CNRS UPR 1524, Université René Descartes, Institut Cochin de Génétique Moléculaire (ICGM), CHU Cochin-Port Royal, Paris, France

^aAuthor for correspondence: CHU Cochin, 3ème étage, 24 rue du Faubourg Saint-Jacques, 75014 Paris, France

The small cell lung carcinoma (SCLC) cell line DMS-79 has been used as a model for studying the molecular mechanism underlying the ectopic ACTH syndrome. We previously showed that two domains of the human Proopiomelanocortin (POMC) gene promoter were specifically active in DMS-79 cells. The present study focuses on the more distal one, Domain IV (-376/-417). DNaseI footprinting experiments identified a single binding site for DMS-79 cell proteins in this domain. Gel-shift and sequence analysis indicated that E2F proteins might bind this site. Indeed, proteins from DMS-79 cells which bind this site (i) have in vitro DNA binding properties indistinguishable from those of E2F proteins (ii) form, like E2F proteins, multiprotein complexes which can be dissociated by sodium deoxycholate and (iii) are recognized by antibodies directed against E2F proteins. Further, we show that the rat POMC distal promoter domain contains a homologous sequence which constitutes a natural mutant of the human POMC E2F binding site, since it does not bind E2F. We show by transient transfection that this natural mutant, in the context of the rat POMC promoter, is not active in DMS-79 cells by contrast to the human POMC E2F binding site. We conclude that E2F binding is required for the activity of Domain IV in DMS-79 cells and contributes to the expression of the POMC gene in SCLC. Further studies are required to elucidate the role of E2F factors in POMC gene transcription in SCLC cells, but our results have identified mechanisms different from those in pituitary corticotroph cells that are used by these SCLC tumor cells.

Ectopic ACTH syndrome; SCLC; proopiomelanocortin; gene expression regulation; E2F; DNA binding proteins

Introduction

Ectopic hormone secretion syndromes are intriguing situations where a bioactive hormone is produced by a tumor arising outside the specialized endocrine gland. Among the peptide hormones that are most frequently ectopically expressed are ACTH, calcitonin and, to a lesser extent vasopressin (Abe et al., 1984; Baylin and Mendelsohn, 1980; Gould et al., 1984). The ectopic ACTH syndrome is most frequently associated with neuroendocrine tumors of the lung such as carcinoids or small cell lung carcinoma (SCLC) and occurs in 9 - 18% of patients with Cushing's syndrome (Becker and Gazdar, 1984; Orth, 1995; Wajchenberg et al., 1994). The rapidly growing SCLCs usually display limited neuroendocrine properties, and those responsible for ectopic ACTH secretion contain only low to very low levels of mRNA coding for proopiomelanocortin (POMC), the precursor to ACTH (de Keyzer et al., 1985, 1996). We investigated the mechanisms of POMC gene transcription in non-pituitary tumors, using the SCLC cell line DMS-79 (Bertagna et al., 1978; Sorenson et al., 1981) as an in vitro model for the study of the ectopic ACTH syndrome.

The activity of the POMC gene promoter in the pituitary context has been extensively studied both in the murine corticotroph cell line AtT-20 and in transgenic mice models (Hammer et al., 1990; Jeannotte et al., 1987; Tremblay et al., 1988). Multiple cis-regulatory elements have been identified whose coordinate activities were required for POMC gene transcription (Therrien and Drouin, 1991). Two trans-acting factors exert important synergistic actions: CUTE (Corticotroph Upstream Transcription Element-binding protein), which belongs to the basic helix - loop - helix family and Ptx1 (Pituitary homeo box 1) of the homeodomain family (Lamonerie et al., 1996; Poulin et al., 1997; Therrien and Drouin, 1993).

Our previous functional analysis of the human POMC gene promoter (Picon et al., 1995) showed that the same (+21/-417) region conferred full activity both in DMS-79 and in AtT-20 cells. However, two main differences were noted between the two cell lines: (i) the synergy between CUTE and Ptx1 was not observed in DMS-79 cells, and (ii) the sequence -376/-417, designated Domain IV, was only active in DMS-79 cells. These results demonstrated that POMC gene transcription in DMS-79 cells was achieved through mechanisms distinct from those in the pituitary.

We have now investigated which trans-acting factor(s) interact with Domain IV sequence and may contribute to the aberrant activity of the POMC gene promoter in DMS-79 cells. We show that Domain IV can bind factors of the E2F family of transacting factors, which are known for their role in the control of cell proliferation and differentiation (Lam and La Thangue, 1994). In addition, the presence of a binding site for E2F is required for the activity of Domain IV in DMS-79 cells.

