THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, No. 19, Issue of July 5, pp. 14003-14010,1993 Printed in U. S. A.

Characterization of Upstream Activation Elements Essential for the Expression of Germ Cell Alkaline Phosphatase in Human Choriocarcinoma Cells*

(Received for publication, December 22,1992, and in revised form, March 19, 1993)

Naohiro Wada and Janice Yang Chout From the Human Genetics Branch, National Institute of Child Health and Human Deuelopment, National Institutes of Health, Bethesda, Marylnnd 20892

Expression of the germ cell alkaline phosphatase is a highly regulated process tied to malignant transfor- mation of the human placenta. Human chorioearci- noma cells (malignant trophoblasts) express primarily the germ cell alkaline phosphatase gene and only low or nondetectable levels of the placental alkaline phos- phatase normally found in the human placenta. Here, we show that nucleotides -156 to -1 region relative to the gene transcription start site (+1) contain cis- acting DNA elements that direct germ cell alkaline phosphatase expression in choriocarcinoma cells. Within the minimal activator region, at least three nuclear protein-binding sites, I (-63/-44), I1 (-87/ -67), and I11 (-136/-103), were identified by DNase I footprinting analysis. All three sites are GC-rich. Sites I and I1 contain a sequence known to bind the transcription factor AP-2; the AP-2 site in site I1 over- laps a consensus motif for the transcription factor Spl. Gel retardation experiments showed that similar nu- clear protein factor(s) in JEG-3 choriocarcinoma cells bind to all three sites, with highest affinity to sites I and 11. Site-directed mutagenesis that prevents binding of nuclear proteins to either site I or 11, or both sites I and 11, resulted in the ioss of factor binding and re- duced activator activity. The germ cell alkaline phos- phatase promoter that contains an intact binding site I11 but altered sites I and 11 had little activator activity, suggesting that protein-protein interaction is impor- tant for germ cell alkaline phosphatase gene activa- tion.

Genetic, biochemical, and molecular cloning studies have revealed the existence of four human alkaline phosphatases (APs)’: the placental (PL), the PLAP-like germ cell (GC), the intestinal, and the liver/bone/kidney isozymes (1-8). Each is encoded by a separate gene. The structures of GCAP and PLAP genes are very similar, both composed of 11 exons and

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in thispaper h been submitted

L12591. to the GenBankTM/EMBL Data Bank with accession number(s)

$ T o whom correspondence should be addressed Bldg. 10, Rm. 98242, NIH, Bethesda, MD 20892. Tel.: 301-496-1094; Fax: 301-402- 0234.

’ The abbreviations used are: AP, alkaline phosphatase; PL, pla- cental; GC, germ cell; CAT, chloramphenicol acetyltransferase; PCR, polymerase chain reaction; bp, base pair(s); a-MEM, a-modified minimal essential medium; DTT, dithiothreitol.

10 introns (3-6) and clustered on the long arm of chromosome 2 (9). GCAP and PLAP are synthesized as preproproteins containing both amino- and carboxyl-terminal signal peptides which are cleaved post-translationally to yield mature pro- teins of 484 amino acids (10, 11). During enzyme synthesis and processing, the phosphatidylinositol-glycan moiety is CO-

valently attached to the Asp residue 484 of nascent GCAP or PLAP which is then anchored to the plasma membrane (10- 12). Mature GCAP and PLAP proteins share 98% amino acid sequence identity. However, GCAP is more strongly inhibited by heat, EDTA, and L-leucine than PLAP (13, 14). The differential reactivity of GCAP and PLAP was shown to depend critically on a single amino acid at position 429 (15).

PLAP is expressed in the human placenta beginning late in the first trimester of pregnancy (16, 17). In addition to the PLAP mRNA, the human term placenta also expresses a low level (approximately 2%) of the GCAP message (18). GCAP is also found in trace amounts in the testis, cervix, and thymus (19, 20). Both PLAP and GCAP are oncodevelopmental pro- teins, and ectopic expression of these genes is frequently associated with human malignant tumors (13, 21-24). For example, in patients with germ cell tumors the GCAP gene is preferentially expressed (23, 24). Placental cell lines derived from human choriocarcinoma, the extraembryonic germ cell tumor of the placenta (25), express primarily the GCAP isozyme rather than the normal PLAP isozyme (6). On the other hand, these malignant trophoblasts express high levels of the placental glycoprotein hormone, human chorionic go- nadotropin, like the human placenta (26, 27). Therefore, malignant transformation of placenta appears to involve in- activation of PLAP and activation of GCAP expression. Stud- ies on the control of GCAP expression may yield insight into the mechanisms of malignant transformation.

Despite the isolation and characterization of cDNA and genomic clones encoding PLAP as well as GCAP, little is known about the regulation of these genes. Using transient expression assays, Deng et al. (28) recently showed that nu- cleotides -170 to -1 relative to the transcription start site of the GCAP gene directed the expression of this gene in a colon cancer cell line. The present report is a comprehensive analy- sis of the cis-acting regulatory sequences in the 5”flanking region of GCAP gene that are responsible for its expression in human choriocarcinoma cells. We show that the -156 to -1-bp region upstream of the transcription initiation site of the GCAP gene contains activation elements that are essential for its expression. Furthermore, there are transcription factors present in choriocarcinoma cells that bind to the GCAP promoter elements. Mutagenesis studies delineate the core DNA sequence essential for the expression of GCAP in human choriocarcinoma cells.

14003

14004 GCAP Activation Elements in Choriocarcinoma Cells

MATERIALS AND METHODS

Construction of Promoter-CAT Fusion Genes-The GCAP pro- moter-CAT fusion gene constructs were synthesized by polymerase chain reaction (PCR) using a GCAP (6) genomic clone as a template and a GeneAmp DNA amplification kit (Perkin-Elmer Cetus). The 3’ primer for the GCAP 5’ deletion mutants is nucleotides -18 to -1, and the 5’ primers are nucleotides -551 to -533, -402 to -385, -302 to -285, -201 to -184, -156 to -139, -100 to -83, and -49 to -32 (see Fig. 1). The 5’ and 3’ primers for GCAP(+1/-401)CAT are nucleotides -17 to +1 and -401 to -386, respectively. Each primer contains an additional HindIII or XbaI site at the 5’ end. After digestion with XbaI and HindIII, the amplified fragments were in- serted upstream of the bacterial CAT gene of a promoterless and enhancerless pCAT-basic plasmid (pBCAT, Promega Biotech, Mad- ison, WI). Some of the constructs were inserted in the pCAT-en- hancer plasmid, which contains the SV40 enhancer. The constructs were verified by DNA sequencing by the Sanger dideoxy chain- termination method (29).

