Characterization of Proximal Transcription Regulatory Elements in the Rat Phospholamban Promoter

J Mol Cell Cardiol 31, 2137–2153 (1999) Article No. jmcc.1999.1042, available online at http://www.idealibrary.com on

Characterization of Proximal Transcription Regulatory Elements in the Rat Phospholamban Promoter Charles F. McTiernan, Bonnie H. Lemster, Carole S. Frye, David C. Johns∗ and Arthur M. Feldman Cardiovascular Institute, University of Pittsburgh Medical Center, Pittsburgh, PA 15213, USA (Received 21 June 1999, accepted in revised form 14 September 1999) C. F. MT, B. H. L, C. S. F, D. C. J  A. M. F. Characterization of Proximal Transcription Regulatory Elements in the Rat Phospholamban Promoter. Journal of Molecular and Cellular Cardiology (1999) 31, 2137–2153. Phospholamban is a major regulator of cardiac diastole, with alterations in expression associated with modified cardiac relaxation. To study transcriptional regulation of phospholamban expression, we made reporter constructs that expressed luciferase under control of putative promoter sequences from the rat phospholamban gene. When transfected into neonatal rat cardiomyocytes, constructs containing at least 159 nucleotides preceding the transcription start site were equally active, while truncation to −66/+64 removed all promoter activity. Constructs were more active in cardiomyocytes than in HeLa cells (which do not express phospholamban), but did not show absolute cell-type specificity of expression. Addition of sequences upstream to −4032, all of the intron (7.4 kb), or 3′UTR sequences (0.8 kb) did not enhance cell-specific expression. To focus on the basal promoter region (−159/−66), a series of deletion constructs were made that identified a novel 35 bp region (−159/−125; Phospholamban Promoter Element 1, PPE1) required for promoter activity in cardiomyocytes. Site-specific mutations identified nucleotides −150/−133 as containing most of the promoterenhancing activity. While the rat PPE1 is highly conserved (>70%) in four other mammalian phospholamban genes, it does not contain previously characterized regulatory elements. In cardiomyocytes the PPE1 sequence markedly enhanced activity of the SV40 early promoter. A conserved CCAAT element (−83/−79) was also required for promoter activity in both cardiomyocytes and HeLa cells. Exonuclease III footprinting identified protein/DNA interactions in both the extended CCAAT box and PPE1 domains. Gel shift studies identified the  1999 Academic Press CCAAT elements as binding CBF/NF-Y. K W: Phospholamban; Rat; Transcriptional regulation; Promoter; Gene expression.

Introduction In the mammalian heart, removal of calcium from the cytosol into the sarcoplasmic reticulum (SR) is critical to the regulation of diastolic function (Morgan et al., 1990). This process involves the coordinate activities of two SR proteins, sarcoplasmic reticulum calcium ATPase (SERCA) and phospholamban. SERCA serves as a calcium pump, while phospholamban regulates calcium movement by

decreasing the affinity of SERCA for calcium, thus reducing the rate of calcium uptake into the SR (Hicks et al., 1979). Following b-adrenergic stimulation, phospholamban is phosphorylated via cAMP-dependent protein kinases, which reverses inhibition of the SR calcium pump (Lindemann et al., 1983). A direct demonstration of this pathway is seen in phospholamban-deficient mice generated by targeted gene ablation. These mice demonstrate increased affinity of SERCA for calcium, display

∗ Present address: Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA. Please address all correspondence to: Charles F. McTiernan, Cardiovascular Institute, University of Pittsburgh, Biomedical Science Tower 1744.1, 200 Lothrop St., Pittsburgh, PA 15213, USA.

0022–2828/99/122137+17 $30.00/0

 1999 Academic Press

2138

C. F. McTiernan et al.

enhanced myocardial contractile parameters, and a loss of responsiveness to b-adrenergic stimulation (Luo et al., 1994). Phospholamban is encoded by a single gene and is expressed in three different muscle lineages: cardiac, slow skeletal, and smooth muscle. The level of expression varies greatly, with a high level of expression in the ventricular myocardium and less expression in the atrium, aorta, slow skeletal muscle, and various smooth muscles (Fujii et al., 1991; McTiernan et al., 1999). Within the developing myocardium, phospholamban displays a spatial and temporal order of expression distinct from that of SERCA, implying different regulatory mechanisms (Moorman et al., 1995). Besides tissuespecific differences, phospholamban expression is modified in various physiologic states associated with alterations of contactile function, with expression regulated by T3 (Arai et al., 1991), interleukin 1b (McTiernan et al., 1997), skeletal muscle contractile frequency (Leberer et al., 1989), and differentiation state of smooth muscle cells (Shanahan et al., 1993). In addition, the pathologic states of pressure overload-induced hypertrophy and heart failure are associated with decreased cardiac phospholamban transcript and protein levels in both animal models of hypertrophy and heart failure (Rockman et al., 1994; Kiss et al., 1995) and in failing human hearts (Feldman et al., 1991; Arai et al., 1993), although these observations remain controversial (Meyer et al., 1995; Schwinger et al., 1995). Regulation of phospholamban expression may be directed largely by promoter activity. While splice variants within the 3′ untranslated region (UTR) of phospholamban transcripts exist (Fujii et al., 1991; Toyofuku and Zak, 1991), they do not encode different phospholamban isoforms. Although it is technically possible that different transcript isoforms are translated with different efficiencies, there is currently no evidence that this contributes to the regulation of phospholamban expression. Thus, characterization of the phospholamban promoter may reveal important regulatory mechanisms for this gene. Promoter sequences from chicken (Toyofuku and Zak, 1991), rabbit (Fujii et al., 1991), rat (Johns and Feldman, 1992), mouse (Haghighi et al., 1997), and human (McTiernan et al., 1999) phospholamban genes have been presented, with a partial functional dissection of the mouse and rabbit phospholamban promoter reported (Haghighi et al., 1997; Yabuki et al., 1998). Our previous studies identified sequences for putative basal regulatory elements (TATAA and CCAAT boxes), as well as the transcription start site of the rat (Johns and

Feldman, 1992) and human phospholamban genes (McTiernan et al., 1999). In this report, we have prepared and examined various rat phospholamban promoter constructs that direct expression of luciferase when transfected into neonatal cardiomyocytes in order to identify the cis-acting elements required for phospholamban gene expression.

