Retinoblastoma protein acts as Pax 8 transcriptional coactivator

Oncogene (2005) 24, 6993–7001

& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc

ORIGINAL PAPERS

Retinoblastoma protein acts as Pax 8 transcriptional coactivator Stefania Miccadei1,4, Claudia Provenzano2,4, Martin Mojzisek1,3, Pier Giorgio Natali1 and Donato Civitareale*,1,2 1 Molecular Pathology Laboratory, Regina Elena Cancer Institute, Via delle Messi d’Oro 156, 00158 Rome, Italy; 2Institute of Neurobiology and Molecular Medicine, National Council Research, Via del Fosso del Cavaliere 100, 00133 Rome, Italy; 3Department of Medical Biology and Genetics, Charles University, Faculty of Medicine in Hradec Kra´love´, Simkova 870, 500 01 Hradec Kralove, Czech Republic

Control of cell proliferation and differentiation by the retinoblastoma protein (pRb) depends on its interactions with key cellular substrates. Available data indicate that pRb and the transcription factor Pax 8 play a crucial role in the differentiation of thyroid follicular cells. In this study, we show that pRb takes part in the complex assembled on the thyroperoxidase gene promoter acting as a transcriptional coactivator of Pax 8. Accordingly, pRb interacts with and potentiates Pax 8 transcriptional activity. In addition, we show that the downregulation of pRb gene expression, in thyrocytes, through RNA interference results in a reduction of the thyroperoxidase gene promoter activity mediated by the Pax 8-binding site. In agreement with these results and with the ability of the adenoviral protein E1A to bind pRb, we show that E1A downregulates Pax 8 activity and that such inhibition requires the E1A–Rb interaction. Furthermore, we show that the Pax 8/pRb synergy plays a role on the sodium/ iodide symporter gene expression as well. Oncogene (2005) 24, 6993–7001. doi:10.1038/sj.onc.1208861; published online 27 June 2005 Keywords: Pax 8; retinoblastoma protein; thyroid; transcription; coactivator

Introduction Terminal differentiation invariably involves two closely linked phenomena: permanent withdrawal from the cell cycle and biochemical differentiation characterized by the tissue-specific gene expression. The retinoblastoma gene product, pRb has been implicated in mediating both the permanent cell cycle arrest and upregulation of tissue-specific genes in a wide variety of tissues (Weinberg, 1995; Lipinski and Jacks, 1999). In deciphering the function of pRb, it has been crucial to identify its *Correspondence: D Civitareale, Institute of Neurobiology and Molecular Medicine, Italian National Council Research, Via del Fosso del Cavaliere 100, 00133 Rome, Italy; E-mail: [email protected] 4 These two authors contributed equally to this work Received 1 July 2003; revised 16 May 2005; accepted 18 May 2005; published online 27 June 2005

binding partners. It interacts with the virally encoded transforming proteins Large T antigen (Tag), E1A and E7 of the DNA tumor viruses SV40, adenovirus and human papilloma virus (Chellappan et al., 1992). Mutations in the pRb-binding domain prevent cellular transformation by these transforming viral proteins. These results led to the preliminary conclusion that pRb binds E2F transcription factors and negatively regulates their activity in cell cycle progression (Nevins, 2001). Like Large Tag, E1A and E7, E2F binds only the hypophosphorylated forms of pRb present during the G1 phase of the cell cycle. Thus, phosphorylation of pRb, mainly mediated by cyclinD/cdk4 and cyclin E/cdk2, increases the activity of promoters containing the E2F-binding sites and thereby induces the transcription of several important genes involved in cell cycle regulation (Adams, 2001). Thus, most of the growth suppressive properties of pRb and of its related ‘pocket proteins’ p107 and p130 are mediated through their modulation of the activity of E2F factors (Harbour and Dean, 2000). Furthermore, a regulatory loop has been proposed as one of the mechanisms by which pRb couples permanent exit from mitosis with tissue-specific gene expression. pRb also induces the expression of p21 (Decesse et al., 2001). This cell cycle-dependent kinase inhibitor leads to inactivation of multiple cyclin/cdk complexes and allows pRb to remain hypophosphorylated and active. Studies with skeletal muscle cells have shown that the ability of MyoD and other members of the basic helix– loop–helix family of myogenic transcription factors to regulate differentiation depends on the presence of pRb (Puri and Sartorelli, 2000). Similarly, pRb positively regulates terminal adipocyte differentiation through direct interaction with C/EBPs (Chen et al., 1996a). In addition, in hematopoietic cells, it regulates NF-IL6 activity (Chen et al., 1996b). Furthermore, it has been demonstrated that, in osteogenic differentiation, the central role exerted by the transcription factor CBFA1 requires pRb and that loss of pRb blocks late osteoblast differentiation. Coexpression of CBFA1 and pRb results in the association of both proteins with an osteoblastspecific promoter in vivo and consequent transcriptional activation. Thus, in osteoclast cells, pRb works as a positive transcriptional coactivator (Thomas et al., 2001).

