Genomic structure and promoter activity of the mouse polysialic acid synthase gene (mST8Sia II). Brain-specific expression from a TATA-less GC-rich sequence.

The mouse ST8Sia II (mST8Sia II/STX) gene encodes a neural cell adhesion molecule-specific polysialic acid synthase whose expression is regulated during the developmental stages of mouse brain. To elucidate the molecular mechanism by which the expression is tissue-specifically and developmentally regulated, we isolated the complete genomic DNA and characterized the promoter of the gene for mST8Sia II. The gene encoding mST8Sia II was found to span about 80 kilobases and to be composed of six exons. Primer extension and S1 nuclease protection analyses revealed that the transcription started from 167 nucleotides upstream of the translational initiation site. Promoter analyses of the 5′-flanking region of the mST8Sia II gene using a luciferase gene reporter system revealed strong promoter activity in retinoic acid-induced differentiated P19 cells, which highly express the mST8Sia II gene. Deletion analyses demonstrated that the minimal promoter activity detected for the proximal region 325 base pairs upstream from the translational initiation codon (−158 to +167) could be modulated by various sequences within the 9.5-kilobase 5′-flanking region. The minimal promoter was embedded in a GC-rich domain (74%, GC content), in which two Sp1 binding motifs as well as a long purine-rich region were found, but it lacked TATA and CAAT boxes. The positive regulatory region located between −159 and −659 contained two additional Sp1 binding motifs and a long pyrimidine-rich region. We also found that the minimal promoter region of the mST8Sia II gene was sufficient for expression of a reporter gene in mST8Sia II gene-expressing neural differentiated P19 cells but not in nonexpressing ones. Thus the TATA-less GC-rich minimal promoter region of mST8Sia II probably controls the cell type-specific expression of the mST8Sia II gene.

The mouse ST8Sia II (mST8Sia II/STX) gene encodes a neural cell adhesion molecule-specific polysialic acid synthase whose expression is regulated during the developmental stages of mouse brain. To elucidate the molecular mechanism by which the expression is tissuespecifically and developmentally regulated, we isolated the complete genomic DNA and characterized the promoter of the gene for mST8Sia II. The gene encoding mST8Sia II was found to span about 80 kilobases and to be composed of six exons. Primer extension and S1 nuclease protection analyses revealed that the transcription started from 167 nucleotides upstream of the translational initiation site. Promoter analyses of the 5flanking region of the mST8Sia II gene using a luciferase gene reporter system revealed strong promoter activity in retinoic acid-induced differentiated P19 cells, which highly express the mST8Sia II gene. Deletion analyses demonstrated that the minimal promoter activity detected for the proximal region 325 base pairs upstream from the translational initiation codon (؊158 to ؉167) could be modulated by various sequences within the 9.5-kilobase 5-flanking region. The minimal promoter was embedded in a GC-rich domain (74%, GC content), in which two Sp1 binding motifs as well as a long purine-rich region were found, but it lacked TATA and CAAT boxes. The positive regulatory region located between ؊159 and ؊659 contained two additional Sp1 binding motifs and a long pyrimidine-rich region. We also found that the minimal promoter region of the mST8Sia II gene was sufficient for expression of a reporter gene in mST8Sia II gene-expressing neural differentiated P19 cells but not in nonexpressing ones. Thus the TATA-less GC-rich minimal promoter region of mST8Sia II probably controls the cell type-specific expression of the mST8Sia II gene.
Polysialic acid (PSA) 1 is a linear homopolymer of ␣2, 8-sialic acid residues mainly associated with the neural cell adhesion molecule (N-CAM) in mammalian cells, and is implicated in the reduction of N-CAM adhesion through its large negative charge (1). The expression of PSA on N-CAM is developmentally regulated in the central and peripheral nervous systems (2)(3)(4)(5), i.e. PSA on N-CAM is more abundant in the fetal than in the adult stage. Recent data imply important functions of PSA in the pathfinding and targeting on innervation of axons, migration of neuronal cells and tumor cells, and spatial learning and memory (6 -8). PSA on N-CAM was shown to be synthesized through the action of specific sialyltransferase(s) (9,10), but the mechanisms underlying the regulation of PSA synthesis and PSA expression remain poorly understood.