Materials and Methods Reporter expression constructs The organization of the rat phospholamban gene is presented in Figure 1(A). Nucleotide sequence of the rat phospholamban gene can be found in the GenBank Nucleotide Sequence Database (Accession numbers L03381 and L03382). Restriction digestion fragments from plasmids containing rat phospholamban genomic sequences (Johns and Feldman, 1992) were cloned into the promoterless luciferase construct pGL2-Basic (Promega, Madison, WI, USA) to generate a series of plasmids (such as −159/+64), identified by their 5′ and 3′ phospholamban nucleotides, designated as + or − relative to the transcription start site. Constructs with the 5′ end between −159 and −92 and the downstream end fixed at +64 were created by PCR amplification from plasmid template −1212/+64 using primers containing artifical restriction site sequences and the desired region of the rat phospholamban promoter. PCR products were restriction digested and cloned into pGL2-Basic. Plasmids that contained nucleotides −159 to −142, −92 to +64, and variable lengths between −141 and −93 were made with a four primer overlap extension PCR amplification strategy (Horton et al., 1989), using plasmid −159/+64 as a template to create plasmid −159/+64C93, which contains a unique NcoI site arising from a G to C base change at −93. Variable length sequences between −159 and −93 were PCR amplified, digested with Kpn and NcoI and cloned into KpnI/NcoI digested −159/+64C93. A fortuitous cloning error generated plasmid −159/+64C93G81 in which the CCAAT box A at position −81 was replaced with G. A similar PCR-based strategy was used to create a deletion construct (−1212d159/126) containing sequences −1212 to +64 with nucleotides −159 to −126 replaced with a 6 bp oligomer containing a KpnI site. A substitution construct (−1212/ scram159/125) was made by inserting a double stranded oligomer (scram159/125, sense strand 5′CATGGTTAGAAATTCTAAGCCAT-GGTAC-3′) into

Rat Phospholamban Promoter Elements

2139

Figure 1 Activity of rat phospholamban promoter in transiently transfected cells. (A) Organization of the rat phospholamban gene. Boxed areas indicate exons. Hatched areas identify 5′UTR. CDS depicts coding region. Unshaded area indicates 3′UTR. Size of regions or distances between features are indicated below map in base pair (bp) or kilobase pair (kb). Location of CCAAT and TATAA boxes are indicated. Plasmids containing phospholamban promoter regions serially deleted at the 5′ end and fused to the luciferase coding sequence were transfected into neonatal rat cardiomyocytes (B) or HeLa cells (C). Luciferase activity was normalized to the activity of co-transfected internal control reporters, and then normalized to the level of activity for construct −159/+64 (arbitrarily set at 1.0). Data is presented as mean±standard error of the mean (...). Data was collected from at least four independent transfection experiments containing two to six (typically three) replicates of each construct.

the KpnI site of −1212d159/126, regenerating the original spacing and GC content between nucleotides −160 and −126. Using similar PCR amplification and sub-cloning techniques, construct PLB Mini Intron was made which contained nucleotide −1212 through −1, all of exon 1 (+1 through +87) which contains 5′UTR sequences, 240 nucleotides at the 5′ end of the 7.4 kb intron (Johns and Feldman, 1992) joined to 139 nucleotides at the 3′ end of the intron, and the noncoding sequences of the second exon which contains

5′UTR (109 bp) and 3′UTR (818 bp) sequences. In addition the phospholamban coding region was replaced with the firefly luciferase coding sequence amplified by PCR from pGl2-Basic using the primers luc sense (5′-GCTGCCATGGCA-TGACCGGTACTGTTGGTAAAAATG-3′) and luc antisense (5′-GCTGCCATGGTTACAATT-TGGACTTTCCGCC-3′). A derivative of PLB Mini Intron containing all of the 7.4 kb rat phospholamban intron (PLB Intron), as well one containing all of the intron and 4032 bp upstream of the transcription start site (PLB Intron

2140

C. F. McTiernan et al.

Plus) were created by isolation of various restriction fragments from the genomic clone and ligation into the plasmid PLB Mini Intron. Enhancer activity of the −159/−125 region was tested by cloning a 35 bp oligonucleotide (−159/ −125, sense strand 5′-GATCTGAAGCACAACATGTTACCGACATGTCCAAACCTA-3′) into BglII or BamH1 digested pGL2-Promoter (Promega). pGL2Promoter expresses firefly luciferase under control of the SV40 early promoter; the BglII site immediately precedes the SV40 early promoter, while the BamHI site is downstream from the SV40 poly-adenylation site. Plasmids pPPE(s)Pro (sense), pPPE(a)Pro (antisense), pProPPE(s) (sense), and pProPPE(a) (antisense) have a single insert of the 35 bp fragment in the indicated orientation at the BglII or BamHI sites, respectively. Plasmid pPPE(a,a,s)Pro contains three inserts in the BglII site. A series of 12 mutant constructs were made, each containing consecutive three nucleotide changes in the rat phospholamban gene. These mutants covered the region −159 through −124. The nucleotide changes were selected to differ from the nucleotide sequences of the rat phospholamban gene and four other mammalian phospholamban genes in the comparable location. Mutations were generated by PCR using a 5′ oligomer containing various mutated phospholamban sequences starting at nucleotide −159, a 3′ oligomer complementary to sequences +46 to +64, and plasmid −159/+64 as an amplification template. PCR products were restriction digested and cloned into pGL2-Basic. PCR amplifications were performed using Taq polymerase (Life Technologies Inc., Gaithersburg, MD, USA) and the manufacturer’s recommended reaction buffer. All subcloned PCR products had nucleotide sequences confirmed via automated dideoxy sequence analysis.

Rat/human chimeric constructs A KpnI/NcoI fragment containing human phospholamban sequences −169/−93 from plasmid H−169/+64 (McTiernan et al., 1999) was subcloned into KpnI/NcoI digested −159/+64C93 to generate plasmid HRluc, which contains human phospholamban sequences −169/−93 followed by rat phospholamban sequences −92/+64 and the luciferase coding region. Plasmid RHluc in which the rat −159/−93 sequence is followed by the human −92/+64 was created in a similar fashion.

Cell culture, DNA transfections, and reporter gene assays Cardiomyocytes were prepared from ventricles of 1-day-old Sprague–Dawley rats (Harlan Sprague– Dawley, Indianapolis, IN, USA) exactly as previously described (McTiernan et al., 1997, 1999), with approval of the Animal Care and Use Committee of the University of Pittsburgh. HeLa and rat fibroblast cells (Rat2) were grown, plasmid DNAs purified, and cells transfected as previously reported (McTiernan et al., 1999). Constructs were usually tested with at least two different plasmid DNA preparations using triplicate wells and at least three different cell preparations. Plasmid construct DNAs were mixed with internal control DNA (1:0.33 mass ratio) prior to addition of cationic liposomes (Lipofectin or Lipofectamine, Life Technologies). Internal control plasmids expressed beta-galactosidase (RSV b-gal), or chloramphenicol acetyl transferase (RSV CAT) under control of the Rous Sarcoma virus LTR, or sea pansy luciferase under control of the SV40 early enhancer/promoter, or CMV promoter (pRLSV40 and pRLCMV; Promega). Parallel control plasmids include pGL2-Promoter (see above) and pGL2-Control (Promega), a derivative of pGL2-Promoter which also contains the SV40 early gene enhancer. Cell lysates were prepared and assayed using commercial assay substrates (firefly and sea pansy luciferases, Promega; beta-galactosidase, Tropix, Bedford, MA, USA) as previously reported (McTiernan et al., 1999). Chloramphenicol acetyl transferase activity was measured as described (Seed and Sheen, 1988). In a few experiments luciferase values were normalized to extract protein concentration, determined via a colorimetric assay (BioRad, Hercules, CA, USA) using bovine IgG as a protein standard. Backgound values were determined using extracts from parallel non-transfected cultures or assays containing reaction buffer only, and subtracted from the raw sample values prior to calculating the ratio of luciferase to internal control reporter activity so as to normalize for variations in transfection efficiency. Due to the negligible luciferase activity of the promoterless construct pGL2-Basic in cardiomyocytes, this ratio was reported as the percentage of the value obtained from parallel cultures transfected with plasmid −159/+64, the minimal phospholamban construct displaying significant promoter activity in cardiomyocytes. When assessing the enhancer activity of phospholamban sequences, the ratio of luciferase to internal control activity was reported relative to cultures transfected with the parent plasmid (pGL2-Promoter). When assessing the

Rat Phospholamban Promoter Elements

replacement of the PPE1 element with a KpnI linker or the scram159/125 oligomer, the ratio was reported relative to the parent plasmid −1212/ +64.