The GCAP(-2000/-1)CAT plasmid was constructed by ligating a SalI-SstI fragment (1500 bp upstream of nucleotide -551 in the GCAP gene) into the 5’ SstI site in GCAP(-551/-1)CAT. The pSVCAT (pCAT-control, Promega), which contains both SV40 en- hancer and promoter, and pBCAT plasmids were used as positive and negative controls, respectively.

The ability of the GCAP gene fragments to drive transcription from the heterologous SV40 early promoter was evaluated by using the SV40 early promoter-CAT gene fusion vector (pCAT-promoter, Promega) which contains a BglII site upstream of the SV40 promoter. GCAP(-100/-40) or GCAP(-156/-101) fragments with additional BglII sites at each end were synthesized by PCR and were subcloned in both orientations into the pCAT-promoter plasmid in the BglII site in one or multiple copies.

Site-directed Mutagenesis by PCR-Mutagenesis was carried out as described by Higuchi (30). Mutant GCAP(-201/-1)SIAA contains 2 base substitutions (GG to AA at nucleotides -53 and -52, see Fig. l ) , whereas mutant GCAP(-20I/-l)SIIA contains a single base substitution (G to A at nucleotide -78, see Fig. 1). The two outside

The two inside primers are 5”CAAGGTGGTAACAAGGGGAGA- PCR primers are nucleotides -201 to -184 and -18 to -1 (Fig. 1).

- AGCCAGGACAC-3’ and 5”GGGAAAACTGTGTCCTGGCTTCT CCCCTTG-3’ (GCAP(-201/-1)SIAA) or 5”TCAGGTCmAG- GCTGGGCAGGGTCAAGG-3’ and 5”CCTTGTTACCACCTTGA- CCCTGCCCAGCC-3’ (GCAP(-201/-1)SIIA). The outside primers contain either an additional HindIII or XbaI linker. After digestion with XbaI and HindIII, the amplified fragments were inserted up- stream of the CAT gene in the pCAT-Basic plasmid.

Transfection and CAT Assays-JEG-3 human choriocarcinoma cells (27) were grown at 37 “C in a-modified minimal essential me- dium (a-MEM, Meditech, Inc., Herndon, VA) supplemented with 4% fetal bovine serum, streptomycin (100 pg/ml), and penicillin (100 units/ml). After reaching 90% confluence, cells in 150-cm2 flasks were transfected in suspension by incubating with 2 ml of a calcium phosphate-DNA coprecipitate containing 50 pg of plasmid DNA and 20 p~ chloroquine (31, 32). To correct for transfection efficiency, 2 pg of pSV(3gal (33) was cotransfected with the GCAP plasmid DNA (50 pg). (3-Galactosidase enzyme activities were assayed as described (34). The cell suspension was divided equally into two 75-cm2 flasks in a-MEM-4 containing 10 p~ chloroquine. After an overnight in- cubation, cultures were re-fed with a-MEM-4 medium and harvested after an additional 2 days incubation. Cell lysates were prepared by three freeze-thaw cycles and heated at 60 “C for 10 min. The CAT activity was assayed by incubating total cellular protein in a buffer containing 250 mM Tris, pH 7.8, 4 mM acetyl coenzyme A and 0.1 pCi of [14C]chloramphenicol (35). Routinely, the assay was run for 1 h with the amount of extract required to convert 0.5%-50% of the substrate to the acetylated forms. Assays outside this range were repeated using the appropriate amount of extract. The acetylated compounds were separated from chloramphenicol by thin-layer chro- matography (95% chloroform, 5% methanol, v/v) on silica gel IB2 (Gilman Sciences). Spots were quantitated on an AMBIS Radioan- alytic Imaging System (San Diego, CA).

Preparation of Nuclear Extracts-JEG-3 nuclear extracts were prepared essentially as described by Ohlsson and Edlund (36) and Dignam et al. (37). Briefly, cells (10’ to IO9) were incubated at 37 ”C for 10 min in a buffer containing 40 mM Tris-HC1, pH 7.4, 140 mM NaCl, and 1 mM EDTA and then scraped off with a rubber policeman. After centrifugation, cell pellets were suspended in 5 packed volumes

of cold buffer A (10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, and 0.5 mM dithiothreitol (DTT) and incubated on ice for 10 min. After centrifugation, the pellets were resuspended in 2 packed vol- umes of buffer A and homogenized by 10 strokes with a Kontes glass Dounce homogenizer (pestle B). The homogenate was centrifuged at 2000 X g for 10 min to separate the cytoplasmic extract. The pellet was recentrifuged at 25,000 X g for 20 min, and the nuclear pellet was resuspended in ice-cold buffer C (20 mM HEPES, pH 7.9, 1.5 mM MgClz, 0.42 M NaCl, 0.2 mM EDTA, 0.5 mM phenylmethylsul- fonyl fluoride, 0.5 mM DTT, and 25% glycerol) (3 ml for lo0 cells) and homogenized by 10 strokes with a glass Dounce homogenizer. After gently stirring for 30 min, the homogenates were centrifuged and the supernatant was dialyzed for 4 h against 50 volume of buffer D (20 mM HEPES, pH 7.9, 0.1 M KCl, 0.2 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM DTT, and 25% glycerol). The dialyzed homogenate was centrifuged at 25,000 X g for 20 min, and aliquots of the supernatant solution were frozen and stored in liquid N:! freezer.

DNase I Footprinting Assay-The GCAP(-201/-1)CAT plasmid was linearized with either HindIII or XbaI, end-labeled with Klenow DNA polymerase and a-[32P]dCTP, and then digested with either XbaI (the HindIII digested-plasmid) or HindIII (the XbaI-digested plasmid). The single end-labeled 201-bp GCAP fragment was purified on a 5% polyacrylamide gel. DNase I footprinting analysis was performed using the Pharmacia SureTrack Footprinting Kit (Phar- macia LKB Biotechnology Inc.). Briefly, 2 ng of end-labeled DNA fragment was incubated in 50 pl of binding reaction buffer for 30 min at room temperature in the presence of 10-40 pg of nuclear extract. Then MgCl, and CaC12 were added to final concentrations of 1 and 0.5 mM, respectively, and incubated for 1 min at room temperature. The reaction mixture was digested with 1.25-5 units of DNase I for 1 min at room temperature and stopped with 140 p1 of a solution containing 192 mM sodium acetate, 32 mM EDTA, 0.14% SDS, and 64 pg/ml of yeast tRNA. After phenol-chloroform extraction and ethanol precipitation, the pellet was resuspended in a sample dye buffer and analyzed on an 8% polyacrylamide-urea sequencing gel.