Results

Identification of the IVA binding site

Among the four domains previously defined in the human POMC gene promoter (Picon et al., 1995), Domain IV (-376/-417) had the distinctive property of being active in DMS-79 cells but not in AtT-20 cells. We explored the regions of this domain interacting with DMS-79 cell proteins by in vitro DNaseI footprinting experiments (Figure 1a). With DNA fragment (-227/-417) on either the upper (Up) or lower (Low) strand, DMS-79 cell protein extracts (D) modified the DNaseI digestion pattern when compared to the free probe (f). Two regions, IIIB and IVA, indicated by vertical lines, were detected on both strands and had the typical features of protein footprints: they were totally protected from DNaseI digestion and delineated by hypersensitive sites. Sites IIIB and IVA extended from -312 to -340 and from -371 to -403, respectively, in the human POMC gene promoter sequence (Figure 1b). Thus a single large footprint was present in Domain IV, extending slightly downstream of the -376 border.

The IVA site binds several proteins

We performed gel-shift assays using IVA as a probe to characterize DMS-79 cell proteins that bound to the protected sequence. Three specific complexes were detected (Figure 2a, horizontal arrows) that could be competed for by an excess of IVA oligonucleotide but not by a similar excess of an unrelated sequence. A fourth complex (ns) was observed in all experiments but appeared to be nonspecific, since it was either unaffected or barely affected by any competitor or treatment.

To determine the core recognition sequence in site IVA, we used a series of block-mutated oligonucleotides in gel-shift assays (Figure 2b). Several complexes formed with MUT1 and MUT2 probes (Figure 2c), but none was competed by an excess of IVA oligonucleotide. The absence of a IVA specific complex with these probes indicated that the block-mutation had disrupted their IVA binding site. By contrast, the MUT3 probe generated complexes with the same retardation pattern as the IVA probe. These complexes were competed for similarly by an excess of MUT3 and IVA oligonucleotides. Therefore, the region disrupted in the MUT3 probe was not important for binding of IVA specific factors. In agreement with these data, competitive binding assays showed that oligonucleotides MUT1 and MUT2 did not compete for IVA complex formation, in contrast to MUT3. The observation that all IVA specific complexes were affected equally by the various mutants IVA oligonucleotides suggested that their formation was dependent on the same sequence motif and, therefore, that they all contained an identical protein important for specific DNA binding or, closely related proteins. The ineffectiveness of MUT1 and MUT2 indicated that the IVA binding site motif is CCGCCAAAT, itself contained in a larger symmetrical sequence: ATATTTACCGCCAAAT. Published work on the rat POMC gene promoter (Philips et al., 1997) and database search suggested that this sequence might constitute a binding site for two distinct protein families. The IVA sequence containing the symmetrical site is highly homologous (20 of 22 bp) to the rat POMC promoter element DE2A/B, which has been shown to bind in vitro translated Nur77 proteins and to represent a new binding site for this family, called the NurRE (Philips et al., 1997) (Table 1). The IVA binding site also has sequence homologies with binding sites for the E2F factors, since it is composed of a GC-rich core sequence flanked on both sides by AT-rich stretches.

IVA complexes have DNA binding properties similar to that of E2F

We tested in competitive binding studies various oligonucleotides known to bind Nur77 or E2F factors (Figure 3). Several Nur binding oligonucleotides (NBRE, COUP (Wilson et al., 1991) and hP-NBRE (Okabe et al., 1998)) had no effect on IVA complex formation. In contrast, the E2F oligonucleotide (Helin et al., 1992) was as efficient as the IVA sequence itself, whereas E2FMUT, an E2F mutant oligonucleotide unable to bind E2F (Helin et al., 1992), had no effect on IVA proteins binding. In addition, the E2F probe formed specific complexes with DMS-79 cell proteins that had the same apparent mobility as IVA complexes and were abolished by an excess of E2F or IVA oligonucleotides, but not by an unrelated sequence. Thus, proteins present in DMS-79 cells were able to bind the E2F or the IVA probes in a similar fashion in vitro.