Nuclear extract preparations and DNA/protein interaction analyses Extracts from ventricles of 1-day-old neonatal rats were prepared as described (Navankasattusas et al., 1992). HeLa nuclear extracts were commercially available (Promega). Proteins were dialyzed into 15 m HEPES pH 7.9, 40 m KCl, 1 m EDTA, 0.5 m dithiothreitol, 20% glycerol (v/v), and 0.5 m phenylmethylsulfonyl fluoride (PMSF) and stored in aliquots at −80°C. Protein concentrations were determined as described above. Multiple double-stranded oligomers or PCR-generated templates were utilized to study protein/DNA interactions. A 174 base pair template (−193/−20) was prepared by PCR amplification (using sense oligomer −193/−172 5′-CCTGTCATTCACAGCTCTACAT-3′ and antisense oligomer −20/−40 5′-GGAGGAGAAAAAGCATTCTAG-3′). Radiolabelled templates were prepared by PCR using one oligomer labeled via T4 polynucleotide kinase. Unlabelled double-stranded phospholamban promoter competitor templates were prepared in a similar fashion. DNA templates were purified by gel electrophoresis, electroelution, and ethanol precipitation. Additional double-stranded competitor oligomers were commercially available (AP1-3, CREB, GRE, NFjB, OCT-1, and SP1; Promega). AlbD and M-CAT element double-stranded competitor DNAs (Mueller et al., 1990; Karns et al., 1995) were generously provided by A. Stewart, University of Pittsburgh. DNA/protein interactions were performed by preincubating 2–40 lg extract protein and 0.4–0.8 lg poly dIdC in 10 ll binding buffer (10 m TrisCl pH 8.0, 0.5 m EDTA, 0.5 m DTT, 40 m KCl, 5% glycerol, 1 m MgCl2) for 15 min at room temperature. When present, competitor DNA (10–200fold molar excess relative to radiolabelled DNA) was incubated with extract at room temperature for 15 min prior to addition of the radiolabeled probe. Complex formation was initiated by adding 1 ll containing 1 ng (about 20 000 cpm) radiolabeled double-stranded oligomer followed by incubation at room temperature for 20 min. For exonuclease III digestions, additional MgCl2 (final concentration 13 m) and 455 U of exonuclease III (Life Technologies) were added and samples incubated at room temperature for an

2141

additional 20 min. Samples were ethanol precipitated and recovered DNA separated on 8% acrylamide DNA sequencing gels. A sequencing ladder, created by PCR cycle sequencing of the −1212/+64 template using the same oligomer that had been radiolabeled for the ExoIII reactions, was prepared to identify the location of protected bands. Gels were dried and images obtained on a storage phosphor screen (Molecular Dynamics, Sunnyvale, CA, USA). For electrophoretic mobility shift assays (EMSA), protein/DNA interactions were performed as described above. When utilized, antiNF-YA antibody (Rockland, Gilbertsville, PA, USA) was pre-incubated with nuclear extract (2 ll antisera per 8 lg extract protein) for 1 h at 4°C prior to initiation of binding reactions. Complexes were electrophoresed at 4°C on 4% (29:1 acryl:bis) acrylamide gels in 0.25X TBE (1X TBE is 89 m Tris/89 m boric acid/2 m EDTA). Images were obtained as described above.

Statistical analysis Results of experiments are reported as the mean± standard error of the mean (...). Statistical comparisons were performed by Student’s t-test or Mann–Whitney non-parametric tests as indicated, or using one-way analysis of variance with P values adjusted by the Bonferroni method. Statistical significance was accepted at P<0.05.

Results Functional analysis of 5′-flanking region of the rat phospholamban gene Our previous work isolated rat genomic phospholamban DNA sequences (Johns and Feldman, 1992), and identified the putative transcription start site and intron/exon organization (Fig. 1A). This information and clones were utilized to create luciferase reporter constructs. Transient transfection assays were performed to identify regions of 5′-flanking sequences of the rat phospholamban gene required for promoter activity. Transfections were performed in primary cultures of neonatal rat cardiomyocytes, which express their endogenous phospholamban gene, and HeLa cells which do not. Truncations of 5′ upstream promoter sequences from −2100 to −159 were expressed at similar levels in neonatal rat cardiomyocytes (Fig. 1B). Truncations to −66 removed the putative CCAAT

2142

C. F. McTiernan et al.

box region (−83 through −79) (Johns and Feldman, 1992) and reduced expression to near background levels. Since the parent promoterless plasmid pGL2-Basic was completely inactive in neonatal cardiomyocytes, the promoter activity of the phospholamban constructs was compared to that of two additional control plasmids, pGL2-Promoter and pGL2-Control. In neonatal cardiomyocytes, the phospholamban reporter constructs longer than −159/+64 were five times more active than pGL2Promoter, and equally active with the pGL2-Control plasmid. In HeLa cells the phospholamban constructs were markedly less active, though not inactive; approximately 50% as active as pGL2-Promoter, and 25-fold less active than pGL2-Control (Fig. 1C). Deletion of 5′ sequences to −66 removed nearly all promoter activity of the phospholamban constructs in HeLa cells. Three additional constructs (PLB Mini Intron, PLB Intron, and PLB Intron Plus) were made to assess whether the inclusion of sequences within the far upstream region (between −1212 and −4032), the 7.4 kb intron, and 3′UTR region were capable of directing greater cardiomyocyte expression and/or limiting expression in non-cardiomyocytes, which in these experiments included Hela and Rat2 cells. While the addition of the intronic sequences reduced expression of the constructs in HeLa cells relative to both the −159/ +64 and −1212/+64 phospholamban promoter constructs, they did not reduce promoter activity in Rat2 cells to levels lower than the −159/+64 construct (Fig. 2), hence additional tissue-restrictive regulatory elements or mechanisms remain to be identified. Although not identifying tissue-specific elements, these results identified a small region (between −159 and −66) required for maximal phospholamban promoter activity in cardiomyocytes.