Recently, two cDNAs encoding ␣2,8-sialyltransferases named ST8Sia II/STX (11)(12)(13) and ST8Sia IV/PST-1 (14 -16) were cloned. Mouse ST8Sia II (mST8Sia II) exhibits overall amino acid sequence identity of 56% with mouse ST8Sia IV (mST8Sia IV). Both sialyltransferases can synthesize PSA on ␣2,3-linked sialic acids on N-CAM without any initiators (16 -20). Northern blot analysis indicated that expression of the mST8Sia II gene was restricted to the brain and testis, whereas the mST8Sia IV gene was expressed strongly in lung and heart rather than brain (13,16). The expression of the mST8Sia II gene in the brain was strictly regulated during developmental stages (13), i.e. an mST8Sia II transcript could be detected in brain on embryonic day 14, which peaked on embryonic day 20 and then decreased progressively to an almost undetectable level by 2 weeks after birth. The expression of the mST8Sia IV gene was also higher in fetal than adult brain but was less regulated during brain development compared with that of the mST8Sia II gene (16). To elucidate the mechanisms underlying of their tissue-and development-specific expression, it is important to know the structures and activities of their promoters.
We showed recently that expression of the mST8Sia II gene increased in parallel with the increased activity of PSA synthase and the quantity of PSA N-CAM during neuronal differentiation of mouse teratocarcinoma P19 cells, which provides us with an in vitro model system for studying its promoter activity in the neural differentiation processes (21). In this paper, we describe the genomic structure and promoter sequence of the mST8Sia II gene and determination of the 5Јflanking region responsible for the cell type-specific promoter activity by means of transfection experiments using differentiated P19 cells.

EXPERIMENTAL PROCEDURES
Isolation of Genomic and cDNA Clones Encoding mST8Sia II-A mouse genomic cosmid library was constructed and screened as described previously (22). The locations of the exons of the mST8Sia II gene were determined by PCR (GeneAmp XL PCR kit; Perkin-Elmer) with specific oligonucleotide primers or by hybridizing radiolabeled sialyltransferase cDNA to the same blots.
PCR Amplification of the 5Ј cDNA End (RACE)-mRNAs from newborn mouse brain were extracted by the guanidinium isothiocyanate method and purified with Oligotex-dT30 (Takara-Shuzo, Japan). Amplification of the 5Ј end of mST8Sia II cDNA was performed essentially according to the procedure of Frothman et al. (24). cDNA was synthesized by reverse transcription (Superscript II; Life Technologies, Inc.) of 5 g of mouse brain poly(A) RNA using a primer O1-N6, 5Ј-TTAGATT-TGCTATGTAAGCTGTT-3Ј, which is complementary to nucleotides ϩ327 to ϩ303 of the mST8Sia II gene. The excess primers and deoxynucleotides were removed by passage of the cDNA through a Micro-Spin S-400 column (Pharmacia Biotech Inc.). The cDNA was A-tailed with 0.6 unit of terminal deoxynucleotidyltransferase (Boehringer Mannheim), using 0.05 mM dATP. Two consecutive PCRs were performed with two nested sets of primers. For pair 1, the forward primer was NotI-(dT) 18 (Pharmacia), and the reverse primer was O1-N6. For pair 2, the forward primer was as above but without the T-tail, 5Ј-AACTGGAAGAATTCGCGGCCGCAGGAA-3Ј, and the reverse primer was O1-EX3 (5Ј-CTTCGATCGCTGAGATGTCTGCGAAGATGAGG-3Ј; complementary to nucleotides ϩ255 to ϩ224 of the mST8Sia II gene). The cDNA was amplified for 35 cycles of a step program (94°C, 40 s; 55°C, 40 s; 72°C, 60 s). The amplification products were subcloned into pUC119 and then sequenced.