To compare further IVA complexes to E2F complexes, we used site-specific methylation of the E2F oligonucleotide. Yee et al. (1987) demonstrated that methylation of cytosines within the E2F binding site with HhaI methylase inhibited E2F binding. Since the HhaI methylase recognition sequence is not present in the IVA sequence, we used mE2F, an E2F oligonucleotide with a HhaI-like methylated sequence, in competitive binding studies. In contrast to E2F oligonucleotide, mE2F oligonucleotide did not inhibit formation of IVA complexes (Figure 4), suggesting that, like E2F-binding proteins, the IVA-binding proteins had no affinity for this methylated E2F sequence.

E2F proteins can bind to DNA not only as dimers but also as complexes of higher order that can be dissociated in the presence of DOC (Wong et al., 1995). Treatment of DMS-79 cell extracts with DOC abolished the slower migrating complexes (Figure 4), indicating that these IVA protein-DNA complexes also contained higher order complexes.

In vitro translated DP-1 and E2F-1 proteins bind site IVA

To demonstrate that the IVA site is actually an E2F binding site, we performed gel-shift assays with in vitro translated chE2F-1 together with chDP-1, one of the usual partners of the E2F family of trans-acting factors (Loiseau et al., 1997), and the IVA probe. In vitro translated chDP-1 and chE2F-1 interacted with the IVA probe and formed a specific complex (Figure 5, horizontal arrow) that was efficiently competed for by an excess of unlabeled IVA or unlabeled E2F oligonucleotide.

E2F-4 and DP-1 are part of the IVA complexes

Antibodies against E2F-4 and DP-1 were used in supershift gel-shift assays to establish their presence in IVA complexes. Addition of the anti E2F-4 antibodies to the E2F probe gel-shift assay decreased the amount of the upper complex simultaneously with appearance of a supershifted complex (Figure 6, star). A supershifted complex of slower mobility (Figure 6, vertical bar) was also formed with the anti-DP-1 antibody and the intensity of all IVA complexes was decreased. Similar effects were observed on complexes formed with the IVA probe, demonstrating that the upper IVA complex, at least, contained E2F-4 proteins and that DP-1 proteins were present in all IVA complexes. Altogether, these results suggested that other E2F proteins (E2F-1, E2F-2, E2F-3 or E2F-5) were present in all IVA specific complexes.

The rat POMC promoter is not a target for E2F

As previously indicated, the IVA sequence is highly conserved in the rat POMC gene sequence. However, the rat POMC gene promoter IVA (rIVA) sequence constitutes a natural mutant, since it exhibits a transversion (GT) in the center of the E2F binding site. A 200-fold excess of rIVA oligonucleotide had no effect on the formation of E2F complexes, like an unrelated sequence, whereas a similar excess of either E2F or IVA oligonucleotide completely abolished it (Figure 7a). Therefore, the rIVA sequence, although highly homologous to the IVA sequence, was unable to bind E2F proteins.

The rIVA sequence was functionally analysed in DMS-79 cells by transient transfection. The alignment of rat and human sequences analysed in this assay is shown in Figure 7b. The rat POMC minimal promoter (construct 1) was eightfold less active than the sequence extending up to -323 (construct 2, Figure 7c). Addition of the distal rat POMC gene promoter domain (up to -480) did not increase activity in DMS-79 cells (construct 3). In contrast, addition of Domain IV (construct 6) to human POMC promoter Domains I - III (construct 5), resulted in a ca. 1.3-fold increase as previously described (Picon et al., 1995). Thus, in the context of the rat POMC gene promoter, a sequence highly homologous to the IVA sequence that does not bind E2F proteins has no functional activity in DMS-79 cells, suggesting that the activity of Domain IV in DMS-79 cells is dependent on E2F proteins binding.

Discussion

The DMS-79 human SCLC cell line was taken here as a model for the ectopic expression of the POMC gene, that is occasionally observed in highly proliferative bronchial tumors such as SCLCs, and is responsible for the ectopic ACTH syndrome. Indeed, DMS-79 cells have many of the abnormal features of POMC gene expression detected in such tumors: (i) very low level of POMC mRNA; (ii) greatly reduced or altered processing of the precursor protein (Bertagna et al., 1978) and (iii) constitutive secretion of POMC peptides that does not respond to either glucocorticoid inhibition (Gaitan et al., 1995) or activation of the cAMP pathway (A Picon, unpublished data).

