Characterization of the bidirectional promoter region between the human genes encoding VLCAD and PSD-95

Available online at www.sciencedirect.com R

Genomics 82 (2003) 660 – 668

www.elsevier.com/locate/ygeno

Characterization of the bidirectional promoter region between the human genes encoding VLCAD and PSD-95 Li-Feng Zhang,a Jia-Huan Ding,a,* Bing-Zhi Yang,a Guo-Cheng He,b and Charles Roea a

Kimberly H. Courtwright & Joseph W. Summers Institute of Metabolic Disease, Baylor University Medical Center, Dallas, TX 75226, USA b Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, TX 75390-9052, USA Received 5 February 2003; accepted 23 June 2003

Abstract Bidirectional promoters are widely known among lower organisms but rare in mammals. A shared promoter between the two human genes encoding very long chain acyl-CoA dehydrogenase (VLCAD) and postsynaptic density protein 95 (PSD-95) is an ideal model to investigate bidirectional transcription in mammals. VLCAD associates with the inner mitochondrial membrane and catalyzes the initial step in mitochondrial long-chain fatty acid ␤-oxidation. PSD-95, a component protein of the PSD, plays an essential role in clustering the transmembrane proteins in synaptic membranes. Interestingly, the human genes encoding VLCAD (ACADVL) and PSD-95 (DLG4) are adjacently located in the head-to-head orientation on chromosome 17p. The transcribed regions of the two genes overlap, while the two transcription start sites stand ⬃220 bp apart. To analyze the common transcriptional control region shared by the two genes, we generated serial promoter partial deletion constructs using firefly luciferase as the reporter gene. Our results showed that the essential promoter activity of PSD-95 is carried within an ⬃400-bp region, which covers the entire ⬃270-bp minimal promoter of VLCAD. The results from di-(2-ethylhexyl) phthalate (DEHP)-treated HepG2 cells revealed that the minimal VLCAD promoter is able to up-regulate VLCAD expression in response to DEHP treatment. Site-directed mutagenesis experiments showed that a mutated activator protein 2-binding site markedly reduced the transcriptional activity of both promoters and abolished the minimal VLCAD promoter’s response to DEHP treatment. © 2003 Elsevier Inc. All rights reserved. Keywords: Very long chain acyl-CoA dehydrogenase; PSD-95; Bidirectional promoters; AP-2; Fatty acid ␤-oxidation

The initial step of the mitochondrial fatty acid ␤-oxidation spiral is catalyzed by four acyl-CoA dehydrogenases that have different, but overlapping, substrate chain length specificities. Very long chain acyl-CoA dehydrogenase (VLCAD), one of the four acyl-CoA dehydrogenases, is a homodimer of a 71-kDa polypeptide containing 2 mol FAD/ mol enzyme [1], which associates with the mitochondrial inner membrane and plays a major role in long-chain acylCoA dehydrogenation [1,2]. The other three acyl-CoA dehydrogenases are mitochondrial matrix proteins and homotetramers of approximately 40-kDa polypeptides

* Corresponding author: Kimberly H. Courtwright & Joseph W. Summers Institute of Metabolic Disease, Baylor University Medical Center, 3812 Elm Street, Dallas, TX 75226, USA. Fax: ⫹1-214-820-4853. E-mail address: [email protected] (J.-H. Ding). 0888-7543/$ – see front matter © 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0888-7543(03)00211-8

containing 4 mol FAD/mol enzyme [3,4]. Human VLCAD cDNA encodes a protein of 655 amino acids, including a 40-amino-acid leader peptide [5]. The VLCAD protein presents some homology with the other acyl-CoA dehydrogenases along its amino-terminal region, but it has about 180 amino acid residues of extra polypeptide at the carboxylterminal end, which shares no homology with any other acyl-CoA dehydrogenase. The sequence of the unique extra polypeptide revealed a repetitive pattern with alternating hydrophobic/hydrophilic regions. It is assumed that the unique C-terminal region of VLCAD might mediate the association of the protein with the mitochondrial inner membrane [6]. Discs large homolog 4 (DLG4), named as the fourth human protein identified as showing significant similarity to the Drosophila tumor suppressor DLG (Discs large) [7–9], is a member of the membrane-associated guanylate kinase

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

family of proteins. First purified as a component protein of the PSD (postsynaptic density), a tiny, amorphous structure located beneath the postsynaptic membrane of synapses in the CNS, the protein was originally named PSD-95 or SAP (synapse-associated protein)-90 [10,11]. PSD-95 has no detectable enzyme activity and acts as an adapter protein to mediate the localization and clustering of transmembrane receptors, therefore mediating the assembly of signaling complexes [12–26]. As an adaptor protein, PSD-95 contains multiple protein–protein interaction motifs: three PDZ domains (PSD-95, DLG, ZO-1), an SH3 domain, and a guanylate kinase-like domain. The PDZ domain, first recognized in PSD-95 as a novel protein–protein interaction domain, is now an important protein motif identified in a growing list of proteins. The PDZ domain classically interacts with the C-terminal 4 residues of proteins with the sequence X(S/ T)X(I/L/V). In addition, two PDZ domains can also interact with each other. The N-terminal of PSD-95, which carries the three PDZ domains, is posttranslationally palmitoylated via thioester bonds to specific cysteine residues [27–30]. Palmitoylation of PSD-95 mediates its membrane association [30]. PSD-95 is expressed in multiple tissues [31] and is involved in protein–protein interactions in tissues other than neurons [32]. To study the gene structure of VLCAD, a DNA fragment covering the VLCAD promoter region has been cloned in our laboratory (GenBank Accession No. AF244932). Consistent with the human genome data from GenBank (GenBank Accession No. NT_010747), the sequence of the cloned DNA fragment revealed that the genes for human VLCAD and PSD-95 were adjacently located in head-tohead orientation on chromosome 17p. The transcribed regions of the two genes overlap. The transcription start sites of the gene for VLCAD [33] (HGMW-approved symbol ACADVL) and the gene for PSD-95 [34] (HGMW-approved symbol DLG4) stand ⬃220 bp apart. In this study, we have characterized the transcriptional control region shared by the two genes, using firefly luciferase as the reporter gene. Our study not only provides better understanding of transcriptional regulation of both VLCAD and PSD-95, but also adds one more element to the growing database of the identified bidirectional promoters in mammalian genomes.

