- Email: [email protected]
Importance of Sp1 consensus motifs in the MYCN promoter
Importance of Sp1 consensus motifs in the MYCN promoter Thomas H. Inge, MD, PhD, Lavona K. Casson, BS, Waldemar Priebe, PhD, John O. Trent, PhD, Keith E. Georgeson, MD, Donald M. Miller, MD, PhD, and Paula J. Bates, PhD, Cincinnati, Ohio, Louisville, Ky, Houston, Tex, and Birmingham, Ala
Background. MYCN (N-myc) amplification in neuroblastoma is associated with poor clinical outcome. Factors that regulate MYCN expression have not been elucidated. MYCN is considered a TATA-less promoter, whereas significant promoter activity resides within 160 bp 5´ of the major transcription start site. This region contains two GC-rich motifs and a CT box, regions for potential transcription factor interaction. Methods. To characterize DNA-protein interactions in this region of the MYCN promoter, electrophoretic mobility shift assays, and promoter-reporter were used. Results. A MYCN promoter fragment was incubated with HeLa nuclear extract, with or without competitors. Three major protein/DNA complexes were formed. Formation of 2 complexes could be inhibited by unlabeled Sp1 consensus duplex and by the Sp1 site-specific drug WP631. Purified Sp1 protein produced a complex similar to that formed with HeLa extract. To determine whether these DNA/protein interactions could be blocked in a sequence-specific fashion, a triplex forming oligonucleotide (TFO) was used. This TFO was designed to bind in the major groove of the promoter, covering the CT-box (putative Sp1 binding) motif. When triplex formation was followed by addition of nuclear extract, protein binding was indeed inhibited. Functional significance of this inhibition was tested with pE/Bnmyc-luc, a promoter-reporter plasmid containing the human MYCN promoter driving luciferase expression. Incubation with TFO, but not control oligodeoxynucleotides, completely inhibited luciferase activity. Conclusions. These data suggest that protein binding does occur in regions of the MYCN promoter containing GC and CT box elements and that this interaction is important for MYCN promoter activity. By inference, these data also suggest that the proteins that bind in this region are Sp1 family members. (Surgery 2002;132:232-8.) From The Children’s Hospital Research Foundation and Department of Pediatric Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio; the Brown Cancer Center, University of Louisville, Louisville, Ky; the MD Anderson Cancer Center, Houston, Tex; and the Division of Pediatric Surgery, the Children’s Hospital of Alabama, Birmingham, Ala
NEUROBLASTOMA (NB), the most common extracranial solid malignancy of childhood, is a rapidly progressive, fatal tumor in more than half of all patients. In contrast to other childhood malignancies, there have been only incremental advances in Supported in part by The Children’s Hospital of Alabama Research Institute Faculty Development Award, and the Division of Pediatric Surgery, The Children’s Hospital of Alabama, Birmingham, University of Alabama, Birmingham, Birmingham, Ala, and the Department of Pediatric Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio. Presented at the 63rd Annual Meeting of the Society of University Surgeons, Honolulu, Hawaii, February 14-16, 2002. Reprint requests: Thomas H. Inge, MD, PhD, Department of Pediatric Surgery, Children’s Hospital and Medical Center, 3333 Burnet Ave, Cincinnati, OH 45229. © 2002, Mosby, Inc. All rights reserved. 0039-6060/2002/$35.00 + 0 11/6/125387 doi:10.1067/msy.2002.125387
232 SURGERY
outcomes for patients with high-risk NB, and overall survival remains poor.1 Treatment advances will depend on a fundamental understanding of the biology of the disease. The best characterized genetic abnormality of NB is amplification of the MYCN oncogene.2 MYCN has been shown to participate in the malignant transformation of mammalian cells and was shown to be directly responsible for development of NB in mice by use of directed expression of a MYCN transgene,3 confirming an important role for MYCN in tumorigenesis of NB. Conversely, inhibition of MYCN expression can result in the loss of the transformed phenotype of NB cell cultures. When NB cells were treated with retinoic acid (which is known to down-regulate MYCN expression)4 or with MYCN antisense oligodeoxyribonucleotide (ODN),5 impaired proliferation and phenotypic changes associated with differentiation resulted. Thus, inhibition of MYCN is associated with cellu-
Surgery Volume 132, Number 2
Inge et al 233
Fig 1. Sequences of duplex promoter fragments and single stranded ODN used in these studies. Line 1 shows the longer promoter fragment used for protein binding experiments in Figs 2, 3, and 5. Line 2 shows the HNT4 specific target duplex promoter fragment used for TFO titration in Fig 4. HNT4 TFO (line 3) binds to the target duplex in an antiparallel orientation. Control ODNs were constructed as parallel to the target sequence (HNT4P, line 4, has opposite 5´/3´ polarity) or scrambled (HNT4S, line 5). Sp1 and Oct-1 duplex competitors are also shown (lines 6 and 7).
lar differentiation and loss of malignant characteristics. MYCN belongs to a small gene family that also includes MYCC (c-myc) and MYCL (L-myc). The MYC proteins are helix-loop-helix DNA binding phosphoproteins that have been independently conserved throughout evolution, suggesting that these molecules perform distinct and biologically important functions. These transcription factors play critical roles in cellular differentiation, proliferation, and apoptosis.6 The expression of MYCN is tightly regulated in normal cells during differentiation and during early B cell maturation, and lack of MYCN expression during embryogenesis is a lethal condition for the mouse. In normal human and murine somatic cells, no MYCN expression is detected, whereas in several tumor types, this locus is amplified. The regulation of MYCN gene expression is a complex process and involves multiple levels of control. Control at the level of transcriptional initiation, elongation, mRNA processing and translation have all been described in various systems.7,8 Despite the importance of understanding the mechanisms controlling MYCN expression, to date the specific nuclear factors that participate in the control of MYCN expression have not been defined. GC-rich elements, such as GGGCGG, and CT boxes have been described in the promoters of many eukaryotic genes and are thought to be recognition sites for transcription regulatory factors.9 Sp1 is a ubiquitous, general transcription factor that is a member of a much larger family of zinc finger DNA binding proteins. These transcription factors have been found to modulate
mRNA transcription.10 Sp1 was found to transactivate the human MYCC promoter, whereas Sp3 repressed Sp1-dependent activation of MYCC.11 Both human and murine MYCN promoters have independently conserved GC-rich motifs that are thought to play a role in promoter regulation. Although there has been speculation about a role of Sp1 in MYCN promoter function, no studies have defined a role for Sp1 family members in MYCN transcriptional regulation. The MYCN promoter has been studied extensively by using deletion mutagenesis. Essentially no detectable promoter activity (~2% of maximal activity) can be demonstrated within the first 55 base pair (bp) upstream of the major cap site, a region that contains a TAATAA sequence.12 Basal promoter activity (~15% of maximal activity) was demonstrated with a construct that retains 121 bp of the promoter, a region that contains a GC-rich motif (GGG GCG GGG GG) and a CT box (CCCTCCCCC). A striking increase in promoter activity (~65% of maximal activity) was seen with a construct that retains 160 bp of the promoter, a region that contains an additional GC-rich motif (GGG CGG GG). These observations suggest that there is no functional TATA box in this gene, and that cis elements are present immediately 5´ of the major mRNA cap site that are critical for transcriptional initiation. These findings show that high levels of promoter activity can be seen with constructs containing only 160 bp upstream of exon 1 and raise the possibility that transcription factor binding to regulator sequences in this area may strongly influence the activity of the MYCN promoter. We therefore sought to determine whether MYCN pro-
234 Inge et al
Fig 2. Protein binding EMSA. The sequence of human MYCN promoter used in this EMSA is shown in Fig 1. The sense strand of the promoter fragment was 32P-labeled. SS, Single (sense) strand alone; DS, double strand; HeLa NE, 4.4-µg HeLa nuclear extract; Sp1 Comp, Sp1 competitor duplex added at increasing concentrations in nmol/L; Sp1, 5 footprint units of purified Sp1 protein. a, b, and c along the left side of the figure designate individual protein/DNA complexes seen in these experiments. These data are representative of those obtained in 3 separate experiments.
