MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oda, E.
Right arrow Articles by Oda, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oda, E.
Right arrow Articles by Oda, K.

Mol Cell Biol, August 1998, p. 4772-4782, Vol. 18, No. 8
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.

Cloning and Characterization of a GC-Box Binding Protein, G10BP-1, Responsible for Repression of the Rat Fibronectin Gene

Eri Oda, Kenna Shirasuna, Mitsuhiro Suzuki, Kuniko Nakano, Takuma Nakajima, and Kinichiro Oda*

Department of Biological Science and Technology, Science University of Tokyo, Noda-shi, Chiba 278, Japan

Received 2 January 1998/Returned for modification 15 May 1998/Accepted 26 May 1998

    ABSTRACT
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Fibronectin (FN) is an extracellular matrix protein that connects the extracellular matrix to intracellular cortical actin filaments through binding to its cell surface receptor, alpha 5beta 1, a member of the integrin superfamily. The expression level of FN is reduced in most tumor cells, facilitating their anchorage-independent growth by still unclarified mechanisms. The cDNA clone encoding G-rich sequence binding protein G10BP-1, which is responsible for repression of the rat FN gene, was isolated by using a yeast one-hybrid screen with the G10 stretch inserted upstream of the HIS3 and lacZ gene minimal promoters. G10BP-1 comprises 385 amino acids and contains two basic regions and a putative zipper structure. It has the same specificity of binding to three G-rich sequences in the FN promoter and the same size as the G10BP previously identified in adenovirus E1A- and E1B-transformed rat cells. Expression of G10BP-1 is cell cycle regulated; the level was almost undetectable in quiescent rat 3Y1 cells but increased steeply after growth stimulation by serum, reaching a maximum in late G1. Expression of FN mRNA is inversely correlated with G10BP-1 expression, and the level decreased steeply during G1-to-S progression. This down regulation was strictly dependent on the downstream GC box (GCd), and base substitutions within GCd abolished the sensitivity of the promoter to G10BP-1. In contrast, the level of Sp1, which competes with G10BP for binding to the G-rich sequences, was constant throughout the cell cycle, suggesting that the concentration of G10BP-1 relative to that of Sp1 determines the expression level of the FN gene. Preparation of glutathione S-transferase pulldowns of native proteins from the cell extracts containing exogenously or endogenously expressed G10BP-1, followed by Western blot analysis, showed that G10BP-1 forms homodimers through its basic-zipper structure.

    INTRODUCTION
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Fibronectin (FN) is a large glycoprotein of the extracellular matrix that binds to its cell surface receptor, alpha 5beta 1, a member of the integrin superfamily, as a dimer (2, 15, 39). FN consists of multiple rodlike domains, and each domain binds to a component of the extracellular matrix such as collagen and heparin and to its receptor (16). The cytoplasmic domain of the receptor binds to bundles of actin filaments known as stress fibers indirectly through binding to the attachment proteins (14). The binding of FN to its receptor at so-called focal contact sites therefore connects the cytoskeleton with the extracellular matrix (3, 17, 21, 39), governing cell shape, adhesion, and movement. FN is also involved in the regulation of cell proliferation and differentiation through organization of the extracellular matrix and cytoskeleton (7, 15).

The expression level of FN is closely linked to the growth potential of cells. The level increases when cells cease growing, and senescent cells inevitably express FN at high levels (11, 28, 33). On the contrary, FN expression is greatly inhibited upon neoplastic transformation (10, 20, 35), and most tumor cells express very low levels of FN, resulting in disorganization of the cytoskeleton and extracellular matrix, which facilitates tumor metastasis.

Rat 3Y1-derivative cell line XhoC, transformed by the adenovirus E1A and E1B genes, also expresses a very low level of FN (35). Analysis of transcription factors in XhoC cells revealed an adenovirus E1A-inducible negative regulator, G10BP, which binds to three G-rich sequences in the FN promoter (43). One of the sequences, located at positions -239 to -230, consists of only G residues (G10 stretch), and the other two, located at positions -105 to -95 and -54 to -44, consist of the G10 stretch with one internal C residue insertion (GC boxes). Transcription factor Sp1 also binds to these G-rich sequences. Transcription of an FN promoter-CAT (chloramphenicol acetyltransferase) fusion gene in HeLa cell extract, which contains abundant Sp1, was inhibited by the addition of purified G10BP. The downstream GC box (GCd) seemed to be the most critical site, and G10BP inhibited FN promoter activity primarily by excluding the binding of Sp1 to this site (43). The human FN promoter also contains multiple G-rich sequences (5).

In the present study, to correlate the function of G10BP with its structural features and control of expression, cDNA clones whose protein products bind to the G10 stretch were searched for by using the yeast one-hybrid system (3, 40). A tester strain containing the HIS3 and lacZ genes fused to the G10 stretch and minimal promoters was transformed with a pGAD-XhoC cell cDNA library, which directs the synthesis of fusion proteins between the cDNA-encoded polypeptides and the transcriptional activation domain of Gal4. A positive clone that was isolated encodes a G-rich sequence binding protein with a putative zipper structure, and the protein was designated G10BP-1. G10BP-1 comprises 385 amino acids and has properties identical to those of the previously identified G10BP protein with respect to specificity of binding to three G-rich sequences and electrophoretic mobility. Expression of G10BP-1 was undetectable in quiescent 3Y1 cells but was induced steeply after growth stimulation by serum or adenovirus E1A concomitant with the decrease in FN promoter activity. Glutathione S-transferase (GST) pulldown experiments performed with cell extracts containing exogenously or endogenously expressed G10BP-1 indicated that G10BP-1 forms homodimers.

    MATERIALS AND METHODS
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Cell lines. 3Y1-B cell line clone 1-6 is a clonal line of Fischer rat embryo fibroblasts (24). The XhoC cell line was established by transformation of 3Y1 cells with the adenovirus type 2 E1A and E1B genes (35). g12 cells were established by transfection of 3Y1 cells with PM12SG, in which the adenovirus type 2 E1A 12S cDNA (49) was placed downstream of the mouse mammary tumor virus long terminal repeat (29). The YG1 cell line was established by introducing pCMV-G10BP-1 and pSV2neo (42) into 3Y1 cells and expresses G10BP-1 constitutively. These cell lines were cultivated at 37°C in Dulbecco's modified Eagle's minimal essential medium with 10% fetal calf serum (FCS).

Construction of a pGAD-XhoC cDNA library. Total cellular RNA was prepared from XhoC cells as described by Okayama et al. (37), and poly(A)+ RNA was reverse transcribed by using an oligo(dT) primer with an XhoI site. The RNA strand of the mRNA-cDNA hybrid was replaced with the corresponding DNA strand by using Escherichia coli RNase H, E. coli DNA polymerase I, and E. coli DNA ligase (37), and the cDNA was ligated to EcoRI linkers after both termini were blunt ended. After cleavage with EcoRI and XhoI, the cDNA (100 ng) was ligated to 100 ng of EcoRI/SalI-digested yeast expression plasmid pGAD424 (Clontech) by incubation at 16°C for 48 h. The ligated cDNA was extracted with phenol-chloroform, ultrafiltered by using micron 10 (Millipore), and electrotransformed into E. coli DH10B with an E. coli Pulser (Bio-Rad) to generate the pGAD-XhoC cDNA library (Fig. 1B). This library directs the expression of fused proteins between the DNA-binding domain of Gal4 and cDNA-encoded polypeptides from the crippled ADH1 promoter and replicates autonomously as plasmids in yeasts. The library contained 3.1 × 105 primary recombinants with an average cDNA size of about 0.9 kb.