Primer Extension-The reverse primer O1-EX3, which is complementary to a portion of the first exon of the mST8Sia II gene, was end labeled with [␥-32 P]ATP using T4 polynucleotide kinase. The 32 P-labeled O1-EX3 was hybridized with 5 g of poly(A) RNA prepared from mouse brain or 5 g of yeast tRNA in a 10-l hybridization solution (250 mM KCl, 10 mM Tris-HCl, pH 8.3, and 1 mM EDTA) for 1.5 h at 37°C after denaturation of the mixture for 1 h at 60°C. Following hybridization, 60 l of reverse transcriptase buffer and 200 units of Superscript II (Life Technologies, Inc.) were added (final buffer concentrations: 75 mM KCl, 20 mM Tris-HCl, pH 8.3, 10 mM dithiothreitol, 10 mM MgCl 2 , and 0.25 mM dNTPs), followed by incubation for 1 h at 45°C. The protected products were separated on a 6% sequencing gel along with a dideoxy chain termination sequencing reaction of pO1-22E4.8, using O1-EX3 as the primer. S1 Analysis-pO1-22E4.8 was constructed by subcloning a 4.8-kb EcoRI fragment from COS O1-22 into the pUC119 plasmid. This fragment contains the 1.7-kb 5Ј-flanking region, all of the first exon, and part of the first intron of the mST8Sia II gene. A 32 P-labeled S1 probe complementary to nucleotides Ϫ393 to ϩ255 was generated using a GeneAmp XL PCR kit, with pO1-22E4.8 as the template and O1-EX3 as the primer. The extension products were digested with SmaI and denatured at 100°C for 3 min, and then the 648-nucleotide 32 P-labeled S1 probe was separated from the template by electrophoresis on a 6% sequencing gel. The probe was isolated by soaking gel pieces in the elution buffer (300 mM sodium acetate, pH 5.0, and 10 mM EDTA) at room temperature for 8 h, followed by ethanol precipitation.
The S1 probe was hybridized with 5 g of mouse brain poly(A) RNA or 5 g of yeast tRNA in 10 l of hybridization buffer (40 mM PIPES, pH 6.4, 1 mM EDTA, 0.4 M NaCl, and 50% formamide) overnight at 50°C, after denaturation of the mixture for 10 min at 80°C. Then 200 l of S1 mapping buffer (280 mM NaCl, 50 mM sodium acetate, pH 5, 4.5 mM ZnSO 4 , 10 mg/ml salmon sperm DNA, and 1000 units/ml S1 nuclease) was added to each sample, and the mixtures were incubated for 1 h at 37°C. Digestion was stopped by the addition of 1 l of 0.5 M EDTA, and the nucleic acids were recovered by ethanol precipitation. The protected products were separated on a 6% sequencing gel along with a dideoxy chain termination sequencing reaction of pO1-22E4.8, using O1-EX3 as the primer.
DNase I Footprinting Assays-Recombinant Sp1 was purchased from Promega. Plasmid pO1-FT was constructed by subcloning the PCR product using pO1-22E4.8, as the template, the O1-FTA primer (5Ј-AGCAAAGCTGTCAAACTGCGCCTGGAGCCCAG-3Ј; mST8Sia II coding strand, nucleotides Ϫ124 to Ϫ93) and O1-FTB primer (5Ј-TC-CACGCGCACGAGGGACACACACCCTGCGCT-3Ј; complementary to the mST8Sia II coding strand, nucleotides ϩ94 to ϩ63) into pBluescript SKϩ. 32 P-Labeled DNA fragments were prepared by PCR using 32 Plabeled O1-FTA and cold O1-FTB, with pO1-FT as the template. The binding reaction was carried out for 30 min at 25°C in 50 l of a mixture comprising 25 mM HEPES-KOH, pH 7.9, 0.5 mM EDTA, 50 mM KCl, 10% glycerol, 1 mM phenylmethanesulfonyl fluoride, 1 mM dithiothreitol, 0.3 g of poly(dI-dC), 10 kcpm of 32 P-labeled DNA fragment, and 1 footprinting unit of recombinant Sp1. Five l of freshly diluted DNase I (50 g/ml) was added to the mixture for 60 s at 25°C, and then the reaction was stopped by the addition of 100 l of 1% sodium dodecyl sulfate containing 20 mM EDTA and 200 mM NaCl. After phenol-chloroform extraction and ethanol precipitation, the reacted products were separated on a 6% sequencing gel along with a dideoxy chain termination sequencing reaction of pFT-O1, using O1-FTA as the primer.
Analysis of Promoter Activity-NIH 3T3 and undifferentiated P19 cells were seeded at 5 ϫ 10 4 cells/60-mm-diameter dish in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum 24 h prior to transfection, respectively. For the differentiation of P19 cells into neuronal cells, the cells were seeded into and aggregated in bacteriological grade dishes in the presence of 1 M retinoic acid at the cell density of 1 ϫ 10 5 /ml. After 3 days, the aggregates were trypsinized, and approximately 1 ϫ 10 5 cells/60-mm diameter dish (tissue culturegrade dishes) were plated in Dulbecco's modified Eagle's medium, 10% fetal calf serum 24 h prior to transfection.
