Transcriptional Regulation of the Human Transforming Growth Factor-a Gene
Rampyari H. Raja, Andrew J. Paterson, Tae Ho Shin, and Jeffrey E. Kudlow Departments of Medicine and Cell Biology University of Alabama Birmingham, Alabama 35294
INTRODUCTION
We and others have previously reported that transforming growth factor-a (TGFa) expression is hormonally responsive and its expression is coregulated with that of its receptor [the epidermal growth factor (EGF) receptor]. The 5'-flanking region of the TGFa gene was characterized to determine whether it could confer hormone responsiveness to a reporter gene (luciferase) in human mammary carcinoma cells (MDA468). This segment of the gene is GC rich and contains an element strikingly similar to the core element of the EGF receptor gene that has been shown to mediate both basal and hormonestimulated expression of the EGF receptor. We now report that a 313-basepair (bp) proximal element of the TGFa 5'-flanking region (-373 to - 5 9 relative to the TGFa translation start codon) is capable of conferring responses to phorbol ester and EGF. This gene segment does not contain the EGF receptor gene homolog or potential AP-2-binding sites, suggesting that these elements are not necessary for basal and EGF- or phorbol ester-responsive TGFa gene expression. This 313-bp proximal element also confers proper transcriptional initiation to the chimeric TGFa-luciferase reporter construct, indicating it is the TGFa promoter. A 1.1-kilobase segment of the TGFa 5'-flanking region also confers retinoic acid, thyroid hormone, and glucocorticoid responsiveness despite the absence of recognizable steroid hormone receptor-binding sites. These hormones stimulate reporter expression 1.5- to 2-fold in a dose-dependent manner. Extension of the 5'flanking region to -3500 results in marked suppression of reporter gene expression. These results indicate that the TGFa gene 5'-flanking sequence contains the elements responsible for hormonal responsiveness of this gene and that these elements are distinct from those that regulate the expression of the EGF receptor gene. (Molecular Endocrinology 5: 514-520, 1991)
Transforming growth factor-a (TGFa), a homolog of epidermal growth factor (EGF), uses the EGF receptor to exert its effects on cells (see Ref. 1 for review). While originally believed to be exclusively expressed in oncofetal tissues (2), TGFa has since been shown to be present in a variety of normal adult tissues, including the anterior pituitary gland (3-6), ovary (7, 8), brain (9, 10), skin (11), macrophages (12), and vascular smooth muscle cells (13). In each of these tissues, TGFa expression can potentially be controlled at either the transcriptional (6, 7, 11) or posttranslational (13) level. Hormones are known to be among those factors capable of altering TGFa expression. Estrogen (14), phorbol ester (6, 15, 16), and EGF (6, 11, 16) have each been shown to stimulate TGFa mRNA accumulation in a variety of cell types. We have previously focused on anterior pituitary (6) and mammary carcinoma cells (1719), in which we showed that EGF and phorbol ester simultaneously stimulate EGF receptor and TGFa mRNA accumulation. The simultaneous increase in receptor and ligand expression in the same cell could have a multiplicative effect on signal transduction through this receptor system. While considerable progress has been made in defining the locus in the EGF receptor gene that mediates responsiveness to hormones (20-22), the same is not the case for the TGFa gene. While the 5'-flanking region of both the human (23) and rat (24) TGFa genes has been cloned and shown to drive basal reporter gene expression, the presence of hormonal responsiveness in these TGFa genes has not been reported. Striking similarities exist in the 5'-flanking regions of the EGF receptor and TGFa genes. Both genes are very GC rich, contain potential AP-2- and Sp1-binding sites, and lack a TATA box. In addition, an EGF receptor gene core element, as defined by Hudson and co-workers (22) mediates basal and hormonal-stimulated transcription. The existence of a highly homologous element within the 5'-flanking region of the TGFa gene and the coordinate regulation of the EGF receptor and TGFa genes in many cells suggested that these genes shared a similar core element involved in hormonal regulation.
0888-8809/91/0514-0520$03.00/0 Molecular Endocrinology Copyright © 1991 by The Endocrine Society
514
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
515
TGFa Gene Regulation
However, we have cloned 13 kilobases (kb) of the 5'flanking region of the TGFa gene and show in this paper that the proximal 313-basepair (bp) segment of the gene, devoid of the EGF receptor gene homolog, is capable of conferring basal expression and EGF and TPA responsiveness to a luciferase reporter gene. These studies establish that these agents mediate their transcriptional effects on TGFa gene expression through a regulatory element distinct from the element involved in EGF receptor gene expression.
TPA
JVhJY EGF
RESULTS C
C
1
2
3
RA
C
1
2
T3
JM\J\J\M\JV C
C
1
2
3
4
5
C
C
1
2
3
4
5
DEX
B
An
Rl Amp
TGFcr 5' Region Fig. 1. A, Effects of Various Concentrations of TPA, EGF, Retinoic Acid, Thyroid Hormone, and Dexamethasone on TGFa Promoter Expression MDA468 cells were transiently transfected with a construct containing 1.1 kb of the TGFa 5'-flanking region up-stream of the luciferase reporter gene. The transfected pool of cells was divided equally onto 6-cm culture dishes and, 40 h later, treated for 5 h with the stated hormones before harvest.
Preliminary experiments determined that the 1.1-kb fragment of the TGFa 5'-flanking region could direct luciferase expression in the MDA468 cells. The transfected cells were then tested to determine if reporter expression was hormone responsive. We had previously shown by Northern blotting that the TGFa mRNA content of MDA468 cells increased in response to EGF and phorbol ester [12-O-tetradecanoylphorbol-13-acetate (TPA)] (19). Furthermore, the similarity in the structures of the EGF receptor and TGFa genes suggested that the TGFa gene might confer responses to steroid hormones similar to those of the EGF receptor gene (21). In concert with these earlier observations, exposure of the transfected cells to EGF and TPA for 5 h resulted in a dose-dependent increase in cellular luciferase activity (Fig. 1A). Optimal exposure to TPA was 4-5 h, with a decline in activity to subbasal levels after 24 h, whereas EGF exposure resulted in a continuously rising luciferase activity which had not reached a plateau even by 48 h (data not shown). Integration of the light signals revealed 4- to 5-fold stimulations of reporter activity after a 5-h exposure to maximally effective doses of EGF (1 nM) and TPA (10 nM). We also tested the transient transfectants for responsiveness to retinoic acid, thyroid hormone (T3), and the glucocorticoid dexamethasone. In each case, there was a small (1.5to 2-fold) reproducible increase in reporter activity at doses of 10 nM for retinoic acid, 1 nM for thyroid hormone, and 0.5 nM for dexamethasone. These experiments were repeated in duplicate on three separate occasions and gave the same qualitative results. These
Luciferase activity was assayed in a solubilized extract of the cells. Shown is a chart recorder tracing of the luminometer output (45 sec) for each dose of hormone for a typical experiment. In each case, C designates the unstimulated control. The doses of hormone were as follows: TPA, 1 = 0.5 nM, 2 = 1 nM, 3 = 10 nM, 4 = 300 nM; EGF, 1 = 0.1 nM, 2 = 0.5 nM, 3 = 1 nM, 4 = 3 nM, 5 = 30 nM; retinoic acid, 1 = 1.0 nM, 2 = 10 nM, 3 = 100 nM, 4 = 1 HM; thyroid hormone, 1 = 0.1 nM, 2 = 0.5 nM, 3 = 1 nM, 4 = 10 nM, 5 = 30 nM; dexamethasone, 1 = 0.5 nM, 2 = 1 nM, 3 = 10 nM, 4 = 30 nM, 5 = 100 nM. The peaks were integrated for 10 sec, and the areas did not vary by more than 7% for each replicate point. B, The pXP1 mammalian expression vector into which the TGFa Gene segments were inserted.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 516
results imply that the previously observed accumulation of TGFa mRNA in MDA468 cells in response to EGF and TPA results from increased transcription of this gene. The finding that the gene is also responsive to thyroid hormone, retinoic acid, and glucocorticoid is further indication of the hormonal responsiveness of the TGFa gene. The responses of the TGFa gene to these steroid hormones occurred at similar doses and were of similar magnitude to those previously reported for the EGF receptor gene (21). Our earlier studies had shown simultaneous stimulation of EGF receptor and TGFa gene expression by EGF and TPA, prompting us to search for common regulatory elements. A computer comparison of the genes demonstrated the presence of a strikingly homologous 26-bp segment, of which 20 bases were identical, and five of the six nonidentical bases were GC interchanges (Fig. 2). This homology took on possible functional significance when Hudson and co-workers (22) recently showed that this segment in the EGF receptor gene was the core element that conferred hormonal responsiveness to that gene. To test whether the EGF receptor gene-like element in the TGFa gene was necessary for EGF and TPA responsiveness, we assessed reporter constructs that contained or lacked this segment. We found that the reporter construct containing the 313-bp insert (-373 to - 5 9 relative to the TGFa translation start site), which deletes the entire EGF receptor gene-like element, responded to EGF (Fig. 3A) and TPA (Fig. 4) with 22- and 4-fold increments, respectively. There are also seven potential AP2 sites in the 1.1-kb segment of the the TGFa 5'flanking region (Fig. 3B) that could confer EGF/TPA responsiveness. However, the 313-bp fragment of the TGFa gene does not contain any of these AP-2 sites. The ability of this proximal segment to confer basal and EGF/TPA-stimulated reporter expression suggests that AP-2 and the factor(s) that binds to EGF receptor gene core element are not necessary for basal or EGF/TPAstimulated TGFa expression. We could detect no reporter activity in the construct containing the additional 2.4 kb of distal 5'-flanking sequence even after EGF stimulation (Fig. 3A). Biasband and co-workers (24) have reported a similar find-
Homologous Segments in 5'-Flanking Regions of the EGF Receptor and TGFa Genes -410
-400
-390
TTTTCGTCGGCCCGGGTGCCCGGCTCGCGCC CGCGTCCGCCCGAGTCCCCGCCTCGCCGCCA -110
-100
TGFa EGF-R
-90
Fig. 2. A Comparison of the EGF Receptor Gene Core Element with a Similar Segment in the TGFa Gene The positions shown above and below the bases indicate the distance from the A in the ATG start codon of both genes.
