The human synapsin II gene promoter. Possible role for the transcription factor zif268/egr-1, polyoma enhancer activator 3, and AP2.

Synapsin II is a neuron-specific phosphoprotein that selectively binds to small synaptic vesicles in the presynaptic nerve terminal. Here we report the cloning and sequencing of the 5′-flanking region of the human synapsin II gene. This sequence is very GC-rich and lacks a TATA or CAAT box. Two major transcriptional start sites were mapped. A hybrid gene consisting of the Escherichia coli chloramphenicol acetyltransferase gene under the control of 837 base pairs of the synapsin II 5′-upstream region was transfected into neuronal and non-neuronal cells. While reporter gene expression was low in neuroblastoma and non-neuronal cells, high chloramphenicol acetyltransferase activities were monitored in PC12 pheochromocytoma cells. However, there was no correlation between reporter gene expression in the transfected cells and endogenous synapsin II immunoreactivity. Using DNA-protein binding assays we showed that the transcription factors zif268/egr-1, polyoma enhancer activator 3 (PEA3), and AP2 specifically contact the synapsin II promoter DNA in vitro. Moreover, the zif268/egr-1 protein as well as PEA3 were shown to stimulate transcription of a reporter gene containing synapsin II promoter sequences. In the nervous system, zif268/egr-1 functions as a “third messenger” with a potential role in synaptic plasticity. PEA3 is expressed in the brain and its activity is regulated by proteins encoded from non-nuclear oncogenes. We postulate that zif268/egr-1 and PEA3 couple extracellular signals to long-term responses by regulating synapsin II gene expression.

The synapsins are a family of neuronal phosphoproteins that are localized on the cytoplasmic surface of small synaptic vesicles (for review, see Greengard et al. (1993) and Thiel (1993)). This family consists of four proteins, synapsin Ia and synapsin Ib (collectively termed synapsin I) and synapsin IIa and synapsin IIb (collectively termed synapsin II). These isoforms are generated via alternative splicing from two different genes (Sü dhof et al., 1989). Molecular cloning of bovine, human, and rat synapsins revealed striking homologies in the amino-terminal 420 amino acids of all four synapsins. The major difference between synapsins I and II is the presence of a very proline-rich COOH-terminal domain in the synapsin I isoforms that contains clusters of basic amino acids as well as two recognition sites for Ca 2ϩ /calmodulin-dependent protein kinase II (Sü dhof et al., 1989).
Synapsin I has been postulated to link synaptic vesicles to the cytoskeleton, thus regulating the availability of synaptic vesicles for exocytosis. In addition, a role for synapsin I in the regulation of short-term plasticity has recently been suggested (Rosahl et al., 1993). The functional role of synapsin II is not yet clear. However, the fact that synapsin II interacts with both synaptic vesicles (Thiel et al., 1990;Siow et al., 1992) and actin filaments (Chilcote et al., 1994) in a similar fashion as synapsin I suggests that all synapsin isoforms share basic functional properties. This hypothesis was supported by results obtained with mice lacking synapsin I (Rosahl et al., 1993). These mice seem to have no serious impairment of brain function indicating that synapsin II might be compensating the loss of synapsin I.
Synapsins I and II have no non-neuronal counterpart in contrast to the synaptic vesicle proteins synaptobrevin and synaptophysin (Zhong et al., 1992;McMahon et al., 1993). The synapsin genes are therefore good candidates for an investigation of neuron-specific gene expression. Virtually all neurons express the synapsins, but not every isoform of synapsin is expressed in every nerve cell (Sü dhof et al., 1989). The human and rat synapsin I genes have been analyzed and it was shown that the 5Ј-flanking region contains neuron-specific enhancer elements (Sauerwald et al., 1990;Thiel et al., 1991). Here, we report the sequence of the human synapsin II promoter. Our studies have revealed similarities as well as distinct differences in comparison with the synapsin I promoter. Moreover, we present data indicating that the transcription factors zif268/ egr-1, PEA3, 1 and AP2 regulate synapsin II gene expression.

EXPERIMENTAL PROCEDURES
DNA Cloning-A genomic clone termed pSyC58 -10 that contained the synapsin II promoter region was kindly provided to us by Thomas Sü dhof, University of Texas, Dallas, TX.
Mapping of the Start Site of Transcription-Human brain RNA (Clontech) was analyzed by S1 nuclease protection assay. To synthesize a single-strand DNA probe complementary to the synapsin II mRNA, a PstI/PvuII-fragment of the genomic clone pSyC58-10 was inserted into M13 mp18, generating M13SyII-1. Using the oligonucleotide 5Ј-CCG-GCGCCTCAGGAAGTTCATCATCTG-3Ј, a primer extension reaction was performed in the presence of radiolabeled dCTP. The extension product was digested with EcoRI and purified on a denaturating polyacrylamide gel. The S1 digestion was done with 10 g of human brain RNA as described (Wille et al., 1991).
