INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE BENEFICIAL EFFECT of female gender on the development and progression of atherosclerosis may reflect, in part, reduced accumulation of vascular wall extracellular matrix (27). Consistent with this hypothesis, estrogen administration reduces collagen deposition in the aorta of hypertensive and hypercholesterolemic animals in vivo and reduces type I and type III collagen synthesis by vascular smooth muscle cells in vitro (22). Analogous to its effects on atherosclerosis, female gender has been associated with a slower rate of progression of renal disease (27). The accumulation of glomerular extracellular matrix after renal injury is a precursor to the development of glomerular obsolescence and progressive loss of renal function (18). In this context, the ability of estradiol to suppress mesangial cell collagen synthesis may translate into reduced accumulation of collagen after glomerular injury and thereby limit the development of glomerulosclerosis (18). It should be noted, however, that female gender has been associated with an accelerated course in several experimental models of renal disease, whereas in other models, male gender is responsible for enhanced renal injury (2, 27).

Transforming growth factor- (TGF-) has been found to play an important role in mediating progressive renal injury (5). Accumulation of collagen in the glomerular mesangium and renal interstitium contributes to the adverse effects of TGF- on the progression of chronic renal disease (5). In mesangial cells, TGF- increases steady-state mRNA and immunoreactive protein for the ₁- and ₂-chains of types I and IV collagen (11, 28). This stimulation occurs at the transcriptional level as reflected in an increase in steady-state mRNA in the absence of an increase in mRNA transcript stability (11, 28). In addition, TGF- inhibits collagen degradation by decreasing the synthesis of matrix degrading proteases and by increasing the synthesis of protease inhibitors (5).

In the present study, we examined the interaction between estradiol and TGF-1 on collagen synthesis. We found that estradiol inhibited the stimulatory effect of TGF-1 on ₁-collagen IV (COL4A1) gene transcription and on collagen IV protein synthesis. We also showed that treatment with TGF-1 enhanced the binding of mesangial cell nuclear extracts to an Sp1 consensus binding sequence oligonucleotide and to an Sp1 binding site in the collagen IV promoter and that this enhanced binding was reversed by estradiol. The ability of estradiol to inhibit TGF-1-stimulated collagen IV synthesis may contribute to the protective effect of female gender on the progression of renal disease by reducing glomerular collagen accumulation.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Isolation and characterization of murine mesangial cells. Mesangial cells were isolated from kidneys of 8- to 10-wk-old naive male SJL/J (H-2^S) mice by differential glomerular sieving. The present studies were performed on an immortalized, differentiated murine mesangial cells line transformed with nonreplicating, noncapsid-forming SV40 virus (strain Rh 911). The cells express receptors for angiotensin II and stain positive for Thy-1.1 antigen, desmin, vimentin, and types I and IV collagens but fail to bind antibody directed against a proximal tubular antigen (29). Receptors for estradiol have been demonstrated in mesangial cells isolated from male rats and in murine renal tissue (22).

Construction of reporter gene plasmids. A plasmid containing portions of the gene encoding murine ₁(IV) collagen linked to luciferase coding sequences to form a reporter construct HB35 has previously been described (10). A 2.3-kb Xho I fragment derived from p184 (gift of Dr. P. Killen, University of Michigan) (6, 15) spanning the first three exons, the first two introns, and a portion of the third intron of the ₂(IV) collagen gene, the intergenic bidirectional promoter region, and a portion of the ₁(IV) collagen transcription unit was cloned into pBluescript (Stratagene, La Jolla, CA) and was truncated within the third intron of the ₂(IV) gene by excising the appropriate fragment with Hind III. The fragment was then ligated into the Hind III site of the luciferase expression plasmid, pSVOAlucD5' (9), and its orientation was determined by restriction mapping (17). An additional 3.2-kb fragment of the first intron of the ₁(IV) gene previously shown to modulate the activity of the ₁(IV) promoter (6, 15), corresponding to a BamH I-linked 3.2-kb Xba fragment of the first intron of ₁(IV), was inserted into the BamH I site downstream from the 3' end of the luciferase coding sequences following partial BamH I digestion (17). Intron fragment orientation and position were confirmed by restriction mapping (17).

