Abstract
Pit-1, a POU domain-containing transcription factor, is involved in two functions in the pituitary: PRL and GH tissue-specific expression and somato-lactotroph cells expansion. To analyze the molecular basis of the latter function, we tested whether Pit-1 can directly transactivate expression of an early marker of cell cycle initiation, the c-fos gene. We show that Pit-1 overexpression in PC12 cells, which do not express Pit-1, increases c-fos expression. Moreover, cAMP-induced c-fos promoter activity is decreased in the somato-lactotroph cell line GH3 when Pit-1 expression is reduced by hybrid arrest with an antisense sequence complementary to Pit-1 cDNA. In contrast to hormonal genes regulation, where it has been shown that any Pit-1 phosphorylation site is involved, we show that the Pit-1 phosphorylation sites are required to allow increase of c-fos promoter activity by Pit-1. We further show, by gel shift analyses, that Pit-1 is able to specifically bind the serum response element sequence present within the c-fos promoter but with a lesser affinity than the Pit-1 response element. Taken together, these results demonstrate that the tissue-specific transcription factor Pit-1 is able to enhance expression of genes involved in cell cycle initiation, suggesting that this mechanism allows Pit-1 to increase somato-lactotroph cell proliferation.
INTRODUCTION
GH and PRL are mainly pituitary hormones produced by somatotrophs and lactotrophs, respectively. The study of tissue-specific expression of GH and PRL genes has led to a characterization of the transcription factor Pit-1/GHF-1, which is required for tissue-specific expression of these genes (1–7). Pit-1 is the first member of the POU transcription factor family (8). The POU proteins contain two highly conserved domains: the first, which is highly homologous to the homeodomain (POUh), and a second, only present in the POU family members, the POU-specific domain (POUs) (9–11). The POU homeodomain is a minimal region required for sequence-specific DNA binding (12), but the POUs domain contributes also to the high-affinity binding and participates in sequence recognition (13). However, the POUh alone is able to bind consensus homeodomain sites (13). Pit-1 binds its site as a homodimer or a heterodimer with OCT-1 (14, 15), and each domain, POUh and POUs, recognize one part of the consensus sequence (9, 11, 15). It has been proposed that the POUh domain binds the sequence TAT and the POUs domain binds the sequence TACN. Within the GH promoter, two cis-acting elements, centered around −80 and− 120 and required for somatotroph-specific expression, bind Pit-1/GHF-1 (3, 16, 17). Similar observations have been obtained with the rat PRL promoter, where Pit-1 binds two elements located in the first 200 bp of the proximal promoter and one at position −1580/−1720 in the distal enhancer (1). Pit-1 itself is regulated at the transcriptional level by different mechanisms, including a positively acting autoregulatory mechanism (18). Pit-1 is a substrate for different kinases: protein kinase C, protein kinase A, and a still unknown kinase activated during the cell cycle (19–23). At the molecular level, the functional relevance of these phosphorylation events has not been determined precisely, although it appears that phosphorylation lowers DNA binding of Pit-1 (21). In addition, at the cellular level, the precise role of Pit-1 in the establishment of the somatotroph, lactotroph, and thyrotroph cell types remains also poorly understood. Several observations suggest that expression of Pit-1 could be involved in the regulation of cell proliferation. First, in the GC somatotroph cell line, microinjection of Pit-1 antisense sequences block cell growth (6). Second, physiopathological evidence for such a role comes from Pit-1-deficient dwarf mice (7). In these mice, severe growth reduction is associated not only with PRL and GH deficiency, but also with a marked failure in development of somato-lactotroph cells in the pituitary. To gain some insight into the molecular mechanisms underlying the growth-promoting effects of Pit-1, we analyzed the transcriptional modulation by Pit-1 of the immediate early gene (IEG) c-fos, a gene associated with cell cycle initiation. Indeed, c-fos expression increase is an early marker of cell cycle initiation after growth factor treatment (24, 25). The role of Fos in cell proliferation is also suggested by its involvement in the transcriptional complex that stimulates expression of genes that act directly in the cell cycle, like cyclin D1 (26–30).
