Transcriptional regulation of human placental ... - Semantic Scholar

Molecular Human Reproduction Vol.7, No.9 pp. 887–894, 2001

Transcriptional regulation of human placental leucine aminopeptidase/oxytocinase gene T.Ito1, S.Nomura1,3, M.Okada1, Y.Katsumata1, A.Iwase1, F.Kikkawa1, M.Tsujimoto2 and S.Mizutani1 1Department

of Obstetrics and Gynecology, Nagoya University School of Medicine, Nagoya, 466-8550, Japan and 2Laboratory of Cellular Biochemistry, Institute of Physical and Chemical Research (RIKEN), Saitama, 351-0148, Japan 3To

whom correspondence should be addressed. E-mail: [email protected]

Human placental leucine aminopeptidase (P-LAP) plays a major role in the clearance of oxytocin, which is a key hormone in regulating labour pain. To explore the transcriptional regulation of P-LAP gene expression in placenta, we performed systematic studies using human choriocarcinoma cells, BeWo and JEG-3, as a model of placental trophoblastic cells. Transient transfection and luciferase assays using various 5⬘-deleted P-LAP-luciferase constructs showed that the region from –297 to ⍣49 of the transcription start site was responsible for promoter activity in these cells. Footprinting analysis with nuclear extracts from both cell lines demonstrated at least four sites for nucleoprotein interactions in this region (FP1 to FP4). Site-directed deletion of FP1–4 in luciferase assays indicated the significance of the FP3 region (–214 to –183) for high promoter activity in the cells. Electrophoretic mobility shift assays to identify the proteins interacting with DNA at FP3 revealed three retarded bands, one of which was generated by activator protein-2 (AP-2) binding. Our findings suggest that AP-2 may be one of the important factors regulating P-LAP gene expression in human placenta. Key words: aminopeptidase/AP-2/oxytocinase/placenta/promoter

Introduction Levels of bioactive peptides depend on synthesis and degradation. Oxytocin is a key hormone in labour pain during pregnancy, and the enzyme responsible for hydrolysing oxytocin is termed oxytocinase. Previous studies demonstrated that oxytocinase is cystine aminopeptidase (EC 3.4.11.3), which is identical with human placental leucine aminopeptidase (P-LAP) (Tsujimoto et al., 1992). Maternal serum P-LAP activity is increased with gestational age just before the onset of labour (Mizutani et al., 1976), suggesting that P-LAP may be involved in suppressing labour pain by hydrolysing oxytocin. Moreover, P-LAP plays an important role in regulating the functional levels of several other hormones, such as vasopressin and angiotensin III (Tsujimoto et al., 1992; Mizutani and Tomoda, 1996) which are involved in controlling feto-placental circulation by vasoconstriction. Thus P-LAP has been believed to be critical for maintaining the physiological condition of gestation. Previously, we have isolated the P-LAP cDNA clone (Rogi et al., 1996) and revealed that this enzyme is a homologue of rat insulin-regulated membrane aminopeptidase (IRAP) (Keller et al., 1995) which is present in glucose transporter isotype GLUT4 vesicles of rat adipocytes (Kandror et al., 1994; Mastick et al., 1994). Since P-LAP is co-translocated from © European Society of Human Reproduction and Embryology

the cytosol to the cell membrane with GLUT4 by insulin stimulation in adipocytes and skeletal muscle cells, P-LAP may also be involved in glucose homeostasis by insulininduced trafficking of GLUT4 vesicles. Recent studies concerning P-LAP gene expression elicit two interesting points. One is the wide distribution of P-LAP in tissues besides placenta as demonstrated by Northern blot analysis (Rogi et al., 1996; Laustsen et al., 1997) and immunohistochemistry (Nagasaka et al., 1997). According to Northern blot analysis, P-LAP mRNA is expressed in placenta, brain, heart and skeletal muscle, while little or no P-LAP mRNA can be detected in lung, liver and kidney. Immunohistochemical analysis both in the adult and fetus has revealed a similar tissue distribution to that found by Northern blotting. Although these results indicate that P-LAP tissue distribution is regulated at transcriptional levels in a tissue-specific manner, the detailed mechanism of P-LAP gene regulation has not been clarified. It should be noted that the widespread distribution of P-LAP does not contradict the significance of serum P-LAP activities during pregnancy, because the soluble form P-LAP is thought to be present only in the serum of pregnant women (Nakanishi et al., 2000; Yamahara et al., 2000). There is also an exponential increase in P-LAP mRNA expression levels in human placenta throughout gestation (Yamahara et al., 2000). This is compat887

T.Ito et al.