View larger version (20K):
[in this window]
[in a new window]
  		FIG. 1.   Three G-rich sequences in the rat FN promoter and cloning of the G10BP cDNA with a yeast one-hybrid screen. (A) The arrow designated +1 indicates the start site of transcription (35). The positions of the TATA box, GCu, GCd, the G10 stretch (G10), a cAMP-responsive element (TGACGTCA), two CAT boxes, and the AP-1-like motif are shown together with the nucleotide numbers. The affinity of purified G10BP binding to G10, GCu, GCd, and these G-rich sequences carrying four-base substitutions outside the Sp1 motif (GGGCGG) (G10m, GCum, and GCdm) is from reference 44. +++, very strong affinity; ++, strong affinity; +, weak affinity; -, no affinity. (B) A yeast tester strain, YMHL-G10, used for screening of the G10BP cDNA was constructed by introduction of pHISi-G10 and pLacZi-G10, which carry three multimerized 24-bp oligonucleotides containing the G10 stretch upstream of the minimal promoters of the HIS3 and lacZ genes. A pGAD-XhoC cDNA library was constructed with poly(A)+ RNA from XhoC cells expressing a high level of G10BP by using a GAD-GH vector which directs the synthesis of fusion proteins consisting of cDNA-encoded polypeptides and the transcriptional activation domain (AD) of Gal4. Transformation of the tester strain with the cDNA library would result in binding of the Gal4 AD-G10BP fusion protein to the G10 stretch and activation of the HIS3 and lacZ genes.

One-hybrid screening. Reporter plasmids pHISi-G10 and pLacZ-G10 (Fig. 1B) were constructed by inserting three head-to-tail-ligated copies of 24-bp oligonucleotide ACCAAAGGGGGGGGGGAAGTTCTC with an EcoRI recognition sequence at the 5' end into the EcoRI-SmaI site of pHISi and pLacZ located upstream of the HIS3 and CYC1 promoters. The oligonucleotide contains the rat FN promoter sequence from positions -245 to -222, including the G10 stretch which is underlined. Saccharomyces cerevisiae YM4271 (his- ura- leu-) was transformed with these two plasmids and plated on synthetic dropout (SD) medium lacking histidine, and clones that grew slowly due to residual expression of the HIS3 gene but were unable to grow in the presence of 60 mM 3-aminotriazole (3-AT) were selected. These clones were further tested for residual expression of beta -galactosidase (beta -Gal) as stated below, and a clone designated YMHL-G10, which integrated two reporter plasmids, was finally established.

Strain YMHL-G10 was transformed with the pGAD-XhoC cDNA library, and his+ leu+ transformants grown in SD medium containing 30 mM 3-AT were scored for beta -Gal activity. Each colony was streaked onto a square area printed on the nylon filter and incubated by placing the filter on SD agar medium without histidine but containing 30 mM 3-AT at 30°C for 2 days. The filter was then dipped in liquid nitrogen, and the colonies were frozen and thawed three times. The filter was overlaid onto Whatman 3 MM filters that had been soaked in Z buffer (60 mM Na2HPO4 · 7H2O, 60 mM NaH2PO4 · 7H2O, 10 mM KCl, 1 mM MgSO4 · 7H2O, 50 mM beta -mercaptoethanol, pH 7.0) containing 0.01% 5-bromo-4-chloro-3-indolyl-beta -D-galactopyranoside (X-Gal) at 30°C for 1 to 2 h. Positive clones that were blue were selected.

Screening of a lambda ZapII cDNA library. A cDNA library was constructed from XhoC cells by using lambda ZapII as a cloning vector. The mRNA (5 µg) was reverse transcribed by using lambda ZapII cloning kit (Stratagene), and the cDNA with an EcoRI site at the 5' end and an XhoI site at the 3' end was inserted into 100 ng of lambda ZapII arms. The recombinant DNA was packaged by using GigapackII Packaging Extract (Stratagene) to yield a lambda ZapII-XhoC cDNA library. The cDNA library was plated on Luria-Bertani agar plates at about 5 × 104 recombinants per 137-mm-diameter dish and incubated at 42°C for 3 to 4 h until visible plaques developed. Nitrocellulose filters soaked with 20 mM isopropyl-beta -D-thiogalactopyranoside (IPTG) were overlaid on the plates, and expression of fusion proteins consisting of beta -Gal and the polypeptide encoded by the cDNA was induced by incubation at 37°C overnight. The filters were washed in binding buffer (20 mM HEPES, 40 mM KCl, 10 mM MgCl2, 1 mM dithiothreitol [DTT], 0.1 mM ZnSO4) (31), and the proteins were denatured by incubating the filter in 50 ml of binding buffer containing 6 M guanidinium hydrochloride at 4°C for 15 min. The proteins were renatured in binding buffer containing a 1:2 serial dilution of guanidinium hydrochloride from 6 to 0.1875 M at 4°C for 5 min. The filters were then blocked in binding buffer supplemented with sterilized 5% skim milk at 4°C for 30 min and incubated in binding buffer containing 10-µg/ml poly(dI-dC) · poly(dI-dC) and a 106-cpm/ml 32P-labeled 27-bp oligonucleotide containing the G10 stretch. The filters were washed in binding buffer and autoradiographed with an intensifying screen at -80°C.

Construction of G10BP-1 expression vectors. For construction of pCMV-G10BP-1, Bluescript SK (-) G10BP-1 was excised from the lambda ZapII recombinant after coinfection of XL-1 Blue cells with the recombinant and the helper phage in accordance with the manufacturer's (Stratagene) instructions. The DNA was cleaved with ApaI and BamHI, and the 1.5-kb fragment containing the G10BP-1 cDNA was inserted into the ApaI-BamHI site of pCMV to generate pCMV-G10BP-1.

pGEX-G10BP-1 was constructed with PCR products. The G10BP-1 cDNA was synthesized with pCMV-G10BP-1 DNA as the template and the upstream sense primer from positions 1 to 18 fused to the BamHI recognition sequence and the downstream antisense primer from positions 1144 to 1163 fused to the EcoRI recognition sequence. The PCR product was cleaved with BamHI and EcoRI and inserted into the BamHI-EcoRI site of pGEX-2TK to generate pGEX-G10BP-1. pGEX-G10BP-1Delta b-Zip was constructed by joining two PCR products of the N- and C-terminal portions of the G10BP-1 cDNA. The N-terminal portion was synthesized with the upstream sense primer from positions 1 to 18 fused to the BamHI recognition sequence and the downstream antisense primer from positions 497 to 517 fused to the SmaI recognition sequence and inserted into the BamHI-SmaI site of pGEX-2TK to generate pGEX-G10BP-1N. The C-terminal portion was synthesized with the upstream sense primer from positions 738 to 756 fused to the SmaI recognition sequence and the downstream antisense primer from positions 1144 to 1163 fused to the EcoRI recognition sequence and inserted into the SmaI-EcoRI site of pGEX-G10BP-1N to generate pGEX-G10BP-1Delta b-Zip.

Transient transfection and analysis of gene expression. DNA transfection was performed by using the CaPO4 coprecipitation procedure as modified by Chen and Okayama (4). For analysis of FN promoter activity during cell cycle progression, subconfluent monolayers of 3Y1 and YG1 cells were transfected with 10 µg each of FN promoter-luciferase constructs carrying base substitutions in the G-rich sequences. The total amount of the transfected DNA was adjusted to 20 µg per dish with pRSV0 DNA. The cells were maintained in low-serum (0.5% FCS) medium for 48 h, and growth was stimulated by replacing the medium with fresh medium containing 10% FCS. Cells harvested at various intervals were assayed for luciferase activity with 120 µg of protein from the cell extract and 100 µl of the luciferin substrate (Nippongene) with an LB9501 luminometer (Berthold).

