The Ets1 Proto-Oncogene Is Upregulated by Retinoic Acid: Characterization of a Functional Retinoic Acid Response Element in the Ets1 Promoter

Oncogene (2000) 19, 1969 ± 1974 ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc The Ets1 proto-oncogene is upregulated by retinoic acid: characterization of a functional retinoic acid response element in the Ets1 promoter

Afshin Raouf1, Vincent Li1, Ismail Kola2, Dennis K Watson3 and Arun Seth*,1

1Department of Laboratory Medicine and Pathobiology, MRC group in Periodontal Physiology, University of Toronto, and Sunnybrook and Women's College Health Sciences Centre, Toronto, Ontario, ON, Canada; 2Molecular Genetics and Development Group, Monash University, Melbourne, Australia; 3Center for Molecular and Structural Biology, Hollings Cancer Center, Medical University of South Carolina, Charleston, South Carolina, SC, USA

The v-ets oncoprotein and its progenitor Ets1 belong to a and cytokines (Janknecht and Nordheim, 1993; Seth et family of transcription factors that are related by an 85 al., 1992; Sharrocks et al., 1997; Wasylyk et al., 1993). amino acid conserved DNA binding domain, the ets Ets genes have been shown to transform NIH3T3 cells domain. Ets1 plays important role(s) in control of cell and induce tumor formation in nude mice (Hart et al., proliferation, dierentiation and apoptosis. Abnormal 1995; Seth and Papas, 1990; Seth et al., 1989). expression of Ets1 could lead to disruption of these Similar to the human ETS1 5' ¯anking sequence, the processes and contribute to development of malignancy. mouse Ets1 gene promoter contains a GC-rich region Retinoic acid (RA) inhibits proliferation, induces dier- upstream of multiple mRNA initiation sites and lacks entiation and regulates apoptosis in many dierent cell both TATA sequences and a CAAT box (Jorcyk et al., types. Here, we demonstrate that RA treatment increases 1991, 1997; Majerus et al., 1992; Oka et al., 1991). We the expression of Ets1 mRNA, but not that of Ets2, Elk1 have identi®ed consensus recognition sites for several or Fli1 in MC3T3-E1 cells. Ets1 induction is detectable transcription factors that may regulate the mouse Ets1 after 4 h, can be maintained for at least 14 days, and is promoter, including AP1, RARs, Ets1 and NFkB inhibited by Actinomycin D, which suggests that RA (Jorcyk et al., 1997). Several other agents that are regulation of Ets1 occurs at the transcriptional level. The known to upregulate Ets1 transcription, include RA, promoter region of Ets1 contains four retinoic acid TNF-alpha, VEGF, and TPA (Chen et al., 1997; Gilles response element (RARE) half sites located at 794, et al., 1996; Wang et al., 1997). 7152, 71765 and 72252 from the translation start site. Ets1 gene can be activated by several mechanisms to We show that RARb is expressed by MC3T3-E1 cells in become an ecient transforming gene. For example, in the presence of RA and demonstrate that it binds to the MuLV-induced rat T-cell lymphomas, Ets1 expression 794 RARE half site. Furthermore, RA induces transcrip- is upregulated due to the proviral insertion upstream of tion of Ets1 promoter-reporter constructs containing this the ®rst exon of the gene (Bellacosa et al., 1994). In RARE half site. Oncogene (2000) 19, 1969 ± 1974. contrast, the ectopic expression of the ETS1 gene in human colon cancer cell lines reduced the rate of the Keywords: Ets1 promoter; retinoic acid; osteoblast; anchorage-independent growth in a dose dependent transcription factor manner (Suzuki et al., 1995). It has been shown that RA reduces the anchorage-independent growth of the human colon cancer cells (Hoosein et al., 1988; Niles et Introduction al., 1988) suggesting that RA may reduce the tumorigenicity of