ible with immunohistochemical analysis of P-LAP in human placenta showing that P-LAP staining is restricted to syncytiotrophoblast cells throughout gestation (Nagasaka et al., 1997; Yamahara et al., 2000). It is likely that P-LAP expression in placenta is associated with cell differentiation, as has been demonstrated in rat adipocytes (Ross et al., 1998). Gene regulation through cell differentiation is generally controlled by the interplay between transcription factors and the gene promoter/enhancer region. Investigation of the transcriptional regulation of the P-LAP gene, which has remained unclear, may provide clues to understanding the mechanism of trophoblast differentiation. To obtain insights into the molecular mechanisms underlying the tissue-specific and gestation-induced regulation of human P-LAP gene expression, we first isolated genomic clones containing the 5⬘-upstream region of the human P-LAP gene (Horio et al., 1999). The P-LAP promoter contains GC-rich regions and the consensus sequences for the binding of several transcription factors, such as activator protein-2 (AP-2) and selective promoter factor 1 (Sp1); however, the interplay of these proteins on the P-LAP promoter gene has not been determined. We have also found that the human P-LAP gene is assigned to the chromosomal region 5q14.2–q15. Recently, the P-LAP gene was reported to span ~75 kb containing 18 exons and 17 introns (Rasmussen et al., 2000). In order to begin a systematic study of the human P-LAP gene regulation in placenta, we performed transient transfection assays of the P-LAP gene promoter linked to the luciferase gene, DNase I footprinting analysis, and an electrophoretic mobility shift assay (EMSA) using BeWo and JEG-3 choriocarcinoma cells. These choriocarcinoma cells have retained several properties of the placenta (Ringler and Strauss, 1990) and express P-LAP mRNA (Horio et al., 1999), and are therefore considered to be a suitable model to investigate the molecular mechanism of P-LAP gene regulation in the placenta. We identified a functional region for high promoter activity located between –214 bp and –183 bp from the transcription initiation site. At least three nuclear factors bind to this region, one of which was identified as AP-2.

Materials and methods Materials The following items were purchased: Roswell Park Memorial Institute (RPMI) 1640 medium and fetal calf serum (FCS) from Sigma (St Louis, MO, USA); LipofectAMINE PLUSTM Reagent and customsynthesized oligonucleotides from Life Technologies Inc. (Gaithersburg, MD, USA); the luciferase reporter vector pGL3-Enhancer, the control reporter vector pRL-TK, Dual-Luciferase reporter assay kit and RQ1 DNase from Promega Corp. (Madison, WI, USA); restriction enzymes and Klenow DNA polymerase from New England Biolabs Inc. (Beverly, MA, USA); Poly (dI-dC), [α-32P]dCTP, [α-32P]dTTP, ProbeQuant G-50TM columns from Amersham Pharmacia Biotech Inc. (Piscataway, NJ, USA); the BCA Protein assay kit and bovine serum albumin (BSA) from Pierce (Rockford, IL, USA); polyclonal rabbit anti-NF-1 antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

888

Table I. Sequences of the primers used in this study Primer Sense –752 –396 –297 –265 –242 –216 –172 –144 –94 –28 Antisense ⫹49

GCGGTACCCAATGCGCATACCGAGAGGATA ATGGTACCTAGAGGCCGCAAGGCAGCCGCC ATGGTACCTCCTGAGGCACTCTGATGCGGG ATGGTACCCTGCCGAAAACACCTACAGCGC ATGGTACCTCTCCCTCGGTCCTTCGAGCAGTGA ATGGTACCTCCCTCCGCGGCCAACCT ATGGTACCTTCAGATAGCGACCTTCGCCCG ATGGTACCTGCTCTGCCTTCAGACTTCGGC GCGGTACCCGTCTAACCAAGCCATCCGAAT ATGGTACCAGCGGGGCACTGACTTTACACT ATGGTACCTCCCAGGAGCGGACATTCCAGA

The KpnI restriction site is underlined.

