required for strong transcriptional stimulation by distant enhancers (P. Wen, N. Crawford, and J. Locker,. Nucleic Acids Res. 21:1911-1918, 1993). In this study ...
MOLECULAR AND CELLULAR BIOLOGY, OCt. 1994, p. 6616-6626
Vol. 14, No. 10
0270-7306/94/$04.00+0 Copyright C) 1994, American Society for Microbiology
A Novel Hepatocytic Transcription Factor That Binds the a-Fetoprotein Promoter-Linked Coupling Element PING WEN
AND
JOSEPH LOCKER*
Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 Received 8 February 1994/Returned for modification 14 June 1994/Accepted 18 July 1994
We recently characterized a promoter-linked coupling element (PCE) in the rat a-fetoprotein (AFP) gene required for strong transcriptional stimulation by distant enhancers (P. Wen, N. Crawford, and J. Locker, Nucleic Acids Res. 21:1911-1918, 1993). In this study, oligonucleotide gel retardation and competition experiments defined the PCE as a 12-bp binding site, TGTCCTTGAACA, an imperfect inverted repeat from -166 to -155 near the AFP promoter. A factor that bound this site (PCF) was abundant in HepG2 nuclear extracts and detectable in extracts from several other AFP-producing hepatocarcinoma cell lines and fetal liver. Hepatocytic cell lines that did not express AFP, nonhepatocytic cell lines, adult liver, and fetal brain did not show the factor. Experiments excluded the possibility that PCF activity was due to binding of glucocorticoid receptor or an APi-like factor that bound overlapping sites. Competition experiments with several mutant oligonucleotides determined that the optimum PCF binding site was TGTCCTTGAAC(A/T). Mutations decreased binding or totally abolished binding activity. In expression plasmids, PCE mutations strongly cross-linking to a PCE probe identified peptide bands near 34 kDa. PCF was reduced gene expression. purified by heparin-Sepharose chromatography followed by affinity binding to oligomerized PCE DNA. The product resolved as a complex of three peptides (PCFol, PCFt2, and PCF3, 32 to 34 kDa) on sodium dodecyl sulfate-acrylamide gels. The peptide sizes and gel patterns are unlike those of any of the well-described hepatic transcription factors, and the binding site has not been previously reported. PCF thus appears to be a novel transcription factor. moved closer to the promoter, and deletion of this region had little effect on transcriptional activation by the promoter and near upstream region in the absence of the distant enhancers. The lack of independent activation by this element, as well as its dispensability when the enhancers were moved closer to the promoter, led us to hypothesize that the primary function was to couple the distant enhancers to the promoter; we named it the promoter-linked coupling element (PCE). Several factorbinding sites are possible in the PCE region (8), including a glucocorticoid response element (GRE) (- 175 to - 161). Glucocorticoid receptor (GR) binding at this site repressed AFP gene expression in some cell types (18). However, footprint analysis revealed a distinct protected site from -153 to -166 bp. That the sequence of this site is 100% conserved among rat, mouse, and human genomes (8) indicates the importance of its function. This study was designed to further define the PCE region and the factors that bind it in hepatocytic cells. This study indicated a novel promoter-coupling factor (PCF) with a distinctive binding site. PCF expression was found only in hepatocytic cells that express AFP.
a-Fetoprotein (AFP) is a serum protein produced at very high levels by fetal liver but repressed about 1,000-fold to a barely detectable level in adult liver (28, 32, 36). As a result, the AFP gene has provided an important model system of high-level tissue-specific and developmental stage-specific gene expression in mammals. AFP gene expression is controlled mainly at the level of transcription. The 6-kbp 5'-flanking region of the rat and mouse AFP genes contains three discrete upstream enhancers at -6.0, -4.2, -2.3 kbp (16, 41, 43) as well as an unusually complex promoter region (42). Both the enhancers and the promoter are required for high-level hepatocyte-specific gene expression, but developmental silencing of the gene is mediated only by elements near the promoter, since the enhancers do not silence when combined with other promoters (9, 41, 43). Each enhancer is independently active, and full gene activity is the sum of contributions from the individual enhancers simultaneously simulating the promoter (43). The mechanism of silencing has not been worked out but presumably is closely related to regulation at functional elements within the near upstream region associated with the promoter. Recently, we demonstrated that an element near the rat AFP gene promoter was critical for the stimulation of the promoter by the distant AFP enhancers (42). The promoter deleted to 178 was fully active, but further deletion to -155 resulted in a 70% loss of total gene expression. Though residual gene activity indicated that the enhancers still stimulated the promoter, the dramatic reduction demonstrated the importance of an element near -155. The element was not required for full gene expression when the enhancers were
MATERIALS AND METHODS
-
Cell lines. Hepatic cell lines were maintained in Williams E medium plus 5% fetal calf serum (FCS) and 2 mM glutamine, except cell line McARH8994, which was maintained in 15% FCS. HepG2, Hep3B, and HuH7 cells are human hepatoma cell lines (5, 25, 29). McARH7777 (7777), McARH8994 (8994), Reuber H4IIEC3 (H4C3), EOCST1269 (EOC), and HTCSR (HTC) are rat hepatoma cell lines (32). HeLa (human cervical carcinoma), W138 (human fetal lung epithelium), and COS-7 (monkey kidney fibroblasts) were cultured in Eagle's minimum essential medium with 10% FCS. THP1 (human
* Corresponding author. Mailing address: University of Pittsburgh School of Medicine, Department of Pathology, Scaife Hall, Room A-725, Pittsburgh, PA 15261. Phone: (412) 648-8253. Fax: (412) 648-1916.