moter sequences lying just 5´ of the TATA box were capable of nuclear protein binding and also to assess the importance of such interaction on promoter function. METHODS ODN. Phosphodiester ODNs were obtained from Life Technologies, Inc, Bethesda, Md. The ODN sequences and relationship to the promoter duplex are shown in Fig 1. HNT4 triplex forming oligonucleotide (TFO) was designed by O’Donoghue13 to bind as a third strand in the antiparallel orientation in the major groove of its triplex target region in MYCN promoter. HNT4P was designed to be parallel to the coding strand in the triplex forming region (and thus the
Surgery August 2002
reverse of the HNT4 molecule); HNT4S represents a scrambled ODN with G and T content identical to HNT4. Both of these represent control ODN. Protein binding electrophoretic mobility shift assay. Duplex probes were constructed by first 5´-labeling the sense ODN strand (Fig 1, lines 1, 2) with [γ32P] adenosine triphosphate (ATP) by using T4 polynucleotide kinase. Unincorporated ATP was removed by using a G25 column (5 Prime-3 Prime Inc, Boulder, Colo). Next, labeled ODNs were annealed to complementary strands by heating to 95°C and cooling slowly to room temperature. Labeled duplex (50,000 cpm, equivalent to approximately 0.4 nmol/L duplex per reaction) was incubated with either 5 units of purified Sp1 (Promega Corp, Madison, Wis) or 4.4 µg of HeLa cell nuclear extract (HeLa NE; Promega) and 0.25 µg of poly (dI-dC) nonspecific competitor in buffer consisting of 25 mmol/L HEPES pH 7.4, 12.5 mmol/L MgCl2, 70 mmol/L KCl, 1 µmol/L ZnSO4, l mmol/L dithiothreitol, 0.1% NP-40, 10% (v/v) glycerol, with or without duplex Sp1 or Oct1 competitors for 1 hour at 37°C. WP631 was synthesized as previously described.14 For experiments that examined triplex interference with protein binding, TFO or control ODN (Fig 1, lines 3 and 5) were preincubated overnight with promoter duplexes in TBM buffer (89 mmol/L Tris, 89 mmol/L boric acid [pH 7.4], 20 mmol/L MgCl) at 4°C before addition of NE in the above buffer. The samples were analyzed on a non-denaturing 5% polyacrylamide gel at 90 V in TBM buffer. Electrophoretic mobility shift assay of triplex formation. The sense strand of the specific HNT4 target (Fig 1, line 2) was 5´-labeled with [γ32P] ATP by using T4 polynucleotide kinase and then annealed to its complementary purine-rich strand as described above. TFO or control ODN was incubated overnight at 4°C with the labeled duplex in TBM buffer. Samples were analyzed on 15% nondenaturing polyacrylamide TBM gels at 90 to 120 V. Transient transfection and luciferase assay. A liposome vehicle consisting of 1:1 mixture of cationic lipid 1,2-dioleoyloxy-3-(trimethylammonium)-propane (DOTAP) and the neutral lipid dioleoyl-phosphatidylethanolamine (DOPE) was obtained from Avanti Polar Lipids (Alabaster, Ala) and used for transient transfections. DOTAP/ DOPE liposomes were prepared by mixing 0.5 mg of DOTAP and 0.5 mg of DOPE and evaporating the chloroform solvent. After the addition of 500 µL of cyclohexane, the mixture was placed on dry ice and lyophilized. One mL of sterile water was added to the lyophilized lipids, and the solution
Surgery Volume 132, Number 2
Inge et al 235
Fig 4. EMSA of HNT4. To determine the affinity of triplex formation at this site, a 25-bp duplex target was constructed (Fig 1, line 2) and EMSA performed. For EMSA, the probe (DS) was incubated without (lane 1) or with various nanomolar concentrations of HNT4 TFO, and gel electrophoresis was used to test for triplex formation (lanes 29). Control ODNs are shown in the far right 2 lanes (P, HNT4P; S, HNT4S). These results are representative of data obtained in 3 individual experiments.
Fig 3. Sp1 competitors interfere with protein binding to MYCN promoter. EMSA was performed with 32P-labeled MYCN promoter duplex (DS) as a probe. Experimental conditions are as above in Fig 2. a, b, and c designate individual protein/DNA complexes seen in these experiments. These results are representative of those obtained in 2 separate experiments.
free Opti-Mem I (Life Technologies) for 30 minutes at room temperature. DMEM was aspirated from HeLa cells, and 500 µL of transfection solution was added to each well and incubated for 4 hours after which the medium was aspirated, and 2 mL of complete DMEM was replaced in each well, followed by 18 hours of incubation. The medium was then removed by aspiration, cells were washed, and luciferase and β-galactosidase activity were detected with their respective assay kits (Promega).
was vortexed every 5 minutes for 30 minutes. HeLa cells were obtained from the American Type Culture Collection (Rockville, Md) and grown in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum and 1% antibiotics at 37°C with 85% humidity and 5% C02. Cells in logphase growth were harvested, washed, and seeded at 40% confluency in 24-well plates 1 day before use. The pE/Bnmyc-luc plasmid was generously provided by Dr William Carroll (University of Utah, Salt Lake City, Utah).15 This luciferase reporter construct contains a 1830 bp SacI/BamH1 human MYCN promoter fragment cloned into the pGL3basic vector (Promega). To transfect triplicate wells, 3 µg of pE/Bnmyc-luc (and 0.3 µg of pSV40βGal control plasmid) were incubated with or without 10 µmol/L ODN for 30 minutes at room temperature in TBM. DOTAP/DOPE was added at a ratio of 4 µg of lipid per µg of plasmid DNA. Liposomes were then dissolved in 1.5 mL serum-
RESULTS Prior work has documented the importance of the first 200 bases 5´ of exon 1 for MYCN promoter activity. To further characterize protein binding in this region, a homologous promoter fragment was synthesized (Fig 1, line 1). Protein complex formation within this promoter fragment was examined by incubation with HeLa NE, a rich source of eukaryotic transcription factors. Three major complexes (a, b and c) were formed with the promoter duplex (Fig 2, lane 3). To further investigate the identity of these complexes, unlabeled oligonucleotide representing the Sp1 consensus duplex was titrated into the protein binding reaction. As shown in Fig 2, Sp1 consensus duplex completely competed for complex a binding when added in excess concentration. Complex b was also competed away by the Sp1 competitor consensus duplex, although to a lesser extent than complex a. Minimal competition for complex c was seen. Electrophoretic mobility shift
236 Inge et al
Surgery August 2002
Fig 6. Effect of TFO on MYCN promoter activity. The effect of HNT4 TFO was compared to either parallel or scrambled control ODN. Vector alone consisted of pGL-3 basic without a promoter inserted into polylinker site.