Preparation of cell extracts. Whole-cell extracts were prepared essentially by the method of Manley et al. (30). The cells were washed in phosphate-buffered saline (PBS) containing 0.5 mM MgCl2 and suspended in 4 volumes of hypotonic buffer (10 mM Tris-hydrochloride [pH 7.9] at 4°C, 1 mM EDTA, 5 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride [PMSF]). After 20 min, the cells were homogenized and 4 volumes of sucrose-glycerol solution (50 mM Tris-hydrochloride [pH 7.9] at 4°C, 10 mM MgCl2, 25% [wt/vol] sucrose, 50% [vol/vol] glycerol, 2 mM DTT, 0.5 mM PMSF) was added. After gentle stirring, 1 volume of saturated (NH4)2SO4 was added dropwise and the homogenate was centrifuged at 53,000 rpm and 4°C for 3 h in a Hitachi RP65T rotor. To the supernatant was added solid (NH4)2SO4 to a final concentration of 0.33 g/ml, and the suspension was centrifuged at 53,000 rpm in a Hitachi RP65T rotor for 30 min. The precipitate was dissolved in a minimal volume of HM buffer (20 mM HEPES [pH 7.9], 100 mM KCl, 12.5 mM MgCl2, 0.1 mM EDTA, 17% [vol/vol] glycerol, 2 mM DTT, 0.5 mM PMSF). The sample was dialyzed against two changes of 1 liter of HM buffer for at least 10 h and centrifuged at 37,000 rpm for 1 h in a Hitachi RP100AT rotor. The supernatant was quickly frozen in dry ice-ethanol and stored at -80°C. Protein concentrations were determined by a dye-binding assay (Bio-Rad Laboratories).

Western blotting. For preparation of anti-G10BP-1 antibody, GST-G10BP-1 produced in E. coli was purified through a glutathione-Sepharose column (41) and used to immunize rabbits. The rabbit antiserum obtained was first passed through a glutathione-Sepharose column preloaded with GST, and the flowthrough fraction was collected. The fraction was then loaded onto a column made by coupling GST-G10BP-1 to CNBr-Sepharose 4B (Pharmacia LKB Biotechnology), and the antiserum was eluted with 50% ethylene glycol containing 1 M KCl.

Eight to 10 µg of cell extract protein was electrophoresed on 15 and 8% polyacrylamide gels with Laemmli running buffer (25 mM Tris · glycine [pH 8.3], 0.1% sodium dodecyl sulfate [SDS]). Proteins were electrophoretically transferred to nitrocellulose membrane filters (BA85; Schleicher & Schuell) and incubated in immunoblotting diluent solution (5% skim milk, 5% FCS, 1% Tween 20 in PBS) at room temperature for 1 h to minimize nonspecific antibody binding. The filter was incubated with the primary antibody at an appropriate dilution, as indicated in the figure legends, at room temperature for 1 h and washed three times in PBS containing 1% Tween 20 for 15 min each time. The filter was then incubated with a secondary antibody at an appropriate dilution at room temperature for 1 h and washed three times in PBS containing 1% Tween 20 for 15 min. Immune complexes were detected by enhanced chemiluminescence (ECL) by treating the membrane with the ECL detection system in accordance with the manufacturer's (Amersham) protocol and exposed to X-ray film.

Preparation of 32P-labeled GST fusion proteins. Expression of GST fusion proteins in E. coli and purification on glutathione-Sepharose were performed as described by Kaelin et al. (22). The fusion protein was phosphorylated in a 60-µl reaction mixture containing HMK buffer (20 mM Tris-hydrochloride [pH 7.5], 100 mM NaCl, 12 mM MgCl2), 10 µg of the GST fusion protein, 185 kBq of [gamma -32P]ATP, 1 mM DTT, and 50 U of the catalytic subunit of cyclic AMP (cAMP)-dependent protein kinase (Sigma) at room temperature for 30 min. The reaction was terminated by addition of 40 µl of stop buffer (10 mM sodium phosphate [pH 8.0], 10 mM sodium pyrophosphate, 10 mM EDTA, 1-mg/ml bovine serum albumin), and the mixture was applied to a Sephadex G50 column. 32P-labeled GST fusion protein was eluted by centrifugation of the column at 4°C for 5 min.

In vitro protein binding assay. GST fusion proteins were electrophoresed on SDS-15% polyacrylamide gels at 150 V for 3 h and transferred to nitrocellulose filters. The proteins were denatured by incubating the filters in 50 ml of HBB buffer (20 mM HEPES-KOH [pH 7.5], 50 mM KCl, 5 mM MgCl2, 1 mM DTT, 0.1% Nonidet P-40) containing 6 M guanidinium hydrochloride at 4°C for 5 min and renatured successively in HBB buffer containing 1:2 serial dilutions of guanidinium hydrochloride of 6 to 0.1875 M at 4°C for 5 min. The filters were blocked with 5% sterilized skim milk (Difco) at 4°C for 30 min and reacted with 32P-labeled GST fusion proteins in HBB buffer supplemented with 1% skim milk at 2.5 × 105 cpm/ml at 4°C for 4 h. The filters were washed in PBS-Tween 20 buffer at 4°C for 10 min and autoradiographed with an intensifying screen at -80°C.

For GST pulldown assays, the YG1 0-h and 3Y1 16-h extracts were prepared from quiescent YG1 cells and quiescent 3Y1 cells serum stimulated for 16 h by lysing the cells in TNE buffer (10 mM Tris-hydrochloride [pH 7.8], 1 mM EDTA, 0.15 M NaCl, 1% Nonidet P-40, 10-µg/ml aprotinin) at 4°C for 30 min, and the supernatant was collected by centrifugation at 14,000 × g for 20 min at 4°C. GST-G10BP-1 and GST-G10BP-1Delta b-Zip (50 µg of each) were preincubated with 10 µl of glutathione-Sepharose 4B beads (Pharmacia) at 4°C for 1 h, and the beads were then incubated with the cell extracts (700 µg of protein each) at 4°C for 1 h (25) and collected by centrifugation. The beads were washed five times with TNE buffer, and the bound proteins were eluted from the beads by boiling in 30 µl of 2× Laemmli buffer for 5 min. The sample (10 µl) was subjected to SDS-polyacrylamide gel electrophoresis (PAGE). The presence of G10BP-1 in the eluate was analyzed by Western blotting.

    RESULTS
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Molecular cloning of a cDNA encoding the G10 stretch binding protein with a yeast one-hybrid system. The expression of the rat FN gene is regulated by competitive binding of transcription factor Sp1 and its negative regulator G10BP to three G-rich sequences present in the promoter (43). These sequences are the G10 stretch between positions -239 and -230 and two GC boxes consisting of the G10 stretch with one internal C residue insertion between positions -105 and -95 and between -54 and -44 (Fig. 1A). GCd is the most critical site, and G10BP represses the promoter activity primarily by exclusion of Sp1 binding to this site.

To correlate the function of G10BP with its structural features and control of expression, a cDNA clone encoding G10BP was isolated by using a yeast one-hybrid screen. A yeast tester strain, YMHL-G10, was established by introducing two reporter plasmids, pHISi-G10 and pLacZi-G10, into strain YM4271 (His- Ura- Leu-) (Fig. 1B). Both pHISi-G10 and pLacZi-G10 contain three multimerized 26-bp sequences from the FN promoter containing the G10 stretch upstream of the minimal promoters of the HIS3 and lacZ genes. A pGAD-XhoC cDNA library was constructed from XhoC cells, a rat 3Y1 derivative transformed by the adenovirus E1A and E1B genes, by using a yeast expression vector tagged with the transcriptional activation domain from the Gal4 transcription factor. The vector thus directs the synthesis of fusion proteins composed of cDNA-encoded polypeptides and the Gal4 transcriptional activation domain. Strain YMHL-G10 was transformed with an aliquot of the pGAD-XhoC cDNA library which yields about 2 × 106 Leu+ colonies and plated on medium lacking leucine and histidine and containing 20 mM 3-AT. The His+ colonies (116 clones) developed were streaked on nylon filters placed on medium containing tryptophan and leucine, and the colonies grown were quickly frozen in liquid nitrogen and dipped in Z buffer containing X-Gal. Three positive clones that were blue and thus expressed beta -Gal were isolated. The library plasmids were recovered from each of these clones, amplified in E. coli, and used for retransformation of the tester strain. The isolation of His+ beta -Gal+ colonies was repeated twice more, and two clones were finally selected. The clone containing a 6.6-kb cDNA insert has homology with the human elastin gene, and the other clone, containing a 1.7-kb insert, has the sequence with the basic-zipper motif described below. Elastin is a main component of elastic fibers in the extracellular matrix and was not applicable to the present study.