6616
AFP PROMOTER-COUPLING FACTOR
VOL. 14, 1994
6617
TABLE 1. Oligonucleotide competitors Oligonucleotidea
Factor
HNF3 HNF4
HNF5 GRE AP1
NFKB HNF1
C/EBP LFA1
LTFE HBLF
Reference(s) 12
tcgaAGTTGACTAAGTCAATAATCAGAATCAG TCAACTGATTCAGTTATTAGTCTTAGTCagct tcgaGCGCTGGGCAAAGGTCACCTGC CGCGACCCGTTTCCAGTGGACGagct tcgaCTAGAACAAACAAGTCCTGCGT GATCTTGTTTGTTCAGGACGCAagct tcgaTGCTGTACAGGATGTTCTAGCTAC ACGACATGTCCTACAAGATCGATGagct tcgaAGGGGCCATGTGACTCATTACACCAG TCCCCGGTACACTGAGTAATGTGGTCagct tcgaTCGAGGGCTGGGGATTCCCCATCTC AGCTCCCGACCCCTAAGGGGTAGAGagct tcgaTGTGGTTAATGATCTACAGTTA ACACCAATTACTAGATGTCAATagct
33
17 22 44
34 7
12,13
tcgaTTTTCCATCTTACTCAACATCCTCC AAAAGGTAGAATGAGTTGTAGGAGGagct tcgaCAGCCAGTGGACTTAGCCCCTGTTTG GTCGGTCACCTGAATCGGGGACAAACagct tcgaCCGAACGTGTTTGCCTTGGCCAGTTTTCCATGTACATGC GGCTTGCACAAACGGAACCGGTCAAAAGGTACATGTACGagct
31
26
39
tcgaAGTAAACAGTA TCATTTGTCATagct
ALBD
38
gTATGATTTTGTAATGGGGTA ATACTAAAACATTACCCAATgaatt
SPi
tcgaGGGGCGGGG
30
CF1
CCCCGCCCCagct tcgaGCGAGAAGAGAAAATGGTCG CGCTCTTCTCTTTTACCAGCagct
24
Serum response factor
tcgaGATGTCCATATTAGGACATC
37
CTACAGGTATAATCCTGTAGagct
a Double-stranded oligonucleotides were constructed to represent known strong binding sites for individual transcription factors (uppercase) with additional bases for labeling or cloning (lowercase).
monocytes) and A3M (human B lymphocytes) were cultured in RPMI 1640 medium with 10% FCS (40, 45). Nuclear extract preparation. Crude nuclear extracts were prepared as described previously (10, 15); some additional modifications are described in Results. Cultured cells or tissues (rat adult liver, fetal liver, and fetal brain) were disrupted by incubation in a hypotonic buffer (10 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid [HEPES; pH 7.9], 1.5 mM MgCl2, 0.5 mM spermidine, 1 mM dithiothreitol [DTT], 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 ,ug each of leupeptin, aprotinin, and antipain per ml) for 10 min at 4°C followed by Dounce homogenization. Tissue homogenates were filtered through cheesecloth. Released nuclei were pelleted by centrifugation at 800 x g for 5 min, washed, and extracted in 5 volumes of hypertonic extraction buffer (30 mM HEPES [pH 7.9], 25% glycerol, 450 mM NaCl, 0.3 mM EDTA, 6 mM DTT, 12 mM MgCl2, 1 mM PMSF, 1 ,ug each of leupeptin, aprotinin, and antipain per ml) for 45 min. Following centrifugation at 40,000 x g for 30 min, the supernatant was recovered, diluted with 2 volumes of extraction buffer without NaCl or dialyzed against extraction buffer containing 150 mM NaCl, aliquoted, and stored at -80°C. Protein concentration was determined by the Bradford method (4). Gel shift assays. Double-stranded oligonucleotides used as either probes or competitors (Table 1) were chemically synthesized and purified by passage through a 1-ml Sephadex G-50 spin column. Labeling was performed by incubating the double-stranded oligonucleotides with 2 U of Klenow DNA polymerase I and 20 ,uCi of [at-32P]dCTP and [at-32P]dATP (each at 650 Ci/mmol) in 50 nM dGTP-50 nM dTTP-10 mM Tris (pH 7.5)-5 mM MgCl2-7.5 mM DTT at 37°C for 60 min.