ing in the rat TGFa gene, in that a 1.9-kb fragment of 5'-flanking region of the rat gene conferred no basal activity to the reporter. These results suggest the presence of an up-stream suppressor sequence whose functional significance and location in these gene segments have yet to be determined. We could also detect neither basal nor EGF-stimulated activity in the reporter construct containing the 183-bp (-242 to -59) TGFa gene segment (Fig. 3A), suggesting that the sequence up-stream of the Alu\ site at -242 bp relative to the translational start site is required for basal and EGFstimulated transcription. The luciferase expression is specifically driven by the TGFa 5'-flanking sequence contained in the expression plasmids, as indicated by the observation that promoterless constructs using the parental pXP1 plasmid and the construct containing the -242 to - 5 9 segment of the TGFa gene showed no detectable luciferase activity in the transiently transfected cells. The responses to steroid hormones (Fig. 1A) are also specific, because the construct containing the 313-bp insert (-373 to -59), while EGF/TPA responsive, was nonresponsive to retinoic acid, thyroid hormone, and dexamethasone (data not shown). The failure of the 313-bp insert to confer steroid hormone responsiveness suggests that the observed responses from the 1.1-kb insert were transcriptional and not due to stabilization of the luciferase mRNA. While the receptors for these steroids have not been directly detected in MDA468 cells, the transcriptional responses suggest that they are indeed present. Specificity of the responses is also ensured by the design of pXPI. This plasmid contains two polyadenylation signals up-stream of the promoter insertion site which are designed to terminate any nonspecifically initiated transcripts before they extend into the luciferase reporter sequence (25). This vector also does not appear to contain cryptic enhancer sequences, since constitutive promoters are not rendered hormone responsive (21). Transfection of a Rous sarcoma virus (RSV)-long terminal repeat-luciferase chimera resulted in luciferase activity 20-fold higher than the basal activity seen with the 1.1 -kb fragment of the TGFa gene (data not shown). However, the EGF receptor promoter (1.1 kb) drove expression of the luciferase reporter to levels similar to those seen with the 1.1-kb fragment of the TGFa gene (Fig. 4) and responded similarly to TPA, suggesting that the enhancer-promoter activity of the TGFa gene segment is similar in potency in these cells to that in the corresponding EGF receptor gene segment. The ability of the 5'-flanking region of the human TGFa gene to confer EGF and TPA responsiveness to the luciferase reporter gene similar to that seen for the endogenous TGFa gene suggests that this region of the gene contains the key enhancers and the promoter. Further support derives from the prior observation (23) that the 5'-flanking region of the TGFa gene confers initiation to a chimeric TGFa-chloramphenicol acetyltransferase construction which corresponds to the native gene. Indeed, using an essentially identical TGFa
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
517
TGFa Gene Regulation
Insert Limits
• -EGF
Nil
E+EGF
-242 to -59 -373 to -59 -930 to -59 -1139 to -59 -3500 to -59
500
1000 1500 2000
2500
Relative Luciferase Activity
-3500
-1139 -930
-373 -242
o
Potential AP2 binding sites
M
EGF Receptor Gene Homologue
Luciferase
Fig. 3. A, EGF Responsiveness of Various Segments of the TGFa Gene MDA468 cells were transfected, as described in Fig. 1, with constructs containing various segments of the TGFa gene. All TGFa gene segments had the same 3'-end, corresponding to - 5 9 relative to the translation start site and extending 5' of this site to the designated position shown in the figure. EGF was added to the cells 20 h after transfection. Plotted is the mean of duplicate determinations of the numerical value of the integrated signal from the luminometer resulting from assays of cell extracts that were derived from cells that were treated with or without 1 nM EGF for 48 h, as indicated on the figure. B, The construction of the TGFaluciferase chimeric plasmid is shown. The positions of the EGF receptor gene homolog and potential AP-2-binding sites are shown in relation to the positions in the gene to which the 5'-deletions extended.
gene segment in a chimeric TGFa-luciferase construct, a dominant transcription initiation site was also detected by primer extension. This site (at position -115 relative to the TGFa translation start site) was in agreement with the 5'-end of the transcript from the native gene in MDA468 cells (Fig. 5). These primer extension data were also confirmed by RNAse protection, using a riboprobe spanning the junction between the TGFa (-373) and luciferase (99) genes (data not shown). Other minor sites were detected on longer exposures of the gels. The TGFa gene is probably capable of initiating transcription from multiple sites, consistent with other genes whose promoters are CG rich and contain no TATA box (24-27). Under different circumstances, one or more of the potential initiation sites may dominate.
DISCUSSION In previous studies we have shown that the expression of TGFa and its receptor (EGF receptor) are coordi-
nately regulated in MDA468 cells by EGF in a manner that appears to depend on protein kinase-C (16, 18, 19). With the recent elucidation of the sequences of the promoter regions of both of these genes (20-24), it became evident that there were common structural features, raising the hypothesis that the coordinate regulation of these genes could result from these structural similarities. Hudson and co-workers (22) have determined that a 36-bp element in the EGF receptor gene mediates responses to EGF and TPA. We found a highly homologous sequence in the TGFa gene. This EGF receptor gene-like element then became an attractive candidate for the EGF/TPA-responsive element in the TGFa gene. In this study we clearly demonstrate that deletion of this EGF receptor gene homolog from the TGFa 5'-flanking sequence does not alter the ability of the TGFa promoter to confer basal and EGF/TPAstimulated luciferase expression from the transiently transfected chimeric construct. Thus, this element in the TGFa gene is not necessary for TGFa expression. This result is consistent with the finding of Hudson and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 518
-350
"S
1 2
3
4
5
.