human colon cancer cells through The ets oncogene (v-ets) was originally discovered as upregulation of ETS1 expression. part of a fusion protein (gag-myb-ets) expressed by an Retinoic acid is a natural morphogen and plays a key avian transforming retrovirus, E-26, which transforms role(s) in development and dierentiation of various cell myeloblasts and erythroblasts in vitro and causes mixed types (Glass et al., 1990). RA exerts its biological erythroid-myeloid leukemia in vivo (Seth et al., 1992; response through interaction with nuclear retinoic acid Sharrocks et al., 1997; Watson et al., 1990). V-ets receptors, which belong to the steroid hormone receptor contains an 85 amino acid sequence, the ets domain, super family of transcription factors (Linney, 1992). which has been also identi®ed in various cellular There are two families of RA receptors: the retinoid X proteins including Ets1, Ets2, Elk1, Fli1 and others receptors (RXR) and the retinoic acid receptors (RAR) (Seth et al., 1992; Sharrocks et al., 1997). The ets (Lohnes et al., 1994). Both RARs and RXRs consist of family genes encode transcription factors, which three receptor subtypes, alpha, beta and gamma, which regulate gene expression by interacting with speci®c are ligand-inducible transcription factors (Mangelsdorf purine-rich core sequences (C/AGGAA/T) that have et al., 1992). Once activated, these receptors regulate been found in a large number of cellular and viral transcription by interacting with RA response elements enhancers and promoters including proto-oncogenes Pu-G-T/G-T-C-A, directly repeated 1 ± 5 bp apart in the promoter of target genes (Leid et al., 1992). Although the mechanism(s) by which RA is able to regulate the function of osteoblasts is not as yet fully *Correspondence: A Seth ascertained, the RA mediated induction of at least one We dedicate this manuscript to the memory of Takis S Papas, a kind bone related gene product, parathyroid hormone- andgenerouscolleagueandfriend,whowasknownforhis pioneering work on the Ets family genes related peptide, has been shown to involve interaction Received 22 November 1999; revised 31 January 2000; accepted 31 via a functional binding site, for the transcription January 2000 factor Ets1 (Karperien et al., 1997). Moreover, Ets1 is Ets1 is upregulated by retinoic acid A Raouf et al 1970 highly expressed in bone during bone formation and To study the role of Ets1 in bone development, we remodeling and is associated with tissue remodeling have examined the mechanism of its activation by and branching morphogenesis, processes that involve retinoic acid in the osteoblastic MC3T3-E1 cell line. the degradation and remodeling of the extracellular Here, we demonstrate that all-trans-retinoic acid matrix (Karperien et al., 1997; Kola et al., 1993; (atRA) induces the expression of Ets1 mRNA by 19- Maroulakou et al., 1994). fold in 48 h in osteoblast-like cells and that the Ets1 Ets1 has been shown to interact with the `quintes- promoter contains a functional RAR responsive sential' osteoblast transcription factor CbfA1 (Ducy et element. al., 1997; Kim et al., 1999; Sun et al., 1995). Thus the genetic programs regulated by Ets1 may signi®cantly aect the development and dierentiation of osteo- Results blasts. We have shown that Ets1 is induced by the potent bone morphogen retinoic acid in dierentiating Ets1 mRNA expression is induced by retinoic acid in P19 embryonic carcinoma cells (Kola et al., 1993). MC3T3-E1 cells Elucidating regulatory mechanisms for direct interaction of RA with the Ets1 promoter would demonstrate In order to investigate the eects of retinoic acid on that Ets1 can mediate retinoic acid eects. Ets1 expression in osteoblast like cells, MC3T3-E1