Role of pRb in thyroid-specific gene expression S Miccadei et al

6994

Iuliano et al. have reported that pRB plays a pivotal role in thyroid cell differentiation and transformation. They have shown that the interaction of E1A with pRb is necessary to block differentiation, while an E1A mutant unable to bind pRb does not affect thyroid cell differentiation. Furthermore, they have demonstrated that pRb is required for the expression of the thyroid differentiation markers thyroglobulin (Tg) and thyroperoxidase (TPO) (Iuliano et al., 2000). The activity of Pax 8 is crucial in these pathways being a paired-domain-containing transcription factor with a key role in thyrocyte differentiation (Pasca di Magliano et al., 2000). Pax 8 gene knockout mice have smaller thyroids with normal calcitonin-producing parafollicular C cells but no follicular cells, thus they suffer from severe hypothyroidism (Mansouri et al., 1998). Furthermore, in humans, mutations in the Pax 8 gene are associated with dysgenesis of the thyroid gland (Macchia et al., 1998). It has been shown that Pax 8 activates the gene expression of the thyrocyte differentiation marker genes: Tg, TPO and sodium/iodide symporter (NIS) gene (Fabbro et al., 1998; Zannini et al., 1992; Esposito et al., 1998; Ohno et al., 1999). The Pax 8 role in TPO gene transcription is very well characterized. Both TPO gene promoter and TPO enhancer element are bound and activated by Pax 8 (Zannini et al., 1992; Esposito et al., 1998). Recently, it has been described that Pax 8 and a second tissue-specific transcription factor termed thyroid transcription factor 1 (TTF-1) act synergistically in the activation of both TPO and Tg gene transcription (Miccadei et al., 2002; Di Palma et al., 2003). Furthermore, we have previously reported that Pax 8 activation of the TPO promoter requires, as a transcriptional coactivator, p300 (De Leo et al., 2000). Based on the above-mentioned findings, pointing to a possible role of pRb as transcriptional coactivator and regulator of TPO gene expression, we have envisaged that Pax 8 and pRb could act cooperatively. This hypothesis is also supported by studies that show a direct interaction between pRb and factors containing paired-like homeodomains (Wiggan et al., 1998). Interestingly, Eberhard and Busslinger (1999) have demonstrated the direct interaction between Pax 5 and pRb both in vitro and in vivo and it has been reported that pRb interacts with Pax 2 (Yuan et al., 2002). It is worth to point out that structure–function analysis revealed that Pax 2, Pax 5 and Pax 8 proteins are structurally and functionally very similar (Dorfler and Busslinger, 1996). We show here that Pax 8 and pRb not only do interact but that pRb acts as a Pax 8 transcriptional coactivator on the TPO gene promoter.

Results

amino acids 379–928) fusion protein was expressed in Escherichia coli and bound to the glutathione-Sepharose resin (Amersham-Pharmacia). On this affinity column, we have loaded the protein cell extract from HeLa cell transfected with Pax 8-expressing vector. We collected three fractions from the column: the flow-through, the wash and the elution step at 1 M NaCl. To identify the fraction containing Pax 8, they were subsequently analysed in a band-shift assay. As shown in Figure 1A, Pax 8 was identified in the elution step at high salt concentration. In a control experiment, where the affinity column was prepared with the GST protein only, hence in absence of pRbLP, Pax 8 was identified in the flow-through fraction (data not shown). In addition, to demonstrate that pRb and Pax 8 directly interact, in Figure 1B we show the pull-down experiments performed using the chimeric protein GST-pRbLP and the in vitro translated Pax 8, Figure 1B, lane WT. Using the same assay and deletion mutants of Pax 8, we have been able to identify in the C-terminal region of Pax 8, the domain involved in the interaction, Figure 1B, lane Dad. Therefore, differently from the pRb/Pax 5 interaction where pRb binds the residual homeodomain of Pax 5 (Eberhard and Busslinger, 1999), our experiment suggests that pRb interacts with the activation domain of Pax 8, Figure 1B. As shown in Figure 1B, lane DHD, Pax 8 deleted of the residual homeodomain is still able to interact with pRbLP, although the reduced intensity of the Pax 8 band (lane 3) compared to the input (lane 1) would suggest that the DHD domain could partially contribute to the interaction. Moreover, to verify the interaction in vivo, we expressed the two proteins in transiently transfected HeLa cells. The Pax 8–pRb complexes were immunoprecipitated with an anti-Pax 8 antibody from the HeLa cells extracts and detected by Western blot analysis with an anti-Rb antibody, Figure 1C, lane b. In Figure 1C, lane a, is shown the same experiment using unspecific IgG in the immunoprecipitation. Furthermore, in order to demonstrate that the pRb/Pax 8 interaction is not the result of the high protein concentration achieved in transfected Hela cells, the antibody against pRb conjugated with agarose beads (Santa Cruz Biothecnology) was used to precipitate the endogenous pRb from nuclear proteins extracted from FRTL-5 cells. This is a stable cell line, derived from rat thyrocytes, able to maintain in culture the differentiated characteristic of the follicular thyroid cells (AmbesiImpiombato et al., 1980). The FRTL-5 nuclear extract and the immunoprecipitated proteins were analysed by Western blot with the Pax 8-specific antibody, Figure 1C, lane c and lane d, respectively. The pRb/ Pax 8 complex is demonstrated by the presence of the Pax 8 in the immunoprecipitated proteins. Hence, the Pax8/pRb complex is present in vivo in the thyrocytes nucleous.

Pax 8/pRb interaction In order to provide experimental evidence for the Pax 8/ pRb protein–protein interaction, we have followed a biochemical approach. A GST-RbLP (glutathione Stransferase fused to the large pocked of human pRB, Oncogene

Pax8/pRb synergism In order to assign a functional role to the Pax 8/pRb interaction, we took advantage of the strong activity shown by Pax 8 on the TPO promoter in non-thyroid

