 1998. Iodide symporter gene expression in human thyroid tumors. J. Clin. Endocrinol. Metab. 83, 2493–2496. Calkhoven, C.F., Ab, G., 1996. Multiple steps in the regulation of transcription-factors level and activity. Biochemical J. 317, 329–342. Chun, Y.S., Saji, M., Zeiger, M.A., 1998. Overexpression of TTF-1 and PAX-8 restores thyroglobulin gene promoter activity in ARO and WRO cell lines. Surgery 124, 1100–1105. Clifton-Bligh, R.J., Wentworth, J.M., Heinz, P., Crisp, M.S., John, R., Lazarus, J.H., Ludgate, M., Chatterjee, V.K., 1998. Mutation of the gene encoding human TTF-2 associated with thyroid agenesis, cleft palate and choanal atresia. Nat. Genet. 19, 399–401. Damante, G., Di Lauro, R., 1994. Thyroid-specific gene expression. Biochim. Biophys. Acta 1218, 255–266. Damante, G., Tell, G., Di Lauro, R., 2001. A unique combination of transcription factors controls differentiation of thyroid cells. Prog. Nucl. Acid Res. Mol. Biol. 66, 307–356. Dasen, J.S., Rosenfeld, M.G., 2001. Signaling and transcriptional mechanisms in pituitary development. Annu. Rev. Neurosci. 24, 327–355. De Felice, M., Ovitt, C., Biffali, E., Rodriguez-Mallon, A., Arra, C., Anassiatidis, K., Macchia, P.E., Mattei, M.G., Mariano, A., Scholer, H., Macchia, V., Di Lauro, R., 1998. A mouse model for hereditary thyroid dysgenesis and cleft palate. Nat. Genet. 19, 395–398. D’Elia, A.V., Tell, G., Russo, D., Arturi, F., Puglisi, F., Manfioletti, G., Gattei, V., Mack, D.L., Cataldi, P., Filetti, S., Di Loreto, C., Damante, G., 2002. Expression and localization of the homeodomain-containing protein Hex in human thyroid tumors. J. Clin. Endocrinol. Metab. 87, 1376–1383. Denson, L.A., McClure, M.H., Bogue, C.W., Karpen, S.J., Jacobs, H.C., 2000. HNF3beta and GATA-4 transactivate the liver-enriched homeobox gene. Hex. Gene 246, 311–320. Di Palma, T., Nitsch, R., Mascia, A., Nitsch, L., Di Lauro, R., Zannini, M., 2003. The paired domain-containing factor Pax8 and the

C. Puppin et al. / Molecular and Cellular Endocrinology 214 (2004) 117–125 homeodomain-containing factor TTF-1 directly interact and synergistically activate transcription. J. Biol. Chem. 278, 3395–3402. Evans, A.R., Limp-Foster, M., Kelley, M.R., 2000. Going APE over ref-1. Mutat. Res. 461, 83–108. Fabbro, D., Di Loreto, C., Beltrami, C.A., Belfiore, A., Di Lauro, R., Damante, G., 1994. Expression of thyroid-specific transcription factors TTF-1 and Pax-8 in human thyroid neoplasms. Cancer Res. 54, 4744– 4749. Ghislain, J., Desmarquet-Trin-Dinh, C., Gilardi-Hebenstreit, P., Charnay, P., Frain, M., 2003. M Neural crest patterning: autoregolatory and crest-specific elements co-operate for Krox20 transcriptional control. Development 130, 941–953. Jiang, J., Hoey, T., Levine, M., 1991. Autoregulation of a segmentation gene in Drosophila: combinatorial interaction of the even-skipped homeobox protein with a distal enhancer element. Genes Dev. 5, 265– 277. Kimura, S., Hara, Y., Pineau, T., Fernandez-Salguero, P., Fox, C.H., Ward, J.M., Gonzalez, F.J., 1997. The T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev. 10, 60–69. Kingston, R.E., Green, M.R., 1994. Modeling eukaryotic transcriptional activation. Curr. Biol. 4, 325–332. Krude, H., Schutz, B., Biebermann, H., von Moers, A., Schnabel, D., Neitzel, H., Tonnies, H., Weise, D., Lafferty, A., Schwarz, S., De Felice, M., von Deimling, A., van Landeghem, F., Di Lauro, R., Gruters, A., 2002. Choreoathetosis, hypothyroidism, and pulmonary alterations due to human NKX2-1 haploinsufficiency. J. Clin. Invest. 109, 475–480. Lai, E., Darnell Jr., J.E., 1991. Transcriptional control in hepatocytes: a window on development. Trends Biochem. Sci. 16, 427–430. Macchia, P.E., Lapi, P., Krude, H., Pirro, M.T., Missero, C., Chiovato, L., Souabni, A., Baserga, M., Tassi, V., Pinchera, A., Fenzi, G., Gruters, A., Busslinger, M., Di Lauro, R., 1998. PAX8 mutations associated with congenital hypothyroidism caused by thyroid dysgenesis. Nat. Genet. 19, 83–85. Mansouri, A., Chowdury, K., Gruss, P., 1998. Follicular cells of the thyroid gland require Pax8 gene function. Nat. Genet. 19, 87–90. Martinez Barbera, J.P., Clements, M., Thomas, P., Rodriguez, T., Meloy, D., Kioussis, D., Beddington, R.S., 2000. The homeobox gene Hex is required in definitive endodermal tissues for normal forebrain, liver and thyroid formation. Development 127, 2433–2445.