Reporter Constructs-Plasmid pSyCAT-10 has been described previously (Jü ngling et al., 1994). Plasmid pSyIICAT contains the sequence * This work was supported by the Deutsche Forschungsgemeinschaft through SFB 274, the Bayer AG (Bayer International Investigatorship), and a grant from the Fritz Thyssen Stiftung. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EMBL Data Bank with accession number(s) X89851.
‡ To whom correspondence should be addressed: Institute for Genetics, University of Cologne, Zü lpicher Str. 47, from Ϫ837 to ϩ110 of the human synapsin II gene inserted into the HindIII and XbaI sites of pCAT-Basic (Promega). Plasmids OVEC, JH514ref, and EBS-2 2 /OVEC have been described . To construct plasmid EBS-3 2 /OVEC that contains 2 copies of the putative zif268/egr-1 binding site of the human synapsin II promoter, the oligonucleotides 5Ј-AGCTTGAGCTCGAGCTGGGTCGCCCCCTCC-TGCAAG-3Ј and 5Ј-TCGACTTGCAGGAGGGGGCGACCCAGCTCGA-GCTCA-3Ј were annealed and ligated into the SstI and SalI sites of OVEC and the sites duplicated as described (Westin et al., 1987). Reporter plasmids SyIIPEA3 4 CAT and p38PEA3 4 CAT, containing PEA3 binding sites derived from synapsin II and synaptophysin promoters, were constructed by inserting synthetic annealed oligonucleotides 5Ј-TCGAGCTGCCTTCCTCCCTAG-3Ј and 5Ј-TCGACTAGGGAG-GAAGGCAGC-3Ј as well as 5Ј-TCGAGTCGCAGGAAGGAGGG-3Ј and 5Ј-TCGACCCTCCTTCCTGCGAC-3Ј into the SalI site and SalI/XhoI site, respectively, of pHIVTATA-CAT. This plasmid will be described elsewhere. 2 In brief, it contains the open reading frame of the Escherichia coli CAT gene under control of a minimal promoter consisting of the HIV TATA box and the adenovirus major late promoter initiator (Inr) element. The PEA3 binding sites were subsequently multimerized. 2 Plasmid EBS-2 2 CAT was generated similarly using two copies of the synapsin I promoter zif268/egr-1 binding site.
Expression Plasmids-Plasmid pCMVzif has been described . PEA3 expression plasmid pCMVPEA3 was constructed by cloning PEA3 cDNA derived from plasmid MLP.PEA3K (Xin et al., 1992) into pCMV5. The PEA3 cDNA was a kind gift of John A. Hassell, McMaster University, Canada.
Expression of AP2 and PEA3 in E. coli-The human AP2 cDNA derived from plasmid pP adh AP2 (Williams et al., 1988), a kind gift of Robert Tjian, University of California at Berkley, was cloned into the NcoI site of pRSETB (Kroll et al., 1993), thus generating plasmid pRSET-AP2. This plasmid was introduced into E. coli strain BL21(DE3) and expression was induced by adding 1.0 mM isopropyl-␤-D-thiogalactopyranoside. The recombinant AP2 was subsequently purified by affinity chromatography on Ni 2ϩ -nitrilotriacetic acid columns using 200 mM imidazole for eluting the bound protein. To express the DNAbinding domain of PEA3 as a fusion protein with glutathione S-transferase (GST), a BclI/EcoRI fragment of plasmid pMLP.PEA3K (Xin et al., 1992) was cloned into pGEX-3X (Pharmacia), generating plasmid pGEX-PEA3. This plasmid was introduced into the protease-deficient E. coli strain NB42 (Mayer et al., 1992). The GST-PEA3 fusion protein was expressed and purified as described (Smith and Johnson, 1988).
Expression of Zif268/egr-1 and PEA3 in 293 Human Kidney Cells-293T/17 cells were transfected with expression plasmids encoding zif268/egr-1 and PEA3, respectively, using the calcium phosphate coprecipitation method. Cells were harvested 2 days later and nuclear extracts were prepared as described .
DNase I Footprinting-Binding reactions (50 l) contained 25 mM HEPES, pH 7.6, 50 mM KCl, 6.25 mM MgCl 2 , 0.5 mM EDTA, 0.5 mM dithiothreitol, 10% glycerol, 20 g/ml poly(dI-dC), 20,000 -30,000 cpm of radiolabeled probe, and 25 g of recombinant GST or GST-PEA3 protein. The samples were incubated on ice for 10 min and shifted to room temperature for 2 min. 50 l of a room temperature solution containing 5 mM CaCl 2 and 10 mM MgCl 2 was subsequently added and the incubation continued for 1 min. DNase I (Boehringer Mannheim) was then added for 1 min. The reaction was terminated by adding 100 l of DNase I stop solution (0.2 M NaCl, 0.03 M EDTA, 1% SDS, 100 g/ml yeast RNA). The samples were incubated for 30 min at 37°C, extracted with phenol/chloroform, precipitated with ethanol, and finally separated on a sequencing gel.