Our previous studies demonstrated that POMC gene transcription in DMS-79 cells was achieved through a mechanism distinct from that occurring in the pituitary since the synergy between CUTE and Ptx1 did not occur in DMS-79 cells (Picon et al., 1995). Two domains of the human POMC gene promoter were active in DMS-79 cells: Domain II (-95/-161) and Domain IV (-376/-417). The main contribution to POMC gene transcription was obtained with Domain II, but all the factors that bound to it in DMS-79 cells were also found in AtT-20 cells (A Picon, unpublished data). We therefore focused on Domain IV where only one footprint was observed. Gel-shift assays with the IVA oligonucleotide and nested oligonucleotides (-357/-382), (-367/-394), (-384/-412), (-396/-422), spanning the whole domain, identified the same IVA site as a single binding site for DMS-79 cell proteins. ATATTTACCGCCAAAT was defined as the binding motif which potentially binds two distinct protein families. The Nurr1/Nur77 subgroup which belongs to the nuclear receptor superfamily contains at least three members. These factors were shown to participate in the activity of the POMC gene promoter (Murphy and Conneely, 1997; Okabe et al., 1998; Philips et al., 1997). Particularly, Philips et al. (1997) described in the distal part of the rat POMC gene promoter an element, called the Nur Response Element (NurRE), that is bound by dimers of in vitro translated Nur77 proteins and mediates transactivation by Nur77. The distinct Nur proteins are believed to bind as monomers or dimers to the NurRE, so several specific complexes with the same binding specificity could be expected, consistent with the gel-shift pattern observed with the IVA probe.

The second candidate is the E2F family, which plays an important role in growth regulatory processes by controlling numerous cell cycle regulated promoters (for a review see Slansky and Farnham, 1996). E2F binding units are heterodimers composed of an E2F subunit and a DP subunit. There are at least five distinct E2F proteins and three DP proteins, and E2F dimers are also known to bind DNA as multiprotein complexes (Slansky and Farnham, 1996). Thus, the interactions between E2F factors and the IVA site could also result in a multiple band gel-shift pattern like that obtained with DMS-79 cell proteins.

Competitive binding studies showed that Nur proteins were not part of the IVA complexes. Gel-shift studies with antibodies to Nur proteins, the high-affinity NBRE binding site (Wilson et al., 1991) and DMS-79 cell proteins showed there was little or no Nurr1/Nur77 proteins in DMS-79 cell extracts, although the same experiments detected these proteins in both mouse AtT-20 and human HeLa cell lines (A Picon, unpublished data). In contrast, we demonstrated not only that in vitro translated E2F proteins bound to the IVA sequence, but also that E2F proteins were part of each DMS-79 cell IVA-specific complex on the basis of binding, biochemical (DOC) and immunological (supershift) evidences.

Transactivation via E2F binding sites has been demonstrated in the promoters of N-myc, c-myc, cdc-2 and IGF-1 genes (for a review see Slansky and Farnham, 1996). Thus, it was possible that E2F binding between -381 and -394 also induced the activity of the POMC gene promoter in SCLC cells. Despite high sequence homologies between the distal rat POMC gene promoter domain and the IVA site, the rat POMC gene sequence did not bind E2F proteins, thus constituting a natural mutant of the human POMC E2F binding site. In contrast to the human IVA site, addition of rIVA sequence (-323/-480) to the remaining rat POMC promoter did not increase activity, suggesting that an efficient E2F binding site is required for the activity of distal POMC gene promoter sequences in DMS-79 cells. Therefore, it is tempting to speculate that the sequence difference between the human and rat IVA binding sites renders the human POMC gene promoter sensitive to E2F proteins in tumor cells that can form active E2F complexes. However, additional studies are needed to confirm the lack of activity of Domain IV mutants in which the E2F binding site is inactivated.

