(COL7A1) Promoter by Transforming Growth Factor

THE JOURNAL

OF

BIOLOGICAL CHEMISTRY

Vol. 273, No. 21, Issue of May 22, pp. 13053–13057, 1998 Printed in U.S.A.

Smad-dependent Transcriptional Activation of Human Type VII Collagen Gene (COL7A1) Promoter by Transforming Growth Factor-b* (Received for publication, February 10, 1998, and in revised form, April 1, 1998)

Laurence Vindevoghel‡§, Atsushi Kon‡§, Robert J. Lechleider¶, Jouni Uitto‡§i**, Anita B. Roberts¶, and Alain Mauviel‡§‡‡ From the Departments of ‡Dermatology and Cutaneous Biology and iBiochemistry and Molecular Pharmacology, Jefferson Medical College, §Jefferson Institute of Molecular Medicine, and the **Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, Pennsylvania 19107 and ¶Laboratory of Cell Regulation and Carcinogenesis, NCI, National Institutes of Health, Bethesda, Maryland 20892

The collagens comprise a superfamily of proteins that play a critical role in the maintenance of extracellular matrix integrity. Type VII collagen is found primarily in the basement membrane zone of specialized squamous epithelia, such as in the skin, various mucous membranes, and the cornea of the eye (1, 2). It is the predominant, if not the exclusive, component of anchoring fibrils, attachment structures that play a critical role in ensuring stability to the association of the epithelial

* This work was supported in part by National Institutes of Health Grants 29-AR43751 (to A. M.) and RO1-AR41439 and T32-AR07651 (to J. U.) and by a Research Career Development Award from the Dermatology Foundation (to A. M.). The costs of publication of this article were defrayed in part by the payment of page charges. This 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 Dermatology and Cutaneous Biology, Jefferson Medical College, Thomas Jefferson University, 233 South 10th St., Rm. 430, Philadelphia, PA 19107. Tel.: 215-503-5775; Fax: 215-923-9354; E-mail: [email protected]. This paper is available on line at http://www.jbc.org

basement membrane zone to the underlying papillary dermis (3, 4). Synthesis of functional anchoring fibrils is of critical importance in providing integrity to the cutaneous basement membrane zone, and abnormalities in these adhesion structures clinically manifest as dystrophic forms of epidermolysis bullosa, a group of heritable bullous diseases characterized by cutaneous fragility and the tendency to sub-basal lamina densa blister formation (5). In fact, our laboratory recently demonstrated that mutations within the COL7A1 gene are associated with different forms of dystrophic epidermolysis bullosa (6, 7). Analysis of the 59-end sequences of the human COL7A1 gene has revealed a promoter devoid of a canonical TATA or CAAT box (GenBank/EMBL accession no. L23982). Its expression requires the integrity of an Sp1-binding site located between residues 2512 and 2505, relative to the transcription start site (8). Despite these structural features usually associated with promoters of so-called “housekeeping” genes, type VII collagen expression has been shown to be transcriptionally regulated by several cytokines, including transforming growth factor-b (TGF-b)1 (9 –11), and by other biological response modifiers, such as ultraviolet irradiation (12). Interestingly, we have previously shown that cytokine-mediated regulation of COL7A1 gene expression is strikingly different from that of type I collagen, as evidenced by synergistic activation of COL7A1 gene expression by TGF-b and tumor necrosis factor-a (11). Thus far, however, little is known about the transcriptional mechanisms underlying the above mentioned regulation of COL7A1 gene expression. In this study, we have investigated the molecular mechanisms by which TGF-b up-regulates the activity of human COL7A1 in dermal fibroblasts. We report, for the first time, evidence for Smad-mediated, immediate-early activation of a human gene by TGF-b through direct interaction of a Smadcontaining transcription complex with the TGF-b-responsive region of the COL7A1 promoter. MATERIALS AND METHODS