The luciferase plasmid (5 g) used as the reporter and the pSR␤-Gal plasmid (0.5 g) used as an insertional control for transfection efficiency were transfected into the cells by means of Lipofectamine (Life Technologies, Inc.). After 48-h transfection, the cells were washed three times with phosphate-buffered saline and then lysed with cell lysis buffer (PG␤-50; Toyo-ink). Luciferase activity was measured using a PicaGene Luciferase assay system (Toyo-ink) and a Luminescencer AB-2000 (ATTO, Japan). Light activity measurements were performed in quadruplicate, averaged, and then normalized as to ␤-galactosidase activity to correct for the transfection efficiency. ␤-Galactosidase activity was measured using a Luminescent ␤-Galactosidase Detection Kit (Clontech).

RESULTS
Isolation of mST8Sia II Genomic Clones-Screening of an NIH 3T3 cell cosmid library for mST8Sia II cDNA resulted in the isolation of four independent clones. A restriction map of the approximately 100-kb region containing the mST8Sia II gene is shown in Fig. 1.
We sequenced the exons, to determine their exact sizes, as well as the sequences of the intron/exon junctions (Table I). The sequences of all the intron-exon splice junctions conformed to the GT-AG rule (25). Introns were present only in the protein coding region of mST8Sia II. All of the splicing junctions of the mST8Sia II gene occurred after the second nucleotide of the amino acid codon. The mST8Sia II cDNA was divided into six exons, ranging from 63 to 4,340 bp, with intron sizes of about 4 -39 kb and spanning approximately 80 kb of genomic DNA (Fig. 1). Exon 1 contained the entire 5Ј-untranslated region and the beginning of the coding region to amino acid residue 32. This exon contained a cytoplasmic domain, a hydrophobic signal anchor sequence, and part of a stem domain. Exons 4 -6 encoded the putative active domain of the enzyme, and exon 6 contained a large 3Ј-untranslated region. We reported previously mST8Sia II cDNA sequences lacking the whole 3Ј-untranslated region. Therefore, to determine the 3Ј end of the 5-kb mST8Sia II transcript, the mouse brain cDNA library was screened by PCR using primers distributed along the 3Ј part of the gene. Sequence analyses of the PCR products revealed that the size of the transcribed RNA was 5,350 bp, thus it included a large 3Ј-untranslated region of 4055 bp. Poly(A) addition had occurred 18 nucleotides downstream of the sequence at the T residue of the polyadenylation signal (AATAAA).
Mapping of the Transcription Start Site-The transcription start site was determined by S1 nuclease mapping and primer extension with RNA recovered from 1-day mouse brain, in which the mST8Sia II gene was highly expressed (Fig. 2). A single stranded 32 P-labeled S1 probe, corresponding to nucleotides Ϫ393 to ϩ205 of the mST8Sia II gene, was hybridized with poly(A) RNA, followed by S1 nuclease digestion. The protected fragments were analyzed on a 6% sequencing gel. Yeast tRNA was used as a control to protection specificity. The end points of the protected fragments were determined by comparison with a sequence ladder derived from the same genomic DNA template (pO1-22E4.8) and the original primer, O1-EX3 (nucleotides ϩ255 to ϩ224, which is complementary to the mST8Sia II mRNA), used for synthesizing the S1 probe. The end point was determined to be at cytosine (ϩ1), which corresponded to a position 167 nucleotides upstream from the initiation codon, ATG ( Fig. 2A). Control experiments involving tRNA showed no such band. Primer extension with the same primer (O1-EX3) as used for S1 analysis resulted in extension products corresponding to the same cytosine (ϩ1) (Fig. 2B). Consistent with the results of primer extension and S1 nuclease experiments, a predominant initiation site was found 167 nucleotides upstream of ATG, which gave the only comigrating product in both S1 nuclease protection and primer extension experiments. Moreover, we performed 5Ј RACE-PCR on newborn mouse brain poly(A) RNA to identify the 5Ј end of the ST8Sia II gene, and the longest RACE-PCR products corresponded to the transcription initiation site determined in the S1 nuclease protection and primer extension experiments.