Fig 5. HNT4 TFO competes for Sp1 binding. Protein binding EMSA was performed with the MYCN promoter fragment as probe (Fig 1, line 1). DS, Double strand; Sp1, 100 nmol/L; HNT4, 1 µmol/L; S, 1 µmol/L HNT4S; P, 1 µmol/L HNT4P. These results are representative of data obtained in 3 separate experiments.
assay (EMSA) was also performed by using purified Sp1 protein to ascertain directly whether Sp1 was capable of interacting with the promoter. Incubation of the MYCN promoter fragment with Sp1 resulted in a single mobility-shifted complex (Fig 2, panel A, lane 9), with mobility similar to complex a in experiments with HeLa NE. There was no specific competition for complexes when control duplexes corresponding to the Oct1 consensus sequence were used (Fig 3). WP631 is a bis-intercalating anthracycline antibiotic that strongly binds to GC-rich sequences of DNA.16 This compound has been shown to bind selectively to the Sp1 recognition sequences in DNA and block Sp1-activated transcriptional initiation.17 When EMSA was performed with the MYCN promoter fragment, formation of complexes a and b was strikingly inhibited by addition of WP631 in a dose-dependent fashion (Fig 3). Of note, complex a was more efficiently competed than complex b, and the more rapidly migrating complex c was again not inhibited by either Sp1 consensus duplex or WP631. Collectively, these data suggest that complexes a and b represent binding of Sp1 or Sp1-like molecules to important control regions of the MYCN promoter. TFOs have been used previously to block transcription factor binding to the promoter of various genes, and, as opposed to chemotherapeutic
agents, TFOs are not toxic to cells.18 Within the MYCN promoter fragment used here, there is a 25 bp polypurine/polypyrimidine tract (–130 to –105), which has been previously shown to be capable of triplex formation.13 HNT4 formed triplex DNA when added to duplex at concentrations as low as 12.5 nmol/L (Fig 4). With titration of HNT4, increasing triplex formation was seen with an apparent dissociation constant of 50 nmol/L in accord with a prior report.13 This binding was specific, because neither parallel nor scrambled control ODNs demonstrated any binding to the duplex. EMSA was next used to determine whether triplex formation interfered with protein binding. HNT4 and control ODN were incubated with promoter duplex before addition of HeLa NE. HNT4 competed for binding to the slowest migrating complex (a*) as did Sp1 consensus duplex (Fig 5). Neither Oct1 consensus duplex (data not shown) nor parallel or scrambled versions of HNT4 competed for binding of complex a*. Complex b* was weak in the absence of competitor but increased in the presence of HNT4. Thus, formation of complex a* was clearly inhibited by both Sp1 consensus duplex and HNT4, suggesting that HNT4 blocks binding of Sp1 or Sp1-like molecules to the MYCN promoter. To assess the biologic relevance of interference with protein binding, a promoter-reporter construct was used. The pE/Bnmyc-luc plasmid contains an 1830 nucleotide SacI-BamHI human MYCN promoter fragment cloned into the pGL3basic vector (Promega). This construct was incubated with HNT4 TFO or control ODN for 30 minutes before transient transfection of HeLa cells. The MYCN promoter drove active transcription of the luciferase gene when pretreated with
Surgery Volume 132, Number 2
either parallel or scrambled control ODN. However, when the plasmid was treated with HNT4 before transfection, profound inhibition of luciferase production occurred (Fig 6). The level of luciferase activity in HNT4 groups was similar to background luciferase activity seen when a promoter-less pGL3-basic vector was used. To control for DNA transfection efficiency, an internal control was included. Specifically, pSV40βgal was cotransfected with pE/Bnmyc-luc, and luciferase results were normalized to β-galactosidase activity for each sample. To rule out nonspecific toxicity of the HNT4 TFO, an irrelevant promoter (which has no HNT4 target sequences) was also tested. pSV40-luc consists of the SV40 viral promoter driving luciferase production. This promoter contains known Sp1 binding sites that are critical for promoter activity. When this pSV40-luc plasmid was incubated with ODN, there was no inhibition of luciferase activity associated with HNT4 TFO or HNT4P, demonstrating that the inhibition of pE/Bnmyc-luc seen with HNT4 was not due to nonspecific toxicity of the ODN. Collectively, these data provide evidence that TFO can inhibit protein binding in the MYCN promoter, and this inhibition leads to marked down-regulation of transcriptional activity of the promoter. DISCUSSION This report underscores the importance of putative transcription factor binding sites within a functionally important region of the MYCN promoter. Because the minimal promoter for efficient transcription of MYCN is embodied in the 3´-most 160 bp upstream of the first exon, this region was chosen to examine for protein binding. We found that this region does bind nuclear proteins, including Sp1. In fact, 2 discrete protein/DNA bands were detected that were specifically competed by Sp1 consensus duplex, one of which (a) comigrated with purified Sp1 and one of which (b) most likely represents an Sp1 family member. A third complex (c) most likely represents protein/DNA interactions not associated with the Sp1 binding regions, because it was seen despite incubation with competitor duplex or WP631. WP631 is a bis-intercalating DNA ligand that specifically interacts with Sp1 binding regions of DNA and competes with Sp1 for DNA binding, thereby blocking transcriptional initiation. We used WP631 as a second experimental approach to confirm the identity of proteins interacting with the promoter. Coincubation of MYCN promoter duplex with WP631 blocked formation of the same 2 complexes that were competed by the Sp1 con-
Inge et al 237
