Differential Activation of the Rat Phenylethanolamine N-Methyltransferase Gene by Sp1 and Egr-1статья из журнала
Аннотация: The rat phenylethanolamine N-methyltransferase (PNMT) gene contains overlapping consensus elements for the Sp1 and Egr-1 transcription factors located at −45 bp and −165 bp in the PNMT promoter. In the present study, we show that Sp1 and Egr-1 can specifically bind to these overlapping elements, that this binding appears to be mutually exclusive, and that binding site occupancy is dependent upon the concentration of each factor and its binding affinity for each site. Egr-1 binds to the −165 bp site with relatively high affinity (IC50 = 14 nM) and to the −45 bp site with relatively low affinity (IC50 = 1360 nM), whereas Sp1 binds to both sites with intermediate affinities (IC50 = 210 and 140 nM, respectively). Consistent with the DNA-binding data, Egr-1 stimulates PNMT promoter activity primarily through interaction with the −165 bp site, while Sp1 stimulates PNMT promoter activity by interacting with both the −45 bp and the −165 bp sites. These results show that Sp1 and Egr-1 are capable of differentially activating PNMT gene expression, thereby suggesting that different stimuli may control the activity of the PNMT gene by selectively regulating Sp1 and/or Egr-1. The rat phenylethanolamine N-methyltransferase (PNMT) gene contains overlapping consensus elements for the Sp1 and Egr-1 transcription factors located at −45 bp and −165 bp in the PNMT promoter. In the present study, we show that Sp1 and Egr-1 can specifically bind to these overlapping elements, that this binding appears to be mutually exclusive, and that binding site occupancy is dependent upon the concentration of each factor and its binding affinity for each site. Egr-1 binds to the −165 bp site with relatively high affinity (IC50 = 14 nM) and to the −45 bp site with relatively low affinity (IC50 = 1360 nM), whereas Sp1 binds to both sites with intermediate affinities (IC50 = 210 and 140 nM, respectively). Consistent with the DNA-binding data, Egr-1 stimulates PNMT promoter activity primarily through interaction with the −165 bp site, while Sp1 stimulates PNMT promoter activity by interacting with both the −45 bp and the −165 bp sites. These results show that Sp1 and Egr-1 are capable of differentially activating PNMT gene expression, thereby suggesting that different stimuli may control the activity of the PNMT gene by selectively regulating Sp1 and/or Egr-1. Phenylethanolamine N-methyltransferase (PNMT)1 1The abbreviations used are: PNMTphenylethanolamine N-methyltransferasefpufootprinting unit(s). catalyzes the final step in the catecholamine biosynthetic pathway, converting norepinephrine to epinephrine. During periods of stress, epinephrine is secreted into the bloodstream from the adrenal medulla, where it is produced. Following epinephrine release, PNMT levels increase (10Ciaranello R.D. Dornbusch J.N. Barchas J.D. Science. 1971; 175: 789-790Crossref Scopus (7) Google Scholar; 21Kvetnansky R. Gerwitz G.P. Weise V.K. Kopin I.J. Am. J. Physiol. 1971; 220: 928-931Crossref PubMed Scopus (71) Google Scholar; 9Ciaranello R.D. Catecholamines and Stress. Elsevier Science Publishing Co., Inc., New York1980: 317-328Google Scholar; 34Weisberg E.P. Baruchin A. Stachowiak M.K. Stricker E.M. Zigmond M.J. Kaplan B.B. Mol. Brain Res. 1989; 6: 159-166Crossref PubMed Scopus (25) Google Scholar; 36Wong D.L. Lesage A. Siddall B. Funder J.W. FASEB J. 1992; 6: 3310-3315Crossref PubMed Scopus (61) Google Scholar; 33Viskupic E. Kvetnansky R. Sabban E.L. Fukuhara K. Weise V.K. Kopin I.J. Schwartz J.P. J. Neurochem. 1994; 63: 808-814Crossref PubMed Scopus (58) Google Scholar), suggesting that control of PNMT expression may be an important component of the mammalian stress response. phenylethanolamine N-methyltransferase footprinting unit(s). In an effort to identify the factors that regulate PNMT gene expression, we have recently shown that the immediate early gene transcription factor Egr-1 (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar), also known as NGFI-A (27Milbrandt J. Science. 1987; 238: 797-799Crossref PubMed Scopus (936) Google Scholar), Krox 24 (23Lemaire P. Revelant O. Bravo R. Charnay P. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 4691-4695Crossref PubMed Scopus (452) Google Scholar), zif 268 (8Christy B.A. Lau L.F. Nathans D. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 7857-7861Crossref PubMed Scopus (573) Google Scholar) and TIS8 (25Lim R.W. Varnum B.C. Herschman H.R. Oncogene. 1987; 1: 263-270PubMed Google Scholar), can stimulate transcription of the rat PNMT gene. Stimulation appears to occur through direct binding of Egr-1 protein to PNMT promoter elements located at −165 bp and to a lesser extent, at −45 bp, relative to the transcriptional initiation site at +1 bp (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). A perfect (9/9) match to the Egr-1 consensus DNA binding sequence (CGCCCCCGC) (5Cao X. Koski R.A. Gashler A. McKiernan M. Morris C.F. Gaffney R. Hay R.V. Sukhatme V.P. Mol. Cell. Biol. 1990; 10: 1931-1939Crossref PubMed Scopus (298) Google Scholar) is located at the −165 bp site, while an 8/9 consensus match (GCGGGGGGG) is located at the −45 bp site. Overlapping the 5′ end of these Egr-1 consensus elements are consensus binding sites for another transcription factor, Sp1 (consensus = (G/T)(G/A)GGC(G/T)(G/A)(G/A)(G/T) (12Faisst S. Meyer S. Nucleic Acids Res. 1992; 20: 3-26Crossref PubMed Scopus (938) Google Scholar), indicating that Sp1 may also influence PNMT gene expression. Sp1 is a ubiquitous mammalian transcription factor thought to be involved in the regulation of a large number of genes (18Kadonaga J.T. Jones K.A. Tjian R. Trends Biochem. Sci. 1986; 11: 20-23Abstract Full Text PDF Scopus (878) Google Scholar; 20Kriwacki R.W. Schultz S.C. Steitz T.A. Caradonna J.P. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 9759-9763Crossref PubMed Scopus (138) Google Scholar). For housekeeping genes that lack TATA and CAAT boxes, it has been suggested as being the signal initiating transcriptional activation (3Blake M.C. Jambon R.C. Swick A.G. Kahn J.W. Azizkhan J.C. Mol. Cell. Biol. 1990; 10: 6632-6641Crossref PubMed Scopus (233) Google Scholar; 17Joliff K. Li Y. Johnson L.F. Nucleic Acids Res. 1991; 19: 2267-2274Crossref PubMed Scopus (51) Google Scholar; Bosclair et al., 1993). Sp1 binding sites have also been shown to overlap the binding sites for other transcription factors, and it is thought that competition between Sp1 and these factors for binding site occupancy may be an important mechanism for controlling the expression of specific genes (22Lemaigre F.P. Lafontaine D.A. Courtois S.J. Durviaux S.M. Rousseau G.G. Mol. Cell. Biol. 1990; 10: 1811-1814Crossref PubMed Scopus (39) Google Scholar; 1Ackerman S.L. Minden A.G. Williams G.T. Bobonis C. Yeung C.Y. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 7523-7527Crossref PubMed Scopus (116) Google Scholar; 6Cao X. Mahendran R. Guy G.R. Tan Y.H. J. Biol. Chem. 1993; 268: 16949-16957Abstract Full Text PDF PubMed Google Scholar). Using gel mobility shift assays, we have previously shown that Sp1, present in nuclear extracts from a PC12-derived cell line (RS1 cells), can bind to DNA sequences representing the overlapping Sp1/Egr-1 consensus elements from the rat PNMT promoter (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). Furthermore, mutations designed to disrupt Egr-1 binding to its cognate consensus elements in the PNMT promoter also decreased base-line (uninduced) PNMT promoter activity. We speculated that this decline might be due to a reduction in Sp1 binding. The present study was undertaken to determine the role Sp1 may play in the regulation of PNMT gene expression and to distinguish its action from that of Egr-1. We demonstrate that Sp1 can bind to both overlapping consensus elements (−45 bp and −165 bp) with similar affinities and that this binding appears to stimulate transcriptional activity from the PNMT promoter. Egr-1, which also binds to both consensus elements, has ∼100-fold higher affinity for the distal site (−165 bp), and the integrity of this site is critical for Egr-1 activation of PNMT gene expression. These findings suggest that both Sp1 and Egr-1 can activate PNMT gene expression at these overlapping consensus elements, raising the possibility that selective activation of Sp1 or Egr-1 in response to specific stimuli may be biologically important for the activation and responsiveness of the PNMT gene. Eight duplex 21-mer oligonucleotides were prepared by annealing individually synthesized complimentary DNA fragments as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). These duplex oligonucleotides represent wild-type and mutant forms of the overlapping