Identification and Functional Analyses of Molecular Haplotypes of the Human Osteoprotegerin Gene Promoter
Objective— Osteoprotegerin (OPG) has been reported to be involved in the development of atherosclerotic disease, and OPG gene variation has been associated with plasma OPG levels and different cardiovascular disease phenotypes. However, the genetic architecture of the OPG promoter and its transcriptional regulation are poorly characterized.
Methods and Results— We identified 1008 bp of the OPG 5′-flanking region to be sufficiently transcriptionally active in osteosarcoma cell lines and generated serial promoter deletion constructs. Individual subcloning revealed the existence of 3 molecular haplotypes (MolHaps): [T−960-A−946-G−900-T−864; MolHap1, wild type], [T−960-G−946-G−900-T−864; MolHap2], [C−960-G−946-A−900-G−864; MolHap4]. Compared to MolHap1, transcriptional activities of MolHaps 2 and 4 were significantly reduced (P=0.0018). Whereas introduction of the −159C allele reduced transcriptional activities of the full-length constructs (P=0.0014), it significantly increased activities of the deletion constructs (P=0.0005). Electrophoretic mobility shift, competition, and chromatin immunoprecipitation assays revealed specific DNA:protein interactions for the MolHaps with Sp1 and NF-1, and identified Egr1 interacting exclusively with the −159T allele.
Conclusions— We propose new structural and transcriptional features within the OPG promoter region and identified MolHaps being differentially transcriptionally active and allele-dependently interacting with a proximal polymorphic site.
- genetic variants
- functional promoter analyses
- deletion constructs
- transcriptional control
As shown for the matrix Gla protein,1,2 osteopontin,3,4 and others—in human genetic and animal studies—osteoprotegerin (OPG) has been reported to be involved in the development of vascular calcification/atherosclerotic phenotypes.5–7 Increased OPG serum levels have been reported to be significantly related to cardiovascular disease (CVD) phenotypes,8 and in apparently healthy individuals, with an increased future risk of coronary events.9
Genetic variation at the human OPG gene locus has been associated with colorectal10/prostate cancer risk11 and bipolar disorders12 in whole genome approaches; in single variant analyses, OPG T-159C has been associated with elevated OPG serum levels,13 CVD,13,14 intima media thickness, and reduced maximal forearm blood flow in healthy subjects.15
The human OPG gene (8q24) consists of 5 exons and 4 introns, and is transcribed into 4 transcripts (2.4 kb, 3.0 kb, 4.2 kb, and 6.5 kb in lengths). The 2.4-kb form represents the major transcript and originates from transcriptional initiation from a TATA box 30 bp upstream of the transcription start site (TSS). Whereas the 3.0-kb transcript is derived from the alternative use of TSS, the 4.2-kb and 6.5-kb transcripts result from alternative splicing events.16
Because little is known about the structural architecture or functional aspects of the OPG promoter region, we (1) reanalyzed putative TSS in osteosarcoma cell lines SaOs-2 and U2Os by rapid amplification of 5′ cDNA Ends (5′ RACE); (2) directly sequenced 1008 bp of the OPG 5′-flanking region in a set of 114 chromosomes of patients with CVD to identify positions of single-nucleotide polymorphisms (SNPs); (3) determined the presence of molecular haplotypes (MolHaps) by individual DNA subcloning; and (4) performed molecular and functional profiling assays using reporter gene experiments, bandshift assays, chromatin immunoprecipitation (ChIP) assays, and coexpression analyses in 2 different cell lines.
The current investigation was based on the Münster Molecular Functional Profiling for Mechanism Detection (MolProMD) Study,17 a running study of CVD patients (MI, essential hypertension). The study was approved by the ethics committee of the Medical Faculty, Westfälische Wilhelms University of Münster, and written informed consent was obtained from all study subjects.
