HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 40, 30765-30773, October 6, 2000

This Article Abstract Full Text (PDF) An addition or correction has been published All Versions of this Article: 275/40/30765    most recent M000039200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yue, J. Articles by Mulder, K. M. Search for Related Content PubMed PubMed Citation Articles by Yue, J. Articles by Mulder, K. M. Social Bookmarking                      What's this?

Requirement of Ras/MAPK Pathway Activation by Transforming Growth Factor  for Transforming Growth Factor ₁ Production in a Smad-dependent Pathway^*

Jianbo Yue and Kathleen M. Mulder

From the Department of Pharmacology, Pennsylvania State University College of Medicine, Hershey, Pennsylvania 17033

Received for publication, January 3, 2000, and in revised form, June 7, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our previous results have shown that transforming growth factor  (TGF) rapidly activates Ras, as well as both ERKs and SAPKs. In order to address the biological significance of the activation of these pathways by TGF, here we examined the role of the Ras/MAPK pathways and the Smads in TGF₃ induction of TGF₁ expression in untransformed lung and intestinal epithelial cells. Expression of either a dominant-negative mutant of Ras (RasN17) or a dominant-negative mutant of MKK4 (DN MKK4), or addition of the MEK1 inhibitor PD98059, inhibited the ability of TGF₃ to induce AP-1 complex formation at the TGF₁ promoter, and the subsequent induction of TGF₁ mRNA. The primary components present in this TGF₃-inducible AP-1 complex at the TGF₁ promoter were JunD and Fra-2, although c-Jun and FosB were also involved. Furthermore, deletion of the AP-1 site in the TGF₁ promoter or addition of PD98059 inhibited the ability of TGF₃ to stimulate TGF₁ promoter activity. Collectively, our data demonstrate that TGF₃ induction of TGF₁ is mediated through a signaling cascade consisting of Ras, the MAPKKs MKK4 and MEK1, the MAPKs SAPKs and ERKs, and the specific AP-1 proteins Fra-2 and JunD. Although Smad3 and Smad4 were not detectable in TGF₃-inducible AP-1 complexes at the TGF₁ promoter, stable expression of dominant-negative Smad3 could significantly inhibit the ability of TGF₃ to stimulate TGF₁ promoter activity. Transient expression of dominant-negative Smad4 also inhibited the ability of TGF₃ to transactivate the TGF₁ promoter. Thus, although the Ras/MAPK pathways are essential for TGF₃ induction of TGF₁, Smads may only contribute to this biological response in an indirect manner.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Transforming growth factor  (TGF)¹ is a natural growth inhibitor for epithelial-derived cells and a pleiotropic polypeptide for a variety of other cell types (1). TGF initiates its signaling by binding and activating TGF receptor types I (RI) and II (RII), which then form heterocomplexes and activate downstream components (1-3). Recently, members of the Sma and Mad homologue (Smad) family of signaling intermediates have been cloned and appear to play an important role in TGF signal transduction (1-3). Thus far, nine mammalian Smads (1-9) have been identified (3-5). The binding of TGF to TGF receptor complexes induces the phosphorylation of receptor-activated Smads, including Smads 1, 2, and 3 (3, 6, 7). The phosphorylated receptor-activated Smads form a heteromeric complex with the co-Smad Smad4 and translocate to the nucleus (3-5). This heteromeric complex may either directly bind the promoters of its target genes, or associate with other transcription factors to induce gene transcription (8, 9).

The mitogen-activated protein kinases (MAPKs) represent another major type of signaling intermediate for TGF (1, 2). We were the first to demonstrate that TGF could activate Ras, ERK1/2, and Sapks/JNKs within 3-5 min of TGF addition (10-13). Recently, other groups have confirmed the finding that TGF can activate ERKs and Sapks/JNKs (14-17). We were also the first to demonstrate that Ras was required for TGF-mediated activation of ERKs (18) and for the up-regulation of p21^Cip1 and p27^Kip1 (19). In addition, we have provided evidence of transmodulation of the Smad1 pathway by the Ras/MAPK pathways (6, 7).² However, the cellular events targeted by these TGF-mediated kinase activation events have not been widely studied.

TGF regulates the growth of cancer cells in both an autocrine and a paracrine fashion (1, 21). In TGF-sensitive tumor cells, autocrine TGF inhibits the growth and diminishes the tumorigenic potential of the cells (2, 21-24). In TGF-resistant tumor cells, which still secrete large amounts of TGF, the secreted TGF can enhance tumorigenesis by increasing cell migration, connective tissue formation, immunosuppression, and angiogenesis in a paracrine fashion (1, 21). Specifically blocking the production of TGF would inhibit the paracrine, tumor-enhancing effects of TGF in adenocarcinomas that have become refractory to TGF-mediated growth inhibition. Thus, it is important to explore the signaling cascades mediating TGF production.