Figure 1 Retinoic Acid induces Ets1 mRNA. MC3T3-E1 cells were treated with or without 1 mM all-trans-retinoic acid, and 10 mg of total RNA was examined at the indicated times. Expression of Ets1 mRNA was detected by Northern blot hybridization and quanti®ed relative to GAPDH expression (see Materials and methods). (a) Induction of Ets1 transcript by atRA for up to 48 h. The Northern blots used for 0.5 ± 4 h atRA treatment were exposed for 3 days. The Northern blots used for 24 ± 48 h treatment were exposed overnight to the X-ray ®lm. (b) Long term eects of atRA on Ets1 expression. Lanes 1 ± 6; untreated controls, lanes 7 ± 12; 1 mM all-trans-RA. The Northern blot was exposed overnight to the X-ray ®lm. (c) atRA is necessary for sustained induction of Ets1. MC3T3-E1 cells grown in medium with 1 mM atRA for 48 h and then in atRA free medium for another 24 or 48 h

Oncogene Ets1 is upregulated by retinoic acid A Raouf et al 1971 cultures were treated with all-trans-retinoic acid (atRA) atRA treatment (Figure 4a,b). However, we found that for 0.5 ± 48 h and the expression of Ets1 transcript was the expression of RARg is moderately enhanced in the determined by Northern blot analysis (Figure 1a). presence of atRA, whereas RXRg is undetectable Relative to GAPDH, we observed a 3.4-fold increase in (Figure 4b). the Ets1 mRNA levels after 4 h of treatment with atRA and 19-fold after 48 h. This induction of Ets1 The Ets1 promoter has a functional RA responsive mRNA can be maintained for at least 14 days (Figure element 1b). In the absence of atRA at the level of Ets1 transcript was substantially reduced over 24 h and Our results showed that RA regulates Ets1 gene returned to basal levels after 48 h (Figure 1c) expression at the level of transcription. Therefore, we suggesting that atRA is required for the sustained analysed 2.1 kb 5' ¯anking region of the mouse Ets1 induction of Ets1 mRNA levels. Interestingly, the gene promoter for consensus retinoic acid response- mRNA expression of Ets2 (Figure 2), Fli1 and Elk1 elements. Interestingly, we found four half-sites located (data not shown) transcription factors was not induced at 794 (RARE I), 7152 (RARE II), 71765 (RARE by atRA. III), and 72252 (RARE IV) base pairs up stream of the translation start (Figure 5a). Other regulatory elements that we identi®ed within the Ets1 promoter Actinomycin D inhibits atRA mediated induction of Ets1 include the consensus binding sites for AP1, AP2, SP1 mRNA expression and an EBS. The proximal RARE I site is conserved It has been shown that the retinoic acid regulation of between human, rat and mouse Ets1 promoters, gene expression occurs at both transcriptional and suggesting that this site may be important for RA- post-transcriptional levels (Inoue et al., 1996; LaRosa mediated regulation of this gene. and Gudas, 1998; Mahadevan and Edwards, 1991). In To determine whether retinoic acid receptors directly order to investigate the nature of induction of Ets1 bind Ets1 RARE sites, electrophoretic mobility shift mRNA levels by atRA, MC3T3-E1 cells were treated assays (EMSA) were performed with in vitro expressed with a transcription initiation inhibitor, actinomycin D RARb protein and 32P-labeled oligonucleotide contain- (Act D) and a protein synthesis inhibitor, cycloheximide (CHX) subsequent to a 24 h atRA administra- tion. Treatment of cells with CHX for 4 h increased Ets1 mRNA levels. This superinduction is not unique to Ets1, as CHX treatment also induced the Ets2 gene (Figure 2). Cycloheximide is known to increase the mRNA levels of other transcription factors such as c- myc,c-fos, and egr-1 by stabilizing their message (Greenberg et al., 1986; Suva et al., 1991). In contrast, actinomycin D inhibited the induction of Ets1 mRNA by atRA yet had no eect on CHX stabilization of Ets1 mRNA for up to 4 h (Figure 3).

Figure 2 Cycloheximide (CHX) superinduces Ets1 and Ets2 RA receptor expression profile in the MC3T3-E1 cells mRNA. MC3T3-E1 cells were pretreated with 1 mM all-trans- retinoic acid for 24 h and then treated with 1 mg/ml CHX in the To determine the expression status of the RA receptors presence of atRA as indicated above lanes. Ten mg of total RNA in the MC3T3-E1 cells, we treated the culture with was analysed at indicated times and the same blot was hybridized atRA for various times and the expression of RARs serially with probes for Ets1, Ets2, and GAPDH. Lanes 1, no and RXRs was examined by Northern blot analysis treatment; 2, 24 h atRA treatment; 3, 25 h treatment with atRA; 4, 1 h treatment with CHX in presence of atRA; 5, 28 h treatment (Figure 4). Our results indicate that RARb expression with atRA; 6, 4 h treatment with CHX in presence of atRA; 7, is induced by atRA and that RXRa and RXRb 48 h treatment with atRA; 8, 24 h treatment with CHX in transcript levels are not signi®cantly altered following presence of atRA

Figure 3 Transcriptional activation of Ets1 by atRA. MC3T3-E1 cells were treated with either 1 mM all-trans-retinoic acid for 24 h or with CHX (1 mg/ml) for 4 h prior to treatment with Act D (2 mg/ml). Total RNA was extracted on the indicated times after actinomycin D treatment, Northern blotted and hybridized with a probe for Ets1, stripped, and reprobed for GAPDH expression