Miscellaneous Techniques-For Western blot analysis, proteins were separated on a 10% SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (pore size 0.2 m). The antibodies recognizing synapsin I and synapsin II were kind gifts of Thomas Sü dhof. Blots were developed using iodinated protein A (Amersham). Computer analysis was performed according to the program manual for the GCG package, version 7, Genetics Computer Group.

Cloning of the 5Ј-Flanking Region of the Human Synapsin II
Gene-The sequence of the 5Ј-flanking region of the human synapsin II gene is shown in Fig. 1A. The promoter does not contain a TATA or a CAAT box. The two major transcriptional initiation sites as well as consensus binding sites for the transcription factors zif268/egr-1, PEA3, and AP2 are depicted. The NH 2 -terminal amino acid sequence of human synapsin II as deduced from the genomic clone revealed typical synapsin II specific amino acid changes in comparison to synapsin I, i.e. Phe 4 instead of Tyr, Ser 13 instead of Asn, Ile 15 instead of Met, and Glu 30 instead of Gln. These changes are conserved for mouse, rat, and human synapsin II. The sequence surrounding the transcriptional start sites, as well as the proximal promoter region, is very GC-rich ( Fig. 1B), suggesting the possibility that methylation might have an effect upon synapsin II gene regulation (Tate and Bird, 1993). A dot matrix analysis between the human synapsin I and II promoters revealed low homology in comparison to high conservation of the coding region (Fig. 2).
Mapping of Transcription Initiation Sites in the Human Synapsin II Gene-Identification of the synapsin II transcription initiation site was revealed by S1 nuclease mapping. A cDNA probe was synthesized by primer extension from a M13 template that contained synapsin II gene sequence from Ϫ301 to ϩ241 (Fig. 3A). Human brain total RNA was hybridized to this probe as well as total RNA from HeLa cells and yeast tRNA as negative controls. Incubation with S1 nuclease revealed two major protected fragments in the reaction containing human brain RNA ( Fig. 3B) but not in the control reactions. Sequences are numbered with ϩ1 being the first transcribed nucleotide corresponding to the start site most 3Ј to the ATG codon.
Promoter Activity of the 5Ј-Flanking Region-Plasmid pSyII-CAT contains 837 nucleotides of 5Ј-flanking sequence, the transcription initiation site as well as 110 nucleotides of 5Ј-untranslated region of the human synapsin II gene fused to a promoterless CAT gene. This plasmid was transiently transfected into neuronal and non-neuronal cell lines together with plasmid pCMV␤ to correct for variations in transfection efficiencies. Fig. 4A shows that the synapsin II promoter is weak in the neuronal cell lines NS20Y, NS26, and NG108-15 as well as in the fibroblasts CHO and HeLa. There is virtually no difference in CAT activity between these cell lines. In addition, we analyzed three different clones of the pheochromocytoma cell line PC12. CAT activity was 4 -12 times higher in these cells than in the neuroblastoma cells or fibroblasts. As a control we introduced plasmid pSyCAT-10 into these cells (Fig. 4B). The synapsin I promoter-CAT gene was highly expressed in both neuroblastoma and PC12 cells, and poorly expressed in nonneuronal cells.
Immunoblot analysis of the endogenous synapsins in neuronal cells used here revealed that both synapsins I and II are present in NS20Y, NS26, and NG108 -15 cells (Fig. 4C). The expression of synapsins I and II in PC12 cells is in general much lower. A quantitative estimation of the bound iodinated protein A by PhosphorImager analysis revealed that PC12/HD and PC12/SF cells contain only 20 or 13% of the synapsin IIb expressed in NG108 -15 cells. There was no correlation between reporter gene expression and endogenous synapsin II immunoreactivity. The synapsin II promoter-CAT gene is tran-scribed in PC12/NY cells 10 times more efficiently than in NS26 or NG108 -15 cells although the endogenous synapsin II levels are comparable. Furthermore, transfection experiments of PC12/HD cells revealed 8 times higher CAT activities than in NS26 or NG108 -15 cells despite the fact that endogenous synapsin II levels are much lower in the PC12 clone. Based on the low CAT activities in neuroblastoma cells as well as the Mapping of transcriptional initiation sites by S1 nuclease protection assay. A, the cartoon shows the strategy in synthesizing a human synapsin II specific DNA probe by primer extension. B, S1 nuclease protection mapping using human brain total RNA or as controls yeast tRNA and total RNA from HeLa cells. An aliquot of undigested probe is depicted as well as molecular weight markers (HaeIII cut pBR322). The arrowheads indicate the two transcriptional initiation sites. On the right-hand site, a sequencing reaction is depicted that was generated with the primer used to synthesize the S1 probe. lack of correlation between endogenous synapsin II expression and CAT activity following transfection experiments of a synapsin II promoter-CAT construct, we conclude that the elements directing neuron-specific expression of synapsin II are located outside the analyzed region encompassing the sequence Ϫ837 to ϩ110.