Previous work has demonstrated that TGF₁ can induce its own production (25-27). In the current report, we demonstrate for the first time that TGF₃ induction of TGF₁ is mediated through the Ras   MAPKKs (MKK4 and MEK1)  MAPKs (Sapks and ERKs) signaling cascades. Moreover, we demonstrate that these pathways are required for the ability of TGF to regulate specific AP-1 proteins, namely Fra-2 and JunD, thereby leading to TGF₁ production. Finally, although the Smads did not directly bind the relevant AP-1/SBE site in the TGF₁ promoter, Smads 3 and 4 may be indirectly involved in TGF₃ induction of TGF₁.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell Culture-- The untransformed rat intestinal epithelial cell (IEC) clone IEC 4-1 (TGF-sensitive) was isolated as described previously (28). Cells were routinely maintained in SMIGS medium, consisting of McCoy's 5A (Life Technology, Inc.) supplemented with amino acids, pyruvate, and antibiotics (streptomycin, penicillin), and containing insulin (4 µg/ml), glucose (4.5 mg/ml), and 5% fetal bovine serum. RasN17-transfected IEC 4-1 clone E3, isolated and characterized as described previously (18), was routinely maintained in SMIGS plus G418 (131 µg/ml). DN MKK4-transfected IEC 4-1 clones N10 and N12, isolated and characterized as described previously,³ were routinely maintained in SMIGS plus G418 (500 µg/ml). CCL64-L20 and CCL64-Smad3C mink lung epithelial cells, gifts from Dr. H. Lodish (Cambridge, MA), were routinely maintained in Dulbecco's modified Eagle's medium plus 10% fetal bovine serum.

Multiprobe RNase Protection Assay-- The rCK-3 (45631P, PharMingen) and hCK-3 (45033P, PharMingen) Multi-Probe template sets contain multiple probes, including TGF₁, L32, and glyceraldehyde-3-phosphate dehydrogenase. The probes were synthesized as described in the user manual (Multi-probe RNase Protection Assay System, PharMingen). Total RNA (15 µg) from proliferating cultures of IEC 4-1, IEC-P3, RasN17 E3, DN MKK4 clones N10 or N12, CCL64-L20, or CCL64-Smad3C cells were incubated with the labeled probe set overnight at 56 °C. The samples were then digested by RNase and resolved on 5% denaturing gels as described in the user manual (Multi-probe RNase protection assay system, PharMingen). Gels were dried and exposed to x-ray film at 70 °C.

Luciferase Reporter Assays-- CCL64 cells were transfected with 0.5 µg of phTG5-Lux or 3TP-Lux, and 0.125 µg of renilla luciferase control reporter (pRL-SV40), using SuperFect (301305, QIAGEN) as described in the user manual. TGF₃ (10 ng/ml) was added 21 h after transfection, and luciferase activity was measured at 24 h after TGF treatment. The dual luciferase assay (E1910, Promega) was performed according to the manufacturer's instructions. Transfection efficiency was determined by co-transfecting renilla luciferase.

Site-directed Mutagenesis-- The plasmid phTG5-Lux, containing a 450-base pair fragment of the human TGF₁ gene promoter, was provided by S. J. Kim (Bethesda, MD). The AP-1 consensus site (362 to 355, 5'-TGTCTCA-3') was changed into 5'-TGcagCA-3' by Quick Change site-directed mutagenesis kit (Stratagene, La Jolla, CA). The primer used was: 5'-CCTCTGGTCGGCTCCCCTGTGCAGCATCCCCCGGATTAAGCCTTC-3'. The mutant clones were verified by DNA sequencing.

Electrophoretic Mobility Shift Assays (EMSAs)-- Nuclear extracts were prepared as described previously (30). Briefly, cells were plated and treated with TGF as described above. Cells were washed with ice-cold phosphate-buffered saline twice, and lysed in 1 ml of ice-cold hypotonic lysis buffer (20 mM Hepes, pH 7.4, 20% glycerol, 0.01 M NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.1% Triton X-100, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 mM glycerophosphate, 1 mM NaV₃O₄, 1 × protease inhibitor mixture). After a 20-min incubation on ice, samples were centrifuged at 10,000 × g at 4 °C. The pellets were resuspended in 200 µl of nuclear extraction buffer (hypotonic buffer + 500 mM NaCl). The nuclear lysates were cleared by centrifugation, and protein concentrations were determined by the bovine serum albumin assay as described previously (19).

Oligonucleotides were labeled with [-³²P]ATP by T4 polynucleotide kinase. Nuclear extracts (6 µg) were incubated with binding buffer (20% glycerol, 5 mM MgCl₂, 2.5 mM EDTA, 2.5 mM dithiothreitol, 250 mM NaCl, 50 mM Tris-HCl pH 7.5, 0.25 mg/ml poly(dI-dC)-poly(dI-dC)) for 10 min at room temperature, followed by addition of 1 µl (500,000 cpm) of ³²P-labeled oligonucleotides to each reaction. The reactions were incubated at room temperature for 20 min. For supershift assays, 1 µl of antibodies anti-pan-Jun (sc-44X, Santa Cruz, CA), anti-pan-Fos (sc-253x), anti-c-Jun (sc45x), anti-c-Fos (sc-52x), anti-FosB (sc-48x), anti-Fra-1 (sc-183x), anti-Fra-2 (sc-604x), anti-JunB (sc-46X), anti-JunD (sc-74x), anti-Smad4 (sc-1909x), anti-Smad4 (sc-7154x), or anti-Smad3 (51-1500, Zymed Laboratories Inc., South San Francisco, CA) were then added to reactions. The reactions were incubated at room temperature for 45-60 min, stopped by addition of 1 µl of gel loading (× 10) buffer, and analyzed by native polyacrylamide gels (4%) at 150 volts for 2 h. The gels were dried and exposed to x-ray film at 80 °C. The TGF₁ probe corresponding to the AP-1 site (362 to 355) was: ³⁷²GGCTCCCCTGTGTCTCATCCCCCGGAT³⁴⁵. The mutant TGF₁ AP-1 probe was: ³⁸¹GAAGGCTTAATCCGGGGGATgctgCACAGGGGAGCCGACCAGAGG³³⁶. The TGF₁ probe corresponding to the second potential SBE (+21 to +25) was: ⁺⁸TCCGCGGAGCAAGACAGCGAGGGCCC⁺³⁸. The control SBE probe used was: 5'-GGAGTATGTCTAGACTGACAATGTAC-3' (38).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Requirement of Ras, MEK1, and MKK4 for TGF₃ Induction of TGF₁-- Untransformed IECs, which are exquisitely sensitive to the growth inhibitory effects of TGF (28), have been shown to display autoinduction of TGF₁ mRNA expression (31). Furthermore, we have previously demonstrated that TGF resulted in a rapid activation of Ras, ERKs, and SAPK/JNK in these cells (10-13). However, the biological significance of the activation of these components by TGF was not entirely clear. Thus, it was of interest to determine whether TGF₃ induction of TGF₁ was mediated through the Ras/MAPK signaling cascades, thereby linking the activation of these cytoplasmic effects to an important biological response of TGF.