Isolation of the cDNA clone encoding G10BP was also performed by screening approximately 5 × 105 recombinant phages from a lambda ZapII XhoC cell cDNA library with the 32P-labeled oligonucleotide containing the G10 stretch as a probe. The plaques transferred to nitrocellulose filters were dipped in an IPTG solution, and the clones that expressed fusion proteins consisting of a cDNA-encoded polypeptide and beta -Gal were detected. One positive clone contained the 1.6-kb cDNA insert whose sequence is identical to that of the 1.7-kb cDNA clone isolated from the pGAD-XhoC cDNA library. The protein encoded by this cDNA was named G10BP-1.

Structural features of G10BP-1. The G10BP-1 cDNA insert of a lambda ZapII clone was sequenced as shown in Fig. 2. The predicted amino acid sequence derived from the open reading frame comprises 385 amino acids. The basic regions shown by the dark boxes are located in the middle and on the C-terminal side of the coding region. In the middle basic region, 9 of the 24 amino acid residues are arginine and lysine. The stretch of four basic amino acids adjacent to proline (underlined in the box) might be the nuclear localization signal. A putative zipper structure consisting of a heptad repeat of four hydrophobic amino acids is located downstream of the middle basic region. In the C-terminal side basic region, 11 of the 26 residues are basic amino acids. The C-terminal portion of G10BP-1 indicated by the open box is unusually acidic in nature, and 17 of the 49 amino acid residues are glutamic acid or aspartic acid. A search of the GenBank protein database with the predicted sequence revealed that the N-terminal half of G10BP-1 from codons 1 to 235 is identical to the amino acid sequence of resiniferatoxin-binding protein (RBP) (36). The RBP cDNA was isolated from a rat ganglion cDNA library during isolation of the cDNA clone for a channel-like RBP. However, RBP lacks channel-like characteristics, as the authors discussed, and is present ubiquitously in nonneural cells. Comparison of the RBP sequence with that of G10BP-1 revealed that the A residue in codon 230 adjacent to the first Leu codon in the zipper structure is missing in the RBP sequence. If the A residue is inserted, the open reading frame of RBP extends to codon 499. Although the N-terminal half of G10BP-1 is identical to that of RBP, the length and sequence of the C-terminal half of G10BP-1 are both considerably different from those of RBP.


View larger version (66K):
[in this window]
[in a new window]
  		FIG. 2.   Structural features of G10BP-1. The nucleotide (n.t.) and predicted amino acid (a.a.) sequences of G10BP-1 are shown. The basic regions, enriched with lysine and arginine residues, are shaded. A possible nuclear localization signal in the middle basic region is underlined. Four hydrophobic amino acids present every seven residues downstream of the middle basic region are black boxed. The acidic region enriched with glutamic acid and aspartic acids is open boxed.

Affinity of binding of G10BP-1 to three G-rich sequences. To analyze the specificity of binding of G10BP-1 to three G-rich sequences in the FN promoter, the lambda ZapII clone encoding the G10BP-1-beta -Gal fusion protein was plated and its synthesis was induced by covering the plate with a filter dipped in an IPTG solution. After denaturation and renaturation of the protein, the filter was cut into four pieces and each piece was probed with 32P-labeled oligonucleotides containing either the G10 stretch (G10), the upstream GC box (GCu), GCd, or G10 with four base substitutions (G10m), as shown in Fig. 3A. G10BP-1 bound strongly to G10 and GCd and very weakly to GCu and did not bind to G10m. The pattern of the affinity of binding to three G-rich sequences is the same as that of the G10BP protein previously purified from E1A- and E1B-transformed 3Y1 (XhoC) cells, as shown in Fig. 1A.


View larger version (24K):
[in this window]
[in a new window]
  		FIG. 3.   Affinity of G10BP-1 binding to three G-rich sequences in the FN promoter. (A) The G10BP-1 lambda ZapII clone was plated, and the beta -Gal-G10BP-1 fusion protein was induced by covering the plate with a nitrocellulose filter dipped in a 10 mM IPTG solution. The protein was denatured and renatured as described in Materials and Methods. The filter was cut into four pieces, and each piece was probed with a 32P-labeled oligonucleotide containing the G10 stretch (G10), GCu, GCd, or the G10 stretch carrying four base substitutions (G10m). Probe localization was determined by autoradiography. (B) Comparison of the sizes of G10BP and G10BP-1. Aliquots of 50 µg of protein from XhoC cell extract and from quiescent YG1 cell extract were electrophoresed, transferred to a filter, and probed with a 32P-labeled oligonucleotide containing the G10 stretch. As a control, quiescent (0-h) and serum-stimulated (24-h) 3Y1 cell extracts were similarly treated.

The molecular mass of G10BP-1 predicted from the amino acid sequence (42 kDa) is larger than that of G10BP (~30 kDa) estimated by SDS-PAGE under nonreducing conditions in the absence of DTT (43). The molecular mass of G10BP was presumably underestimated under the nonreducing conditions, since it has two Cys residues at codons 42 and 67 and may form a secondary structure. The sizes of G10BP-1 and G10BP were therefore directly compared by Southwestern blotting. The extracts were prepared from XhoC cells expressing G10BP and quiescent and serum-stimulated YG1 cells. The quiescent cells express only the exogenously introduced G10BP-1 cDNA, as shown later (see Fig. 6A and B). These extracts were electrophoresed and probed with the 32P-labeled G10 oligonucleotide. As controls, extracts prepared from quiescent (0 h) and growth-stimulated (24 h) 3Y1 cells were treated similarly (Fig. 3B). G10BP-1 showed the same mobility as G10BP. G10BP-1 was not detected in the 3Y1 0-h extract but was detected in the 3Y1 24-h extract. This suggests that G10BP-1 is identical or very closely related to G10BP.

Suppression of FN promoter activity by G10BP-1 during G1-to-S progression depends on the genotypes of the G-rich sequences. To analyze the negative role of G10BP-1 in FN promoter activity during cell cycle progression and the requirement of three G-rich sequences in the promoter for this negative regulation, subconfluent cultures of 3Y1 and YG1 cells were transfected with FN promoter-luciferase cDNA constructs carrying base substitutions in the G-rich sequences and maintained in low-serum medium for 48 h to synchronize the cells in the quiescent state. The YG1 cell line was established by introduction of pCMV-G10BP-1 into 3Y1 cells and expresses G10BP-1 constitutively. The cells were then growth stimulated by replacing the medium with fresh medium containing 10% FCS. Progression of the cell cycle was analyzed by the incorporation of [3H]thymidine into the acid-insoluble fraction in untransfected cells similarly synchronized in the quiescent state. Incorporation began to increase steeply in both 3Y1 and YG1 cells at about 12 h after stimulation, localizing the G1-to-S boundary at around this time (Fig. 4A). The level of endogenous FN mRNA, which was extremely high in quiescent 3Y1 cells (0 h), decreased steeply, reaching a minimal level of about one-fifth of the original level. The level of FN mRNA in YG1 cells was low, but the level decreased gradually, reaching the minimal level.