Role of pRb in thyroid-specific gene expression S Miccadei et al

6995

HeLa cells (Zannini et al., 1992). As shown in Figure 2a, cotransfection of a plasmid containing the luciferase reporter gene under the control of the TPO gene promoter (TPO-Luc, De Leo et al., 2000) with the Pax 8 encoding plasmid results in strong activation of the TPO promoter. Cotransfection of pRb and TPO-Luc results in very weak reporter gene expression, whereas the combined activity of pRb and Pax 8 results in cooperative activation of TPO gene promoter activity. This synergism requires Pax 8 bound to the promoter since a mutated TPO promoter, TPO-Pm, which does not bind Pax 8 (Zannini et al., 1992) cannot be activated by Pax 8 and the coexpression of pRb and Pax 8 does not affect TPO-Pm promoter activity (data not shown). As shown in Figure 2a, pRb synergy with Pax 8 is dose dependent and is not limited to HeLa cells but can also be demonstrated in SAOS-2 cells, Figure 2a. Since Pax 8 activates the Tg and NIS gene expression (Fabbro et al., 1998; Ohno et al., 1999), we addressed the question whether pRb is required on these gene promoters as well as on TPO promoter. In Figure 2b we

show that Rb does not cooperate, in the tranfection assays, with Pax 8 in the activation of the Tg gene promoter, whereas it enhances Pax 8 activity on NIS gene transcription. Hence, the Pax 8/pRb synegy is promoter specific. Since the adenoviral protein E1A has been shown to block thyrocyte differentiation (Berlingieri et al., 1993), to downregulate TPO gene expression (Iuliano et al., 2000) and to bind the pRb pocket domain (Whyte et al., 1988), we wondered whether E1A and Pax 8 could compete for the binding of pRb. To test this hypothesis, we overexpressed, in transient transfection experiments in HeLa cells, E1A and Pax 8. As shown in Figure 2c, E1A inhibits Pax 8 trans-activation whereas the E1A mutant (E1AM), unable to interact with pRb, does not affect Pax 8 activity. To further support our hypothesis, we show that the E1A inhibition of the Pax 8 activity is reduced by the forced expression of pRb in a dosedependent way. Thus, E1A inhibition of Pax 8 activity is mediated by pRb. In Table 1, we list the values obtained from transfection experiments in HeLa cells using the phosphorylation mutant of pRb, pRbDcdk (Lukas et al., 1999) and the other pocket proteins (Grana et al., 1998). Although to different extents, both p130 and p107 can cooperate with Pax 8. In addition, the results with the pRbDcdk mutant suggest that the pRb nonphosphorylated form can cooperate with Pax 8. This result is in agreement with the pull-down experiments shown in Figure 1 where the bacterially expressed pRb binds Pax 8.

Figure 1 In vitro and in vivo interaction of Pax 8 and pRb. (A) Band-shift assay of the input protein nuclear extract and fractions of the GST affinity column. Lane 1 shows the band-shift assay with the nuclear extract from HeLa cells overexpressing Pax 8. Lane 2 shows the same assay with nuclear extract from HeLa cells transfected with an empty vector. Lanes 3, 4 and 5 show the bandshift assays performed with the flow-through, wash and high salt elution step, respectively, of the GST affinity column loaded with the nuclear extract from HeLa cells overexpressing Pax 8. (B) Pulldown experiments with the full-length Pax 8 (lane WT); with Pax8Dad (lane Dad) and with Pax8DHD (lane DHD). Row 1 shows 10% of the input protein. Row 2, pull-down experiment with the GST control protein. Row 3, pull-down experiment with the GSTpRbLP protein. Schematic diagram of the Pax 8 full-length and mutants are shown on the right. The paired domain (pd), octapeptide (o), residual homeodomain (hd) and activation domain (ad) are indicated as well. In the Supplementary Information, we provide further controls to the pull-down experiment. (C) In vivo interaction of Pax 8 and pRb. Cell lysates from HeLa overexpressing both Pax 8 and pRb were precipitated with unspecific IgG, lane a, or with Pax 8 specific antibody, lane b. The immunoprecipitated proteins were analysed by Western blot and the filter was probed with the pRb-specific antibody C15. The arrowheads indicate the migration of the molecular weight markers of 85 and 118 kDa as indicated. In lane d, we show the coimmunoprecipitation from 10 mg of FRTL-5 nuclear extracts using the beads-conjugated anti-Rb antibody. The immunoprecipitated proteins were analysed by Western blot and the filter was probed with the Pax 8-specific antibody. As control, we show in lane c the Western blot of the FRTL-5 nuclear proteins probed with the same Pax 8 antibody. The arrowheads indicate the migration of the molecular weight markers of 47 and 85 kDa, as indicated Oncogene

Role of pRb in thyroid-specific gene expression S Miccadei et al

6996

Figure 2 pRb cooperates with Pax 8 in TPO gene promoter activation and it is required for E1A inhibition of Pax 8 activity. HeLa cells were transiently transfected with the TPO-Luc plasmid and were cotransfected with the vector encoding the indicated proteins. The relative luciferase activity of the cells transfected with TPO-Luc only was normalized to 1, and the other activities are expressed relative to this. Values of fold of activation and standard deviations are reported. The amount of the transfected plasmids is reported in the Meterials and methods section unless specifically indicated in the figure. (a) We show the synergy between Pax 8 and pRb in the activation of TPO gene promoter. Furthermore, we show the Pax 8/pRb synergy in SAOS-2 cells as well. Data obtained with HeLa cells are shown with black bar-graphs and the experiments with SAOS-2 cells are shown in gray. The results from three independent experiments are reported. (b) We have tested, in transfection experiments in HeLa cells, the role of pRb on Pax 8 activity on the thyroid-specific gene promoters, Tg and NIS. The plasmid containing the NIS transcriptional regulatory elements in front of the luciferase gene, termed pNISLUC9, was kindly provided by R Di Lauro, Naples (Ohno et al., 1999). The standard deviations were in the same range as those reported in the experiments shown in Figure 2a; they have been omitted because they are negligible. (c) Pax 8 activity is downregulated by E1A and such inhibition requires the E1A capability to bind pRb. E1AM is the point mutant of E1A gene at nucleotide 928 inhibiting the pRb binding (Moran, 1993). The pRb-encoding vector was tranfected at 10 ng per well and the experiment with 50 ng is indicated. The results from three independent experiments are reported

pRb binds on and is required for the activity of the thyroid-specific TPO promoter In order to further substantiate the role of pRb as Pax 8 coactivator, we show by that pRb is involved in the complex assembled on the TPO gene promoter. Hence, Oncogene

we have performed chromatin immunoprecipitation (ChIP) experiments using subconfluent FRTL5 cells. In Figure 3a we show that an anti-pRb antibody immunoprecipitates, from FRTL5 chromatin, the TPO gene promoter whereas the same antibody does not precipitate neither a GAPDH gene sequence nor the