Identification of OPG Promoter Variants and Molecular Haplotypes
Genomic DNA was prepared from white blood cells (Qiagen), and 1008 bp of the 5′-flanking region of the OPG locus were scanned for genetic variation by sequencing both DNA strands (ABI Prism 3770, Perkin Elmer). Two hundred forty-nine bp and 275 bp of the promoter region (Acc# AB008822; harboring T-960C, A-946G, G-900A, T-864G, or T-159C) of genomic DNA from 57 CVD patients were amplified, subcloned, and sequenced twice.
In Silico Analyses of Putative Transcription Factor Binding Sites
All identified SNPs were subjected to computer-aided analyses using AliBaba 2.1 (www.gene-regulation.com),18 a position weight matrix algorithm, predicting physical transcription factor binding sites (TFBS) in a given DNA sequence based on the TRANSFAC database of eukaryotic transcription factors (TFs; TRANSFAC 10).
Detailed protocols on the human osteosarcoma cell lines SaOs-2 and U2Os are presented in the supplemental materials (available online at http://atvb.ahajournals.org). For the identification of endogenously OPG expressing cells, semiquantitative PCR using cDNA from different cell lines was performed.
Rapid Amplification of 5′ RACE
5′ RACE was performed as described previously19 using RNA extracted from SaOs-2 and U2Os cells.
ChIP was performed as described previously,17,20 using specific antibodies against specificity protein 1 (Sp1), nuclear factor-1 (NF-1), and early growth response factor-1 (Egr1).
Reporter Gene Constructs and Transient Transfections
The plasmids pOPG-MolHap1/luc, pOPG-MolHap2/luc, and pOPG-MolHap4/luc as well as the deletion constructs were constructed following standard cloning procedures. All reporter gene constructs were transfected in SaOs-2 and U2Os cells using Effectene transfection reagent (Qiagen).
Electrophoretic Mobility Shift Assay
Nuclear protein extracts were harvested by a modified procedure of the protocol published by Schreiber et al.21 To cover the promoter portion comprising T-960C, A-946G, G-900A, T-864G, a 154-bp fragment was amplified by PCR using GoTaq (Promega) and primers 5′-ACCACACTTTACAAGTCATCAAG-3′ (sense) and 5′-TTCCTACGCGCTGAACTTC-3′ (antisense). A 72-bp fragment, representing the cis-active element, was amplified by PCR using GoTaq (Promega) and primers 5′-ACCACACTTTACAAGTCATCAAG-3′ (sense) and 5′-CTCTAGGGTTCGCTGTCT-3′ (antisense). Double-stranded probes, harboring either MolHap1, 2, or 4, were 3′-biotinylated using biotin-16-ddUTP (Roche). Electrophoretic mobility shift assays (EMSEs) were performed using the LightShift Chemiluminescent EMSE kit (Thermo Fisher).
A detailed description of all methods is provided in the supplemental material.
Endogenous OPG mRNA Expression and Determination of TSS
Using different stimulatory regimes, OPG mRNA expression was detectable in both SaOs-2 and U2Os cells (supplemental Figure IA). Using 5′ RACE, we identified 2 TSS in both cell lines; TSS1 being located 64 bp upstream of the ATG codon, and TSS2 225 bp upstream of the ATG codon. We confirmed these results by diacritic PCR with specific primers, located directly at the respective TSS. Compared to TSS2 (2.4-kb transcript), which was used to a lesser extent, TSS1 (2.4-kb transcript) was equally and quantitatively used in both cell lines. We defined a third TSS (≈3.0-kb transcript), 667 bp upstream of the ATG codon by diacritic PCR in both cell lines, which was used as frequently as TSS1 in SaOs-2 but slightly less than TSS1 in U2Os. As TSS1 is usually described as major TSS, positions of SNPs are indicated relative to this site (Figure 1).
Identification of OPG Gene Variants and MolHaps
We identified 1 proximal (T-159C [rs2073617]) and 4 distal OPG promoter variants (T-960C [rs3134071], A-946G [rs3102735], G-900A [rs3134070], T-864G [rs3134069]). By individual subcloning, we identified 3 MolHaps in the MolProMD sample defined by the allelic constellation of the distal OPG SNPs. The OPG-MolHap frequencies were as follows: MolHap1 [T−960-A−946-G−900-T−864] ≈79%, MolHap2 [T−960- G−946- G−900- T−864] ≈12%, and MolHap4 [C−960- G−946- A−900- G−864] ≈9%. Alleles −960C, −900A, and −864G were always transmitted together. To define cis-active elements, we designed serial deletion constructs by 5′-truncation (supplemental Figure II).