RESULTS
The Tissue-Specific Transcription Factor Pit-1 Stimulates c-fos Transcription
We analyzed the effect of Pit-1 overexpression on c-fos transcription in PC12 cell lines, which have no endogenous Pit-1. As reporter gene we used a construct containing the human c-fos promoter (bp −700 to +42) fused upstream of the chloramphenicol acetyltransferase (CAT) cDNA. In a first series of experiments, this c-fos-CAT vector is cotransfected with an expression vector coding for the wild-type Rat Pit-1 (pRSV Pit-1; Fig. 1A). Overexpression of Pit-1 stimulates c-fos promoter activity more than 10-fold, as compared with control expression vector (pRSV) (Fig. 1A). Furthermore, activation by forskolin treatment of the cAMP pathway enhanced c-fos promoter activity in control cells and in cells overexpressing Pit-1. We confirm the ability of Pit-1 to increase c-fos transcription by measuring c-fos mRNA levels after Pit-1 overexpression into PC12 cells. In these experiments, the mRNA was quantified only in the transfected cells using pHook expression vector (Invitrogene, Invitrogen BV, Groningen, The Netherlands) system. This vector allows the expression of a chimeric protein containing the transmembrane part of the PDGF receptor fused to the variable region of a antibody directed against the phOx hapten (4-ethoxymethylene-2-phenyl-oxazolin-5-one). Transfected cells are selected by using phOx hapten cross-linked to magnetic beads (Capture Tech pHook-2 System, Invitrogene). Selected cells were lysed, and c-fos mRNA levels were assessed by RT-PCR. After transfection (48 h), endogenous c-fos expression increased when Pit-1 is ectopically expressed (Fig. 1B). Using the cyclophilin expression as an internal standard, we can estimate that the c-fos mRNA level is increased 2.8- fold by Pit-1 (Fig. 1B). It is interesting to note that Pit-1 stimulation of the c-fos promoter activity and the c-fos mRNA levels appears unequal. This discrepancy could be the result of the absence in the reporter gene construct of elements that could control positively the c-fos transcription and/or mRNA stability. To further analyze the physiological relevance of c-fos promoter activation by Pit-1, Pit-1 expression was inhibited in the somato-lactotroph-derived GH3 cell line, which expresses high level of Pit-1, by using an expression vector coding for an antisense Pit-1 sequence (pCMV Tip; Fig. 1C). The basal c-fos promoter activity, as detected by the c-fos-CAT reporter, was reduced in cells transfected with the Pit-1 antisense vector (Fig. 1C) by 30–50% in three independent experiments, although this inhibition could not be proven statistically different because of the variability of low CAT activity of basal c-fos promoter transcription. In contrast, inhibition of Pit-1 expression strongly reduces (70%) forskolin induction of c-fos promoter activity, demonstrating that Pit-1 is an essential element of transcriptional mechanisms involved in the cAMP-mediated regulation of c-fos promoter activity. To further validate this approach, we show that in cells cotransfected with Pit-1 antisense and pHook vectors and further selected, the expression of Pit-1 decreases after 48 h to only 30% of the original level. Under these conditions the protein level for actin is not affected, showing the specificity of the Pit-1 antisense vector.