Role of pRb in thyroid-specific gene expression S Miccadei et al

6997 Table 1 Role of pRbDcdk, p130 and p107 on Pax 8 activity Reporter vector TPO-Luc TPO-Luc TPO-Luc TPO-Luc

FA 1 1 1 1

+Pax 8 Pax Pax Pax Pax

8 8 8 8

FA

+Pocket protein

FA

+Pax8+pocket protein

FA

27.5 28.6 27.9 26.9

pRb pRbDcdk p130 p107

1.6 1.4 1.8 1.4

Pax Pax Pax Pax

59.7 55.0 46.1 38.2

8+pRb 8+pRbDcdk 8+p130 8+p107

We have performed, in HeLa cells, the same experiments shown in Figure 2a expressing the other pocket proteins p130 or p107 or pRbDcdk. The results with pRb are reported as well. FA stands for fold of activation. Similarly to the experiments shown in Figure 2, the FA of the reporter vector alone was arbitrarily set to 1. The standard deviations were in the same range as those reported in the experiments shown in Figure 2; they have been omitted because they are negligible

Table 2 pRb downregulation, in FRTL5 cells, decreases TPO promoter activity via the Pax 8-binding site Transfected vectors pGL3 Promoter+pSUPER pGL3 Promoter+pSUPER-iRbA pTPO-Luc+pSUPER pTPO-Luc+pSUPER-iRbA pTPO-Pm-Luc+pSUPER pTPO-Pm-Luc+pSUPER-iRbA

Luc gene expression 100 9878 100 3976 100 105710

The FRTL5 cells were transiently transfected with the indicated luciferase-encoding plasmids. Each plasmid was cotransfected with pSuper-iRbA or with pSuper as control. The luciferase activity measured in the control experiments was set as 100%,, and the luciferase activity in the experiments with pSuper-iRbA refers to the respective control. The means of the row luciferase/b-galactosidase data for pGL3 Promoter, pTPO-Luc and pTPO.Pm-Luc in the control experiments were 850 000/260, 20 000/260 and 3000/280, respectively. We show, as Supplementary Information, some experiments that we have performed to control the pSuper-iRbA activity, Figures 4 and 5 Figure 3 pRb is involved in the complex assembled on the TPO gene promoter. Chromatin from crosslinked subconfluent FRTL5 cells was immunoprecipitated with anti-pRb antibody. (a) GAPDH, lanes 1 and 4, or TPO promoter, lanes 2 and 5, or TSHr promoter sequences were detected, by PCR analysis, using input DNA, lanes 1, 2 and 3, or immunoprecipitated DNA, lanes 4, 5 and 6. (b) As control of the PCR conditions utilized in ChIP experiments, we show that there is linearity between the PCR products and the amount of DNA template (input DNA). Lane 1, 2 and 3 are with 0.5, 1.0 and 2.0 ml of input DNA, respectively

promoter of the thyroid-specific TSH receptor gene. We used these negative controls because the GAPDH gene expression is not activated by pRb and the TSHr gene promoter is not activated by Pax 8 (Civitareale et al., 1993; Decary et al., 2002). Figure 3b shows the linear dependence of the amount of PCR amplicons to the amount of DNA template. In order to obtain experimental evidence for the role of pRb on TPO gene transcription in the context of the thyroid follicular cells, we took advantage of the possibility to inhibit gene expression by short interfering RNA (Sharp, 1999). We utilized the pSUPER plasmid (Brummelkamp et al., 2002) to construct the pSUPERiRbA vector which directs the synthesis of small interfering RNAs containing the rat pRb gene target sequence. This plasmid was used in transient transfection experiments with FRTL-5 cells. The data in Table 2

show that pSUPER-iRbA does not affect luciferase gene expression controlled by SV40 promoter in pGL3 Promoter vector but downregulates the same gene expression when transcription occurs under the TPO gene promoter control. The same interfering plasmid does not inhibit the activity of the TPO gene promoter mutated in the Pax 8-binding site, TPO.Pm (Zannini et al., 1992). This suggests that in thyrocytes, pRb stimulates TPO gene transcription through Pax 8 activity.