sensus duplex, providing support for the hypothesis that Sp1 binding occurs in this region. TFOs have been found to block expression of numerous genes including MYCC, 19 c-Ki-ras, 20 HER-2/neu,21 and cyclin D1 genes.22 The mechanism by which this inhibition occurs has been attributed to inhibition of regulatory protein binding.18 In vitro, HNT4 TFO was found to bind MYCN promoter DNA and deliver a DNA-fragmenting dose of radionuclide. The HNT4 target site contains a CT box element (CCCTCCCC at –114 to –107), and the 5´ most end of the HNT4 site lies only 12 bp 3´ of the GC-rich region at –147 of the MYCN promoter. It seemed plausible that HNT4 may therefore interfere with Sp1 binding to the promoter. We found that both HNT4 and Sp1 consensus duplexes interfered with formation of the same DNA/protein complex (complex a). Triplex formation with HNT4 therefore most likely directly blocks protein binding to the CT element within the triplex target site. DesJardins and Hay found that Sp1 recognizes identical CT boxes in the MYCC promoter, which were critical for promoter function. 23 Therefore, we reason that complex a (Figs 2 and 3) represents the promoter fragment bound to Sp1, whereas complex b could represent binding of another Sp1 family member binding at the GC-rich motif. Studies to confirm this possibility are underway. The importance of protein binding at the CT box motif (complex a*) was apparent when promoter function assays were performed. As compared to controls, HNT4 TFO potently inhibited MYCN promoter-dependent luciferase production when promoter constructs were treated with TFO before transient transfection. TFOs targeting MYCC have been used to induce terminal differentiation of myeloid leukemia24 and to inhibit growth of cervical and ovarian carcinoma cells.25 It is certainly possible that regulation of MYCN in NB cells is more complex than can be modeled by using transient transfection into HeLa cells. Thus, further studies are ongoing to determine whether TFO can be similarly used to block endogenous MYCN expression in NB cells. If so, TFO may represent a nontoxic method to manipulate MYCN expression and thereby alter NB growth or differentiation programs. In summary, 4 separate methods (purified Sp1 binding, inhibition with Sp1 consensus ODN, inhibition with Sp1 site-selective drug, and inhibition with site-specific TFO) were used to show that Sp1 or Sp1-like proteins interact with a region of the MYCN promoter known to be critical for basal pro-
238 Inge et al
moter activity. When the CT-box Sp1 binding site was blocked with TFO, the promoter was silenced, suggesting a biologically important function of this motif in the MYCN promoter. REFERENCES 1. Castleberry RP. Biology and treatment of neuroblastoma. Pediatr Clin North Am 1997;44:919-37. 2. Seeger RC, Brodeur GM, Sather H, Dalton A, Siegel SE, Wong KY, et al. Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas. N Engl J Med 1985;313:1111-6. 3. Weiss WA, Aldape K, Mohapatra G, Feuerstein BG, Bishop JM. Targeted expression of MYCN causes neuroblastoma in transgenic mice. EMBO J 1997;16:2985-95. 4. Thiele CJ, Reynolds CP, Israel MA. Decreased expression of N-myc precedes retinoic acid-induced morphological differentiation of human neuroblastoma. Nature 1985;313: 404-6. 5. Negroni A, Scarpa S, Romeo A, Ferrari S, Modesti A, Raschella G. Decrease of proliferation rate and induction of differentiation by a MYCN antisense DNA oligomer in a human neuroblastoma cell line. Cell Growth Differ 1991;2:511-8. 6. Thompson EB. The many roles of c-Myc in apoptosis. Annu Rev Physiol 1998;60:575-600. 7. Smith RK, Zimmerman K, Yancopoulos GD, Ma A, Alt FW. Transcriptional down-regulation of N-myc expression during B-cell development. Mol Cell Biol 1992;12:1578-84. 8. Krystal G, Birrer M, Way J, Nau M, Sausville E, Thompson C, et al. Multiple mechanisms for transcriptional regulation of the myc gene family in small-cell lung cancer. Mol Cell Biol 1988;8:3373-81. 9. Letovsky J, Dynan WS. Measurement of the binding of transcription factor Sp1 to a single GC box recognition sequence. Nucleic Acids Res 1989;17:2639-53. 10. Udvadia AJ, Rogers KT, Higgins PD, Murata Y, Martin KH, Humphrey PA, et al. Sp-1 binds promoter elements regulated by the RB protein and Sp-1-mediated transcription is stimulated by RB coexpression. Proc Natl Acad Sci U S A 1993;90:3265-9. 11. Majello B, De Luca P, Suske G, Lania L. Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 1995;10:1841-8. 12. Wada RK, Seeger RC, Reynolds CP, Alloggiamento T, Yamashiro JM, Ruland C, et al. Cell type-specific expression
Surgery August 2002 and negative regulation by retinoic acid of the human Nmyc promoter in neuroblastoma cells. Oncogene 1992;7:711-7. 13. O’Donoghue JA. Strategies for selective targeting of Auger electron emitters to tumor cells. J Nucl Med 1996; 37:3S-6S. 14. Chaires JB, Leng F, Przewloka T, Fokt I, Ling YH, PerezSoler R, et al. Structure-based design of a new bisintercalating anthracycline antibiotic. J Med Chem 1997;40:261-6. 15. Sivak LE, Tai KF, Smith RS, Dillon PA, Brodeur GM, Carroll WL. Autoregulation of the human N-myc oncogene is disrupted in amplified but not single-copy neuroblastoma cell lines. Oncogene 1997;15:1937-46. 16. Robinson H, Priebe W, Chaires JB, Wang AH. Binding of two novel bisdaunorubicins to DNA studied by NMR spectroscopy. Biochemistry 1997;36:8663-70. 17. Martin B, Vaquero A, Priebe W, Portugal J. Bisanthracycline WP631 inhibits basal and Sp1-activated transcription initiation in vitro. Nucleic Acids Res 1999;27:3402-9. 18. Reddoch JF, Miller DM. Inhibition of nuclear protein binding to two sites in the murine c-myc promoter by intermolecular triplex formation. Biochemistry 1995;34:7659-67. 19. Kim HG, Miller DM. Inhibition of in vitro transcription by a triplex-forming oligonucleotide targeted to human c-myc P2 promoter. Biochemistry 1995;34:8165-71. 20. Vigneswaran N, Mayfield CA, Rodu B, James R, Kim HG, Miller DM. Influence of GC and AT specific DNA minor groove binding drugs on intermolecular triplex formation in the human c-Ki-ras promoter. Biochemistry 1996;35: 1106-14. 21. Ebbinghaus SW, Gee JE, Rodu B, Mayfield CA, Sanders G, Miller DM. Triplex formation inhibits HER-2/neu transcription in vitro. J Clin Invest 1993;92:2433-9. 22. Kim HG, Miller DM. A novel triplex-forming oligonucleotide targeted to human cyclin D1 (bcl-1, proto-oncogene) promoter inhibits transcription in HeLa cells. Biochemistry 1998;37:2666-72. 23. Ponzoni M, Guarnaccia F, Corrias MV, Cornaglia-Ferraris P. Uncoordinate induction and differential regulation of HLA class-I and class-II expression by gamma-interferon in differentiating human neuroblastoma cells. Int J Cancer 1993;55:817-23. 24. Gee JE, Miller DM. Structure and applications of intermolecular DNA triplexes. Am J Med Sci 1992;304:366-72. 25. Helm CW, Shrestha K, Thomas S, Shingleton HM, Miller DM. A unique c-myc-targeted triplex-forming oligonucleotide inhibits the growth of ovarian and cervical carcinomas in vitro. Gynecol Oncol 1993;49:339-43.