-101
6
Fig. 4. TPA Responsiveness of Various Segments of the TGFa Gene Compared to the EGF Receptor Gene MDA468 cells were transfected, as described in Fig. 1, with constructs containing the 313-bp (3, 4) and 1.1-kb (5, 6) segments of TGFa gene 5'-flanking sequence or with a construct containing the EGF receptor promoter sequence (1.1 kb) (1, 2). Twenty hours after transfection, the cells were treated with (2, 4, 6) or without (1, 3, 5) 10 nM TPA for 4 h, and luciferase activity in cell extracts was assayed. The tracings from the luminometer are shown for one of each of the duplicate samples. The integrated peak areas from the duplicates agreed within 7% of each other.
co-workers (22) that the EGF receptor gene homolog in the TGFa gene is also not sufficient to drive reporter expression. Together, these experiments suggest that EGF receptor and TGFa gene expression are mediated via distinct c/s-regulatory elements. The 5'-flanking region of the TGFa gene contains seven consensus sequences for AP-2 (CCCCAGGG) (28). Since AP-2 has been implicated as a mediator of the transcriptional responses to TPA in the metallothionein gene (28), these potential AP-2-binding sites in the TGFa gene could have contributed to the observed TPA responses. However, deletion of all seven of these sites had no observable effect on the TPA or EGF responses, indicating that AP-2 is not necessary for TPA/EGF induction of the TGFa gene. The TGFa gene also contains seven potential Sp1-binding sites (CCGCCC) (29), five of which are within the transcriptionally active 313-bp proximal segment of the gene. Deletion of the three most distal sites from this proximal segment results in its inactivation as both a basal and EGF-inducible promoter. While these results suggest that Sp1 might play a role in TGFa gene transcription, other, as yet unidentified, factors may be involved. The human TGFa gene, like the rat TGFa gene (24), does not contain any sequences bearing strong similarities to the consensus sequences for AP-1 or the steroid receptors. However, the 1.1-kb segment of the TGFa gene does confer glucocorticoid, retinoid, and thyroid hormone responsiveness of a magnitude similar to that reported for the EGF receptor gene (1.5- to 2-fold). The site of action of these hormones on the TGFa gene and
Fig. 5. Primer Extension Analysis of the 5'-End of Transcripts from the Native TGFa Gene (A) and the Chimeric TGFaLuciferase Transfected Gene (B) A, An antisense primer, corresponding to bases 217-235 relative to the TGFa translation start site was extended with reverse transcriptase, using MDA468 cell RNA as a template. A major band, 350 bp long, was detected. This band corresponds to a transcription initiation site that is 115 bases upstream of the translational start site. B, An antisense primer corresponding to bases 54-75 of the published luciferase sequence (30) was extended with reverse transcriptase using RNA isolated from MDA468 cells that had been transiently transfected with the TGFa-luciferase chimeric construct containing the 313-bp TGFa segment. The major extension product was 101 bp long and consisted of 45 bases derived from the luciferase sequence and 56 bases from the TGFa 5'flanking region. Since the transfected plasmid contained the TGFa sequences up-stream of - 5 9 , the 5'-end of the extension product corresponded to -115 relative to the translational start site.
the mechanism by which they mediate their effect on transcription need further investigation. Our experiments have demonstrated that a proximal 313-bp segment of 5'-flanking sequence within the TGFa gene is able to confer both EGF and TPA responsiveness to a heterologous reporter gene. This segment of the gene, which is particularly GC rich and lacks CCAAT or TATA boxes, is characteristic of genes normally classified as housekeeping genes. Such genes generally are neither highly regulated nor expressed in a tissue-specific fashion. Despite the structure of its 5'flanking sequence, however, TGFa expression is both regulatable and somewhat restricted in its tissue distribution. In the pituitary gland, its expression is confined to cells of GH-PRL lineage (5); in the ovary, to the thecal cells (7, 8); in the brain, to a limited array of neurons (10); and in the skin, to the keratinocytes (11). In mammary cancer cells, its expression is estrogen (14), TPA, and EGF responsive (19); in pituitary cells (6) and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
519
TGFa Gene Regulation
keratinocytes (11, 15) it is TPA and EGF responsive; and in the ovary, it is FSH responsive (7). This limited tissue distribution and regulatable expression is more in keeping with the controlled expression that is required of hormones and other regulatory peptides. Our elucidation of the proximal promoter-enhancer region of the TGFa gene will allow further studies into the regulation of expression of this gene.
MATERIALS AND METHODS Isolation of Genomic Clones
The final conditions used near-confluent 10-cm plates of cells. The cells were trypsinized and washed twice by centrifugation in DME containing 10% calf serum. The cells from each plate were then resuspended in 0.4 ml DME containing 20% calf serum, and the tube containing the cells was placed on ice. Plasmid DNA (40 ^g) was added to the cell suspension 5 min before electroporation. Electroporation of the entire cell sample was performed in a cuvette with an electrode gap of 4 mm at 350 V and 500 ^F. The electroporated cells were resuspended in 10 ml growth medium and distributed to 10 culture dishes (6 cm) which had been prefilled with 3 ml growth medium. This approach ensured that all cells in a given set of secondary plates were transfected at the same efficiency. Hormones were added to the cells at the indicated times after the electroporation. Luciferase Assay
6
Phage plaques (4 x 10 ) from a human genomic library (AGEM11, Promega Corp., Madison, Wl) were screened with a T4 polynucleotide kinase-labeled oligonucleotide corresponding to the first 24 bases of the TGFa-coding sequence, 5'GGTCCCCTCGGCTGGACAGCTCGC-3'. Ten clones were ultimately identified and plaque purified. The inserts were mapped by restriction digestion and probing with segments of the TGFa cDNA (10) and oligonucleotides corresponding to sequences previously described to be in the TGFa 5'-flanking region (5'-GAGCTCCGGGTACCTGG-3' and 5'TCGTCGGCCCGGGTGCCC-3') (23). One clone was determined to contain 13 kb of sequence 5' to the TGFa translation start site. Restriction fragments from this clone were subcloned into a plasmid (pT7T3). A 1.1-kb Sacl to Sacll fragment of this clone was sequenced using the Sequenase kit (U.S. Biochemical Corp., Cleveland, OH) and found to be identical to the genomic fragment described previously. This 1.1-kb fragment was subcloned into the H/ndlll site of the pXP1 luciferase expression vector (American Type Culture Collection, Rockville, MD) (25) using two double stranded oligonucleotide adapters that preserved the previously described (23) transcription initiation site and the 3' Sacll and 5' Sacl sites. The adapter used at the 3'-end had the following sequence: 5'-GGCGCCGCTCCGCCA-3' 3'-CGCCGCGGCGAGGCGGTTCGA-5'. This adapter brought the 3'-end of the TGFa promoter sequence to a position 1 bp short of the Aval site used as the 3'-limit of the TGFa segment in the previously described promoter-reporter construct (23). In addition, shorter and longer fragments of the gene, each having the same 3'-end (-59 relative to the A of the ATG start codon of the TGFacoding sequence), were subcloned into the expression vector. An 870-bp insert (—930 to -59) was generated by exonuclease III digestion of the 1.1-kb fragment from its 5'-end. The size of this fragment was determined by agarose gel electrophoresis after subcloning into the expression vector. The generation of the other inserts made use of convenient restriction sites at their 5'-ends (183-bp Alu\, 313-bp BamHI, 3.5-kb A/col). All plasmids used for transfection were purified twice by cesium chloride gradient centrifugation, quantified by UV absorption at 260 niu, and assessed for integrity by agarose gel electrophoresis. Cell Culture and Transfection The MDA468 cells were maintained in a CO2 incubator on 10cm Falcon tissue culture plates (Falcon Plastics, Oxnard, CA) in Dulbecco's Modified Eagle's Medium (DME) containing 10% bovine calf serum plus penicillin and gentamicin (19). Transfection of plasmid DNA was accomplished by electroporation using a Bio-Rad Gene Pulser (Bio-Rad, Richmond, CA). Electroporation conditions were optimized using an RSV-long terminal repeat-luciferase construct (pRSV/L) kindly provided by Dr. Donald Helinski (30), University of California-San Diego.