Results Structure of the putative core promoters A 1959-bp DNA fragment (Fig. 1) covering from VLCAD exon 3 to PSD-95 exon 2 was cloned. Computer analysis on the 1.9-kb common 5⬘ region shared by VLCAD and PSD-95 using PROSCAN version 1.7 (http://bimas. dcrt.nih.gov/molbio/proscan/function.html) showed two putative minimal promoters (Fig. 1): PVLCAD (nt ⫹184 to

661

⫺66) and PPSD-95 (nt ⫺280 to ⫺31). It should be noted that the two putative minimal promoters had a higher GC content than the rest of the 1.9-kb region. The GC% of PVLCAD, PPSD-95, and the1.9-kb region were 72, 76, and 61%, respectively. No TATA motif was recognized in the entire 1.9-kb region using the software TFSEARCH (http://pdap1.trcrwcp.or.jp/research/db/TESEARCH.html), with the cut-off score higher than 75. Many other putative transcription factor binding sites, including an Sp1 binding site, were recognized by the program (Fig. 1B). Sp1 binds to GC-rich sequences and is responsible for fixing the transcriptional start site at TATA-less promoters [35]. Deletion analyses of the regulatory promoters To identify regions that are important for promoter activity, a series of promoter partial deletion constructs was prepared by cloning various promoter fragments upstream of the firefly luciferase reporter gene in the promoterless pGL3-Basic plasmid (Promega). The promoter–luciferase chimeric plasmids were transiently transfected into human MIA PaCa-2 pancreatic carcinoma cells. The PaCa-2 cell line was chosen because the pancreas tissue is one of the tissues from which both genes are expressed at a relatively high level and a human cell line is easily available. We also carried out the same experiments using the 293 cell line (human kidney) and obtained consistent results (data not shown). As shown in Fig. 2, in the VLCAD direction, 5⬘ deletion between positions ⫹680 and ⫹360 had no significant effects on promoter activities (pVL4, pVL3, and pVL2). Further deletion of a 158-bp fragment (⫹360 to ⫹202) resulted in a significant increase in the promoter activity (pVL1). This indicates that a negative regulatory DNA element lies between positions ⫹360 and ⫹202. To narrow down the boundary of the minimal promoter region further, we made another 38-bp deletion at the 3⬘ end (pVL5). The results showed that the 3⬘ deletion led to an approximately 60% decrease in the promoter activity. These observations suggest that the region ⫹202 to ⫺68 contains the minimal VLCAD promoter activity, while a negative regulatory element might be located between positions ⫹360 and ⫹202. In the direction of PSD-95 (Fig. 2), 5⬘ truncation (⫺520 to ⫺280) had no effects on the promoter activities (pDLG5 and pDLG4). The truncation between positions ⫺280 and ⫺170 led to an approximately fivefold increase in activity (pDLG3). This implies that a negative regulatory element lies within the region ⫺280 to ⫺170. Further 5⬘ truncation (⫺170 to ⫹151) significantly decreased promoter activity (pDLG2 and pDLG1). To narrow down the boundary of the core promoter region further, we made a 95-bp truncation from the 3⬘ end (pDLG6). The results showed that the 3⬘ truncation resulted in an approximately 30% decrease in the promoter activity. Thus, the results suggest that the essential

662

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

Fig. 1. Structure and sequence of the bidirectional transcriptional control region between the human VLCAD and PSD-95 genes. (A) Structure of human VLCAD and PSD-95 genes in the head-to-head orientation on chromosome 17p. The exons of the two genes are presented as filled boxes, with the exons of VLCAD as hatched boxes and the exons of PSD-95 as black boxes. The 5⬘ untranslated regions of the two transcripts are both shown as open boxes, with the bent arrows indicating the two transcription start sites. The two putative minimal promoter regions, located using PROSCAN version 1.7 (http:// bimas.dcrt.nih.gov/molbio/proscan/function.html), are indicated by arrows. (B) Nucleotide sequence of the 1.9-kb region between VLCAD exon 3 and PSD-95 exon 2. The sequences of the coding regions are written in capital letters. Putative transcription factor binding sites, predicted using TFSEARCH (http://pdap1.trcrwcp.or.jp/research/db/TESEARCH.html), are shown with arrows indicating the functional direction. The translation start codon of VLCAD is designated as “⫺1”. The start codons and the putative minimal promoters of VLCAD and PSD-95 are boxed with dashed line and solid line, respectively. The sequence within the square brackets is the actual PSD-95 minimal promoter determined by the experiments. The transcription start sites are indicated by bent arrows. The 15-bp DNA motif, which exists as a single-motif-or-double-motif DNA polymorphism, is also boxed with solid line.

promoter activities of PSD-95 are carried within the region ⫺170 to ⫹297 (Fig. 1B), while the region ⫺280 to ⫺170 contains a negative regulatory element for the PSD-95 gene.