Sp1/Egr-1 DNA-binding elements from the rat PNMT promoter. In general, wild-type DNA sequences are designated with the prefix "wt" while mutants have the prefix "mut" followed by an "A" or "B" to indicate which of the overlapping elements from the rat PNMT promoter are represented by each duplex oligomer (A, −159 to −179 bp; B, −39 to −59 bp) (see Fig. 2). Each mutation carries an additional annotation indicating whether the mutation is specific for the Sp1 consensus sequence (designated by the number 1) or the Egr-1 consensus sequence (designated by the number 2). Thus, "wtA" refers to the wild-type overlapping consensus elements for Sp1 and Egr-1 located between −159 and −179 bp in the rat PNMT promoter, while "mutA1" and "mutA2" indicate that this same DNA fragment harbors a mutation that specifically targets the Sp1 and the Egr-1 consensus elements, respectively. The DNA sequences for the various oligonucleotides are as follows (mutant nucleotides are distinguished by lowercase letters): wtA, 5′-CCTCCCCGCCCCCGCGCGTCC-3′; mutA1, 5′-CCTCaaCGCCCCCGCGCGTCC-3′; mutA2, 5′-CCTCCCCGCCCCataGCGTCC-3′; wtB, 5′-GTCTGGGCGGGGGGGAGGGGA-3′; mutB1, 5′-GTCTaaGCGGGGGGGAGGGGA-3′; and mutB2, 5′-GTCTGGGCGGGGaaaAGGGGA-3′. As a positive control for Sp1 binding, a sequence previously shown to be a high affinity Sp1 binding site (Christy and Nathans, 1989), "Sp1" (5′-TAGAGGGGCGGGGCTCTAGAC-3′) was generated. As a negative control, an oligomer containing the PNMT glucocorticoid response element, "GRE" (5′-GCCAGAACAGAGTGTCCTTTC-3′), was also prepared. Finally, for site-directed mutagenesis, the following mutagenic 33-mer oligonucleotide primers were synthesized: mutA1P, 5′-CCACATCTCCCCCCTCaaCGCCCCCGCGCGTCC-3′; mutA2P, 5′-CCCCCTCCCCGCCCCataGCGTCCGTCCGGCGC-3′; mutB1P, 5′-GTGATGGAGGGGTCTaaGCGGGGGGGGAGGGG-3′; and mutB2P, 5′-GGGGTCTGGGCGGGGaaaAGGGGACCCAGTAG-3′. All oligonucleotides were obtained from DNA International, Inc. (Lake Oswego, OR). The wild-type PNMT-luciferase plasmid construct, pRP863LUC, contains the region of the rat PNMT promoter extending from −863 bp to +20 bp fused to the firefly luciferase reporter gene (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). Mutant PNMT-luciferase constructs were created by incorporating each of the mutagenic primers independently into pRP863LUC using the Transformer™ site-directed mutagenesis kit from Clontech (Palo Alto, CA) as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). The identity and integrity of the resulting plasmids, pRPmutA1LUC, pRPmutA2LUC, pRPmutB1LUC, and pRPmutB2LUC, were verified by DNA sequence analysis (Sequenase 2.0, U.S. Biochemical Corp.). The Egr-1 expression vector, pCMVEgr-1, and the corresponding nonfunctional control plasmid, pCMVETTL, have been described elsewhere (14Gupta M.P. Gupta M. Zak R. Sukhatme V.P. J. Biol. Chem. 1991; 266: 12813-12816Abstract Full Text PDF PubMed Google Scholar). All plasmids were amplified in the XL1-Blue strain of E. coli (Stratagene, Inc., La Jolla, CA) and purified through two successive CsCl density gradients. Following purification, plasmid integrity and identity was again verified by restriction digestion, agarose gel electrophoresis, and DNA sequence analysis. Using [γ-32P]ATP (DuPont NEN, Wilmington, DE) and T4 polynucleotide kinase (Life Technologies, Inc.), a DNA fragment from the rat PNMT promoter was labeled at either the HindIII site (+20 bp) or the NheI site (−391 bp, 32Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989: 13.78-13.104Google Scholar). Fragments labeled at the HindIII site were digested further with NheI, while those labeled at the NheI site were digested with HindIII, and radiolabeled DNA fragments (+20 bp to −391 bp) isolated by polyacrylamide gel electrophoresis as described previously (30Muller M.T. J. Virol. 