Cell cultures BeWo (ATCC CCL-98) and JEG-3 (ATCC HTB-36) human choriocarcinoma cells were maintained in monolayer cultures in RPMI 1640 supplemented with 10% FCS. Preparation of luciferase reporter constructs A human P-LAP genomic clone containing the 5⬘-flanking region sequence was obtained and sequenced (Horio et al., 1999). The P-LAP promoter-luciferase constructs were prepared by cloning polymerase chain reaction (PCR)-derived fragments of the P-LAP 5⬘-flanking region into the vector pGL3-Enhancer. PCR was performed with 10 different sense oligonucleotides and one antisense oligonucleotide (Table I). A KpnI restriction site was added to the 5⬘-portion of all the primers. PCR fragments were digested with KpnI and subcloned into a similarly digested pGL3-Enhancer vector. Clones –752/⫹49, –396/⫹49, –297/⫹49, –265/⫹49, –242/⫹49, –216/⫹49, –172/⫹49, –144/⫹49, –94/⫹49 and –28/⫹49 (the transcription initiation site was numbered as ⫹1) were all confirmed by restriction enzyme and sequencing analyses. Transient transfections of deletion constructs and luciferase assay BeWo and JEG-3 cells were plated (0.8⫻106 cells/6-well plates) 24 h before transfection. Firefly luciferase reporter constructs (3 µg) and pRL-TK (0.3 µg) as an internal control to standardize transfection efficiency were transiently co-transfected into cultured cells at 50– 60% confluence using LipofectAMINE PLUSTM Reagent according to the manufacturer’s recommendations. Three hours after transfection, media were replaced by fresh normal growth media. Cells were harvested 36 h later, lysed and assayed for Firefly and Renilla luciferase activities using a dual-luciferase reporter assay system. Preparation of nuclear extracts Crude nuclear extracts were prepared from exponentially growing BeWo cells as described previously (Andrews and Faller, 1991). The concentration of the protein in the nuclear extracts was determined by the BCA protein assay. The nuclear extracts, which contained 2–6 mg/ml protein, were stored in aliquots at –70°C until use. DNase I footprinting analysis Probe preparation and DNase I footprinting analysis were performed as previously described (Galas and Schmitz, 1978; Nomura et al., 1997) with slight modification. Briefly, the luciferase vector containing the P-LAP –297/⫹49 fragment was digested with Bsu36I and the 3⬘-ends were labelled by Klenow fill-in reactions incorporating [α-32P]dCTP and [α-32P]dTTP. A second digestion by SacI generated

Transcriptional regulation of the human P-LAP gene

Table II. Sequences of the oligonucleotides used in this study wt3 mNF-1-wt3 mAP-2-wt3 Consensus NF-1 AP-2 MZF-1 Mutant mNF-1 mAP-2

MZF-1 5⬘-GATCCCTCCGCGCCAACCTCCCCAGGGAAACCAG-3⬘ NF1 AP-2 5⬘-GATCCCTCCGCGTTTAACCTCCCCAGGGAAACCAG-3⬘ 5⬘-GATCCCTCCGCGCCAACCTCCCCATTGAAACCAG-3⬘ 5⬘-GTATTTTGGATTGAAGCCAATATGATAATGA-3⬘ 5⬘-GAACTGACCGCCCGCGGCCCGTG-3⬘ 5⬘-GATCCGGCTGGTGAGGGGGAAT-3⬘ 5⬘-GTATTTTAAATTGAATAAAATATGATAATGA-3⬘ 5⬘-GAACTGACCGTTTGCGGCCCGTG-3⬘

Putative regulatory sites are underlined. Bold letters indicate the bases mutated.

the P-LAP insert labelled solely on the coding strand, which was then purified by agarose gel elecrophoresis. An aliquot of this probe (2–3⫻104 c.p.m.) was incubated with nuclear proteins or 10 µg BSA in the buffer composed of 20 mmol/l Tris–HCl (pH 7.9), 2 mmol/l MgCl2, 50 mmol/l NaCl, 1 mmol/l EDTA, 10% glycerol, 0.1% NP-40, 1 mmol/l dithiothreitol, followed by RQ1 DNase (0.25 unit) digestion. The final samples were run on an 8% sequencing gel with Maxam and Gilbert G tracks of the same DNA in an adjacent lane. The gels were exposed to X-ray film with intensifying screens at –70°C. Electrophoretic mobility shift assay Complementary single-stranded oligonucleotides designed with a 5⬘-overhang were annealed, labelled by end-filling using Klenow fragment and [α-32P]dCTP, and purified by ProbeQuant G-50TM columns. 100 fmol of labelled probe (3⫻104 c.p.m./reaction) was incubated with nuclear extracts from BeWo cells in the binding buffer used for footprinting analysis. In the competition experiments, a 150-fold molar excess of the unlabelled competing DNA fragments was added before the addition of labelled probe. The oligonucleotides used as probes or competitors are indicated in Table II (with mutation sites underlined). For supershift experiments, 10 µg of rabbit polyclonal NF-1 antibodies which cross-react NF-1 isoforms were preincubated with the nuclear extracts for 20 min at room temperature. The protein–DNA complexes were separated on 5% non-denaturing polyacrylamide gels which were then dried and autoradiographed at –70°C. Statistical analysis The transfection assays were performed three times in triplicate. All values (in the case of Figures 2 and 4) were pooled for calculation of mean ⫾ SD for paired t-test statistics using Statview 4.5 software (Abacus Concepts, Berkeley, CA, USA).