Taken together, these data show that the essential promoter activity of PSD-95 is carried within an ⬃400-bp region, which covers the entire ⬃270-bp minimal promoter of VLCAD (Fig. 2).

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

663

Fig. 2. 5⬘-Deletion analysis of the human VLCAD–PSD-95 bidirectional transcriptional control region in MIA PaCa-2 cells. A series of 5⬘-deleted putative promoter fragments was inserted upstream of the firefly luciferase gene in the promoterless pGL3-basic vector. The promoter–luciferase fusion constructs were transiently transfected into MIA PaCa-2 cells. The pRL-TK plasmid carrying the Renilla luciferase gene was cotransfected as the internal control. The firefly luciferase activities were normalized to the Renilla luciferase activities to correct the transfection efficiency. Data are presented with standard deviation from at least three independent experiments. The length and the position of each promoter insert are shown in scale with the structure of the genes shown at the top. The minimal promoters determined by these data are presented as arrows.

Function analysis of a 15-bp DNA motif A 15-bp DNA motif (Fig. 1B) located within the core promoter regions of both genes was identified as a DNA polymorphism. Among normal persons, some had a single 15-bp motif in the promoter region, while others had double 15-bp motifs lying side by side (data not shown). To study the functional effect of the 15-bp motif on the promoter activity, we compared the reporter gene activities driven by the essential promoter fragments of different polymorphs. As shown in Fig. 3, promoter inserts carrying the double 15-bp motif (pVL-d and pDLG-d) had higher activities than the promoter inserts carrying a single motif (pVL-s, pDLGs). The results indicate that the 15-bp motif has a positive regulatory effect on transcription in both directions. Function analysis of the AP-2 binding motif Deletion analysis has previously shown that an AP-2 binding site positively regulates VLCAD promoter activity [33]. To characterize further the functional role of the AP-2

binding site (Fig. 1B) in two transcriptional directions, we carried out site-directed mutagenesis studies (Fig. 4) in a human hepatic cell line (HepG2). Four point mutations were introduced into the promoter fragment by mismatched PCR primers to destroy the AP-2 binding motif (Table 1). The results showed that the mutated AP-2 binding site (pVLAp2gone and pDLGAp2gone) led to an approximately 60% decrease in promoter activity (Fig. 4) in both directions. Therefore, the AP-2 binding motif positively regulates promoter activities in both directions. Studies in rat have shown that oral intake of di-(2ethylhexyl) phthalate (DEHP), a common environmental toxin in drinking water, which accumulates in liver after oral intake, enhanced VLCAD mRNA content fivefold in liver [6]. In our studies, we treated HepG2 cells with Dulbecco’s modified Eagle medium (DMEM) containing 500 ␮M DEHP for 48 h. The results (Fig. 4) showed that the minimal VLCAD promoter was able to respond positively to the DEHP treatment (pVLnormal ⫹ DEHP). However, the promoter fragment carrying the mutated Ap2-binding site lost the ability to respond to DEHP treatment

664

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668 Table 1 List of primers used for cloning of human VLCAD–PSD-95 bidirectional promoter fragments

Fig. 3. The effect of a 15-bp motif DNA polymorphism on promoter activity. The plasmid constructs carrying the promoter insert with a single 15-bp DNA motif (Fig. 1B) (pVL-s; pDLG-s) and a double 15-bp DNA motif (pVL-d; pDLG-d) were transiently transfected into MIA PaCa-2 cells. The pRL-TK plasmid carrying the Renilla luciferase gene was cotransfected as the internal control. The firefly luciferase activities were normalized to the Renilla luciferase activities to correct the transfection efficiency. Data are presented with standard deviations from at least three independent experiments.

(pVLAp2gone ⫹ DEHP). In the direction of PSD-95, DEHP treatment had no significant effect on the activities of both normal and mutated promoter fragment (pDLGnormal ⫹ DEHP; pDLGAp2gone ⫹ DEHP).

Discussion The transcription start sites of both the VLCAD and the PSD-95 genes have been mapped by two previous studies

Fig. 4. Functional significance of the AP-2 binding motif. The luciferase reporter constructs containing normal promoter fragments (pVLnormal and pDLGnormal) and the constructs carrying promoter fragments with mutated AP-2 binding sites (pVLAp2gone and pDLGAp2gone) were transiently transfected into HepG2 cells. Luciferase assay and DEHP treatment were carried out as described under Materials and methods.