Discussion In this study, we have provided experimental evidence that pRb interacts and functionally cooperates with the thyrocyte-specific transcription factor Pax 8. Hence, pRb acts as a positive transcriptional coactivator in thyroid follicular cells. This finding identifies a molecular mechanism underlying the relevant role of pRb in thyrocytes differentiation (Iuliano et al., 2000). pRb has been shown to regulate differentiation in several other cell types (Lipinski and Jacks, 1999). In osteoclasts, pRb acts as a transcriptional coactivator of CBFA1 and, in muscle cells, it cooperates with the myogenic bHLH Oncogene

Role of pRb in thyroid-specific gene expression S Miccadei et al

6998

transcription factors (Puri and Sartorelli, 2000). Furthermore, direct or indirect interaction of pRb with several transcription factors, including C/EBPs, NF-IL6, Ap1, ATF2, Sp1 and Sp3 was demonstrated to enhance their transcriptional activity (Lipinski and Jacks, 1999). The synergy of pRb with these transcription factors suggests a potential mechanism for integrating the process of cellular proliferation and differentiation. It is conceivable that a molecule such as pRb capable of inhibiting cell cycle progression is also required for the activity of the differentiation ‘master genes’. Thus, the synergy between Pax 8 and pRb could indicate a molecular mechanism resulting in the permanent withdrawal from the cell cycle and the expression of the differentiation markers in thyrocytes. pRb directly interacts with several Pax genes, it binds all the members of the Pax subfamily, Pax 2/5/8 (Eberhard and Busslinger, 1999; Yuan et al., 2002; and this study). Since it has been reported that pRb increases the DNA-binding activity of several transcription factors (Lipinski and Jacks, 1999), we have tested this hypothesis but we did not identified such a role of pRb on Pax 8 DNA-binding activity (data not shown). Therefore, how pRb can enhance Pax 8 transcriptional activity remains an open question. In this regards, it is worth mentioning that Pax 8 activity is inhibited by Id proteins (Roberts et al., 2001), and that pRb binds Id2 (Lasorella et al., 1996). Although we have not addressed this hypothesis experimentally, it is possible to imagine that pRb could displace Id2 from Pax 8 and thereby rescue the Pax 8 transcriptional activity. In analogy, it has been shown that pRb binds the inhibitory factor EID-1 rescuing the transcriptional activity of MyoD (MacLellan et al., 2000; Miyake et al., 2000). We have shown that Pax 8/pRb synergy is promoterspecific. It can be detected on both TPO and NIS gene transcription but not in the Tg gene promoter activity. These findings are in agreement with other published results (Perrone et al., 2000), indicating that Pax 8 activates its target genes using different molecular requirements. A remarkable aspect of the Pax 8/pRb synergy emerged from the experiments with E1A (Figure 2b). We have shown that the downregulation of thyrocyte differentiation, mediated by E1A, correlates with its ability to inhibit Pax 8 activity and that this inhibition relies on pRb binding. We suggest that repression by E1A results from the sequestration of pRb that is required for Pax 8-mediated activation. Such a role of pRb in thyroid-specific gene expression would explain how E1A can abrogate thyroid follicular cell differentiation with concomitant downregulation of the thyroidspecific gene expression (Berlingieri et al., 1993; Iuliano et al., 2000). The adenoviral protein can sequester the coactivator, thus hampering the activity of Pax 8. We have shown that E1A inhibits Pax 8 activity, whereas the E1A mutant, lacking the pRb interaction domain, does not. In line with these results, we have previously reported that Pax 8 and p300 cooperate in the induction of TPO gene expression and that the E1A mutant deficient in p300 binding was not able to inhibit Oncogene

Pax 8 activity. Hence, our results would indicate that Pax 8, like E1A, simultaneously interacts with both p300 and pRb. Several transcription factors, including CBFA1 (Sierra et al., 2003), MyoD, MEF2, AP2, C/ EBP, require both p300 and pRb as coactivators (Grana et al., 1998; Goodman and Smolik, 2000). These observations are compatible with the hypothesis that pRb and p300 function as molecular matchmakers, assembling different protein complexes in various cells and as crucial regulators of the differentiation process (Wang, 1997). Furthermore, it has been shown that E1A is able to simultaneously recruit both p300 and pRb into a multimeric-protein complex, which results in the stimulation of pRb acetylation by the acetylase activity of p300 (Chan et al., 2001). This post-translational modification appears to positively regulate the growthsuppressing and the differentiation-inducing capabilities of pRb, given that acetylated pRb is a poorer substrate for phosphorylation by cyclin dependent kinases. Thus, it is tempting to speculate that in thyrocytes, the Pax 8– p300–pRb complex could allow, on TPO gene promoter, both the formation of the active transcription complex and the preservation of the pRb active form, resulting in the maintenance of the differentiated state and the exit from the cell cycle.

Materials and methods Plasmid construction and cell transfection To construct pcDNA3k-Pax8, the Pax 8 gene was amplified from pCMV5-Pax 8 (Zannini et al., 1992) with the oligonucleotides bP8F (50 -cgcggatccgccaccatgcctcacaactcgatcag-30 ) and xP8R (50 -ggtttctagactacagatggtcaaaggctg-30 ), the amplicon, digested with BamH1 and Xba1, was cloned in pcDNA3 digested with the same enzymes. The plasmid pcDNA3kPax8Dad was similarly constructed using the oligonucleotide dP8ad (50 -cttatctagattatgagaggagggcctggc-30 ) instead of xP8R. The plasmid encoding Pax 8 lacking the residual homeodomain (pcDNA3k-Pax 8DHD) was constructed, according to Ho et al. (1989), using pcDNA3-mPax8 as template in the polymerase chain reaction (PCR) with the oligonucleotides bP8F and P8-HDR (50 -ccctttggtgtggctgggggaggcggtgtccgtac gaaggtgctttcg-30 ) and with the oligonuclotides P8-HDF (50 -cgaaagcaccttcgtacggacaccgcctcccccagccacaccaaaggg-30 ) and xP8R. The two amplicons were mixed and used as template in a PCR with the oligonucleotides bP8F and xP8R. The final PCR product, digested with BamH1 and Xba1, was cloned in pcDNA3 digested with the same enzymes. Thus, the resulting plasmid encodes a mutated mouse Pax 8 lacking the amino acids 229– 250. pSuper-iRbA was constructed annealing the oligonucleotides Ri-RbA-Fw (50 -gatcccccacatcatctggactctgtttcaagagaacagagtccagatgatgtgtttttggaaa-30 ) and Ri-RbA-Rv (50 -agcttttccaaaaacacatcatctggactctgttctcttgaaacagagtccagatga tgtgggg-30 ) and cloning the resulting double-stranded oligonucleotide in pSuper (Brummelkamp et al., 2002) digested with Bgl2 and Hind3. The pTg-Luc plasmid expressing the luciferase under the rat Tg gene promoter was constructed cloning in pGL3 Basic, digested with Bgl2 and Hind3, the BamH1–Hind3 fragment obtained from 50 –41 plasmid, kindly provided by R Di Lauro, Naples (Musti et al., 1987). The new constructs were confirmed by restriction enzyme digestions or by direct sequencing analysis.