1987; 61: 858-865Crossref PubMed Google Scholar). Approximately 7.5 × 104 dpm of each DNA fragment was incubated with 0-5 footprinting units (fpu) of purified Sp1 (1 fpu, ∼50 ng of Sp1 protein; Promega, Madison, WI) in 50 μl on ice using the binding buffer supplied in the Core Footprinting System™ (Promega). After incubation for 1 h, DNase I footprinting was performed according to the manufacturer's recommendations. The optimal RQ1 DNase I (Promega) concentration (0.1 unit/reaction) was determined empirically in preliminary experiments. DNase I cleavage patterns were identified by polyacrylamide gel electrophoresis (7% sequencing gel), and protected sequences were mapped by comparison to Maxam and Gilbert G and G/A chemical sequencing ladders generated from these same DNA probes and run in adjacent lanes on the gels (32Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989: 13.78-13.104Google Scholar). The banding patterns were visualized by autoradiography. Gel mobility shift assays were performed as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar) using 1 ng of radiolabeled probe (∼3 nM, specific activity = 2.5 × 108 dpm/μg) and purified proteins in place of crude nuclear extracts. Typically, 0.5 fpu of purified Sp1 protein (Promega) was used per binding reaction, while 100 ng of the Egr-1 DNA binding domain protein, V5 (13Gashler A.L. Swaminathan S. Sukhatme V.P. Mol. Cell. Biol. 1993; 13: 4556-4571Crossref PubMed Scopus (213) Google Scholar), kindly provided by Dr. Vikas P. Sukhatme (Department of Medicine, Harvard Medical School), was utilized per reaction. For DNA competition experiments, excess concentrations of specific unlabeled duplex competitor 21-mers were included in the binding reactions. Relative binding was quantitated from the intensity of the protein-DNA complexes as determined densitometrically from the autoradiographic images using Image 1.52 software (National Institutes of Health) on a MacIntosh FX computer. The concentration of competitor oligomer that effectively inhibited maximum binding by 50% (IC50) was calculated by plotting the concentration of each competitor versus the intensity of radiolabeled DNA-protein complexes. RS1 cells, a subclone of the PC12 cell line (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar), were maintained in Dulbecco's modified Eagle's medium (DMEM) containing 5% defined iron-supplemented bovine calf serum (Hyclone, Inc., Logan, UT), 5% donor equine serum (Hyclone, Inc.), gentamycin sulfate (50 μg/ml; U.S. Biochemical Corp.), and hygromycin B (200 units/ml; Calbiochem) in a 37°C incubator (7% CO2, 93% air) as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). Transient transfection assays were performed by calcium-phosphate co-precipitation as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). Total protein and luciferase and β-galactosidase activities were also determined as described previously (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). Experiments were repeated at least twice with six replicates/sample for each experiment. Statistical significance was determined by one way analysis of variance. The rat PNMT promoter contains two overlapping Sp1/Egr-1 consensus binding elements (Fig. 1). Both consensus sequences share six base pairs, with the Sp1 consensus sequences extending an additional three bp in the 5′ direction and the Egr-1 consensus sequences extending three bp in the 3′ direction. The distal overlapping site, located at −165 bp, contains a perfect match to both the Egr-1 (5Cao X. Koski R.A. Gashler A. McKiernan M. Morris C.F. Gaffney R. Hay R.V. Sukhatme V.P. Mol. Cell. Biol. 1990; 10: 1931-1939Crossref PubMed Scopus (298) Google Scholar) and Sp1 consensus sequences (Faisst and Meyer, 1992). The proximal site, located at −45 bp, contains an 8/9 match to the Egr-1 consensus DNA binding sequence and a perfect match to the Sp1 consensus sequence. We have previously shown that Egr-1 can bind to these two sites in the rat PNMT promoter, and Sp1 appeared to bind as well (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). To identify the DNA sequences involved in Sp1 binding to the rat PNMT promoter, DNase I footprinting analysis was performed. As shown in Fig. 2, Sp1 binds at two sites, both of which were predicted from the overlapping Sp1/Egr-1 consensus elements. The distal or "A" site spans the region between −176 and −162 bp, while the proximal or "B" site includes the nucleotide sequences between −55 and −39 bp. No other protected regions were observed. To determine the relative binding affinity of Sp1 for these sites, competitive gel mobility shift assays were performed. A representative experiment, utilizing a probe corresponding to the wild-type A site (wtA), is shown in Fig. 3. The results demonstrated that Sp1 binding to the wtA probe was effectively competed by either wtA itself or the wild-type B oligomer sequence (wtB) with similar relative affinities (IC50 for wtA = 210 nM; IC50 for wtB = 140 nM). However, the wtA and wtB sequences appeared