Results Analysis of the P-LAP 5⬘-flanking region for promoter activity We have previously reported the sequences of P-LAP 5⬘flanking region (Acc. No. AF145455) (Horio et al., 1999) and repeated analyses of the sequence revised the three nucleotides (Figure 1, shown in underlined bold type, ‘A’ instead of ‘G’ at –674; ‘A’ removed from ‘TAG’ at –529; ‘A’ instead of ‘G’ at –344). No typical mammalian TATA boxes are present but consensus sequences for several putative regulatory elements, Sp1, AP-2 and NF-1 (nuclear factor-1) are present. To define the DNA sequences of the 5⬘-flanking region responsible for

promoter activity of the human P-LAP gene, we prepared chimera luciferase reporter constructs containing serially deleted fragments of the 5⬘-flanking sequence of P-LAP and transiently introduced these constructs into BeWo and JEG-3 cells for measuring promoter activity. Each plasmid construct and its corresponding relative luciferase activity, which was corrected for variations in transfection efficiency by reference to Renilla luciferase activity, is shown in Figure 2. In both BeWo and JEG-3 cells, the elimination of sequences from position –752 to –396 or –297 had no significant effects on promoter activity; however, further deletion up to position –172 decreased promoter activity to 50% of the initial level. Subsequent deletions to position –94 reduced the activity to background levels and the shortest fragment, –28/⫹49, behaved similarly to fragment –94/⫹49. Since these results suggested that the region from –297 to –94 may contain positive control elements that are essential for efficient transcription of the human P-LAP gene, we focused on this region. DNase I footprinting analysis of the promoter region identifies four protected regions To locate potential regulatory elements of the human P-LAP gene promoter, DNase I footprinting analysis was performed. As shown in Figure 3, the addition of increasing amounts of crude nuclear extracts from BeWo cells resulted in at least four distinct regions protected from DNase I: FP1, –253 to –241; FP2, –233 to –219; FP3, –214 to –183; FP4, –170 to –150. A similar protection pattern was obtained with JEG-3 nuclear proteins (data not shown). DNase hypersensitive sites were clearly detected at positions –241, –233, –219, –214, –183, –170 and –150 (Figure 3, shown in asterisks). Functional analysis of protein-binding regions for P-LAP promoter activity To determine the functional significance of the protein-binding regions detected by footprinting analysis, we constructed a series of serially deleted mutants extending from –297 to –94 and tested their ability to promote transcription in BeWo and JEG-3 cells (Figure 4). Deletion of Sp1-binding sequence (in construct –265/⫹49), FP1 (in construct –242/⫹49) and FP2 (in construct –216/⫹49) had little effect on the relative promoter activity, while abolition of FP3 (in construct –172/ ⫹49) resulted in 62% suppression of luciferase activity in BeWo cells. In JEG-3 cells, similar results were obtained. These results indicate that the promoter region from –216 to ⫹49 is sufficient to express efficient luciferase activity in these cells and suggest that functional elements for high level promoter activity reside within the boundaries of FP3. Identification of proteins interacting with DNA at FP3 by EMSA To study protein binding to FP3, we performed EMSA experiments using BeWo nuclear extracts. The labelled doublestranded oligomer which encompasses the FP3 footprint, wt3, (Table II) formed three major complexes and one minor complex (Figure 5A, lane 2 and 3). The major complexes C1– C3 are specific, since the formations were completely abolished by 150-fold molar excess of unlabelled wt3, whereas the fastest 889

T.Ito et al.

Figure 1. Nucleotide sequence of the gene around the human P-LAP transcription initiation site. Exon sequences are shown in upper-case letters and 5⬘-flanking region and intron sequences are shown in lower case letters. The arrow marks the major transcriptional initiation site (nucleotide numbered ⫹1 in the sequence). Revised sites are in underlined bold face. Consensus sequences for Sp1, NF-1 and AP-2 are underlined.

Figure 2. Luciferase activity in choriocarcinoma cells transfected by P-LAP-luciferase constructs. Serial deleted fragments of the 5⬘-flanking region of P-LAP were linked to the luciferase reporter gene (LUC). Luciferase expression was measured 36 h after transfection of these constructs into BeWo and JEG-3 cells. Transfection efficiency was normalized by Renilla luciferase activities. The results are given as a relative percentage of the activity of construct –752/⫹49 in each cell type and represent the mean ⫾ SD of at least three independent experiments.

complex was hardly affected and therefore represented a nonspecific interaction (Figure 5A, lane 4). Database analysis for transcriptional factors revealed that the FP3 region contains consensus binding sequences for NF-1, AP-2 and MZF-1 (myeloid zinc finger-1), which prompted us to test the binding ability of these factors to FP3. Synthetic oligonucleotides used in EMSA are shown in Table II. MZF-1 oligonucleotide was inefficient as a competitor for any of the retarded complexes (Figure 5A, lane 10). The NF-1 oligonucleotide appeared to 890

compete weakly with C1 and C3 complexes, and more so with C2 (Figure 5A, lane 5), while mutated NF-1 oligonucleotide did not compete with any of the complexes (Figure 5A, lane 7). Therefore we tested the effect of anti-NF-1 antibodies on the mobility of the complexes (Figure 5A, lane 6), and revealed no marked shift of any complex. The AP-2 oligonucleotide competed weakly with C1 and C3 complexes and strongly with C2 (Figure 5A, lane 8), while the mutated AP-2 oligonucleotide was much less competitive (Figure 5A, lane 9).