At the stated times after the addition of hormone, the cell culture plates were placed on ice. The monolayer was washed twice with cold PBS (calcium and magnesium free), and the cells were lysed on the plate with 0.3 ml of a 1 % Triton X-100 buffer, as described by Brasier et al. (31). A 200-M' aliquot of extract was assayed, as described by Brasier ef al. (31), in an LKB1250 luminometer (LKB, Rockville, MD). The luminometer signal was monitored by both the printer-integrator supplied with the luminometer and a Shimadzu C-R3A chromatographic plotter-integrator. The signal was integrated on the LKB apparatus for 10 sec to obtain a numerical value for the light intensity. The cell extracts were generally assayed within 30 min of cell lysis; however, if flash frozen, they could be stored at - 8 0 C for 2 weeks without appreciable loss of activity. All assays were performed on duplicate plates, and in a given experiment, agreement between samples was within 7%. Primer Extension Poly(A)+ RNA was prepared from MDA468 cells that had either not been transfected or had been transiently transfected with the TGFa-luciferase construct containing the 313-bp TGFa 5'flanking region. The RNA (10 fig) from the untransfected cells was annealed to 10 ng of a T4 polynucleotide kinase-labeled primer with the following antisense sequence: 5'-GTAAGACCCATGCAACCAC-3', corresponding to bases 217-235 (relative to the ATG start codon) of the TGFa cDNA. The RNA (10 ng) from the transfected cells was annealed to 10 ng labeled primer with the following antisense sequence: 5'CCTTCTGCGGI I I I IGTATTTC-3', corresponding to bases 54-75 of the luciferase cDNA (30). After annealing the primers on their corresponding RNA, primer extension was carried out using avian myeloblastosis virus reverse transcriptase. After phenol extraction and ethanol precipitation, the extension products were analyzed on a 6% DNA sequencing gel. The size of the reverse transcription product was determined by comparison with an M13 sequencing ladder generated using the Sequenase kit and [a-35S]dATP.
Acknowledgments We would like to thank Dr. Gordon Gill, University of CaliforniaSan Diego, for providing us with the sequence of the EGF receptor gene core element before its publication and for the EGF-receptor-luciferasechimeric construct used in these studies.
Received December 17, 1990. Revision received January 30,1991. Accepted February 1,1991. Address requests for reprints to: Jeffrey E. Kudlow, Department of Medicine, Division of Endocrinology and Metabolism, University of Alabama, UAB Station, Birmingham, Alabama 35294.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 520
REFERENCES 16. 1. Lyons RM, Moses HL 1990 Transforming growth factors and the regulation of cell proliferation. Eur J Biochem 187:467-473 2. Goustin AS, Leof EB, Shipley GD, Moses HL 1986 Growth factors and cancer. Cancer Res 46:1015-1029 3. Samsoondar J, Kobrin MS, Kudlow JE 1987 Alpha-transforming growth factor secreted by untransformed bovine anterior pituitary cells in culture. I. Purification from conditioned medium. J Biol Chem 261:14408-14413 4. Kobrin MS, Samsoondar J, Kudlow JE 1987 Alpha-transforming growth factor secreted by untransformed bovine anterior pituitary cells in culture. II. Indentification using a sequence specific monoclonal antibody. J Biol Chem 261:14414-14419 5. Kobrin MS, Asa SL, Samsoondar J, Kudlow JE 1987 Alpha-transforming growth factor in the bovine anterior pituitary gland: secretion by dispersed cells and immunohistochemical localization. Endocrinology 121:1412-1416 6. Mueller SG, Kobrin MS, Paterson AJ, Kudlow JE 1989 Transforming growth factor-alpha expression in the anterior pituitary gland: regulation by epidermal growth factor and phorbol ester in dispersed cells. Mol Endocrinol 3:976-983 7. Kudlow JE, Kobrin MS, Purchio AF, Twardzik DR, Hernandez FR, Asa SL, Adashi EY1987 Ovarian transforming growth factor-alpha gene expression: immunohistochemical localization to the theca-interstitial cells. Endocrinology 121:1577-1579 8. Lobb DK, Kobrin MS, Kudlow JE, Dorrington JH 1989 Transforming growth factor-alpha in the adult bovine ovary: identification in growth follicles. Biol Reprod 8:1087-1093 9. Wilcox JN, Derynck R 1988 Localization of cells synthesizing transforming growth factor-alpha mRNA in the mouse brain. J Neurosci 8:1910-1904 10. Kudlow JE, Leung AWC, Kobrin MS, Paterson AJ, Asa SL 1989 Transforming growth factor-a in the mammalian brain: immunohistochemical detection in neurons and characterization of its mRNA. J Biol Chem 264:38803883 11. Coffey Jr RJ, Derynck R, Wilcox JN, Bringman TS, Goustin AS, Moses HL, Pittelkow MR 1987 Production and auto-induction of transforming growth factor-alpha in human keratinocytes. Nature 328:817-820 12. Rappolee DA, Mark D, Banda MJ, Werb Z 1988 Wound macrophages express TGFa and other growth factors in vivo: analysis by mRNA phenotyping. Science 241:708712 13. Mueller SG, Paterson AJ, Kudlow JE 1990 Transforming growth factor in arterioles: cell surface processing of its precursor by elastases. Mol Cell Biol 10:4596-4602 14. Bates SE, Davidson NE, Valverius E, Freter CE, Dickson RB, Tarn JP, Kudlow JE, Lippman ME, Salomon DS 1988 Expression of transforming growth factor alpha and its messenger ribonucleic acid in human breast cancer; its regulation by estorgen and its possible functional significance. Mol Endocrinol 6:543-555 15. Pittelkow MR, Lindquist PB, Abraham RT, Graves-Deal R, Derynck R, Coffey Jr RJ 1989 Induction of transforming
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
growth factor-alpha expression in human keratinocytes by phorbol esters. J Biol Chem 264:5164-5171 Bjorge JD, Paterson AJ, Kudlow JE 1989 Phorbol ester and epidermal growth factor (EGF) stimulate the concurrent accumulation of mRNA for the EGF-receptor and its ligand transforming growth factor-alpha in a human breast cancer cell line. J Biol Chem 264:4021-4027 Bjorge JD, Kudlow JE, Mills GB, Paterson AJ 1989 Inhibition of stimulus-dependent epidermal growth factor receptor and transforming growth factor-alpha mRNA accumulation by the protein kinase C inhibitor staurosporine. FEBS Lett 243:404-408 Kudlow JE, Cheung C-YM, Bjorge JD 1986 Epidermal growth factor stimulates the synthesis of its own receptor in a human breast cancer cell line. J Biol Chem 261:41344138 Bjorge JD, Kudlow, JE 1987 Epidermal growth factor receptor synthesis is stimulated by phorbol ester and epidermal growth factor: evidence for a common mechanism. J Biol Chem 262:6615-6622 Kageyama R, Merlino GT, Pastan I 1988 A transcription factor active on the epidermal growth factor receptor gene. Proc Natl Acad Sci USA 85:5016-5020 Hudson LG, Santon JB, Gill GN 1989 Regulation of epidermal growth factor gene expression. Mol Endocrinol 3:400-408 Hudson LG, Thompson KL, Xu J, Gill GN 1990 Indentification and characterization of a regulated promoter element in the epidermal growth factor receptor gene. Proc Natl Acad Sci USA 87:7536-7540 Jakobovits EB, Schlokat U, Vannice JL, Derynck R, Levinson AD 1988 The human transforming growth factor alpha promoter directs transcription from a single site in the absence of a TATA sequence. Mol Cell Biol 8:55495554 Blasband AJ, Rogers KT, Chen X, Azizkhan JC, Lee DC 1990 Characterization of the rat transforming growth factor alpha gene and identification of promoter sequences. Mol Cell Biol 10:2111-2121 Reynolds GA, Basu SK, Osborne TF, Chin DJ, Gil G, Brown MS, Goldstein JL, Luskey 1984 HMG CoA reductase: a negatively regulated gene with unusual promoter and 5' untranslated regions. Cell 38:275-285 Ueda K, Pastan I, Gottesman MM 1987 Isolation and sequence of the promoter region of the human multidrugresistance (P-glycoprotein) gene. J Biol Chem 262:1743217436 Ishii S, Merlino GT, Pastan 11985 Promoter region of the human Harvey ras proto-oncogene: similarity to the EGF receptor proto-oncogene promoter. Science 230:13781381 Imagawa MR, Chiu R, Karin M 1987 Transcription factor AP-2 mediates induction by two different signal-transduction pathways: protein kinase C and cAMP. Cell 51:251260 Kadonaga JT, Jones KA, Tjian R 1986 Promoter-specific activation of RNA polymerase II transcription by Sp1. Trends Biochem Sci 11:20-23 DeWet JR, Wood KV, DeLuca M, Helinski DR, Subramani S 1987 Firefly luciferase gene: Structure and expression in mammalian cells. Mol Cell Biol 7:725-737 Brasier AR, Tate JE, Habener JF 1989 Optimum use of the firefly luciferase assay as a reporter gene in mammalian cell lines. Biotechnique 7:1116-1122
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Rampyari H. Raja, Andrew J. Paterson, Tae Ho Shin, and Jeffrey E. Kudlow Departments of Medicine and Cell Biology University of Alabama Birmingham, Alabama 35294
INTRODUCTION