Gel Retardation Assay-End-labeled oligonucleotide probes (1 ng; 0.5-1 X IO5 cpm) were incubated for 20 min at room temperature in binding reaction buffer (10 mM Tris-HC1, pH 7.5,50 mM NaCl, 0.05% Nonidet P-40, 1 mM EDTA, 0.5 mM DTT, and 10% glycerol) con- taining 1 pg of poly(d1-dC) and 3 pg of nuclear extract. Following binding, the mixture was electrophoresed through a 5% nondenatur- ing polyacrylamide gel in a buffer containing 7 mM Tris-HC1, pH 7.4, 3 mM sodium acetate, and 1 mM EDTA for 2 h at 30 mA, dried, and autoradiographed. For competition experiments, competitor DNA was incubated in the mixture prior to the addition of nuclear extract. For gel supershift assays, following binding, 1 pg of anti-Spl antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was then added and the reaction mixture was incubated at 4 “C for 2 h before gel electrophoresis. Purified Spl and AP-2 were obtained from Promega and the AP-2 consensus oligonucleotide (5”TCGACTACGTAGCCT- CAGGGTGTTTGTTG-3’) was a gift from Dr. T. Sargent (National Institutes of Health).

RESULTS

The -156 to -1-bp GCAP 5‘-Flanking Region Drives Expression of the GCAP Gene in Choriocarcinoma Cells- Earlier studies showed that human malignant trophoblasts (choriocarcinoma cells) express primarily the GCAP gene (6). To delineate the DNA sequences responsible for GCAP expression, we sequenced its 5”flanking region and examined the promoter activity by transient expression in JEG-3 human choriocarcinoma cells. The GCAP 5”flanking sequence re- ported in the present study differs from the sequence reported for the Caco2 colon carcinoma cell line (28) only by the presence of one additional nucleotide A at -316 (Fig. 1). A putative TATA box is located at -30 to -24 bp upstream of the transcription start site. Four CAGCC direct repeats are present between nucleotides -150 to -130 and three CACCC motifs, a common structural and functional feature of the promoters of globin (38) and some non-globin (39) genes, are present at nucleotides -204 to -200, -177 to -173, and -128 to -124. The GCAP promoter contains two sites (GN&GG) shown to bind AP-2 (40) located at nucleotides -82 to -75

GCAP Activation Elements in Choriocarcinoma Cells 14005

GAGCTCTGGGCCTAGATTCCTGCCAGCCCCACCTGTCCAGGCCMGGCCAGATOTTGGAGC -491

FIG. 1. Nucleotide sequence of the putative promoter region of GCAP gene. The numbers indicate the dis- tance in nucleotides from the transcrip- tion initiation site (6) which is indicated by an arrow. The TATA box and ATG codon are shaded, CAGCC sequences are underlined, and the CACCC boxes are boldface and underlined. The AP-2 and Spl motifs are double underlined. The nucleotide that differs from that re- ported by Deng et al. (28) was high- lighted.

AAGGGGGTOCCAGAGCGGCAAAOCCCCCCCCATGTGCCCCTTTCCCACC~CCA~CCT

GGCATCAGGAQGCTGMCCCAGGCCCTGGCCAGACTGTGTWTTCCAGCCTCCCCTCCTCT

CGACACCAGMCAGAGCCTGGCCCCCAGCTCCCAGGAAATACAG-TGGTWA

T G A A C G A G T G A C A G G G T G T C T T G T T C C A C A C M G A C A C A G T G A G C A ~ G ~ ~ A ~

GCCCCTGGCQGCAGGATGCACACTGCACTATACCCAAAATC~C$iLTTCCCTGQQQQACA

CCTGGTC~~TMGCTGCCTTTCTCAGGACCC~CCAGCCCAGCCCAGCCA~~

C_GCCACTCCCTTCAGCCAGTGTOGCTTCAGGTCMGAGGCTGQGCGGQGTCMGGTGGTA 8pl /AP-2

ACMGGGGAGGG~CCAGGACACAOTTTTTCCCTGXW%#~%CCCAGGCAGCCTGGAGTQCAO

CTCATACTCCATACCTGGGATTTCCGCCTCGCCGCTCTCCGACTGCTTCCAGAC~

AP-2

r’

and -58 to -51. The -82 to -75 AP-2 site in the GCAP promoter overlaps with a consensus Spl motif (GGGCGG) (41) located at nucleotides -82 to -77.

To define the DNA sequences that direct GCAP expression in choriocarcinoma cells, we created GCAP promoter-CAT fusion genes containing progressive 5’ deletions and their promoter activities were examined after transfection into JEG-3 cells (Fig. 2). CAT activity, corrected for either protein alone or both protein and @-galactosidase activity, was pre- sented as activity relative to the GCAP(-2000/-1)CAT ac- tivity. The relative CAT activities were essentially the same, regardless of the correction method.

The GCAP(-2000/-1)CAT construct exhibited significant CAT activity (Fig. 2). The CAT activity was then determined after transfecting into choriocarcinoma cells the GCAP(-551/

GCAP(-201/-1)CAT, or GCAP(-156/-1)CAT plasmids. Results in Fig. 2 show that deletion of -2000 to -157-bp region in the GCAP gene had little effect on CAT expression. On the other hand, the GCAP promoter-CAT construct in the reverse orientation, GCAP(+1/-401)CAT, had little pro- moter activity (data not shown). GCAP constructs containing 5’ deletions further than -156 decreased CAT expression markedly. GCAP(-lOO/-l)CAT expressed only 50% of the CAT activity expressed by GCAP(-156/-1)CAT; GCAP(-49/-1)CAT had little promoter activity (Fig. 2). Our data indicate that DNA regions at -156 to -101 and -100 to -50 contain cis-acting elements that positively regulate GCAP gene expression in choriocarcinoma cells. An increase in CAT activity to 204% of the activity expressed by the GCAP(-2000/-1)CAT was achieved after fusing a SV40 en- hancer downstream of GCAP(-49/-1)CAT, locating the min- imal GCAP promoter a t nucleotides -49 to -1.