Following phenol extraction, the unincorporated nucleotides were removed by using a Sephadex G-50 minicolumn. Specific activities were about 5 x 107 cpm/,ug. The binding reaction mixtures (10 ,ul) contained 0.2 to 2 ng of double-stranded oligonucleotide probe, 0.75 to 5 ,ug of poly(dI-dC) poly(dIdC), and 1 to 10 ,ug protein in 25 mM HEPES (pH 7.9)-100 mM NaCl-9 mM MgCl2-0.25 mM EDTA-18% glycerol-5 mM DTT-0.75 mM PMSF-0.75 pug each of leupeptin, antipain, and aprotinin per ml. In some reaction mixtures, unlabeled competitor oligonucleotides were added at 100-fold molar excess. After a 20-min incubation at room temperature, the mixture was electrophoresed through a 6% polyacrylamide gel in 0.5 x Tris-borate-EDTA running buffer for 1.5 h. The dried gel was -
autoradiographed. Transfection and CAT assay. Plasmids were purified by banding twice in CsCl gradients. Fifteen micrograms of plasmid was added to approximately 4 x 106 HepG2 cells seeded on 75-cm2 plates, conditions determined to be nonsaturating. Comparisons were carried out on identical plates prepared at the same time. Cells were treated with 15% glycerol 4 h after addition of the DNA-Ca3(PO4)2 precipitate and further cultured in fresh medium for 2 days. Plasmids pAFP6000, which contained the entire AFP gene transcription control region, and pPAFP900, which contained only the AFP gene promoter region, were used as controls. Chloramphenicol acetyltransferase (CAT) activity was analyzed as described previously (43). Plasmid construction. Two intermediate plasmids were constructed for replacement of the PCE region with oligonucleotides. Plasmid pAFP6000 was constructed by ligating a 7.9-kb SacII-EcoRV (-6.1 to +1.8 kbp) fragment from pAFP7300 to
6618
WEN AND LOCKER
SacII-EcoRV linker sites of pBluescript II SK+. pPCAT was constructed by inserting a 2.2-kb NdeI-HindIII (-2408 to -2391 and -179 bp to +2.0 kbp) fragment derived from an AFP gene plasmid which was deleted from -2391 to -178 (42) into the EcoRV-HindIII site of pBluescript II SK+. Doublestranded oligonucleotides with a PstI site at the 5' end and an AflIl site at the 3' end were synthesized and substituted for the 40-bp PstI-AflII fragment of pPCAT. Subsequently, modified promoter regions were substituted into pAFP6000 by replacing its 1.1-kb PstI-BspEI (-890 to +252 bp) fragment with the equivalent 0.4-kb fragments from the oligonucleotide-modified pPCAT plasmids. In these PCE-substituted plasmids, the most proximal enhancer was placed at -1.6 kb, compared with its location at -2.3 kbp in the intact gene region of pAFP6000. All PCE modifications were confirmed by sequencing. Purification of PCF. Column buffers contained 25 mM HEPES (pH 7.9), 1 mM EDTA, 1 mM DTT, 10% glycerol, 0.5 mM PMSF, 0.05% Nonidet P-40, and variable NaCl concentrations.