Stable transfection. Murine mesangial cells were cotransfected with HB35 and a selection plasmid pSV2-Neo encoding for neomycin-resistance at a molar ratio of 10:1 using the CaPO₄-DNA precipitation procedure (1). Cells were then grown in selection medium containing Geneticin (G-418). Cells surviving in medium with Geneticin were expanded and reselected with the neomycin analog. Stable transfectants exhibited patterns of growth behavior and protein synthesis similar to the parent cell line (10).

Luciferase assay. SV40-transformed mesangial cells were grown in RPMI media (GIBCO; Life Technologies) containing 10% fetal calf serum and 1% penicillin/streptomycin (GIBCO). The estradiol concentration in the 10% fetal calf serum-supplemented media was 9 × 10¹¹ M. Cells were plated in six-well plates and grown in phenol-free, serum-free RPMI. Cells were exposed to TGF-1 (2 ng/ml; R & D Systems, Minneapolis, MN) in the presence and absence of 17-estradiol (10¹⁰-10⁷ M; Sigma Chemical, St. Louis, MO) or 17-estradiol (10⁷ M; Sigma) for 24 h. In some experiments cells were also exposed to ICI-182,780 (Tocris Cookson, Ballwin, MO), a estrogen receptor antagonist. Cells were washed twice with phosphate-buffered saline, then lysed with 100 µl reporter lysis buffer (Promega, Madison, WI) at room temperature for 15 min. Wells were then scraped, and the cell lysate was transferred to a microcentrifuge tube and placed on ice. Tubes were vortexed and microcentrifuged for 2 min at 4°C. The suspension was transferred to a new microcentrifuge tube and stored at 70°C until assayed. Twenty microliters of cell extract was mixed with 10 µl of assay reagent [20 mM tricine, 1.07 mM (MgCO₃)₄Mg(OH)₂ · 5H₂O, 2.67 mM MgSO₄, 0.1 mM EDTA, 33.3 mM dithiothreitol (DTT), 270 µM coenzyme A, 470 µM luciferin, and 530 µM ATP, pH 7.8] at room temperature (all reagents from Sigma). Light emission was measured directly at room temperature over a 10-s period in a luminometer (Promega). Blank reactions were determined with equivalent volumes of lysis buffer substituted for cell lysates, and these values were subtracted from experimental values. Luciferase activity was expressed per milligram of cellular protein in the supernatant as determined by the Bio-Rad protein assay (Bio-Rad, Richmond, CA). Relative luciferase units were calculated as percentages of control values, where 100% is the value obtained with control media (serum free, no added agents).

Immunoprecipitation. Cells were plated on six-well plates (Becton-Dickinson, Franklin Lakes, NJ) and grown in phenol-free, serum-free DMEM media (GIBCO). Cells were exposed to TGF-1 (2 ng/ml; R & D Systems) in the presence or absence of estradiol (Sigma) or control media for 24 h in the presence of ascorbic acid (50 µg/ml; Sigma), -aminopropionitrile (80 µg/ml; Sigma), and [³H]proline (15 µCi/ml; Amersham Life Science, Arlington Heights, IL). Aliquots of the culture medium were collected and mixed with equal volumes of RIPA buffer consisting of 50 mM Tris (pH 7.5), 150 mM NaCl, 0.1% SDS, 1% Triton X-100, 0.5% deoxycholate, and 1 mM phenylmethylsulfonyl fluoride (PMSF) (all from Sigma). Normal goat serum (Sigma) was added followed by a 10% suspension of protein A-positive Staphylococcus aureus (Boehringer Mannheim, Indianapolis, IN). This mixture was incubated on a rocker for 1 h at 4°C. Nonspecifically bound material was removed by centrifugation for 10 min at 5,000 rpm. Collagens remaining in the supernatant after centrifugation were specifically precipitated using goat anti-human type I collagen (1:250 dilution; Southern Biotechnology, Birmingham, AL) or goat anti-bovine type IV collagen (1:250 dilution; Southern Biotechnology). After 1 h of incubation, rabbit anti-goat antibody (Sigma) and the protein A suspension were added, and the immunoprecipitates were recovered by centrifugation. The resulting pellet was washed three times with cold RIPA buffer and then resuspended in Laemmli sample buffer by boiling (2 min). Tritium content of 10-µl aliquots of the suspension were measured by liquid scintillation counting. The suspension was then run on a 7.5% SDS polyacrylamide gel. The resultant bands were enhanced by autoradiography (26).