Cell Cultures—Human dermal fibroblast cultures, established by explanting tissue specimens obtained from neonatal foreskins, were utilized in passages 3– 6. The cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum, 2 mM glutamine, and antibiotics (100 units/ml penicillin, 50 mg/ml streptomycin-G, and 0.25 mg/ml FungizoneTM). Human recombinant TGF-b2 was a kind gift from Dr. David Olsen, Celtrix Co., Palo Alto, CA. It is referred to as TGF-b throughout the text. 1 The abbreviations used are: TGF-b, transforming growth factor-b; ARE, activin-responsive element; CAT, chloramphenicol acetyltransferase; EMSA, electrophoresis mobility shift assay; TBSB, TGF-b-specific band; TBRS, TGF-b-responsive sequence; bp, base pair(s).

13053

Downloaded from http://www.jbc.org/ by guest on December 27, 2015

We have previously shown that transforming growth factor-b (TGF-b) increases type VII collagen gene (COL7A1) expression in human dermal fibroblasts in culture (Mauviel, A., Lapie`re, J.-C., Halcin, C., Evans, C. H., and Uitto, J. (1994) J. Biol. Chem. 269, 25–28). To gain insight into the molecular mechanisms underlying the up-regulation of COL7A1 by this growth factor, we performed transient cell transfections with a series of 5*-deletion promoter/chloramphenicol acetyltransferase reporter gene constructs. We identified a 68-base pair region between nucleotides 2524 and 2456, relative to the transcription start site, as critical for TGF-b response. Using electrophoresis mobility shift assays (EMSAs) with an oligonucleotide spanning the region from 2524 to 2444, we discovered that a TGF-b-specific protein-DNA complex was formed as early as 11 min after TGF-b stimulation and persisted for 1 h after addition of the growth factor. Deletion analysis of the TGF- bresponsive region of the COL7A1 promoter by EMSA identified segment 2496/2444 as the minimal fragment capable of binding the TGF-b-induced complex. Furthermore, two distinct segments, 2496/2490 and 2453/2444, appeared to be necessary for TGF-b-induced DNA binding activity, suggesting a bipartite element. Supershift experiments with a pan-Smad antibody unambiguously identified the TGF-b-induced complex as containing a Smad member. This is the first direct identification of binding of endogenous Smad proteins to regulatory sequences of a human gene.

13054

Smad-dependent Collagen Gene Regulation

RESULTS

TGF-b Up-regulates Human COL7A1 Expression at a Transcriptional Level—We and others have previously demonstrated that TGF-b is a potent inducer of type VII collagen gene expression in dermal fibroblasts, as determined at both protein and mRNA levels (10, 11). To investigate whether TGF-b upregulates COL7A1 gene expression at the transcriptional level by activation of the promoter, human dermal fibroblast cultures were transiently transfected with several 59-end deletion/ CAT reporter gene constructs spanning the COL7A1 promoter region from positions 2722 to 192, relative to the transcription start site, 11 (8). Cells were subsequently treated with TGF-b (10 ng/ml) for 40 h, at which point CAT activity (representing COL7A1 promoter activity) was determined. As shown in Fig. 1A, a stimulatory effect of TGF-b was observed with constructs 2722COL7A1/CAT (;4.4-fold induction) and 2524COL7A1/ CAT (;5.3-fold induction), indicating that the up-regulation of type VII collagen gene expression by TGF-b occurs, at least in part, at the transcriptional level through activation of the promoter. Subsequent 59-deletion to position 2456 abolished the COL7A1 promoter responsiveness to TGF-b, which was similarly lost when further 59-deletions extended to positions 2396 and 2230 (Fig. 1A). These data indicate that the DNA sequences between residues 2524 and 2456 of the COL7A1 promoter are essential in providing TGF-b responsiveness in fibroblasts.