Sequence analysis of the 5′-flanking regions of the rat phospholamban gene We previously assessed the sequence conservation among five mammalian phospholamban promoter regions (McTiernan et al., 1999) to identify areas of similarity and previously characterized transcription regulatory motifs (Fig. 3). This analysis revealed a GATA box (Ko and Engel, 1993) (A/T GATA A/G −102/−97), a conserved CP-1-like CCAAT motif (Chodosh et al., 1988) (−92 through −74), M-CAT-like (Mar and Ordahl, 1990) (CATTCCT, −67/−62), TATA-like (−52/−48), and E-box (Murre et al., 1989) (CANNTG,

Figure 2 Influence of intronic and far-upstream sequences on phospholamban promoter activity. Plasmids containing approximately 1.2 kb of upstream sequence and a portion of the rat phospholamban intron (Mini Intron), the entire intron (PLB Intron), or 4 kb of upstream sequence and the entire intron (Intron Plus) were transiently transfected into neonatal rat cardiomyocytes, Rat2, and HeLa cells. Promoter activity (mean±...) is reported relative to that of the rat phospholamban construct −159/+64 (set as 1.0). Data from four to seven independent experiments for each cell type, each using two different DNA preparations for each construct, transfected in triplicate.

−11/−6) elements. As the extensive sequence conservation in the −159 through −66 region only contained a probable CCAAT-box and possible GATA element, a more restricted series of reporter construct mutations were created to functionally identify novel sequences required for phospholamban promoter activity. Also identified in Figure

Rat Phospholamban Promoter Elements

2143

Figure 3 Alignment of mammalian upstream phospholamban sequences. Nucleotide sequences (206 bases including first six nucleotides corresponding to rat exon 1) from rat (GenBank L03381; Johns and Feldman, 1992), mouse (Haghighi et al., 1997), rabbit (GenBank M63600; Fujii et al., 1991), human (McTiernan et al., 1999; and GenBank Z99496), and dog (GenBank AF037348) were aligned using the BLAST program. Numbers correspond to rat nucleotides upstream of exon 1. Gaps indicated by dot. Conserved phospholamban promoter element 1 (−159/−125), GATA (−101/ −96), CP1/NFY (−82/−78), M-CAT-like (−67/−62), TATAA-like (−51/−47), and E-box (−11/−6) elements are boxed; reported exon sequences underlined.

3 is a region of marked conservation between the five mammalian species, further characterized and described below as Phospholamban Promoter Element 1 (PPE1). Novel basal regulatory region between −159 and −125 In order to focus on the region upstream from the CP-1-like CCAAT domain (−92/−71) to search for additional novel regulatory elements, a series of truncations and deletions between −159 and −93

were created and tested in neonatal cardiomyocytes and HeLa cells (Fig. 4A). The results indicated the importance of sequences between −159 and −125 in allowing maximal phospholamban promoter activity in neonatal cardiomyocytes as all constructs with the phospholamban 5′ end downstream from −159 had expression reduced to less than 15% of the −159/+64 construct (Fig. 4B). Additional constructs containing internal deletions fixed at −93 and increasing in the 5′ direction indicated the requirement for sequences upstream of −125 in allowing maximal promoter activity in neonatal

2144

C. F. McTiernan et al.

Figure 4 Importance of sequences from −159 to −125 for phospholamban promoter activity. (A) Schematic map of promoter constructs. Location of CCAAT box, luciferase coding regions (Luc), and an NcoI site introduced by point mutation (Nco) are indicated. Rat phospholamban sequences are indicated relative to transcription initiation site at +1. After transfection into neonatal rat cardiomyocytes (B) or HeLa cells (C), luciferase activity was normalized to internal control plasmid activity and then normalized to the activity of the −159/+64 construct, arbitrarily set at 1.0. Values are mean±... Neonatal rat cardiomyocyte results are the summation of 4 to 15 different experiments using two to six replicates per construct (usually three). HeLa results are from transfections of at least two different cell preparations using two to six replicates per construct (typically three).

cardiomyocytes, and the non-essential nature of sequences between −93 and −124 (including the potential GATA motif at −101/−96) for promoter activity during transient transfection assays (Fig. 4). Interestingly, the complete loss of even low levels of promoter activity in constructs containing the −107/+64 region was reversed when the −107 through −94 sequence was removed (as for −93/ +64), suggesting that the −107/−94 region, which contains the GATA motif, may have negative regulatory functions in the absence of sequences upstream of nucleotide −107. As before, in HeLa cells the phospholamban promoter constructs appear less active than in neonatal cardiomyocytes when compared to the

SV40 promoter control plasmids. However, the phospholamban 5′ truncation constructs (−159/ +64, −142/+64, and −93/+64) were equally active in HeLa cells indicating that sequences upstream of −93 may not have a marked role in regulating activity of the phospholamban promoter in HeLa cells (Fig. 4C). We previously reported a similar observation for human phospholamban promoter constructs (McTiernan et al., 1999). The requirement of the −159 through −125 region (PPE1) for promoter activity in neonatal cardiomyocytes was further demonstrated by constructs which had −159/−125 replaced with either a 6 bp linker containing a Kpn I site (−1212d159/125), or a 34 base pair scrambled

Rat Phospholamban Promoter Elements

2145

While not exhausting all possible sequence mutations, it was striking that five consecutive mutants between −150 and −133 (mutants 4–9) could each individually remove 80–90% of the promoter activity of construct −159/+64 (Fig. 5B). The rat −159/−125 sequence can activate a heterologous promoter in cardiomyocytes

Figure 5 Site-directed mutation analysis of phospholamban promoter region −159 through −125. (A) Twelve different nucleotide triplet frames covered the rat phospholamban region −159 through −124 (wild-type sequence). Triplet mutations of each separate frame were made, creating mutations M1 through M12; mutant triplet sequences indicated. (B) Mutant reporters were transiently transfected into neonatal rat cardiomyocytes. Luciferase activity was normalized to internal control plasmid activity and then normalized to the activity of the −159/+64 construct, arbitrarily set at 1.0. Values are mean±... Data from three independent experiments each using one or two different DNA preparations for each construct, transfected in triplicate.

sequence containing a base composition similar to the endogenous −159/−125 region (−1212scram159/125). These mutants both lost 90% or more promoter activity relative to the parent −1212/+64 plasmid (−1212/+64, 100± 7.46%, n=39; −1212d159/125, 10.5±1.19%, n=39; −1212scram159/125, 1.81±0.25%, n= 18; P<0.0001 ANOVA). Additional evidence for the importance of the −159/−125 region for phospholamban promoter activity in cardiomyocytes was obtained through a series of 12 mutations in which three consecutive nucleotides were systematically mutated to sequences not conserved with other mammalian phospholamban genes within the −159/−125 region (Fig. 5A).