Simultaneous addition of oligonucleotide NF-1 and oligomer AP-2 completely out-competed only the C2 complex. In order to clarify the above findings, we mutated AP-2 and NF-1 sites in the wt3 oligonucleotide and tested the ability of these mutated sequences to interact with BeWo nuclear extracts. The labelled wt3 oligonucleotide with point mutations in the AP-2 binding site (mAP-2-wt3) only formed C1 and C3 complexes (Figure 5B, lane 6), which were not competed by AP-2 and NF-1 oligonucleotides (Figure 5B, lanes 8 and 9). On the other hand, binding of nuclear proteins to the labelled wt3 oligonucleotide mutated in the NF-1 binding site (mNF1-wt3) generated the same three complexes as those generated by wt3 (Figure 5B, lane 2). The AP-2 oligonucleotide efficiently out-competed the C2 complex (Figure 5B, lane 5), although addition of the competing NF-1 oligonucleotide did not modify any of the complexes (Figure 5B, lane 4). These results indicated that wt3 was able to produce an AP-2 complex, but not an NF-1 complex, with BeWo extracts.

Discussion The placenta is thought to be a main source of serum P-LAP during pregnancy (Mizutani et al., 1976; Yamahara et al., 2000) and play important roles in controlling labour pain through inactivating oxytocin (Mitchell and Wong, 1995; Mizutani and Tomoda, 1996). However, the basal transcriptional regulation of the P-LAP gene in any of the tissues, including placenta, has not been investigated. In the current study, we identified the sequence responsible for high promoter activity of the P-LAP gene and demonstrated the possible involvement of AP-2 transcription factor. Our previous isolated clone containing 5⬘-flanking region of the human P-LAP gene (Horio et al., 1999) lacks obvious TATA boxes but possesses the initiator element (YY1; a consensus sequence; PyPyA⫹1NT/APyPy, where Py is a pyrimidine) and GC-rich region as seen in most TATA-less promoters (Smale, 1997). Although the human P-LAP promoter has highly conserved consensus binding sequences for Sp1, elimination of these sites did not significantly affect promoter activity in this study. A similar result was reported in the other gene promoters (Haun et al., 1993, Nishikawa et al., 2000). Sp1 is a minor protein in normal term human placenta (Hu et al., 1996), and this may account for the lower correlation between P-LAP promoter activity and Sp1 binding in choriocarcinoma cells. As another explanation, it is reported that Sp1 can bind to the GC-rich sequence within a few hundred base pairs upstream of the transcription start sites and activate transcription of certain genes, especially housekeeping genes

Figure 3. DNase I footprinting analysis of the –297/⫹49 promoter region of human P-LAP gene. A DNA fragment spanning –297/ ⫹49 was labelled with 32P on the coding strand. The amounts of BeWo nuclear extract proteins used are indicated at the tops of lanes. Free DNA (lane 1) contains no protein and shows the natural sensitivity to DNase I. Positions of the four footprints are designated as FP1, FP2, FP3 and FP4 respectively. The numbers on the right side of the autoradiogram indicate the boundaries of the footprints. G (lane 9) indicates Maxam and Gilbert chemical sequence ladder. Asterisks identify DNase I-hypersensitive sites.

891

T.Ito et al.

Figure 4. Analysis of essential regions for the transcription of human P-LAP gene. Transfection efficiency was normalized by Renilla luciferase activities. The results are given as a relative percentage of the activity of construct –297/⫹49 in each cell type and represent the mean ⫾ SD of at least three independent experiments.

Figure 5. Electrophoretic mobility shift assay (EMSA) with BeWo nuclear extracts. (A) EMSA using 32P-labelled double-stranded oligonucleotide wt3 spanning –214 to –183. The amounts of nuclear extracts used were 0, 5 and 10 µg, corresponding to lane 1, lanes 2 and 3–11 respectively. Arrows indicate the specific DNA–protein complexes C1–C3. NS indicates a non-specific complex, and F shows the position of free oligonucleotide. Competitors (150-fold molar excess) or anti-NF-1 antibody are indicated above the lanes; –, no competitors. (B) EMSA using the probes mutated in their NF-1- and AP-2-binding sites. 32P-labelled mNF-1-wt3 and mAP-2-wt3 were incubated with 10 µg of nuclear extracts from BeWo in the presence or absence of 150-fold molar excess of unlabelled competitors.