View larger version (27K):
[in this window]
[in a new window]
  		FIG. 4.   Repression of FN promoter activity by G10BP-1 during G1-to-S-phase progression depends on GCd. (A) Subconfluent cultures of 3Y1 and YG1 cells were maintained in low-serum (0.5% FCS) medium for 48 h, and growth was stimulated by changing the medium with fresh medium containing 10% FCS. At the times indicated, the cells were labeled with 4 µCi of [3H]thymidine per ml (814 GBq/mmol) for 1 h and the radioactivity in an aliquot of the cell lysates was counted after trichloroacetic acid precipitation. The levels of endogenous FN mRNA were analyzed by Northern blotting with 25 µg of total cellular RNAs. The EcoRI-BglII fragment (nucleotides 5947 to 6679) of pFH1 (27) was labeled with 32P and used as a probe. The autoradiogram was quantified by densitometer scanning, and relative values were plotted. (B to E) Subconfluent cultures of 3Y1 and YG1 cells in 80-mm-diameter dishes were transfected with 10 µg each of the following expression plasmids. pFGGGluc and pRSV-Sp1 or pRSV 0 (B), pFgggluc and pRSV 0 (C), pFgGgluc and pRSV0 (D), or pFggGluc and pRSV0 (E). The luc constructs contain the FN promoter sequence up to base -414 and the three G-rich sequences are represented by three letters G or g. The wt sequence is shown by a capital G, and the base-substituted sequence is shown by a small g. pRSV0 contains only the promoter. The transfected cells were maintained in low-serum (0.5% FCS) medium for 48 h, and growth was stimulated by replacing the medium with fresh medium containing 10% FCS. The cells were harvested at the times indicated, and luciferase activities were assayed with 120 µg of protein from the cell extracts. The lowest level of activity expressed by the 0-h extract from unstimulated YG1 cells was taken as 1.

The luc constructs contain the FN promoter sequence up to base -414 and base substitutions in the three G-rich sequences (Fig. 4). The wild type (wt) sequence is shown by a capital G, and the base-substituted sequence is shown by a small g. The first letter represents the G10 stretch, and the second and third letters represent GCu and GCd. When 3Y1 cells were transfected with pFGGGluc (Fig. 4B), the promoter activity decreased during cell cycle progression and reached a minimal level at 20 to 24 h. Cotransfection with pRSV-Sp1 had little effect on the decrease, although the activity was elevated significantly. In YG1 cells, the promoter activity was very low, even in the quiescent state (time zero), irrespective of cotransfection with pRSV-Sp1, indicating that exogenously expressed G10BP-1 suppressed the promoter activity almost completely. When 3Y1 and YG1 cells were transfected with pFgggluc (Fig. 4C), which carries base substitutions in all of the G-rich sequences, the promoter activities were low in both cells throughout progression of the cell cycle. No significant difference was observed between 3Y1 and YG1 cells, indicating that G10BP-1 is unable to interact with any of the base-substituted G-rich sequences. The reduction in the basal promoter activity at 0 h is likely to be caused by base substitutions, although the substitutions were introduced outside the Sp1 motif GGGCGG. The affinity of Sp1 binding to its motif is influenced greatly by the surrounding sequence, as we found in the human FN promoter (44) and as found in the alpha 2-integrin promoter (48). The promoter activity of pFgGgluc carrying the base substitutions in the G10 stretch and GCd but containing wt GCu (Fig. 4D) was similar to that of pFgggluc in both 3Y1 and YG1 cells. In contrast, expression of the promoter of pFggGluc (Fig. 4E), which carries base substitutions in the G10 stretch and GCu but contains wt GCd, was at a high level in the quiescent state and decreased during G1-to-S progression, just like that observed with pFGGGluc (Fig. 4B). The activity expressed in YG1 cells was very low throughout cell cycle progression. These results indicate that FN promoter activity and its suppression by G10BP-1 are regulated primarily through GCd and GCu plays only a minor role.

To correlate the changes in FN promoter activity during G1-to-S progression with the levels of G10BP-1 and Sp1 expression, the same cultures of 3Y1 and YG1 cells were similarly made quiescent and growth stimulated by replacing the medium. Aliquots of the cell extracts prepared at 4-h intervals were electrophoresed, and the amounts of G10BP-1 and Sp1 were analyzed by Western blotting (Fig. 5). G10BP-1 could not be detected in 0- and 4-h extracts prepared from 3Y1 cell in the quiescent state and early G1 but began to be detected in mid-G (Fig. 5A). The level reached a maximum in late G1 (12 h), and the high level was maintained until mid-S. The level decreased steeply thereafter. In YG1 cells (Fig. 5B), the exogenously introduced G10BP-1 cDNA was expressed constitutively, and a significant level of G10BP-1 was detected in quiescent and early G1 cells. The expression of endogenous G10BP-1 began to increase after mid-G1, as observed in 3Y1 cells. The level of G10BP-1 endogenously expressed was higher than that expressed exogenously. In contrast, Sp1 was expressed constitutively in both 3Y1 and YG1 cells (Fig. 5C and D), and the levels did not change significantly throughout cell cycle progression. These results show a good correlation between the induction of G10BP-1 expression and the decrease in FN promoter activity (Fig. 4), suggesting that FN promoter activity is determined by the concentration of G10BP-1 relative to that of Sp1.


View larger version (35K):
[in this window]
[in a new window]
  		FIG. 5.   Levels of G10BP-1 and Sp1 in 3Y1 and YG1 cells during G1-to-S-phase progression. Quiescent 3Y1 and YG1 cells were growth stimulated with serum as described in the legend to Fig. 4A, and cell extracts were prepared by lysing the cells in SDS-containing buffer at the times indicated. Aliquots of 10 µg (for G10BP-1 assay) or 8 µg (for Sp1 assay) of protein from the extracts were electrophoresed, and the amounts of G10BP-1 and Sp1 were analyzed by Western blotting. Anti-G10BP-1 rabbit polyclonal antibody (A and B) and anti-Sp1 rabbit polyclonal antibody (PEP2; Santa Cruz Biotechnology) (C and D) were used at dilutions of 1:10,000 and 1:15,000, respectively. Horseradish peroxidase-conjugated anti-rabbit immunoglobulin G goat antibody was used as the secondary antibody. The filters were treated with the ECL detection system and exposed to X-ray film.

Induction of G10BP-1 by adenovirus E1A. To test the induction of G10BP-1 by adenovirus E1A, a 3Y1 derivative cell line, g12, which expresses the exogenously introduced E1A 12S cDNA in response to dexamethasone (dex) (18) was made quiescent by maintenance in low-serum (0.5% FCS) medium, and cell cycle progression was induced by addition of 10-6 M dex. The levels of E1A and G10BP-1 were monitored by Western blotting (Fig. 6). Under these conditions, the rate of [3H]thymidine uptake into the acid-insoluble fraction began to increase after about 8 h. E1A expression was fully induced within 7 h, and the level was maintained throughout cell cycle progression (Fig. 6A). G10BP-1 was not expressed in the quiescent state (0 h) but was maximally induced within 7 h, concomitant with E1A expression (Fig. 6B). The level decreased gradually during cell cycle progression. This indicates that the G10BP-1 gene is a target of E1A. Similar treatment of quiescent 3Y1 cells with dex did not result in the induction of G10BP-1.


View larger version (35K):
[in this window]
[in a new window]
  		FIG. 6.   Induction of G10BP-1 by adenovirus E1A. Confluent monolayers of g12 cells were maintained in low-serum (0.5% FCS) medium for 48 h, and the expression of E1A 12S was induced by the addition of 10-6 M dex. At the times indicated, whole-cell extracts were prepared and the amounts of E1A 12S (A) and G10BP-1 (B) were quantified by Western blotting. Aliquots of 10 µg of protein were electrophoresed on SDS-15% polyacrylamide gels. Anti-E1A antibody (13s-5; Santa Cruz Biotechnology) and anti-G10BP-1 antibody were used at a dilution of 1:10,000, and horseradish peroxidase-conjugated anti-rabbit immunoglobulin G goat antibody was used as the secondary antibody at a dilution of 1:10,000.