Western blotting for types I and IV collagen. Cells were grown as described above. Media was collected and concentrated using an Amicon Centriprep-10 concentrator (Grace, Beverly, MA). Protein content was measured in a 0.1-ml aliquot (Bio-Rad Protein Assay). To prepare the samples for SDS-PAGE electrophoresis, 1 ml of 10% trichloroacetic acid was added to 4 mg protein of concentrated media, and the final volume was brought to 2 ml with distilled water. The sample was vortexed and then centrifuged at 2,000 rpm for 10 min. The pellet was dissolved in 1 ml of loading buffer (2% SDS, 10% glycerol, 50 nM Tris, 3% bromophenol blue, and 2% -mercaptoethanol, pH 7.6), boiled for 5 min, then immediately placed on ice. A quantity of 25, 50, 75, or 100 µg of sample was loaded for electrophoretic separation of proteins. SDS-PAGE electrophoresis was performed by standard techniques, and proteins were transferred to a polyvinylidene difluoride microporous membrane.

After blotting, the membrane was immediately placed into blocking buffer [2% bovine serum albumin in wash buffer (10 mM Tris, 100 mM NaCl, 0.1% Tween 20, distilled H₂O added to 1 liter)] on a shaking apparatus for 30 min at 37°C. The blocking buffer was discarded and replaced with new blocking buffer containing goat anti-bovine type I or type IV collagen antibody (1:250 dilution; Southern Biotechnology). The membrane was incubated with the primary antibody solution on a shaking apparatus for 30 min at 37°C. The membrane was next washed for 30 min with agitation. The membrane was then placed in 5% nonfat milk in wash buffer containing goat anti-mouse IgG conjugated to horseradish peroxidase (Sigma). The antibody conjugate was allowed to incubate for 30 min at 37°C with agitation. After washing, the membrane was treated with enhanced chemiluminescence reagent (Amersham Life Science) according to the instructions of the manufacturer. Kodak X-Omat 4R film was exposed to the blot for 1 min. Bands were quantitated by laser densitometry. Human type IV collagen (Sigma) was used as a positive control in Western blotting experiments. Appropriate negative controls using irrelevant antibodies were also performed.

Preparation of nuclear extracts. Mesangial cells were scraped into phosphate-buffered saline and pelleted by centrifuging at 800 g for 10 min at 4°C. The cells were resuspended in hypotonic buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 0.2 mM PMSF, 0.5 mM DTT; all reagents from Sigma), recentrifuged at 3,000 g for 5 min, resuspended in the hypotonic buffer, and allowed to swell for 10 min on ice. The cells were homogenized in a glass Dounce homogenizer. Nuclei were collected by centrifuging at 3,300 g for 15 min. The nuclear pellet was resuspended in low-salt buffer (20 mM HEPES, 25% glycerol, 1.5 mM MgCl₂, 0.02 M KCl, 0.2 mM EDTA, 0.2 mM PMSF, and 0.5 mM DTT). With gentle stirring, a volume of high-salt buffer (KCl concentration of 1.2 M) equal to one-half the packed nuclear volume was added dropwise. The protein concentration was measured by colorimetric assay (Bio-Rad).

Synthesis of a COL4 promoter oligodeoxynucleotide. An oligodeoxynucleotide containing the first 60 base pairs of the collagen type IV promoter, relative to the ₂(IV) transcription start site, was synthesized by the core DNA facility of the Albert Einstein College of Medicine Human Genetics Program with an automated solid-phase DNA synthesizer (Applied Biosystems, Foster City, CA) and purified with HPLC.

Electrophoretic mobility shift assay (gel shift assay). Four micrograms of nuclear extract were mixed with 2 mg of poly(dI:dC) in 20 µl of a reaction buffer consisting of 25 mM HEPES, pH 7.5, 1.2 mM DTT, 4 mM MgCl₂, 150 mM NaCl, 5% glycerol, 0.005% bromphenol blue, and 0.05% Nonidet P-40 (24). The mixture was incubated on ice for 15 min, followed by the addition of 10 fmol of ³²P end-labeled Sp1 consensus binding sequence oligonucleotide or the ³²P end-labeled COL4A1 promoter oligonucleotide. The incubation was continued for 30 min. The incubation mixture was then subjected to electrophoresis on a 6% polyacrylamide gel in Tris-glycine buffer. The gels were dried, and autoradiography was performed at 70°C with an intensifying screen. Bands were quantitated by laser densitometry (Molecular Dynamics, model 300S). Competition experiments were performed with a 200-fold excess of unlabeled Sp1 consensus binding sequence oligonucleotide (Promega). Supershift assays were performed by preincubating nuclear extracts with anti-Sp1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) for 0.5 h prior to the binding reaction.