We investigated whether the rat −159/−125 (PPE1) sequence could activate a heterologous promoter when introduced into two different locations and orientations on the SV40 early promoter-luciferase expression construct pGL2-Promoter (Fig. 6A). These constructs were transfected into neonatal cardiomyocytes and their promoter activity compared to that of pGL2-Promoter (Fig. 6B). Insertion of the −159/−125 sequence in either orientation in the 5′ proximal region of the SV40 early promoter markedly enhanced activity of the SV40 early promoter in cardiomyocytes. Three repeats of the PPE1 element [pPPE(a, a, s)Pro] yielded greater activation, approximately 15-fold higher than the parent plasmid (pGL2-Promoter, P<0.001), and significantly higher than the constructs containing a single PPE1 insert [P<0.01 compared to pPPE(a or s)Pro]. However, insertion of PPE1 in either orientation at the 3′ end of the luciferase gene [pProPPE(a) or pProPPE(s)] did not activate the SV40 early promoter. These constructs were also transfected into HeLa cells to assess their cell-type specificity. In contrast to the studies in cardiomyocytes, the PPE1 sequence showed a much smaller, non-significant activation of the SV40 promoter in HeLa cells (Fig. 6C). Our previous studies with the human phospholamban promoter (McTiernan et al., 1999) also indicated the importance of sequences between −169 and −93 (equivalent to rat −159/−93) in allowing maximal human phospholamban promoter activity during transient transfection in neonatal rat cardiomyocytes, although the magnitude of the effect, while significant, was much lower. Indeed, parallel transfection assays indicated that the human phospholamban promoter was approximately 10 times less active than the equivalent rat promoter construct when transfected into rat cardiomyocytes. To identify the region of the human promoter which was responsible for the limitation of the human promoter in rat cells, we made chimeric constructs containing nucleotides −159 through −93 from the rat promoter ligated to the human −92 through +64 (which contains the human putative CCAAT box; construct RHluc), or human

2146

C. F. McTiernan et al.

Figure 6 The PPE1 element activates expression of the SV40 promoter in neonatal rat cardiomyocytes. (A) A synthetic oligomer encoding rat phospholamban sequences −159 through −125 was cloned either upstream or downstream of the SV40 early promoter (SV Pro) in the plasmid pGL2-Promoter. SV40 enhancer (SV Enh), rat phospholamban promoter sequences −159/+64 (−159/+64), and −159/−125 in the sense (s; −159/−125) and antisense (a; −125/−159) orientations are indicated. Constructs were transfected into cardiomyocytes (B) or HeLa cells (C), luciferase activity normalized to an internal control plasmid, and results in turn normalized to that of pGL2-Promoter, arbitrarily set at 1.0. Values are reported as mean±... Results are from at least five different cell preparations of each cell type, with the number of replicate wells per construct within an experiment ranging from two to eight (typically three). ∗ Bonferroni corrected P values <0.001 compared to pGL2-promoter; † Bonferroni corrected P values <0.01 compared to pPPE(a or s)Pro.

−169 through −93 nucleotides ligated to rat −93 through +64 (which contains the rat putative CCAAT box; construct HRluc). These studies indicated that the rat sequences −159 through −93 could enhance promoter activity of the human −92/+64 fragment, whereas the human −169 through −93 region could not enhance the rat −92 through +64 fragment, suggesting that sequences or spacing parameters of the rat −159/−93 region were critical to activation of the rat phospholamban promoter in rat cells (Fig. 7).

Importance of the CCAAT motif Both neonatal cardiomyocytes and HeLa cells require sequences between −93 and −66 for minimal rat phospholamban promoter activity. As this region contains a conserved CCAAT-box motif, we assessed whether a point mutation in this element would affect basal promoter activity. A plasmid (−159/+64C93G81) containing a point mutation in the CCAAT box motif (CCGAT) loses approximately 85% of promoter activity in both neonatal cardiomyocytes (−159/+64C93, 100±9.9, n=

Rat Phospholamban Promoter Elements

2147

Figure 7 Rat phospholamban sequences −159/−93 activates the basal human phospholamban promoter in rat cardiomyocytes. Neonatal rat cardiomyocytes were transfected with reporter constructs and analysed as described (see Materials and Methods and legends to Figs 1, 4–6). Organization of chimeric plasmids of rat and human phospholamban genes (RHLuc and HRLuc) are described in Materials and Methods. Luciferase activity was normalized to internal control plasmid activity and then normalized to the activity of the −159/+64 construct, arbitrarily set at 1.0. Values are mean±... Data from three independent experiments each using two different DNA preparations for each construct, transfected in triplicate.

9; −159/+64C93G81, 15.35±1.8, n=15, P<0.0001) and HeLa cells (−159/+64C93, 100±9.17, n=12; −159/+64C93G81, 10.4± 1.04, n=18, P<0.0001). The parent plasmid (−159/+64C93) which contains a single nucleotide change at −93 was as active as the equivalent wild type construct (−159/+64). Thus while both neonatal cardiomyocytes and HeLa cells require an intact CCAAT-box element for activity of the phospholamban promoter, the 35 bp region between −159 and −125 (PPE1) appears more active in neonatal rat cardiomyocytes. In contrast, the GATA-like element (−100 through −96) is not required for basal, unregulated activity of the phospholamban promoter when transiently transfected into either neonatal cardiomyocytes or in HeLa cells (Fig. 4).

Protein-binding domains in the phospholamban promoter To localize the sites of interaction between −159/ −125 and trans-acting factors, nuclear extracts were prepared from whole neonatal hearts and tested for the presence of specific phospholambanpromoter binding factors using both Exonuclease III (Exo III) protection and electrophoretic mobility shift assays. Exo III protection shows the boundary of protein/DNA interactions which limit further exonuclease digestions but does not reveal the entire

footprint of the protein/DNA interaction. However, it is more sensitive than methods such as DNAse I footprinting as new exonuclease-resistant fragments can be identified even if only a small proportion of the template DNA forms specific complexes (Wu, 1985). For these studies a 174 bp template containing phospholamban promoter region −193 through −20 (−193/−20) was used; the template was designed to include the CCAATbox to serve as an internal control for detection of protein/DNA interactions. Template DNA was radiolabelled on the 5′ end of the antisense strand and interacted with extracts prepared from whole neonatal hearts, digested with Exo III, and analysed on sequencing gels. In the absence of neonatal rat heart nuclear extract or in the presence of bovine serum albumin the template was digested to a series of fragments consistent with digestion to nucleotide −170/−168, −160/−159, −155/−154, and −136/−135 (Fig. 8). However, addition of neonatal rat heart nuclear extract led to the appearance of additional bands consistent with halting of the Exo III digestions around nucleotides −174/−173, and more prominently −148/−143 and −91/ −84. Increasing the amount of rat heart nuclear extract enhanced the appearance of these bands. Addition of a 100–200-fold molar excess of unlabelled −193/−20 greatly reduced the appearance of these protected bands whereas addition of excess oligomer encoding an unrelated sequence

2148

C. F. McTiernan et al.

Figure 8 Exonuclease III footprint analysis of phospholamban proximal promoter sequences. A 174-bp template containing phospholamban promoter nucleotides −193 through −20 was radiolabelled at the 5′ end of the anti-sense strand and subjected to Exonuclease III digestion after interaction with various amounts of neonatal rat heart (NRH) nuclear extracts or BSA and products separated on 8% polyacrylamide sequencing gels. Undigested template (lane 1); Exo III digested without extract (lane 2); Exo III digested after interaction of template with BSA (lanes 3, 7); Exo III digested after interaction with 2, 4, or 8 lg of NRH nuclear extract (lanes 4–6). Exo III digestions performed with NRH (lanes 8–11) or HeLa (lanes 12–15) extracts, and with specific (Self, −165/−135) or non-specific (AP3) competitor DNAs. Extracts were pre-incubated with 100-fold molar excess non-radiolabelled labeled competitor DNA (none, lanes 1–8, 12); non-specific competitor AP3 oligomer (lanes 10, 14), phospholamban promoter sequences −165/−135 (lanes 11, 15), or Self (lanes 9, 13) before addition of radiolabelled probe and Exo III digestion. Numbers indicate nucleotide position relative to transcription start site. Nucleotides encompassing PPE1 and CP-1 domains indicated between panels. (A,G) Cycle-sequencing reactions terminated at A or G residues to determine product sizes.