The role of Ras in mediating TGF₃ induction of TGF₁ was examined by RNase protection assays (RPAs) (Fig. 1A). For these studies, we utilized the IEC 4-1 clone E3, which had been stably transfected with a dominant-negative mutant of Ras (RasN17) under the control of an inducible metallothionein promoter (18). Our previous results demonstrated that a 4-fold induction of RasN17 expression, after a 36-h treatment with ZnCl₂ in these cells, was sufficient to completely block TGF downstream events mediated by the Ras pathway (18, 19).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Expression of a dominant-negative mutant of Ras (RasN17) or a dominant-negative mutant of MKK4 (DN MKK4), or addition of the MEK1 inhibitor PD98059, inhibits TGF3 induction of TGF1. A, proliferating cultures of IEC RasN17 E3 cells were incubated in serum-free medium either in the absence or presence of ZnCl2 for 36 h to induce RasN17 expression. Cells were then treated with or without TGF3 (10 ng/ml) for 4 h. Total RNA was isolated. RNase protection assays were performed as described under "Materials and Methods." Representative of four experiments. B, proliferating cultures of IEC 4-1 cells were incubated in serum-free medium with or without TGF3 (10 ng/ml) for 4 h in the presence of the MEK1 inhibitor PD98058 (10 µM). Total RNA was isolated and RNase protection assays were performed. Representative of two experiments. C, the DN MKK4 expressing IEC clones N10 and N12, and empty vector-transfected IEC-P3 cells were plated and treated with TGF for 4 h. Total RNA was isolated and RNase protection assays were performed. Representative of two experiments. Bottom panels, densitometric scan of results shown in top panels. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

We performed RPAs to examine the effects of RasN17 on TGF₃ induction of TGF₁ expression. As shown in Fig. 1A, left side, in the absence of ZnCl_2, TGF₁ mRNA expression was increased to values 10-fold above initial baseline levels after 4 h of TGF₃ treatment. In contrast, in the presence of ZnCl₂, TGF₁ mRNA expression was increased by only 2.5-fold after the same time period of TGF₃ treatment (Fig. 1A, right side). Thus, the induction of RasN17 by ZnCl₂ inhibited the ability of TGF₃ to induce TGF₁ mRNA expression by 75%. Similar results have been observed for TGF₁ mRNA expression after 24 h of TGF₃ treatment and by Northern blot analysis (data not shown). Taken together, our results clearly demonstrate that TGF activation of Ras is required for TGF₃ induction of TGF₁.

TGF activates ERKs through a Ras-dependent pathway (18). Thus, it is conceivable that TGF₃ may regulate TGF₁ production by activating the MEK1/ERK pathway as one of the pathways downstream of Ras. We have previously employed the MEK1 inhibitor PD98059 to block Erk1 activation by TGF in IEC 4-1 cells. A concentration of PD98059 of 10 µM resulted in complete blockade of the ability of TGF to activate Erk1, without affecting basal Erk1 activity levels (7). Here, we utilized this MEK1 inhibitor in RPAs to examine the requirement of MEK1 activation for TGF₃ induction of TGF₁ expression. As shown in Fig. 1B, in the absence of PD98059, TGF₁ mRNA expression was increased to levels 9.8-fold above initial baseline values by 4 h after TGF₃ treatment. In contrast, in the presence of PD98059, TGF₁ mRNA expression was increased by only 3-fold above initial baseline levels over the same time period. Thus, MEK1 blockade by PD98059 inhibited the ability of TGF₃ to induce TGF₁ mRNA expression by 70%. Accordingly, our results indicate that TGF₃ induction of TGF₁ is mediated through the MEK1/ERK pathway as one of the events downstream of Ras.

We have previously shown that TGF activated the SAPK/JNK pathway (11, 12), and that Ras was required for TGF-mediated SAPK/JNK activation.³ Thus, it is conceivable that TGF could regulate TGF₁ production by activating the MKK4/SAPK pathway as another pathway downstream of Ras. For these studies, we utilized the IEC 4-1 clones N10 and N12, which had been stably transfected with a dominant-negative mutant of MKK4 (DN MKK4). We have shown that overexpression of DN MKK4 in these clones completely blocked the ability of TGF to activate SAPK/JNK and phosphorylate c-Jun.³ Thus, expression of DN MKK4 in these DN MKK4 clones was sufficient to completely inhibit TGF-mediated SAPK/JNK activation and its downstream events.