Pit 1 Regulates c-fos Transcription. Left panel, Overexpression of the Pit-1 transcription factor stimulates c-fos transcription (A) and c-fos mRNA levels (B) in transfected PC12 cells. A, Pit-1 expression vector (pRSV Pit 1) was transfected into PC12 cells along with the reporter gene containing the first 700 bp of the c-fos promoter (−700 c-fos-CAT). A pRSVβ -globin construct was used to keep the total amount of transfected DNA to avoid unspecific promoter effects. Histograms represent means ± sd (n = 3) in fold induction from one representative experiment of three performed in triplicate. Statistical differences (Student’s t test, P ≤ 0,01) between basal level (pRSVβ -globin/pRSV Pit-1) and forskolin-stimulated (Fk) levels (pRSVβ -globin/pRSV Pit 1 in the presence of Fk 5 10−6m) are indicated by an asterisk (*). B, PC12 cells were cotransfected with pRSV Pit-1 and pHook (Invitrogen) or pRSV β-globin and pHook. After 48 h, transfected cells were selected according to manufacturer’s instruction and c-fos and cyclophilin mRNA were quantified by RT-PCR on ethidium bromide gel electrophoresis. Numbers below indicate means ± sd of three independent experiments. Values are corrected by cyclophilin mRNA levels, which served as control. Right panel, Expression of Pit-1 antisense cDNA inhibits c-fos transcription (C) and decreases Pit-1 protein level (D). C, GH3 cells were cotransfected with the −700 c-fos CAT reporter construct and a Pit-1 antisense construct (pCMV TIP-1; Pit-1 CDNA fragment used is indicated) or with the parental expression plasmid pCMV lacking the Pit-1 insert. One typical experiment of three is shown, and histograms are means ± sd (n = 3) of three transfected cultures. Asterisks indicate statistically significant differences (Student’s t test, P ≤ 0.01). Fk indicates forskolin treatment as describes in panel A. D, GH3 cells were transfected with pHook and either pCMV or antisense expression vector pCMV Tip-1. Selection was done after 24 h and 48 h as described above (Fig. 1B). Pit-1 and actin protein levels were quantified by Western blot. Relative Pit-1 protein levels are indicated. Actin protein was used as internal control.
The Pit-1 Acceptor Sites for Phosphorylation Are Required for Full c-fos Promoter Activation
Pit-1 has been shown to be phosphorylated at various acceptor sites (19, 20). These sites have been described to be not important for transcriptional activation of the GH and PRL genes by the cAMP pathway (22, 23). However, it is known that the Pit-1 phosphorylation status is regulated during the cell cycle (21), suggesting that Pit-1 phosphorylation might be required for cell cycle regulation and particularly for c-fos activation. To test whether the potential phosphorylation sites of Pit 1 contribute to the control of c-fos transcription, we compared the transactivation properties of wild-type Pit 1 and of a mutant protein lacking phosphorylation sites (see Fig. 2A). In preliminary control experiments, we show that upon transfection into PC12 cells, both expression vectors (pRSV Pit 1 and pRSV Pit-1 3A) lead to the accumulation of comparable levels of Pit 1 proteins (Fig. 2B). This result indicates that differences in transcriptional activation observed with the expression of wild-type and mutated forms of Pit-1 do not result from different levels of expressed proteins. We next confirmed, in our cellular model, i.e. PC12 cells, earlier data (see above) showing that the phospho-acceptor sites are not required for PRL transcription. Using a reporter gene containing the 250 proximal base pairs of the rat PRL promoter, we show that wild-type and mutated Pit 1 equally stimulate both basal and forskolin (Fk)-induced PRL transcription (Fig. 2C). This does not hold true for c-fos transcription, as shown in Fig. 2D. Indeed, if wild-type Pit 1 clearly enhances basal and Fk-stimulated c-fos transcription, mutated Pit-1 exerted only minor effects on basal c-fos transcription and no effect at all on Fk-induced transcription. Taken together, these results show that, in contrast to the PRL gene, regulation of c-fos transcription by Pit-1 is dependent on the three main phospho-acceptor sites of Pit-1. It remains to be established whether phosphorylation of these residues is required for transactivation or whether they are involved in a protein-protein interaction independently of their phosphorylation status. However, the fact that these three residues are phosphorylated upon activation of the cAMP pathway (19, 20) and are required for the c-fos tansactivation suggests strongly that phosphorylation of these residues is involved in the transactivating processes.