The -156 to -101- and -100 to -40-bp GCAP Gene Regions Function as Transcription Activators-To demonstrate that nearly all of the transcription activity of the GCAP 5’- flanking region in JEG-3 cells is contained within the -156 to -1 DNA region, we examined transcriptional enhancement mediated by -156 to -101 and -100 to -40-bp regions of the GCAP promoter. GCAP(-100/-40) was chosen because this fragment contains GC-rich sequences, including two AP-2 motifs (-82/-75 and -58/-51) and one consensus Spl bind- ing site (-82/-77). CAT activity was examined after fusing one or multiple copies of either fragment, in both sense and antisense orientations, upstream of the SV40 promoter in the pCAT-promoter plasmid (Fig. 3). Although one copy of either the GCAP(-156/-101) or GCAP(-100/-40) fragment in- creased pCAT-promoter activity only slightly, a 2-2.4-fold increase in CAT expression was observed when two copies of either GCAP(-156/-101) or GCAP(-100/-40) were fused

-1)CAT, GCAP(-402/-1)CAT, GCAP(-302/-1)CAT,

-430

-369

-309

-241

-186

-125

-64

-3

+55

B

-2180

-1 ” +GJH

I I l I l I I 1 L I Z 0 4 11.6 1l.X I I1 I 2 I 4 I h

Rcl:riwc C A T Aclivity

FIG. 2. Promoter activity of the GCAP 5‘-flanking region. Nucleotides -2OOO/-1, -551/-1, -402/-1, -302/-1, -2Ol/-1, -156/-1, -loo/-I, or -49/-1 region 5’ of the transcription initia- tion site of the GCAP gene were inserted upstream of a promoterless and enhancerless CAT gene (pBCAT). The pSVCAT plasmid which contains both SV40 enhancer and promoter and pBCAT plasmid were used as positive and negative controls, respectively. The CAT activity was determined for each construct after transient transfec- tion into JEG-3 choriocarcinoma cells. A, thin-layer chromatographic separation of acetylated compound from chloramphenicol. The amount of protein in each reaction is 100 pg except for pSVCAT, which is 5 pg. B, relative CAT activity values represent CAT/@-gal activity ratios relative to that of construct GCAP(-2000/-1)CAT. Data are the mean of four independent experiments using two batches of plasmids, and bars are the S.E.

14006 GCAP Activation Elements in Choriocarcinoma Cells

GCAP(-100/-40)

1 I I I I

1.0 2.0 3.0 4.0

Relative CAT Activity

FIG. 3. Expression of CAT gene under the control of the SV40 early promoter and different fragments from the GCAP gene 5”flanking region. Plasmid constructs with the CAT gene under the control of the SV40 early promoter (pCAT-promoter, indicated by SV40) and the GCAP -100 to -40 (shaded boxes) or GCAP -156 to -101 (open bores) DNA were tested for CAT activity by transfection into JEG-3 choriocarcinoma cells. Arrows indicate the orientation of the fragment relative to GCAP transcription. CAT activity is expressed as a ratio of the activity of pCAT-promoter. Data are the mean of three independent experiments, and bars are the S.E.

upstream of the SV40 promoter. A 3.2-3.8-fold increase in CAT activity was achieved when three copies of either GCAP fragment were fused to the SV40 promoter (Fig. 3). Both fragments function in either orientation relative to the SV40 promoter, demonstrating clearly that both GCAP(-156/ -101) and GCAP(-100/-40) function as activators in chorio- carcinoma cells.

Identification of Protein-binding Sites by DNase I Footprint- ing-The promoter activity of the 156-bp 5”flanking region in the GCAP gene in human choriocarcinoma cells raises the possibility that these cells contain transcription factors that bind to sequences within this region. DNase I footprinting analysis using JEG-3 cell nuclear extract was performed to identify specific nuclear protein-binding sites in nucleotides -201 to -1 in the GCAP gene (Fig. 4). Three discrete regions, I (-63 to -44), I1 (-89 to -67), and I11 (-136 to -103), were reproducibly protected from DNase I digestion on the coding (upper) strand by JEG-3 nuclear extract (Fig. 4A). Only region I (-65 to -44) and I1 (-87 to -67) were reproducibly protected on the noncoding (lower) strand, suggesting a weaker binding at site I11 (Fig. 4A). All three binding sites (-63/-44, -87/-67, and -136/-103) contained GC-rich se- quences. Two AP-2 motifs are contained within site I (-58/ -51) and I1 (-82/-75). The AP-2 motif in GCAP promoter region I1 overlaps with a consensus Spl-binding site at -82 to -77.

Characterization of Nuclear Protein Binding to GCAP Ac- tiuation Elements-To characterize individual protein-bind- ing sites within the GCAP activators, gel retardation assays were performed using JEG-3 cell nuclear extracts and three double-stranded synthetic oligonucleotides corresponding to binding sites I (-67/-38, oligonucleotide I), I1 (-94/-65, oligonucleotide 11), and I11 (-136/-101, oligonucleotide 111) in the GCAP promoter (Fig. 5). Specific protein-DNA inter- actions were identified by the ability to block complex for- mation by homologous oligonucleotides but not by unrelated ones. Two protein-DNA complexes were detected with both GCAP oligonucleotides I and I1 (Fig. 5 , A and B). A similar complex with greatly reduced binding was also observed with the GCAP oligonucleotide I11 (Fig. 5C).