Oncogene Ets1 is upregulated by retinoic acid A Raouf et al 1972 ing the RARE I site. The results in Figure 5b show transfected with the control vector lacking Ets1 that RARb form a complex with RARE I and excess promoter sequences did not alter the expression of unlabeled RARE I oligonucleotide completely the reporter gene. abolishes the protein : DNA complex formation, de- monstrating the speci®city of this interaction. In contrast, neither a mutant oligonucleotide, nor any of the other RARE sites (RARE II, III or IV) were able to signi®cantly compete with this binding site, suggesting that RARb binds speci®cally to the RARE I site in the Ets1 promoter (Figure 5b). DNA binding with 32P- labeled oligonucleotides corresponding to other RARE sites (II, III and IV) was also tested. However, no protein : DNA complexes were detected with RARb, RARg, or RXRa (data not shown) which further indicates that only the RARE I binds to the retinoic acid receptor proteins. To demonstrate the functionality of RARE I site we constructed three Ets1-Luciferase promoter constructs, pLucE1P2.1, and pLucE1P0.7 and pLucE1P1.5 (Figure 6). pLucE1P2.1 contains all four consensus RARE sites and pLucE1P0.7 contains the two proximal RAREs (RARE I and RARE II). pLucE1P1.5 lacks the proximal RAREs but does contain the distal RAREs (RARE III and RARE IV). Retinoic acid treatment of transfected MC3T3-E1 cells increased the transcriptional activity of pLucE1P2.1 and pLucE1P0.7 by 2- and 2.3-fold respectively and did not aect the activity of pLucE1P1.5 (Figure 6). atRA treatment of cells

Figure 5 RARb binds to the RARE I (at 794) in the Ets1 promoter. (a) The promoter region of the mouse Ets1 promoter contains four consensus sites RARE I (GGTTCA), RARE II (CGGTCA), RARE III (TGGTCA), RARE IV (TGGTCA). (b) Gel mobility shift assay using in vitro expressed RARb protein and labeled RARE I oligonucleotide (cgcgtctgtccccttccactgGGTT- CAaaaatccccatt). Lane 1, no competitor. Unlabeled competitors: Lane 2, RARE I. Lane 3, Mutant RARE I (cgcgtctgtccccttcca- ctgGGTATAaaaatccccatt) Lane 4, RARE II (agctcactgatGGT- CAttggtg), Lane 5 RARE III (agctttcctctGGTCAcccaag), Lane 6 RARE IV (agctcgcgggcGGTCAgcggga). Arrow indicates the location of the protein : DNA complex. The RARE consensus sequences are shown in bold. The Ets1 mouse promoter sequence b has been deposited in the GenBank database (accession number AF221504)

Figure 6 RA induces the transcriptional activity of the Ets1 Figure 4 Expression of RARs and RXRs. Northern blots of promoter. MC3T3-E1 cells were transfected with the luciferase total RNA extracted at various days from MC3T3-E1 cultures constructs pLucE1P2.1 and pLucE1P0.7 and pLucE1P1.5, grown with or without all-trans-retinoic acid as indicated above containing various Ets1 promoter fragments as described in the each lane. The blots were probed with labeled full-length cDNAs Materials and methods section. Open rectangle in pLucE1P1.5 corresponding to RA receptors as indicated. Ethidium bromide represents the TATA box. The expression of the reporter gene staining of gels is shown below each autoradiograph. (a) RARb was normalized relative to total protein content and the fold expression in the absence or presence of atRA. (b) Expression of induction was calculated relative to untreated cells. Results of RARg and RXRs in the absence or presence of at RA three experiments are presented