125

McKinnon, C.M., Docherty, K., 2001. Pancreatic duodenal homeobox-1, PDX-1, a major regulator of beta-cell identity and function. Diabetologia 44, 1203–1214. Melloul, D., Marshak, S., Cerasi, E., 2002. Regulation of pdx-1 gene expression. Diabetes 51 (Suppl. 3), S320–S325. Oguchi, H., Kimura, S., 1998. Multiple transcripts encoded by the thyroid-specific enhancer-binding protein (T/EBP)/thyroid-specific transcription factor-1 (TTF-1) gene: evidence of autoregulation. Endocrinology 139, 1999–2006. Pellizzari, L., D’Elia, A., Rustighi, A., Manfioletti, G., Tell, G., Damante, G., 2000. Expression and function of the homeodomain-containing protein Hex in thyroid cells. Nucl. Acid Res. 28, 2503–2511. Pellizzari, L., Tell, G., Damante, G., 1999. Co-operation between the PAI and RED subdomains of Pax-8 in the interaction with the thyroglobulin promoter. Biochem. J. 337, 253–262. Puppin, C., D’Elia, A.V., Pellizzari, L., Russo, D., Arturi, F., Presta, I., Filetti, S., Bogue, C.W., Denson, L.A., Damante, G., 2003. Thyroid-specific transcription factors control Hex promoter activity. Nucl. Acid Res. 31, 1845–1852. Russo, D., Arturi, F., Bulotta, S., Pellizzari, L., Filetti, S., Manzini, G., Damante, G., Tell, G., 2001. Ape1/Ref-1 expression and cellular localization in human thyroid carcinoma cell lines. J. Endocrinol. Invest. 24, RC10–RC12. Sekiguchi, K., Kurabayashi, M., Oyama, Y., Aihara, Y., Tanaka, T., Sakamoto, H., Hoshino, Y., Kanda, T., Yokoyama, T., Shimomura, Y., Iijima, H., Ohyama, Y., Nagai, R., 2001. Homeobox protein Hex induces SMemb/nonmuscle myosin heavy chain-B gene expression through the cAMP-responsive element. Circ. Res. 88, 52– 58. Struhl, K., 1999. Fundamentally different logic of gene regulation in eukaryotes and prokaryotes. Cell 98, 1–4. Tanaka, T., Inazu, T., Yamada, K., Myint, Z., Keng, V.W., Inoue, Y., Taniguchi, N., Noguchi, T., 1999. cDNA cloning and expression of rat homeobox gene, Hex, and functional characterization of the protein. Biochem. J. 339, 111–117. Tell, G., Pellizzari, L., Cimarosti, D., Pucillo, C., Damante, G., 1998. Ref-1 controls Pax-8 DNA-binding activity. Biochem. Biophys. Res. Commun. 252, 178–183. Tell, G., Pellizzari, L., Pucillo, C., Puglisi, F., Cesselli, D., Kelley, M.R., Di Loreto, C., Damante, G., 2000. TSH controls Ref-1 nuclear translocation in thyroid cells. J. Mol. Endocrinol. 24, 383–390.