Background. MYCN (N-myc) amplification in neuroblastoma is associated with poor clinical outcome. Factors that regulate MYCN expression have not been elucidated. MYCN is considered a TATA-less promoter, whereas significant promoter activity resides within 160 bp 5´ of the major transcription start site. This region contains two GC-rich motifs and a CT box, regions for potential transcription factor interaction. Methods. To characterize DNA-protein interactions in this region of the MYCN promoter, electrophoretic mobility shift assays, and promoter-reporter were used. Results. A MYCN promoter fragment was incubated with HeLa nuclear extract, with or without competitors. Three major protein/DNA complexes were formed. Formation of 2 complexes could be inhibited by unlabeled Sp1 consensus duplex and by the Sp1 site-specific drug WP631. Purified Sp1 protein produced a complex similar to that formed with HeLa extract. To determine whether these DNA/protein interactions could be blocked in a sequence-specific fashion, a triplex forming oligonucleotide (TFO) was used. This TFO was designed to bind in the major groove of the promoter, covering the CT-box (putative Sp1 binding) motif. When triplex formation was followed by addition of nuclear extract, protein binding was indeed inhibited. Functional significance of this inhibition was tested with pE/Bnmyc-luc, a promoter-reporter plasmid containing the human MYCN promoter driving luciferase expression. Incubation with TFO, but not control oligodeoxynucleotides, completely inhibited luciferase activity. Conclusions. These data suggest that protein binding does occur in regions of the MYCN promoter containing GC and CT box elements and that this interaction is important for MYCN promoter activity. By inference, these data also suggest that the proteins that bind in this region are Sp1 family members. (Surgery 2002;132:232-8.) From The Children’s Hospital Research Foundation and Department of Pediatric Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio; the Brown Cancer Center, University of Louisville, Louisville, Ky; the MD Anderson Cancer Center, Houston, Tex; and the Division of Pediatric Surgery, the Children’s Hospital of Alabama, Birmingham, Ala
NEUROBLASTOMA (NB), the most common extracranial solid malignancy of childhood, is a rapidly progressive, fatal tumor in more than half of all patients. In contrast to other childhood malignancies, there have been only incremental advances in Supported in part by The Children’s Hospital of Alabama Research Institute Faculty Development Award, and the Division of Pediatric Surgery, The Children’s Hospital of Alabama, Birmingham, University of Alabama, Birmingham, Birmingham, Ala, and the Department of Pediatric Surgery, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio. Presented at the 63rd Annual Meeting of the Society of University Surgeons, Honolulu, Hawaii, February 14-16, 2002. Reprint requests: Thomas H. Inge, MD, PhD, Department of Pediatric Surgery, Children’s Hospital and Medical Center, 3333 Burnet Ave, Cincinnati, OH 45229. © 2002, Mosby, Inc. All rights reserved. 0039-6060/2002/$35.00 + 0 11/6/125387 doi:10.1067/msy.2002.125387
232 SURGERY
outcomes for patients with high-risk NB, and overall survival remains poor.1 Treatment advances will depend on a fundamental understanding of the biology of the disease. The best characterized genetic abnormality of NB is amplification of the MYCN oncogene.2 MYCN has been shown to participate in the malignant transformation of mammalian cells and was shown to be directly responsible for development of NB in mice by use of directed expression of a MYCN transgene,3 confirming an important role for MYCN in tumorigenesis of NB. Conversely, inhibition of MYCN expression can result in the loss of the transformed phenotype of NB cell cultures. When NB cells were treated with retinoic acid (which is known to down-regulate MYCN expression)4 or with MYCN antisense oligodeoxyribonucleotide (ODN),5 impaired proliferation and phenotypic changes associated with differentiation resulted. Thus, inhibition of MYCN is associated with cellu-
Surgery Volume 132, Number 2
Inge et al 233
Fig 1. Sequences of duplex promoter fragments and single stranded ODN used in these studies. Line 1 shows the longer promoter fragment used for protein binding experiments in Figs 2, 3, and 5. Line 2 shows the HNT4 specific target duplex promoter fragment used for TFO titration in Fig 4. HNT4 TFO (line 3) binds to the target duplex in an antiparallel orientation. Control ODNs were constructed as parallel to the target sequence (HNT4P, line 4, has opposite 5´/3´ polarity) or scrambled (HNT4S, line 5). Sp1 and Oct-1 duplex competitors are also shown (lines 6 and 7).
lar differentiation and loss of malignant characteristics. MYCN belongs to a small gene family that also includes MYCC (c-myc) and MYCL (L-myc). The MYC proteins are helix-loop-helix DNA binding phosphoproteins that have been independently conserved throughout evolution, suggesting that these molecules perform distinct and biologically important functions. These transcription factors play critical roles in cellular differentiation, proliferation, and apoptosis.6 The expression of MYCN is tightly regulated in normal cells during differentiation and during early B cell maturation, and lack of MYCN expression during embryogenesis is a lethal condition for the mouse. In normal human and murine somatic cells, no MYCN expression is detected, whereas in several tumor types, this locus is amplified. The regulation of MYCN gene expression is a complex process and involves multiple levels of control. Control at the level of transcriptional initiation, elongation, mRNA processing and translation have all been described in various systems.7,8 Despite the importance of understanding the mechanisms controlling MYCN expression, to date the specific nuclear factors that participate in the control of MYCN expression have not been defined. GC-rich elements, such as GGGCGG, and CT boxes have been described in the promoters of many eukaryotic genes and are thought to be recognition sites for transcription regulatory factors.9 Sp1 is a ubiquitous, general transcription factor that is a member of a much larger family of zinc finger DNA binding proteins. These transcription factors have been found to modulate
mRNA transcription.10 Sp1 was found to transactivate the human MYCC promoter, whereas Sp3 repressed Sp1-dependent activation of MYCC.11 Both human and murine MYCN promoters have independently conserved GC-rich motifs that are thought to play a role in promoter regulation. Although there has been speculation about a role of Sp1 in MYCN promoter function, no studies have defined a role for Sp1 family members in MYCN transcriptional regulation. The MYCN promoter has been studied extensively by using deletion mutagenesis. Essentially no detectable promoter activity (~2% of maximal activity) can be demonstrated within the first 55 base pair (bp) upstream of the major cap site, a region that contains a TAATAA sequence.12 Basal promoter activity (~15% of maximal activity) was demonstrated with a construct that retains 121 bp of the promoter, a region that contains a GC-rich motif (GGG GCG GGG GG) and a CT box (CCCTCCCCC). A striking increase in promoter activity (~65% of maximal activity) was seen with a construct that retains 160 bp of the promoter, a region that contains an additional GC-rich motif (GGG CGG GG). These observations suggest that there is no functional TATA box in this gene, and that cis elements are present immediately 5´ of the major mRNA cap site that are critical for transcriptional initiation. These findings show that high levels of promoter activity can be seen with constructs containing only 160 bp upstream of exon 1 and raise the possibility that transcription factor binding to regulator sequences in this area may strongly influence the activity of the MYCN promoter. We therefore sought to determine whether MYCN pro-
234 Inge et al
Fig 2. Protein binding EMSA. The sequence of human MYCN promoter used in this EMSA is shown in Fig 1. The sense strand of the promoter fragment was 32P-labeled. SS, Single (sense) strand alone; DS, double strand; HeLa NE, 4.4-µg HeLa nuclear extract; Sp1 Comp, Sp1 competitor duplex added at increasing concentrations in nmol/L; Sp1, 5 footprint units of purified Sp1 protein. a, b, and c along the left side of the figure designate individual protein/DNA complexes seen in these experiments. These data are representative of those obtained in 3 separate experiments.