Oncogene Ets1 is upregulated by retinoic acid A Raouf et al 1973 Discussion by loss of transcriptional repressors and/or mRNA degrading enzymes. RA is a natural morphogen, which can function as a The multiple eects of RA on cell function suggests dierentiating agent in many cell types. For example, the existence of regulatory interactions with other ETS1 mRNA expression is upregulated during RA factors, such as Ets proteins, which may be key induced dierentiation of both the human neuroblas- mediators of its outcomes. toma cell line, LA-N-5 and the embryonic carcinoma For example, the RA induction of PTHrP, a bone cell line P19 (Kola et al., 1993; Thiele et al., 1988). In related gene, which lacks RAREs, has been shown to this study we demonstrated that atRA increases the be mediated by a functional Ets1 binding site in its expression of Ets1 mRNA by 19-fold in 48 h in the promoter (Karperien et al., 1997). Thus the mechan- MC3T3-E1 cells. The inhibition of this induction with isms by which RA regulates gene expression may actinomycin D suggests that RA regulation of Ets1 include the activation of Ets1 and its target genes. expression occurs at the transcriptional level. The transcriptional regulation of Ets1 by RA is further demonstrated through EMSA and transient transfection assays. Through gel retardation assays we have Materials and methods identi®ed a RA-responsive element (RARE I) which is located 94 bp upstream of the translational start site, Cell culture which forms protein : DNA complexes with RARb, RARg, and RXRa. However, RARE II, III, and IV MC3T3-E1 cells were maintained in alpha modi®ed minimal essential medium supplemented with 10% fetal calf serum failed to form protein : DNA complexes. (FCS, Gibco ± BRL) and 10% antibiotic mix. Cultures were Similar to our ®ndings with Ets1 promoter, Suva et maintained at 378C in fully humidi®ed atmosphere of 5% al. (1991, 1994) have shown that induction of egr1 CO2 and 95% oxygen. For the purpose of this study, cells promoter by RA requires only a single half-site, and were plated at a density of 2.56104 cell/cm2. Once 80% that this half site is recognized by all three RARs. con¯uent, cells were treated with 2 or 10% FCS and either Transient transfection of MC3T3-E1 cells revealed 1 mM all-trans-retinoic acid (atRA, Sigma) or 2 mgof that RA induces the transcriptional activity of Actinomycin D (dactinomycin, Sigma) per ml or 1 mgof pGL3E1P0.7 plasmid containing *700 bp of the cycloheximide (Sigma) per ml or DMSO (Sigma). To study identi®ed Ets1 regulatory region placed in front of the eects of Act D and CHX on the Ets1 expression the luciferase gene. Our transient transfection and regulation by atRA, MC3T3-E1 cells were treated with all- trans-RA for 24 h prior to addition of CHX or Act D to the EMSA results demonstrate that RARE I is a culture media. functional RA-responsive element, mediating the regulatory eect of RA on the Ets1 gene expression. The RA activation of the Ets1 promoter shown RNA isolation, Northern blot analysis and quantification through our transient transfection assays is signi®- Total cellular RNA was isolated by guanidinium isothiocya- cantly less than the mRNA induction of the nate and phenol extraction as described (Chomczynski and endogenous Ets1 gene detected by Northern blots. Sacchi, 1987). Total RNA (7 ± 15 mg) was size fractionated This dierence could be due to the presence of using a 1.2% agarose gel containing 2% formaldehyde and additional RA-responsive motifs present outside of transferred to a nylon membrane using capillary transfer the 2.1 kb promoter region analysed here. Previous method. Northern blot analysis was carried out using probes speci®c to Ets1, Ets2, Elk1, Fli1. Probes (26106 c.p.m./ml) studies have demonstrated that the ®rst intron of Ets1 were made using random priming kit (Amersham-Pharmacia contains elements necessary for tissue speci®c expres- Biotech) and hybridization was carried out at 428C over sion (Jorcyk et al., 1997). However, our data suggests night. The blots were washed at 658C for total of 1 h and that induction of Ets1 gene transcription by RA is exposed to Kodak X-ray ®lm. mediated in part through direct interaction between The autoradiographs were scanned on a HP Scan Jet RARs with RARE I in the Ets1 promoter. Induction 5200C and bands were quantitated by the IP Lab Gel of Ets1 mRNA by atRA is maximal after 48 h Scienti®c Image Processing program (Signal Analytics). Ets1 suggesting that de novo protein synthesis may be bands were normalized to their respective GAPDH bands. required for initiation of transcription. In addition, cycloheximide treatment resulted in superinduction of Electrophoretic mobility shift assay Ets1 mRNA, as similarly shown for transcription RA receptor proteins were made using an in vitro factors, egr1 and c-myc (Davido and Mendelow, transcription and translation kit, according to the manufac- 1994; Lau and Nathans, 1987; Suva et al., 1991). turer's protocol (TNT kit, Promega). Retinoic acid receptor Previously it was reported that the Ets2 mRNA and expression vectors were generous donations of Dr TM not the Ets1 mRNA was superinduced in presence of Underhill (University of Western Ontario, Canada). To CHX in regenerating liver (Bhat et al., 1987). Here we analyse protein : DNA interactions, synthetic oligonucleotides show that both Ets1 and Ets2 mRNAs are super- corresponding to RARE I were annealed and labeled with induced by CHX, which may simply be due to cell- a32P-dCTP and gel puri®ed (Mandel). Oligonucleotides speci®c dierences similar to those previously demon- corresponding to other Ets1 RARE half-sites located at strated for c-myc (Greenberg et al., 1986). In mouse 794, 7152, 71765 and 72252 were used as competitors. 5 ml of freshly in vitro synthesized protein was allowed to ®broblastic cell line BALB/c-3T3, c-myc mRNA is bind to oligonucleotide probes in the presence or absence of superinduced by CHX whereas in the rat pheochro- competitors for 20 min on ice (106binding buer: 200 mM mocytoma cell line PC12, CHX has no eect on the c- Tris-HCl pH 7.5, 500 mM NaCl, 10 mM MgCl2 2mM EDTA myc mRNA expression (Greenberg et al., 1986). and 5 mM DTT). Subsequently protein : DNA complexes Superinduction of Ets1 mRNA by CHX is likely were analysed through a non-denaturing 4% polyacrylamide due to the enhanced mRNA stability, brought about gel containing 0.46TBE. The polyacrylamide gel was ®xed