(AP3 element) did not prevent appearance of the protected bands (Fig. 8). Furthermore, excess unlabelled oligomer encoding phospholamban sequences −165/−135 almost completely inhibited formation of protected bands around nucleotides −148/−143, but only incompletely inhibited protection around bases −91/−89. Similar results were obtained for at least five different experiments utilizing at least three different preparations of neonatal rat heart extracts. Studies performed with

HeLa extracts did not protect regions within the PPE1 domain but led to a somewhat weaker protection of −91/−87 region upstream of the CCAAT-box, indicating a cell-specificity of the factors that interact with and protect the PPE1 region from Exo III digestions (Fig. 8). Additional bands that appeared to be protected between −128/−120 and −100 were neither consistently observed nor competed with excess self template and thus are unlikely to represent specific

Rat Phospholamban Promoter Elements

complexes. Furthermore, although digestion with decreasing amounts of Exo III led to less complete digestion of the −193/−20 template and the appearance of higher molecular weight bands, none of the partial digestion products corresponded to the bands that appeared in the presence of neonatal rat heart extracts (data not presented). To complement the evidence for nuclear factor interactions with phospholamban promoter sequences obtained from exonuclease III protection assays, electrophoretic mobility shift assays were also performed. The template containing phospholamban promoter sequences −193/−20 consistently formed three complexes (termed A, B, C) when interacted with neonatal rat heart nuclear extracts (Fig. 9A). These complexes could not be competed for by the preincubation of extract with a 100-fold excess of consensus oligomers for nine different transcriptional regulatory elements (AP13, CREB, GRE, NFjB, OCT-1, M-CAT, or SP1), while they could be competed for by a 100-fold excess of the unlabelled template DNA. Promoter sequences between −66 to −20 appear not to be involved in these complexes as a 100-fold excess of an oligomer encoding this region could not compete for formation of complexes. However, an oligomer encoding the CCAAT-box region (−98/−64) competed for formation of complex B with little competition for complex A and C. To better ascertain the nature of the CCAAT-box motif, competition was performed with consensus oligomers for three different CCAAT-box containing elements; CP-1, NF1/CTF, and AlbD. Only the CP-1 consensus element competed complex B formation, indicating that CP-1/NF-Y factors bind to the CCAAT-box motif between −83 and −79 of the rat PLB promoter. Further confirmation of CP-1/NF-Y binding to this region was obtained by preincubation of nuclear extract with an anti NF-YA antiserum, which supershifted complex B to a slower mobility (Fig. 9B). Additional studies sought to identify complexes arising from the interaction of nuclear extract with the −159/−125 region. Oligomers containing PPE1 sequences (−165/−135 and −159/−125) competed formation of the C complex, whereas competitor −107/−20 (as well as −98/−64, −66/−20, and 12 other consensus site oligomers; see Fig. 9A) did not compete for complex C formation (Fig. 9A,C). Complex A appears to be nonspecific as it could only be competed with lengthy DNA fragments regardless of sequence (data not presented). HeLa cell nuclear extract interacted with the −193/−20 template also formed complexes resembling complex B and C obtained with neonatal rat heart extract (Fig. 9). Complex B could also

2149

be competed with −107/−20 (data not shown) suggesting interaction with the CCAAT- motif. In addition, the PPE1-related oligomers −165/−135 and −159/−125 also competed complex C similar to that observed with neonatal rat heart extracts.

Discussion To initiate studies on transcriptional regulation of phospholamban expression, we have characterized elements of the 5′-proximal region of the rat phospholamban promoter required for functional activity in transient transfection assays. Constructs containing phospholamban promoter sequences ranging from −159/+64 to −2100/+64 show significant activity in neonatal rat cardiomyocytes with lesser activity observed in HeLa cells, which do not express phospholamban. In an attempt to localize more robust cell-specific regulatory elements we examined 12 kb of the rat phospholamban gene region but were unable to achieve cell-type restricted expression. However, using a series of truncated and mutagenized templates we have identified both a novel 35 bp region (PPE1, −159/ −125) and a more ubiquitous motif (CP-1/NF-Y; −79/−83) required for maximal phospholamban promoter activity in neonatal rat cardiomyocytes. The importance of the −159/−125 region (PPE1) is supported by several lines of evidence: deletion of this 35 bp region, or mutations within the −150/−133 region markedly reduced promoter activity in transiently transfected cardiomyocytes; the 35 bp PPE1 sequence enhanced promoter activity of a heterologous promoter in cardiomyocytes; and exonuclease III footprint analysis indicated sequence specific interactions with nucleotides adjacent to −148/−143 when using rat heart, but not Hela nuclear extracts. Furthermore, the 35 bp region shows >70% conservation of nucleotide sequences with other mammalian phospholamban promoters. Notably, a recent analysis of the murine phospholamban promoter by transient transfection of cells (L6 and H9c2) which weakly express their endogenous phospholamban gene observed a five-fold increase of promoter activity when sequences between −200 and −100 were included in the expression constructs (Haghighi et al., 1997), similar to that observed in this report when comparing the rat expression constructs −93/+64 to −159/+64 in neonatal cardiomyocytes. We previously reported a similar effect but of a lower agnitude for the equivalent region (−169 through −93) of the human phospholamban gene (McTiernan et al.,

2150

C. F. McTiernan et al.

1999), although we have noted that the human phospholamban promoter is markedly less active than the corresponding rat promoter in neonatal rat cardiomyocytes. Our chimeric constructs indicated the importance of the rat −159/−93 region in activation of the human phospholamban basal promoter in rat cardiomyocytes. Indeed, the heterologous species combination of rabbit phospholamban promoter and rat cardiomyocytes may have helped to mask the importance of the PPE1 region in studies of the rabbit phospholamban promoter (Yabuki et al., 1998). Despite the identification of a strong positive regulatory element in the rat phospholamban promoter we did not observe absolute cell specificity of phospholamban promoter construct activity, which may arise from several factors. While the CP-1/NFY region is required for phospholamban promoter activity in both neonatal cardiomyocytes and HeLa cells, it alone is sufficient for activity in HeLa cells, suggesting the absence of negative regulatory elements within the regions that we have examined. Although gene sequences such as distal inhibition elements may not be present in our reporter constructs, such regions would need to be outside of the 12 kb of the rat phospholamban gene assessed in this study. In addition, despite the common use of transient transfection assays to identify promoter elements, several recent studies (reviewed in Smith and Hager, 1997) suggest that transiently transfected reporter constructs may not acquire the repressed chromatin structure of endogenous genes or stably replicated templates, allowing ubiquitous factors to bind and modify expression of the transfected DNA template. The PPE1 sequence does not resemble other known transcriptional regulatory motifs, raising the possibility that this element interacts with novel regulatory factors in a sequence-specific fashion to effect gene expression in cardiomyocytes. Such a sequence-specific interaction of rat heart nuclear factors with the PPE1 region was suggested both by the analysis of mutated −159/−125 sequences,