E2F is part of a pathway that is involved in the control of cell proliferation and differentiation and is regulated at different levels by cyclins and their dependent kinases. E2F is the direct target of retinoblastoma protein (pRB) and related proteins, which block the transactivating activity of E2F factor by interacting with it, but without affecting its DNA binding ability (for a review see Lam and La Thangue, 1994). Inactivation of the Rb gene, believed to be a key step in tumorigenesis, is particularly frequent in SCLCs; 80 - 90% of these tumors do not express a functional pRB (Harbour et al., 1988; Helin et al., 1997). Rygaard et al. (1990) established by RNA and protein blotting analysis that Rb gene products are undetectable in DMS-79 cells. Disruption of regulation of the E2F pathway might explain the aberrant activity of Domain IV in DMS-79 cells. This domain accounts for 25 - 30% of POMC promoter activity in transfected DMS-79 cells and E2F is clearly not the sole determinant of POMC gene transcription in these cells. It is possible that general neuroendocrine factors, such as those binding in Domain II, provide a basal activity, further enhanced in those tumoral cells with inactivated pRB and active E2F. Whether or not E2F factor is a key component of POMC gene expression in these tumor cells, the presence of activated E2F complexes satisfies at least one condition for aberrant POMC gene expression. Because of the additional frequent lack of pRB control in SCLCs, one might expect frequent expression of the POMC gene in this tumor type. In fact since ectopic ACTH production associated with SCLCs may cause only mild hypokalemia and occurs in patients with rapidly progressive cancer, it is estimated to represent the most underdiagnosed form of Cushing's syndrome, accounting for as much as three-quarters of all cases of ectopic ACTH secretion (Orth, 1995). It will be interesting to determine if loss of pRB function is associated with POMC gene expression in these tumors.

Materials and methods

Plasmids and oligonucleotides

Plasmids 417hPOLU, 376hPOLU and 39hPOLU were previously described (Picon et al., 1995). Oligonucleotides used for gel-shift analysis are shown in Figure 2b or have the following sequences: GGGGTAGGAACCAATGAAATGAAAGGTTA (NFY from the rat albumin gene promoter, (Raymondjean et al., 1988)), CGGGAAGGTCAAAGTCC-CGCGCCCA (hP-NBRE, human POMC sequence from -47 to -71), GAGTTTTAAAAGGTCATGCTCAATTT (NBRE consensus, (Wilson et al., 1991)), GATCAAAGTCAGGTCACAGTCACCTGATCAAAGTT (ERE, (Hedvat and Irving, 1995)), TATGGTGTCAAAGGTCAAACTTCT (COUP from the chicken ovalbumin upstream promoter, (Therrien and Drouin, 1991)), ATTTAAGTTTCGCGCCCTTTCTCAA (E2F consensus, (Helin et al., 1992)) and ATTTAAGTTTCGATCCCTTTCTCAA (E2F MUT, a mutant of the E2F consensus sequence, (Helin et al., 1992)). For interference by methylation, a double-stranded E2F oligonucleotide containing a Hhal methylation pattern (mE2F) was constructed by annealing two oligonucleotides, each containing one 5-methyl-2'-deoxycytidine (5-Me-dC) with the following sequences: ATTTAAGTTTCG(5-Me-dC)GCCCT-TTCTCAA and TTGAGAAAGGG(5-Me-dC)GCGAAA-CTTAAAT.

Cells and whole-cell protein extracts

DMS-79 cells were grown as previously described (Picon et al., 1995). Whole-cell microextracts were prepared essentially as described (Helin et al., 1993), with slight modifications. 2´10⁷ cells were collected by centrifugation with PBS. Cells were washed twice in PBS and resuspended at 4°C in 20 mM HEPES (pH 7.6), 100 mM NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 25% glycerol, 0.5 mM DTT, 5 mg/ml aprotinin, 0.5 mg/ml leupeptin, 5 mg/ml pepstatin, 0.5 mM PMSF. This suspension was adjusted to 300 mM NaCl, submitted to three freeze/thaw cycles and incubated on ice for 20 min. Cellular protein extracts were obtained after centrifugation for 15 min at 12 000 g and stored at -80°C.

Whole-cell extracts were prepared from 1.5 - 2´10⁸ cells by the same method, except that proteins were precipitated by addition of (NH₄)₂SO₄ to 45% saturation. Proteins were then suspended in dialysis buffer containing 20 mM HEPES (pH 8.0), 0.2 mM EDTA, 0.2 mM EGTA, 25% glycerol, 60 mM KCl, 0.5 mM DTT, 5 mg/ml aprotinin, 0.5 mg/ml leupeptin, 5 mg/ml pepstatin, 0.5 mM PMSF and (NH₄)₂SO₄ was eliminated by dialysis twice for 2 h against 250 volumes. After centrifugation 5 min at 12 000 g, the supernatant containing the proteins was stored at -80°C. Protein concentrations were determined with the protein bioassay kit (Bio-Rad Laboratories, Richmond, CA, USA).