Statistics. For each individual experiment, the mean of replicate determinations in three individual wells was calculated. The data are expressed as means ± SE. Differences among groups were tested by analysis of variance with the Scheffé correction or Student's two-tailed t-test for independent variables. P < 0.05 was considered a significant difference.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

TGF-1 significantly increased COL4A1 gene transcription (140.5 ± 6.2 relative luciferase units, expressed as a percent of control untreated cells, P < 0.001, n = 11) (Fig. 1). Estradiol reversed the stimulatory effects of TGF-1 on reporter gene transcription in a dose-dependent manner [for 2.5 × 10⁹ M, 114.2 ± 0.2, P < 0.002 vs. TGF-1; for 10⁷ M, 89.5 ± 4.0, P < 0.001 vs. TGF-1 and P = not significant (NS) vs. control, n = 9] (Fig. 1). 17-Estradiol (10⁷ M), a time-dependent mixed agonist/antagonist of 17-estradiol, failed to reverse TGF-1-stimulated gene transcription (135.5 ± 5.5, P = NS vs. TGF-1, n = 3). ICI-182,780 (10⁶ M), a high-affinity estrogen receptor antagonist, prevented estradiol from reversing TGF-1-stimulated gene transcription (139.7 ± 8.1, P = NS vs. TGF-1, n = 3). In contrast, neither estradiol (10¹⁰-10⁷ M), 17-estradiol (10⁷ M), nor ICI-182,780 (10⁶ M) had any effect on reporter gene transcription in serum-free media in the absence of TGF-1 (data not shown). Neither TGF-1 nor estradiol had any effect on total cellular protein [control, 221 ± 40 µg/well; TGF-1, 195 ± 35 µg/well; estradiol (10⁷ M), 187 ± 35 µg/well; TGF-1 + estradiol, 205 ± 28 µg/well, P = NS].

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Effects of estradiol on transforming growth factor-1 (TGF-1)-stimulated type IV collagen gene transcription by mesangial cells grown in serum-free media as assessed by the activity of a collagen IV/luciferase gene construct. All groups were treated with TGF-1 (2 ng/ml) and the indicated concentrations of 17-estradiol. The group represented by the last bar was also treated with ICI-182,780 (ICI, 106 M). Results are expressed as a percent of control, untreated cells. * P < 0.002 vs. TGF-1 (first bar). ** P < 0.02 vs. TGF-1.  P < 0.001 vs. TGF-1.

We measured the effects of TGF-1 on mesangial cell type I and IV collagen synthesis by immunoprecipitation techniques. TGF-1 (2 ng/ml for 24 h) significantly increased both type I and type IV collagen synthesis (type I, 162.2 ± 16.2, expressed as a percent of control untreated cells, P < 0.023 vs. control, n = 7; type IV, 175.3 ± 14.7, P < 0.001 vs. control, n = 9) (Figs. 2 and 3). Estradiol (10⁷ M) reversed the stimulatory effects of TGF-1 on type IV collagen synthesis (175.3 ± 14.7 vs. 111.6 ± 7.1, P < 0.001, n = 9) (Fig. 2) but did not affect TGF-1-stimulated type I collagen synthesis (162.2 ± 16.2 vs. 166.9 ± 18.8, P = NS, n = 7) (Fig. 3). Estradiol alone had no effect on collagen synthesis in serum-free media (type I, 110.0 ± 12.3, P = NS vs. control; type IV, 101.7 ± 7.7, P = NS vs. control) (Figs. 2 and 3). Results obtained with Western blotting closely paralleled those obtained with immunoprecipitation techniques (Fig. 4).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effects of TGF-1 and estradiol on type IV collagen synthesis by mesangial cells grown in serum-free media as assessed by immunoprecipitation techniques. Inset: representative autoradiogram. Con, control; E, estradiol. * P < 0.001 vs. control and P < 0.001 vs. TGF-1+E.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of estradiol and TGF-1 on type I collagen synthesis by mesangial cells grown in serum-free media as assessed by immunoprecipitation techniques. Inset: representative autoradiogram. * P < 0.023 vs. control. ** P < 0.013 vs. control.