Transcriptional Stimulation of c-fos by Pit-1 Requires Pit-1 Phospho-Acceptor Sites A, Schematic representation of the Pit-1 expression vectors. pRSV Pit-1 and pRSV Pit-1 3A contain, respectively, the wild-type Pit-1 cDNA or the mutated Pit-1 3A cDNA lacking the phospho-acceptor at position S115, T220, and T219. B, Expression analysis of the wild-type and mutated Pit-1 proteins in PC12 cells. PC12 cells were transfected in 10-cm diameter dishes with 10 μg of the Pit-1 expression plasmids. Thirty hours after transfection, cells were harvested and Western blot analysis was performed with 20 μg of nuclear extract proteins. GH3 nuclear extracts (GH3) and purified Pit-1 protein (Pit-1) were used as positive controls. NT are PC12 cells transfected with the pRSVβ -globin expression vector. C, The PRL promoter does not discriminate between wild-type and Pit-1 3A mutant. Experimental procedures were as described in Fig. 1A. Cells were cotransfected with the −250 PRL-CAT reporter gene construct along with either wild-type Pit-1 (pRSV Pit-1) or the mutant Pit-1 3A (pRSV Pit-1 3A) expression vectors. Wild-type and the Pit-1 3A mutant equally stimulate basal level (Ct) and forskolin-induced (Fk; 5 × 10−6m) PRL transcription. Histograms are means ± sd (n = 3), and asterisks indicate significant differences (Student’s t test, P ≤ 0.01) when compared with respective controls (pRSV β-globin with or without Fk treatment). D, The Pit-1 phospho-acceptor sites are required for efficient c-fos transcription. Cells were cotransfected with the −700 c-fos-CAT reporter gene construct along with either wild-type (pRSV Pit-1) or the mutant (pRSV Pit-1 3A) Pit-1 expression vector. In contrast to wild-type, the Pit-1 3A only modestly stimulates basals and did not further stimulate forskolin- induced (Fk) c-fos transcription. Histograms are means ± sd (n = 3), and asterisks indicate significant differences (Student’s t test, P ≤ 0.01) when the efficiency of pRSV Pit-1 and pRSV Pit-1 3A are compared.
The Serum Response Element (SRE) of the c-fos Promoter Is the Main Target of Pit-1
To determine which sequence(s) is the Pit-1 target within the c-fos promoter, we performed a rough deletion analysis of the c-fos promoter and analyzed these deletions by transfection into PC12 cells. As shown above, the construct containing the first 700 bp of the c-fos promoter (Fig. 3A) can be stimulated by Pit-1 more than 10-fold (Fig. 3B). In contrast, the activity of a construct containing only the first 99 bp, with a cAMP response element and a retinoblastoma response element, is only induced 3-fold. Further deletion of these two elements (−53 c-fos-CAT) leads to a non-Pit-1 responsive construct. These results show that the distal part of the c-fos promoter, upstream position −99, contains the major Pit-1-responsive element. This region contains various elements, including SIS response element, FAP (Fos AP1-like element), and SRE. Moreover, analysis of the region centered on SRE shows noticeable homologies to part of the consensus Pit-1-binding site (Fig. 4A). To analyze the responsiveness of this region to Pit-1, we used reporter constructs containing either the c-fos-FAP or the c-fos-SRE sequences fused upstream to the minimal unresponsive c-fos promoter (−53 c-fos-CAT). In contrast to the FAP construct showing no induction, the SRE construct is strongly induced by Pit-1 (Fig. 3B). To further strengthen these results, chimeric reporter genes containing either the SRE or the FAP fused to the heterologous SV40 promoter were used. Overexpression of Pit-1 is, again, only able to stimulate the activity of the reporter gene containing the SRE element and not those with the FAP element or without any element upstream of the SV40 promoter (Fig. 3C). These results indicate that Pit-1 transactivates c-fos mainly through the SRE.