In silico analyses of MolHaps 1, 2, and 4 using AliBaba 2.1 predicted several TFBS, some of which were altered in the presence of nucleotide substitutions (supplemental Figure IB). MolHap1 [T−960-A−946-G−900-T−864], MolHap2 [T−960-G−946-G−900-T−864], and MolHap4 [C−960-G−946-A−900-G−864] showed a consensus sequence similarity to NF-κB.
Transcriptional Activity of the OPG 5′-Flanking Region
The reporter vector pOPG-Del1/luc represents the full-length construct (1008 bp) with MolHap1 sequence [T−960-A−946-G−900-T−864-T−159]. The remaining constructs were 5′-truncated, each harboring MolHap1 sequences: pOPG-Del2/luc (924 bp [G−900-T−864-T−159]), pOPG-Del3/luc (665 bp [T−159]), and pOPG-Del4/luc (275 bp [T−159]). As shown in Figure 2, construct pOPG-Del1/luc displayed a sufficient transcriptional activity in SaOs-2. Truncation of 85 bp (−1020/−936) resulted in a complete abrogation of transcriptional activity (pOPG-Del2/luc, P<0.0001). Further truncated constructs pOPG-Del3/luc (−936/−677) and pOPG-Del4/luc (−677/−287) did not restore activity. We subsequently introduced the identified SNPs in the context of the MolHaps into both full-length and deletion constructs (Figure 3). In SaOs-2, the activity of the MolHap1 full-length construct was significantly decreased for MolHap2 (OPG-D1H2, P=0.0014), and MolHap4 (OPG-D1H4, P<0.0001). Transcriptional activities were significantly decreased for MolHaps1 and 2, when the −159C allele was introduced into the respective full-length constructs (OPG-Del1–159c, P<0.0001; OPG-D1H2–159c, P=0.0058). Presence of the −159C allele resulted in a nonsignificant decrease of transcriptional activity of MolHap4 (OPG-D1H4–159c, P=0.2749; Figure 3). Introduction of the −159C allele into the truncated constructs (Figure 4) increased transcriptional activity (OPG-Del2–159c, P=0.0005). Similar results were obtained for deletion constructs OPG-Del3 and OPG-Del4 (Figure 4). Transcriptional activities of the full-length and deletion constructs were similar in U2Os cells (supplemental Figure III and IV).
To investigate the transcriptional activity of the three MolHaps separately, we linked the −1020/−772 portion of the human OPG 5′-flanking region to the reporter gene vector ptk81-luc3, which contains a minimal promoter allowing for the assembly of the transcriptional machinery, with only very little remnant activity. In both cell lines, MolHap1 displayed poor transcriptional activity, being as active as the mock vector ptk81-luc3 (supplemental Figure VC). A highly significant decrease of transcriptional activity occurred for MolHap2 (SaOs-2, P=0.0027; U2Os, P<0.0001) and MolHap4 (SaOs-2, P=0.0097; U2Os, P=0.0006). This isolated promoter portion comprising the 3 MolHaps appeared to be poorly involved in the transcriptional regulation of OPG.
SaOs-2 nuclear extracts revealed a specific shift for all 3 MolHaps (Figure 5A, left), being less pronounced for MolHaps2 and 4 (black arrow). An additional weak band was visible for MolHap2 and MolHap4, being competed with the unlabeled PCR fragment (open arrows). Because in silico analyses predicted altered binding sites for Sp1 and NF-1 (supplemental Figure IB), we performed EMSAs using the consensus sites as competitors (Figure 5A, right panel). The prominent band (black arrow) was competed for all 3 MolHaps by NF-1, but not by Sp1 consensus oligos. The additional binding pattern for MolHap2 and MolHap4 (open arrows) was competed by a Sp1 consensus site. Because 5′-truncation of 84 bp (−1020/−936) resulted in significantly decreased transcriptional activity, this portion was also subjected to EMSA (Figure 5B). We detected 3 specific shifts for all 3 MolHaps in SaOs-2, being competed with the homologue unlabeled probe. Whereas 2 of the shifted bands were competed by Sp1 consensus sites (black arrows), 1 was competed by NF-1 consensus sites (white arrow). ChIP assays with specific antibodies against Sp1, Egr1, and NF-1 revealed no binding to the MolHap portion in SaOs-2 (supplemental Figure VIIIA, left).