(Bohm et al., 1995). Taking account of the positions of Sp1 sites in the P-LAP promoter region, the effects of Sp1 may be less potent on P-LAP promoter activity. The 5⬘-deletion analysis of the P-LAP promoter region in BeWo and JEG-3 cells suggested that the sequence between 892

–297 and –94 was important for driving high promoter activity in choriocarcinoma cells. P-LAP promoter analysis has also been performed with HEK293 cells (Rasmussen et al., 2000), demonstrating a functional region for high promoter activity upstream from our sites. Although the reasons for this differ-


ence are not clear, it is interesting that the sequence responsible for the high promoter activity in HEK 293 cells contains two Sp1 binding sites. The cell types used in the studies may contribute to this discrepancy, because HEK 293 cells are derived from the kidney and presumably possess different amounts of AP-2 and Sp1 proteins compared with choriocarcinoma cells. In fact, AP-2 is known to be abundant in placenta but in poor supply in kidney (Shi et al., 1997). Further, it is noteworthy that P-LAP gene expression in kidney is relatively low according to Northern blot analysis (Rogi et al., 1996). Footprinting analyses on the region from –297 to –94 indicated that the sequence can be subdivided into at least four discrete protein binding regions surrounding the DNase I-hypersensitive sites, leading us to examine which protected region is essential for the high promoter activity. We therefore prepared a new series of P-LAP-luciferase constructs according to the results of the footprinting analyses and found that the FP3 region was important for the high level luciferase expression in choriocarcinoma cells. A computer search for potential binding sites of transcription factors in FP3 region showed the presence of a putative NF-1 half-site, as well as AP-2 and MZF-1 binding sites. In EMSA, binding of BeWo nuclear proteins to FP3 generated three retarded complexes C1, C2 and C3. Several oligomers were used as competitors in EMSA to test the ability of AP-2, MZF-1 and NF-1 factors to interact with FP3. The three pieces of evidence from EMSA indicated that the second retarded band, C2, was generated by AP-2 binding to FP3: (i) C2 complex was competed out by the consensus sequence for AP-2; (ii) mutation in the AP-2 binding site in the competitor failed to compete with the C2 complex; (iii) the labelled FP3 oligomer with point mutations in the AP-2 binding site formed only C1 and C3 complexes. AP-2, originally described as the trophoblast-specific element-binding protein, has been identified as important for the placental expression of genes including those for human chorionic gonadotrophin (HCG) α and β subunit (Johnson et al., 1997), human growth hormone (HGH) (Courtois et al., 1990), human placental lactone (PL) (Richardson et al., 2000) as well as other genes (Yamada et al., 1995; Shi et al., 1997). It has been reported that AP-2 gene expression is induced during trophoblast differentiation (Johnson et al., 1997), and that differentiation-dependent transcription of the CG-α gene in villous trophoblasts is mainly governed by increasing expression of AP-2 (Knofler et al., 2000). AP-2 has also been investigated for its role in mediating cAMP regulation of many genes including HCG (Johnson and Jameson 1999; LiCalsi et al., 2000). Although further experiments (such as investigating the effects of the mutation in AP-2 binding sites on P-LAP promoter activity) are required, the enhancement of human P-LAP gene expression during placental development, as demonstrated by Northern blot and immunohistochemical analyses may be regulated by AP-2. Interestingly, AP-2 binding sites are also present on the promoters of other genes that may link to the functions of P-LAP, i.e. the genes for oestrogen receptor (McPherson and Weigel, 1999), oxytocin receptor (Inoue et al., 1994) and vasopressin receptor (Rabadan-Diehl et al., 2000). Recently, AP-2 has also been reported to have

implications for oestrogenic regulation (Perissi et al., 2000). Thus, AP-2 may also play an important role in controlling the activity of bioactive peptides such as oxytocin and vasopressin through the regulation of their receptor levels. MZF-1 is preferentially expressed in haematopoietic cell lines, while recently it has been reported that MZF-1 may regulate LIMK2 gene expression in choriocarcinoma cell lines (Nomoto et al., 1999). Our competitive gel shift assays using a consensus binding sequence for MZF-1 (Kida et al, 1999) indicated little possibility of it binding to FP3. AP-2 protein may have a higher affinity with FP3 and interfere with MZF-1 binding at overlapping sites. Alternatively, a few nucleotides of MZF-1 binding site on the P-LAP gene are different from those on the LIMK2 gene, so MZF-1 may not bind to the P-LAP gene effectively. The binding of AP-2 and NF-1 in a mutual way, which was reported on the HGH promoter (Courtois et al., 1990), was also not indicated on the P-LAP promoter, because the NF-1 consensus oligonucleotide did not behave as a competitor to any of the complexes. One possible explanation for our findings may relate to the different NF-1 isoforms. At least four proteins constitute a family of NF-1 transcription factors (Gronostajski, 2000) and the quantity of the different forms of NF-1 varies with the growth condition of the cells. To our knowledge, a detailed analysis concerning NF-1 in BeWo cells has not been performed. Another possibility is that the affinity of NF-1 to FP3 region may be rather weak, because the FP3 sequence contains just a half-site sequence for NF-1 binding. NF-1 binds strongly to the palindromic sequence, but weakly to a halfsite sequence. Until now we have been unable to identify the two other factors which form C1 and C3. AP-2 has been reported as an interactive partner of YY1 (Wu and Lee, 2000) and so in the human P-LAP promoter, YY1 and AP-2 might also be involved in regulatory network. We are currently performing a study to identify other proteins interacting with FP3. In conclusion, our study is the first detailed analysis of the human P-LAP gene promoter and indicates that possible interactions between AP-2 and the FP3 region should modulate P-LAP gene expression. Elucidation of the underlying mechanisms for P-LAP gene regulation will pave the way to understanding the functional roles of P-LAP in vivo in labour pain. Further investigations of extracellular stimuli that could act as specific inducers or depressors of P-LAP gene expression may also facilitate the eventual development of therapeutic strategies for controlling oxytocin levels based on the modulation of P-LAP activities.