Oncogene Ets1 is upregulated by retinoic acid A Raouf et al 1974 with 10% acetic acid, 20% methanol, 5% glycerol and 65% pLucE1P1.5 or the empty pGL3 vector using 5 ml lipofecta- water, dried and exposed to Kodak X-ray ®lm. mine agent according to the manufacturer's protocol (Gibco ± BRL). Cells were incubated in media containing DNA-lipofectamine complexes for 5 h and immediately after Plasmid construction and transfections transfection, cells were treated with either control medium or To study the eects of RA on the Ets1 transactivation, three medium supplemented with 1 mM all-trans-RA and 2% or Ets1 promoter constructs were made. pLucE1P2.1 construct 10% FCS for 48 h. Cell extracts were prepared using was generated by inserting a Fok1-Fok1 (2.1 kb) fragment of luciferase detection kit according to the manufacturer's the Ets1 promoter isolated from Lambda DBA1 (Jorcyk et protocol (Promega). Luciferase activity was determined with al., 1997) at the BglII site of the promoterless luciferase a luminometer and normalized relative to total protein reporter gene in the plasmid pGL3-basic (Promega). content as measured by Bradford assay (BioRad). The pLucE1P0.7 construct was made by HindIII digestion of transfection experiments were repeated three times. the pLucE1P2.1 and inserting the 692 bp fragment at the HindIII site of the pGL3-basic vector. PGL-3 plasmid contains no minimal promoter. pLucE1P1.5 containing RAREs III and IV was constructed by inserting the 1.5 kb HindIII digested Ets1 promoter fragment into a pGL3 Acknowledgments derivative that contains a minimal promoter. MC3T3-E1 This work was supported by a Medical Research Council cells were transfected with 2 mg of pLucE1P2.1, pLucE1P0.7, (Canada) group grant to A Seth.