Delineation of a Pit-1-Responsive Element in the c-fos Promoter A, Schematic representation of the c-fos reporter genes. The construct −700 c-fos-CAT contains the c-fos promoter region from −700 bp to +42 bp upstream of the CAT cDNA. The responsive elements present within the −700 to +42 region are indicated: SRE; Fos AP1-like element (FAP); SIS response element (SISRE); retinoblastoma response element (RbRE); and cAMP response element (CRE). The −99 c-fos-CAT and −53 c-fos-CAT contain, respectively, the c-fos −99 to +42 bp and −53 to +42 bp. The constructs SRE c-fos-CAT and FAP c-fos-CAT contain, respectively, the SRE or the FAP-responsive element inserted in the− 53 c-fos-CAT construct. B, Pit-1 expression vector was transfected in PC12 cells along with the various c-fos reporter genes. Experimental procedures were as described in Fig. 1. Histograms represent means ± sd in fold induction. Asterisk indicates statistical differences between pRSVβ -globin and pRSV Pit-1 (Student’s t test, P ≤ 0.05) in fold induction as compared with their respective controls (same reporter gene but cotransfected with pRSVβ -globin plasmid). C, The SRE is sufficient to confer Pit-1 responsiveness to heterologous promoter. Left, Schematic representation of the different luciferase reporter genes used. The SRE and FAP element were cloned into the SV40-luc plasmid in front of the minimal unresponsive SV40 promoter. Right, Same experimental protocol and mode of representation is used as in panel B.
Gel Shift Analysis of Pit-1 Interactions with the SRE A, Alignment of the different sequences used in gel shift experiments. Sequences homologies between SRE-containing sequences and Pit-1 RE are boxed, and the region mutated in the SRE*/FAP oligonucleotide is underlined. B, Autoradiography of competition experiments; Pit-RE probe (104 cpm) was incubated with purified Pit-1 protein (50 ng) either in the absence (−) or in the presence of increasing concentrations of cold Pit-1 RE duplex (homologous competition, left), SRE/FAP duplex (heterologous competition, middle), and the mutated SRE*/FAP duplex (no competition, right). A 5- and 50-fold excess was used for the Pit-1 RE homologous competition and 50-, 100-, 500-, and 1000-fold excess was used for the SRE/FAP and SRE*/FAP competition. M and D indicate the two bands corresponding to binding of Pit-1 as monomer (M) or dimer (D). C, Shifts were analyzed on a phosphosimager (BAS 2000 bioimager, Fuji Photo Film Co., Ltd., Stamford, CT), and quantification was assisted by MAC BAS software (Bio Image, Fuji Film Co., Ltd.). Results are expressed as percent of Pit-1 shift (100%). Values are calculated on as many as three independent experiments where each lane is corrected with free probe input.
Pit-1 Binds to the c-fos SRE
To determine whether Pit-1 physically interacts with the c-fos SRE site highlighted by the deletion experiments, we performed gel shift analysis and competition experiments with purified Pit-1 protein and various oligonucleotide sequences (Fig. 4A). Purified Pit-1 protein interacts with the PRL Pit-1 RE by generating two high molecular weight complexes, according to previously published data (15), that are efficiently competed by a homologous cold duplex (Fig. 4B, left panel). To test whether Pit-1 protein binds specifically the SRE, the Pit-1 RE/Pit-1 complex was competed by various cold duplexes spanning the SRE and FAP sequence of the c-fos promoter (e.g. for SRE/FAP on Fig. 4B, middle panel). These sequences clearly compete for Pit-1 binding although with a lesser affinity than Pit-1 RE itself (Fig. 4C). Moreover, mutation of the SRE (SRE1/FAP oligos) eliminates the competition on Pit-1RE/Pit-1 shift (Fig. 4, B and C), thus showing the specificity of the competition obtained with the wild-type sequence. Finally, a short sequence spanning solely the SRE is sufficient to compete Pit-1 RE with almost equal efficiency than the SRE/FAP sequence (Fig. 4C). The SRE region has been described to bind several transcription factors, such as SRF, TCF, or SRE-binding protein, which interact with the SRE in many case as ternary transcriptional complexes (25, 31). A requirement of such ternary complexes including Pit-1 could explain the relatively low affinity of Pit-1 to the SRE when present alone. To analyze how Pit-1 interacts with the SRE sequence in the context of native nuclear proteins, we performed gel shift analysis with GH3 nuclear extracts. Using either SRE/FAP or SRE probes, we obtain a number of bands and, to discriminate within this pattern which one could contain Pit-1 protein, we performed competition experiments with an excess of cold Pit-1 RE (Fig. 5A). In both experiments, two bands can be competed by the Pit-1 RE sequence, suggesting the presence of Pit-1 protein within this shift. To further verify that Pit-1 present in nuclear extracts is able to interact with SRE/FAP probe, we performed supershift experiments with a Pit-1-directed antibody (Ab) (Fig. 5B). Only two major bands are seen when Pit-1 RE probe is used with nuclear extract, and incubation with Pit-1-directed antibodies allowed visualization of supershifted material (Fig. 5B, left panel). A clear supershifted band is also visible when the SRE/FAP probe is used. This supershifted band is further competed with cold Pit-1 RE, indicating also the presence of Pit-1-related material. In control experiments, antibodies unrelated to Pit-1 (e.g. anti-cAMP response element binding protein antibody) did not produce any supershift (data not shown). Taken together, all these results indicate that Pit-1 present in GH3 nuclear extracts is able to interact with the SRE sequence of the c-fos promoter.