FIG. 1. Effect of TGF-b on the COL7A1 promoter activity. A, confluent fibroblast cultures were transiently transfected with various 59-deletion constructs of the human COL7A1 promoter linked to the CAT gene and treated with TGF-b, as described under “Materials and Methods.” After measuring b-galactosidase activity in each extract, identical amounts were used for constructs 2722 and 2524, whereas extracts corresponding to five times the b-galactosidase activity were used for constructs 2456, 2390, and 2230 to obtain a clearly detectable activity, as the basal activity of these shorter constructs is significantly lower (see Ref. 8). CAT activity was assayed using [14C]chloramphenicol as a substrate, as described under “Materials and Methods.” B, fibroblast cultures were transfected with either 2524COL7A1/CAT or 2524mCOL7A1/CAT promoter constructs. The GT box mutation in the 2524m construct prevents Sp1 binding and reduces promoter activity by ;80% (8). CAT assays were normalized for identical b-galactosidase activity in each sample. In each panel, a representative autoradiogram is shown. Quantitative results are the mean 6 S.D. of six (A) or three (B) separate experiments, each performed with duplicate samples. Values are expressed as relative CAT activity. AC and C indicate the acetylated and unacetylated forms, respectively, of [14C]chloramphenicol.

We have recently demonstrated that a GT box, 2512/2505, binds the transcription factor Sp1 and is crucial for high expression of COL7A1 (8). Sp1 has been shown to play a role in TGF-b-mediated up-regulation of COL1A2 (18), although other transcription factors, such as AP-1, are also involved (18 –20). We, therefore, tested the effect of a functional mutation in this GT box on the TGF-b responsiveness of COL7A1 promoter. For this purpose, the TGF-b responsiveness of mutant 2524mCOL7A1/CAT construct, harboring the GT box mutation, was compared with that of 2524COL7A1/ CAT in transient cell transfection experiments. As anticipated from our previous studies (8), the mutation drastically reduced the basal activity of the promoter. It did not, however, alter its up-regulation by TGF-b. On the other hand, 59-deletion up to position 2456 not only reduced basal promoter activity by ;90% but also totally eliminated TGF-b responsiveness, confirming the data presented in Fig. 1A. These data indicate that the 2512/2505 GT box is not involved in the TGF-b response. TGF-b Induces the Rapid Formation of a Nuclear ProteinDNA Complex Binding to the 2524/2444 Fragment of COL7A1 Promoter—To determine whether the region of COL7A1 promoter shown above to confer TGF-b responsive-


Plasmid Constructs—To study the transcriptional regulation of human type VII collagen gene (COL7A1) expression, transient transfection experiments were performed with various COL7A1 promoter 59deletion fragments cloned into promoterless pBS0CAT vector (13), as described previously (8). Transient Cell Transfections and CAT Assays—Transient cell transfections of human dermal fibroblasts were performed with a calcium phosphate/DNA co-precipitation procedure (14). Briefly, cultured cells were transfected with 10 mg of plasmid DNA and 2 mg of the pRSV-bgalactosidase plasmid DNA to monitor the transfection efficiencies (15). After glycerol shock, the cells were placed in Dulbecco’s modified Eagle’s medium containing 1% fetal calf serum, and TGF-b was added 3 h later. After 40 h of incubation, the cells were rinsed once with phosphate-buffered saline, harvested by scraping, and lysed in 200 ml of reporter lysis buffer (Promega, Madison, WI). The b-galactosidase activities were measured according to a standard protocol (15). Unless stated otherwise, aliquots corresponding to identical b-galactosidase activity were used for each CAT assay with [14C]chloramphenicol as substrate (16), using thin layer chromatography. Following autoradiography, the plates were cut and counted by liquid scintillation to quantify the acetylated [14C]chloramphenicol. Electrophoresis Mobility Shift Assays—Several fragments spanning the region between nucleotides 2524 and 2444 of the COL7A1 promoter responsive to TGF-b were generated by polymerase chain reaction amplification using the plasmid 2722COL7A1 as template and purified by electroelution after electrophoresis in a 2% agarose gel. Each oligonucleotide was used either as a probe or as a competitor in electrophoresis mobility shift assay (EMSA) experiments. Nuclear extracts were isolated from human dermal fibroblasts using a small scale preparation (17), aliquoted in small fractions to avoid repetitive freezethawing, and stored at 280 °C until use. The protein concentration in the extracts was determined using a commercial assay kit (Bio-Rad). Nuclear extracts (5 mg) were incubated for 20 min on ice in binding reaction buffer (10 mM HEPES-KOH, pH 7.9, at 4 °C, 4% glycerol, 40 mM KCl, 0.4 mM EDTA, and 0.4 mM dithiothreitol) in the presence of 1 mg of poly(dI-dC), prior to the addition of [32P]59-end-labeled oligomers (0.05– 0.1 pmol, 2– 6 3 104 cpm) for another 20 min of incubation at 4 °C. For competition experiments, a 1– 60-fold molar excess of unlabeled DNA was added to the binding reaction. For supershift experiments, nuclear extracts were incubated overnight with antisera prior to the binding reaction. Sp1 and c-Jun/AP-1 antibodies were from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). The pan-Smad antibody 367 was described previously (21). Samples were then separated by electrophoresis on a 4% polyacrylamide gel in 0.5 3 Tris borate-EDTA buffer at 200 V for 2 h at 4 °C, fixed for 1 h in 30% methanol, 10% acetic acid, vacuum-dried, and autoradiographed.