We and others have previously reported that transforming growth factor-a (TGFa) expression is hormonally responsive and its expression is coregulated with that of its receptor [the epidermal growth factor (EGF) receptor]. The 5'-flanking region of the TGFa gene was characterized to determine whether it could confer hormone responsiveness to a reporter gene (luciferase) in human mammary carcinoma cells (MDA468). This segment of the gene is GC rich and contains an element strikingly similar to the core element of the EGF receptor gene that has been shown to mediate both basal and hormonestimulated expression of the EGF receptor. We now report that a 313-basepair (bp) proximal element of the TGFa 5'-flanking region (-373 to - 5 9 relative to the TGFa translation start codon) is capable of conferring responses to phorbol ester and EGF. This gene segment does not contain the EGF receptor gene homolog or potential AP-2-binding sites, suggesting that these elements are not necessary for basal and EGF- or phorbol ester-responsive TGFa gene expression. This 313-bp proximal element also confers proper transcriptional initiation to the chimeric TGFa-luciferase reporter construct, indicating it is the TGFa promoter. A 1.1-kilobase segment of the TGFa 5'-flanking region also confers retinoic acid, thyroid hormone, and glucocorticoid responsiveness despite the absence of recognizable steroid hormone receptor-binding sites. These hormones stimulate reporter expression 1.5- to 2-fold in a dose-dependent manner. Extension of the 5'flanking region to -3500 results in marked suppression of reporter gene expression. These results indicate that the TGFa gene 5'-flanking sequence contains the elements responsible for hormonal responsiveness of this gene and that these elements are distinct from those that regulate the expression of the EGF receptor gene. (Molecular Endocrinology 5: 514-520, 1991)
Transforming growth factor-a (TGFa), a homolog of epidermal growth factor (EGF), uses the EGF receptor to exert its effects on cells (see Ref. 1 for review). While originally believed to be exclusively expressed in oncofetal tissues (2), TGFa has since been shown to be present in a variety of normal adult tissues, including the anterior pituitary gland (3-6), ovary (7, 8), brain (9, 10), skin (11), macrophages (12), and vascular smooth muscle cells (13). In each of these tissues, TGFa expression can potentially be controlled at either the transcriptional (6, 7, 11) or posttranslational (13) level. Hormones are known to be among those factors capable of altering TGFa expression. Estrogen (14), phorbol ester (6, 15, 16), and EGF (6, 11, 16) have each been shown to stimulate TGFa mRNA accumulation in a variety of cell types. We have previously focused on anterior pituitary (6) and mammary carcinoma cells (1719), in which we showed that EGF and phorbol ester simultaneously stimulate EGF receptor and TGFa mRNA accumulation. The simultaneous increase in receptor and ligand expression in the same cell could have a multiplicative effect on signal transduction through this receptor system. While considerable progress has been made in defining the locus in the EGF receptor gene that mediates responsiveness to hormones (20-22), the same is not the case for the TGFa gene. While the 5'-flanking region of both the human (23) and rat (24) TGFa genes has been cloned and shown to drive basal reporter gene expression, the presence of hormonal responsiveness in these TGFa genes has not been reported. Striking similarities exist in the 5'-flanking regions of the EGF receptor and TGFa genes. Both genes are very GC rich, contain potential AP-2- and Sp1-binding sites, and lack a TATA box. In addition, an EGF receptor gene core element, as defined by Hudson and co-workers (22) mediates basal and hormonal-stimulated transcription. The existence of a highly homologous element within the 5'-flanking region of the TGFa gene and the coordinate regulation of the EGF receptor and TGFa genes in many cells suggested that these genes shared a similar core element involved in hormonal regulation.
0888-8809/91/0514-0520$03.00/0 Molecular Endocrinology Copyright © 1991 by The Endocrine Society
514
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
515
TGFa Gene Regulation
However, we have cloned 13 kilobases (kb) of the 5'flanking region of the TGFa gene and show in this paper that the proximal 313-basepair (bp) segment of the gene, devoid of the EGF receptor gene homolog, is capable of conferring basal expression and EGF and TPA responsiveness to a luciferase reporter gene. These studies establish that these agents mediate their transcriptional effects on TGFa gene expression through a regulatory element distinct from the element involved in EGF receptor gene expression.
TPA
JVhJY EGF
RESULTS C
C
1
2
3
RA
C
1
2
T3
JM\J\J\M\JV C
C
1
2
3
4
5
C
C
1
2
3
4
5
DEX
B
An
Rl Amp
TGFcr 5' Region Fig. 1. A, Effects of Various Concentrations of TPA, EGF, Retinoic Acid, Thyroid Hormone, and Dexamethasone on TGFa Promoter Expression MDA468 cells were transiently transfected with a construct containing 1.1 kb of the TGFa 5'-flanking region up-stream of the luciferase reporter gene. The transfected pool of cells was divided equally onto 6-cm culture dishes and, 40 h later, treated for 5 h with the stated hormones before harvest.
Preliminary experiments determined that the 1.1-kb fragment of the TGFa 5'-flanking region could direct luciferase expression in the MDA468 cells. The transfected cells were then tested to determine if reporter expression was hormone responsive. We had previously shown by Northern blotting that the TGFa mRNA content of MDA468 cells increased in response to EGF and phorbol ester [12-O-tetradecanoylphorbol-13-acetate (TPA)] (19). Furthermore, the similarity in the structures of the EGF receptor and TGFa genes suggested that the TGFa gene might confer responses to steroid hormones similar to those of the EGF receptor gene (21). In concert with these earlier observations, exposure of the transfected cells to EGF and TPA for 5 h resulted in a dose-dependent increase in cellular luciferase activity (Fig. 1A). Optimal exposure to TPA was 4-5 h, with a decline in activity to subbasal levels after 24 h, whereas EGF exposure resulted in a continuously rising luciferase activity which had not reached a plateau even by 48 h (data not shown). Integration of the light signals revealed 4- to 5-fold stimulations of reporter activity after a 5-h exposure to maximally effective doses of EGF (1 nM) and TPA (10 nM). We also tested the transient transfectants for responsiveness to retinoic acid, thyroid hormone (T3), and the glucocorticoid dexamethasone. In each case, there was a small (1.5to 2-fold) reproducible increase in reporter activity at doses of 10 nM for retinoic acid, 1 nM for thyroid hormone, and 0.5 nM for dexamethasone. These experiments were repeated in duplicate on three separate occasions and gave the same qualitative results. These
Luciferase activity was assayed in a solubilized extract of the cells. Shown is a chart recorder tracing of the luminometer output (45 sec) for each dose of hormone for a typical experiment. In each case, C designates the unstimulated control. The doses of hormone were as follows: TPA, 1 = 0.5 nM, 2 = 1 nM, 3 = 10 nM, 4 = 300 nM; EGF, 1 = 0.1 nM, 2 = 0.5 nM, 3 = 1 nM, 4 = 3 nM, 5 = 30 nM; retinoic acid, 1 = 1.0 nM, 2 = 10 nM, 3 = 100 nM, 4 = 1 HM; thyroid hormone, 1 = 0.1 nM, 2 = 0.5 nM, 3 = 1 nM, 4 = 10 nM, 5 = 30 nM; dexamethasone, 1 = 0.5 nM, 2 = 1 nM, 3 = 10 nM, 4 = 30 nM, 5 = 100 nM. The peaks were integrated for 10 sec, and the areas did not vary by more than 7% for each replicate point. B, The pXP1 mammalian expression vector into which the TGFa Gene segments were inserted.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 516
results imply that the previously observed accumulation of TGFa mRNA in MDA468 cells in response to EGF and TPA results from increased transcription of this gene. The finding that the gene is also responsive to thyroid hormone, retinoic acid, and glucocorticoid is further indication of the hormonal responsiveness of the TGFa gene. The responses of the TGFa gene to these steroid hormones occurred at similar doses and were of similar magnitude to those previously reported for the EGF receptor gene (21). Our earlier studies had shown simultaneous stimulation of EGF receptor and TGFa gene expression by EGF and TPA, prompting us to search for common regulatory elements. A computer comparison of the genes demonstrated the presence of a strikingly homologous 26-bp segment, of which 20 bases were identical, and five of the six nonidentical bases were GC interchanges (Fig. 2). This homology took on possible functional significance when Hudson and co-workers (22) recently showed that this segment in the EGF receptor gene was the core element that conferred hormonal responsiveness to that gene. To test whether the EGF receptor gene-like element in the TGFa gene was necessary for EGF and TPA responsiveness, we assessed reporter constructs that contained or lacked this segment. We found that the reporter construct containing the 313-bp insert (-373 to - 5 9 relative to the TGFa translation start site), which deletes the entire EGF receptor gene-like element, responded to EGF (Fig. 3A) and TPA (Fig. 4) with 22- and 4-fold increments, respectively. There are also seven potential AP2 sites in the 1.1-kb segment of the the TGFa 5'flanking region (Fig. 3B) that could confer EGF/TPA responsiveness. However, the 313-bp fragment of the TGFa gene does not contain any of these AP-2 sites. The ability of this proximal segment to confer basal and EGF/TPA-stimulated reporter expression suggests that AP-2 and the factor(s) that binds to EGF receptor gene core element are not necessary for basal or EGF/TPAstimulated TGFa expression. We could detect no reporter activity in the construct containing the additional 2.4 kb of distal 5'-flanking sequence even after EGF stimulation (Fig. 3A). Biasband and co-workers (24) have reported a similar find-
Homologous Segments in 5'-Flanking Regions of the EGF Receptor and TGFa Genes -410
-400
-390
TTTTCGTCGGCCCGGGTGCCCGGCTCGCGCC CGCGTCCGCCCGAGTCCCCGCCTCGCCGCCA -110
-100
TGFa EGF-R
-90
Fig. 2. A Comparison of the EGF Receptor Gene Core Element with a Similar Segment in the TGFa Gene The positions shown above and below the bases indicate the distance from the A in the ATG start codon of both genes.
ing in the rat TGFa gene, in that a 1.9-kb fragment of 5'-flanking region of the rat gene conferred no basal activity to the reporter. These results suggest the presence of an up-stream suppressor sequence whose functional significance and location in these gene segments have yet to be determined. We could also detect neither basal nor EGF-stimulated activity in the reporter construct containing the 183-bp (-242 to -59) TGFa gene segment (Fig. 3A), suggesting that the sequence up-stream of the Alu\ site at -242 bp relative to the translational start site is required for basal and EGFstimulated transcription. The luciferase expression is specifically driven by the TGFa 5'-flanking sequence contained in the expression plasmids, as indicated by the observation that promoterless constructs using the parental pXP1 plasmid and the construct containing the -242 to - 5 9 segment of the TGFa gene showed no detectable luciferase activity in the transiently transfected cells. The responses to steroid hormones (Fig. 1A) are also specific, because the construct containing the 313-bp insert (-373 to -59), while EGF/TPA responsive, was nonresponsive to retinoic acid, thyroid hormone, and dexamethasone (data not shown). The failure of the 313-bp insert to confer steroid hormone responsiveness suggests that the observed responses from the 1.1-kb insert were transcriptional and not due to stabilization of the luciferase mRNA. While the receptors for these steroids have not been directly detected in MDA468 cells, the transcriptional responses suggest that they are indeed present. Specificity of the responses is also ensured by the design of pXPI. This plasmid contains two polyadenylation signals up-stream of the promoter insertion site which are designed to terminate any nonspecifically initiated transcripts before they extend into the luciferase reporter sequence (25). This vector also does not appear to contain cryptic enhancer sequences, since constitutive promoters are not rendered hormone responsive (21). Transfection of a Rous sarcoma virus (RSV)-long terminal repeat-luciferase chimera resulted in luciferase activity 20-fold higher than the basal activity seen with the 1.1 -kb fragment of the TGFa gene (data not shown). However, the EGF receptor promoter (1.1 kb) drove expression of the luciferase reporter to levels similar to those seen with the 1.1-kb fragment of the TGFa gene (Fig. 4) and responded similarly to TPA, suggesting that the enhancer-promoter activity of the TGFa gene segment is similar in potency in these cells to that in the corresponding EGF receptor gene segment. The ability of the 5'-flanking region of the human TGFa gene to confer EGF and TPA responsiveness to the luciferase reporter gene similar to that seen for the endogenous TGFa gene suggests that this region of the gene contains the key enhancers and the promoter. Further support derives from the prior observation (23) that the 5'-flanking region of the TGFa gene confers initiation to a chimeric TGFa-chloramphenicol acetyltransferase construction which corresponds to the native gene. Indeed, using an essentially identical TGFa
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
517
TGFa Gene Regulation
Insert Limits
• -EGF
Nil
E+EGF
-242 to -59 -373 to -59 -930 to -59 -1139 to -59 -3500 to -59
500
1000 1500 2000
2500
Relative Luciferase Activity
-3500
-1139 -930
-373 -242
o
Potential AP2 binding sites
M
EGF Receptor Gene Homologue
Luciferase
Fig. 3. A, EGF Responsiveness of Various Segments of the TGFa Gene MDA468 cells were transfected, as described in Fig. 1, with constructs containing various segments of the TGFa gene. All TGFa gene segments had the same 3'-end, corresponding to - 5 9 relative to the translation start site and extending 5' of this site to the designated position shown in the figure. EGF was added to the cells 20 h after transfection. Plotted is the mean of duplicate determinations of the numerical value of the integrated signal from the luminometer resulting from assays of cell extracts that were derived from cells that were treated with or without 1 nM EGF for 48 h, as indicated on the figure. B, The construction of the TGFaluciferase chimeric plasmid is shown. The positions of the EGF receptor gene homolog and potential AP-2-binding sites are shown in relation to the positions in the gene to which the 5'-deletions extended.
gene segment in a chimeric TGFa-luciferase construct, a dominant transcription initiation site was also detected by primer extension. This site (at position -115 relative to the TGFa translation start site) was in agreement with the 5'-end of the transcript from the native gene in MDA468 cells (Fig. 5). These primer extension data were also confirmed by RNAse protection, using a riboprobe spanning the junction between the TGFa (-373) and luciferase (99) genes (data not shown). Other minor sites were detected on longer exposures of the gels. The TGFa gene is probably capable of initiating transcription from multiple sites, consistent with other genes whose promoters are CG rich and contain no TATA box (24-27). Under different circumstances, one or more of the potential initiation sites may dominate.
DISCUSSION In previous studies we have shown that the expression of TGFa and its receptor (EGF receptor) are coordi-
nately regulated in MDA468 cells by EGF in a manner that appears to depend on protein kinase-C (16, 18, 19). With the recent elucidation of the sequences of the promoter regions of both of these genes (20-24), it became evident that there were common structural features, raising the hypothesis that the coordinate regulation of these genes could result from these structural similarities. Hudson and co-workers (22) have determined that a 36-bp element in the EGF receptor gene mediates responses to EGF and TPA. We found a highly homologous sequence in the TGFa gene. This EGF receptor gene-like element then became an attractive candidate for the EGF/TPA-responsive element in the TGFa gene. In this study we clearly demonstrate that deletion of this EGF receptor gene homolog from the TGFa 5'-flanking sequence does not alter the ability of the TGFa promoter to confer basal and EGF/TPAstimulated luciferase expression from the transiently transfected chimeric construct. Thus, this element in the TGFa gene is not necessary for TGFa expression. This result is consistent with the finding of Hudson and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 518
-350
"S
1 2
3
4
5
.