The formation of specific complexes with labeled GCAP

A 0 10 20 40

Il l

I I

G I 0 20 40

1 2 3 4 5

upper Strand

ii ”

6 7 8 9

Lower Strand

B

5*-CAGCCCA(ICCACACCCMCCACTCCCTTCA(ICCAQTQTGGCTTCAGGT -139 I11 -92

3~-GTCGGGTCGGTGTGGGACGGTGAGGGMGTCGGTCACACCGlUGTCCA

-91 11 I ClUGAG(ICTG~CQGQGTC~GGTQGT~C~GGGGAG~QCCAQGACACAGTT-3’ GTTCTCCGACCCGCCCCAGTTCCACCATTG~CCCCTCCCCGGTCCTGTGTClU-5’

-38

I1 I

FIG. 4. DNase I footprinting analysis of choriocarcinoma nuclear protein-binding sites in the GCAP 5”flanking region. A , The GCAP -201 to -1 fragment (flanked by 5’-HindIII and 3’- XbaI sites) was asymmetrically labeled at the XbaI site for analysis of the upper strand (lanes 1-5) and at the Hind111 site for analysis of the lower strand (lanes 6-9). Labeled fragments were incubated with JEG-3 choriocarcinoma nuclear extract, digested with DNase I, and analyzed on a 8% polyacrylamide-urea gel. Lanes I and 6, G + A sequencing reaction; lanes 2 and 7, control reaction lacking nuclear extract; lane 3, 4, 5, 8, and 9, reaction containing 10 pg (lane 3 ) , 20 pg (lanes 4 and 8), 40 pg (lanes 5 and 9) of JEG-3 nuclear extract. B, nucleotides -139 to -38 GCAP 5”flanking region. The DNase I protected regions (I, II, and III) are shaded.

oligonucleotide I was efficiently blocked by an excess of un- labeled GCAP oligonucleotide I and oligonucleotide I1 (Fig. 5A). The formation of GCAP oligonucleotide I-protein com- plex was also inhibited by high amounts (50-200-fold excess) of GCAP oligonucleotide I11 (-136/-101). The close relation- ship between the GCAP DNA-protein complexes was further demonstrated by the competition of labeled GCAP oligonu- cleotide I1 with an excess of unlabeled GCAP oligonucleotide I (Fig. 5B). Again, the GCAP oligonucleotide I11 was a weaker competitor for binding than either GCAP oligonucleotide I or 11.

GCAP oligonucleotide I11 bound only weakly to protein factors in the JEG-3 nuclear extract (Fig. 5C). Moreover, the GCAP oligonucleotide 111-DNA complex could be effectively competed by GCAP oligonucleotides I and I1 as well as the homologous oligonucleotide I11 (Fig. 5C). In all cases, a non- specific oligonucleotide competitor had no effect on the bind- ing of GCAP sequences to factors in choriocarcinoma nuclear extracts (Fig. 5). Our data indicate that the transcription factors that bind to sites I, 11, and I11 in the GCAP promoter

GCAP Activation Elements in Choriocarcinoma Cells 14007

n n r 01 T I ~ I 0 iT Cowtitor: - - S 5: (I 5: cu V) cv - , z z z g z z : V ) C U 0 0 0 0 0 0

x x x x x x x x x x x x x x x x x x Extract: - + + + + + + + + + + - + + + + + + + + + +

1

1 2

Probe:

C

Competitor: - Extract: -

+ +

3 4 5 6 7 8 9 1 0 1 1 1 2 3 4 5 6 7 8 9 1 0 1 1

GCAP Oligo I GCAP Oligo II

" ~ V ) V ) 0 0 0 0

x x x x + + + + +

1 2 3 4 5 6

D

GCAP Oligo I (-67/-38) 5'-GGTAACAAGGGGAGGGGCCAGGACACAG-3'

3"TTGTTCCCCTCCCCGGTCCTGTGTCM-5'

GCAP Oligo I1 (-94/-65) 5'-GGTCAAGAGGCTGGGCGGGGTCMGGTG-3'

3"AGTTCTCCGACCCGCCCCAGTTCCACCA-5'

GCAP Oligo I11 (-136/-101) S'-CCCAGCCACACCCTGCCACTCCCTTCAGCCAGTG-3'

~*-GTCGGTGTGGGACGGTGAGGGMGTCGGTCACAC-~'

Unrelated Oligo 5 l - G e M ~ A G G G G A ~ G C @ G G A C A C A G - 3 '

3'-TTOC_TCCCCTCIT_CG@CCTGTGTCM-5'

Probe: GCAP Oligo lU FIG. 5. Binding of JEG-3 nuclear proteins to distinct sites in the GCAP gene activator. GCAP(-67/-38) oligonucleotide I ( A ) ,

GCAP(-94/-65) oligonucleotide I1 (B), and GCAP(-136/-101) oligonucleotide I11 ( C ) were end-labeled and used in gel retardation assays with JEG-3 nuclear extract. Specific complexes are indicated with arrows. The amount of competitor used is indicated as fold molar excess over the corresponding probe. D, oligonucleotide probes or competitors.

14008 GCAP Activation Elements in Choriocarcinoma Cells

are closely related. The high affinity ones are sites I and I1 and the low affinity one, site 111.

Nucleotides -82 to -75 and -58 to -51 in the GCAP promoter contain putative binding sites (GN,GGG) for AP-2 (40). The -82 to -75 AP-2 site overlaps with a consensus Spl motif (GGGCGG) (41) located a t nucleotides -82 to -77. To delineate the core sequence essential for binding to cho- riocarcinoma factors, we examined the binding abilities of mutant GCAP oligonucleotides whose GC-rich stretch have been disrupted (Fig. 6). A site I mutant, GCAP(-67/-38)AA which contains two nucleotides substitutions a t positions -53 and -52 (GG to AA), did not bind to choriocarcinoma factors (Fig. 6A). Moreover, GCAP(-67/-38)AA oligonucleotide failed to compete for complex formation between GCAP(-67/ -38) oligonucleotide I and choriocarcinoma nuclear factors, demonstrating that nucleotides -53 and -52 constitute the core sequence of the GCAP activator in site I (Fig. 6A).

A site I1 mutant oligonucleotide, GCAP(-94/-65)A which contains a single G to A substitution a t position -78, bound extremely weakly to protein factors in the JEG-3 nuclear extract (Fig. 6A). GCAP(-94/-65)A mutant also competed weakly for GCAP oligonucleotide 11-protein binding (Fig. 6A).

Exb.ct + - + + + +

* J MY * Y b u

Extract - + - + + + + + . . "WP

The G to A substitution a t nucleotide 78 disrupted the con- sensus Spl binding site (GGGCGG, nucleotides -82 to -77) in the GCAP promoter. This suggests that site I1 may be a Spl-like binding site.