to represent only moderate affinity Sp1 binding sites, since the high affinity Sp1 competitor sequence (Christy and Nathans, 1989) more effectively competed for Sp1 binding than either the wtA or wtB sequences (IC50 for Sp1 = 12 nM). In contrast, the GRE oligonucleotide, representing the glucocorticoid response element, did not compete for Sp1 binding (IC50 for GRE > 2000 nM). Similar results were obtained when wtB was used as the probe instead of wtA (data not shown). Thus, while Sp1 bound specifically to the A and B sites in the PNMT promoter, it did so with only moderate affinity. Since Sp1 binding in the PNMT promoter overlaps those regions bound by Egr-1, Sp1 and Egr-1 may potentially compete for binding at a given site. To examine this possibility, the wtA and wtB probes were incubated with increasing concentrations of Sp1 in the presence of a constant amount of the purified Egr-1 DNA binding domain protein (V5) (Fig. 4). V5 bound to the probe and formed a complex that migrated between the unbound probe and the Sp1-DNA complex. As the concentration of Sp1 was increased, Sp1-DNA complex formation correspondingly increased while V5-DNA complex formation diminished. Conversely, when the Sp1 concentration was kept constant and the V5 concentration increased, Sp1-DNA complex formation appeared to be displaced in favor of V5-DNA complex formation (Fig. 5). Since no additional or intermediate protein-DNA complexes were observed, Sp1 and V5 binding to either the A or the B site appeared to be mutually exclusive, with occupancy dependent upon the relative concentration of these two factors.Figure 5:V5 displacement of Sp1 binding at the A site. Gel mobility shift assays were performed using the 21 bp wtA probe in the presence of a constant amount of Sp1 protein (0.5 fpu/reaction) and increasing quantities of V5 protein. Lane1, probe alone; lanes2-6, probe and Sp1 protein plus 100, 200, 400, and 1000 ng of V5, respectively. The specific complexes are identified to the left of the autoradiogram.View Large Image Figure ViewerDownload Hi-res image Download (PPT) In addition, occupancy of these promoter elements by Sp1 or Egr-1 appeared to be dependent upon the relative binding affinities of these two factors for each site. For example, V5 binding to the A site was much stronger than its binding to the B site (compare V5-DNA complex intensities in Fig. 4, lanes5 and 13). Consistent with this observation is the fact that equivalent concentrations of Sp1 could displace V5 binding at the B site much more effectively than they could at the A site (Fig. 4, compare lanes14-16 with lanes6-8). Similarly, Sp1 binding at the A site was readily displaced by V5 (Fig. 5), whereas equivalent concentrations of V5 only partially interfered with Sp1 binding at the B site (data not shown). Hence, binding site occupancy was dependent upon both the relative concentration of Sp1 and Egr-1 and the relative binding affinity of each factor for the two overlapping sites. To discriminate between the actions of Sp1 and Egr-1 in the PNMT promoter, we attempted to selectively interfere with the binding of either factor by incorporating specific mutations into the nonoverlapping region of each binding site (Fig. 6). The mutations were then evaluated for their ability to bind to Sp1 and Egr-1 using competitive gel mobility shift assays (Fig. 6). The mutA1 mutation targets the Sp1 consensus binding element at the A site by changing the second and third cytosine bases to adenosine (Fig. 6, A and B), while the mutA2 mutation converts the 3′ CGC triplet of the Egr-1 binding site into ATA (Fig. 6, A and B). A similar set of mutations was generated for the overlapping consensus elements at the B site. The mutB1 mutation targets Sp1 binding by converting the second and third guanosine residues to adenosine (Fig. 6, C and D), whereas the mutB2 mutation specifically alters the Egr-1 consensus element by replacing the 3′ terminal GGG triplet with AAA (Fig. 6, C and D). IC50 values for these competitors are shown in Table 1.TABLE I Open table in a new tab As is apparent from Fig. 6A and Table 1, the mutA1 mutation reduced the binding affinity for Sp1 approximately 5-fold compared with the wtA sequence, yet this mutation had very little influence on Egr-1 binding (Fig. 6B). In contrast, the mutA2 mutation eliminated Egr-1 binding (Fig. 6B). In addition, although its ability to compete for