-101
6
Fig. 4. TPA Responsiveness of Various Segments of the TGFa Gene Compared to the EGF Receptor Gene MDA468 cells were transfected, as described in Fig. 1, with constructs containing the 313-bp (3, 4) and 1.1-kb (5, 6) segments of TGFa gene 5'-flanking sequence or with a construct containing the EGF receptor promoter sequence (1.1 kb) (1, 2). Twenty hours after transfection, the cells were treated with (2, 4, 6) or without (1, 3, 5) 10 nM TPA for 4 h, and luciferase activity in cell extracts was assayed. The tracings from the luminometer are shown for one of each of the duplicate samples. The integrated peak areas from the duplicates agreed within 7% of each other.
co-workers (22) that the EGF receptor gene homolog in the TGFa gene is also not sufficient to drive reporter expression. Together, these experiments suggest that EGF receptor and TGFa gene expression are mediated via distinct c/s-regulatory elements. The 5'-flanking region of the TGFa gene contains seven consensus sequences for AP-2 (CCCCAGGG) (28). Since AP-2 has been implicated as a mediator of the transcriptional responses to TPA in the metallothionein gene (28), these potential AP-2-binding sites in the TGFa gene could have contributed to the observed TPA responses. However, deletion of all seven of these sites had no observable effect on the TPA or EGF responses, indicating that AP-2 is not necessary for TPA/EGF induction of the TGFa gene. The TGFa gene also contains seven potential Sp1-binding sites (CCGCCC) (29), five of which are within the transcriptionally active 313-bp proximal segment of the gene. Deletion of the three most distal sites from this proximal segment results in its inactivation as both a basal and EGF-inducible promoter. While these results suggest that Sp1 might play a role in TGFa gene transcription, other, as yet unidentified, factors may be involved. The human TGFa gene, like the rat TGFa gene (24), does not contain any sequences bearing strong similarities to the consensus sequences for AP-1 or the steroid receptors. However, the 1.1-kb segment of the TGFa gene does confer glucocorticoid, retinoid, and thyroid hormone responsiveness of a magnitude similar to that reported for the EGF receptor gene (1.5- to 2-fold). The site of action of these hormones on the TGFa gene and
Fig. 5. Primer Extension Analysis of the 5'-End of Transcripts from the Native TGFa Gene (A) and the Chimeric TGFaLuciferase Transfected Gene (B) A, An antisense primer, corresponding to bases 217-235 relative to the TGFa translation start site was extended with reverse transcriptase, using MDA468 cell RNA as a template. A major band, 350 bp long, was detected. This band corresponds to a transcription initiation site that is 115 bases upstream of the translational start site. B, An antisense primer corresponding to bases 54-75 of the published luciferase sequence (30) was extended with reverse transcriptase using RNA isolated from MDA468 cells that had been transiently transfected with the TGFa-luciferase chimeric construct containing the 313-bp TGFa segment. The major extension product was 101 bp long and consisted of 45 bases derived from the luciferase sequence and 56 bases from the TGFa 5'flanking region. Since the transfected plasmid contained the TGFa sequences up-stream of - 5 9 , the 5'-end of the extension product corresponded to -115 relative to the translational start site.
the mechanism by which they mediate their effect on transcription need further investigation. Our experiments have demonstrated that a proximal 313-bp segment of 5'-flanking sequence within the TGFa gene is able to confer both EGF and TPA responsiveness to a heterologous reporter gene. This segment of the gene, which is particularly GC rich and lacks CCAAT or TATA boxes, is characteristic of genes normally classified as housekeeping genes. Such genes generally are neither highly regulated nor expressed in a tissue-specific fashion. Despite the structure of its 5'flanking sequence, however, TGFa expression is both regulatable and somewhat restricted in its tissue distribution. In the pituitary gland, its expression is confined to cells of GH-PRL lineage (5); in the ovary, to the thecal cells (7, 8); in the brain, to a limited array of neurons (10); and in the skin, to the keratinocytes (11). In mammary cancer cells, its expression is estrogen (14), TPA, and EGF responsive (19); in pituitary cells (6) and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
519
TGFa Gene Regulation
keratinocytes (11, 15) it is TPA and EGF responsive; and in the ovary, it is FSH responsive (7). This limited tissue distribution and regulatable expression is more in keeping with the controlled expression that is required of hormones and other regulatory peptides. Our elucidation of the proximal promoter-enhancer region of the TGFa gene will allow further studies into the regulation of expression of this gene.
MATERIALS AND METHODS Isolation of Genomic Clones
The final conditions used near-confluent 10-cm plates of cells. The cells were trypsinized and washed twice by centrifugation in DME containing 10% calf serum. The cells from each plate were then resuspended in 0.4 ml DME containing 20% calf serum, and the tube containing the cells was placed on ice. Plasmid DNA (40 ^g) was added to the cell suspension 5 min before electroporation. Electroporation of the entire cell sample was performed in a cuvette with an electrode gap of 4 mm at 350 V and 500 ^F. The electroporated cells were resuspended in 10 ml growth medium and distributed to 10 culture dishes (6 cm) which had been prefilled with 3 ml growth medium. This approach ensured that all cells in a given set of secondary plates were transfected at the same efficiency. Hormones were added to the cells at the indicated times after the electroporation. Luciferase Assay
6
Phage plaques (4 x 10 ) from a human genomic library (AGEM11, Promega Corp., Madison, Wl) were screened with a T4 polynucleotide kinase-labeled oligonucleotide corresponding to the first 24 bases of the TGFa-coding sequence, 5'GGTCCCCTCGGCTGGACAGCTCGC-3'. Ten clones were ultimately identified and plaque purified. The inserts were mapped by restriction digestion and probing with segments of the TGFa cDNA (10) and oligonucleotides corresponding to sequences previously described to be in the TGFa 5'-flanking region (5'-GAGCTCCGGGTACCTGG-3' and 5'TCGTCGGCCCGGGTGCCC-3') (23). One clone was determined to contain 13 kb of sequence 5' to the TGFa translation start site. Restriction fragments from this clone were subcloned into a plasmid (pT7T3). A 1.1-kb Sacl to Sacll fragment of this clone was sequenced using the Sequenase kit (U.S. Biochemical Corp., Cleveland, OH) and found to be identical to the genomic fragment described previously. This 1.1-kb fragment was subcloned into the H/ndlll site of the pXP1 luciferase expression vector (American Type Culture Collection, Rockville, MD) (25) using two double stranded oligonucleotide adapters that preserved the previously described (23) transcription initiation site and the 3' Sacll and 5' Sacl sites. The adapter used at the 3'-end had the following sequence: 5'-GGCGCCGCTCCGCCA-3' 3'-CGCCGCGGCGAGGCGGTTCGA-5'. This adapter brought the 3'-end of the TGFa promoter sequence to a position 1 bp short of the Aval site used as the 3'-limit of the TGFa segment in the previously described promoter-reporter construct (23). In addition, shorter and longer fragments of the gene, each having the same 3'-end (-59 relative to the A of the ATG start codon of the TGFacoding sequence), were subcloned into the expression vector. An 870-bp insert (—930 to -59) was generated by exonuclease III digestion of the 1.1-kb fragment from its 5'-end. The size of this fragment was determined by agarose gel electrophoresis after subcloning into the expression vector. The generation of the other inserts made use of convenient restriction sites at their 5'-ends (183-bp Alu\, 313-bp BamHI, 3.5-kb A/col). All plasmids used for transfection were purified twice by cesium chloride gradient centrifugation, quantified by UV absorption at 260 niu, and assessed for integrity by agarose gel electrophoresis. Cell Culture and Transfection The MDA468 cells were maintained in a CO2 incubator on 10cm Falcon tissue culture plates (Falcon Plastics, Oxnard, CA) in Dulbecco's Modified Eagle's Medium (DME) containing 10% bovine calf serum plus penicillin and gentamicin (19). Transfection of plasmid DNA was accomplished by electroporation using a Bio-Rad Gene Pulser (Bio-Rad, Richmond, CA). Electroporation conditions were optimized using an RSV-long terminal repeat-luciferase construct (pRSV/L) kindly provided by Dr. Donald Helinski (30), University of California-San Diego.