Gel Shift Analysis of Pit-1 Interactions within GH3 Nuclear Extract A, GH3 nuclear extract (20 μg) were incubated with either SRE/FAP or SRE probe in the absence (−) or in the presence of increasing amounts of Pit-1 RE cross-competitor. Pit-1 RE duplex competed bands are indicated (*). B, GH3 nuclear extract (20 μg) was incubated with either the Pit-1 RE (left) or SRE/FAP probe (right). A supershifted band is observed with both probe (*) and is specifically competed by Pit-1 RE duplex (5-, 50-, and 500-fold excess were used).
DISCUSSION
The observation that Pit-1-deficient mice display somatotroph, lactotroph, and thyrotroph aplasia suggests that Pit-1 could be involved in the development of these three pituitary cell lineages (7). However, only a limited amount of information is available about the molecular mechanism involving Pit-1 in the establishment of the pituitary lineages. Two mechanisms have been suggested: 1) Pit-1 could be involved in the replication of DNA in pituitary cells, as suggested by the observation that Pit-1 is able to increase replication of adenovirus in vitro (32); 2) Pit-1 is required for GHRH receptor expression (33), a stimulatory input for somatolactotroph proliferation (34, 35), suggesting that the role of Pit-1 could be indirect. In this paper, we show that Pit-1 could act on proliferation of somatolactotrophs by using a third mechanism based on its ability to increase the expression of the c-fos gene, which is involved in cell cycle initiation. Our results show that overexpression of Pit-1, in a cell line that does not express Pit-1 (PC12 cells), leads to c-fos promoter activation. The physiological relevance of this effect is demonstrated by the fact that reduction of Pit-1 expression in the GH3 somatolactotroph-derived cell line induced a strong inhibition of c-fos promoter induced by cAMP. These two experiments show that Pit-1 regulates c-fos promoter and that Pit-1 may be involved in a tissue-specific regulation of c-fos expression in the somatolactotroph lineage. The Fos protein is involved in the cell cycle (36–38), and the AP1 transcription factor, a complex of Fos and Jun proteins, is also required for the proliferation of fibroblasts (39). This appears to be true for GH3 cells, since transfection of double-stranded AP1 oligonucleotides, which are expected to bind and to inactivate endogenous AP1, strongly reduces the rate of proliferation (C. Gaiddon and J.-P. Loeffler, unpublished observation). The AP1-binding sites are present in the promoter regions of various genes that are directly involved in cell cycle progression (26, 27, 29). Indeed, it has been demonstrated that Fos regulates cyclin D1 expression (28, 30). Thus, by increasing c-fos expression, Pit-1 could stimulate the cell cycle by activating expression of genes from the cyclin family. The Pit-1 protein is a substrate for different kinases, protein kinase A or protein kinase C, and was shown to be phosphorylated under various physiological conditions, such as activation of the cAMP pathway, treatment with epidermal growth factor, or during the cell cycle (19–21). However, these phosphorylation sites do not seem to play a role in hormonal gene regulation since wild-type and mutated Pit-1 stimulate transcription through the Pit-1-binding site. In contrast, the present study shows that regulation of c-fos transcription by Pit-1 requires its phospho-acceptor sites. Although we cannot formally exclude that site-specific mutation of the Pit-1 phospho-acceptor sites might also affect protein-protein intecation, these results suggest that Pit-1 phosphorylation events are necessary for the c-fos induction by Pit-1. To determine how Pit-1 regulates c-fos transcription and to localize the cis-elements required for this regulation, we