Role of pRb in thyroid-specific gene expression S Miccadei et al

6999 Hela, Saos-2 and FRTL-5 cells were transfected by Lipofectamine 2000 Reagent (Invitrogen) according to the protocol suggested by the manufacturer. The following amount of the indicated plasmids were transfected: 0.4 mg, luciferase reporter plasmids; 10 ng of the Pax 8, wild type or mutant, encoding plasmids; 10 ng of the E1A, wild type or mutant, expressing vectors; 10 ng of pocket protein-expressing plasmids (pCMV-Rb, pCMV-p107 and pCMV-p130). We have tested in control experiments that the proteins, encoded by the transfected plasmids, were expressed in comparable amounts by EMSA and/or Western blot experiments, these experiments are presented as Figure 6 in Supplementary Information. In HeLa and Saos-2 cells, the efficiency of transfection was assayed with 20 ng of the pCMV-b-galactosidase plasmid, whereas in FRTL-5 cells it was tested transfecting 100 ng of pRSV-CAT (chloramphenicol acetyltransferase) plasmid. At 48 h after transfection, cell extracts were prepared using the Reporter Lysis Buffer (Promega). Luciferase, b-galactosidase and CAT assays were performed as previously reported (Civitareale et al., 1993; De Leo et al., 2000). Transfection experiments were carried out in duplicate or in triplicate and repeated at least three times. For each experiment, we report the mean of three independent experiments and the standard deviations are shown. The statistical analysis performed utilizing the row data of all transfection experiments resulted in Po0.05. For each experiment, the t-test (the probability associated with the t-test) was calculated versus the respective control group. Affinity column and band-shift assay GST and GST-pRbLP were expressed in the E. coli strain BL21. Bacteria growth, expression induction and purification of the recombinant protein was performed according to Eberhard and Busslinger, 1999. Bacterially expressed GSTpRbLP fusion protein was bound to Glutathione-Sepharose 4B resin (Amersham Biosciences). The column was equilibrated in buffer A (30 mM HEPES, 0.1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF) þ 80 mM NaCl and loaded with the nuclear extract prepared from HeLa cells overexpressing Pax 8. The nuclear extract, prepared according to Suzuki et al. (1998), was diluted with buffer A to decrease the salt concentration to 80 mM. The column was washed with buffer A þ 80 mM NaCl and eluted with buffer A þ 1 M NaCl. The flow-through, wash and high salt step elution fractions were concentrated on Centricon 30 and assayed in band-shift experiments using the double-strand oligonucleotide with the Pax 8-binding site, oligo C (Civitareale et al., 1989). The bandshift assays were performed as previously reported (Esposito et al., 1998). Protein–protein interaction assay Glutathione-Sepharose beads coated with 3 mg of GST or GST-pRbLP recombinant proteins were incubated for 2 h at 41C with 5 ml of 14C-proteins that were synthesized by a coupled in vitro transcription–translation system (TNT, Promega) in the presence of [14C]Leucine. Binding assay were performed in buffer BC100 (20 mM Tris-Cl pH 8.0, 100 mM KCl, 0.1 mM EDTA 5 mM MgCl2, 20% glycerol, 1 mM DTT, 0.5 mM PMSF) containing 0.2% NP40 and 0.5 mg/ml BSA. After washing with 400 volumes of binding buffer, bound proteins were eluted from the beads by boiling in 2  SDS (sodium dodecyl sulfate) sample buffer and were applied to SDS–PAGE (polyacrylamide gel electrophoresis). Image analysis was performed by using a Molecular Dynamics Model Storm (860) PhosphorImager. For coprecipitation of