ness to a CAT reporter construct contained a putative TGF-b specific cis-element, EMSAs were performed using a radiolabeled 81-bp oligonucleotide spanning the region located between residues 2524 and 2444, encompassing the entire TGFb-responsive COL7A1 promoter region. In the first set of experiments, fibroblast cultures were incubated without or with TGF-b (10 ng/ml) for 24 h prior to nuclear extract preparation. No difference in the EMSA pattern could be detected between nuclear extracts from control cultures and those from TGF-b-treated cultures (not shown). In the second set of experiments, fibroblast cultures were treated with TGF-b (10 ng/ml) and nuclear extracts were prepared at several times thereafter, between 1 min and 3 h. As shown in Fig. 2A, incubation of the nuclear extracts with the 2524/2444 oligonucleotide resulted in the formation of two major, TGF-b-independent DNA-protein complexes, identified as shifts 1 and 2, and present in nuclear extracts from either control or TGF-b-stimulated fibroblasts. These two bands have been previously shown to contain the transcription factor Sp1 (8). An additional complex migrating between complexes 1 and 2 was induced as early as 15 min after TGF-b stimulation, persisted with a similar intensity until the 1-h time point, and disappeared at 3 h. This new complex was not present in nuclear extracts from control fibroblast cultures and will be referred to as the TGF-b-specific band (TBSB).