View larger version (46K): [in this window] [in a new window]   Fig. 4.   Representative Western blot for type IV and type I collagen in the supernatant of mesangial cells treated with the indicated agents. Estr, estradiol.

Nuclear extracts from mesangial cells grown in the presence of TGF-1 (2 ng/ml for 24 h) showed enhanced binding to an Sp1 consensus binding sequence oligonucleotide and to an Sp1 binding site contained within the first 60 base pairs of the 5' end of the type IV collagen promoter (Fig. 5, Sp1 consensus binding sequence oligonucleotide; Figs. 6 and 7: COL4A1 promoter oligonucleotide). Specific binding of Sp1 was examined by supershift assay. Preincubation with antibody to Sp1 reduced the binding of nuclear extracts from cells treated with TGF-1 to the Sp1 consensus binding sequence oligonucleotide (Fig. 5) and to the COL4A1 promoter oligonucleotide (not shown). Addition of unlabeled Sp1 consensus binding sequence oligonucleotide to the binding reaction in 200-fold excess markedly reduced the binding of nuclear extracts from cells treated with TGF-1 to the labeled COL4A1 promoter oligonucleotide (Fig. 6). Incubation of mesangial cells with estradiol (10⁷ M) reversed TGF-1-stimulated binding of nuclear extracts to the Sp1 consensus binding sequence oligonucleotide (Fig. 5) and to the COL4 promoter oligonucleotide (Figs. 6 and 7).

View larger version (94K): [in this window] [in a new window]   Fig. 5.   Left: binding of nuclear extracts from mesangial cells grown in presence of TGF-1 (2 ng/ml) or in control media to a 32P-labeled Sp1 consensus binding sequence oligonucleotide. The last lane shows a supershift assay in which nuclear extracts from cells exposed to TGF-1 were preincubated with anti-Sp1 antibody prior to the binding reaction. Right: binding of nuclear extracts from mesangial cells grown in presence of control media, TGF-1 (2 ng/ml), estradiol (107 M), or estradiol + TGF-1 to a 32P-labeled Sp1 consensus binding sequence oligonucleotide.

View larger version (126K): [in this window] [in a new window]   Fig. 6.   Binding of nuclear extracts from mesangial cells grown in presence of control media, TGF-1 (2 ng/ml), TGF-1 + estradiol (107 M), or estradiol alone (107 M) to a 32P-labeled COL4 promoter oligonucleotide. The last lane shows nuclear extracts from TGF-1-treated cells that were incubated with the labeled COL4A1 promoter oligonucleotide in presence of a 200-fold excess of unlabeled Sp1 consensus binding sequence oligonucleotide. Oligo, oligonucleotide; COL4A1, 1 chain of type IV collagen.

View larger version (10K): [in this window] [in a new window]   Fig. 7.   Densitometry readings from gel shift assays (n = 3) measuring the binding of nuclear extracts from mesangial cells grown in presence of control media, TGF-1 (2 ng/ml), TGF-1 + estradiol (107 M), or estradiol (107 M) to a 32P-labeled COL4A1 promoter oligonucleotide. The last bar represents nuclear extracts from TGF-1-treated cells that were incubated with the labeled COL4A1 promoter oligonucleotide in presence of a 200-fold excess of unlabeled Sp1 consensus binding sequence oligonucleotide. * P < 0.001 vs. control, P < 0.001 vs. TGF-1 + E, P < 0.001 vs. E, and P < 0.001 vs. TGF- + excess cold Sp1 oligonucleotide. ** P < 0.01 vs. control, P < 0.001 vs. TGF-1, P < 0.001 vs. TGF-1 + E, and P < 0.001 vs. E.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We examined the effects of estradiol on the increase in COL4A1 gene transcription and type IV collagen synthesis induced by exogenous TGF-1. Estradiol suppressed TGF-1-stimulated COL4A1 reporter gene transcription. Estradiol also reversed TGF-1-stimulated type IV collagen synthesis as assessed by immunoprecipitation and Western blotting techniques.