References

Bellacosa A, Datta K, Bear SE, Patriotis C, Lazo PA, Lau LF and Nathans D. (1987). Proc. Natl. Acad. Sci. USA, Copeland NG, Jenkins NA and Tsichlis PN. (1994). J. 84, 1182 ± 1186. Virol., 68, 2320 ± 2330. Leid M, Kastner P and Chambon P. (1992). Trends Biochem. Bhat NK, Fisher RJ, Fujiwara S, Ascione R and Papas TS. Sci., 17, 427 ± 433. (1987). Proc. Natl. Acad. Sci. USA, 84, 3161 ± 3165. Linney E. (1992). Curr. Top. Dev. Biol., 27, 309 ± 350. Chen Z, Fisher RJ, Riggs CW, Rhim JS and Lautenberger Lohnes D, Mark M, Mendelsohn C, Dolle P, Dierich A, JA. (1997). Cancer Res., 57, 2013 ± 2019. Gorry P, Gansmuller A and Chambon P. (1994). Chomczynski P and Sacchi N. (1987). Anal. Biochem., 162, Development, 120, 2723 ± 2748. 156 ± 159. Mahadevan LC and Edwards DR. (1991). Nature, 349, 747 ± Davido AN and Mendelow BV. (1994). Anticancer Res., 14, 748. 1199 ± 1201. Majerus MA, Bibollet-Ruche F, Telliez JB, Wasylyk B and Ducy P, Zhang R, Georoy V, Ridall AL and Karsenty G. Bailleul B. (1992). Nucleic Acids Res., 20, 2699 ± 2703. (1997). Cell, 89, 747 ± 754. Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong Gilles F, Raes MB, Stehelin D, Vandenbunder B and Fafeur ES, Oro AE, Kakizuka A and Evans RM. (1992). Genes V. (1996). Exp. Cell. Res., 222, 370 ± 378. Dev., 6, 329 ± 344. Glass CK, Devary OV and Rosenfeld MG. (1990). Cell, 63, Maroulakou IG, Papas TS and Green JE. (1994). Oncogene, 729 ± 738. 9, 1551 ± 1565. Greenberg ME, Hermanowski AL and Zi EB. (1986). Mol. Niles RM, Wilhelm SA, Thomas P and Zamcheck N. (1988). Cell. Biol., 6, 1050 ± 1057. Cancer Invest., 6, 39 ± 45. HartAH,CorrickCM,TymmsMJ,HertzogPJandKolaI. Oka T, Rairkar A and Chen JH. (1991). Oncogene, 6, 2077 ± (1995). Oncogene, 10, 1423 ± 1430. 2083. Hoosein NM, Brattain DE, McKnight MK, Childress KE, Seth A, Ascione R, Fisher RJ, Mavrothalassitis GJ, Bhat NK Chakrabarty S and Brattain MG. (1988). Cancer Lett., 40, and Papas TS. (1992). Cell Growth Dier., 3, 327 ± 334. 219 ± 232. Seth A and Papas TS. (1990). Oncogene, 5, 1761 ± 1767. Inoue A, Otsuka E, Hiruma Y, Hirose S, Furuya M, Tanaka Seth A, Watson DK, Blair DG and Papas TS. (1989). Proc. S and Hagiwara H. (1996). Biochem. Biophys. Res. Natl. Acad. Sci. USA, 86, 7833 ± 7837. Commun., 228, 182 ± 186. Sharrocks AD, Brown AL, Ling Y and Yates PR. (1997). Int. Janknecht R and Nordheim A. (1993). Biochim. Biophys. J. Biochem. Cell Biol., 29, 1371 ± 1387. Acta, 1155, 346 ± 356. Sun W, Graves BJ and Speck NA. (1995). J. Virol., 69, Jorcyk CL, Garrett LJ, Maroulakou IG, Watson DK and 4941 ± 4949. Green JE. (1997). Cell. Mol. Biol., (Noisy-le-grand) 43, Suva LJ, Ernst M and Rodan GA. (1991). Mol. Cell. Biol., 211 ± 225. 11, 2503 ± 2510. Jorcyk CL, Watson DK, Mavrothalassitis GJ and Papas TS. Suva LJ, Towler DA, Harada S, Gaub MP and Rodan GA. (1991). Oncogene, 6, 523 ± 532. (1994). Mol. Endocrinol., 8, 1507 ± 1520. Karperien M, Farih-Sips H, Lowik CW, de Laat SW, Suzuki H, Romano-Spica V, Papas TS and Bhat NK. (1995). Boonstra J and De®ze LH. (1997). Mol. Endocrinol., 11, Proc. Natl. Acad. Sci. USA, 92, 4442 ± 4446. 1434 ± 1448. Thiele CJ, Deutsch LA and Israel MA. (1988). Oncogene, 3, Kim WY, Sieweke M, Ogawa E, Wee HJ, Englmeier U, Graf 281 ± 288. T and Ito Y. (1999). EMBO J., 18, 1609 ± 1620. Wang DY, Yang VC and Chen JK. (1997). In Vitro Cell. Dev. Kola I, Brookes S, Green AR, Garber R, Tymms M, Papas Biol. Anim., 33, 248 ± 255. TS and Seth A. (1993). Proc. Natl. Acad. Sci. USA, 90, Wasylyk B, Hahn SL and Giovane A. (1993). Eur. J. 7588 ± 7592. Biochem., 211, 7±18. LaRosa GJ and Gudas LJ. (1988). Mol. Cell. Biol., 8, 3906 ± Watson DK, Ascione R and Papas TS. (1990). Crit. Rev. 3917. Oncog., 1, 409 ± 436.

Oncogene