transfected HeLa cells, 24 h after transfection, the cells were resuspended in hypotonic buffer (10 mM HEPES pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM PMSF, 0.5% NP40), the nuclei were pelletted by 30 s centrifugation at maximum speed in the microcentrifuge. Crude nuclear extracts were prepared in 20 mM HEPES pH 7.9, 350 mM NaCl, 30 mM MgCl2, 0.5% NP-40, 5 mM bglycerophosphate, 0.1 mM Na3VO4, 10 mM NaF, 1 mM DTT), supplemented with Roche protease inhibitors. After 30 min centrifugation at maximum speed in the microcentrifuge, the supernatant was diluted in the same buffer deprived of NaCl and NP40 to lower the salt concentration to 150 mM and the detergent to 0.2%. Crude HeLa nuclear extract were immunoprecititated with 2 mg of normal rabbit IgG (Santa Cruz Biotechnology) or with 2 mg affinity-purified polyclonal rabbit anti-Pax 8 antibody (kindly provided by R Di Lauro, Naples, Italy). The FRTL5 nuclear proteins were extracted according to Shapiro et al. (1998). The co-immunoprecipitation from FRTL5 nuclear proteins were performed according to Klenova et al., 2002. For Western blot analysis, the extracts or the immunoprecipitated products were separated by SDS– PAGE followed by liquid transfer to Hybond P membranes (Amersham-Pharmacia Biotech). Membranes were blocked with 5% nonfat dry milk in Tris-buffered-saline (TBS), 0.1% Tween 20 and incubated with primary antibody, anti-pRb C15 (Santa Cruz Biotechnology). Membranes were washed and incubated with horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology). Filters were then processed for enhanced chemiluminescence detection (ECL kit from Amersham-Pharmacia Biotech). Chromatin immunoprecipitation Chromatin immunoprecipitations were performed using the anti-pRb C-15 (Santa Cruz Biotechnology) according to Ferreira et al. (2001). Subconfluent FRTL-5 cells were treated with formaldehyde at a final concentration of 1% for 7 min at room temperature. Chemical crosslinking was terminated by addition of glycine to a final concentration of 0.125 M, followed by additional incubation for 5 min. After a wash with cold phosphate-buffered saline, cells were suspended in lysis buffer (5 mM piperazine N,N0 -bis(2-ethanesulfonic acid) (pH 8.0), 85 mM KCl, 0.5% NP-40) and disrupted using a Dounce homogenizer. Nuclei were then pelleted and suspended in nuclear lysis buffer (50 mM Tris-HCl (pH 8.1), 10 mM EDTA, 1% SDS. Chromatin was sonicated with 16 10-s pulses (50 W; amplitude, 80%). After centrifugation, the supernatant was diluted 10-fold with TNE buffer (16.7 mM Tris-HCl (pH 8.1), 167 mM NaCl, 0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA). Diluted chromatin was precleared with protein A/G agarose beads (Santa Cruz) saturated with bovine serum albumin and salmon sperm DNA and then incubated overnight at 41C using anti-Rb C-15 (Santa Cruz) and immunoprecipitated with protein A/G agarose beads. The beads were extensively washed, and then chromatin was eluted from beads by incubation during vortexing in elution buffer (50 mM NaHCO3, 1% SDS). Crosslinks were then reversed by overnight incubation at 651C in elution buffer containing in addition 300 mM NaCl and 30 mg of RNase A/ml. An equivalent amount of diluted chromatin was similarly processed without immunoprecipitation and noted as ‘input’ afterwards. DNA samples were then purified by phenol– chloroform extraction, ethanol precipitated, and further analysed by PCR. In each experiment, linearity of the signal was verified by amplifying increasing amount of the DNA template. For each experiment, PCRs were performed with different numbers of cycles or with dilution series of input Oncogene

Role of pRb in thyroid-specific gene expression S Miccadei et al

7000 DNA to determine the linear range of the amplification; all results shown fall within this range. The primer sequences to amplify the TPO gene promoter are: TPO2Fw (50 -gctaaca cacctagcaggaaggg-30 ) and TPO.Pr4 (50 -gtgaatctcgagtactttctg gagacttggttacccaccatataaatggactccatgc-30 ). To amplify the minimal TSHr gene promoter, we used the previously described oligonucleotides (Civitareale et al., 1993). The PCR to amplify the GAPDH gene fragment was performed with the oligonucleotides GAPDH 1F (50 -cctggccaaggtcatccatgacaa-30 ) and GAPDH 1R (50 -gcctgcttcaccaccttcttgatg-30 ). PCR conditions were: 951C for 5 s followed by 35 cycles at 941C for 40 s, 561C for 30 s and 721C for 30 s. The images from the agarose gels were acquired by using the Chemi-Doc (Bio Rad).

Acknowledgements We thank Thomas Wagner and Armando Felsani for critical reading of the manuscript. Various colleagues kindly provided many of the plasmids used in this study and we thank them for their kindness. The E1A- and E1A-Rb-encoding plasmids were a gift of B Moran (Moran, 1993). pCMV-Rb and pGSTRbLP, pCMV-107 and pCMV-130 were provided by A Felsani e M Caruso. pRbDcdk was kindly provided by WJ Harbour. We thank MR Nicotra for immunohistochemistry. MM was supported by a fellowship of Marie Curie Training site program number HPMT-CT-2000-0010. This study was in part supported by MURST-FIRB n.RBAU19HHA-001 and by grant no. 210 Life Sciences 2002 from ASI to DC.

References Adams PD. (2001). Biochim. Biophys. Acta, 1471, M123– M133. Ambesi-Impiombato FS, Parks LAM and Coon HG. (1980). Proc. Natl. Acad. Sci. USA, 77, 3455–3459. Berlingieri MT, Santoro M, Battaglia C, Greco M and Fusco A. (1993). Oncogene, 8, 249–255. Brummelkamp TR, Bernards R and Agami R. (2002). Science, 296, 550–553. Chan HM, Krstic-Demanacoes M, Smith L, Demonacos C and La Thangue NB. (2001). Nat. Cell Biol., 3, 667–674. Chellappan S, Kraus VB, Kroger B, Munger K, Howley PM, Phelps WC and Nevins JR. (1992). Proc. Natl. Acad. Sci. USA, 89, 4549–4553. Chen PL, Riley DJ, Chen Y and Lee WH. (1996a). Genes Dev., 10, 2794–2804. Chen PL, Riley DJ, Chen-Kiang S and Lee WH. (1996b). Proc. Natl. Acad. Sci. USA, 93, 465–469. Civitareale D, Castelli MP, Falasca P and Saiardi A. (1993). Mol. Endocrinol., 7, 1589–1595. Civitareale D, Lonigro R, Sinclair AJ and Di Lauro R. (1989). EMBO J., 8, 2537–2542. Decary S, Decesse JT, Ogryzko V, Reed JC, Naguibneva I, Harel-Bellan A and Cremisi CE. (2002). Mol. Cell. Biol., 22, 7877–7888. Decesse JT, Medjkane S, Datto MB and Cremisi CE. (2001). Oncogene, 20, 962–971. De Leo R, Miccadei S, Zammarchi E and Civitareale D. (2000). J. Biol. Chem., 275, 34100–34105. Di Palma T, Nitsch R, Mascia A, Nitsch L, Di Lauro R and Zannini M. (2003). J. Biol. Chem., 278, 3395–3402. Dorfler P and Busslinger M. (1996). EMBO J., 15, 1971–1982. Eberhard D and Busslinger M. (1999). Cancer Res. (Suppl.), 59, 1716–1725. Esposito C, Miccadei S, Saiardi A and Civitareale D. (1998). Biochem. J., 331, 37–40. Fabbro D, Pellizzari L, Mercuri F, Tell G and Damante G. (1998). J. Mol. Endocrinol., 21, 347–354. Ferreira R, Naguibneva I, Mathieu M, Ait-Si-Ali S, Robin P, Pritchard LL and Harel-Bellan A. (2001). EMBO Rep., 2, 794–799. Goodman RH and Smolik S. (2000). Genes Dev., 14, 1553–1577. Grana X, Garriga J and Mayol X. (1998). Oncogene, 17, 3365–3383. Harbour JW and Dean DC. (2000). Genes Dev., 14, 2393–2409. Ho SN, Hunt HD, Horton RM, Pullen JK and Pease LR. (1989). Gene, 77, 51–57. Oncogene