Functional Analysis of Protein-binding Sites within the GCAP Actiuation Region-Gel retardation analysis demon- strated that nucleotides -53 and -52 in GCAP site I (-67/-38) and nucleotide -78 in GCAP site I1 (-94/-65) are essential for nuclear protein binding. To determine the role of protein-binding sites I and I1 in transcriptional en- hancement, we constructed three GCAP(-201/-1)CAT mu- tants, GCAP(-201/-1)CATSIAA, GCAP(-ZOl/-l)CATSIIA, and GCAP(-201/-1)CATSIAASIIA. The GCAP(-201/-1) CATSIAA mutant contains 2 base substitutions within site I (GG to AA conversion a t nucleotides -53 and -52), the GCAP(-201/-1)CATSIIA mutant contains a single base sub- stitution within site I1 (G to A conversion at nucleotide -78), and the GCAP(-ZOl/-l)CATSIAASIIA double mutant con- tains mutations a t both sites I and 11. The effect of mutation in these regions on enhancement of CAT expression directed by the GCAP promoter was then examined in JEG-3 cells (Fig. 7). Mutation in either site I (GCAP(-201/

A

GCAP(-67/-38)U S'-GGTUCMGGGGA-QCCAGGACACAG-3*

3"TTGTTCCCCTCrrCGGTCCTGTGTCM"

G C A P ( - ~ ~ / - ~ ~ ) A 5"GGTCAAGAGGCTGGG~GGGTCAAGGTG-3*

3*-AGTTCTCCGACCCmCCCAGTTCCACCA-S*

FIG. 6. A, binding of JEG-3 nuclear proteins to mutant GCAP oligonucleotides. Mutant GCAP(-67/-38)AA (lane 1 ), GCAP(-67/ -38) oligonucleotide I (lanes 2-6), mutant GCAP(-94/-65)A (lanes 7 and 8) , and GCAP(-94/-65) oligonucleotide I1 (lanes 9-14) were end-labeled and used in gel retardation assays with JEG-3 nuclear extract. Specific complexes are indicated with arrows. The amount of competitor used is indicated as fold molar excess over the correspond- ing probe. B, mutant GCAP(-67/-38)AA differs from GCAP(-67/ -38) oligonucleotide I by 2 bases and mutant GCAP(-94/-65)A differs from GCAP(-94/-65) oligonucleotide I1 by 1 base. The sub- stitutions were indicated by boldface and underline.

. c 0 cu - n 0 4

a

. r 0

E N

0 4

0

t Y U v) 4

t m, 0

B I 11 -44

GCAP AGGCTGGGCGGGGTCAAGGTGGTAACAAGGGGAGGGGCCAGGAC GCAPSIAA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GCAPSIIA GCAPSIAASIIA r**+****rA.r.crrrr..r+r.+r+rrrrrr*AAlzrr**~*

**** . * * * * A * * * * * * * * * . . * * * * * * * * * * * . * ~ * * * * * ~ ~ ~ ~

-87

FIG. 7. A, functional analysis of mutant GCAP promoter-CAT constructs. GCAP(-201/-1)CAT, GCAP(-201/-1)CATSIAA con- taining mutations in site I, GCAP(-POl/-l)CATSIIA containing a mutation in site 11, and GCAP(-201/-1)CATSIAASIIA double mu- tant were transfected into JEG-3 choriocarcinoma cells. pSVCAT and pBCAT were used as positive and negative controls, respectively. The amount of protein in each reaction is 100 pg except for pSVCAT, which is 5 pg. CAT activity is expressed as a ratio of the activity of GCAP(-201/-l)CAT. Data are the mean f S.E. of three independent experiments. B, nucleotide sequences of mutant GCAP promoter- CAT constructs. The mutated bases are shaded.

GCAP Activation Elements in Choriocarcinoma Cells 14009

-l)CATSIAA) or I1 (GCAP(-201/-1)CATSIIA) resulted in a marked reduction (62-64% inhibition) in CAT expression as compared with the GCAP(-201/-1)CAT plasmid. A near complete abrogation of activator activity was achieved with the double mutant, GCAP(-201/-1)CATSIAASIIA, which disrupts both binding sites I and 11.

The Nuclear Protein(s) That Binds to the GCAP Activators Is Closely Related to Spl-The three nuclear protein-binding sites in the GCAP promoter consist of GC-rich sequences; site I contains an AP-2 motif and site I1 contains an AP-2 motif which overlaps with a consensus Spl-binding site. To further characterize the protein factors involved in transac- tivating GCAP expression, we performed DNase I footprint- ing analysis using the GCAP(-201/-1) fragment and purified Sp l or AP-2 protein. No DNase I-protected region was ob- served with the AP-2 protein under the experimental condi- tions (data not shown). However, S p l protein binds to two regions in the GCAP promoter coinciding with sites I (-63 to -44) and I1 (-87 to -67) protected by the JEG-3 nuclear extract (Fig. 8A) .

The relationship between Spl , AP-2, and the JEG-3 nuclear factor was further examined by gel retardation assays using GCAP(-67/-38) oligonucleotide I and GCAP(-94/-65) oli- gonucleotide 11. GCAP oligonucleotides I and I1 formed com- plexes with proteins in the JEG-3 nuclear extract and the Spl protein (Fig. 8 B ) but not with the AP-2 protein (data not shown). Under the assay conditions used, the AP-2 protein strongly bound to a consensus AP-2 oligonucleotide. However, the Spl-DNA complexes migrated slower than the JEG-3

A

-89-

-67 - -63 -

-44 -

1 2 3 4

Upper

Strand

factor-DNA complexes regardless of whether GCAP oligonu- cleotide I or I1 was used.

To further demonstrate that Spl is involved in GCAP activation, we performed gel supershift assays using an anti- body to Spl. The GCAP oligonucleotide 11-Spl complex was supershifted by the anti-Spl antibody (Fig. 8C). Furthermore, in the presence of this antibody, the DNA-protein complexes formed between JEG-3 factors and GCAP oligonucleotide I or I1 was also supershifted (Fig. 8C). It appears that the anti- S p l antibody preferentially supershifted the complex of the higher molecular mass.

DISCUSSION

Humans and great apes are the only mammals that express both PLAP and the PLAP-like GCAP genes, in addition to expressing the intestinal and the liver/bone/kidney AP genes (42). PLAP is produced in high levels only in the human term placenta which expresses only low levels (about 2%) of the GCAP mRNA (18). Malignant transformation of placenta apparently extinguishes PLAP expression because human choriocarcinoma cells (malignant trophoblasts) produce low or nondetectable levels of this phosphatase (6). Instead, cho- riocarcinoma cells express primarily the GCAP gene. In the present report, we demonstrate that GCAP expression in choriocarcinoma cells is positively regulated by cis-acting elements in the 5'-flanking region of the gene. The GCAP promoter contains three activator sequences capable of bind- ing to nuclear protein factors in JEG-3 choriocarcinoma cells resulting in transactivation of GCAP expression.