At the stated times after the addition of hormone, the cell culture plates were placed on ice. The monolayer was washed twice with cold PBS (calcium and magnesium free), and the cells were lysed on the plate with 0.3 ml of a 1 % Triton X-100 buffer, as described by Brasier et al. (31). A 200-M' aliquot of extract was assayed, as described by Brasier ef al. (31), in an LKB1250 luminometer (LKB, Rockville, MD). The luminometer signal was monitored by both the printer-integrator supplied with the luminometer and a Shimadzu C-R3A chromatographic plotter-integrator. The signal was integrated on the LKB apparatus for 10 sec to obtain a numerical value for the light intensity. The cell extracts were generally assayed within 30 min of cell lysis; however, if flash frozen, they could be stored at - 8 0 C for 2 weeks without appreciable loss of activity. All assays were performed on duplicate plates, and in a given experiment, agreement between samples was within 7%. Primer Extension Poly(A)+ RNA was prepared from MDA468 cells that had either not been transfected or had been transiently transfected with the TGFa-luciferase construct containing the 313-bp TGFa 5'flanking region. The RNA (10 fig) from the untransfected cells was annealed to 10 ng of a T4 polynucleotide kinase-labeled primer with the following antisense sequence: 5'-GTAAGACCCATGCAACCAC-3', corresponding to bases 217-235 (relative to the ATG start codon) of the TGFa cDNA. The RNA (10 ng) from the transfected cells was annealed to 10 ng labeled primer with the following antisense sequence: 5'CCTTCTGCGGI I I I IGTATTTC-3', corresponding to bases 54-75 of the luciferase cDNA (30). After annealing the primers on their corresponding RNA, primer extension was carried out using avian myeloblastosis virus reverse transcriptase. After phenol extraction and ethanol precipitation, the extension products were analyzed on a 6% DNA sequencing gel. The size of the reverse transcription product was determined by comparison with an M13 sequencing ladder generated using the Sequenase kit and [a-35S]dATP.
Acknowledgments We would like to thank Dr. Gordon Gill, University of CaliforniaSan Diego, for providing us with the sequence of the EGF receptor gene core element before its publication and for the EGF-receptor-luciferasechimeric construct used in these studies.
Received December 17, 1990. Revision received January 30,1991. Accepted February 1,1991. Address requests for reprints to: Jeffrey E. Kudlow, Department of Medicine, Division of Endocrinology and Metabolism, University of Alabama, UAB Station, Birmingham, Alabama 35294.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
Vol 5 No. 4
MOL ENDO-1991 520
REFERENCES 16. 1. Lyons RM, Moses HL 1990 Transforming growth factors and the regulation of cell proliferation. Eur J Biochem 187:467-473 2. Goustin AS, Leof EB, Shipley GD, Moses HL 1986 Growth factors and cancer. Cancer Res 46:1015-1029 3. Samsoondar J, Kobrin MS, Kudlow JE 1987 Alpha-transforming growth factor secreted by untransformed bovine anterior pituitary cells in culture. I. Purification from conditioned medium. J Biol Chem 261:14408-14413 4. Kobrin MS, Samsoondar J, Kudlow JE 1987 Alpha-transforming growth factor secreted by untransformed bovine anterior pituitary cells in culture. II. Indentification using a sequence specific monoclonal antibody. J Biol Chem 261:14414-14419 5. Kobrin MS, Asa SL, Samsoondar J, Kudlow JE 1987 Alpha-transforming growth factor in the bovine anterior pituitary gland: secretion by dispersed cells and immunohistochemical localization. Endocrinology 121:1412-1416 6. Mueller SG, Kobrin MS, Paterson AJ, Kudlow JE 1989 Transforming growth factor-alpha expression in the anterior pituitary gland: regulation by epidermal growth factor and phorbol ester in dispersed cells. Mol Endocrinol 3:976-983 7. Kudlow JE, Kobrin MS, Purchio AF, Twardzik DR, Hernandez FR, Asa SL, Adashi EY1987 Ovarian transforming growth factor-alpha gene expression: immunohistochemical localization to the theca-interstitial cells. Endocrinology 121:1577-1579 8. Lobb DK, Kobrin MS, Kudlow JE, Dorrington JH 1989 Transforming growth factor-alpha in the adult bovine ovary: identification in growth follicles. Biol Reprod 8:1087-1093 9. Wilcox JN, Derynck R 1988 Localization of cells synthesizing transforming growth factor-alpha mRNA in the mouse brain. J Neurosci 8:1910-1904 10. Kudlow JE, Leung AWC, Kobrin MS, Paterson AJ, Asa SL 1989 Transforming growth factor-a in the mammalian brain: immunohistochemical detection in neurons and characterization of its mRNA. J Biol Chem 264:38803883 11. Coffey Jr RJ, Derynck R, Wilcox JN, Bringman TS, Goustin AS, Moses HL, Pittelkow MR 1987 Production and auto-induction of transforming growth factor-alpha in human keratinocytes. Nature 328:817-820 12. Rappolee DA, Mark D, Banda MJ, Werb Z 1988 Wound macrophages express TGFa and other growth factors in vivo: analysis by mRNA phenotyping. Science 241:708712 13. Mueller SG, Paterson AJ, Kudlow JE 1990 Transforming growth factor in arterioles: cell surface processing of its precursor by elastases. Mol Cell Biol 10:4596-4602 14. Bates SE, Davidson NE, Valverius E, Freter CE, Dickson RB, Tarn JP, Kudlow JE, Lippman ME, Salomon DS 1988 Expression of transforming growth factor alpha and its messenger ribonucleic acid in human breast cancer; its regulation by estorgen and its possible functional significance. Mol Endocrinol 6:543-555 15. Pittelkow MR, Lindquist PB, Abraham RT, Graves-Deal R, Derynck R, Coffey Jr RJ 1989 Induction of transforming
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
growth factor-alpha expression in human keratinocytes by phorbol esters. J Biol Chem 264:5164-5171 Bjorge JD, Paterson AJ, Kudlow JE 1989 Phorbol ester and epidermal growth factor (EGF) stimulate the concurrent accumulation of mRNA for the EGF-receptor and its ligand transforming growth factor-alpha in a human breast cancer cell line. J Biol Chem 264:4021-4027 Bjorge JD, Kudlow JE, Mills GB, Paterson AJ 1989 Inhibition of stimulus-dependent epidermal growth factor receptor and transforming growth factor-alpha mRNA accumulation by the protein kinase C inhibitor staurosporine. FEBS Lett 243:404-408 Kudlow JE, Cheung C-YM, Bjorge JD 1986 Epidermal growth factor stimulates the synthesis of its own receptor in a human breast cancer cell line. J Biol Chem 261:41344138 Bjorge JD, Kudlow, JE 1987 Epidermal growth factor receptor synthesis is stimulated by phorbol ester and epidermal growth factor: evidence for a common mechanism. J Biol Chem 262:6615-6622 Kageyama R, Merlino GT, Pastan I 1988 A transcription factor active on the epidermal growth factor receptor gene. Proc Natl Acad Sci USA 85:5016-5020 Hudson LG, Santon JB, Gill GN 1989 Regulation of epidermal growth factor gene expression. Mol Endocrinol 3:400-408 Hudson LG, Thompson KL, Xu J, Gill GN 1990 Indentification and characterization of a regulated promoter element in the epidermal growth factor receptor gene. Proc Natl Acad Sci USA 87:7536-7540 Jakobovits EB, Schlokat U, Vannice JL, Derynck R, Levinson AD 1988 The human transforming growth factor alpha promoter directs transcription from a single site in the absence of a TATA sequence. Mol Cell Biol 8:55495554 Blasband AJ, Rogers KT, Chen X, Azizkhan JC, Lee DC 1990 Characterization of the rat transforming growth factor alpha gene and identification of promoter sequences. Mol Cell Biol 10:2111-2121 Reynolds GA, Basu SK, Osborne TF, Chin DJ, Gil G, Brown MS, Goldstein JL, Luskey 1984 HMG CoA reductase: a negatively regulated gene with unusual promoter and 5' untranslated regions. Cell 38:275-285 Ueda K, Pastan I, Gottesman MM 1987 Isolation and sequence of the promoter region of the human multidrugresistance (P-glycoprotein) gene. J Biol Chem 262:1743217436 Ishii S, Merlino GT, Pastan 11985 Promoter region of the human Harvey ras proto-oncogene: similarity to the EGF receptor proto-oncogene promoter. Science 230:13781381 Imagawa MR, Chiu R, Karin M 1987 Transcription factor AP-2 mediates induction by two different signal-transduction pathways: protein kinase C and cAMP. Cell 51:251260 Kadonaga JT, Jones KA, Tjian R 1986 Promoter-specific activation of RNA polymerase II transcription by Sp1. Trends Biochem Sci 11:20-23 DeWet JR, Wood KV, DeLuca M, Helinski DR, Subramani S 1987 Firefly luciferase gene: Structure and expression in mammalian cells. Mol Cell Biol 7:725-737 Brasier AR, Tate JE, Habener JF 1989 Optimum use of the firefly luciferase assay as a reporter gene in mammalian cell lines. Biotechnique 7:1116-1122
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 17:52 For personal use only. No other uses without permission. . All rights reserved.