Homodimerization of G10BP-1. The presence of the zipper motif in G10BP-1 suggested that it forms homo- and/or heterodimers. To test homodimer formation and the involvement of the basic-zipper motif in dimerization, the G10BP-1 and G10BP-1Delta b-Zip cDNAs were cloned into the pGEX-2TK vector (22), in which the nucleotide sequence corresponding to the amino acid sequence RRASV for recognition by the catalytic subunit of cAMP-dependent protein kinase A (1, 38) was placed between the GST sequence and the cDNA cloning site. The G10BP-1Delta b-Zip cDNA lacks the middle basic region and the zipper motif from codons 172 to 244. GST-G10BP-1 and GST-G10BP-1Delta b-Zip produced in E. coli were purified and subjected to SDS-PAGE. The proteins were denatured and renatured after transfer to a nitrocellulose filter and probed with 32P-labeled GST-G10BP-1 prepared by phosphorylation with cAMP-dependent protein kinase A. As shown in Fig. 7A, the probe bound to GST-G10BP-1 but not to GST-G10BP-1Delta b-Zip or to GST alone, indicating that G10BP-1 forms the homodimer through its basic-zipper region.


View larger version (21K):
[in this window]
[in a new window]
  		FIG. 7.   Homodimerization of G10BP-1. (A) GST-G10BP-1, GST-G10BP-1Delta b-Zip lacking codons 172 to 244, and GST (5 µg of protein each) were electrophoresed on an SDS-15% polyacrylamide gel and transferred to a nitrocellulose filter. The proteins were denatured, renatured, and probed with 32P-labeled GST-G10BP-1. (B) Aliquots of 10 µg of the protein in extracts prepared from quiescent 3Y1 and YG1 cells (0 h), serum-stimulated 3Y1 and YG1 cells (24 h), and XhoC cells were similarly electrophoresed and transferred to a filter. As a control, 0.1 µg of G10BP-1 prepared from GST-G10BP-1 by cleavage with thrombin was electrophoresed. The proteins were probed with 32P-labeled GST-G10BP-1. (C) Aliquots of 50 µg each of GST-G10BP-1 and GST-G10BP-1Delta b-Zip were bound to glutathione-Sepharose 4B beads and incubated with 700 µg of protein from the YG1 0-h extract (lanes 3 and 4) and the 3Y1 16-h extract (lanes 5 and 6) at 4°C for 1 h. After washing, retained proteins were eluted together with the GST fusion proteins by boiling the beads in 2× Laemmli buffer and separated by SDS-PAGE. As controls, 5 µg of GST-G10BP-1 (lane 1), 5 µg of GST-G10BP-1Delta b-Zip (lane 2), and 40 µg of the YG1 0-h extract (lane 7) were electrophoresed simultaneously. The proteins were analyzed by Western blotting with anti-G10BP-1 antibody. The arrow indicates a fast-migrating band that corresponds to G10BP-1.

To determine whether G10BP-1 also forms a dimer with the G10BP previously identified in XhoC cells (43) and a heterodimer with some cellular factors, extracts prepared from quiescent and serum-stimulated 3Y1 and YG1 cells and from XhoC cells were electrophoresed and a species of protein that binds to G10BP-1 was similarly analyzed by far-Western blotting by using 32P-labeled GST-G10BP-1. As shown in Fig. 7B, only one discrete band was detected with all of the extracts, except the 3Y1 0-h extract, which lacks G10BP-1. The bands migrated to the same position as did G10BP-1 prepared by cleavage of GST-G10BP-1 with thrombin. This suggests that G10BP-1 forms a homodimer but may not form a heterodimer under the conditions employed and that G10BP-1 is identical or closely related to G10BP.

To confirm the homodimerization of G10BP-1 in its native state, in which it had not been subjected to in vitro denaturation and renaturation, GST pulldown assays were performed (Fig. 7C). GST-G10BP-1 and GST-G10BP-1Delta b-Zip were preloaded onto glutathione-Sepharose beads and incubated with the YG1 0-h extract, which contains only exogenous G10BP-1, and the 3Y1 16-h extract, which contains endogenous G10BP-1, as shown in Fig. 5. Specifically retained proteins were eluted together with the GST fusion protein by boiling the beads in SDS loading buffer and resolved by SDS-PAGE. The proteins that bound to G10BP-1 were detected by Western blotting using anti-G10BP-1 antibody. The eluates obtained from the YG1 0-h and 3Y1 16-h extracts after incubation with preloaded GST-G10BP-1 showed few bands (lane 3 and 5). The two slow-migrating bands corresponded to GST-G10BP-1 and its cleavage product, judging from the bands in lane 1, in which purified GST-G10BP-1 was electrophoresed. The fast-migrating band shown by the arrow corresponded to G10BP-1, because it showed the same mobility as the G10BP-1 protein present in the YG1 0-h extract (lane 7). No G10BP-1 was detected in the eluates recovered after incubation with preloaded GST-G10BP-1Delta b-Zip (lanes 4 and 6). The slow-migrating band corresponded to GST-G10BP-1Delta b-Zip (lane 2). These results indicate that both exogenous G10BP-1 in the YG1 0-h extract and endogenous G10BP-1 in the 3Y1 16-h extract bound to GST-G10BP-1 in its native state through the basic-zipper region.

    DISCUSSION
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

In the present study, a cDNA encoding an Sp1 negative regulator, G10BP (43), which suppresses FN promoter activity through binding to three G-rich sequences was cloned by the yeast one-hybrid system. A tester strain, YMHL-G10, containing the HIS3 and lacZ genes fused to the G10 stretch and minimal promoters was transformed by a pGAD-XhoC cDNA library constructed in a yeast expression vector tagged with the transcriptional activation domain from the Gal4 transcription factor. One positive clone encodes a G-rich sequence binding protein, designated G10BP-1, which comprises 385 amino acids. G10BP-1 has regions rich in basic residues and a zipper structure; however, it may not belong to the basic-zipper DNA binding protein family. First, the basic region, usually located adjacent to the helically stacked hydrophobic residues, is located at a considerable distance in G10BP-1. Second, there is a proline in the first heptad and no hydrophobic residues are present at positions 3 and 4 in the first and the third heptads, which are found in true zippers.

The following features of G10BP-1 suggest that it is identical or closely related to the G10BP protein previously purified from XhoC cells, which were derived from rat 3Y1 cells transformed by the adenovirus E1A and E1B genes (43). (i) The affinity of G10BP-1 binding to the three G-rich sequences, the G10 stretch (G10), GCu, GCd, and these sequences containing base substitutions is identical to that of G10BP (Fig. 3A). (ii) The promoter activity of FN promoter-luciferase cDNA constructs is repressed by G10BP-1, and the repression is strictly dependent on GCd, as previously shown with purified G10BP (43) (Fig. 4). (iii) The expression of G10BP-1 is induced by E1A and serum factors (Fig. 5 and 6). (iv) The size of G10BP-1 analyzed by Southwestern blotting was the same as that of G10BP (Fig. 3B).

The expression of G10BP-1 is cell cycle regulated and dependent on cell growth arrest rather than cell density. The level of G10BP-1 was very low or undetectable in quiescent 3Y1 cells but increased steeply after growth stimulation by serum, reaching a maximum in late G1. Transcription factor Sp1, which competes with G10BP-1 for binding to three G-rich sequences, is expressed at a constant level throughout cell cycle progression from G1 to S phase. The elevation of the expression level of G10BP-1 may exclude the binding of Sp1 to these sites after mid-G1 and suppresses FN promoter activity. E1A mutants dl646N and dl922/947 (47), containing deletions of codons 30 to 85 and 122 to 129, respectively, lose the ability to repress FN gene expression (34). Since these regions are essential for binding of E1A to the retinoblastoma protein (pRB) and related proteins p107 and p130 (46, 47), the expression of G10BP-1 might be repressed by pRB family members and E1A may overcome this repression through complex formation with pRB family members. Negative regulation of G10BP-1 expression by pRB family members may also be released through phosphorylation of these members by cyclin-dependent kinases, cyclin D/cdk4,6, and cyclin E/cdk2 (6, 8, 9, 23, 26, 32), since in quiescent 3Y1 cells, cyclin D1 is induced in mid-G1 and cyclin E and cdk2 are induced in late G1 after growth stimulation by serum (18). The time course of the induction of these cyclins and cdks is well correlated with that of G10BP-1 accumulation.