B

w z o 4 %

5 6 7 8

Lower

Strand ma 4 5 6

5-a 1 2 3

GCAP

GCAP 01igon Oligo 1 2 3 4 5 8 7 8

GCAP GCAP Ollgo I Ollgo 11

FIG. 8. DNase I footprinting, gel retardation, and gel supershift analysis of Spl binding sites in the GCAP 5"flanking region. A, DNase I footprinting assay. The assay conditions are the same as described in the legend to Fig. 4 except Spl was analyzed in addition to the JEG-3 nuclear extract. The locations (relative to the GCAP transcription initiation site) of DNase I-protected sites ( I and I I ) are indicated on the left of each panel. Lanes 1 and 5, G + A sequencing reaction; lanes 2 and 6 , control without nuclear extract, lanes 3 and 7, 20 pg of JEG-3 nuclear extract; and lanes 4 and 8, 15 ng of purified Spl . E , gel retardation assay. GCAP(-6T/-38) oligonucleotide I or GCAP(-94/-65) oligonucleotide I1 was end-labeled and incubated with no nuclear extract (lanes 1 and 4 ) , 3 pg of JEG-3 nuclear extract (lanes 2 and 5 ) , or 7.5 ng of purified Sp l (lanes 3 and 6). C , gel supershift assay. GCAP(-67/-38) oligonucleotide I or GCAP(-94/-65) oligonucleotide I1 was end-labeled and incubated with no nuclear extract (lanes 1 and 4 ) , 3 pg of JEG-3 nuclear extract (lanes 2 and 51, 7.5 ng of purified Sp l (lane 7), JEG-3 nuclear extract plus 1 pg of anti-Spl antibody (lanes 3 and 6 ) , purified Sp l plus 1 pg of anti-Spl antibody (lane 8 ) .

14010 GCAP Activation Elements in Choriocarcinoma Cells

Using DNase I footprinting and gel retardation assays with JEG-3 nuclear extract, we have located in the GCAP promoter two high affinity sites, I (-63/-44) and I1 (-87/47), and one low affinity site 111 (-136/-103) for cellular transcription factors. All three sites are GC-rich. Moreover, we showed that substitution of nucleotides G, G, and G at positions -53, -52 (site I), and -78 (site 11) with A, A, and A, which disrupt the GC-rich stretch within site I and the consensus Spl motif (-82/-77) within site 11, abolished or greatly diminished binding to choriocarcinoma factors. Studies have shown that mammalian cells contain several sequence-specific DNA- binding proteins, including Spl and AP-2, that recognize GC- rich sequences (40, 41, 43-45). Spl is a well characterized transcription factor that binds to GC boxes (GGGCGG) and stimulates transcription from promoters that contain these sites (40). AP-2 is a 52-kDa enhancer-binding protein that will bind to a GC-rich recognition sequence (GN,GGG) pres- ent in the promoters of several viral and cellular genes (41). Although sites I and I1 each contain this AP-2 binding motif, purified AP-2 protein neither protects the GCAP -201 to -1 DNA against DNase I digestion nor binds to site I or TI within the GCAP promoter. However, purified Spl protein behaves like the choriocarcinoma nuclear protein factor in both DNase I footprinting, gel retardation, and gel supershift analyses, except that Spl differs from the choriocarcinoma factor in electrophoretic mobility on a nondenaturing polyacrylamide gel. A consensus Spl binding site (-82/-77) is contained within site 11. Studies have shown that Spl protein is modified post-translationally by protein glycosylation (46) and be- comes phosphorylated upon binding to promoter elements (47). Both processes alter the apparent molecular mass of Spl. Whether the choriocarcinoma nuclear factor is a Spl- like protein is currently under investigation.

Transient transfection experiments showed that interac- tions of transcription factors binding to the three activator sequences within the GCAP promoter are important for tran- scription activation. Deletion of site I11 reduced CAT activity by 50% and further elimination of sites I and I1 completely abolished CAT expression. Site-directed mutagenesis studies followed by functional analysis demonstrated that alteration of site I or 11, in a manner that abrogated choriocarcinoma nuclear protein binding, also diminished the ability of the GCAP promoter to direct CAT expression. A GCAP pro- moter-CAT mutant, GCAP(-201/-1)CATSIAASIIA, which contains mutations in both sites I and 11, but an intact binding site 111, directed only low levels of CAT expression in chorio- carcinoma cells. Site I11 appears to be necessary for GCAP promoter activity, because deletion of nucleotides -156 to -101 markedly inhibited GCAP promoter-CAT expression. Our data suggest that nuclear protein binding to site I11 in the absence of a functional site I or I1 weakly transactivates GCAP expression. It has been well established that the tran- scriptional control regions in eukaryotic genes consist of multiple binding sites for the same or for several different sequence-specific DNA-binding proteins (48, 49). Therefore, in the GCAP promoter, one or more transcription factors may work in concert to activate GCAP expression in choriocarci- noma cells.

In transient expression assays using GCAP promoter-CAT constructs, Deng et al. (28) have shown that nucleotides -170 to -1 in the GCAP 5'-flanking region could direct CAT expression in a colon carcinoma cell line. Furthermore, in these malignant colon cells, nucleotides -341 to -182 in the GCAP promoter appear to inhibit GCAP activity because deletion of this region increased CAT expression. We could not detect a region in the GCAP promoter that negatively

regulates GCAP expression in human choriocarcinoma cells. The apparent difference in GCAP expression between chorio- carcinoma and colon carcinoma cells suggests that this gene is differentially regulated in the two types of cancer cells. We are now undertaking studies to delineate the molecular mech- anisms that differentially regulate GCAP expression.

Acknowledgments-We thank Drs. M. Chamberlin, L. Shelly, and J. Ritter for critical reading of the manuscript.

REFERENCES

2. 1.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

16.

17.

18. 19.

20.

21.

22. 23.

24.

25.

26. 27.

28.

29.

30.

31. 32. 33. 34.

35.

37. 36.

38.

39.

40. 41. 42.

43. 44. 45.

46. 47.

48. 49.

Harris, H. (1982) Harvey Leet. 76,95-124 St%brand, T., Millan, J. L., and Fishman, W. H. (1982) Isozymes: Curr.