The effects of DN MKK4 on TGF₃ induction of TGF₁ were examined by RPAs as shown in Fig. 1C. TGF₃ increased TGF₁ mRNA expression by 8.5-fold in empty vector-transfected IEC P3 cells. In contrast, in the DN MKK4-expressing N10 or N12 clones, TGF₃ increased TGF₁ mRNA expression by only 3- or 3.6-fold above initial baseline levels, respectively (Fig. 1C). Thus, expression of DN MKK4 significantly inhibited the ability of TGF₃ to induce TGF₁ mRNA expression. Accordingly, our results indicate that TGF activation of SAPK/JNK is also required for TGF₃ induction of TGF₁.

Requirement of the AP-1 Site in the TGF₁ Promoter and the ERK Pathway for TGF₃ Stimulation of TGF₁ Promoter Activity-- The TGF₁ promoter contains an AP-1 site at 362 to 355 (25-27). Although previous results have suggested that TGF₁ autoinduction could be mediated through this AP-1 site in the TGF₁ promoter (25-27), there is no definitive evidence that this AP-1 site is essential for this effect. Here, we examined the requirement for this AP-1 site in the transactivation of the TGF₁ promoter by TGF in CCL64 mink lung epithelial cells. As shown in Fig. 2, TGF₃ treatment increased TGF₁ promoter luciferase activity by 7-fold in CCL64 cells. However, TGF₃ failed to increase the activity of the TGF₁ promoter containing a mutated AP-1 site in the same cell type (the right two bars in Fig. 2A). In addition, the basal TGF₁ promoter activity was decreased by 50% when the AP-1 site was mutated (Fig. 2A). Taken together, our results demonstrate that the AP-1 site in the TGF₁ promoter is essential for TGF transactivation of the TGF₁ promoter.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Requirement of the AP-1 site in the TGF1 promoter and the ERK pathway for TGF3 stimulation of TGF1 promoter activity. A, proliferating cultures of CCL64 cells were transfected with 0.5 µg of phTG5-Lux or phTG5-Mutant-AP-1-Lux, and 0.125 µg of renilla luciferase control reporter (pRL-SV40) by Superfect (301305, QIAGEN). Twenty-one hours after transfection, cells were incubated in serum-free medium for 1 h. Thereafter, TGF3 (10 ng/ml) was added, and luciferase activity was measured 24 h after TGF treatment using the dual luciferase assay (E1910, Promega) according to the manufacturer's instructions. The results plotted represent the  ± S.D. for triplicate transfections. The number above the bar represents the fold change by comparing the luciferase activity in the presence or absence of TGF. B, proliferating cultures of CCL64 cells were transfected with 0.5 µg of phTG5-Lux and 0.125 µg of renilla luciferase control reporter (pRL-SV40). TGF3 (10 ng/ml) was added 21 h after transfection in the presence of MEK inhibitor PD98059 (20 µM). Luciferase activity was measured as described in the legend to Fig. 2A. Representative of four experiments.

Since PD98059 significantly inhibited TGF₃ induction of TGF₁ mRNA, it was of interest to examine whether PD98059 could inhibit the ability of TGF₃ to stimulate TGF₁ promoter activity. As shown in Fig. 2B, TGF₃ treatment increased TGF₁ promoter luciferase activity by 6.5-fold in CCL64 cells. In the presence of PD98059 (20 µM), TGF₃ treatment only increased TGF₁ promoter luciferase activity by 3.5-fold; basal TGF₁ promoter activity was not affected by this concentration of PD98059. Although higher concentrations of PD98059 resulted in a further inhibition of TGF₃ stimulation of TGF₁ promoter activity, basal TGF₁ promoter activity was also decreased at concentrations above 20 µM (data not shown). These results indicate that the ERK pathway is, indeed, required for TGF₃ induction of TGF₁. The data also suggest that the ERK pathway may contribute to TGF₃ induction of TGF₁ through the AP-1 site in the TGF₁ promoter.

Temporal Relationship between TGF₃-stimulated Transcription Factor Binding to the TGF₁ Promoter and TGF₁ mRNA Expression-- If TGF₃ induction of TGF₁ is mediated through the AP-1 site in the TGF₁ promoter, then TGF induction of complex formation at this site should kinetically precede the increase in production of TGF₁ mRNA. In order to verify whether this was the case, we performed EMSAs using an oligonucleotide (372 to 345) spanning the AP-1 site in the TGF₁ promoter as a probe. The results in the left panel of Fig. 3A indicate that the levels of Complex I began increasing by 15 min after TGF treatment and were maximal by 2 h post-TGF addition. After this time the levels began to decline. However, the levels of Complex I were still much higher than initial baseline levels by 4 and 24 h after TGF addition. In contrast, a gel-shifted Complex I was not observed in response to treatment of the cells with TGF when an oligonucleotide containing a mutant AP-1 site was used as a probe (Fig. 3A, right panel). Similar results were observed in CCL64 mink lung cells (Fig. 3B). Therefore, the AP-1 site in the TGF₁ promoter is essential for TGF to stimulate the formation of Complex I. 

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Kinetics for TGF3 induction of transcription factor binding at the TGF1 promoter and for TGF3 induction of TGF1 mRNA expression. Proliferating cultures of IEC 4-1 cells (A and C) or CCL64 cells (B) were incubated in serum-free medium for 1 h. Cells were then treated with or without TGF3 (10 ng/ml) for the indicated times. A and B, nuclear extracts were prepared as described under "Materials and Methods." EMSAs were performed using oligonucleotides containing the AP-1 site in the TGF1 promoter (372 to 345) or the mutant AP-1 site in the TGF1 promoter (381 to 336) as probes. C, total RNA was isolated, and RPAs were performed as described under "Materials and Methods." Bottom panel, densitometric scan of results shown in top panel. GADPH, glyceraldehyde-3-phosphate dehydrogenase.