and by exonuclease III protection assays. In addition, exonuclease III protection studies demonstrated that while both rat heart or HeLa cell extracts protected the CP-1/NF-Y domain, only rat heart extracts showed significant interaction with the PPE1 region. However, for reasons that are not clear, this differential interaction of rat heart and HeLa nuclear extracts with the PPE1 domain was not readily detectable by EMSA. What is clear from this study is that the PPE1 is a novel and necessary element for the activity of the phospholamban promoter in cardiomyocytes, and that this element can activate heterologous promoters in cardiomyocytes. The phospholamban gene also requires a CP-1/ CCAAT box element for activity, suggested by the following evidence: both neonatal cardiomyocytes and HeLa cells require the core CCAAT-box region for phospholamban promoter activity; rat heart extracts form a mobility shift complex that can be competed with a CP-1-like consensus oligomer but not with consensus oligomers for other CCAAT-box motifs such as AlbD or NF-I; this complex can be super-shifted by antibodies directed against NF-Y subunit A; and exonuclease III protection assays suggest protein/DNA interactions around −91, 8 bp upstream of the consensus CCAAT motif. It is of interest to note that purified CP-1/NF-Y contacts the a1(I) and a2(I) collagen promoter 7–9 bp upstream of the CCAAT motif (Bi et al., 1997). These results are compatible with those recently reported for the rabbit phospholamban promoter (Yabuki et al., 1998). While proteins that bind to the CP-1-like motif (CP-1, also known as NF-Y ) are ubiquitously expressed (Hooft van Huijsduijnen et al., 1990) several studies report the importance of NF-Y in regulated gene expression (Alonso et al., 1996; Jump et al., 1997; Orita et al., 1997), thus additional studies will be required to assess the potential role of CCAAT-box binding proteins in phospholamban gene expression. The basal phospholamban promoter (−159 to +64) that shows activity in neonatal cardio-

Figure 9 Interaction of nuclear extracts with CCAAT motif of phospholamban proximal promoter. (A) The 174 bp template containing phospholamban promoter nucleotides −193 through −20 was radiolabelled at the 5′ end of the anti-sense strand, interacted with neonatal rat heart extract (NRH) and subjected to EMSA as described. Except where indicated, extracts were pre-incubated with 100-fold excess of non-radiolabelled oligomers before addition of radiolabelled DNA. Competitor oligomers as described in “Materials and Methods”. −66/−20 and −98/−64 refer to doublestranded oligomers encompassing the indicated phospholamban promoter sequences. (B) NRH extracts pre-incubated with antibody directed against NF-Y subunit A supershifts Complex B (indicated by ∗). (C) The 174 bp template was interacted with neonatal rat heart (NRH, left panel) or HeLa (right panel) extracts and subjected to EMSA. Extracts were pre-incubated with non-radiolabelled competitor DNAs (100-fold excess) where indicated; Self (−193/−20); −165/−135, −107/−20, and −159/−125 refer to double-stranded oligomers encompassing the indicated phospholamban promoter sequences.

Rat Phospholamban Promoter Elements

2151

2152

C. F. McTiernan et al.

myocytes does not contain five other previously described regulatory elements utilized in skeletal muscle or cardiac-expressed genes [CArG (Minty and Kedes, 1986), MEF1 (Buskin and Hauschka, 1989), MEF2 (Gossett et al., 1989), GArC (Mably et al., 1993), CCAC-box (Bassel-Duby et al., 1992)]. In addition, two other motifs [GATA (Ko and Engel, 1993) and E-boxes (Murre et al., 1989)] are not conserved at the same location among the five mammalian phospholamban promoters (McTiernan et al., 1999), and can be altered or removed without effect on the activity of transfected phospholamban promoter constructs in neonatal cardiomyocytes. However, these results do not preclude the possible importance of these motifs in phospholamban promoter activity in the intact adult organism or during cardiac development as the functional significance of a regulatory element may not always be revealed through transient transfection or in vitro DNA-protein interaction assays. Indeed, the results of this study are limited by several aspects of the available experimental systems. The importance of the PPE1 and CCAAT motifs in the rat phospholamban gene were analysed in the context of transient transfection assays in neonatal rat cardiomyocytes, and by interaction with extracts from neonatal hearts. Since routine transient transfection assays of cultured adult cardiomyocytes is not currently practical, the importance of these motifs will need to be confirmed in adult cardiac tissues, perhaps through use of transgenic approaches. The importance of these elements in the regulated expression of phospholamban, such as in response to thyroid hormones (Arai et al., 1991), or interleukin-1 (McTiernan et al., 1999) also remains uninvestigated. Furthermore, the complex physiological signaling of heart failure, commonly associated with a decreased expression of phospholamban (Rockman et al., 1994; Kiss et al., 1995) may not be readily recapitulated by the use of cultured neonatal cardiomyocytes, hence the approaches used in this study to analyse “basal” elements may not reveal regulatory elements important in pathological settings. However, within the bounds of these limitations, the current study points to the essential role of both a novel (PPE1) and previously characterized element (CP1/NF-Y) for phospholamban promoter activity in cardiomyocytes. Considering the combinatorial nature of motifs and factors contributing to promoter activity, it is probable that additional elements regulate phospholamban expression in cardiomyocytes and other cell types.

Acknowledgements We thank G. Bounoutas and B. Will for cultured cardiomyocytes, and Dr A. Stewart for oligonucleotides. This research was supported by the Pennsylvania Affiliate of the American Heart Association (Grant in Aid to C. McTiernan).

References A CR, P CG, K AR, 1996. The CCAATbinding proteins CP1 and NF-I cooperate with ATF-2 in the transcription of the fibronectin gene. J Biol Chem 271: 22271–22279. A M, O K, ML DH, A NR, P M, 1991. Effect of thyroid hormone on the expression of mRNA encoding sarcoplasmic reticulum proteins. Circ Res 69: 266–276. A M, A NR, ML DH, B P, P M, 1993. Alterations in sarcoplasmic reticulum gene expression in human heart failure. A possible mechanism for alterations in systolic and diastolic properties of the failing myocardium. Circ Res 72: 463–469. B-D R, H MD, G MA, K JK, W RS, 1992. A 40-kilodalton protein binds specifically to an upstream sequence element essential for muscle-specific transcription of the human myoglobin promoter. Mol Cell Biol 12: 5024–5032. B W, W L, C F,  C B, M SN, 1997. DNA binding specificity of the CCAAT-binding factor CBF/NF-Y. J Biol Chem 272: 26562–26572. B JN, H SD, 1989. Identification of a myocyte nuclear factor that binds to the muscle-specific enhancer of the mouse muscle creatine kinase gene. Mol Cell Biol 9: 2627–2640. C LA, B AS, C RW, S PA, 1988. Human CCAAT-binding proteins have heterologous subunits. Cell 53: 11–24. F AM, R PE, S CM, M JA, M W, B MR, 1991. Selective gene expression in failing human heart. Quantification of steady-state levels of messenger RNA in endomyocardial biopsies using the polymerase chain reaction. Circulation 83: 1866–1872. F J, Z-H A, W HF, T M, ML DH, 1991. Structure of the rabbit phospholamban gene, cloning of the human cDNA, and assignment of the gene to human chromosome 6. J Biol Chem 266: 11669–11675. G LA, K DJ, S EA, O EN, 1989. A new myocyte-specific enhancer-binding factor that recognizes a conserved element associated with multiple muscle-specific genes. Mol Cell Biol 9: 5022–5033. H K, K VJ, K KL, L W, H JM, P S, Z Z, K EG, 1997. In vitro and in vivo promoter analyses of the mouse phospholamban gene. Gene 203: 199–207. H MJ, S M, K AM, 1979. Mechanism by which cyclic adenosine 3′:5′-monophosphate-dependent protein kinase stimulates calcium transport in cardiac sarcoplasmic reticulum. Circ Res 44: 384–391. H  H R, L XY, B D, M H,

Rat Phospholamban Promoter Elements B C, M D, 1990. Co-evolution from yeast to mouse: cDNA cloning of the two NF-Y (CP- 1/CBF) subunits. Embo J 9: 3119–3127. H RM, H HD, H SN, P JK, P LR, 1989. Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77: 61–68. J DC, F AM, 1992. Identification of a highly conserved region at the 5′ flank of the phospholamban gene. Biochem Biophys Res Commun 188: 927–933. J DB, B MV, T A, 1997. The CCAAT box binding factor, NF-Y, is required for thyroid hormone regulation of rat liver S14 gene transcription. J Biol Chem 272: 27778–27786. K LR, K K, S PC, 1995. M-CAT, CArG, and Sp1 elements are required for alpha 1-adrenergic induction of the skeletal alpha-actin promoter during cardiac myocyte hypertrophy. Transcriptional enhancer factor-1 and protein kinase C as conserved transducers of the fetal program in cardiac growth. J Biol Chem 270: 410–417. K E, B NA, K EG, W RA, 1995. Differential changes in cardiac phospholamban and sarcoplasmic reticular Ca(2+)-ATPase protein levels. Effects on Ca2+ transport and mechanics in compensated pressure-overload hypertrophy and congestive heart failure. Circ Res 77: 759–764. K LJ, E JD, 1993. DNA-binding specificities of the GATA transcription factor family. Mol Cell Biol 13: 4011–4022. L E, H KT, B CJ, F J, T M, ML DH, P D, 1989. Slow/cardiac sarcoplasmic reticulum Ca2+-ATPase and phospholamban mRNAs are expressed in chronically stimulated rabbit fast-twitch muscle. Eur J Biochem 185: 51–54. L JP, J LR, H DR, H BG, W AM, 1983. beta-Adrenergic stimulation of phospholamban phosphorylation and Ca2+-ATPase activity in guinea pig ventricles. J Biol Chem 258: 464–471. L W, G IL, H J, P S, G G, D JJ, D T, K EG, 1994. Targeted ablation of the phospholamban gene is associated with markedly enhanced myocardial contractility and loss of betaagonist stimulation. Circ Res 75: 401–409. M JD, S MJ, L CC, 1993. Characterization of the GArC motif. A novel cis-acting element of the human cardiac myosin heavy chain genes. J Biol Chem 268: 476–482. M JH, O CP, 1990. M-CAT binding factor, a novel trans-acting factor governing muscle-specific transcription. Mol Cell Biol 10: 4271–4283. MT CF, L BH, F C, B S, C A, F AM, 1997. Interleukin-1 beta inhibits phospholamban gene expression in cultured cardiomyocytes. Circ Res 81: 493–503. MT CF, F CS, L BH, K EA, O-H ML, M CS, F AM, 1999. The human phospholamban gene: structure and expression. J Mol Cell Cardiol 31: 679–692. M M, S W, P B, H C, H C, P H, K G, M K, J H, H G, 1995. Alterations of sarcoplasmic reticulum proteins in failing human dilated cardiomyopathy. Circulation 92: 778–784.

2153

M A, K L, 1986. Upstream regions of the human cardiac actin gene that modulate its transcription in muscle cells: presence of an evolutionarily conserved repeated motif. Mol Cell Biol 6: 2125–2136. M AF, V JL, K MU, S K, L WH, B KR, 1995. Patterns of expression of sarcoplasmic reticulum Ca(2+)-ATPase and phospholamban mRNAs during rat heart development. Circ Res 76: 616–625. M JP, E RE, A PD, G W, G JK, 1990. Abnormal intracellular calcium handling, a major cause of systolic and diastolic dysfunction in ventricular myocardium from patients with heart failure. Circulation 81: III21–III32. M CR, M P, S U, 1990. DBP, a liverenriched transcriptional activator, is expressed late in ontogeny and its tissue specificity is determined posttranscriptionally [published erratum appears in Cell 1991 May 31;65(5): following 914]. Cell 61: 279–291. M C, MC PS, B D, 1989. A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and myc proteins. Cell 56: 777–783. N S, Z H, G AV, E SM, C KR, 1992. A ubiquitous factor (HF-1a) and a distinct muscle factor (HF-1b/MEF-2) form an E-box-independent pathway for cardiac muscle gene expression. Mol Cell Biol 12: 1469–1479. O T, S K, M H, N S, 1997. Binding of NF-Y transcription factor to one of the ciselements in the myeloperoxidase gene promoter that responds to granulocyte colony-stimulating factor. J Biol Chem 272: 23216–23223. R HA, O S, R RS, J LR, K M, B V, R J J, C KR, 1994. Molecular and physiological alterations in murine ventricular dysfunction. Proc Natl Acad Sci USA 91: 2694–2698. S RH, B M, S U, K P, B U, F M, K EG, E E, 1995. Unchanged protein levels of SERCA II and phospholamban but reduced Ca2+ uptake and Ca(2+)ATPase activity of cardiac sarcoplasmic reticulum from dilated cardiomyopathy patients compared with patients with nonfailing hearts. Circulation 92: 3220– 3228. S B, S JY, 1988. A simple phase-extraction assay for chloramphenicol acyltransferase activity. Gene 67: 271–277. S CM, W PL, M JC, 1993. Isolation of gene markers of differentiated and proliferating vascular smooth muscle cells. Circ Res 73: 193–204. S CL, H GL, 1997. Transcriptional regulation of mammalian genes in vivo. A tale of two templates. J Biol Chem 272: 27493–27496. T T, Z R, 1991. Characterization of cDNA and genomic sequences encoding a chicken phospholamban. J Biol Chem 266: 5375–5383. W C, 1985. An exonuclease protection assay reveals heat-shock element and TATA box DNA-binding proteins in crude nuclear extracts. Nature 317: 84–87. Y M, T T, O K, N M, K T, H M, T M, 1998. Involvement of NF-Y in transcriptional regulation of the phospholamban gene. Eur J Biochem 258: 744–751.