Accurate determination of the earliest time point at which the TBSB could be observed was performed in a third set of experiments. For this purpose, fibroblast nuclear extracts were prepared every 2 min, from 5 to 15 min, after addition of TGF-b to the fibroblast cultures and tested in EMSAs, using the COL7A1 promoter fragment 2524/2444 as a probe. The results indicate that binding of the TGF-b-specific complex appears with maximal intensity as early as 11 min poststimulation (not shown). To address the specificity of the TBSB and to attempt to identify the protein(s) that it may contain, competition experiments were carried out with several oligonucleotides spanning the 2524/2444 region of COL7A1 promoter, as well as oligonucleotides containing consensus sequences for known transcription factors, Sp1, AP-2, and AP-1. As shown in Fig. 2B, a 30-fold molar excess of unlabeled 2524/2444 fragment efficiently competed all binding associated with the probe (lane 3). Fragment 2524/2491, containing the GT box described above, efficiently competed Sp1-related binding (lane 4), confirming our previous observations (8), whereas fragments 2490/2457 (lane 5) and 2500/2475 (lane 6) failed to compete either Sp1 or TBSB binding. Fragment 2504/2444 (lane 7) competed TBSB binding and most of the Sp1 binding, suggesting that the TBRS is, indeed, comprised within this fragment. As expected from our previous observations (8), an Sp1 oligonucleotide (60-fold molar excess) abolished shifts 1 and 2 but did not alter the TBSB. Also, an AP-2 oligonucleotide significantly displaced both shifts 1 and 2 but without altering TBSB formation (lane 9). Competition of Sp1 binding is likely due to the sequence similarity between Sp1 and AP-2 consensus sequences. Finally, an AP-1 oligonucleotide (lane 10) did not displace either Sp1 binding or the TBSB. Formation of the TBSB Requires Two Distinct Sites on the 2524/2444 Region of COL7A1 Promoter—The next set of experiments was designed to further refine the cis-acting element(s) responsible for the appearance of the TBSB. Toward this end, a series of oligonucleotides was generated, representing a stepwise deletion from either the 59- or 39-ends of the 2504/2444 promoter fragment. Their sequences and relative positions are depicted in Fig. 3A. First, EMSA experiments were performed with nuclear extracts from fibroblast cultures treated for 30 min with TGF-b, using each of the various stepwise deletion oligonucleotides as probes. As expected from the competition experiments presented above, the 2504/2444 fragment efficiently bound the TBSB (Fig. 3B, lane 2). 59-End deletion from nucleotide 2504 to nucleotide 2496 did not affect TBSB formation with oligonucleotides extending to position 2444 in 39 (lanes 4 versus lane 2). However, further 59-end deletion to residue 2490 or 39-end deletion to residue 2453 resulted in complete loss of TGF-b-induced binding activity (lanes 6 and 8, respectively). Three conclusions could be drawn from these experiments. First, the minimal COL7A1 promoter fragment capable of binding the TBSB is 2496/2444. Second, it is likely that two distinct sites, one located within the sequences surrounding nucleotides 2496/2490 at the 59-end of the TBRS and the other between nucleotides 2453/2444 at the 39-end of the TBRS, are simultaneously required for providing the TGF-b-induced binding activity. Third, deletion of the Sp1 binding GT box between 2512 and 2505 did not influence the formation of the TBSB, further indicating that the GT box is not involved in TGF-b response. Finally, despite the removal of the GT box, residual Sp1 binding activity was still observed, suggesting the presence of a secondary Sp1 binding sequence within the 2490/2453 DNA fragment of COL7A1 promoter. The latter observation was confirmed in two subsequent experiments. First, recombinant Sp1 protein was capable of binding


FIG. 2. Binding of fibroblast nuclear proteins to the TGF-bresponsive region of human COL7A1 promoter. Gel mobility shift assays were performed with a labeled oligonucleotide spanning the region from 2524 to 2444 of the COL7A1 promoter. A, nuclear proteins were prepared from fibroblast cultures treated with TGF-b (10 ng/ml) for various lengths of time, as indicated, and binding reactions were carried out as described under “Materials and Methods.” DNA-protein complexes were separated from unbound oligonucleotides by nondenaturing 4% acrylamide gel electrophoresis. Note the appearance of a TBSB as early as 15 min after growth factor addition to the culture medium. Shifts 1 and 2 have been previously characterized as Sp1 transcription factor binding to the DNA probe (Ref. 8). B, competition experiments were carried out using either a 30-fold molar excess (oligonucleotides 2524/2444 and 2504/2444) or 60-fold molar excess (all other nucleotides) of unlabeled competitor. Free probe refers to the unbound radiolabeled oligonucleotides.

13055

13056


FIG. 3. TGF-b-induced transcription factor binding to the responsive segment of human COL7A1 promoter requires two distinct binding sites. EMSAs were performed with several oligonucleotides spanning the region from 2524 to 2444 of the COL7A1 promoter, as probes to determine their ability to bind the TBSB. Nuclear proteins were prepared from fibroblast cultures treated with TGF-b (10 ng/ml) for 30 min, and binding reactions were carried out as described under “Materials and Methods.” DNA-protein complexes were separated from unbound oligonucleotides by nondenaturing 4% acrylamide gel electrophoresis. A, nucleotide sequence of the 59- and 39-end deletions. B, representative autoradiogram using four of the oligonucleotides described in A. Note that the shortest fragment binding the complex is 2496/2444.

this fragment, as determined in gel mobility shift assays, and second, an antibody specifically directed against Sp1 supershifted the weak protein-DNA complexes 1 and 2 generated with the 2504/2444 COL7A1 probe (not shown). Confirmation of the data presented above was provided in a series of competition experiments in which the deletion fragments described in Fig. 3A were utilized to compete the TGFb-induced protein-DNA complex formation with the longer 2524/2444 COL7A1 fragment. Only the fragments containing both discrete regions 2496/2490 and 2453/2444 were able to compete binding of the TBSB to the 2504/2444 probe, whereas shorter fragments lacking either one or both of these elements were unable to compete (not shown). Identification of Smad as Part of the TGF-b-induced DNAProtein Complex—We next attempted to delineate the signaling pathway by which TGF-b induces the formation of the TBSB. For this purpose, fibroblast cultures were first incubated with several inhibitors of different signaling pathways, prior to TGF-b stimulation (30 min) and subsequent nuclear protein preparation. All inhibitors were tested at three concentrations, which have been shown in previous publications to effectively block their target pathways. Protein kinase C inhibitors (GF10923X, staurosporine, calphostin, H7), phospholipase C inhibitors (D609, U73122), tyrosine kinase inhibitors (genistein, tyrphostin 51), the tyrosine phosphatase inhibitor sodium orthovanadate, mitogen-activated protein kinase pathway inhibitors (SB203580, PD 98059), calcium ionophore, pertussis and cholera toxins, and okadaic acid all failed to block TBSB formation (not shown). This lack of effect from all inhibitors tested suggests a rapid mechanism that may be triggered directly through the TGF-b receptors, without a complex cyto-

plasmic cascade of events, involving the several pathways listed above. In this context, Smad proteins were recently identified as immediate-early response factors, which are translocated into the nucleus following their phosphorylation by TGF-b receptor type I (reviewed in Refs. 22–25). To identify the transcription factor(s) participating in the TGF-b-induced complex, we performed supershift assays with antibodies specific for Sp1, c-Jun, and a recently developed pan-Smad antibody, 367 (21).2 As shown in Fig. 4, an Sp1 antibody recognized and supershifted both shifts 1 and 2 without altering the TGF-b-specific band (lane 4). These results were anticipated from our previous demonstration of the essential role played by Sp1 in high basal activity of the COL7A1 promoter (8). Anti c-Jun had no effect on any of the three DNA-protein complexes (lane 5). Interestingly, the anti-Smad polyclonal antibody 367 was able to supershift part of, and to reduce the total amount of, the TBSB (lane 3). Collectively, these data indicate that a member of the Smad family is rapidly complexing to the TGF-b-responsive region of the COL7A1 promoter following treatment of fibroblasts with TGF-b. DISCUSSION

Our studies of the human COL7A1 promoter reported here demonstrate that this gene is an immediate-early target of TGF-b transcriptional activation in dermal fibroblasts, with evidence of formation of a TGF-b-specific transcriptional complex within 11 min after TGF-b addition in a ligand-dependent manner. Smad proteins have recently been identified as critical intracellular mediators of TGF-b family-induced signals (reviewed in Refs. 22–25). Upon ligand binding to TGF-b type I and II receptors, two Smad isoforms, Smad2 and Smad3, are recruited to the type I receptor where they undergo phosphorylation on conserved serine residues in the C terminus. Following activation, these proteins associate with the tumor suppressor Smad4/DPC4 and are translocated to the nucleus where they presumably act in transcriptional complexes (reviewed in Refs. 22–25). In our experiments, supershift assays using a pan-Smad antibody (21) show that the protein-DNA complex, TBSB, contains one or more Smad proteins. Based on signaling specificity, this complex likely contains Smad2 or Smad3 along with Smad4. 2

R. J. Lechleider and A. B. Roberts, unpublished data.


FIG. 4. A Smad member participates in the protein-DNA complex rapidly induced by TGF-b. Supershift experiments were carried out with nuclear extracts from fibroblast cultures treated without (2) or with (1) TGF-b for 30 min. Prior to protein-DNA binding reactions, the nuclear extracts were incubated overnight with polyclonal antibodies against Sp1, c-Jun/AP-1, and a pan-Smad antibody, 367 (Ref. 21). Sp1 and Smad supershifts are indicated.


without additional binding proteins. Experiments are under way to explore these possibilities. Acknowledgment—The expert technical assistance of Ying-Jee Song is gratefully acknowledged. REFERENCES 1. Wetzels, R. H. W., Robben, H. C. M., Leigh, I. M., Schaafsma, H. E., Vooijs, G. P., and Ramaekers, F. C. S. (1991) Am. J. Pathol. 139, 451– 459 2. Uitto, J., Chung-Honet, L. C., and Christiano, A. M. (1992) Exp. Dermatol. 1, 2–11 3. Sakai, L. Y., Keene, D. R., Morris, N. P., and Burgeson, R. E. (1986) J. Cell Biol. 103, 1577–1586 4. Burgeson, R. E. (1993) J. Invest. Dermatol. 101, 252–255 5. Christiano, A. M., and Uitto, J. (1996) Curr. Opin. Dermatol. 3, 225–232 6. Christiano, A. M., Greenspan, D. S., Hoffman, G. G., Zhang, X., Tamai, Y., Lin, A. N., Dietz, H. C., Hovnanian, A., and Uitto, J. (1993) Nat. Genet. 4, 62– 66 7. Christiano, A. M., Ryyna¨nen, M., and Uitto, J. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 3549 –3553 8. Vindevoghel, L., Chung, K.-Y., Davis, A., Kouba, D., Kivirikko, S., Alder, H., Uitto, J., and Mauviel, A. (1997) J. Biol. Chem. 272, 10196 –10204 9. Chen, Y.-Q., Mauviel, A., Ryyna¨nen, J., Sollberg, S., and Uitto, J. (1994) J. Invest. Dermatol. 102, 205–209 10. Ko¨nig, A., and Bru¨ckner-Tuderman, L. (1992) J. Cell Biol. 117, 679 – 685 11. Mauviel, A., Lapie`re, J.-C., Halcin, C., Evans, C. H., and Uitto, J. (1994) J. Biol. Chem. 269, 25–28 12. Chen, M., Petersen, M., Li, H.-L., Cai, X.-Y., O’Toole, E. A., and Woodley, D. T. (1997) J. Invest. Dermatol. 108, 125–128 13. Fazio, M. J., Ka¨ha¨ri, V.-M., Bashir, M.-M., Saitta, B., Rosenbloom, J., and Uitto, J. (1990) J. Invest. Dermatol. 94, 191–196 14. Graham, F. L., and van der Eb, A. J. (1973) Virology 54, 536 –539 15. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 16. Gorman, C., Moffat, L. F., and Howard, B. H. (1982) Mol. Cell. Biol. 2, 1044 –1051 17. Andrews, N. C., and Faller, D. V. (1991) Nucleic Acids Res. 19, 2499 18. Inagaki, Y., Truter, S., and Ramirez, F. (1994) J. Biol. Chem. 269, 14828 –14834 19. Inagaki, Y., Truter, S., Tanaka, S., Di Liberto, M., and Ramirez, F. (1995) J. Biol. Chem. 270, 3353–3358 20. Chung, K.-Y., Agarwal, A., Uitto, J., and Mauviel, A. (1996) J. Biol. Chem. 271, 3272–3278 21. Lechleider, R. J., de Caestercker, M. P., Dehejta, A., Polymeropoulos, M. H., and Roberts, A. B. (1997) J. Biol. Chem. 271, 17617–17620 22. Wrana, J. L., and Attisano, L. (1996) Trends Genet. 12, 493– 496 23. Derynck, R., and Zhang, Y. (1996) Curr. Biol. 6, 1226 –1229 24. Massague´, J., Hata, A., and Liu, F. (1997) Trends Cell Biol. 7, 187–192 25. Heldin, C. H., Miyazono, K., and ten Diijke, P. (1997) Nature 390, 465– 471 26. Chen, X., Rubock, M. J., and Whitman, M. (1996) Nature 383, 691– 696 27. Huang, H.-C., Murtaugh, L. C., Vize, P. D., and Whitman, M. (1995) EMBO J. 14, 5965–5973 28. Chen, X., Weisberg, E., Fridmacher, V., Watanabe, M., Naco, G., and Whitman, M. (1997) Nature 389, 85– 89 29. Liu, F., Pouponnot, C., and Massague´, J. (1998) Genes Dev. 12, 3158 –3167 30. Kim, J., Johnson, K., Chen, H. J., Carroll, S., and Laughon, A. (1997) Nature 388, 304 –308 31. Yingling, J., Datto, M. B., Wong, C., Frederick, J. P., Liberati, N. T., and Wang, X. F. (1997) Mol. Cell. Biol. 17, 7019 –7028