In order to determine whether the TGF₃ induction of TGF₁ mRNA expression was preceded by TGF-inducible complex formation at the AP-1 site, we also examined the kinetics for TGF₃ effects on TGF₁ mRNA expression by RPAs. As shown in Fig. 3C, TGF₁ mRNA expression was increased to levels 2.4-fold above initial baseline values by 30 min after TGF₃ treatment. By 2 h of TGF₃ treatment, maximal TGF₁ mRNA expression levels of 9-fold above initial baseline values were reached. TGF₁ mRNA expression levels were then maintained at levels 8.5-fold above initial baseline values for at least 4 h. Thereafter, TGF₁ mRNA expression levels declined, so that expression levels were 5-fold above initial baseline values by 24 h after TGF₃ treatment. Thus, formation of the TGF₃-inducible complex at the AP-1 site in the TGF₁ promoter temporally preceded TGF₃ induction of TGF₁ mRNA expression in IEC 4-1 cells. Similar kinetics have been observed for the ability of TGF₃ to stimulate TGF₁ secretion into the medium of IECs using a TGF₁ enzyme-linked immunosorbent assay system (data not shown). Our results provide strong evidence of a tight association between TGF₃ stimulation of transcription factor binding at the AP-1 site in the TGF₁ promoter and TGF₃ induction of TGF₁. Our results support the requirement of this AP-1 site for mediating this biological response to TGF.

Requirement of Ras, ERK, and SAPK/JNK for TGF₃ Stimulation of Transcription Factor Binding at the TGF₁ Promoter-- We have demonstrated in Figs. 1-3 that Ras/MAPK signaling cascades and the AP-1 site in the TGF₁ promoter are required for TGF₃ induction of TGF₁, and that TGF₃ can induce transcription factor binding at this AP-1 site. Accordingly, it was of interest to examine whether TGF₃ activation of the Ras/MAPK pathways was required for TGF stimulation of Fos and Jun binding at this site. An oligonucleotide (372 to 345) spanning the AP-1 site in the TGF₁ promoter was utilized in EMSAs. In RasN17E3 cells, TGF₃ increased the formation of protein-DNA Complex I (see arrow) after TGF₃ treatment in the absence of ZnCl₂ (Fig. 4A, left side). In contrast, in the presence of ZnCl₂ (Fig. 4A, right side)_, the induction of RasN17 blocked the formation of this TGF₃-inducible complex. In addition, as shown in Fig. 4A, left side, addition of either a pan-Fos or pan-Jun antibody completely blocked the induction of Complex I by TGF₃, while normal rabbit IgG had no effect on this complex. Accordingly, our results indicate that TGF₃ activation of Ras is required for TGF₃ induction of Fos-Jun complex formation at this AP-1 site.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Requirement of Ras, MKK4, and MEK1 for TGF3 induction of transcription factor binding at the TGF1 promoter. A, proliferating cultures of IEC RasN17 E3 cells were incubated in serum-free medium either in the presence or absence of ZnCl2 (100 µg/ml) for 36 h to induce RasN17 expression. Cells were then treated with or without TGF3 (10 ng/ml) for 2 h. Nuclear extracts were prepared as described under "Materials and Methods," and EMSAs were performed using a probe containing the AP-1 site in the TGF1 promoter (372 to 345). Supershifts were performed using antibodies prepared against pan-Jun (sc-44 X) and pan-Fos (sc-413 X). Normal rabbit IgG was used as a control. The TGF-induced protein-DNA Complex I is indicated by the arrow. Representative of three experiments. B, EMSAs were performed using nuclear extracts from IEC 4-1 cells that were pretreated with or without the MEK1 inhibitor PD98058 (10 µM) in the presence or absence of TGF3. The AP-1 site in the TGF1 promoter (372 to 345) was used as a probe. Representative of two experiments. C, EMSAs were performed using nuclear extracts from IEC 4-1 cells or DN MKK4 cells that were treated with TGF3 or left untreated. The AP-1 site in the TGF1 promoter (372 to 345) was used as a probe. Representative of two experiments.

We also examined whether the MEK1 inhibitor PD98058 would inhibit the ability of TGF₃ to induce Fos and Jun binding at the TGF₁ promoter (Fig. 4B). EMSAs demonstrated that in the absence of PD98059, the induction of Fos-Jun complex formation at the TGF₁ promoter (Complex I) was prominent after 2 h of TGF treatment. In the presence of PD98059, however, TGF₃ failed to induce the formation of this complex. Thus, the MEK1/ERK pathway is required for TGF₃ induction of AP-1 protein binding at the TGF₁ promoter as well.

EMSAs were also performed to determine whether MKK4 was required for TGF₃ induction of AP-1 protein binding at the TGF₁ promoter. As shown in Fig. 4C, in the parental IEC 4-1 cells, TGF₃ induced the formation of the Fos-Jun complex, similar to that seen in Fig. 4, A and B. In contrast, in DN MKK4-transfected IECs, TGF₃ failed to increase the formation of Complex I. Thus, overexpression of DN MKK4 blocked the ability of TGF₃ to induce the formation of this complex at the AP-1 site in the TGF₁ promoter. Collectively, our data indicate that TGF activation of SAPK/JNK is also required for TGF₃ stimulation of Fos-Jun complex formation at the TGF₁ promoter.

Specific AP-1 Proteins Are Present in the TGF₃-inducible Complex at the TGF₁ Promoter-- Previous results demonstrated that expression of antisense c-Jun and c-Fos constructs inhibited TGF₁ autoinduction (25). However, there is no definitive evidence to indicate which specific Fos or Jun members constitute this TGF-inducible AP-1 complex. Since the Jun and Fos families include multiple proteins in addition to c-Jun and c-Fos (32), it was of interest to determine which Fos and Jun family members mediated this complex formation. Fig. 5 depicts the results of supershift assays performed using antibodies specific for each Fos and Jun family member. The pan-Jun and pan-Fos antibodies were also used as controls. As expected, the pan-Jun and pan-Fos antibodies blocked the formation of Complex I. Furthermore, specific anti-JunD and anti-Fra-2 antibodies supershifted and blocked Complex I as shown in Fig. 5. Addition of the specific anti-c-Jun or anti-FosB antibodies also resulted in a supershifted band above Complex I, yet the extent of the shift was less prominent than that observed for the anti-JunD and anti-Fra-2 antibodies. In contrast, the specific anti-JunB and anti-Fra-1 antibodies had no significant effect on this complex. In summary, our data indicate that the primary components present in the TGF₃-inducible AP-1 complex at the TGF₁ promoter are Fra-2 and JunD, although c-Jun and FosB are also present. It is noteworthy that TGF can directly phosphorylate JunD, and that expression of RasN17 completely blocked the ability of TGF to phosphorylate JunD (data not shown).

View larger version (78K): [in this window] [in a new window]   Fig. 5.   Involvement of JunD, Fra-2, FosB, and c-Jun in TGF3 stimulation of transcription factor binding at the TGF1 promoter. EMSAs were performed using nuclear extracts from IEC 4-1 cells that were treated with or without TGF3 for 2 h. The AP-1 site in the TGF1 promoter (372 to 345) was used as a probe. Supershifts were performed using specific antibodies against c-Jun, JunB, JunD, c-Fos, Fos-B, Fra-1, or Fra-2. The pan-Jun and pan-Fos antibodies were used as controls. The TGF-inducible protein-DNA Complex I and supershifts are indicated by arrows or asterisks, respectively. Representative of three experiments.

Smad3 and Smad4 Are Not Present in the TGF₃-inducible AP-1 Complex at the TGF₁ Promoter-- Recently, Smad-binding elements (SBEs) have been identified by several groups (33-39). Although the reported SBEs vary among different genes, it would appear that most SBEs contain either AGAC or its complementary sequence GTCT (33-39). We have searched the TGF₁ promoter sequence (26) and found two potential minimal SBE sites. One potential SBE site (GTCT, 363 to 359) overlaps the AP-1 site (TGTCTCA, 362 to 355) in the TGF₁ promoter. The second potential SBE is located from +21 to +25 (AGAC) in the TGF₁ promoter. Thus, it was of interest to determine whether Smads could bind these potential SBEs in the TGF₁ promoter.

A control SBE probe was also prepared, which was a consensus Smad3/Smad4-binding site identified by random oligonucleotide screening (38). As shown in Fig. 6, after 2 h of TGF treatment, TGF increased the formation of the complex at the control SBE probe (Fig. 6, left panel). Addition of either Smad3 or Smad4 antibodies resulted in a supershift, as indicated by the arrow at the top (Fig. 6, left panel). For the TGF₁ probe spanning both the AP-1 site and the first potential SBE (372 to 345), TGF increased the formation of the AP-1 complex after 2 h of TGF treatment (Fig. 6, middle panel), similar to the results in Fig. 3A. However, neither Smad3 nor Smad4 antibodies affected the formation of this complex (Fig. 6, middle panel). Thus, Smads do not appear to be involved in this TGF₃-inducible AP-1 complex at the TGF₁ promoter. For the TGF₁ probe spanning the second potential SBE (+8 to +38), none of the protein-DNA complexes were inducible by TGF, and none of the Smad3 or Smad4 antibodies had any significant effect on complex formation (Fig. 6, right panel). Thus, this potential SBE in the TGF₁ promoter is not utilized for TGF₃ induction of TGF₁.

View larger version (53K): [in this window] [in a new window]   Fig. 6.   Smad3 and Smad4 are not present in the TGF3-inducible AP-1 complex at the TGF1 promoter. EMSAs were performed using nuclear extracts from CCL64 cells that were treated with TGF3 or left untreated. The probes used were: the control SBE (left panel), oligonucleotides spanning the AP-1 site in the TGF1 promoter (372 to 345) (middle panel), or oligonucleotides spanning the second potential SBE in the TGF1 promoter (+8 to +38) (right panel). Supershifts were performed using anti-Smad4 (H-522), anti-Smad4 (C-20), or anti-Smad3 (51-1500, Zymed Laboratories Inc.) antibodies. The TGF-inducible protein-DNA complexes and supershifts are indicated by arrows.