Knoll, B. J., Rothblum, K. N., and LongIey, M. (1987) Gene 60,267-276 Knoll, B. J., Rothblum, K. N., and Longley, M. (1988) J. Biol. Chem. 2 6 3 ,

Millan, J. L., and Manes, T. (1988) P m . Natl. Acad. Sci. U. S. A. 86,3024-

Watanahe, S., Watanabe, T., Li, W. B., Soong, B. W., and Chou, J. Y.

Henthorn, P. S., Raducha, M., Kadesch, T., Weiss, M. J., and Harris, H.

Weiss. M. J.. Rav. K.. Henthorn. P. S.. Lamb. B.. Kadesch. T.. and Harris.

oples 8101. Med. Res. 6,93-117

12020-12027

3028

(1989) J. Biol. Chem. 264,12611-12619

(1988) J. BrOl. Chem. 263.12011-12019

H. (1988) 2. B%l. Chem. 263,'l2002-l2010 '

Martin, D., Tucker, D. F., Gorman, P., Sheer, D., Spurr, N. K., and

Micanovie, R., Bailey, C. A., Brink, L., Gerber, L., Pan, Y.-C. E., Hulmes, Trowsdale, J. (1987) Ann. Hum. Genet. 6 1 , 145-152

J. D.. and Udenfriend. S. (1988) Proc. Natl. Acad. Sci. (I. S. A. 86.1398-

. ,

1 ~ 0 9 ' OgaiiyS., Hayashi, Y., Takami, N., and Ikehara, Y. (1988) J. Biol. Chem.

Low, M. G., and Saltiel, A. R. (1988) Science 239,268-275 Nakayama, T., Yoshida, M., and Kitamura, M. (1970) Clin. Chim. Acta 30 ,

263,10489-10494

5Afi-5AR Sakiyama, T., Robinson, J. C., and Chou, J. Y. (1978) Arch. Biochem.

Watanahe, T.! Wada, N., Kim, E. E., Wyckoff, H. W., and Chou, J. Y.

Fishman, L., Miyayama, H., Driscoll, S. G., and Fishman, W. H. (1976)

Sakiyama, T., Robinson, J. C., and Chou, J. Y. (1979) J. Eiol. Chem. 2 6 4 ,

Povinelli, C. M., and Knoll, B. J. (1991) Plncenta 1 2 , 663-668 Chang, C. H., Angellis, D., and Fishman, W. H. (1980) Cancer Res. 4 0 ,

Goldstein, D. J., Rogers, C., and Harris, H. (1982) Clin. Chim. Acta 1 2 6 ,

Fishman. W. H.. Inelis. N. I.. Stolbach. L. L.. and Krant. M. J. (1968)

-" - " Biophys. 191,782-791

(1991) J. Btol. Chem. 2 6 6 , 21174-21178

Cancer Res. 3 6 , 2268-2273

935-938

1506-1510

63-75

Cancer Res. 2 8 , 1cO-154 '

Fishman, W. H. (1987) Cancer Res. 4 8 , 1-35 Wahren, B., Holmgren, P. A,, and Stigbrand, T. (1979) fnt. J. Cancer 2 4 ,

749-7.53 Lange, P. H., Millan, J. L., Stigbrand, T., Vessella, R. L., Ruoslahti, E.,

Novak, E. R., and Woodruff, J. D. (1979) Gynecologic and Obstetric Pathol-

Pattillo, R. A., and Gey, G . 0. (1968) Cancer Res. 2 8 , 1231-1236 Kohler, P. O., and Bridson, W. E. (1971) J. Clin. Endocrinol. Metab. 3 2 ,

Deng, G., Liu, G., Hu, L., Gum, J. R., Jr., and Kim, Y. S. (1992) Cancer Res. 62,3378-3383

Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. NatL Acad. Sci. U. S. A. 74,5463-5467

Higuchi, R. (1990) in PCR Protocols: A Guide to Methods and Applications (Innis, M. A., Gelfand, D. H., Sninsky, J. J., and White, T. J., eds) pp. 177-183, Academic Press, Inc., San Diego, CA

. " ."

and Fishman, W. H. (1982) Cancer Res. 42,3244.-3247

ogy, pp. 476-485, W. B. Saunders Company, Phdadelphia

683-687

Chu, G., and Sharp, P. A. (1981) Gene (Amst.) 13 , 197-202 Luthman, H., and Magnusson, G. (1983) Nucleic Acids Res. 11,1295-1308

Samhrook: J.: Fritsch, E.'F.,'and Maniatis, T. (1989) Molecular Cloning: A Herbomel P. Bourachot B. and Yaniv M. (1984) Cell 39,653-662

Loboratory Manuul, 2nd Ed., 17.34-17.35, Cold Spring Harbor Labora-

Fordis, C. M., and Howard, B. H. (1986) Methods Enzymol. 161 , 382-397 tory, Cold Spring Harbor, NY

Ohlsson, H., and Edlund, T. (1986) Cell 46, 35-44 Dimam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983) Nucleic Acids Res.

11 , 1475-1489 Mantovani, R., Malgaretti, N., Nicolis, S., Giglioni, B., Comi, P., Cappellini,

Acids Res. 16,4299-4313 N., Bertero, M. T., Caligaris-Cappio, F., and Ottolenghl, S. (1988) Nuclecc

Schule, R., Muller, M., Otsuka-Murakami, H., and Renkawitz, R. (1988)

Williams, T., and Tjian, R. (1991) Genes & Deu. 6,670-682 Nature 332,87-90

Courey, A,, and Tjian, R. (1988) Cell 66,887-898 Doell ast G. J. (1982) in Human Alkahne Phosphatases (Sti brand, T., and

Kageyama, R., and Pastan, I. (1989) Cell 69,815-825 Mermod, N., Williams, T. J., and Tjian, R. (1988) Nature 332,557-561 Gumucio, D. L., Rood, K. L., Blancbard-Mcquate, K. L., Gray, T. A.,

Jackson, S. P., and Tjian, R. (1988) Cell 66,125-133 Jackson, S. P., MacDonald, J. J. , Lees-Miller, S., and Tjian, R. (1990) Cell

Gidoni, D. Dynan, W. S., and Tjian, R. (1984) Nature 312,409-413 Maire, P., 'Wuarin, J., and Schihler, U. (1989) Science 2 4 4 , 343-346

Fisimin, W. H. eds) pp. 25-46, Alan R. Liss, Inc., New $ark

Saulino, A,, and Collins, F. S. (1991) Blood 78, 1853-1863

63,155-165