Primer

Sequence

pVL⫹ pVL1⫺ pVL2⫺ pVL3⫺ pVL4⫺ pVL5⫹ pDLG⫺ pDLG1⫹ pDLG2⫹ pDLG3⫹ pDLG4⫹ pDLG5⫹ pDLG6⫺ AP2gone⫹ AP2gone⫺ p2k⫹ p2k⫺

5⬘-gaaaagcttACAGACCTTCCGCCCCCG-3⬘ 5⬘-tgtagatctTAAGTCAGCGGAACGCAG-3⬘ 5⬘-tgtagatctCTCCCTTCAGTCCACCTC-3⬘ 5⬘-tgtagatctGTACTCACACCCTAGCTT-3⬘ 5⬘-tgtagatctCAGAGGGCAAAGACAGAA-3⬘ 5⬘-aacaagcttCAAGCTCGCGGCCATCCG-3⬘ 5⬘-tgtaagcttATCCTATCCCATCACCCC-3⬘ 5⬘-accagatctCATGCTGGGAGCTGTAGT-3⬘ 5⬘-accagatctCAAGCTCGCGGCCATCCG-3⬘ 5⬘-agaagatctCGGGGCTGCCCCAGGAGC-3⬘ 5⬘-accagatctTGCCGATGGCGGGCGCCG-3⬘ 5⬘-accagatctAGAGCGTCAGAGGGGTGG-3⬘ 5⬘-aacaagcttTAAGTCAGCGGAACGCAG-3⬘ 5⬘-GACGGTTGCGGGAACCGGAACATGCTGGGA-3⬘ 5⬘-TCCCAGCATGTTCCGGTTCCCGCAACCGTC-3⬘ 5⬘-AGAGCGTCAGAGGGGTGGGAATCT-3⬘ 5⬘-TCGCTATGCAGCACTGTGAGGAGT-3⬘

Every cloning primer carries a 9-bp cloning adapter, which contains either a HindlII or a BglII cleavage site. The adapter is represented by lowercase characters. Primers Ap2gone⫹ and Ap2gone⫺ were used in site-directed mutagenesis. The mutated positions are underlined.

[33,34]. Because of the limited sizes of the cloned 5⬘ ends of both genes, the special bidirectional transcription control region between the two genes remained unrecognized until the human genome sequence was published. Our detailed study on the promoter region of both genes characterized one additional bidirectional promoter in mammalian genomics. Our results determined the boundary of the two minimal promoter regions and showed that the core promoter regions of the two genes were truly overlapped. The actual minimal promoter of VLCAD as illustrated by deletion studies is consistent with the computer prediction of PROSCAN (Fig. 1B), while the actual minimal promoter of PSD-95 is different (Fig. 1B), mainly because of the unusual promoter activity found in the 200-bp DNA fragment downstream of the gene’s transcription start site (Fig. 2). The observation that the 200-bp downstream fragment can function as a promoter by itself (pDLG2, Fig. 2) suggests that the fragment is not just a positive regulatory element but part of the core promoter. The 200-bp region contains the putative Sp1 binding site (Fig. 1B). Corroborating this notion, the presence of an Sp1 binding site within the first 100 nucleotides downstream of the transcriptional start site is a characteristic feature of a TATA-less promoter [36]. A 15-bp DNA motif (Fig. 1B) located within the core promoter regions of both genes was identified as a DNA polymorphism that positively regulates the transcription of both genes. It is possible that a person carrying a double motif has more VLCAD and PSD-95 expressed than a person carrying a single motif. This idea awaits further confirmation that the positive effect is still significant in the

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

665

Table 2 Bidirectional promoters identified in mammalian genomes Gene I

Gene II

Organism

TATA box

GC rich (⬎60%)

Sp1 site

Overlap

Distance

Known relationship

HADHA UFD1L BRCA1 UK114 ACADVL OSGEP H2a Th2a Hsp60 Osgep Adprtl2 Gabpa

HADHB CDC45L NBR2 POP1 DLG4 APEX1 H2b Th2b Hspe1 Apex1 Rmrp5 Atp5j

Human Human Human Human Human Human Rat Rat Rat Mouse Mouse Mouse

No No No No No No Yes Yes No No Yes No

Yes Yes No Yes Yes Yes No No Yes Yes No Yes

Yes Yes Yes Yes Yes Yes No No Yes Yes No Yes

No No No No Yes No No No No No No No

350 bp 312 bp 218 bp 102 bp 220 bp 454 bp 300 bp 300 bp 280 bp 74 bp 113 bp 165 bp

Yes No No No No No Yes Yes Yes No No No

Gene names are shown in abbreviation. Sequences with GC content greater than 60% are considered GC rich. “Distance” indicates the distance between the two transcription start sites.

context of the real genome and at the level of protein expression. The AP-2 binding site (Fig. 1B) located within the core promoter regions of the two genes positively regulates promoter activities in both directions. It is also required for transcriptional up-regulation of VLCAD in response to DEHP treatment. Although the minimal promoter of VLCAD was able to respond positively to DEHP, the minimal promoter of PSD-95 was unaffected by DEHP treatment, which shows that although the two genes share a common cis-acting DNA fragment within the core promoter regions, they have different transcription regulation mechanisms, for example, different combination of trans-acting factors and different cis-acting elements within the nonoverlapping promoter region. The tissue expression profiles of the two RNAs also indicated that, in many tissues, the transcription of the two genes is regulated differently, if not separately [31,37]. The two tissues in which the two genes are both expressed at high levels are heart and skeletal muscle. However, an assay-friendly human cell line is not easily available from these tissues. Whether the transcription of the two genes is regulated coordinately in these tissues is an interesting question for future research. Bidirectional promoters are commonly found in prokaryotes, such as virus and bacteria, because of the limited genome sizes of these small organisms. Mammalian genomes are far more spacious than prokaryotic genomes. The size of a single gene in a mammalian genome is often larger than an entire virus genome. In addition, about 90% of a mammalian genome is occupied by sequences that have no known function. In such a spacious genome, it is unnecessary for two unrelated genes to locate adjacent to each other, sharing a common promoter. Consistent with this notion, bidirectional transcription is not common in mammalian genomes. Table 2 summarizes the characters of 12 bidirectional promoters identified in the genomes of human, mouse, and rat [38 – 48]. The names of the genes are shown in abbreviation in Table 2. Table 3 contains the full name or the function of each gene.

There are two possible explanations for the existence of bidirectional transcription in the mammalian genome. First, the bidirectional transcription regions are ancestral sequences that survived in evolution. Corroborating this idea, almost all the bidirectional promoters found in mammalian genomes belong to the TATA-less and GC-rich type of promoters, which is the type of eukaryotic promoter functioning analogous to prokaryotic promoters. Thus, it is likely that the bidirectional promoters found in the mammalian genome are preserved ancestral sequences. Interestingly, the studies on the OSGEP/Osgep-APEX1/Apex1 bidirectional promoter, the bidirectional promoter region studied in detail in both human and mouse, have shown that

Table 3 The full names of the genes presented in Table 2 Abbreviation

Full gene name

Apex1 Atp5j BRCA1 Hspe1 Hsp60 CDC45L HADHA HADHB H2a H2b Gabpa NBR2

Apurinic/apyrimidinic endonuclease 1 ATP synthase coupling factor 6 Breast cancer 1, early onset Heat shock 10-kDa protein 1, chaperonin 10 Heat shock 60-kDa protein, chaperonin 60 Cell division cycle 45-like (Saccharomyces cerevisine) Human trifunctional protein ␣ subunit Human trifunctional protein ␤ subunit Somatic histone 2A Somatic histone 2B GA-binding protein ␣ subunit A novel gene encoding a B-box protein within the BRCA 1 region O-sialoglycoprotein endopeptidase Translational inhibitor p14.5 ADP-ribosy transferase (NAD⫹; poly(ADP-ribose) polymerase)-like 2 Postsynaptic density protein 95 Protein subunit of RNase P RNA subunit of RNase P Testis-specific histone 2A Testis-specific histone 2B Ubiquitin-fusion degradation 1-like Very long chain acyl-CoA dehydrogenase

Osgep UK114 Adprtl2 DLG4 POP1 Rmrp5 Th2a Th2b UFD1L ACADVL

666

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

the distance between the two genes has extended from 74 bp in mouse to 454 bp in human, which indicates that the distance between these two genes was increased during evolution [43,47]. The second possible explanation of the existence of bidirectional transcription in the mammalian genome is that the bidirectional promoter region is shared by two genes that are functionally related to each other. In accordance with this idea, four of the bidirectional promoters in Table 2 are shared by gene pairs with known functional relationship. Interestingly, the functional relationships of all the four gene pairs are similar. The two genes are either two different peptide subunits of a same protein or two different protein members of the same protein family. The genes encoding the two subunits, ␣ subunit (HADHA) and ␤ subunit (HADHB), of the human trifunctional protein (TFP) are controlled by such a bidirectional promoter [44]. Similar to VLCAD, TFP is also a fatty acid ␤-oxidation enzyme controlling long-chain fatty acid substrate. A TFP protein consists of four ␣ and four ␤ subunits [49]. It is reasonable to assume that a bidirectional promoter is involved in this case to make sure the two protein subunits are expressed at the same level in all tissues. VLCAD and PSD-95 obviously do not belong to the same protein family. There is also no known functional relation between the two proteins. It is likely that this unique bidirectional promoter exists through evolution. According to the published genome data, the two genes share similar gene structures at the 5⬘ ends in the mouse genome. Additional work is required to locate the transcription start sites of the two genes in mouse to confirm whether the two genes are also controlled by a bidirectional promoter in the mouse genome.

Materials and methods Cell culture media and conditions HepG2 and MIA PaCa-2 cells were purchased from American Type Culture Collection and grown in DMEM supplemented with 10% fetal bovine serum, 0.5 mM Lglutamine, 50 mg/ml streptomycin, and 100 IU/ml penicillin. All cell lines were maintained at 37°C with 5% CO2. Construction of reporter gene plasmids A 2-kb (nucleotide ⫺520 to ⫹1440) human VLCAD– PSD-95 bidirectional promoter fragment (Fig. 1B) was PCR amplified from human genomic DNA using primers p2k⫹ and p2k⫺ (Table 1). The fragment was then cloned into the pCR2.1-TOPO vector using the TOPO TA cloning kit (Invitrogen). The promoter fragment was sequenced to confirm that the sequence matched the original genomic sequences without PCR-generated errors. The plasmid clone served as PCR template for construction of reporter gene plasmids.

To analyze the overlapping promoter region, we generated serial chimeric plasmids containing various promoter fragments cloned upstream of the firefly luciferase reporter gene in the pGL3-Basic plasmid (Promega). The promoter inserts were PCR amplified using primers carrying a 9-bp cloning adaptor at the 5⬘ end. Depending on the desired orientation, a cloning adaptor contained either a HindIII or a BglII site (Table 1). After enzyme digestion, the PCR product was inserted into the HindIII or BglII site of the pGL3-Basic plasmid using T4 DNA ligase (Promega). The chimeric plasmids, their promoter inserts, and the cloning primers were pVL1 (also named pVLnormal) (⫹202 to ⫺68), pVL⫹ and pVL1⫺; pVL2 (⫹360 to ⫺68), pVL⫹ and pVL2⫺; pVL3 (⫹530 to ⫺68), pVL⫹ and pVL3⫺; pVL4 (⫹680 to ⫺68), pVL⫹ and pVL4⫺; pVL5 (⫹202 to ⫺30), pVL5⫹ and pVL1⫺; pDLG1 (⫹151 to ⫹297), pDLG⫺ and pDLG1⫹; pDLG2 (also named pDLGnormal) (⫺30 to ⫹297), pDLG⫺ and pDLG2⫹; pDLG3 (⫺170 to ⫹297), pDLG⫺ and pDLG3⫹; pDLG4 (⫺280 to ⫹297), pDLG⫺ and pDLG4⫹; pDLG5 (⫺520 to ⫹297), pDLG⫺ and pDLG5⫹; pDLG6 (⫺170 to ⫹202), pDLG6⫺ and pDLG3⫹. The site-directed mutations were introduced into plasmids pVLAp2gone and pDLGAp2gone, using the mismatch PCR primers, AP2gone⫹ and AP2gone⫺. The template plasmids for pVLAp2gone and pDLGAp2gone were pVL1 and pDLG2, respectively. Plasmid constructs carrying the promoter insert with a single 15-bp motif (Fig. 1B) (pVL-s; pDLG-s) and a double 15-bp motif (pVL-d; pDLG-d) were cloned from normal human genomic DNA of the desired genotype. The cloning primers for pVL-s and pVL-d were pVL5⫹ and pVL1⫺. The cloning primers for pDLG-s and pDLG-d were pDLG6⫺ and pDLG2⫹. The sequences of all the PCR primers are presented in Table 1. All the plasmids were sequenced to check the correct insert orientation and to confirm that the sequences matched the GenBank data (GenBank Accession No. NT_010747). Endotoxin-free plasmid DNA was prepared using the EndoFree plasmid maxi kit (Qiagen). Cell transfection and dual luciferase reporter assay The human cell lines HepG2 and MIA PaCa-2 were plated into a 24-well culture plate at a density of 5 ⫻ 104 cells/well and grown overnight to about 60% confluence before transfection. The DNA mixture for transfection of each well was composed of test plasmid (0.2 ␮g) and pRL-TK plasmid (Promega) (0.01 ␮g). pRL-TK plasmid carries a Renilla luciferase gene driven by the herpes simplex virus thymidine kinase promoter and served as an internal control to normalize the effect of transfection efficiencies. Transfections were carried out using an Effectene Transfection Reagent kit (Qiagen). The ratio of Effectene Reagent (␮l) to DNA (␮g) was optimized following the manufacturer’s instructions. Briefly, 0.21 ␮g plasmid DNA, 1.6 ␮l Enhancer, 60 ␮l DNA-condensation buffer, and 5 ␮l

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

Effectene were used in transfection for 1 well of the 24-well culture plate. Transfected cells were lysed 48 h after transfection by applying 150 ␮l Passive Lysis Buffer of the Dual Luciferase Reporter Assay kit (Promega) into each well of the 24-well plate. Twenty microliters of cell lysate was used for the dual luciferase reporter assay with a Dual Luciferase Assay kit (Promega). Light intensity was measured for 10 s with a luminometer (Model OPTOCOMP I; MGM Instruments). The firefly luciferase activity of each test plasmid was normalized to the Renilla luciferase activity of the internal control plasmid. All transfections were performed in triplicate and the lysate from each transfection was measured at least three times. Average relative luciferase activities were determined. DEHP treatment Twenty-four hours after transfection, cell culture media were refreshed with DMEM containing 500 ␮M DEHP and 10% FBS. Seventy-two hours after transfection, cells were lysed. Luciferase assay was carried out as previously described.

Acknowledgments We are grateful to Dr. James L. Matthews for critically reading the manuscript. We thank Mr. Matt Mysliwiec for technical assistance. This work was supported in part by the Kimberly H. Courtwright & Joseph W. Summers Metabolic Disease Fund.

References [1] T. Aoyama, et al., Purification of human very-long-chain acyl-coenzyme A dehydrogenase and characterization of its deficiency in seven patients, J. Clin. Invest. 95 (1995) 2465–2473. [2] K. Izai, Y. Uchida, T. Orii, S. Yamamoto, T. Hashimoto, Novel fatty acid beta-oxidation enzymes in rat liver mitochondria. I. Purification and properties of very-long-chain acyl-coenzyme A dehydrogenase, J. Biol. Chem. 267 (1992) 1027–1033. [3] S. Furuta, S. Miyazawa, T. Hashimoto, Purification and properties of rat liver acyl-CoA dehydrogenases and electron transfer flavoprotein, J. Biochem. (Tokyo) 90 (1981) 1739 –1750. [4] Y. Ikeda, K. Okamura-Ikeda, K. Tanaka, Purification and characterization of short-chain, medium-chain, and long-chain acyl-CoA dehydrogenases from rat liver mitochondria. Isolation of the holo- and apoenzymes and conversion of the apoenzyme to the holoenzyme, J. Biol. Chem. 260 (1985) 1311–1325. [5] T. Aoyama, et al., Cloning of human very-long-chain acyl-coenzyme A dehydrogenase and molecular characterization of its deficiency in two patients, Am. J. Hum. Genet. 57 (1995) 273–283. [6] T. Aoyama, I. Ueno, T. Kamijo, T. Hashimoto, Rat very-long-chain acyl-CoA dehydrogenase, a novel mitochondrial acyl-CoA dehydrogenase gene product, is a rate-limiting enzyme in long-chain fatty acid beta-oxidation system. cDNA and deduced amino acid sequence and distinct specificities of the cDNA-expressed protein, J. Biol. Chem. 269 (1994) 19088 –19094.

667

[7] S.D. Dimitratos, D.F. Woods, D.G. Stathakis, P.J. Bryant, Signaling pathways are focused at specialized regions of the plasma membrane by scaffolding proteins of the MAGUK family, Bioessays 21 (1999) 912–921. [8] A. Fujita, Y. Kurachi, SAP family proteins, Biochem. Biophys. Res. Commun. 269 (2000) 1– 6. [9] D.F. Woods, P.J. Bryant, The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions, Cell 66 (1991) 451– 464. [10] K.O. Cho, C.A. Hunt, M.B. Kennedy, The rat brain postsynaptic density fraction contains a homolog of the Drosophila discs-large tumor suppressor protein, Neuron 9 (1992) 929 –942. [11] U. Kistner, et al., SAP90, a rat presynaptic protein related to the product of the Drosophila tumor suppressor gene dlg-A, J. Biol. Chem. 268 (1993) 4580 – 4583. [12] C.C. Garner, J. Nash, R.L. Huganir, PDZ domains in synapse assembly and signalling, Trends Cell Biol. 10 (2000) 274 –280. [13] T. Hanada, L. Lin, K.G. Chandy, S.S. Oh, A.H. Chishti, Human homologue of the Drosophila discs large tumor suppressor binds to p56lck tyrosine kinase and Shaker type Kv1.3 potassium channel in T lymphocytes, J. Biol. Chem. 272 (1997) 26899 –26904. [14] T. Hanada, L. Lin, E.V. Tibaldi, E.L. Reinherz, A.H. Chishti, GAKIN, a novel kinesin-like protein associates with the human homologue of the Drosophila discs large tumor suppressor in T lymphocytes, J. Biol. Chem. 275 (2000) 28774 –28784. [15] Y.Z. Huang, et al., Regulation of neuregulin signaling by PSD-95 interacting with ErbB4 at CNS synapses, Neuron 26 (2000) 443– 455. [16] H. Husi, M.A. Ward, J.S. Choudhary, W.P. Blackstock, S.G. Grant, Proteomic analysis of NMDA receptor–adhesion protein signaling complexes, Nat. Neurosci. 3 (2000) 661– 669. [17] M. Irie, et al., Binding of neuroligins to PSD-95, Science 277 (1997) 1511–1515. [18] E. Kim, M. Niethammer, A. Rothschild, Y.N. Jan, M. Sheng, Clustering of Shaker-type K⫹ channels by interaction with a family of membrane-associated guanylate kinases, Nature 378 (1995) 85– 88. [19] E. Kim, et al., GKAP, a novel synaptic protein that interacts with the guanylate kinase-like domain of the PSD-95/SAP90 family of channel clustering molecules, J. Cell Biol. 136 (1997) 669 – 678. [20] J.H. Kim, D. Liao, L.F. Lau, R.L. Huganir, SynGAP: a synaptic RasGAP that associates with the PSD-95/SAP90 protein family, Neuron 20 (1998) 683– 691. [21] H.C. Kornau, L.T. Schenker, M.B. Kennedy, P.H. Seeburg, Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95, Science 269 (1995) 1737–1740. [22] M. Niethammer, et al., CRIPT, a novel postsynaptic protein that binds to the third PDZ domain of PSD-95/SAP90, Neuron 20 (1998) 693– 707. [23] M. Passafaro, C. Sala, M. Niethammer, M. Sheng, Microtubule binding by CRIPT and its potential role in the synaptic clustering of PSD-95, Nat. Neurosci. 2 (1999) 1063–1069. [24] A. Savinainen, E.P. Garcia, D. Dorow, J. Marshall, Y.F. Liu, Kainate receptor activation induces mixed lineage kinase-mediated cellular signaling cascades via post-synaptic density protein 95, J. Biol. Chem. 276 (2001) 11382–11386. [25] M. Sheng, C. Sala, PDZ domains and the organization of supramolecular complexes, Annu. Rev. Neurosci. 24 (2001) 1–29. [26] T. Tezuka, H. Umemori, T. Akiyama, S. Nakanishi, T. Yamamoto, PSD-95 promotes Fyn-mediated tyrosine phosphorylation of the Nmethyl-D-aspartate receptor subunit NR2A, Proc. Natl. Acad. Sci. USA 96 (1999) 435– 440. [27] J.T. Dunphy, M.E. Linder, Signalling functions of protein palmitoylation, Biochim. Biophys. Acta 1436 (1998) 245–261. [28] G. Milligan, M. Parenti, A.I. Magee, The dynamic role of palmitoylation in signal transduction, Trends Biochem. Sci. 20 (1995) 181– 187. [29] S.M. Mumby, Reversible palmitoylation of signaling proteins, Curr. Opin. Cell Biol. 9 (1997) 148 –154.