Analysis of the 5Ј-flanking Region-Analysis of the sequence 2.5-kb upstream of the transcription initiation site revealed that the mST8Sia II gene promoter has no typical TATA or CAAT boxes (Fig. 3). As shown in Fig. 3, the sequence of the putative promoter region of the mST8Sia II gene was embedded in a GC-rich domain. The GC content of the sequence between immediately upstream of the transcription initiation site and the translational initiation site (nucleotides Ϫ175 to ϩ168) was 74%. A pyrimidine-rich region (nucleotides Ϫ500 to Ϫ453) and a purine-rich region (nucleotides Ϫ83 to Ϫ11) were found in the 5Ј-flanking region. The TATA-and CAAT-less mST8Sia II gene promoter contains three Sp1 binding sequences, (G/T)GGGCGG(G/A)(G/A)(C/A), at positions Ϫ34 (matching 9 -10), Ϫ80 (matching 9 -10), and Ϫ170 (matching 6 -10), and an inverted Sp1 binding site at position Ϫ573 (matching 8 -10). To determine whether or not the proximal Sp1 binding sites upstream of the transcriptional initiation site are functional, we performed DNase I footprinting analysis with Sp1. As shown in Fig. 4, the Sp1 binding sites at positions Ϫ34 and Ϫ80 were both protected.
Demonstration of Promoter Activity-To determine mST8Sia II gene promoter activities, we used mouse embryonal car- cinoma P19 cells in which the mST8Sia II gene dramatically increases during neural differentiation on retinoic acid treatment (21).
To determine whether or not the 5Ј-flanking sequence of the mST8Sia II gene contains a cell type-specific promoter, we constructed a reporter plasmid, pBO1-EN9.6, containing a 9.6-kb 5Ј-flanking sequence of the mST8Sia II gene (Ϫ9600 to ϩ168) fused to the promoterless luciferase gene in pPGBII (Fig.  5). The construct was assayed for promoter activity by transient transfection into undifferentiated and differentiated P19 cells, and fibroblast NIH 3T3 cells (not expressing the ST8Sia II gene). As a negative control, pPGBII was transfected into parallel cultures of each cell line. The luciferase activity due to each luciferase reporter plasmid was normalized as to the ␤-galactosidase activity by cotransfecting an internal control plasmid, pSR␤-Gal, carrying a ␤-galactosidase gene under the control of the SR␣ promoter. The pBO1-EN9.6 construct showed the highest level of promoter activity when expressed in the differentiated P19 cells, which was 4-fold higher than that in undifferentiated P19 cells. No promoter activity was detected in NIH 3T3 cells. These results suggest that the 9.6-kb 5Ј-flanking sequence contains sufficient information to direct cell type-specific expression of the mST8Sia II gene.
A series of reporter plasmids containing progressive 5Ј dele-tions between nucleotides Ϫ9600 and Ϫ9 was constructed, and the plasmids were examined for their promoter activities in each cell type to map the regions that regulate cell type-specific gene expression (Fig. 5). Sequential deletions of the region between nucleotides Ϫ9600 and Ϫ5400 had little effect on luciferase activity. However, further deletions from nucleotides Ϫ5400 to Ϫ3400 increased the promoter activity in differentiated P19 cells, suggesting the presence of a negative regulatory element in this region. Deletions up to nucleotide Ϫ659 had no effect on the activity in differentiated P19 cells but increased the activity in undifferentiated P19 cells. Deletions of the sequence from positions Ϫ659 to Ϫ293 increased the luciferase gene expression in differentiated P19 cells but caused a decrease to one-half in undifferentiated P19 cells. The region between Ϫ659 and Ϫ293 contained an inverted Sp1 binding site and a pyrimidine-rich region (Fig. 3). Moreover, further deletions to nucleotide Ϫ158 (pBO1-XN0.35) caused the promoter activity to decrease to approximately half of that of pBO1-SN0.45 in both differentiated and undifferentiated P19 cells, suggesting the presence of a positive regulatory element in this region (nucleotides Ϫ158 to Ϫ293), which contains one Sp1 binding site. Extension of the 5Ј deletion to nucleotide Ϫ9 (pBO1-XN0.15) drastically reduced the promoter activity to the level seen with the promoterless control vector pPGBII in all

intron junctions of the mST8Sia II gene