DNaseI footprinting assays

The plasmid carrying the human POMC gene promoter region (+121/-417) was linearized with the appropriate restriction enzyme and labeled with DNA polymerase I large fragment and ³²P-dNTP (Sambrook et al., 1989). The probe was released by a second restriction enzyme and purified on a 5% polyacrylamide gel. 2´10⁴ c.p.m. of probe were incubated with 60 g of whole-cell extract in 16 l of 1 mM HEPES (pH 8.0), 0.1 mM EDTA, 50 mM NaCl, 50 mM KCl, 5 mM MgCl_2,, 18% glycerol, 1 mM DTT and 125 ng poly (dI-dC). The reaction mixtures were preincubated for 15 min on ice, then adjusted to 2.5 mM CaCl₂ and further incubated for 1 min at 25°C before addition of 1 l of DNaseI dilutions selected to produce an even pattern of digestion products. After 1 min at 25°C the reactions were stopped with 10 mM EDTA and 0.1% SDS. Proteins were digested for 30 min at 42°C with proteinase K (150 g/ml), and DNaseI digestion products were separated on 8% polyacrylamide/7 M urea sequencing gels.

In vitro transcription and translation

0.5 g of chicken DP-1 (chDP-1) or chicken E2F-1 (chE2F-1) plasmids (Loiseau et al., 1997) were simultaneously transcribed and translated with [³⁵S]methionine using the TNT kit (PROMEGA) according to the supplier recommendations. The efficiency of the reactions was checked on 8% SDS - PAGE gels.

Gel-shift assays

Double stranded oligonucleotides were 5' end-labeled with T4 polynucleotide kinase and ³²P-ATP (Sambrook et al., 1989) to a specific activity of ca. 10⁸ c.p.m./g. Five fmoles of probes were incubated with either 5 g of whole-cell microextracts or 1 l of in vitro translation reaction mixture for 10 min at 4°C in 20 l of 10 mM HEPES (pH 8.0), 0.1 mM EDTA, 50 mM NaCl, 50 mM KCl, 5 mM MgCl₂ 18% glycerol, 100 g/ml bovine serum albumin and 1 g poly (dA-dT). For competitive binding experiments, the probe and competitor were mixed prior to the addition of proteins. For supershift experiments, antibodies against DP-1 and E2F-4 (Santa Cruz Biotechnology, ref. K-20 and C-108, respectively) were preincubated with the extracts for 1 h on ice prior to probe addition. The samples were separated on 5% polyacrylamide gels buffered with 0.5´TBE.

Sodium deoxycholate treatment

Whole-cell microextracts were treated for 20 min with 0.8% sodium deoxycholate (DOC) and then for 10 min with NP-40 1.5% before addition of the DNA-binding reaction mixture as previously described (Wong et al., 1995).

Transfection and enzymatic assays

Electroporation, cell extracts, luciferase and CAT assays were performed as previously described (Picon et al., 1995).

Acknowledgements

We are indebted to Dr GD Sorenson and Dr O Pettengill for having provided us with the DMS-79 cell line. We wish to thank Drs DN Orth for his valuable comments on the manuscript and M Raymondjean for helpful discussions and comments. We thank Dr J Drouin for rat POMC plasmids and Drs G Brun and L Loiseau for chDP-1 and chE2F-1 plasmids. This work was supported in part by INSERM CJF92-08 and the Association pour la Recherche Contre le Cancer, (contract 6501 to YdK). A Picon is the recipient of a doctoral fellowship of the Association pour la Recherche Contre le Cancer.

Abe K, Kameya T, Yamaguchi K, Kikuchi K, Adachi I, Tanaka M, Kimura S, Kodama T, Shimosato Y and Ishikawa S. (1984). In: The endocrine lung in health and disease. WB Saunders Company: New York, pp.549-595.

Baylin SB and Mendelsohn G. (1980). Endocr. Rev. 1, 45-77. MEDLINE

Becker KL and Gazdar AF. (1984). In: The endocrine lung in health and disease. WB Saunders Company.,

Bertagna X, Nicholson WE, Sorenson GD, Pettengill OS, Mount CD and Orth DN. (1978). Proc. Natl. Acad. Sci. USA 75, 5160-5164. MEDLINE

de Keyzer Y, Bertagna X, Lenne F, Girard F, Luton J-P and Kahn A. (1985). J. Clin. Invest. 76, 1892-1898. MEDLINE

de Keyzer Y, Vieau D, Picon A and Bertagna X. (1996). Endocr. Rel. Cancer 3, 99-112.

Gaitan D, DeBold CR, Turney MK, Zhou P, Orth DN and Kovacs WJ. (1995). Mol. Endocrinol. 9, 1193-1201. MEDLINE

Gould VE, Warren WH and Memoli VA. (1984). In: The endocrine lung in health and disease. Becker KL and Gazdar AF (eds). WB Saunders Company: New York, pp.406-445.

Hammer GD, Fairchild-Huntress V and Low MJ. (1990). Mol. Endocrinol. 4, 1689-1697. MEDLINE

Harbour JW, Lai S-L, Whang-Peng J, Gazdar AF, Minna JD and Kaye FJ. (1988). Science 241, 353-357. MEDLINE

Hedvat CV and Irving SG. (1995). Mol. Endocrinol. 9, 1692-1700. MEDLINE

Helin K, Lees JA, Vidal M, Dyson N, Harlow E and Fattaey A. (1992). Cell 70, 337-350. MEDLINE

Helin K, Wu C-L, Fattaey AR, Lees JA, Dynlacht BD, Ngwu C and Harlow E. (1993). Genes Dev. 7, 1850-1861. MEDLINE

Helin K, Holm K, Niebuhr A, Eiberg H, Tommerup N, Hougaard S, Skovgaard Poulsen H, Spang-Thomsen M and Norgaard P. (1997). Proc. Natl. Acad. Sci. USA 94, 6933-6938. MEDLINE

Jeannotte L, Trifiro MA, Plante RK, Chamberland M and Drouin J. (1987). Mol. Cell. Biol. 7, 4058-4064. MEDLINE

Lam EW-F and La Thangue NB. (1994). Curr. Opin. Cell. Biol. 6, 859-866. MEDLINE

Lamonerie T, Tremblay JJ, Lanctôt C, Therrien M, Gauthier Y and Drouin J. (1996). Genes Dev. 10, 1284-1295. MEDLINE

Loiseau L, Pasteau S and Brun G. (1997). Gene Expr. 6, 259-273. MEDLINE

Maxam AM and Gilbert W. (1980). Methods Enzymol. 65, 499-560. MEDLINE

Murphy EP and Conneely OM. (1997). Mol. Endocrinol. 11, 39-47. MEDLINE

Okabe T, Takayanagi R, Adachi M, Imasaki K and Nawata H. (1998). J. Endocrinol. 156, 169-175. MEDLINE

Orth DN. (1995). N. Engl. J. Med. 332, 791-803. MEDLINE

Philips A, Lesage S, Gingras R, Maira M-H, Gauthier Y, Hugo P and Drouin J. (1997). Mol. Cell. Biol. 17, 5946-5951. MEDLINE

Picon A, Leblond-Francillard M, Raffin-Sanson M-L, Lenne F, Bertagna X and de Keyzer Y. (1995). J. Mol. Endo. 15, 187-194.

Poulin G, Turgeon B and Drouin J. (1997). Mol. Cell. Biol. 17, 6673-6682. MEDLINE

Raymondjean M, Cereghini S and Yaniv M. (1988). Proc. Natl. Acad. Sci. USA 85, 757-761. MEDLINE

Rygaard K, Sorenson GD, Pettengill OS, Cate CC and Spang-Thomsen M. (1990). Cancer Res. 50, 5312-5317. MEDLINE

Sambrook J, Fritsch EF and Maniatis T. (1989). In: Molecular cloning: A laboratory manual. Cold Spring Harbor: Cold Spring Harbor Laboratory Press.,

Slansky JE and Farnham PJ. (1996). Curr. Top. Microbiol. Immunol. 208, 1-30. MEDLINE

Sorenson GD, Pettengill OS, Brinck-Johnsen T, Cate CC and Maurer LH. (1981). Cancer 47, 1289-1296. MEDLINE

Therrien M and Drouin J. (1991). Mol. Cell. Biol. 11, 3492-3503. MEDLINE

Therrien M and Drouin J. (1993). Mol. Cell. Biol. 13, 2342-2353. MEDLINE

Tremblay Y, Tretjakoff I, Peterson A, Antakly T, Zhang CX and Drouin J. (1988). Proc. Natl. Acad. Sci. USA 85, 8890-8894. MEDLINE

Wajchenberg BL, Mendonca BB, Liberman B, Albergaria-Pereira MA, Carneiro PC, Wakamatsu A and Kirschner MA. (1994). Endocr. Rev. 15, 752-787. MEDLINE

Wilson TE, Fahrner TJ, Johnston M and Milbrandt J. (1991). Science 252, 1296-1300. MEDLINE

Wong K-K, Zou X, Merrell KT, Patel AJ, Marcu KB, Chellapan S and Calame K. (1995). Mol. Cell. Biol. 15, 6535-6544. MEDLINE

Yee AS, Reichel R, Kovesdi I and Nevins JR. (1987). EMBO J. 6, 2061-2068. MEDLINE

Figure 1 Binding sites for DMS-79 cell proteins within region (-227/-417) of the human POMC gene promoter by DNaseI footprinting. (a) DNaseI footprinting analysis. Probes extending from -227 to -417 on the upper strand (Up) and lower strand (Low) were used with DMS-79 cells extracts (D). Positions of the footprints are indicated by vertical bars and labels. M: A+G chemical sequencing ladder (Maxam and Gilbert, 1980); f: free probe. (b) DMS-79 cell DNaseI footprints identified in (a) are shown as lines below the DNA sequence and labeled according to their positions within the promoter (domains III and IV). The numbers indicate the position relative to the cap site

Figure 2 Gel-shift analysis of sequence IVA. (a) Gel-shift assays performed with DMS-79 cell extracts and the IVA probe. A 200-fold molar excess of unlabeled oligonucleotide (competitor) was added when specified. (b) Alignment of IVA and IVA mutant oligonucleotide sequences. Bold lowercase indicates that a mutation has been introduced in the sequence. (c) Gel-shift assays performed with DMS-79 cell extracts and the probe indicated at the top. A 200-fold molar excess of unlabeled oligonucleotide (competitor) was added when specified. Arrow: specific complex; ns: nonspecific complex; f: free probe

Figure 3 Competitive gel-shift analysis of sequence IVA. Gel-shift assays were performed with DMS-79 cell extracts and the IVA or E2F probe, as indicated at the top. A 200-fold molar excess of unlabeled competitor oligonucleotide was added when specified. Arrow: specific complex; ns: nonspecific complex; f: free probe

Figure 4 Effect of sequence methylation and of DOC treatment on the formation of IVA complexes. Gel-shift assays were performed with DMS-79 cell extracts and probe IVA. A 100-fold molar excess of unlabeled competitor oligonucleotide was added when specified. DMS-79 cells extracts were treated with DOC when indicated. Arrow: specific complex; ns: nonspecific complex; f: free probe

Figure 5 Gel-shift analysis of in vitro translated chDP-1 and chE2F-1 binding to the IVA sequence. Gel-shift assays were performed with in vitro translated chDP-1 and chE2F-1 and probe IVA. A 200-fold molar excess of unlabeled competitor oligonucleotide was added when specified. Arrow: specific complex; ns: nonspecific complex; f: free probe

Figure 6 Supershift analysis of IVA complexes. Gel-shift assays were performed with DMS-79 cell extracts and the probe indicated at the top. Antibodies against E2F-4 or DP-1 were added when indicated. Arrow: specific complex; ns: nonspecific complex; f: free probe; star: supershifted complex caused by anti-E2F-4 antibodies; vertical bar: supershifted complexes caused by anti-DP-1 antibodies

Figure 7 Comparative analysis of the IVA sequence and its related sequence in rat POMC gene promoter. (a) Gel-shift assays performed with DMS-79 cells extracts and probe E2F. A 200-fold molar excess of unlabeled competitor oligonucleotide was added when specified. Arrow: specific complex; ns: nonspecific complex; f: free probe. (b) Sequence alignment of rat (italic) and human distal POMC gene promoter domains. Sequence of the E2F binding site and its related sequence in the rat promoter are underlined. Bold lowercase letters point to the nucleotide mismatch. Numbers indicate the position relative to the cap site. The dashes indicate sequence homologies. (c) Functional 5'-deletion analysis of rat and human POMC gene promoters in DMS-79 cells. 5'-deletion mutants of the rat POMC gene promoter (Therrien and Drouin, 1991) were tested in constructs 1 - 3 whereas 5'-deletion mutants of the human POMC gene promoter were tested in constructs 4 - 6. Part of the exon and promoter domains are depicted as hatched and open boxes, respectively. The arrow indicates the position of the cap site and the numbers indicate the position relative to the cap site. The activity is plotted relative to that of the minimal construct (constructs 1 or 4). The data represent the means±s.e.m. (n=6). The luciferase activity is normalized with that of an internal control plasmid, RSV - CAT

Table 1 Table 1