Sp1 binding was also compromised, the reduction in Sp1 binding affinity was relatively small compared with the decrease in Egr-1 binding affinity (2.5-fold compared with 140-fold, respectively; Table 1). Thus, mutA2 preferentially reduced Egr-1 DNA-binding affinity, while mutA1 selectively reduced Sp1 binding affinity. Similarly, the mutB1 mutation effectively reduced Sp1 binding, but its ability to compete for Egr-1 binding was relatively unchanged (Fig. 6, C and D, Table 1). As expected, the mutB2 mutation, designed to reduce Egr-1 binding affinity, substantially reduced Egr-1 binding compared with its wild-type counterpart (Fig. 6D). Surprisingly, however, the mutB2 sequence enhanced Sp1 binding 3.5-fold compared with the wild-type B sequence. Thus, the mutations appeared to behave predictably with respect to their ability to interfere with either Sp1 or Egr-1 binding, with the exception of mutB2, which displayed increased affinity for Sp1. To evaluate the effects of these same mutations on PNMT promoter activity, the mutA1, mutA2, mutB1, and mutB2 mutations were incorporated into PNMT-luciferase reporter plasmid constructs, and reporter gene expression was evaluated by transient transfection assays using RS1 cells. Reporter gene constructs with mutations reducing relative Sp1 binding affinity (mutA1, mutA2, and mutB1) produced 30-50% less luciferase activity in comparison with the wild-type control (Fig. 7). In contrast, the reporter gene construct containing the mutation that increased Sp1 DNA-binding affinity, mutB2, also increased luciferase activity approximately 4-fold. Thus, the changes in luciferase expression appeared to reflect the changes in the relative binding affinities for Sp1. Finally, the ability of Egr-1 to stimulate PNMT promoter activity was examined in co-transfection assays using the same complement of PNMT-luciferase plasmids and either a functional or a nonfunctional Egr-1 expression construct. As shown in Fig. 8, Egr-1 significantly increased (p < 0.001) luciferase expression from the wild-type, mutA1, mutB1, and mutB2 PNMT-luciferase plasmids but not from the mutA2 PNMT-luciferase construct. Moreover, the responsiveness of the mutA1 and mutB1 PNMT-luciferase plasmids to Egr-1 was not significantly different from that of the wild-type construct (Fig. 8). As indicated earlier, neither of these mutations substantially changed the relative affinity for Egr-1 binding (Table 1), consistent with their inability to influence Egr-1 activation of luciferase activity. In contrast, the mutB2 mutation reduced Egr-1 binding to the B site (Fig. 6, Table 1), yet Egr-1 still significantly stimulated luciferase activity from the mutB2 PNMT promoter, likely due to the integrity of the A site. These results, therefore, suggest that Egr-1 stimulation of PNMT promoter activity may primarily depend upon Egr-1 binding to PNMT promoter DNA at the A site. The present study extends our previous findings demonstrating that Egr-1 activates the expression of the rat PNMT gene and that Sp1 is a potential transcriptional activator as well. We show that Sp1 binds to both overlapping Sp1/Egr-1 consensus elements (−45 and −165 bp) while Egr-1 binds preferentially to the A site (−165 bp). The integrity of the Egr-1 consensus element at the A site is essential for Egr-1-stimulated PNMT promoter activity, while the B site (−45 bp) does not appear to influence Egr-1-mediated PNMT promoter activation. Mutations that decrease Sp1 binding to its consensus elements in the PNMT promoter reduce basal promoter activity in the RS1 cells, while a mutation that increases Sp1 binding stimulates PNMT promoter activity. Finally, Sp1 displaces Egr-1 binding at the overlapping consensus elements in a concentration-dependent manner. While the evidence suggests that both Egr-1 and Sp1 function as transcriptional activators of the PNMT gene, their specific transcriptional roles remain to be elucidated. Egr-1, a member of the immediate early gene family, is usually present at low levels endogenously but can be rapidly and transiently induced in response to a variety of external stimuli (27Milbrandt J. Science. 