moter sequences lying just 5´ of the TATA box were capable of nuclear protein binding and also to assess the importance of such interaction on promoter function. METHODS ODN. Phosphodiester ODNs were obtained from Life Technologies, Inc, Bethesda, Md. The ODN sequences and relationship to the promoter duplex are shown in Fig 1. HNT4 triplex forming oligonucleotide (TFO) was designed by O’Donoghue13 to bind as a third strand in the antiparallel orientation in the major groove of its triplex target region in MYCN promoter. HNT4P was designed to be parallel to the coding strand in the triplex forming region (and thus the
Surgery August 2002
reverse of the HNT4 molecule); HNT4S represents a scrambled ODN with G and T content identical to HNT4. Both of these represent control ODN. Protein binding electrophoretic mobility shift assay. Duplex probes were constructed by first 5´-labeling the sense ODN strand (Fig 1, lines 1, 2) with [γ32P] adenosine triphosphate (ATP) by using T4 polynucleotide kinase. Unincorporated ATP was removed by using a G25 column (5 Prime-3 Prime Inc, Boulder, Colo). Next, labeled ODNs were annealed to complementary strands by heating to 95°C and cooling slowly to room temperature. Labeled duplex (50,000 cpm, equivalent to approximately 0.4 nmol/L duplex per reaction) was incubated with either 5 units of purified Sp1 (Promega Corp, Madison, Wis) or 4.4 µg of HeLa cell nuclear extract (HeLa NE; Promega) and 0.25 µg of poly (dI-dC) nonspecific competitor in buffer consisting of 25 mmol/L HEPES pH 7.4, 12.5 mmol/L MgCl2, 70 mmol/L KCl, 1 µmol/L ZnSO4, l mmol/L dithiothreitol, 0.1% NP-40, 10% (v/v) glycerol, with or without duplex Sp1 or Oct1 competitors for 1 hour at 37°C. WP631 was synthesized as previously described.14 For experiments that examined triplex interference with protein binding, TFO or control ODN (Fig 1, lines 3 and 5) were preincubated overnight with promoter duplexes in TBM buffer (89 mmol/L Tris, 89 mmol/L boric acid [pH 7.4], 20 mmol/L MgCl) at 4°C before addition of NE in the above buffer. The samples were analyzed on a non-denaturing 5% polyacrylamide gel at 90 V in TBM buffer. Electrophoretic mobility shift assay of triplex formation. The sense strand of the specific HNT4 target (Fig 1, line 2) was 5´-labeled with [γ32P] ATP by using T4 polynucleotide kinase and then annealed to its complementary purine-rich strand as described above. TFO or control ODN was incubated overnight at 4°C with the labeled duplex in TBM buffer. Samples were analyzed on 15% nondenaturing polyacrylamide TBM gels at 90 to 120 V. Transient transfection and luciferase assay. A liposome vehicle consisting of 1:1 mixture of cationic lipid 1,2-dioleoyloxy-3-(trimethylammonium)-propane (DOTAP) and the neutral lipid dioleoyl-phosphatidylethanolamine (DOPE) was obtained from Avanti Polar Lipids (Alabaster, Ala) and used for transient transfections. DOTAP/ DOPE liposomes were prepared by mixing 0.5 mg of DOTAP and 0.5 mg of DOPE and evaporating the chloroform solvent. After the addition of 500 µL of cyclohexane, the mixture was placed on dry ice and lyophilized. One mL of sterile water was added to the lyophilized lipids, and the solution
Surgery Volume 132, Number 2
Inge et al 235
Fig 4. EMSA of HNT4. To determine the affinity of triplex formation at this site, a 25-bp duplex target was constructed (Fig 1, line 2) and EMSA performed. For EMSA, the probe (DS) was incubated without (lane 1) or with various nanomolar concentrations of HNT4 TFO, and gel electrophoresis was used to test for triplex formation (lanes 29). Control ODNs are shown in the far right 2 lanes (P, HNT4P; S, HNT4S). These results are representative of data obtained in 3 individual experiments.
Fig 3. Sp1 competitors interfere with protein binding to MYCN promoter. EMSA was performed with 32P-labeled MYCN promoter duplex (DS) as a probe. Experimental conditions are as above in Fig 2. a, b, and c designate individual protein/DNA complexes seen in these experiments. These results are representative of those obtained in 2 separate experiments.
free Opti-Mem I (Life Technologies) for 30 minutes at room temperature. DMEM was aspirated from HeLa cells, and 500 µL of transfection solution was added to each well and incubated for 4 hours after which the medium was aspirated, and 2 mL of complete DMEM was replaced in each well, followed by 18 hours of incubation. The medium was then removed by aspiration, cells were washed, and luciferase and β-galactosidase activity were detected with their respective assay kits (Promega).
was vortexed every 5 minutes for 30 minutes. HeLa cells were obtained from the American Type Culture Collection (Rockville, Md) and grown in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum and 1% antibiotics at 37°C with 85% humidity and 5% C02. Cells in logphase growth were harvested, washed, and seeded at 40% confluency in 24-well plates 1 day before use. The pE/Bnmyc-luc plasmid was generously provided by Dr William Carroll (University of Utah, Salt Lake City, Utah).15 This luciferase reporter construct contains a 1830 bp SacI/BamH1 human MYCN promoter fragment cloned into the pGL3basic vector (Promega). To transfect triplicate wells, 3 µg of pE/Bnmyc-luc (and 0.3 µg of pSV40βGal control plasmid) were incubated with or without 10 µmol/L ODN for 30 minutes at room temperature in TBM. DOTAP/DOPE was added at a ratio of 4 µg of lipid per µg of plasmid DNA. Liposomes were then dissolved in 1.5 mL serum-
RESULTS Prior work has documented the importance of the first 200 bases 5´ of exon 1 for MYCN promoter activity. To further characterize protein binding in this region, a homologous promoter fragment was synthesized (Fig 1, line 1). Protein complex formation within this promoter fragment was examined by incubation with HeLa NE, a rich source of eukaryotic transcription factors. Three major complexes (a, b and c) were formed with the promoter duplex (Fig 2, lane 3). To further investigate the identity of these complexes, unlabeled oligonucleotide representing the Sp1 consensus duplex was titrated into the protein binding reaction. As shown in Fig 2, Sp1 consensus duplex completely competed for complex a binding when added in excess concentration. Complex b was also competed away by the Sp1 competitor consensus duplex, although to a lesser extent than complex a. Minimal competition for complex c was seen. Electrophoretic mobility shift
236 Inge et al
Surgery August 2002
Fig 6. Effect of TFO on MYCN promoter activity. The effect of HNT4 TFO was compared to either parallel or scrambled control ODN. Vector alone consisted of pGL-3 basic without a promoter inserted into polylinker site.
Fig 5. HNT4 TFO competes for Sp1 binding. Protein binding EMSA was performed with the MYCN promoter fragment as probe (Fig 1, line 1). DS, Double strand; Sp1, 100 nmol/L; HNT4, 1 µmol/L; S, 1 µmol/L HNT4S; P, 1 µmol/L HNT4P. These results are representative of data obtained in 3 separate experiments.
assay (EMSA) was also performed by using purified Sp1 protein to ascertain directly whether Sp1 was capable of interacting with the promoter. Incubation of the MYCN promoter fragment with Sp1 resulted in a single mobility-shifted complex (Fig 2, panel A, lane 9), with mobility similar to complex a in experiments with HeLa NE. There was no specific competition for complexes when control duplexes corresponding to the Oct1 consensus sequence were used (Fig 3). WP631 is a bis-intercalating anthracycline antibiotic that strongly binds to GC-rich sequences of DNA.16 This compound has been shown to bind selectively to the Sp1 recognition sequences in DNA and block Sp1-activated transcriptional initiation.17 When EMSA was performed with the MYCN promoter fragment, formation of complexes a and b was strikingly inhibited by addition of WP631 in a dose-dependent fashion (Fig 3). Of note, complex a was more efficiently competed than complex b, and the more rapidly migrating complex c was again not inhibited by either Sp1 consensus duplex or WP631. Collectively, these data suggest that complexes a and b represent binding of Sp1 or Sp1-like molecules to important control regions of the MYCN promoter. TFOs have been used previously to block transcription factor binding to the promoter of various genes, and, as opposed to chemotherapeutic
agents, TFOs are not toxic to cells.18 Within the MYCN promoter fragment used here, there is a 25 bp polypurine/polypyrimidine tract (–130 to –105), which has been previously shown to be capable of triplex formation.13 HNT4 formed triplex DNA when added to duplex at concentrations as low as 12.5 nmol/L (Fig 4). With titration of HNT4, increasing triplex formation was seen with an apparent dissociation constant of 50 nmol/L in accord with a prior report.13 This binding was specific, because neither parallel nor scrambled control ODNs demonstrated any binding to the duplex. EMSA was next used to determine whether triplex formation interfered with protein binding. HNT4 and control ODN were incubated with promoter duplex before addition of HeLa NE. HNT4 competed for binding to the slowest migrating complex (a*) as did Sp1 consensus duplex (Fig 5). Neither Oct1 consensus duplex (data not shown) nor parallel or scrambled versions of HNT4 competed for binding of complex a*. Complex b* was weak in the absence of competitor but increased in the presence of HNT4. Thus, formation of complex a* was clearly inhibited by both Sp1 consensus duplex and HNT4, suggesting that HNT4 blocks binding of Sp1 or Sp1-like molecules to the MYCN promoter. To assess the biologic relevance of interference with protein binding, a promoter-reporter construct was used. The pE/Bnmyc-luc plasmid contains an 1830 nucleotide SacI-BamHI human MYCN promoter fragment cloned into the pGL3basic vector (Promega). This construct was incubated with HNT4 TFO or control ODN for 30 minutes before transient transfection of HeLa cells. The MYCN promoter drove active transcription of the luciferase gene when pretreated with
Surgery Volume 132, Number 2
either parallel or scrambled control ODN. However, when the plasmid was treated with HNT4 before transfection, profound inhibition of luciferase production occurred (Fig 6). The level of luciferase activity in HNT4 groups was similar to background luciferase activity seen when a promoter-less pGL3-basic vector was used. To control for DNA transfection efficiency, an internal control was included. Specifically, pSV40βgal was cotransfected with pE/Bnmyc-luc, and luciferase results were normalized to β-galactosidase activity for each sample. To rule out nonspecific toxicity of the HNT4 TFO, an irrelevant promoter (which has no HNT4 target sequences) was also tested. pSV40-luc consists of the SV40 viral promoter driving luciferase production. This promoter contains known Sp1 binding sites that are critical for promoter activity. When this pSV40-luc plasmid was incubated with ODN, there was no inhibition of luciferase activity associated with HNT4 TFO or HNT4P, demonstrating that the inhibition of pE/Bnmyc-luc seen with HNT4 was not due to nonspecific toxicity of the ODN. Collectively, these data provide evidence that TFO can inhibit protein binding in the MYCN promoter, and this inhibition leads to marked down-regulation of transcriptional activity of the promoter. DISCUSSION This report underscores the importance of putative transcription factor binding sites within a functionally important region of the MYCN promoter. Because the minimal promoter for efficient transcription of MYCN is embodied in the 3´-most 160 bp upstream of the first exon, this region was chosen to examine for protein binding. We found that this region does bind nuclear proteins, including Sp1. In fact, 2 discrete protein/DNA bands were detected that were specifically competed by Sp1 consensus duplex, one of which (a) comigrated with purified Sp1 and one of which (b) most likely represents an Sp1 family member. A third complex (c) most likely represents protein/DNA interactions not associated with the Sp1 binding regions, because it was seen despite incubation with competitor duplex or WP631. WP631 is a bis-intercalating DNA ligand that specifically interacts with Sp1 binding regions of DNA and competes with Sp1 for DNA binding, thereby blocking transcriptional initiation. We used WP631 as a second experimental approach to confirm the identity of proteins interacting with the promoter. Coincubation of MYCN promoter duplex with WP631 blocked formation of the same 2 complexes that were competed by the Sp1 con-
Inge et al 237