Analysis of the dimerization of G10BP-1 by Western blotting (Fig. 7A) suggested that G10BP-1 forms homodimers through the basic-zipper region. Far Western blotting of the YG1 0- and 24-h extracts with anti-G10BP-1 antibody (Fig. 7B) revealed only one band, which migrated to the same position as did G10BP-1 prepared from GST-G10BP-1 by cleavage with thrombin. The same band was also detected in the XhoC cell extract from which G10BP was purified. Since the YG1 0-h extract contains only exogenously expressed G10BP-1 and the YG1 24-h extract contains both exogenous and endogenous G10BP-1, this result also suggests that G10BP-1 expressed by the expression vector is identical to endogenous G10BP and forms homodimers but not heterodimers. G10BP-1 was also recovered from the YG1 0-h and 3Y1 16-h extracts by GST pulldown assays (Fig. 7C), indicating that G10BP-1 is capable of homodimer formation in its native state. The pattern of complex formation between the G10 stretch and G10BP-1 varies, depending on the concentration of cell extracts. When the amount of the extracts was limited, fast-migrating complex I was formed, but slow-migrating complexes II and III began to be formed as the amounts of extract added increased (data not shown), suggesting that G10BP-1 forms homomultimers depending on its concentration. The formation of three- and four-stranded alpha -helical coiled coils has been reported with the GCN4 leucine zipper protein (12).

The isolation of cDNA clones encoding the G10 stretch binding protein was performed by both a yeast one-hybrid system and Southwestern blotting of a lambda ZapII cDNA library. The G10BP-1 cDNA was isolated by either method, but it was a sole clone encoding a nuclear protein, suggesting that the number of G10 binding proteins in the cells is limited. As the affinity of Sp1 binding to its consensus motif is strongly influenced by adjacent sequences, the affinity of G10BP-1 binding is also influenced by the position of the C residue(s) in the G stretch, since it binds strongly to GCd but poorly to GCu. GC boxes are the most ubiquitous promoter elements, but the recognition of the sequence by positive and negative factors is highly specific, and a particular GC box seems to be the target of both factors. As shown in Fig. 4, GCd is the target of both Sp1 and G10BP-1 and primarily determines FN promoter activity. We recently found that the expression level of the FN gene in NEC14 human embryonal carcinoma cells is very low but is steeply enhanced by the induction of Sp1 following induction of differentiation (44). Among four GC boxes in the human FN promoter, Sp1 binds preferentially to one of them, and the factor UnDF, which is present specifically in undifferentiated cells and competes with Sp1 for binding, also binds preferentially to the same GC box. After induction of NEC14 cell differentiation, the expression of the alpha 2(I) procollagen and alpha 2-integrin genes is also stimulated concomitant with expression of the FN gene. The expression of these genes and of the human alpha 1(1) precollagen gene (19) is also regulated by Sp1 (45, 48). It would be interesting to determine whether the expression of these genes encoding cell adhesion molecules is also negatively regulated by a common factor or a respective factor specific to the base sequence of the GC box.

    ACKNOWLEDGMENT

We thank Y. Fujii for Sp1 expression plasmid pRSV-Sp1.

    FOOTNOTES

* Corresponding author. Mailing address: Department of Biological Science and Technology, Science University of Tokyo, Noda-shi, Chiba 278, Japan. Phone: 81-471-24-1501, ext. 4401. Fax: 81-471-25-1841. E-mail: koda{at}rs.noda.sut.ac.jp.