The concentrations of estradiol used in the reporter gene assay (10¹⁰-10⁷ M) encompass levels observed in normal cycling women and in women receiving hormonal replacement therapy (7). The serum concentration of estradiol in the early follicular phase is 2 × 10¹⁰ M and rises to 1.2-2.6 × 10⁹ M in the late follicular phase (7). Estrogen replacement therapy with a transdermal delivery system achieves a serum estradiol concentration of 0.9 to 4.4 × 10¹⁰ M (7). Oral estrogen replacement therapy produces an estradiol concentration of 0.73 to 2.0 × 10¹⁰ M (21). However, peak serum levels two- to threefold higher are transiently achieved after a single oral dose (21). In our study, estradiol in a concentration of 2.5 × 10⁹ M (equivalent to the peak level achieved in the late follicular phase) significantly antagonized TGF-1-stimulated COL4A1 gene transcription. The lack of effect of the 17-isomer of estradiol, an inactive congener, and the antagonism afforded by ICI-182,780, an estrogen receptor antagonist, suggest that the reversal of the actions of TGF-1 by estradiol is receptor mediated.

We have previously shown that serum stimulates murine mesangial cells to increase type IV collagen gene transcription, in part, by increasing autocrine production of TGF- (19). We also showed that estradiol suppresses type IV collagen synthesis by serum-stimulated mesangial cells but has no effect on collagen synthesis by mesangial cells grown in the absence of serum (18). The results of the present study suggest that the ability of estradiol to inhibit serum-stimulated type IV collagen synthesis may be mediated by the ability of estradiol to antagonize the actions of TGF- (18).

In addition to its effect on mesangial cell collagen synthesis, TGF- induces cellular hypertrophy and increases total cellular protein normalized for cell number (3). Although we failed to find any increase in total cellular protein when expressed per well, our data are nevertheless consistent with cellular hypertrophy since TGF- suppresses mesangial cell proliferation (5).

We have shown that nuclear extracts from mesangial cells treated with TGF-1-1 show increased binding to an Sp1 consensus binding sequence oligonucleotide and to an Sp1 site in the COL4A1 promoter and that estradiol reverses this enhanced binding. These observations are consistent with the hypothesis that Sp1 is involved in the regulation of collagen IV synthesis by TGF-1 and that estradiol reverses the stimulatory effects of TGF-1 on COL4A1 gene transcription via interactions with Sp1. However, further studies are required to confirm this hypothesis. Although our data establish a correlation between Sp1 binding and the effects of estradiol and TGF-1 on COL4A1 gene transcription, they do not establish a cause and effect relationship.

The murine ₁ and ₂ type IV collagen genes are located on chromosome 13 in a head-to-head orientation separated by a common bidirectional 130-bp promoter region that contains two GC box binding sites for Sp1 (GGGCGG), a retinoic acid responsive element and a CAATT box/CTF binding site (6). Kuncio et al. (17) have recently shown that the murine type IV collagen promoter is sufficient to confer positive regulation by TGF-1 in murine renal tubular cells (17). However, there may be other, as yet uncharacterized, enhancer elements involved in the regulation of the intact gene that were not contained in the promoter minigene construct studied by these authors.

Bruggeman et al. (4) showed that nuclear extracts from mouse Engelbreth-Holm-Swarm tumor cells bind to the murine promoter at two regions that contain a novel CTC motif and an Sp1-like binding site. Neither of the two consensus Sp1 binding sites were contained within the protected regions (4). However, binding reactions were found to be cell type specific, as protein/DNA complexes formed with nuclear extracts from Engelbreth-Holm-Swarm cells showed different mobility on gel shift assay than did nuclear extracts from differentiated F9 cells (4).

A protein from human fibrosarcoma cell extracts forms a specific complex with an oligonucleotide containing the central GC box sequence of the human collagen IV promoter (25). Competition studies identified this protein as Sp1 (25). Mutation of the GC box resulted in reduced transcription of a collagen IV promoter reporter gene construct, suggesting that binding of Sp1 to the central GC box motif is necessary for maximal transcriptional activity (25).

Interactions with Sp1 or Sp1-containing complexes mediate TGF--induced transcriptional activation of the genes coding for ₁(I) collagen, ₂(I) collagen, and the cyclin-dependent kinase inhibitors p21 and p15^INK4B (8, 12-14, 20). Estrogen receptors also interact with Sp1 via competition for essential cofactors or via protein/protein interactions that are cell type specific and promoter specific (16, 23). Estrogen-receptor complexes can bind to the transcription factor Sp1 to alter gene transcription (16). The sequences flanking the Sp1 binding site must be important in determining interactions among Sp1, TGF-1, and estradiol, since estradiol does not antagonize the actions of TGF-1 in all genes that contain an Sp1 regulatory motif in their TGF- responsive elements (e.g., type I collagen genes) (12).