Acknowledgements This work was supported by Grants-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan, and from the Ministry of Posts and Telecommunications of Japan for the specific medical research (collaboration with Nagoya Teishin Hospital).

References Andrews, N.C. and Faller, D.V. (1991) A rapid micropreparation technique for extraction of DNA-binding proteins from limiting numbers of mammalian cells. Nucleic Acids Res., 19, 2499.

893

T.Ito et al. Bohm, S.K., Gum Jr, J.R., Ericksen, R.H. et al. (1995) Human dipeptidyl peptidase IV gene promoter: tissue-specific regulation from a TATA-less GC-rich sequence characteristic of a housekeeping gene promoter. Biochem. J., 311, 835–843. Courtois, S.J., Lafontaine, D.A., Lemaigre, F.P. et al. (1990) Nuclear factor-I and activator protein-2 bind in a mutually exclusive way to overlapping promoter sequences and trans-activate the human growth hormone gene. Nucleic Acids Res., 18, 57–64. Galas, D.J. and Schmitz, A. (1978) DNase footprinting: a simple method for the detection of protein–DNA binding specificity. Nucleic Acids Res., 5, 3157–3170. Gronostajski, R.M. (2000) Roles of the NFI/CTF gene family in transcription and development. Gene, 249, 31–45. Haun, R.S., Moss, J. and Vaughan, M. (1993) Characterization of the human ADP-ribosylation factor 3 promoter. Transcriptional regulation of a TATAless promoter. J. Biol. Chem., 268, 8793–8800. Horio, J., Nomura, S., Okada, M. et al. (1999) Structural organization of the 5⬘-end and chromosomal assignment of human placental leucine aminopeptidase/insulin-regulated membrane aminopeptidase gene. Biochem. Biophys. Res. Commun., 262, 269–274. Hu, Y.L., Lei, Z.M. and Rao, C.V. (1996) cis-acting elements and trans-acting proteins in the transcription of chorionic gonadotropin/luteinizing hormone receptor gene in human choriocarcinoma cells and placenta. Endocrinology, 137, 3897–3905. Inoue, T., Kimura,T., Azuma, C. et al. (1994) Structural organization of the human oxytocin receptor gene. J. Biol. Chem., 269, 32451–32456. Johnson, W. and Jameson, J.L. (1999) AP-2 (activating protein 2) and Sp1 (selective promoter factor 1) regulatory elements play distinct roles in the control of basal activity and cyclic adenosine 3⬘,5⬘-monophosphate responsiveness of the human chorionic gonadotropin-beta promoter. Mol. Endocrinol., 13, 1963–1975. Johnson, W., Albanese, C., Handwerger, S. et al. (1997) Regulation of the human chorionic gonadotropin alpha- and beta-subunit promoters by AP2. J. Biol. Chem., 272, 15405–15412. Kandror, K.V., Yu, L. and Pilch, P.F. (1994) The major protein of GLUT4containing vesicles, gp160, has aminopeptidase activity. J. Biol. Chem., 269, 30777–30780. Keller, S.R., Scott, H.M., Mastick, C.C. et al. (1995) Cloning and characterization of a novel insulin-regulated membrane aminopeptidase from Glut4 vesicles. J. Biol. Chem., 270, 23612–23618. Kida, M., Souri, M., Yamamoto, M. et al. (1999) Nucleotide transcriptional regulation of cell type-specific expression of the TATA-less A subunit gene for human coagulation factor XIII. J. Biol. Chem., 274, 6138–6147. Knofler, M., Saleh, L., Bauer, S. et al. (2000) Promoter elements and transcription factors involved in differentiation-dependent human chorionic gonadotrophin-alpha messenger ribonucleic acid expression of term villous trophoblasts. Endocrinology, 141, 3737–3748. Laustsen, P.G., Rasmussen, T.E., Petersen, K. et al. (1997) The complete amino acid sequence of human placental oxytocinase. Biochim. Biophys. Acta., 1352, 1–7. LiCalsi, C., Christophe, S., Steger, D.J. et al. (2000) AP-2 family members regulate basal and cAMP-induced expression of human chorionic gonadotropin. EMBO J., 28, 1036–1043. Mastick, C., Aebersold, R. and Lienhard, G.E. (1994) Characterization of a major protein in GLUT4 vesicles. Concentration in the vesicles and insulinstimulated translocation to the plasma membrane. J. Biol. Chem., 269, 6089–6092. McPherson, L.A. and Weigel, R.J. (1999) AP2-alpha and AP2-gamma: a comparison of binding site specificity and trans-activation of the estrogen receptor promoter and single site promoter constructs. Nucleic Acids Res., 27, 4040–4049. Mitchell, B.F. and Wong, S. (1995) Metabolism of oxytocin in human decidua, chorion, and placenta. J. Clin. Endocrinol. Metab., 80, 2729–2733.