Nucleotide sequences at the intron (lowercase letters) and exon (uppercase letters) junctions are shown. The derived amino acid sequence is displayed below the nucleotide sequence. Exons are numbered from the 5Ј end as described in Fig. 1. Exon and intron sizes are indicated in base pairs. Numbering starts with position ϩ1 at the adenosine of the initiator methionine.  2. Identification of the transcription initiation sites of the ST8Sia II gene of mouse brain. The transcription start site has been mapped by means of both S1 nuclease protection (panel A) and primer extension (panel B) analysis. For S1 mapping, a 648-nucleotide single-strand DNA complementary to nucleotides Ϫ393 to ϩ255 was generated as described under "Experimental Procedures" and hybridized with 5 g of poly(A) RNA isolated from 1-day mouse brain or 5 g of yeast tRNA before the S1 nuclease reaction. For the primer extension reaction, a 32-mer oligonucleotide (O1-EX3) complementary to mST8Sia II mRNA (nucleotides ϩ255 to ϩ224) was [␥-32 P]ATP end labeled, hybridized with 5 g of poly(A) RNA from 1-day mouse brain or 5 g of yeast tRNA, and then reverse transcribed. The S1 nuclease-protected fragment as well as the primer-extended product were run on a sequencing gel along with a sequencing reaction of pO1-22E4.8, using O1-EX3 as the primer. In panel A, lane 1 is S1 nuclease reaction with 1-day mouse brain mRNA; lane 2, with yeast tRNA. In panel B, lane 1 is primer extension with 1-day mouse brain mRNA; lane 2, with yeast tRNA. The arrow indicates the position of the transcription start site. cells, indicating that the sequence between Ϫ158 and Ϫ9 is necessary for the minimal promoter activity. Two Sp1 binding sites, which were revealed to be functional on footprinting analysis (Fig. 4), and a purine-rich region were found in the region between Ϫ9 and Ϫ158. Interestingly, the promoter activities due to the pBO1-XN0.35 and pBO1-SN0.45 constructs were approximately 10-fold higher in differentiated P19 cells than undifferentiated P19 cells. On the other hand, all constructs gave hardly detectable activity in NIH 3T3 cells, which do not express the mST8Sia II gene. These results suggest that the 325-bp sequence upstream from the translational initiation codon (nucleotides Ϫ158 to ϩ167) is necessary for the minimum and cell type-specific promoter activity.

DISCUSSION
In this study, we determined the entire genomic organization of the mST8Sia II gene, characterized the functional promoter activity for the 5Ј-flanking region in transient transfection assays, and investigated the expression of the gene in mouse teratocarcinoma P19 cells. We demonstrated that the minimal promoter region of the mST8Sia II gene is sufficient to express cell type-specific promoter activity, which is correlated with mST8Sia II gene expression in the cells.
Previously, the genomic structures of five other sialyltransferase genes have been reported, i.e. rat galactoside ␣2,6-sialyltransferase (rST6Gal I) (26,27), human galactoside ␣2,3sialyltransferase (hST3Gal I) (28), human Gal␤1,3GalNAc/ Gal␤1,4GlcNAc ␣2,3-sialyltransferase (hST3Gal IV) (29), mouse N-acetylgalactosamide ␣2,6-sialyltransferase (mST6GalNAc II) (30), and mouse sialoside ␣2,8-sialyltransferase (mST8Sia III) (22). The genomic structure of the mST8Sia II gene was most similar to that of the mST8Sia III gene, as follows (Fig. 6). The putative active domains of mST8Sia II and mST8Sia III were encoded by only three and two exons, respectively, whereas those of the other genes consisted of at least five exons. In particular, the positions of the two exon-intron boundaries of the mST8Sia II gene (introns 3 and 5) and the mST8Sia III gene (introns 2 and 3) were identical, and all of their splice junctions occurred after the second nucleotide of the amino acid codon. Although in mST8Sia III sialyl motif L, a highly conserved putative nucleotide sugar binding domain (31) is encoded by one exon, the corresponding motif in mST8Sia II is encoded by discrete exons, like those observed in other genes. The similarity of the genomic organization and amino acid sequences between mouse ST8Sia II and ST8Sia III suggests that these genes constitute a subgroup distinct from other sialyltransferase genes.