Twenty micrograms of HepG2 nuclear extract protein was diluted to a NaCl concentration of 50 mM and passed through a 1-ml heparin-Sepharose column (Pharmacia, Piscataway, N.J.) at 10 ml/h. The column was then washed with 20 ml of 50 mM NaCl buffer and eluted with 1 ml each of 100 mM steps. The PCF-rich fractions identified by gel shift assay were pooled, diluted to 100 mM NaCl, and concentrated to 1 ml in a Centricon-30 microfiltration cell (Amicon, Beverly, Mass.) before DNA affinity column purification. A DNA-Sepharose column was prepared by coupling 12.5 ,ug of concatenated S2 oligonucleotide (see below) to 0.25 ml of swollen cyanogen bromide-activated Sepharose 4B (Pharmacia) according to the manufacturer's protocol, but with an overnight coupling reaction (6). Heparin-Sepharose column-purified PCF was mixed with 500 ,ug of poly(dI-dC) * poly(dI-dC) at 4°C for 30 min and subsequently with 0.25 ml of DNA-Sepharose resin overnight. The mixture was loaded into a 5-ml column tube, and the passed liquid portion was passed through the same column twice at 5 ml/h. The column was washed with 2.5 ml of 300 mM buffer and eluted with 0.5-ml steps from 400 mM. The PCFrich fractions were identified by gel shift assays. The purification was monitored by running fractions through a 10% noncontinuous reducing sodium dodecyl sulfate (SDS)-polyacrylamide gel stained with Coomassie blue. UV cross-linking. The procedure was carried out as described previously (11), with slight modification. A probe was prepared by annealing a 12-bp primer (5'-CTGAAGTGGT CT) to the bottom strand of the P oligonucleotide and filling in 22 bases as described above in the presence of 37.5 ,uM each dATP and 5-bromo-2'-dUTP, 6.7 pRM each [a-32P]dCTP and [a-32P]dGTP, and 1 U of Klenow DNA polymerase I at 16°C for 3 h. The specific activity was 1.4 x 108 dpm/,g. Binding was carried out by incubating protein fractions with 4 ng of DNA probe and 10 ,ug of poly(dI-dC) * poly(dI-dC) in 20 ,ul at room temperature for 20 min. The mixture was irradiated under a 302-nm UV source lamp at a distance of 1 cm. Subsequently, the DNA was digested in the presence of 10 mM CaCl2 for 30 min at 37°C with 10 ,ug of DNase I and 3 U of micrococcal nuclease. The products were resolved by SDS-polyacrylamide gel electrophoresis (PAGE). RESULTS PCE binding activity. In a recent study (42), we localized a specific promoter-linked coupling element (PCE) required for the stimulation of the AFP promoter by the three distant
MOL. CELL. BIOL.
u
.0
'a
u
u
U
.0
u
a so
B
A
.1I
FIG. 1. Specific binding to the PCE. (A) Titration of binding. Increasing amounts of HepG2 nuclear extract were incubated with 0.9 ng (55,000 cpm) of double-stranded oligonucleotide P in the presence of 0.75 pug of poly(dI-dC) * poly(dI-dC). Lane a, 6 ,ug of bovine serum albumin; lanes b to e, 1, 3, 6, and 9 ,ug of nuclear extract protein, respectively. (B) Effects of competitors. All lanes show 2.5 ,ug of HepG2 nuclear extract and 0.2 ng of PCE probe. Lanes a to d, 1, 2, 3, and 4 jig of poly(dI-dC) * poly(dI-dC), respectively; lane e, conditions identical to those for lane d except for the addition of a 100-fold molar excess of unlabeled oligonucleotide P.
enhancers. The activity was found in a promoter deleted to -178 but not to -155. In addition, a footprint was identified from -166 to -153 bp. To further characterize PCE binding activity, including both its binding factor and the binding site, nuclear extract from HepG2 cells was used to gel shift oligonucleotide P, which contained the AFP promoter region from -179 to -150 bp. Gel retardation with HepG2 nuclear extract showed two major bands (Fig. 1A). The upper band increased exponentially, while the lower band increased linearly, with increasing nuclear extract protein. This result suggests that the top band represents cooperative binding while the lower band represents simple stoichiometric binding. To determine specificity, increasing amounts of poly(dIdC) * poly(dI-dC) were added to the binding reaction mixtures as a nonspecific competitor. Alternatively, unlabeled oligonucleotide P was added as a specific competitor (Fig. 1B). Comparison of lanes d and e shows that only the lower band was competed for by a 100-fold molar excess of the unlabeled oligonucleotide P. In contrast, the lower band was retained even when a 20,000-fold molar excess of poly(dI-dC) * poly(dIdC) was present in the binding reaction, while the top band was significantly diminished (lanes a to d). For further experiments, 5 ,ug of poly(dI-dC) poly(dI-dC) per reaction was used to reduce the upper band. To identify the factor that binds oligonucleotide P, a computer search of known transcription factor binding sites (27) localized five potential binding sites, GRE, NFKB, HNF3, HNF4, and HNF5, with significant similarity to the -179 to -150 region (Fig. 2A). In addition, an AP1 site was suggested by the studies of Bacus et al. (1). The GRE has been previously characterized by Guertin et al. (18). DNA oligonucleotides (Table 1) representing each of the six potential binding sites were used as competitors, but none showed appreciable competition for binding activity (Fig. 2B). Additional competitors (Fig. 2C) included HNF1, C/EBP, LTFA1, LTFE, LBLF, and ALBD, binding sites of major -
VOL. 14, 1994
AFP PROMOTER-COUPLING FACTOR
6619
A -179 CTGM
TGGTTTTGTGAACATAAGh NW\A\N'
-150
JM' twss--\X\
"
I
ll
.
..
.
C
B Y
Competitor
10
=
6.0
=
L
-
-
A.
:~ .j
" = w 0. 1
~
i ~=
~L
P.
r,UEr