EMSAs performed with SaOs-2 nuclear extracts resulted in an allele-specific shift for −159T only (Figure 6A, left). Because Sp1 and Egr1 are grouped in “clusters” of TFBS 68 bp upstream of T-159C (supplemental Figure IB), EMSAs were performed using Sp1 and Egr1 consensus sites as competitor. The prominent −159T binding pattern was competed with an Egr1 consensus site, but not with a Sp1 consensus site (Figure 6A, right panel). A recombinant Egr1 binding domain was found to specifically bind to the −159T allele (supplemental Figure VIIB). ChIP assays with specific antibodies against Sp1, Egr1, and NF-1 revealed binding of Sp1 and NF-1 (supplemental Figure VIIIB, left). Similar results were obtained with U2Os nuclear extracts (supplemental Figure VI-VIII).
A detailed description of all results is provided in the supplemental material.
As genome-wide approaches to identify previously unexpected loci associated with complex disease phenotypes still fail to disclose specific functional effects of genetic variation at the proposed locus,22,23 a reliable molecular profiling of genetic variation and a better knowledge of the molecular/functional structure of the locus of interest is mandatory. Our aim was to analyze the transcriptional organization of the (polymorphic) OPG promoter and the impact of previously unpublished individual MolHaps on its activity. We were able to identify 1 novel TSS 225 bp upstream of the ATG codon, and subsequently defined a TSS 64 bp upstream of the ATG codon very close to the major TSS described by Morinaga et al,16 reporting a TSS 67 bp upstream of the ATG codon by primer extension with [γ-32P]dCTP end-labeled primers. Given the very short difference (3 bp) between these 2 major TSS, a technical inaccuracy rather than the actual presence of 2 separate TSS is a likely explanation. By use of diacritic PCR, we further mapped a third TSS, tissue—specifically generated in brain, placenta, spleen, and prostate.24 It is conceivable that transcription of the OPG gene is under control of alternative promoters, activated by tissue-specific TFs. Even if transcription, started at the different TSS of the OPG gene, does not result in an alteration of the amino acid composition of the protein per se, mRNA isoforms with heterogeneous 5′UTR are generated, which may influence transcript stability or translation efficiency.
To define promoter portions being involved in OPG transcriptional activity, we designed serial deletion constructs; these analyses being also useful to mimic the impact of physical inactivation of DNA portions by nuclear histone packaging. Whereas the full-length construct Del1 (−1020/−13) displayed a sufficient transcriptional activity, 5′-truncation of 85 bp (−936/−13) resulted in a complete abrogation of activity. In contrast to haplotypes estimated by maximum likelihood method of inference or other tools, MolHaps are inferred from subcloning of individual DNA providing the allelic constellation in a given DNA strand.17 Introduction of the MolHaps and the proximal T-159C variant in the context of the full-length construct decreased transcriptional activities significantly. Occurrence of the −159C allele in the context of the truncated constructs Del2, Del3, and Del4 revealed a distinct and significant increase of transcriptional activity.