The results of primer extension and S1 protection analyses showed that mST8Sia II gene has a single transcription initiation site. In addition, we demonstrated that the 5Ј-flanking sequence of the mST8Sia II gene contained a functional promoter that was highly active in neural differentiated P19 cells, which strongly express the mST8Sia II gene. Deletion analysis clearly indicated that the minimal promoter, the region contained within the 158 bp upstream of the transcription start site (pBO1-XN0.35), exhibited substantial tissue specificity (Fig. 5). In differentiated P19 cells, an approximately 10-fold increase in promoter activity was observed with the pBO1-XN0. 35 construct compared with that in undifferentiated P19 cells. We showed recently that the expression of the mST8Sia II gene in neural differentiated P19 cells increased FIG. 5. mST8Sia II gene promoter activity and identification of the regulatory regions. Shown is a schematic representation of DNA constructions containing various lengths of the mST8Sia II promoter linked to the luciferase gene (pPGBII). Each DNA fragment subcloned into the luciferase reporter plasmid is defined as to its position in the mST8Sia II gene promoter relative to the transcription start (ϩ1). Five g of each construct was transfected into NIH 3T3 (3T3), undifferentiated P19 (P19), or neural-differentiated P19 (P19(Dif)) cells. Luciferase activity was normalized as to ␤-galactosidase activity of a cotransfected internal control plasmid, pSR␤-Gal. The data are expressed as the fold increase in enzyme activity compared with transfections using a promoterless luciferase plasmid, pPGBII. Results are the average of five experiments.
FIG. 6. Comparison of the exon structure of the mST8Sia II gene with those of other sialyltransferase genes. The intron-exon structures of five sialyltransferase genes are presented (22, 26, 28 -30). The protein domain structure is represented schematically by a rectangle, which is subdivided to show the major structural elements of the protein. Sialyl motifs L and S are underlined. The identical positions of the two introns of the mST8Sia II and mST8Sia III genes are indicated by dotted lines. dramatically to a level approximately 10 -20 times higher than that in undifferentiated cells (21). The increase in minimal promoter activity on neural differentiation is correlated with the gene expression of mST8Sia II in P19 cells. Moreover, the promoter activity was negligible in NIH 3T3 cells, in which the mST8Sia II gene was not expressed. Therefore, the minimal promoter region may possess the ability to initiate transcription in a cell-specific fashion.
The proximal region 158 bp upstream from the transcription initiation site was characterized as a minimal promoter that contained two functional Sp1 binding sites but lacked consensus TATA and CAAT boxes. In addition, the 5Ј-untranslated region of the mST8Sia II gene contained a large GC-rich domain with the characteristics of a CpG island. These structural features are usually associated with housekeeping gene promoters. However, recent studies have shown that many tissuespecific gene promoters, including neural cell-specific promoters, include the structural features of housekeeping gene-like promoters. Neuron-specific TATA-less promoters have been shown to have either a single transcription initiation site, as observed in the synapsin I gene (32), synapsin II gene (33), nerve growth factor receptor gene (34), and several olfactory neuron-specific genes (35), or to have multiple initiation sites, such as those found in the genes encoding synaptophysin, the type II sodium channel (36), the D 1A dopamine receptor (37), and brain-specific aldolase C (38). Therefore, the mST8Sia II promoter may belong to the former group, i.e. GC-rich and TATA-less promoters with a single transcription initiation site.
Two Sp1 binding sites were found in the minimal promoter region, and DNase I protection assays with the purified Sp1 protein revealed that these two sites were functional (Fig. 4). Moreover, two additional Sp1 binding sites were present in the regulatory regions (Ϫ158 to Ϫ293 and Ϫ293 to Ϫ659). It was reported previously that the distantly located site functions synergistically with the promoter-proximal site to activate transcription strongly in vivo, and that this synergism is a direct consequence of interactions between remote and local Sp1s (39). Deletion of the sequence from positions Ϫ158 to Ϫ293 of mST8Sia II reduced the promoter activity, suggesting that Sp1 bound to the site at Ϫ170 probably interacts with another Sp1 in the minimal promoter region. The expression of Sp1 is ubiquitous, but it was reported previously that Sp1 levels appeared to be highest in developing hematopoietic cells, fetal cells, and spermatides in the mouse (40). The regulated expression of the mST8Sia II gene in fetal and newborn brain and testis seemed to correspond to that of Sp1. Although Sp1 was found to be expressed in NIH 3T3 and undifferentiated and differentiated P19 cells (data not shown), the expression level of the reporter gene fused to the minimal promoter sequence of ST8Sia II (Ϫ158/ϩ168) was very low in undifferentiated P19 cells and hardly detectable in 3T3 cells. These results suggest that some cell-specific transcription factor is involved in differentiated P19 cells and that it might be required for interaction with Sp1 in the minimal promoter elements, which allows activation of the mST8Sia II gene promoter. Further studies involving identification of the regulatory elements and their binding factors will facilitate elucidation of the tissue-specific and developmental regulation of ST8Sia II gene expression.