The method of transcriptional activation by Smads has recently begun to be elucidated. Chen et al. (26) identified a DNA-binding protein of the forkhead family (FAST-1) that specifically and inducibly interacts with an activin-responsive element (ARE) in the promoter of the Xenopus homeobox gene Mix.2 (27). FAST-1 also interacts with Smad2 and Smad4/ DPC4 in the ARE binding complex (28). Reconstitution of this system in mammalian cells demonstrates that efficient DNA binding and transcriptional activation require FAST-1, Smad2, and Smad4 in the same complex (29). Our data represent the first example of a mammalian promoter directly regulated by interactions with endogenous Smad proteins and as such make COL7A1 the first example of an immediate-early gene regulated by TGF-b in a Smad-dependent fashion. Although the cis-elements necessary for interaction with Smad transcriptional complexes are not known, some architectural constraints may be deduced from comparison of the TBRS in the COL7A1 promoter with the ARE from the Mix.2 gene. Chen et al. (26) identified a direct 6-bp AAATGT repeat separated by 11 bp, and this sequence was used to clone the DNA binding protein FAST-1. Mutation of either element of this repeat disrupts DNA binding by FAST-1 (26). Similarly, the COL7A1 element contains two 5-bp ATGGC repeats, two adjacent CAGA repeats in the 59end and two pairs of 4-bp repeats, CCCA and ACAG. Deletion studies suggest that two elements are necessary for Smad/DNA interaction, as both the COL7A1 2490/2444 and 2496/2453 probes failed to bind the TGF-b-inducible complex, whereas the 2496/2444 fragment was the minimum required to generate a gel-shifted band. This suggests a mechanism similar to that observed with the Mix.2 ARE, with two separated elements required for full activity (26). The nature of the exact cis-elements required for DNA-Smad interaction is currently under investigation. Whether a Smad protein binds the COL7A1 promoter directly cannot be ascertained from our data. The Drosophila Smad1 homologue, the mad gene product, can bind elements in the vestigial promoter directly, and a consensus Mad binding sequence (GCCGnCGc) that can bind recombinant Mad protein has been identified (30). Similarly, the tumor suppressor Smad4 can bind DNA directly (31), but this has not been convincingly demonstrated in an endogenous complex. In the COL7A1 promoter, an accessory DNA binding protein may be required, as in the Mix.2 ARE, or a Smad or Smads may bind the TGF-b-responsive elements

13057

NUCLEIC ACIDS, PROTEIN SYNTHESIS, AND MOLECULAR GENETICS: Smad-dependent Transcriptional Activation of Human Type VII Collagen Gene (COL7A1) Promoter by Transforming Growth Factor- β

J. Biol. Chem. 1998, 273:13053-13057. doi: 10.1074/jbc.273.21.13053

Access the most updated version of this article at http://www.jbc.org/content/273/21/13053 Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites. Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 30 references, 12 of which can be accessed free at http://www.jbc.org/content/273/21/13053.full.html#ref-list-1


Laurence Vindevoghel, Atsushi Kon, Robert J. Lechleider, Jouni Uitto, Anita B. Roberts and Alain Mauviel