    REFERENCES
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

1. Blanar, M. A., and W. J. Rutter. 1992. Interaction cloning: identification of a helix-loop-helix zipper protein that interacts with c-Fos. Science 256:1014-1018[Abstract/Free Full Text].
2. Buck, C. A., and A. F. Horowitz. 1987. Cell surface receptors for extracellular matrix molecule. Annu. Rev. Cell Biol. 3:179-205.
3. Burridge, K., K. Fath, T. Kelly, G. Nuckolls, and C. Turner. 1988. Focal adhesions: transmembrane junctions between the extracellular matrix and the cytoskeleton. Annu. Rev. Cell Biol. 4:487-526.
4. Chen, C., and H. Okayama. 1987. High-efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol. 7:2745-2752[Abstract/Free Full Text].
5. Dean, D. C., C. L. Bowlus, and S. Bourgeois. 1987. Cloning and analysis of the promoter region of the human fibronectin gene. Proc. Natl. Acad. Sci. USA 84:1876-1880[Abstract/Free Full Text].
6. Dowdy, S. F., P. W. Hinds, K. Louis, S. I. Reed, A. Arnold, and R. A. Weinberg. 1993. Physical interactions of the retinoblastoma protein with human cyclins. Cell 73:499-511[Medline].
7. Dufour, S., J.-L. Duband, A. R. Kornblihtt, and J. P. Thiery. 1988. The role of fibronectins in embryonic cell migrations. Trends Genet. 4:198-207[Medline].
8. Dulic, V., E. Lees, and S. I. Reed. 1992. Association of human cyclin E with a periodic G1-S phase protein kinase. Science 257:1958-1961[Abstract/Free Full Text].
9. Ewen, M. E., H. K. Sluss, C. J. Sherr, H. Matsushime, J. Kato, and D. M. Livingston. 1993. Functional interactions of the retinoblastoma protein with mammalian D-type cyclins. Cell 73:487-497[Medline].
10. Fagen, J. B., M. E. Sobel, K. M. Yamada, B. de Crombrugghe, and I. Pastan. 1981. Effect of transformation on fibronectin gene expression using cloned fibronectin cDNA. J. Biol. Chem. 256:520-525[Free Full Text].
11. Hara, E., T. Yamaguchi, H. Tahara, N. Tsuyama, H. Tsurui, T. Ide, and K. Oda. 1993. DNA-DNA subtractive cDNA cloning using oligo(dT)30-latex and PCR: identification of cellular genes which are overexpressed in senescent human diploid fibroblasts. Anal. Biochem. 214:58-64[Medline].
12. Harbury, P. B., P. S. Kim, and T. Alber. 1994. Crystal structure of an isoleucine-zipper trimer. Nature (London) 371:80-83[Medline].
13. Hill, J. A., K. A. Donald, and D. E. Griffiths. 1991. DMSO-enhanced whole cell yeast transformation. Nucleic Acids Res. 19:5791[Free Full Text].
14. Horowitz, A., K. Duggan, C. Buck, M. C. Beckerle, and K. Burridge. 1986. Interaction of plasma membrane fibronectin with talin: a transmembrane linkage. Nature (London) 320:531-533[Medline].
15. Hynes, R. O., and K. M. Yamada. 1982. Fibronectins: multifunctional modular glycoproteins. J. Cell Biol. 95:369-377[Free Full Text].
16. Hynes, R. O. 1985. Molecular biology of fibronectin. Annu. Rev. Cell Biol. 1:67-90.
17. Hynes, R. O. 1987. Integrins: a family of cell surface receptors. Cell 48:549-554[Medline].
18. Ishii, T., M. Shimizu, Y. Kanayama, S. Nakada, H. Nojima, and K. Oda. 1993. Differential activation of cyclin-dependent kinase genes by adenovirus E1A12S cDNA product. Exp. Cell Res. 208:407-414[Medline].
19. Jimenez, S. A., J. Varga, A. Olsen, L. Li, A. Diaz, J. Herhal, and J. Koch. 1994. Functional analysis of human alpha 1 (I) procollagen gene promoter. J. Biol. Chem. 269:12684-12691[Abstract/Free Full Text].
20. Jochemsen, A. G., R. Bernards, H. J. van Kranen, A. Houweling, J. L. Bos, and A. J. van der Eb. 1986. Different activities of the adenovirus types 5 and 12 E1A regions in transformation with the EJ Ha-ras oncogene. J. Virol. 59:684-691[Abstract/Free Full Text].
21. Jockush, B. M., P. Bubeck, K. Giehl, M. Kroemker, J. Moschner, M. Rothkegel, M. Rudiger, K. Schluter, G. Stanke, and J. Winkler. 1995. The molecular architecture of focal adhesions. Annu. Rev. Cell Dev. Biol. 11:379-416[Medline].
22. Kaelin, W. J., W. Krek, W. R. Sellers, J. A. DeCaprio, F. Ajchebaum, C. S. Fuchs, T. Chittenden, Y. Li, P. J. Farnham, M. A. Blanar, D. M. Livingston, and E. K. Flemington. 1992. Expression cloning of a cDNA encoding a retinoblastoma-binding protein with E2F-like properties. Cell 70:351-364[Medline].
23. Kato, J.-Y., H. Matsushime, S. W. Hiebert, M. E. Ewen, and C. J. Sherr. 1993. Direct binding of cyclin D to the retinoblastoma gene product (pRb) and pRb phosphorylation by the cyclin D-dependent kinase, CDK4. Genes Dev. 7:331-342[Free Full Text].
24. Kimura, G., A. Itagaki, and J. Summers. 1975. Rat cell line 3Y1 and its virogenic polyoma- and SV40-transformed derivatives. Int. J. Cancer 15:694-706[Medline].
25. Klinge, C. M., B. F. Silver, M. D. Driscoll, G. Sathya, R. A. Bambara, and R. Hilf. 1997. Chicken ovalbumin upstream promoter-transcription factor interacts and estrogen receptor binds to estrogen response elements and half-sites and inhibits estrogen-induced gene expression. J. Biol. Chem. 272:31465-31474[Abstract/Free Full Text].
26. Koff, A., A. Giordano, D. Desai, K. Yamashita, J. W. Harper, S. Elledge, T. Nishimoto, D. O. Morgan, B. R. Franza, and J. M. Roberts. 1992. Formation and activation of a cyclin E-cdk2 complex during the G1 phase of the human cell cycle. Science 257:1689-1694[Abstract/Free Full Text].
27. Kornblihtt, A. R., K. Umezawa, Vibe-Pedersen, and F. E. Baralle. 1985. Primary structure of human fibronectin: differential splicing may generate at least 10 polypeptides from a single gene. EMBO J. 4:1755-1759[Medline].
28. Kumazaki, T., R. S. Robetorye, S. C. Robetorye, and J. R. Smith. 1991. Fibronectin expression increases during in vitro cellular senescence: correlation with increased cell area. Exp. Cell Res. 195:13-19[Medline].
29. Lee, F., R. Mulligan, P. Berg, and G. Ringold. 1981. Glucocorticoids regulate expression of dihydrofolate reductase cDNA in mouse mammary tumour virus chimaeric plasmids. Nature (London) 294:228-232[Medline].
30. Manley, J. L., A. Fire, A. Cano, P. A. Sharp, and M. L. Gefter. 1980. DNA-dependent transcription of adenovirus genes in a soluble whole-cell extract. Proc. Natl. Acad. Sci. USA 77:3855-3859[Abstract/Free Full Text].
31. Matsuno, K., C.-C. Hui, S. Takiya, T. Suzuki, and Y. Suzuki. 1989. Transcriptional signals and protein binding sites for sericin gene transcription in vitro. J. Biol. Chem. 264:18707-18713[Abstract/Free Full Text].
32. Matsushime, H., D. E. Quelle, S. A. Shurtleff, M. Shibuya, C. J. Sherr, and J.-Y. Kato. 1994. D-type cyclin-dependent kinase activity in mammalian cells. Mol. Cell. Biol. 14:2066-2076[Abstract/Free Full Text].
33. Murano, S., R. Thweatt, R. J. S. Reis, R. A. Jones, E. J. Moerman, and S. Goldstein. 1991. Diverse gene sequences are overexpressed in Werner syndrome fibroblasts undergoing premature replicative senescence. Mol. Cell. Biol. 11:3905-3914[Abstract/Free Full Text].
34. Nakajima, T., T. Nakamura, S. Tsunoda, S. Nakada, and K. Oda. 1992. E1A-responsive elements for repression of rat fibronectin gene transcription. Mol. Cell. Biol. 12:2837-2846[Abstract/Free Full Text].
35. Nakamura, T., T. Nakajima, S. Tsunoda, S. Nakada, K. Oda, H. Tsurui, and A. Wada. 1992. Induction of E1A-responsive negative factors for transcription of the fibronectin gene in adenovirus E1-transformed rat cells. J. Virol. 66:6436-6450[Abstract/Free Full Text].
36. Ninkina, N. N., J. J. Willoughby, M. M. Beach, P. R. Coote, and J. N. Wood. 1994. Molecular cloning of a resiniferatoxin-binding protein. Mol. Brain Res. 22:39-48[Medline].
37. Okayama, H., M. Kawaichi, M. Brownstein, F. Lee, T. Yokota, and K. Arai. 1987. High-efficiency cloning of full-length cDNA: construction and screening of cDNA expression libraries for mammalian cells. Methods Enzymol. 154:3-28[Medline].
38. Rashidbaigi, A., H. Kung, and S. Pestka. 1985. Characterization of receptors for immune interferon in U937 cells with 32P-labeled human recombinant immune interferon. J. Biol. Chem. 260:8514-8519[Abstract/Free Full Text].
39. Ruoslahti, E. 1988. Fibronectin and its receptors. Annu. Rev. Biochem. 57:375-413[Medline].
40. Sieweke, M. H., H. Tekotte, J. Frampton, and T. Graf. 1996. Maf B is an interaction partner and repressor of Ets-1 that inhibits erythroid differentiation. Cell 85:49-60[Medline].
41. Smith, D. B., and K. S. Johnson. 1988. Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione-S-transferase. Gene 67:31-40[Medline].
42. Southern, P. J., and P. Berg. 1982. Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. Mol. Appl. Genet. 1:327-341[Medline].
43. Suzuki, M., C. Kuroda, E. Oda, S. Tsunoda, T. Nakamura, T. Nakajima, and K. Oda. 1995. G10BP, an E1A-inducible negative regulator of Sp1, represses transcription of the rat fibronectin gene. Mol. Cell. Biol. 15:5423-5433[Abstract].
44. Suzuki, M., E. Oda, T. Nakajima, S. Sekiya, and K. Oda. 1998. Induction of Sp1 in differentiating human embryonal carcinoma cells triggers transcription of the fibronectin gene. Mol. Cell. Biol. 18:3010-3020[Abstract/Free Full Text].
45. Tamaki, T., K. Ohnishi, C. Hartl, E. C. Leroy, and M. Trojanowska. 1995. Characterization of a GC-rich region Sp1 binding site(s) as a constitutive responsive element of the alpha 2 (I) collagen gene in human fibroblasts. J. Biol. Chem. 270:4299-4304[Abstract/Free Full Text].
46. Weinberg, R. A. 1995. The retinoblastoma protein and cell cycle control. Cell 81:323-330[Medline].
47. Whyte, P., N. M. Williamson, and E. Harlow. 1989. Cellular targets for transformation by the adenovirus E1A proteins. Cell 56:67-75[Medline].
48. Ye, J., R. H. Xu, J. Taylor-Papadimitriou, and P. M. Pitha. 1996. Sp1 binding plays a critical role in Erb-B2- and v-ras-mediated downregulation of alpha 2-integrin expression in human mammary epithelial cells. Mol. Cell. Biol. 16:6178-6189[Abstract].
49. Zerler, B., B. Moran, K. Maruyama, J. Moomaw, T. Grodzicker, and H. E. Ruley. 1986. Adenovirus E1A coding sequences that enable ras and pmt oncogenes to transform cultured primary cells. Mol. Cell. Biol. 6:887-899[Abstract/Free Full Text].

Mol Cell Biol, August 1998, p. 4772-4782, Vol. 18, No. 8
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.


This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow