Genomic Organization and Identification of an Enhancer Element Containing Binding Sites for Multiple Proteins in Rat Pituitary Tumor-transforming Gene*

The rat pituitary tumor transforming gene ( PTTG ) genomic structure was characterized in this study. Northern blot analysis showed that PTTG mRNA is highly expressed in testicular cell lines. Transfection of testicular cell lines with fusion constructs containing various portions of PTTG 5 * -flanking sequences linked to luciferase showed that at least 745 base-pair (bp(s)) 5 * -flanking sequences are required for PTTG transcriptional activation. DNaseI footprinting assays indicated that nuclear protein(s) from testicular cell lines interacts with PTTG 5 * -flanking sequence between 2 509 and 2 624 bp, including two consensus Sp1 binding sites. Western and Southwestern blot analysis showed that three nuclear proteins in addition to Sp1 protein specifically interact with this DNA sequence and that two of these proteins are testicular cell-specific. Deletion of this 115-bp sequence from PTTG promoter resulted in complete loss of promoter function. Site-directed mutagenesis within the Sp1 consensus sequences indicated that the Sp1 binding sites are not critical components of the enhancer sequence for PTTG trancriptional activation in testicular cell lines. Finally, the 115-bp enhancer sequence was shown to be able to activate transcription from a heterologous promoter. These results suggest that PTTG transcriptional activation in testicular cell lines involves interactions of multiple nuclear factors with the PTTG 5 * enhancer sequence. from

The rat pituitary tumor transforming gene (PTTG) genomic structure was characterized in this study. Northern blot analysis showed that PTTG mRNA is highly expressed in testicular cell lines. Transfection of testicular cell lines with fusion constructs containing various portions of PTTG 5-flanking sequences linked to luciferase showed that at least 745 base-pair (bp(s)) 5-flanking sequences are required for PTTG transcriptional activation. DNaseI footprinting assays indicated that nuclear protein(s) from testicular cell lines interacts with PTTG 5-flanking sequence between ؊509 and ؊624 bp, including two consensus Sp1 binding sites. Western and Southwestern blot analysis showed that three nuclear proteins in addition to Sp1 protein specifically interact with this DNA sequence and that two of these proteins are testicular cell-specific. Deletion of this 115-bp sequence from PTTG promoter resulted in complete loss of promoter function. Site-directed mutagenesis within the Sp1 consensus sequences indicated that the Sp1 binding sites are not critical components of the enhancer sequence for PTTG trancriptional activation in testicular cell lines. Finally, the 115-bp enhancer sequence was shown to be able to activate transcription from a heterologous promoter. These results suggest that PTTG transcriptional activation in testicular cell lines involves interactions of multiple nuclear factors with the PTTG 5 enhancer sequence.
Recently, we have isolated and characterized a pituitary tumor transforming gene (PTTG) 1 from rat pituitary tumor cell lines (1). PTTG encodes a novel protein of 199 amino acids and does not contain any known functional motifs. Overexpression of PTTG protein in 3T3 fibroblasts resulted in cell transformation in vitro, and injection of transfected 3T3 cells into nude mice resulted in tumor formation, indicating that PTTG is a transforming gene (1). In addition to pituitary tumor cell lines, rat PTTG mRNA is also expressed in a variety of tumor cell lines, including lung carcinoma, melanoma, leukemia, lym-phoma, and HeLa cell lines. 2 However, among adult rat tissues, PTTG mRNA is only expressed in testis (1), suggesting that PTTG may play a role in testicular biological functions. PTTG transcript in testis is shorter than that of tumor cells, indicating that PTTG mRNA is either differentially spliced or it is using an alternate promoter or an alternative polyadenylation signal. The pattern of expression exhibited by PTTG is similar to proto-oncogene c-mos and c-abl. Both also exhibit testisspecific transcripts (2).
To begin to understand the molecular mechanisms that regulate PTTG expression in testis and tumor cells, I have focused on characterizing the rat PTTG gene. I report here the isolation and structural characterization of the entire rat PTTG gene and its expression in different testicular cell lines. By transient transfection of fusion constructs containing various portions of the PTTG 5Ј-flanking region and luciferase gene into testicular cell lines, I have identified a potential enhancer sequence between Ϫ462 and Ϫ745 bp. I show by DNaseI footprint assays that nuclear protein(s) from testicular cells interacts with DNA sequences between Ϫ509 and Ϫ642 bp of PTTG gene. Using Western and Southwestern analysis, I show that four nuclear proteins, including Sp1 protein, specifically interact with this DNA sequence and that two of these proteins are specific to testicular cells. I demonstrate by mutagenesis studies that deletion of this binding site for multiple testicular proteins results in loss of PTTG transcriptional activation, whereas point mutations within the Sp1 binding sites do not reduce transcription remarkably. Finally, I show that the sequence between Ϫ509 and Ϫ624 bp is able to confer transcriptional activation to a heterologous promoter. These results indicate that the sequence between Ϫ509 and Ϫ624 bp constitutes the core enhancer that is responsible for PTTG transcriptional activation in testicular cell lines.

MATERIALS AND METHODS
Isolation of the Rat PTTG Gene-A rat genomic library (using genomic DNA from Sprague-Dawley rat testis) in DASH vector (Stratagene) was screened using the rat PTTG cDNA (1) as a probe. Southern blot analysis identified two SstI fragments that contain the entire PTTG gene and were subcloned into PGEM3Z (Promega) for further characterization. Dideoxy-DNA sequencing was performed using the Sequenase kit (U. S. Biochemical Corp.). Both strands of DNA were sequenced using either internal primers or sequencing subcloned fragments. DNA restriction and modification enzymes were purchased from Life Technologies, Inc.
RNase Protection Assays and Northern Blot Analysis-Total RNA was isolated from rat testis (Leydig, germ, and Sertoli) and pituitary GH4 cell lines, using RNAeasy isolation kit (Qiagen), following manufacturer's instructions. Rat testis poly(A) ϩ RNA was purchased from CLONTECH. A 533-bp fragment spinning ϩ72 to Ϫ461 bp was synthesized by PCR and subcloned into TA vector (Invitrogen). The antisense probe was synthesized from T7 promoter using the MAXIscript kit (Ambion). RNA protection assays were performed with 20 g of total RNA or 0.5 g of poly(A) ϩ RNA and 10 5 cpm of the probe, using the RPA II ribonuclease protection assay kit (Ambion). A X174 HinfI DNA marker was labeled with [␥-32 P]ATP and T4 DNA polynucleotide kinase. The RNase protection assay products were analysis on 6% sequencing gel alongside the labeled DNA marker and a DNA sequencing ladder.
For Northern blot analysis, 20 g of total RNA from each of the three testis cell lines was fractionated on 1% agarose gel, blotted onto nylon membrane, and UV cross-linked. The membrane was hybridized with full-length PTTG cDNA probe in the Quickhyb solution (Stratagene) for 1 h at the 68°C and washed according to manufacturer's instructions. The membrane was exposed to x-ray film for 16 h.
The PCR products were cloned into TA vector (Invitrogen). A SstI-XhoI fragment was isolated from the plasmid and cloned at the corresponding sites in the pGL3 luciferase reporter vector (Promega).
The PCR product was cloned into TA vector, and the insert was excised by digesting the plasmid with SstI and EcoRV. The insert was clone at the SstI/SmaI site of the thymidine kinase (TK)-luciferase vector.
Cell Cultures and Transfection-Mouse testis germ GC2 and rat Leydig cell line were grown in Dulbecco's modified Eagle's medium or minimal essential medium, respectively, supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 0.11 mg/ml sodium pyruvate, and nonessential amino acids. Cells were transfected with 5 g of fusion construct DNA by calcium phosphate precipitation (3). After a 24-h incubation, cells were assayed for luciferase and CAT activities. Each construct was transfected using triplicate plates for each experiment, and each construct was tested in at least three independent experiments. In all experiments, cells were co-transfected with 1 g of TKCAT (thymidine kinase promoter and chloramphenicol acetyl transferase fusion gene) to monitor transfection efficiency. A negative control plasmid containing a promoterless luciferase gene (pGL3) and a positive control plasmid containing Rous sarcoma virus 3Ј long terminal repeat promoter fused to luciferase were included in all experiments.
Luciferase Assays and CAT Assays-Cell lysates were prepared by freezing and thawing cells in 0.25 M Tris, pH 7.8. Protein concentration was determined using a dye binding assay (Bio-Rad). Luciferase activity were determined by adding 100 l of 1 mM luciferin (Analytical Luminescence Laboratory) to the cell lysate (50 g of protein) in assay buffer (4 mM ATP, 0.25 M Tris, pH 7.8, 15 mM MgSO 4 , 1 g/ml bovine serum albumin). Light emission was integrated over 15 s using a luminometer (Autolumat LB953). For CAT assays, 50 g of cell lysate were incubated with 20 l of 4 mM acetyl-CoA, 1 Ci of [ 14 C]chloramphenicol in a total volume of 100 l at 37°C for 90 min. CAT activity was determined by ascending chromatography on TLC plates and quantified using an Ambis Radioanalytic Imaging System (Ambis).
DNaseI Footprinting Assay-Nuclear extracts were prepared from GC2 cells as described by Dignam et al. (4). The probe used in DNase I footprinting assays is the 366-bp PTTG 5Ј-flanking sequence from Ϫ379 to Ϫ745 bp. To generate the probe, primer 5Ј-CAGTAGATTGGTGC-CTCTGACT-3Ј was end-labeled using [␥-32 P]ATP and T4 polynucleotide kinase and was then used in PCR with downstream primer 5Ј-TCAGCAGCGATTTCGTACTTGGATTTTTTG-3Ј and Ϫ745LUC as the template. 10,000 cpm of the probe was used in each reaction. Binding reactions were performed in 25 l of binding buffer (20 mM Tris, pH 7.5, 1 mM EDTA, 2 mM MgCl 2 , 50 mM NaCl, 20 mM dithiothreitol, and 20% glycerol) with 50 ng of purified Sp1 protein (Promega) or with increasing amounts of GC2 cell nuclear extracts. Competitor DNA was mixed with the probe before receptor addition. Binding was for 30 min at room temperature. DNase I digestion was carried out at room temperature for 1 min using 1-10 units of DNase I (Ambion), depending upon protein concentrations. Reactions were terminated by adding 30 l of 2 ϫ stop buffer containing 15 mM EDTA, 0.2% SDS, and 40 g/ml salmon sperm DNA. Nucleic acids were extracted with phenol/chloroform (1:1), ethanol-precipitated, and electrophoresed on 6% sequencing gels alongside of a GϩA DNA sequencing ladder.
Western and Southwestern Blot Analysis-Cell nuclear extracts (50 g) were separated on a 10% polyacrylamide-SDS gel alongside a protein standards (Amersham Life Science, Inc.). The protein was transferred onto a nitrocellulose membrane. For the Western blot, the membrane was blocked with 5% nonfat milk in Tris-buffered saline-Tween, washed in Tris-buffered saline-Tween, and incubated with a affinity-purified polyclonal antibody corresponding to residues 520 -538 of the human Sp1 protein (Santa Cruz Biotechnology) at 4°C overnight. After incubating with the secondary antibody and washing, the signal was detected using ECL detection system (Amersham).
The double Sp1 mutant was generated using plasmid containing the first Sp1 site mutations as template and primers containing the second Sp1 site mutations in the PCR reaction. All mutants were sequenced to confirm the desired mutations.

RESULTS
Genomic Organization of PTTG Gene-The rat PTTG gene was isolated from a genomic library. The exon-intron boundaries and the size of introns were determined by PCR, using primers derived from PTTG cDNA sequence (see "Materials and Methods") to synthesize DNA from phage DNA. The size of the introns were determined by subtracting exon sequences from each PCR fragment (Fig. 1). The exon-intron boundaries were sequenced, and all the exon/intron splice junctions of PTTG gene showed the characteristic splice site consensus sequence GTAG (Fig. 1). These results together with results from Southern blot analysis and restriction mapping indicated that the PTTG gene contains five exons separated by four introns ranging from 640 bp to 1.8 kilobase pairs. Fig. 2 depicts the genomic structure of the rat PTTG gene.
Characterization of PTTG Gene 5Ј Region-To map the transcription start site of the PTTG gene, a 533-bp fragment containing part of exon 1 was used to generate an antisense probe (Fig. 2). As shown in Fig. 3, hybridization of this probe to RNA derived from rat testis, testis Leydig, and pituitary GH4 cell lines results in a protected band of 71 bp (adjusted to reflect the mobility difference between RNA and DNA) in size, whereas no protected band was apparent when the probe was hybridized to yeast RNA. When the protection assay was performed alongside a sequencing ladder, the transcription start site was mapped to a thymidine residue, 44 bp upstream from the ATG initiation codon (Fig. 4).
Sequencing of about 2 kilobases of DNA sequences upstream from the transcription start revealed a potential TATA box sequence at Ϫ62 bp (Fig. 4). Potential binding sites for nuclear factors Sp1, AP-2, EGR-1, and nuclear receptor half sites were also present in this region (Fig. 4).
PTTG mRNA Expression in Testicular Cell Lines-Because testis is the only adult rat tissue that expresses PTTG mRNA, the expression of PTTG mRNA was examined in cell lines representing its three major cell types. Fig. 5 shows that all three cell lines, including Leydig, Sertoli, and germ (GC2) cells expressed PTTG mRNA to high levels. These cell lines provide a convenient tool to study transcriptional regulation of PTTG in the testicular cell culture system.
Activity of the 5Ј-flanking Region of the PTTG Promoter-To identify the DNA sequences involved in transcriptional regulation of the PTTG gene in testicular cells, various portions of the PTTG 5Ј-flanking region were cloned in front of a luciferase reporter gene and transiently transfected into testis germ GC2 and Leydig cells. Constructs beginning at Ϫ1779, Ϫ1530, Ϫ1315, Ϫ1054, Ϫ745, Ϫ462, and Ϫ194 bp relative to the transcription start site were tested. The downstream boundary of each construct was located at ϩ43 bp relative to the start site of transcription. As shown in Fig. 6, constructs containing up to 462 bp of PTTG upstream sequences were not able to induce luciferase activity significantly over a promoterless control in GC2 cells (germ cells). However, construct containing 745 bp of the PTTG 5Ј-flanking region induced luciferase activity approximately 43-fold over control. The luciferase activity reached maximal (79-fold over control) with the construct containing 1,054 bp of the PTTG 5Ј sequences. Inclusion of more upstream sequences resulted in a slight reduction in luciferase activity (Fig. 6, constructs Ϫ1315LUC, Ϫ1530LUC, and Ϫ1779LUC). Similar results were obtained when these fusion constructs were transfected into Leydig cells (Fig. 6), although the overall induction of luciferase activity was lower than that of the GC2 cells. These results indicate that a minimal 745 bp of the 5Ј-flanking sequences are required for transcriptional activation of PTTG in testicular germ and Leydig cells and that a transcriptional enhancer sequence is present between Ϫ462 and Ϫ745 bp.
GC2 Cell Nuclear Protein(s) Interacts with PTTG 5Ј-flanking Sequences-The above cell transfection data indicates that a potential transcriptional enhancer sequence is present between Ϫ462 and Ϫ745 bp of the PTTG 5Ј-flanking region. To determine whether any testicular cell nuclear proteins specifically interact with the DNA sequences in this region, DNaseI foot- The coding region is darkly shaded, and the 5Ј-and 3Ј-untranslated regions are not shaded. The position for the probe used in RNase protection assay is also indicated.

FIG. 3. Transcription initiation site of the rat PTTG gene. The
RNase protection assay was performed as follows. 32 P-labeled riboprobe was hybridized to 20 g of total RNA derived from rat testis Leydig and pituitary GH4 cell lines or to 500 ng of poly(A) ϩ RNA from adult rat testis. Yeast RNA was used as a negative control. The arrow indicates the protected band. Sizes were determined by comparison with molecular weight markers shown at the right side of the figure.
printing assays were performed. As shown in Fig. 7, panel A (plus strand), a footprint between Ϫ509 and Ϫ624 bp on the plus strand was present when GC2 nuclear extracts was added. The protection from DNaseI became more evident with increasing amounts of GC2 nuclear extracts, and the protection was competed by the unlabeled homologous DNA fragment. When purified Sp1 protein was used in the assay, the footprints were confined to the two consensus Sp1 binding sequences between Ϫ638 and Ϫ624 bp and between Ϫ572 and Ϫ551 bp (Fig. 7,  panel A). Similar results were observed when the minus strand DNA was labeled (Fig. 7, panel B). These results suggest that GC2 cell nuclear protein(s), in addition to Sp1, specifically binds to PTTG 5Ј-flanking region between Ϫ509 and Ϫ624 bp and that the interaction between these proteins and the DNA sequence may result in transcriptional activation of PTTG gene in testis germ cells.
To characterize these proteins further, Western and Southwestern blot analysis were performed using GC2 cell nuclear extracts. As shown in Fig. 8, panel A, anti-human Sp1 antibody detected a single protein of about 95 kDa in GC2 nuclear extracts, corresponding to the Sp1 protein in these cells. When the membrane was incubated with the DNA probe for footprint analysis, four proteins were detected (Fig. 8, panel B, lane 3).
The largest protein (about 95 kDa) corresponds to Sp1 protein; three other proteins of about 35, 46, and 60 kDa were also detected by this probe, and the signal of these proteins was stronger than that of Sp1 (Fig. 8, panel B, lane 3). To determine whether any of these three proteins are testicular cell-specific, nuclear extracts from fibroblast and epithelial cell lines were used in Southwestern blot analysis. As shown in Fig. 8, the probe detected three proteins from these cell lines (Fig. 8, panel  B, lanes 1 and 2). Two of these proteins corresponded to Sp1 and the 35 kDa protein, and the third protein was about 50 kDa. These results indicate that at least four nuclear proteins in the GC2 cell interact with PTTG 5Ј-flanking sequences between Ϫ509 and Ϫ624 bp and that the 46 and 60 kDa proteins are specific to GC2 cells.
Deletion between Ϫ509 and Ϫ624 bp Results in Loss of PTTG Promoter Function-To determine whether the binding site for multiple testicular nuclear proteins is responsible for PTTG transcriptional activation, the DNA sequence between Ϫ509 and Ϫ624 bp was deleted from the fusion construct Ϫ745LUC. As shown in Fig. 9, when transiently transfected into GC2 cells, the luciferase activity of the deletion mutant was reduced to background level, suggesting that the sequence between Ϫ509 and Ϫ624 bp constitutes the core enhancer sequence that is critical for PTTG transcriptional activation in testicular cell lines.
Point   made within the two consensus Sp1 binding sites in the core enhancer sequence (see "Materials and Methods"). Fig. 9 shows that mutations within each individual Sp1 site resulted in a 20 -40% reduction in luciferase activity. When both Sp1 sites were mutagenized, luciferase activity decreased to about 50% that of the wild type level (Fig. 9). However, the luciferase activity of the double Sp1 mutant was still 34-fold over the construct, in which the entire 115-bp core enhancer sequence was deleted. These results suggest that the Sp1 binding sites do not play a crucial role in PTTG transcriptional activation.
The PTTG Enhancer Is Capable of Transcriptional Activation of a Heterologous Promoter-To determine whether the 115-bp sequence in the PTTG 5Ј-flanking region is able to confer transcriptional activation to a heterologous promoter, this sequence was cloned in front of a minimal TK promoter linked to luciferase reporter gene (TKLUC) in both orienta- tions. When transfected into GC2 cell, the enhancer sequence in either orientation induced a 10-fold increase in TK transcription (Fig. 10), suggesting that the PTTG enhancer was able to confer transcriptional activation to a heterologous promoter. DISCUSSION The structural organization of the entire rat PTTG gene was characterized in this study. Previous study showed that PTTG transcript in testis is different in size from that of tumor cells (1). In this study, the same transcription start site was observed in both normal testis and in testicular and pituitary cell lines, suggesting that the different transcript size in testis is not a result of using alternative promoters. Whether the different transcript size is a result of differential splicing or using an alternative polyadenylation site needs to be clarified in future studies. The testis commonly gives rise to transcripts that are differentially processed or are derived from alternate promoters, compared with somatic tissues (5)(6)(7)(8)(9)(10)(11)(12)(13). For example, alternative transcriptional initiation gives rise to a different transcript encoding proopiomelanocortin in testis, whereas alternative splicing is responsible for the generation of testisspecific transcripts of prodynorphin and proenkephalin (13).
The original PTTG cDNA was isolated from a rat pituitary somatotroph tumor cell line (1). Surprisingly, PTTG transcript is also present in many other immortalized and malignant cell lines I have examined, regardless of cell lineage (e.g. lymphoid, myeloid, mesenchymal, and epithelial). However, in rat adult tissues, PTTG displays a very selective pattern of expression. PTTG transcript is present only in testis but not in most adult tissues we have analyzed (1). This distinct expression pattern of PTTG mRNA is similar to several other genes including proto-oncogenes c-mos and c-abl (2) and PEM, a homeobox gene (14). Testis consists of multiple cell types, and PTTG mRNA is expressed to high levels in both germ (GC2) and supporting Sertoli and Leydig cell lines. Whether this expression pattern represents PTTG mRNA expression in intact testis will be determined by in situ hybridization in future studies.
Transient transfection of testicular cell lines showed that PTTG transcriptional activation requires at least 745 bp of the 5Ј-flanking sequences. Although a TATA box-like sequence is present near the transcription start site, it is not sufficient for transcriptional activation of PTTG gene; sequences further upstream (i.e. 5Ј of Ϫ462 bp) are required. Several transcription activator binding sites including AP-2 (15, 16) and EGR-1 (17,18) are present in this region. Both AP-2 and EGR-1 are inducible enhancers. AP-2 mediates transcriptional activation by cAMP and phorbol esters (16). EGR-1 can be induced by a variety of stimuli and activates transcription of target genes whose products are required for mitogenesis and differentiation (17)(18)(19)(20). Although neither AP-2 nor EGR-1 site seems to be involved in basal promoter activity of PTTG gene, their roles in hormonal or growth factor-induced transactivation will be investigated in the future.
Results from transient transfection studies also indicated that a potent transcriptional enhancer is present between Ϫ462 and Ϫ745 bp of the PTTG. There are two consensus Sp1 binding sites in this region. Sp1 is an ubiquitously expressed transactivator that binds to GC box sequence and regulates basal transcription of a variety of housekeeping genes (21). Sp1 also plays a role in directing tissue-specific, hormonal, and developmental regulation of gene expression. Sp1 has been shown to regulate expression of erythroid (22)-, lymphocyte (23)-, and monocyte (24)-specific genes and as a modulator of the retinoic acid/cAMP-dependent transcription of the tissueplasminogen activator gene (25). DNaseI footprinting assays demonstrated that although purified Sp1 protein binds to its consensus binding sequence in this region, the region protected by testicular cell nuclear extracts is much larger, indicating that additional proteins in testicular cells other than Sp1 may interact with the DNA sequence. Western and Southwestern blot analysis showed that in addition to Sp1 protein, three distinct proteins of 35-, 46-, and 60-kDa also interacted with this DNA sequence and that the 46-and 60-kDa proteins were testicular cell-specific. These results support the observation from DNaseI footprint assays that multiple nuclear proteins interact with PTTG enhancer sequence and that the two testicular cell-specific proteins may play a role in testis-specific expression of the PTTG gene.
Deletion of this multiple protein binding site completely abolished PTTG transcriptional activation in testicular cells, indicating that the DNA sequence between Ϫ509 and Ϫ624 bp constitutes the core enhancer sequence that interacts with multiple proteins to activate PTTG transcription. That this 115-bp sequence is a true enhancer was demonstrated further by its ability to activate transcription from a heterologous promoter in an orientation-independent manner. Regulation of gene expression by Sp1 through its interactions with transcription factors bound to adjacent cis-acting elements has been demonstrated. Sp1 has been found to interact with nuclear factor-kB (26), Ets (27), and steroid receptors (28,29). My results showed that point mutations within each individual or both Sp1 consensus sequences in the enhancer sequence did not have a remarkable effect on PTTG transcription, suggesting that the Sp1 binding sites are not critical for PTTG transcriptional activation; binding sites for other testicular nuclear proteins within the enhancer sequence are required.
In summary, I have isolated and characterized the structure of the rat PTTG gene. I have also identified an enhancer element in the PTTG 5Ј-flanking region that is the binding site for multiple testicular nuclear proteins. The interaction of these proteins with PTTG promoter may activate its transcription in testicular cells.