These results suggest the presence of a cis-regulatory element within the truncated 85 bp at position −1020/−936, which interacts with the T-159C site. EMSA experiments revealed an allele-specific binding of Egr1 to the −159T allele. By use of ChIP assays, we demonstrated binding of Egr1 to T-159C in U2Os and Sp1 in SaOs-2 cells; binding of NF-1 was detected in both cell lines. In accordance with our transfection analyses, Egr1 seemed to act as a repressor of OPG transcription in the context of the truncated constructs. This repressing feature was absent in the context of the full-length construct. We suppose that an interaction of Egr1 with a TF bound within the truncated 85 bp (−1020/−936) may be responsible for this observation. To confirm this hypothesis, we performed a “reverse” experiment: the isolated OPG MolHaps (−1020/−772) did not display transcriptional activities over the empty vector, indicating an apparent interaction between the distal and the proximal promoter portion, being indispensable for transcriptional activation.
Cis-active elements exert transcriptional activity by interaction with TFs that assemble at the promoter as modules,25 the composition of an individual module being cell type- and context-specific. For OPG, we propose a hypothetical transcriptional module (Figure 6B): (I) Sp1, NF-1, and NF–κB bind to the MolHap portion, and interact with Egr1 (−159T allele) to cooperatively stimulate transcription; (II) The −159C allele leads to a loss of the TFBS for Egr1, the transcriptional module is altered and transcriptional activity is decreased; (III) Truncation leads to a loss of TFBS for Sp1, NF-1, and NFκB, whereas Egr1 alone seems to act as a repressor; (IV) Introduction of the −159C allele into the truncated constructs results in a partial restoration of transcriptional activity.
Under these modular conditions, genetic variation may gain a pronounced tissue-specific impact on OPG transcript levels. With respect to TF interactions, Egr1 and RelA (p65) cooperatively stimulate NFκB1 (p50) gene transcription,26 Chapman and Perkins27 reported an interaction of RelA (p65) with the zinc-finger domain of Egr1, repressing transcriptional activity of RelA, possibly by sequestering RelA from its target promoter.
Sp1 and Egr1 share a similar binding site and compete, dependent on the physiological state of a cell, for binding at the cognate sites. Egr1 has been reported to act as a repressor of a constitutively expressed collagen gene by preventing interactions between Sp1 and the general transcriptional machinery.28 Members of the NF-1 family play important roles in mammary gland function and development. Sequence analyses led to the identification of NF1/A, NF1/B, NF1/C, NF1/L and NF1/X.29,30 NF1/X, for example, was shown to specifically interact with Sp1 and block Sp1 induction of the PDGF A-chain promoter.31 Therefore, interaction of NF-1 with Sp1 may possibly be responsible for the abrogation of transcriptional activities of MolHap2 and MolHap4, even though in silico analyses predicted NF-1 binding site solely for MolHap4 (putative Sp1 was predicted for MolHap1). As in silico analyses cannot integrate the impact of protein:protein interactions on the recruitment of factors to a transcriptional module, further experiments using RNAi, targeting TFs involved, might be useful to cross-check our present findings and help determine putatively altered binding units.
In conclusion, in the present analysis we propose new structural features and the existence of differentially functional portions, interacting with one another within the OPG promoter region. Further, we identified and functionally and structurally characterized different MolHaps and SNPs within the OPG promoter, which might importantly contribute to an understanding of the molecular basis of interindividually different OPG transcriptional regulatory aspects, including control of OPG tissue expression, plasma levels, and predisposition to CVD phenotypes.
We are grateful to Peter Kleine-Katthöfer (St. Franziskus-Hospital Münster) and Peter Baumgart (Clemenshospital GmbH Münster), supplying genomic DNA for the MolProMD Study.
Sources of Funding
E.B. is supported by a Heisenberg professorship from the Deutsche Forschungsgemeinschaft (Br1589/8-1). We are also grateful to the support by the Else Kröner-Fresenius Foundation (P27/05//A24/05//F01). This study was also supported by a grant from the EU-Project Network of Excellence, FP6-2005-LIFESCIHEALTH-6, Integrating Genomics, Clinical Research and Care in Hypertension, InGenious HyperCare proposal No.: 037093 (to E.B. and S.-M.B.-H.; supported Ralph Telgmann) and an ICT in the FP7-ICT-2007-2, project number 224635, VPH2—Virtual Pathological Heart of the Virtual Physiological Human (to S.-M.B.-H.).
Received February 25, 2009; revision accepted July 23, 2009.
Herrmann SM, Whatling C, Brand E, Nicaud V, Gariepy J, Simon A, Evans A, Ruidavets JB, Arveiler D, Luc G, Tiret L, Henney A, Cambien F. Polymorphisms of the human matrix gla protein (MGP) gene, vascular calcification, and myocardial infarction. Arterioscler Thromb Vasc Biol. 2000; 20: 2386–2393.
Speer MY, McKee MD, Guldberg RE, Liaw L, Yang HY, Tung E, Karsenty G, Giachelli CM. Inactivation of the osteopontin gene enhances vascular calcification of matrix Gla protein-deficient mice: evidence for osteopontin as an inducible inhibitor of vascular calcification in vivo. J Exp Med. 2002; 196: 1047–1055.
Schmidt-Petersen K, Brand E, Telgmann R, Nicaud V, Hagedorn C, Labreuche J, Dördelmann C, Elbaz A, Gautier-Bertrand M, Fischer JW, Evans A, Morrison C, Arveiler D, Stoll M, Amarenco P, Cambien F, Paul M, Brand-Herrmann SM Osteopontin gene variation and cardio/cerebrovascular disease phenotypes. Atherosclerosis. In press.
Sandberg WJ, Yndestad A, Oie E, Smith C, Ueland T, Ovchinnikova O, Robertson AK, Muller F, Semb AG, Scholz H, Andreassen AK, Gullestad L, Damas JK, Froland SS, Hansson GK, Halvorsen B, Aukrust P. Enhanced T-cell expression of RANK ligand in acute coronary syndrome: possible role in plaque destabilization. Arterioscler Thromb Vasc Biol. 2006; 26: 857–863.
Bennett BJ, Scatena M, Kirk EA, Rattazzi M, Varon RM, Averill M, Schwartz SM, Giachelli CM, Rosenfeld ME. Osteoprotegerin inactivation accelerates advanced atherosclerotic lesion progression and calcification in older ApoE-/- mice. Arterioscler Thromb Vasc Biol. 2006; 26: 2117–2124.
Ovchinnikova O, Gylfe A, Bailey L, Nordström A, Rudling M, Jung C, Bergström S, Waldenström A, Hansson GK, Nordström P. Osteoprotegerin promotes fibrous cap formation in atherosclerotic lesions of apoE-deficient mice. Arterioscler Thromb Vasc Biol. In press
Kiechl S, Schett G, Wenning G, Redlich K, Oberhollenzer M, Mayr A, Santer P, Smolen J, Poewe W, Willeit J. Osteoprotegerin is a risk factor for progressive atherosclerosis and cardiovascular disease. Circulation. 2004; 109: 2175–2180.
Semb AG, Ueland T, Aukrust P, Wareham NJ, Luben R, Gullestad L, Kastelein JJ, Khaw KT, Boekholdt SM. Osteoprotegerin and soluble receptor activator of nuclear factor-kappaB ligand and risk for coronary events: a nested case-control approach in the prospective EPIC-Norfolk population study 1993–2003. Arterioscler Thromb Vasc Biol. 2009; 29: 975–680.
Zanke BW, Greenwood CM, Rangrej J, Kustra R, Tenesa A, Farrington SM, Prendergast J, Olschwang S, Chiang T, Crowdy E, Ferretti V, Laflamme P, Sundararajan S, Roumy S, Olivier JF, Robidoux F, Sladek R, Montpetit A, Campbell P, Bezieau S, O'Shea AM, Zogopoulos G, Cotterchio M, Newcomb P, McLaughlin J, Younghusband B, Green R, Green J, Porteous ME, Campbell H, Blanche H, Sahbatou M, Tubacher E, Bonaiti-Pellié C, Buecher B, Riboli E, Kury S, Chanock SJ, Potter J, Thomas G, Gallinger S, Hudson TJ, Dunlop MG. Genome-wide association scan identifies a colorectal cancer susceptibility locus on chromosome 8q24. Nat Genet. 2007; 39: 989–894.