1987; 238: 797-799Crossref PubMed Scopus (936) Google Scholar; 5Cao X. Koski R.A. Gashler A. McKiernan M. Morris C.F. Gaffney R. Hay R.V. Sukhatme V.P. Mol. Cell. Biol. 1990; 10: 1931-1939Crossref PubMed Scopus (298) Google Scholar; 15Iwaki K. Sukhatme V.P. Shubeita H.E. Chien K.R. J. Biol. Chem. 1990; 265: 13809-13817Abstract Full Text PDF PubMed Google Scholar; 2Bhat R.V. Cole A.J. Baraban J.M. J. Pharmacol. Exp. Ther. 1992; 263: 343-349PubMed Google Scholar; 26Mailleux P. Zhang F. Vanderhaeghen J.J. Neurosci. Lett. 1992; 147: 182-184Crossref PubMed Scopus (22) Google Scholar; 19Katayama N. Iwata E. Sakurai H. Tsuchiya T. Tsuda M. J. Neurochem. 1993; 60: 902-907Crossref PubMed Scopus (16) Google Scholar). Egr-1 is therefore considered to be a transcriptional factor that acts as a modulator of basal gene expression. Consistent with this role, we have recently shown that Egr-1 expression in the rat adrenal medulla is rapidly activated neurally by reserpine treatment, which also causes significant increases in PNMT mRNA (35Wong D.L. Bildstein C.L. Siddall B. Lesage A. Yoo Y.S. Mol. Brain Res. 1993; 18: 107-114Crossref PubMed Scopus (22) Google Scholar; 11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). We have further shown that the phorbol ester, phorbol 12-myristate 13-acetate, can induce PNMT promoter activity via an Egr-1-dependent mechanism in RS1 cells (29Morita K. Ebert S.N. Wong D.L. J. Biol. Chem. 1995; 270: 11161-11167Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). In contrast, Sp1 is a ubiquitous protein, constitutively expressed at very high levels, and it appears to act as a transcription initiator for many housekeeping genes that lack TATA and CAAT boxes (3Blake M.C. Jambon R.C. Swick A.G. Kahn J.W. Azizkhan J.C. Mol. Cell. Biol. 1990; 10: 6632-6641Crossref PubMed Scopus (233) Google Scholar; 17Joliff K. Li Y. Johnson L.F. Nucleic Acids Res. 1991; 19: 2267-2274Crossref PubMed Scopus (51) Google Scholar; 4Boisclair Y.R. Brown A.L. Casola S. Rechler M.M. J. Biol. Chem. 1993; 268: 24892-24901Abstract Full Text PDF PubMed Google Scholar). Recently, however, Sp1 protein has been demonstrated to be regulated developmentally and in response to specific stimuli (16Jackson S.P. MacDonald J.J. Lees-Miller S. Tjian R. Cell. 1990; 63: 155-165Abstract Full Text PDF PubMed Scopus (518) Google Scholar; 31Saffer J.D. Jackson S.P. Annarella M.B. Mol. Cell. Biol. 1991; 11: 2189-2199Crossref PubMed Scopus (485) Google Scholar), thereby suggesting that this factor may be critical during differentiation and has the potential for regulated expression as well. If the latter is indeed the case, then under appropriate circumstances, Sp1 might also function as a modulator of basal gene transcription. Based on our studies to date, the PNMT promoter functions as a weak promoter in RS1 cells in comparison with other promoters, particularly the viral promoters (11Ebert S.N. Balt S.L. Hunter J.P.B. Gashler A. Sukhatme V. Wong D.L. J. Biol. Chem. 1994; 269: 20885-20898Abstract Full Text PDF PubMed Google Scholar). It does possess sequences consistent with an ATA (−30 bp) box. Hence, if Sp1 participates in constitutive PNMT expression, we would predict that the observed low levels of basal PNMT promoter activity may be due to the fact that the Sp1 consensus binding sites in the PNMT promoter are not high affinity sites. However, when we compared the Sp1 sites within the PNMT promoter with the nine-nucleotide Sp1 sequence identified by Faisst and Meyer(1992), their perfect match suggested that these sites should be very effective Sp1 sites (Table 2). Yet, gel mobility shift competition assays showed that the Sp1 sites were only of moderate binding affinity. A 10-nucleotide Sp1 consensus sequence has been identified as an alternative to the nine-bp Sp1 consensus element and has been confirmed as a high affinity site (18Kadonaga J.T. Jones K.A. Tjian R. Trends Biochem. Sci. 1986; 11: 20-23Abstract Full Text PDF Scopus (878) Google Scholar; 20Kriwacki R.W. Schultz S.C. Steitz T.A. Caradonna J.P. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 9759-9763Crossref PubMed Scopus (138) Google Scholar). If we consider the PNMT Sp1 sequences, the 10th nucleotide, located at the 3′ end, is mismatched and reduces Sp1 binding 10-fold in comparison with the perfect match 10-bp sequence, suggesting that this nucleotide is critical in specifying the binding affinity for Sp1. In fact, the 10-bp Sp1 element at the B site in the PNMT promoter is identical to an Sp1 element in the herpes simplex virus type 1 IE3 gene promoter (18Kadonaga J.T. Jones K.A. Tjian R. Trends Biochem. Sci. 1986; 11: 20-23Abstract Full Text PDF Scopus (878) Google Scholar; 20Kriwacki R.W. Schultz S.C. Steitz T.A. Caradonna J.P. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 9759-9763Crossref PubMed Scopus (138) Google Scholar), which also has intermediate affinity for Sp1 protein. Thus, the wobble in the distal 3′ nucleotide of the PNMT Sp1 consensus elements is consistent with the observed low basal PNMT promoter activity and moderate binding affinity as predicted from the 10-bp Sp1 high affinity consensus sequence.TABLE II Open table in a new tab The relative affinities of the overlapping Sp1/Egr-1 binding sites for their respective proteins and the ability of Sp1 to compete with Egr-1 for binding at the overlapping Sp1/Egr-1 sites in a mutually exclusive, concentration-dependent fashion also suggest important biological ramifications. Since Egr-1 can preferentially displace Sp1 binding at the A site, and in so doing, enhance PNMT promoter activity, the A site may normally function as a high affinity Egr-1 site in the presence of sufficient amounts of Egr-1 protein. However, when Egr-1 levels are low, the A site may alternatively function as an Sp1 site, albeit of moderate affinity. In contrast, the B site where Egr-1 binding seems to have little effect on Egr-1-mediated PNMT promoter activity, may be a biologically active Sp1 site under most circumstances. Consistent with this interpretation is a marked reduction in basal PNMT promoter activity when the B site is mutated. A similar reduction is observed with mutation of the A site, but this occurs at low basal Egr-1 levels, a circumstance where the A site may function as an Sp1 site as predicted above. Several other genes have also been identified with overlapping or closely adjacent Sp1/Egr-1 binding motifs (1Ackerman S.L. Minden A.G. Williams G.T. Bobonis C. Yeung C.Y. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 7523-7527Crossref PubMed Scopus (116) Google Scholar; 24Li Y. Camp S. Rachinsky T.L. Bongiorno C. Taylor P. J. Biol. Chem. 1993; 268: 3563-3572Abstract Full Text PDF PubMed Google Scholar; 28Molnar G. Crozat A. Pardee A.B. Mol. Cell. Biol. 1994; 14: 5242-5248Crossref PubMed Scopus (84) Google Scholar). In the case of the murine adenosine deaminase gene, Sp1 stimulates, while Egr-1 represses, adenosine deaminase promoter activity (1Ackerman S.L. Minden A.G. Williams G.T. Bobonis C. Yeung C.Y. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 7523-7527Crossref PubMed Scopus (116) Google Scholar). The inhibitory action of Egr-1 is opposite to its effects on the PNMT promoter reported here. However, both the stimulatory and inhibitory actions of Egr-1 on transcriptional activity are consistent with the functional domains of Egr-1 protein (13Gashler A.L. Swaminathan S. Sukhatme V.P. Mol. Cell. Biol. 1993; 13: 4556-4571Crossref PubMed Scopus (213) Google Scholar). Egr-1 has been shown to have three activating and one inhibitory domain, although the specific biological roles and activities of these domains remain to be identified. In summary, we have shown that Sp1 and Egr-1 can differentially stimulate PNMT gene expression through interactions with overlapping consensus elements in the PNMT promoter. Sp1 seems to interact with both the A and B sites with moderate binding affinities to induce PNMT promoter activity. In contrast, Egr-1 acts predominantly through the A site, where it appears to effectively exclude the binding of Sp1, thereby increasing PNMT promoter activity. The differences in binding affinities of Sp1 and Egr-1 for their overlapping consensus sequences in the PNMT promoter, their apparent exclusionary binding, and differences in their abilities to activate transcriptional activity from the PNMT promoter suggest that these two factors may play unique roles in the control of PNMT gene expression in response to specific biological stimuli. Further definition of their biological roles will therefore be an important focus for future studies.
Год издания: 1995
Авторы: Steven N. Ebert, Dona L. Wong
Издательство: Elsevier BV
Источник: Journal of Biological Chemistry
Ключевые слова: Synthesis and Biological Activity, Chemical Synthesis and Analysis, Cancer-related gene regulation
Другие ссылки: Journal of Biological Chemistry (PDF)
Journal of Biological Chemistry (HTML)
PubMed (HTML)
Journal of Biological Chemistry (HTML)
PubMed (HTML)
Открытый доступ: hybrid
Том: 270
Выпуск: 29
Страницы: 17299–17305