sensus duplex, providing support for the hypothesis that Sp1 binding occurs in this region. TFOs have been found to block expression of numerous genes including MYCC, 19 c-Ki-ras, 20 HER-2/neu,21 and cyclin D1 genes.22 The mechanism by which this inhibition occurs has been attributed to inhibition of regulatory protein binding.18 In vitro, HNT4 TFO was found to bind MYCN promoter DNA and deliver a DNA-fragmenting dose of radionuclide. The HNT4 target site contains a CT box element (CCCTCCCC at –114 to –107), and the 5´ most end of the HNT4 site lies only 12 bp 3´ of the GC-rich region at –147 of the MYCN promoter. It seemed plausible that HNT4 may therefore interfere with Sp1 binding to the promoter. We found that both HNT4 and Sp1 consensus duplexes interfered with formation of the same DNA/protein complex (complex a). Triplex formation with HNT4 therefore most likely directly blocks protein binding to the CT element within the triplex target site. DesJardins and Hay found that Sp1 recognizes identical CT boxes in the MYCC promoter, which were critical for promoter function. 23 Therefore, we reason that complex a (Figs 2 and 3) represents the promoter fragment bound to Sp1, whereas complex b could represent binding of another Sp1 family member binding at the GC-rich motif. Studies to confirm this possibility are underway. The importance of protein binding at the CT box motif (complex a*) was apparent when promoter function assays were performed. As compared to controls, HNT4 TFO potently inhibited MYCN promoter-dependent luciferase production when promoter constructs were treated with TFO before transient transfection. TFOs targeting MYCC have been used to induce terminal differentiation of myeloid leukemia24 and to inhibit growth of cervical and ovarian carcinoma cells.25 It is certainly possible that regulation of MYCN in NB cells is more complex than can be modeled by using transient transfection into HeLa cells. Thus, further studies are ongoing to determine whether TFO can be similarly used to block endogenous MYCN expression in NB cells. If so, TFO may represent a nontoxic method to manipulate MYCN expression and thereby alter NB growth or differentiation programs. In summary, 4 separate methods (purified Sp1 binding, inhibition with Sp1 consensus ODN, inhibition with Sp1 site-selective drug, and inhibition with site-specific TFO) were used to show that Sp1 or Sp1-like proteins interact with a region of the MYCN promoter known to be critical for basal pro-
238 Inge et al
moter activity. When the CT-box Sp1 binding site was blocked with TFO, the promoter was silenced, suggesting a biologically important function of this motif in the MYCN promoter. REFERENCES 1. Castleberry RP. Biology and treatment of neuroblastoma. Pediatr Clin North Am 1997;44:919-37. 2. Seeger RC, Brodeur GM, Sather H, Dalton A, Siegel SE, Wong KY, et al. Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas. N Engl J Med 1985;313:1111-6. 3. Weiss WA, Aldape K, Mohapatra G, Feuerstein BG, Bishop JM. Targeted expression of MYCN causes neuroblastoma in transgenic mice. EMBO J 1997;16:2985-95. 4. Thiele CJ, Reynolds CP, Israel MA. Decreased expression of N-myc precedes retinoic acid-induced morphological differentiation of human neuroblastoma. Nature 1985;313: 404-6. 5. Negroni A, Scarpa S, Romeo A, Ferrari S, Modesti A, Raschella G. Decrease of proliferation rate and induction of differentiation by a MYCN antisense DNA oligomer in a human neuroblastoma cell line. Cell Growth Differ 1991;2:511-8. 6. Thompson EB. The many roles of c-Myc in apoptosis. Annu Rev Physiol 1998;60:575-600. 7. Smith RK, Zimmerman K, Yancopoulos GD, Ma A, Alt FW. Transcriptional down-regulation of N-myc expression during B-cell development. Mol Cell Biol 1992;12:1578-84. 8. Krystal G, Birrer M, Way J, Nau M, Sausville E, Thompson C, et al. Multiple mechanisms for transcriptional regulation of the myc gene family in small-cell lung cancer. Mol Cell Biol 1988;8:3373-81. 9. Letovsky J, Dynan WS. Measurement of the binding of transcription factor Sp1 to a single GC box recognition sequence. Nucleic Acids Res 1989;17:2639-53. 10. Udvadia AJ, Rogers KT, Higgins PD, Murata Y, Martin KH, Humphrey PA, et al. Sp-1 binds promoter elements regulated by the RB protein and Sp-1-mediated transcription is stimulated by RB coexpression. Proc Natl Acad Sci U S A 1993;90:3265-9. 11. Majello B, De Luca P, Suske G, Lania L. Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 1995;10:1841-8. 12. Wada RK, Seeger RC, Reynolds CP, Alloggiamento T, Yamashiro JM, Ruland C, et al. Cell type-specific expression
Surgery August 2002 and negative regulation by retinoic acid of the human Nmyc promoter in neuroblastoma cells. Oncogene 1992;7:711-7. 13. O’Donoghue JA. Strategies for selective targeting of Auger electron emitters to tumor cells. J Nucl Med 1996; 37:3S-6S. 14. Chaires JB, Leng F, Przewloka T, Fokt I, Ling YH, PerezSoler R, et al. Structure-based design of a new bisintercalating anthracycline antibiotic. J Med Chem 1997;40:261-6. 15. Sivak LE, Tai KF, Smith RS, Dillon PA, Brodeur GM, Carroll WL. Autoregulation of the human N-myc oncogene is disrupted in amplified but not single-copy neuroblastoma cell lines. Oncogene 1997;15:1937-46. 16. Robinson H, Priebe W, Chaires JB, Wang AH. Binding of two novel bisdaunorubicins to DNA studied by NMR spectroscopy. Biochemistry 1997;36:8663-70. 17. Martin B, Vaquero A, Priebe W, Portugal J. Bisanthracycline WP631 inhibits basal and Sp1-activated transcription initiation in vitro. Nucleic Acids Res 1999;27:3402-9. 18. Reddoch JF, Miller DM. Inhibition of nuclear protein binding to two sites in the murine c-myc promoter by intermolecular triplex formation. Biochemistry 1995;34:7659-67. 19. Kim HG, Miller DM. Inhibition of in vitro transcription by a triplex-forming oligonucleotide targeted to human c-myc P2 promoter. Biochemistry 1995;34:8165-71. 20. Vigneswaran N, Mayfield CA, Rodu B, James R, Kim HG, Miller DM. Influence of GC and AT specific DNA minor groove binding drugs on intermolecular triplex formation in the human c-Ki-ras promoter. Biochemistry 1996;35: 1106-14. 21. Ebbinghaus SW, Gee JE, Rodu B, Mayfield CA, Sanders G, Miller DM. Triplex formation inhibits HER-2/neu transcription in vitro. J Clin Invest 1993;92:2433-9. 22. Kim HG, Miller DM. A novel triplex-forming oligonucleotide targeted to human cyclin D1 (bcl-1, proto-oncogene) promoter inhibits transcription in HeLa cells. Biochemistry 1998;37:2666-72. 23. Ponzoni M, Guarnaccia F, Corrias MV, Cornaglia-Ferraris P. Uncoordinate induction and differential regulation of HLA class-I and class-II expression by gamma-interferon in differentiating human neuroblastoma cells. Int J Cancer 1993;55:817-23. 24. Gee JE, Miller DM. Structure and applications of intermolecular DNA triplexes. Am J Med Sci 1992;304:366-72. 25. Helm CW, Shrestha K, Thomas S, Shingleton HM, Miller DM. A unique c-myc-targeted triplex-forming oligonucleotide inhibits the growth of ovarian and cervical carcinomas in vitro. Gynecol Oncol 1993;49:339-43.