TGF₃ Induction of TGF₁ Is Dependent on Smads-- Since Smads play an important role in TGF signaling (1-3) and TGF sometimes activates Sapks/JNKs, JunB, or c-Jun via Smad-dependent pathways (33, 36, 40), it was of interest to determine whether Smads were required in any respect for TGF₃ induction of TGF₁. Accordingly, the ability of TGF₃ to stimulate TGF₁ promoter activity was examined in Smad3C-CCL64 cells, which stably express a dominant-negative form of Smad3 (41). It has been previously reported that TGF failed to induce phosphorylation of Smad3 in this cell line, and that overexpression of this dominant-negative mutant of Smad3 blocked the ability of TGF to inhibit cell growth (41). Thus, normal Smad3 function in this cell line is lost. As expected, TGF₃ stimulated 3TP-Lux activity by 11-fold in the parental CCL64-L20 cells, but only increased 3TP-Lux activity by 3.4-fold in the CCL64-Smad3C cells (Fig. 7A, left panel). Similarly, as shown in the right panel of Fig. 7A, TGF₃ stimulated TGF₁ promoter activity by 9.8-fold in the parental CCL64-L20 cells, but only increased TGF₁ promoter activity by 2.1-fold in the CCL64-Smad3C cells. Thus, overexpression of this dominant-negative mutant of Smad3 inhibited the ability of TGF₃ to increase TGF₁ promoter activity by 75%. The results from RPAs also indicated that overexpression of this dominant-negative mutant of Smad3 significantly inhibited the ability of TGF₃ to stimulate TGF₁ mRNA expression in the CCL64-Smad3C cells (data not shown). Accordingly, these results indicate that TGF₃ induction of TGF₁ requires Smad3. It is possible that expression of this DN Smad3 may also interfere with Smad2 function in CCL64 cells (41). However, previous results have shown that TGF₁ autoinduction was also blocked in Smad3-null macrophages and keratinocytes (42).

View larger version (25K): [in this window] [in a new window]   Fig. 7.   TGF3 induction of TGF1 is dependent upon Smad3 and Smad4 in CCL64 cells. A, proliferating cultures of CCL64-L20 cells or CCL64-Smad3C cells were transfected with 0.5 µg of 3TP-Lux (left panel) or phTG5-Lux (right panel), and 0.125 µg of renilla luciferase control reporter (pRL-SV40) by Superfect. Luciferase activity was measured as described in the legend to Fig. 2A. The results plotted represent the  ± S.D. for triplicate transfections. Representative of three independent experiments. B, proliferating cultures of CCL64 cells were transfected with 0.5 µg of phTG5-Lux, 0.125 µg of renilla luciferase control reporter (pRL-SV40), and different doses of DN Smad4 or empty vector by Superfect. Luciferase activity was measured as described in the legend to Fig. 2A. The results plotted represent the  ± S.D. for triplicate transfections. Representative of three independent experiments.

To determine whether TGF₃ induction of TGF₁ required Smad4, a dominant-negative mutant of Smad4 (DN Smad4) was co-transfected with phTG5-Luc into CCL64 cells. As shown in Fig. 7B, TGF₃ stimulated TGF₁ promoter activity by 5.5-fold; expression of DN Smad4 inhibited the ability of TGF₃ to increase TGF₁ promoter activity in a dose-dependent manner. Thus, Smad4 is required for TGF₃ induction of TGF₁ in CCL64 cells as well.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This report is the first to demonstrate that TGF activation of the Ras/MAPK pathways is essential for TGF₃ induction of TGF₁ expression. Blockade of TGF activation of Ras, MKK4/JNK, or MEK/ERK, each inhibited TGF₃ induction of TGF₁. We also demonstrate that Ras, MKK4/JNK, and MEK/ERK were required for TGF₃-stimulated AP-1 complex formation at the TGF₁ promoter. JunD and Fra-2 were the primary components present in this TGF₃-inducible AP-1 complex, but c-Jun and FosB were also involved. Deletion of the AP-1 site in the TGF₁ promoter abolished TGF₃-inducible Fos-Jun complex formation and TGF₃ transactivation of the TGF₁ promoter. Although neither Smad3 nor Smad4 were detectable in the TGF₃-inducible AP-1 complex in the TGF₁ promoter, expression of a dominant-negative mutant of Smad3 inhibited the ability of TGF₃ to increase TGF₁ promoter activity and TGF₁ mRNA expression. Furthermore, expression of a dominant-negative mutant of Smad4 inhibited TGF₃ stimulation of TGF₁ promoter activity. Thus, TGF₃ induction of TGF₁ is dependent upon TGF activation of the Ras/MAPK pathways, and Smads may also indirectly contribute to this biological response.

Our results reveal that TGF activation of Ras and the MAPK pathways is required for an important biological response to TGF, namely, autocrine production of TGF₁. Although it has been shown previously that the H-ras oncogene can stimulate TGF₁ promoter activity (43), oncogenic Ras may activate additional different downstream targets than those regulated by TGF specifically (2). Furthermore, overexpression of activated ras genes can result in TGF resistance and Smad inactivation (44). Thus, it is important to make the distinction between TGF activation of normal cellular Ras (as we are examining here) versus activation of Ras by overexpression, activating mutations, or other factors (reviewed in Ref. 2).

The activation of the Ras/MAPK pathways by TGF can mediate biological effects of TGF in addition to autocrine amplification of TGF₁ production. For example, we have previously shown that Ras is essential for TGF up-regulation of CKIs that contribute to TGF-mediated growth inhibition (18, 19). Results from two other groups have confirmed our original findings that the Ras/MAPK pathways were required for TGF induction of p21^Cip1 expression (45, 46). Furthermore, we have demonstrated a requirement of the Ras/MEK1 pathway for the positive modulation of the Smad1 pathway (6, 7). TGF activation of SAPK/JNK has also been demonstrated to be required for the induction of fibronectin synthesis by TGF (16). Additional cellular functions of Ras/MAPK activation by TGF are likely to be elucidated as well.⁴

It is of considerable interest that JunD and Fra-2 are the primary components present in the TGF₃-stimulated AP-1 complex at the TGF₁ promoter. TGF has been shown previously to activate AP-1 proteins and to regulate the expression of downstream target genes (47-51). However, with respect to TGF-inducible genes, the involvement of JunD and Fra-2 has only been observed for the murine laminin 3A gene (50). Although JunD has also been reported to be involved in TGF induction of interleukin-6, collagenase-1, and collagenase-3 expression (47, 51), Fra-2 was not involved in the regulation of these genes. Moreover, the transactivation properties of Fra-2·JunD complexes are enhanced relative to JunD homodimers (52). Thus, the involvement of Fra-2/JunD in the regulation of TGF-inducible genes appears to be somewhat specific. Selective targeting of this combination of AP-1 members could block TGF₁ production specifically. Moreover, our results suggest that TGF may activate different AP-1 proteins to specifically regulate the expression of its target genes.