performed promoter deletion analysis and gel shift assays. We showed that Pit-1 is able to enhance the transcriptional activity of the c-fos SRE and that Pit-1 is able to interact with this sequence. The interaction of Pit-1 with the SRE may require a ternary complex, as suggested by the putative presence of Pit-1 in a high molecular weight complex in GH3 nuclear extract (see Fig. 5B) and the lower affinity of purified Pit-1 to the SRE than to the consensus Pit-1 RE. This putative Pit-1-interacting protein on SRE remains to be identified. One potential partner could be SRF, since it has been shown that this protein interacts with homeodomain transcription factors (40). As the POU domain of Pit-1 contains an homeodomain, it is conceivable that this domain may mediate the interaction between Pit-1 and SRF. The effects of Pit-1 on the expression of cell cycle genes could operate at two levels. First, Pit-1 may stimulate genes that contain SRE sequences and second, the products of these genes, if coding for transcription factors such as c-fos, could then stimulate expression of a second set of genes directly involved in the control of cell cycle (e.g. cyclin D1). These mechanisms could have important physiological and physiopathological implications. The direct interaction of Pit-1 with genes that are involved into cell cycle initiation may account for the clonal expansion of the somatolactotroph cell lineage during development. Conversely, it also provides a molecular basis that links reduced proliferation of these cells and the inactivation of Pit-1 in cases of murine and human dwarfism. The same mechanism of IEG transactivation by Pit-1 may further be implicated in the pathogenesis of human pituitary tumors after somatic mutation of Gαs proteins (41). Indeed, chronic activation of the cAMP-signaling pathway in these tumors or in cellular models led to a permanent stimulation of IEGs (42). These effects may not be primarily mediated by the cAMP response element binding protein family of transcription factors as they are rapidly shut off by rapid and potent retrocontrol mechanisms involving phosphatases and various transcription factors with inhibitory effects (43, 44). Taken together with the present finding, this opens the interesting possibility that, in human secretory tumors, Pit-1, as a target of Gαs through the cAMP/protein kinase A signaling pathway, might be the control that stimulates both hormonal production and cell proliferation.
MATERIALS AND METHODS
Cell Culture and Transfections
GH3 and PC12 cells were cultured and grown as described previously (42). Transfections were carried out by employing a lipopolyamine-based method (Transfectam, Promega France, Charbonnieres, France) as described previously (45). Briefly, serum-deprived cell culture in 3.5-cm wells was transfected overnight, followed by culture in serum-free DMEM/F12 for 20 h. In some cases, cells were treated with Fk for the last 8 h. CAT activity was assayed and quantified as described previously (46). Results were corrected for transfection efficiency by using pCMV Luc, a noninducible luciferase reporter gene, and luciferase activity was quantified following manufacturer recommendations (Promega France).
Source of Recombinant Plasmids
The c-fos promoter reporter gene constructs (a gift from Dr. Roeder, New York, NY), were described previously. The pRSV Pit-1 and pRSV Pit-1 3A vectors were kindly provided by Dr. R. Maurer (Portland, OR); and PRL-CAT was provided by Dr. C. Bancroft (New York, NY). The pCMV Tip vector is derived from the pRC CMV (Invitrogen, San Diego, CA) with an insertion of a HindIII-BstXI Pit-1 fragment from the pRSV Pit-1 vector. The SV40-luc, SRE-SV40-luc, and FAP-SV40-luc were the gift of Dr. R. Prywes (New York, NY).