Yeager M, Orr N, Hayes RB, Jacobs KB, Kraft P, Wacholder S, Minichiello MJ, Fearnhead P, Yu K, Chatterjee N, Wang Z, Welch R, Staats BJ, Calle EE, Feigelson HS, Thun MJ, Rodriguez C, Albanes D, Virtamo J, Weinstein S, Schumacher FR, Giovannucci E, Willett WC, Cancel-Tassin G, Cussenot O, Valeri A, Andriole GL, Gelmann EP, Tucker M, Gerhard DS, Fraumeni JF Jr, Hoover R, Hunter DJ, Chanock SJ, Thomas G. Genome-wide association study of prostate cancer identifies a second risk locus at 8q24. Nat Genet. 2007; 39: 645–649.
Dördelmann C, Telgmann R, Brand E, Hagedorn C, Schröer B, Hasenkamp S, Baumgart P, Kleine-Katthöfer P, Paul M, Brand-Herrmann SM. Functional and structural profiling of the human thrombopoietin gene promoter. J Biol Chem. 2008; 283: 24382–24391.
Grabe N. AliBaba2: context specific identification of transcription factor binding sites. In Silico Biol. 2001; 2: S1–15.
Frohman MA, Dush MK, Martin GR. Rapid production of full-length cDNAs from rare transcripts: amplification using a single gene-specific oligonucleotide primer. Proc Natl Acad Sci U S A. 1988; 85: 8998–9002.
Liu S, Spinner DS, Schmidt MM, Danielsson JA, Wang S, Schmidt J. Interaction of MyoD family proteins with enhancers of acetylcholine receptor subunit genes in vivo. J Biol Chem. 2000; 275: 41364–41368.
Schreiber E, Matthias P, Muller MM, Schaffner W. Rapid detection of octamer binding proteins with ‘mini-extracts’, prepared from a small number of cells. Nucleic Acids Res. 1989; 17: 6419.
McPherson R, Pertsemlidis A, Kavaslar N, Stewart A, Roberts R, Cox DR, Hinds DA, Pennacchio LA, Tybjaerg-Hansen A, Folsom AR, Boerwinkle E, Hobbs HH, Cohen JC. A common allele on chromosome 9 associated with coronary heart disease. Science. 2007; 316: 1488–1491.
Yasuda H, Shima N, Nakagawa N, Mochizuki SI, Yano K, Fujise N, Sato Y, Goto M, Yamaguchi K, Kuriyama M, Tsuda E, Morinaga T, Higashio K. Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): a mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology. 1998; 139: 1329–1337.
Telgmann R, Dördelmann C, Brand E, Nicaud V, Hagedorn C, Pavenstädt H, Cambien F, Tiret L, Paul M, Brand-Herrmann SM. Molecular genetic analysis of a human insulin-like growth factor 1 promoter P1 variation. FASEB J. 2009; 23: 1303–1313.
Cogswell PC, Mayo MW, Baldwin AS Jr. Involvement of Egr-1/RelA synergy in distinguishing T cell activation from tumor necrosis factor-alpha-induced NF-kappa B1 transcription. J Exp Med. 1997; 185: 491–497.
Chapman NR, Perkins ND. Inhibition of the RelA(p65) NF-kappaB subunit by Egr-1. J Biol Chem. 2000; 275: 4719–4725.
Tan L, Peng H, Osaki M, Choy BK, Auron PE, Sandell LJ, Goldring MB. Egr-1 mediates transcriptional repression of COL2A1 promoter activity by interleukin-1beta. J Biol Chem. 2003; 278: 17688–17700.
Rupp RA, Kruse U, Multhaup G, Gobel U, Beyreuther K, Sippel AE. Chicken NFI/TGGCA proteins are encoded by at least three independent genes: NFI-A, NFI-B and NFI-C with homologues in mammalian genomes. Nucleic Acids Res. 1990; 18: 2607–2616.
Kruse U, Qian F, Sippel AE. Identification of a fourth nuclear factor I gene in chicken by cDNA cloning: NFI-X. Nucleic Acids Res. 1991; 19: 6641.