Potential Regulatory Sequence Motifs within the Human Synapsin II Gene Promoter Region-The human synapsin II promoter is very GC-rich (Fig. 1B) and lacks TATA and CAAT boxes. Numerous Sp1 sites in the proximal 5Ј-flanking region were detected. Sequence motifs from Ϫ56 to Ϫ47, from Ϫ67 to Ϫ58, from Ϫ86 to Ϫ77, and from Ϫ110 to Ϫ101 are identical to SP1 binding sites described (Jones and Tjian, 1985;Ishii et al., 1986).
The 5Ј-flanking region also contains a repetitive sequence element consisting of multiple GT repeats (nucleotides Ϫ595 to Ϫ540). Similar sequences were found in other neuronally expressed genes, i.e. the genes encoding neuron-specific enolase (Sakimura et al., 1987), synapsin I (Sü dhof, 1990), and GAP-43 (Nedivi et al., 1992). It has been suggested that the GT element activates transcription together with the adenovirus E1A protein (Berg et al., 1989). We could not confirm this observation in transfection experiments of synapsin II-CAT constructs into EIA containing 293 cells (data not shown). Alternatively, the GT element has been suggested to be a structural element in helping to form left-handed or Z-form DNA (Sakimura et al., 1995).
Additionally, consensus sites for the transcription factors zif268/egr-1, PEA3, and AP2 were detected in the synapsin II 5Ј-flanking region. These factors were therefore analyzed for their possible functional roles in synapsin II gene regulation.
The Zinc Finger Transcription Factor Zif268/egr-1 Binds to the Proximal Region of the Synapsin II Promoter and Functions as a Transcriptional Activator-We recently demonstrated that the synapsin I gene is regulated by zif268/egr-1 . The synapsin II 5Ј-flanking region contains a sequence 5Ј-CGCCCCCTC-3Ј that is a 8/9 match to the zif268/egr-1 consensus site 5Ј-CGCCCCCGC-3Ј (Fig. 5A). The zif268/egr-1 binding site in the synapsin I promoter has been termed EBS-2 . By analogy, we refer to the putative zif268/ egr-1 binding site in the synapsin II promoter as EBS-3 (Fig. 5A).
To test whether zif268/egr-1 binds to this motif we performed electrophoretic mobility shift assays. As a source for zif268/ egr-1, 293 human embryonic kidney cells were transfected with an expression vector encoding zif268/egr-1 and nuclear extracts were prepared. As a control, 293 cells were transfected with an expression vector for PEA3. Fig. 5B shows in the left panel the results obtained with the zif268/egr-1 site derived from the synapsin I promoter. Using nuclear extract prepared from 293 cells containing zif268/egr-1, we observed a specific DNA-protein complex that was absent in the reaction with nuclear extracts from 293 cells transfected with an expression vector for PEA3 instead of zif268/egr-1 (compare lanes 2 and 12, marked by an arrowhead). When unlabeled EBS-2 DNA was added to the reaction, this complex disappeared (lanes 3-5). Unlabeled oligonucleotides encompassing the EBS-3 site from the synapsin II promoter also competed efficiently (lanes 6 -8), whereas an unrelated oligonucleotide containing the PEA3 site from the synapsin II promoter did not reduce the intensity of the observed DNA-protein complex (lanes 9 -11). On the right panel of Fig. 5B, the results obtained with the probe containing the EBS-3 site are depicted. Nuclear extracts of 293 cells containing zif268/egr-1 revealed a similar DNA-protein complex (lane 14, marked by an arrowhead) that was competed efficiently by an excess of EBS-3 (lanes 15-17) or EBS-2 DNA (lanes 18 -20). The control probe containing the PEA3 binding site failed to compete (lanes 21-23), suggesting that zif268/ egr-1 recognizes the EBS-3 site in a sequence-specific manner. Moreover, the EBS-3⅐zif268/egr-1 complex was not detectable in nuclear extracts from 293 cells transfected with a PEA3 expression vector (lane 24). The binding experiments depicted in Fig. 5B show, in addition, unspecific complexes that are present in 293 extracts transfected with either zif268/egr-1 or PEA3 expression vectors. However, the retarded band originating from the presence of zif268/egr-1 is clearly defined. We conclude from these results that zif268/egr-1 binds specifically to the promoter region of the synapsin II gene.
To analyze whether zif268/egr-1 functions as a transcriptional activator for synapsin II gene expression, two copies of the EBS-3 sequence were cloned immediately upstream of the TATA box of the OVEC plasmid that contains the rabbit ␤-globin gene as a reporter (Fig. 5C). As a positive control, a plasmid containing two copies of the zif268/egr-1 binding site derived from the synapsin I promoter was used. These plasmids were transfected into CHO-KI cells together with pCMV5 DNA or an expression vector for zif268/egr-1 (Fig. 5C). Plasmid JH514ref containing a mutated ␤-globin gene under the control of the SV40 promoter/enhancer was included in the transfection to control for variability in transfection efficiency . 48 h post-transfection, cytoplasmic RNA of the transfected cells was isolated, hybridized to a ␤-globin derived cRNA probe, and analyzed by RNase protection mapping (Fig. 5D). When zif268/egr-1 was overexpressed in cells transfected with EBS-2 2 /OVEC and EBS-3 2 /OVEC plasmids, transactivation was observed. A quantitative estimation of the transactivation properties of zif268/egr-1 by PhosphorImager analysis revealed FIG. 4. Transfection analysis of synapsin II promoter-CAT and synapsin I promoter-CAT fusion genes. The constructs pSyIICAT (A) and pSyCAT-10 (B) were introduced into the neuronal cell lines NS20Y, NS26, NG108 -15, PC12/NY, PC12/HD, and PC12/SF. To control for the presence of tissue-specific elements, both plasmids were also transfected into CHO-KI and HeLa cells. CAT activity was normalized for variations in transfection efficiency by dividing CAT activity by ␤-galactosidase activity. At least four experiments were done with each cell type and the mean Ϯ S.E. is depicted. C, Western blot analysis of endogenous synapsins I and II. Cell homogenates (70 g/lane) were subjected to 10% SDS-poyacrylamide gel electrophoresis and immunoblotted. As a positive control 3 g of forebrain homogenate was loaded into the left lane.
a 5.8-and 3.6-fold increase in transcription for EBS-2 and EBS-3 binding sites, respectively. These results indicate that zif268/egr-1 functions as a transcriptional activator after binding to the zif268/egr-1 motif of the synapsin II promoter.
PEA3 Interacts in a Sequence-specific Manner with the Synapsin II Promoter and Functions as a Transcriptional Activa-tor-The PEA3 motif 5Ј-AGGAAG-3Ј was first recognized in the polyoma enhancer (Martin et al., 1988). The PEA3 cDNA was recently cloned and the PEA3 mRNA was found to be expressed in brain (Xin et al., 1992). We recognized a sequence identical to the PEA3 motif in the human synapsin II (Ϫ638 to Ϫ633) as well as in the human synaptophysin promoter (Ϫ657/Ϫ652) (Fig. 7A). The synaptophysin promoter is numbered relative to the A of the start codon as ϩ1. To test whether PEA3 binds to the synapsin II and synaptophysin promoters, the PEA3 DNAbinding domain was expressed in E. coli as a fusion protein with Schistosoma japonicum GST (Fig. 6A) and purified by glutathione affinity chromatography (Fig. 6B). The recombinant GST-PEA3 fusion protein was subsequently used in DNAprotein binding assays. Using the PEA3 binding site derived from the synapsin II promoter as a probe, a DNA-protein complex was observed in the absence of competitors (Fig. 7B, middle panel, lane 12). When unlabeled DNA encompassing the synapsin II promoter PEA3 binding site was added to the reaction, the complex disappeared (lanes 13 and 14). Unlabeled oligonucleotides encompassing the PEA3 consensus site as well as the PEA3 site from the synaptophysin promoter also competed with the complex in proportion to the amount of competitor (lanes 15-18), whereas the unrelated oligonucleotide containing the zif268/egr-1-"like" binding site from the synapsin I promoter  interfered only slightly with binding when used at a 100-fold molar excess (lane 20). Incubation of the PEA3 probe derived from the synaptophysin promoter with the GST-PEA3 fusion protein revealed a DNA-protein complex (lane 22) that was competed efficiently by an excess of cold probe or by probes containing the consensus or the synapsin II promoter PEA3 binding site (lanes 23-28). The control oligonucleotide containing a zif268/egr-1 binding site failed to compete (lanes 29 and 30). The PEA3 consensus site (Xin et al., 1993) was tested as a positive control. The GST-PEA3 protein formed a complex with this DNA probe (Fig. 7B, lane 2) that could be competed by an excess of unlabeled probe (lanes 3 and 4) as well as by probes containing the synapsin II and synaptophysin PEA3 binding sites (lanes 5-8). An unrelated DNA probe did not compete with this complex. In the negative con-FIG. 5. Zif268/egr-1 binds to the synapsin II promoter in vitro and transactivates a reporter gene containing zif268/egr-1 binding sites derived from the synapsin II promoter in its upstream regulatory region. A, sequence of zif268/egr-1 consensus sites in the upstream regions of the synapsin I and synapsin II gene; B, electrophoretic mobility shift assay using nuclear extract derived from human 293 cells transfected with expression vectors for zif268/egr-1 (lanes 2-11 and 14 -23) or PEA3 (lanes 12 and 24). The radiolabeled probes were the zif268/egr-1 site of the synapsin I promoter, termed EBS-2 (left panel, lanes 1-12), and the putative zif268/egr-1 site of the synapsin II promoter (Ϫ174 to Ϫ153), termed EBS-3 (right panel, lanes 13-24). Binding experiments without extract are depicted in lanes 1 and 13. Competitor DNA was added to the reaction at 10-fold (lanes 3, 6, 9, 15, 18, and 21), 100-fold (lanes 4, 7, 10, 16, 19, and 22), and 500-fold (lanes 5, 8, 11, 17, 20, and 23) molar excess to the probe. Synthetic oligonucleotides containing the PEA3 binding motif (denoted "PEA3") were used as an unrelated competitor (lanes 9 -11 and 21-23). The arrowheads indicate the protein-DNA complexes originating from zif268/egr-1 binding to its cognate binding site. C, reporter plasmids containing rabbit ␤-globin as reporter gene, a TATA box (plasmid OVEC), and 2 copies of the zif268/egr-1 binding site of the synapsin I promoter (EBS-2 2 /OVEC) and synapsin II promoter (EBS-3 2 /OVEC). The zif268/egr-1 cDNA is expressed under control of the cytomegalovirus IE gene promoter (pCMVzif). D, reporter plasmids, expression plasmids, and the JH514ref internal standard plasmid were introduced into CHO-KI cells. Cytoplasmic RNA was analyzed for ␤-globin mRNA by RNase protection mapping. The bands labeled test indicate correctly initiated ␤-globin transcripts, and the bands labeled ref were generated by the internal standard plasmid JH514ref. RT indicates incorrectly initiated read-through transcripts of the test templates. Also shown is an aliquot of undigested cRNA (riboprobe). Size markers, BstEII cut DNA (left) and HaeIII-digested pBR322, are shown in lanes M. trol experiments no protein was added to the probe. No retarded DNA-protein complex was observed (Fig. 7B, lanes 1,  11, and 21). We conclude from these data that PEA3 interacts specifically with both synapsin II and synaptophysin promoters.
The members of the family of ets-transcription factors bind to the core sequence 5Ј-(A/C)GGAA-3Ј. The flanking sequences are variable and there is evidence that these sequences are necessary to determine which ets protein will bind (Wasylyk et al., 1992(Wasylyk et al., , 1993. We performed DNase I footprinting analysis to identify the sequence of the synapsin II promoter that is pro-tected by PEA3 from DNase I digestion. Fig. 7C shows that the GST-PEA3 protein protects the synapsin II promoter sequence Ϫ628 to Ϫ647 from DNase I digestion, including the core sequence 5Ј-AGGAAG-3Ј. No nuclease protection was observed by adding no protein (control) or GST protein. These data confirm the electrophoretic mobility shift analysis that PEA3 binds specifically to the synapsin II promoter through sequences containing the PEA3 binding motif.
We next asked if PEA3 functions as a transcriptional activator following binding to the synapsin II or the synaptophysin promoter. Using the CAT gene as a reporter, four binding sites for PEA3 from the synapsin II promoter or the synaptophysin promoter were cloned immediately upstream of a minimal promoter, thus generating plasmids SyIIPEA3 4 CAT and p38PEA3 4 CAT, respectively (Fig. 8A). As a negative control, four copies of the zif268/egr-1 site from the synapsin I promoter were inserted into the same vector (plasmid EBS-2 4 CAT). These plasmids were transfected in CHO-KI cells together with the expression vector pCMV5 or an expression vector encoding PEA3. Also transfected in all experiments was a plasmid containing the ␤-galactosidase gene under control of the CMV promoter/enhancer to correct for differences in transfection efficiencies. The results of this experiment are depicted in Fig.  8B. Overexpression of PEA3 increased transcription of the CAT gene 7.5-and 15-fold, respectively, when PEA3 binding sites derived from the synapsin II or the synaptophysin gene promoter were present in the regulatory region. There was no increase in transcription when zif268/egr-1 binding sites were inserted upstream of the CAT gene. These data indicate that PEA3 functions as a transcriptional activator after binding to the synapsin II or synaptophysin promoter.

FIG. 8. Transactivation of synapsin II promoter-CAT and synaptophysin promoter-CAT constructs in CHO-KI cells by PEA3.
A, reporter plasmids containing the E. coli CAT gene as reporter, a minimal promoter consisting of TATA box and an Inr element, and 4 copies of the PEA3 binding site of the synapsin II promoter (SyIIPEA3 4 CAT) or synaptophysin promoter (p38PEA3 4 CAT). As a control, 4 copies of the zif268/egr-1 binding site of the synapsin I promoter were inserted upstream of the TATA box (plasmid EBS-2 4 CAT). PEA3 cDNA is expressed under control of the CMV promoter (pCMVPEA3). B, relative CAT activities obtained after transfection of reporter and expression constructs into CHO-KI cells. Cells were cotransfected with plasmid pCMV␤ to correct for variations in transfection efficiencies. CAT activity was normalized by dividing CAT activity by ␤-galactosidase activity. "Ϫ" denotes the transfections with the "empty" expression vector pCMV5, whereas "ϩ" indicates transfections with the PEA3 expression vector pCMVPEA3. At least four experiments were done and the mean Ϯ S.E. is depicted.
fore decided to test whether these DNA motifs might interact with AP2. Gel mobility shift assays were performed with double-stranded, synthetic oligonucleotides corresponding to the proximal as well as the distal putative AP2 site of the synapsin II promoter. As a control the AP2 site derived from the human metallothionein IIA (hMT-II A ) promoter (Mitchell et al., 1987) was used. With recombinant AP2, a major DNA-protein complex was detected using the hMT-II A probe (Fig. 9B, lane 2). The specificity of this complex was revealed by competition with unlabeled DNA (lanes 3 and 4). This complex was also sensitive to a 100-fold excess of the distal AP2 binding site of the synapsin II promoter (lane 6). In contrast, the proximal putative AP2 site and an unrelated DNA probe encompassing a zif268/egr-1 site failed to compete with this complex (lanes 7-10). Fig. 9B shows in the middle panel the results obtained with the distal AP2 site of the synapsin II promoter. A major DNA-protein complex was observed (lane 12) that was competed efficiently by an excess of unlabeled probe (lanes 13 and 14) as well as by a DNA probe containing the AP2 binding site of the metallothionein promoter (lanes 15 and 16). Again, DNA containing the proximal putative AP2 site of the synapsin II promoter as well as the control oligonucleotides did not compete (lanes 17-20). Finally, the proximal putative AP2 site was analyzed (Fig. 9B, right panel). No shift was observed. These results are in agreement with the previous observations that this probe was unable to compete with the binding of AP2 to its cognate site. We conclude that AP2 interacts only with the distal site of the synapsin II promoter. DISCUSSION We have analyzed a genomic clone containing the human synapsin II 5Ј-flanking region. A comparison with the genes encoding the synaptic vesicle proteins synapsin I , synaptophysin (Ö zcelik et al., 1990), and synaptobrevin II  revealed that the promoter regions of these genes are very GC-rich and lack TATA as well as CAAT boxes. We also observed that there is no extensive promoter sequence homology between the human genes encoding synapsin I and synapsin II. This lack of homology has also been reported for the mouse synapsin promoters (Chin et al., 1994). Transcription factors interact in general with short consensus sequences, thus a lack in overall homology between the regulatory region of the synapsin genes does not necessarily implicate different regulatory mechanisms.
Neuron-specific control elements have been identified in the proximal region of the human, and rat synapsin I promoter (Sauerwald et al., 1990;Thiel et al., 1991). Recently, an analysis of the mouse synapsin II promoter showed that a reporter gene under the control of 5 kilobase pairs of the 5Ј-upstream region is highly transcribed in the neuronal cell lines PC12, NS20Y, NS26, and NG108 -15, but not in the non-neuronal cells HeLa, L6, 3T3, or HepG2 (Chin et al., 1994). In addition, transfection experiments of deletion constructs into HeLa and PC12 cells implicated a neuron-specific core promoter between Ϫ79 and ϩ153 of the mouse synapsin II gene (Chin et al., 1994). Our own transfection experiment in PC12 and HeLa cells using the 5Ј-proximal region of the human synapsin II promoter yielded results similar to those of Chin et al. (1994). There was, however, a striking lack of correlation between the endogenous synapsin II gene expression in PC12 cells and the high transcriptional activity of a transfected synapsin II promoter-CAT construct. We had noted this phenomenon before as problematic in respect to synapsin I gene expression as measured in transfected PC12 cells (Thiel et al., 1991). We therefore found it necessary to test other neuronal cell lines as model systems for measuring synapsin II gene expression. The data we obtained from these neuronal cell lines, along with the non-neuronal cell lines tested, indicated that the 5Ј-proximal region of the synapsin II promoter does not confer tissue specificity of transcription. Thus, we believe that, as was the case with synapsin I promoter transfection experiments (Thiel et al., 1991), the use of PC12 cells as the sole neuronal model system to define the neuron-specificity of a promoter sequence is ill advised since the high level of transcriptional activity from the transfected synapsins I and II promoter-CAT constructs in these cells, 1) show no correlation with endogenous synapsin levels in these cells and 2) is at variance with the other neuronal cell lines tested.
The synapsin II gene promoter contains binding sites for the transcription factors zif268/egr-1, PEA3, and AP2, transcription factors that have been suggested to be involved in connecting extracellular signals to gene expression. A cartoon depicting the binding sites of these proteins in the synapsin II promoter is shown in Fig. 10.
FIG. 10. Landmarks of the synapsin II promoter. The cartoon summarizes the protein binding sites of the synapsin II promoter, including the binding of zif268/egr-1, PEA3, and AP2. In addition, the major start site of transcription is depicted. induced by high-frequency stimulation and this induction is inhibited by N-methyl-D-aspartic acid-receptor antagonists (Cole et al., 1989;Wisden et al., 1990). These findings suggest a potential role for zif268/egr-1 in synaptic plasticity. Recently we presented data indicating that the synapsin I gene is a target for zif268/egr-1 . Here, our results suggest that zif268/egr-1 might also regulate synapsin II gene expression. LTP is known to cause an increase in synaptic contact area as revealed by morphological studies (Desmond and Levy, 1988;Schuster et al., 1990;Geinisman et al., 1991). In addition, LTP-dependent changes in the number of vesicles in the synapse have been reported (Applegate et al., 1987;Meshul and Hopkins, 1990). It seems therefore likely that the concentrations of synaptic vesicle proteins are subject to regulation depending on the actual "need" of a particular neuron. In support of this view, Lynch et al. (1994) showed that LTP induced an increase of synapsin I, synaptotagmin, and synaptophysin immunoreactivity in the dentate gyrus. Thus, we propose a signaling cascade where induction of LTP is accompanied by an increase in zif268/egr-1 mRNA and protein (Abraham et al., 1993). The zif268/egr-1 protein translocates to the nucleus and activates synapsin I gene expression that can be monitored 3 h later by immunoblotting with specific antibodies to synapsin I. This cascade can be blocked at early stage by administration of an N-methyl-D-aspartic acid-receptor antagonist prior to induction of LTP. It will be of interest to determine whether synapsin II immunoreactivity also increases following LTP.
PEA3 was recently cloned and analyzed for tissue-specific expression. Detectable amounts of PEA3 mRNA were found only in brain and epididymis (Xin et al., 1992). Here, using DNA-protein binding experiments and transfection analysis, it was demonstrated that PEA3 binds to the synapsin II promoter and transactivates a reporter gene that has synapsin II promoter sequences in its regulatory region. Similar results were obtained for a PEA3-binding site in the synaptophysin promoter. It has been published that the DNA binding and transcriptional activity of PEA3 is influenced by serum and tumor promoters as well as by the gene products of several nonnuclear oncogenes including Ha-ras, polyoma middle T-antigen, v-raf, and v-src (Wasylyk et al., 1989;Gutman and Wasylyk, 1990;Bruder et al., 1992). Interestingly, many of these proteins are involved in a signal transduction cascade that activates mitogen-activated protein kinase. Upon activation this kinase is translocated to the nucleus where it activates transcription by phosphorylating transcription factors such as p62 TCF , c-myc, c-jun, c-fos, NF-IL6, and ATF-2 (for review, see Davis (1993)). PEA3 was shown to be a substrate of mitogen-activated protein kinase 3 in vitro suggesting that phosphorylation of PEA3 by mitogen-activated protein kinase might result from the various stimuli known to activate PEA3. Taken together we postulate that synapsin II and synaptophysin gene expression are regulated by PEA3 through extracellular signal molecules via the mitogen-activated protein kinase pathway.
Finally, we showed in DNA-protein binding experiments that AP2 interacts in a sequence-specific manner with the synapsin II 5Ј-flanking region. The sequence encompassing the synapsin II promoter from Ϫ238 to Ϫ218 contains two AP2 consensus sites. These sites, however, are very close together making it unlikely that both are occupied by AP2 molecules. It has been proposed that AP2 mediates the effect of protein kinase C and cAMP upon transcription (Imagawa et al., 1987;Hyman et al., 1989). This effect is not executed by an increase in AP2 mRNA but rather by post-translational modifications of AP2 (Lü scher et al., 1989). It will be of interest to determine whether cAMP or tumor promoters increase synapsin II transcription via AP2.
The findings that zif268/egr-1, PEA3, and AP2 bind to the synapsin II promoter and activate transcription implicates the possibility that various signaling pathways may play a role in synapsin II gene expression and indicates possible mechanisms for which future studies should be directed.