mRNA Extraction and RT-PCR Analysis
Total RNA was extracted as previously described by Chomczynski and Sacchi (47). Two micrograms of total RNA were used for reverse transcription reactions with 100 U of MMLV-RT (Promega Corp.) in a 30-μl reaction mix containing 20 U of RNASin (Promega Corp.) and 8 μm random hexamers. Reaction mix was incubated at 42 C for 50 min and terminated by adding 70 μl H2O and boiling for 5 min. PCR was performed with 3 μl of the reverse transcription products and 1 U of AmpliTaq DNA polymerase (Perkin Elmer Corp., Branchburg, CT) in a 50-μl reaction mix containing 15 pmol of each specific primer. Standard cycle parameters were used for 10–25 cycles followed by a final incubation at 72 C for 10 min. Ten microliters of each PCR reaction were electrophoresed on 2% ethidium bromide agarose gel. Gels were subjected to photography and densitometric analysis. Control experiments were performed to determine the range of PCR cycles over which amplification efficiency remains constant and proportional to the amount of input RNA. Oligonucleotides used for c-fos and cyclophilin amplifications are, respectively: 5′-TTTCAACGCGGACTGAGG-3′ and 5′-AGGTCATTGGGGATCTTGCA-3′; 5′-GGGGAGAAAGGATTTGGC-TA-3′ and 5′-ACATGCTTGCCATCCAGCCA-3′.
Nuclear Extracts and Gel Retardation Assay
GH3 cells were grown in 10-cm tissue culture dishes (Falcon, PolyPaba, Strasbourg, France) until 70–80% confluency. Cells were serum deprived for 24 h. Nuclear extracts were isolated by the technique of Dignam et al. (48). Final protein concentrations were typically 2–3 μg/μl as determined by Bradford assay (Bio-Rad Laboratories, Inc., Richmond CA). Equivalent amounts of protein, 1 μg of poly(dI-dC), and one of the different 18-bp 32P-labeled oligonucleotides were incubated in binding buffer (20 mm HEPES, pH 7.9, 6 mm KCl, 5 mm dithiotreitol, 5 mm spermidine, 2% Ficoll, and 8% glycerol) to a final volume of 20 μl. The binding reaction lasted 20 min at room temperature. In competition experiments, an excess of cold duplex was incubated in reaction mixture with either Pit-1 purified protein or GH3 nuclear extract for 0.5 h before probe incubation. In experiments using anti-Pit-1 antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, ref. SC442 X), the antibody is added in the reaction mixture at a final concentration of 1 μg/ml for 1 h after complex formation. The DNA-protein complexes were resolved on nondenaturating 5% polyacrylamide gels.
Western Blot Analysis
Nuclear extracts (20 μg) were loaded on a 12% SDS-PAGE. Transfer was performed at 4 C overnight onto nitrocellulose membrane (0.45 μm, Bio-Rad Laboratories, Inc.). Filters were blocked in 5% Blotto (5% nonfat dried milk, 150 mm NaCl, 50 mm Tris-HCl, pH 8.0, 0.05% Tween 20) for 5 h before incubation with 10 μg/ml of Pit-1 polyclonal antibody (Santa Cruz Biotechnology, Inc., ref. SC442 X) in 3% Blotto overnight. After three washes in TBS (150 mm NaCl, 50 mm Tris-HCl, pH 7.4, 0.05% Tween 20) a 1:500 dilution of anti-mouse Ig horseradish peroxidase-linked whole antibody (twin sheep, NA931, Amersham, Arlington Heights, IL) was added in 3% Blotto for 2 h. The blots were developed using the chemiluminescence system [ECL (Amersham) and Kodak (Eastman Kodak Co., Rochester, NY) BIOMAX MR films]).
Acknowledgments
We acknowledge the generous gifts of c-fos constructs from Dr. Roeder (New York, NY), the expression vectors from Dr. Maurer (Portland, OR) and Dr. C. Bancroft (New York, NY). The SV40 chimeric reporter genes were kindly provided by Dr. R. Prywes (New York, NY).
This work was supported by funds from Association pour la Recherche contre le Cancer (Grants 6989 and 9821).
Present address: Department of Biological Sciences, Columbia University, Fairchild Building, 1212 Amsterdam Avenue, New York, New York 10027.