668

L.-F. Zhang et al. / Genomics 82 (2003) 660 – 668

[30] J.R. Topinka, D.S. Bredt, N-terminal palmitoylation of PSD-95 regulates association with cell membranes and interaction with K⫹ channel Kv1.4, Neuron 20 (1998) 125–134. [31] D.G. Stathakis, K.B. Hoover, Z. You, P.J. Bryant, Human postsynaptic density-95 (PSD95): location of the gene (DLG4) and possible function in nonneural as well as in neural tissues, Genomics 44 (1997) 71– 82. [32] M.J. Ludford-Menting, et al., A functional interaction between CD46 and DLG4: a role for DLG4 in epithelial polarization, J. Biol. Chem. 277 (2002) 4477– 4484. [33] Y. Zhou, D.P. Kelly, A.W. Strauss, H. Sims, Z. Zhang, Characterization of the human very-long-chain acyl-CoA dehydrogenase gene promoter region: a role for activator protein 2, Mol. Genet. Metab. 68 (1999) 481– 487. [34] D.G. Stathakis, et al., Genomic organization of human DLG4, the gene encoding postsynaptic density 95, J. Neurochem. 73 (1999) 2250 –2265. [35] A.R. Black, J.D. Black, J. Azizkhan-Clifford, Sp1 and kruppel-like factor family of transcription factors in cell growth regulation and cancer, J. Cell Physiol. 188 (2001) 143–160. [36] L. Lania, B. Majello, P. De Luca, Transcriptional regulation by the Sp family proteins, Int. J. Biochem. Cell Biol. 29 (1997) 1313–1323. [37] B.S. Andresen, et al., Cloning and characterization of human verylong-chain acyl-CoA dehydrogenase cDNA, chromosomal assignment of the gene and identification in four patients of nine different mutations within the VLCAD gene, Hum. Mol. Genet. 5 (1996) 461– 472. [38] J.C. Ame, et al., A bidirectional promoter connects the poly(ADPribose) polymerase 2 (PARP-2) gene to the gene for RNase P RNA structure and expression of the mouse PARP-2 gene. J. Biol. Chem. 276 (2001) 11982–11099. [39] Y. Chinenov, C. Coombs, M.E. Martin, Isolation of a bi-directional promoter directing expression of the mouse GABPalpha and ATP synthase coupling factor 6 genes, Gene 261 (2000) 311–320.

[40] A.N. Domansky, et al., Solitary HERV-K LTRs possess bi-directional promoter activity and contain a negative regulatory element in the U5 region, FEBS Lett. 472 (2000) 191–195. [41] N.E. Huh, I.W. Hwang, K. Lim, K.H. You, C.B. Chae, Presence of a bi-directional S phase-specific transcription regulatory element in the promoter shared by testis-specific TH2A and TH2B histone genes, Nucleic Acids Res. 19 (1991) 93–98. [42] H. Igaki, et al., Characterization of the bi-directional transcriptional control region between the human UFD1L and CDC45L genes, Biochem. Biophys. Res. Commun. 283 (2001) 569 –576. [43] S. Ikeda, H. Ayabe, K. Mori, Y. Seki, S. Seki, Identification of the functional elements in the bidirectional promoter of the mouse Osialoglycoprotein endopeptidase and APEX nuclease genes, Biochem. Biophys. Res. Commun. 296 (2002) 785–791. [44] K.E. Orii, et al., Genes for the human mitochondrial trifunctional protein alpha- and beta-subunits are divergently transcribed from a common promoter region, J. Biol. Chem. 274 (1999) 8077– 8084. [45] M.T. Ryan, et al., The genes encoding mammalian chaperonin 60 and chaperonin 10 are linked head-to-head and share a bidirectional promoter, Gene 196 (1997) 9 –17. [46] G. Schmiedeknecht, C. Buchler, G. Schmitz, A bidirectional promoter connects the p14.5 gene to the gene for RNase P and RNase MRP protein subunit hPOP1, Biochem. Biophys. Res. Commun. 241 (1997) 59 – 67. [47] Y. Seki, et al., Sequencing analysis of a putative human O-sialoglycoprotein endopeptidase gene (OSGEP) and analysis of a bidirectional promoter between the OSGEP and APEX genes, Gene 285 (2002) 101–108. [48] C.F. Xu, J.A. Chambers, E. Solomon, Complex regulation of the BRCA1 gene, J. Biol. Chem. 272 (1997) 20994 –20997. [49] Y. Uchida, K. Izai, T. Orii, T. Hashimoto, Novel fatty acid betaoxidation enzymes in rat liver mitochondria. II. Purification and properties of enoyl-coenzyme A (CoA) hydratase/3-hydroxyacylCoA dehydrogenase/3-ketoacyl-CoA thiolase trifunctional protein, J. Biol. Chem. 267 (1992) 1034 –1041.