Iuliano R, Trapasso F, Stella A, Le Pera I, Melillo RM, Bruni P, Baldassarre G, Chiarotti L, Santoro M, Viglietto G and Fusco A. (2000). Exp. Cell. Res., 260, 257–267. Klenova E, Chernukhin I, Inoue T, Shamsuddin S and Noerton J. (2002). Methods, 26, 254–259. Lasorella A, Iavarone A and Israel MA. (1996). Mol. Cell. Biol., 16, 2570–2578. Lipinski MM and Jacks T. (1999). Oncogene, 18, 7873–7882. Lukas J, Sorensen C, Lukas C, Santoni-Rugiu E and Bartek J. (1999). Oncogene, 18, 3930–3935. Macchia P, Lapi P, Krude H, Pirro M, Missero C, Chiovato L, Souabni A, Baserga M, Tassi V, Pinchera A, Fenzi G, Gruters A, Busslinger M and Di Lauro R. (1998). Nat. Genet., 19, 83–86. MacLellan WR, Xiao G, Abdellatif M and Schneiter MD. (2000). Mol. Cell. Biol., 20, 8903–8915. Mansouri A, Chowdhury K and Gruss P. (1998). Nat. Genet., 19, 87–90. Miccadei S, De Leo R, Zammarchi E, Natali PG and Civitareale D. (2002). Mol. Endocrinol., 16, 837–846. Miyake S, Sellers WR, Safran M, Li X, Zhao W, Grossman SR, Gan J, DeCaprio J, Adams PD and Kaelin Jr WG. (2000). Mol. Cell. Biol., 20, 8889–8902. Moran E. (1993). Curr. Opin. Genet Dev., 3, 63–70. Musti AM, Ursini VM, Avvedimento E, Zimarino V and Di Lauro R. (1987). Nucleic Acid Res., 15, 8149–8166. Nevins JR. (2001). Hum. Mol. Genet., 10, 699–703. Ohno M, Zannini M, Lavy O, Carrasco N and Di Lauro R. (1999). Mol. Cell. Biol., 19, 2051–2060. Pasca di Magliano M, Di Lauro R and Zannini M. (2000). Proc. Natl. Acad. Sci. USA, 93, 465–469. Perrone L, Pasca di Magliano M, Zannini M and Di Lauro R. (2000). Biochem. Biophys. Res. Commun., 275, 203–208. Puri PL and Sartorelli V. (2000). J. Cell Physiol., 185, 155–173. Roberts EC, Deed RW, Inoue T, Norton JD and Sharrocks AD. (2001). Mol. Cell. Biol., 21, 524–553. Shapiro DJ, Sharp PA, Wahli WW and Keller MJ. (1998). DNA, 7, 47–55. Sharp PA. (1999). Genes Dev., 13, 139–141. Sierra J, Villagra A, Paredes R, Cruzat F, Gutirrez S, Javed A, Arriagada G, Olate J, Imschenetzky M, van Wijnen AJ, Lian JB, Stein GS, Stein JL and Montecino M. (2003). Mol. Cell. Biol., 23, 3339–3351. Suzuki K, Lavaroni S, Mori A, Okajima F, Kimura S, Katoh R, Kawaoi A and Kohn LD. (1998). Mol. Cell. Biol., 18, 7410–7422.

Role of pRb in thyroid-specific gene expression S Miccadei et al

7001 Thomas DM, Carty SA, Piscopo DM, Lee JS, Wang WF, Forrester WC and Hinds PWW. (2001). Cell, 8, 303–316. Wang JY. (1997). Curr. Opin. Genet. Dev., 7, 39–45. Weinberg RA. (1995). Cell, 81, 323–330. Whyte P, Buchkovich KJ, Horowitz JM, Friend SH, Raybuck M, Weinberg RA and Harlow E. (1988). Nature, 334, 124–129.

Wiggan O, Taniguchi-Sidle A and Hamel PA. (1998). Oncogene, 16, 227–236. Yuan SS, Yeh YT and Lee E. (2002). Biochem. Biophys. Res. Commun., 296, 1019–1025. Zannini M, Francis-Lang H, Plachov D and Di Lauro R. (1992). Mol. Cell. Biol., 12, 4230–4241.

Supplementary Information accompanies the paper on Oncogene website (http://www.nature.com/onc)

Oncogene