894

Mizutani, S. and Tomoda, Y. (1996) Effects of placental proteases on maternal and fetal blood pressure in normal pregnancy and preeclampsia. Am. J. Hypertens., 9, 591–597. Mizutani, S., Yoshino, M. and Oya, M. (1976) A comparison of oxytocinase and L-methionine-insensitive leucine aminopeptidase during normal pregnancy. Clin. Biochem., 9, 228. Nagasaka,T., Nomura, S., Okamura, M. et al. (1997) Immunohistochemical localization of placental leucine aminopeptidase/oxytocinase in normal human placental, fetal and adult tissues. Reprod. Fertil. Dev., 9, 747–753. Nakanishi, Y., Nomura, S., Okada, M. et al. (2000) Immunoaffinity purification and characterization of native placental leucine aminopeptidase/oxytocinase from human placenta. Placenta, 21, 628–634. Nishikawa, N.S., Izumi, M., Uchida, H. et al. (2000) Cloning and characterization of the 5⬘-upstream sequence governing the cell cycledependent transcription of mouse DNA polymerase alpha 68 kDa subunit gene. Nucleic Acids. Res., 28, 1525–1534. Nomoto, S., Tatematsu, Y., Takahashi, T. et al. (1999) Cloning and characterization of the alternative promoter regions of the human LIMK2 gene responsible for alternative transcripts with tissue-specific expression. Gene, 236, 259–271. Nomura, S., Lahuna, O., Suzuki, T. et al. (1997) A specific distal promoter controls gamma-glutamyl transpeptidase gene expression in undifferentiated rat transformed liver cells. Biochem. J., 326, 311–320. Perissi, V., Menini, N., Cottone, E. et al. (2000) AP-2 transcription factors in the regulation of ERBB2 gene transcription by estrogen. Oncogene, 19, 280–288. Rabadan-Diehl, C., Lolait, S. and Aguilera, G. (2000) Isolation and characterization of the promoter region of the rat vasopressin V1b receptor gene. J. Neuroendocrinol., 12, 437–444. Rasmussen, T.E., Pedraza-Diaz, S., Hardre, R. et al. (2000) Structure of the human oxytocinase/insulin-regulated aminopeptidase gene and localization to chromosome 5q21. Eur. J. Biochem., 267, 2297–2306. Richardson, B.D., Langland, R.A., Bachurski, C.J. et al. (2000) Activator protein-2 regulates human placental lactogen gene expression. Mol. Cell. Endocrinol., 160, 183–192. Ringler, G.E. and Strauss III, J.F. (1990) In vitro systems for the study of human placental endocrine function. Endocr. Rev., 11, 105–123. Rogi, T., Tsujimoto, M., Nakazato, H. et al. (1996) Human placental leucine aminopeptidase/oxytocinase. A new member of type II membrane-spanning zinc metallopeptidase family. J. Biol. Chem., 271, 56–61. Ross, S.A., Keller, S.R. and Lienhard, G.E. (1998) Increased intracellular sequestration of the insulin-regulated aminopeptidase upon differentiation of 3T3-L1 cells. Biochem. J., 330, 1003–1008. Shi, D., Winston, J.H., Blackburn, M.R. et al. (1997) Diverse genetic regulatory motifs required for murine adenosine deaminase gene expression in the placenta. J. Biol. Chem., 272, 2334–2341. Smale, S.T. (1997) Transcription initiation from TATA-less promoters within eukaryotic protein-coding genes. Biochim. Biophys. Acta, 1351, 73–88. Tsujimoto, M., Mizutani, S., Adachi, H. et al. (1992) Identification of human placental leucine aminopeptidase as oxytocinase. Arch. Biochem. Biophys., 292, 388–392. Wu, F. and Lee, A.S. (2000) YY1 as a regulator of replication-dependent hamster histone H3.2 promoter and an interactive partner of AP-2. J. Biol. Chem., 236, 259–271. Yamada, K., Harada, N., Honda, S. et al. (1995) Regulation of placentaspecific expression of the aromatase cytochrome P-450 gene. J. Biol. Chem., 270, 25064–25069. Yamahara, N., Nomura, S., Suzuki, T. et al. (2000) Placental leucine aminopeptidase/oxytocinase in maternal serum and placenta during normal pregnancy. Life Sci., 66, 1401–1410. Received on January 24, 2001; accepted on July 9, 2001