Although some AP-1 proteins have been shown to be important in mediating transformation of cells induced by oncogenes, JunD has actually been shown to inhibit the transformation of cells in vitro (53). Furthermore, overexpression of JunD can decrease the growth of mesenchymal cells (53). Thus, it is conceivable that the growth inhibitory effects of JunD may partially be due to the role of JunD in mediating TGF₁ production. That is, the induction of TGF₁ production by JunD may amplify the growth inhibitory effects of TGF in TGF-sensitive cells. In contrast, in some cancer cells, JunD and Fra-2 may still stimulate TGF₁ production, yet the cells would be resistant to TGFs growth inhibitory effects. In this case, the elevated TGF would actually enhance the tumor-promoting effects of TGF via paracrine mechanisms. Along these lines, Fra-2 expression levels are high in breast cancer cells (54).

In Drosophila, the production of the TGF superfamily member Dpp requires DJNK, DFos, and DJun during dorsal closure (55-60). Here we demonstrate that Ras, MKK4/JNK, MEK/ERK, and specific AP-1 proteins are required for TGF₃ induction of TGF₁ in mammalian cells. Thus, the requirement of the Ras/MAPK pathways for production of TGF superfamily members appears to be evolutionarily conserved. Moreover, our results have indicated that TGF can directly activate JNK (reviewed in Refs. 1 and 2). Thus, it may also be possible for Dpp to directly activate DJNK and to establish an autocrine loop during dorsal closure. Along these lines, it has been shown that Dpp is required for Dfos expression during dorsal closure (58). However, the ability of Dpp to induce its own expression through DJNK/DJun/DFos has not been reported thus far.

Here, we demonstrate that TGF₃ induction of TGF₁ was also dependent upon Smads in CCL64 cells. Similarly, in Drosophila, it has been found that Mad is required for the maintenance of dpp expression during embryonic midgut development (61, 62). However, we were unable to detect binding of Smad3 or Smad4 to the AP-1 site in either IECs or CCL64 cells (Fig. 6). Moreover, pretreatment of CCL64 cells with cycloheximide (10 µg/ml) did not affect the ability of TGF to increase AP-1 complex formation under conditions where protein synthesis was inhibited by 97% (data not shown). Thus, it is possible that Smads may not function as transcription factors in mediating TGF₃ induction of TGF₁. Instead, Smads, in this context, may function as cytoplasmic modulators for other signaling components. Indeed, it has been reported that Smads can associate with other cytoplasmic proteins, including Smad anchor for receptor activation (SARA) (63), calmodulin (29), and microtubules (20).

In summary, TGF₃ induction of TGF₁ appears to occur via the following mechanism: TGF activates Ras, leading to induction of both the MKK4/SAPK an MEK1/ERK signaling cascades. These cascades, in turn, stimulate AP-1 protein complex formation at the AP-1 site in the TGF₁ promoter. JunD and Fra-2 are the major AP-1 proteins that bind at this site to induce TGF₁ transcription, thereby resulting in increased TGF₁ protein expression and secretion. In addition, although Smads appear to contribute to this important cellular response of TGF, their effects are likely to be indirect in this context. Perhaps Smads assume novel intracellular functions not currently represented by the accepted Smad activation cascade.

	ACKNOWLEDGEMENTS

We thank M. Morin (Pfizer Pharmaceuticals, Groton, CT) for generously supplying the TGF₃, S. J. Kim (National Cancer Institute, Bethesda, MD) for the phTG5 construct, H. F. Lodish (MIT, Cambridge, MA) for the CCL64-Smad3C cells, R. Derynck (University of California, San Francisco, CA) for the DN Smad3 and DN Smad4 constructs, and R. Davis (University of Massachusetts, Worcester, MA) for the DN MKK4 construct.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants CA51424, CA54816, and CA68444 (to K. M. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Pharmacology H078, Pennsylvania State University College of Medicine, 500 University Dr., Hershey, PA 17033. Tel.: 717-531-6789; Fax: 717-531-5013; E-mail: kmm15@psu.edu.

Published, JBC Papers in Press, June 7, 2000, DOI 10.1074/jbc.M000039200

² X. J. Liu and K. M. Mulder, submitted for publication.

³ J. Yue and K. M. Mulder, submitted for publication.

⁴ J. Yue and K. M. Mulder, unpublished results.

	ABBREVIATIONS

The abbreviations used are: TGF, transforming growth factors ; RasN17, dominant-negative Ras mutant; IECs, intestinal epithelial cells; MAPK, mitogen-activated protein kinase; Smads, Sma and Mad homologues; JNK/SAPK, c-Jun N-terminal kinases/stress-activated protein kinases; ERKs, extracellular signal-regulated kinases; EMSAs, electrophoretic mobility shift assays; RPAs, RNase protection assays; MAPKK, mitogen-activated protein kinase kinase; RI, receptor type I; DN, dominant-negative; SBE, Smad-binding elements.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES