Plenary Session Presentations

michihiko tada
mtada@mr-path.med.osaka-u.ac.jp

Summary

To understand the transcriptional regulation of the phospholamban gene, we analyzed a
5'-upstream region of the phospholamban gene. Using a series of deletion constructs, we
demonstrated that the region from -96 to -78 base pairs containing the CCAAT sequence
is essential for transcription of this gene. This region specifically bound to nuclear
proteins extracted from rat hearts, and gel shift assays using competitive
oligonucleotides, antibodies and recombinant proteins showed that this region binds to
the NF-YA and NF-YB among the CCAAT motif-binding proteins. This
region-dependent transcription in cardiac myocytes transfected with antisense cDNAs
encoding NF-YA and NF-YB was decreased to approximately 50% of that seen in cells
transfected with the same sense cDNAs. We therefore conclude that the region from -96
to -78 base pairs plays a critical role in expression of the phospholamban gene which is
regulated by binding of the nuclear protein NF-Y.
 
 

Keywords

Phospholamban, Transcriptional Regulation, CCAAT binding proteins.
 

 
Introduction

    Phospholamban (PLN), a pentameric protein composed of 52 amino acids subunits, is
a major substrate for the cAMP-dependent protein kinase in sarcoplasmic reticulum of
cardiac myocytes [1][2]. PLN is found to regulate activity of cardiac sarcoplasmic
reticulum Ca-ATPase (SERCA2a) by the protein-protein interaction, in that the
de-phosphorylated PLN functions as an inhibitory co-factor of SERCA2a and the
phosphorylation of PLN relieves its inhibition for SERCA2a [3][4]. The subsequent
activation of SERCA2a leads to enhanced muscle relaxation rates, thereby contributing
to the inotropic response to adrenergic stimulation.
    By chemical and molecular biological methods, the protein-protein interaction
between PLN and SERCA2a is found to take place in cytoplasmic and transmembrane
parts of the two proteins [5][6](Fig. 1). At the cytoplasmic level, the charged residues of
two proteins form an electrostatic milieau. Unphosphorylated PLN suppresses the
SERCA2a activity by associating charged amino residues in the cytoplasmic domain of
PLN with the 6 serial amino acid residues KDDKPV402 in the cytoplasmic region of
SERCA2a [7][8][9]. Incorporation of phosphate into Ser16 in the cytoplasmic domain of
PLN by the cAMP-dependent protein kinase alters the charge and induces its
conformational change, resulting in the dissociation from SERCA2a. At the
transmembrane level, mutational study showed that mutations of amino acid residues on
one side of PLN monomer lose the inhibitory effect on SERCA2a activity without
changing monomer-pentamer stability, while mutations of amino acid residues on the
other side augment their inhibitory effect with the enhancement of monomer formation.
We therefore proposed that the PLN monomer is the functional, inhibitory form, and that
the pentamer represents a less active or inactive reservoir of subunits. On the basis of
this idea, reversible inhibition involves dissociation of the PLN pentamer, formation of
the inhibited PLN/SERCA2a heterodimer, and dissociation of the heterodimer. Thus, the
monomer-pentamer conversion is related to the PLN-SERCA2a interaction. Binding of
Ca to the transmembrane helices of SERCA2a and phosphorylation of the cytoplasmic
domain of PLN enhance PLN/SERCA2a heterodimer dissociation and activate
SERCA2a activity [6].
    The PLN gene, which is localized in human chromosome 6, consists of 2 exons and 5'
upstream regulatory region [10]. The expression of PLN is restricted to cardiac,
slow-twitch skeletal [11][12], and smooth muscles [13] (Table 1). There are at least
three kinds of Ca-ATPase genes expressed in sarcoplasmic reticulum. Type 1 gene,
SERCA1, is expressed in fast-twitch skeletal muscle. Type 2 gene, SERCA2, encodes
two alternatively spliced products, SERCA2a and SERCA2b. SERCA2a is expressed in
slow-twitch skeletal muscle and cardiac muscle, while SERCA2b is expressed in
smooth muscle and non-muscle tissues [14]. Type 3 gene, SERCA3, is found to be
expressed in a broad variety of both muscle and non-muscle tissues [15]. When cardiac
muscles are subjected to neurohumoral and hemodynamic stresses, the amounts of PLN
and SERCA2a mRNA expressed in cardiac sarcoplasmic reticulum are greatly altered
[16][17][18].
    In this study, we characterized the transcriptional regulation of the PLN gene. By
measuring the transcriptional activity of deleted PLN gene, we found the region
containing a CCAAT sequence is important for transcriptional activity of the PLN gene.
By using gel shift assays, we showed that this region associates with transcriptional
factors, NF-YA and NF-YB. Suppression of these two proteins in cardiac myocytes by
over-expression of antisense cDNAs encoding them decreased the transcriptional
activity of the PLN gene. We therefore concluded that transcription of the PLN gene is
regulated by the association of this region with NF-Y.

 
Experimental Procedures

Cell Culture
    Neonatal rat cardiac myocytes were isolated and cultured as described previously
[19]. Hearts were removed from 20 2-day-old neonatal Wistar-Kyoto rats. The excised
ventricles were cut into 1 to 2-mm cubes and shaken for 10 min at 37oC in 15 ml of
phosphate-buffered saline (PBS: NaCl, 137 mM; Na2HPO4, 10.6 mM; KH2PO4, 2.1
mM; and K2HPO4, 1.1 mM) containing collagenase (0.1% w/v). The dissociated cells
were suspended in 20 ml of Dulbecco's modified Eagle medium (DMEM) supplemented
with glucose (25 mM) and fetal bovine serum (FBS; 10% v/v) and then pre-plated for 1
h to selectively separate unattached myocytes. Finally, the unattached myocytes were
plated on culture dishes at a density of 3.1* 104 /cm2. After 2 days of culture in DMEM
with 10% FBS and Penicillin G (400 U/ml) under 5% CO2, myocytes were used for
transfections. All cell lines were obtained from American Type Culture Collection.
Mouse muscle myoblast cells (C2C12), rat glial tumor cells (C6) and mouse soleus
muscle myoblast cells (Sol8) were cultured in DMEM with 10% FBS and Penicillin G
(400 U/ml) under 5% CO2.

RNA analysis
    Total RNA was extracted by Isogen (Nippon Gene Co., Toyama, Japan). Fifteen
micrograms of total RNA were electrophoresed and blotted onto a Hybond-N+ nylon
membrane (Amersham Co., Arlington Heights, IL). The blots were hybridized with
32P-labeled, 280 bp cDNA probe of PLN. After hybridization, the blots were washed,
then exposed to x-ray films with an intensifying screen.

Construction of PLN-luciferase Expression Vectors
    A low background promoterless luciferase expression vector, PGV-B (Toyo Ink Co.,
Tokyo, Japan), and the genomic clone of rabbit PLN gene containing the 3121 base pair
(bp) from the transcriptional start site of 5' upstream region were used for preparation of
the expression constructs. A SacI-StyI fragment from -3121 to +84 bp of the clone was
sub-cloned into pBluescript II KS+ (Stratagene, La Jolla, CA). Using EXOIII/Mung bean
nuclease deletion system (Stratagene), successively 5'-truncated segments were
generated and then each segment was ligated into the PGV-B luciferase reporter vector.
The luciferase constructs ligated with the fragment -2080 to +84 bp, -375 to +84 bp, -96
to +84 bp, and -78 to +84bp were designated as pLuc-2080, pLuc-375, pLuc-96, and
pLuc-78, respectively. The internally deleted constructs of pLuc-375, pLuc-375a
(lacking -115 to -96 bp), pLuc-375b (lacking -115 to -78 bp), pLuc-375c (lacking -96
bp to -78 bp), pLuc-375d (lacking -78 to -33 bp), pLuc-375e (lacking -29 to -2 bp), and
pLuc-375f (lacking -22 to -2 bp) were generated by PCR directed mutagenesis.

Transfection and Luciferase Assay
    Cultured rat cardiac myocytes were co-transfected with 1 g of the PLN-luciferase
constructs and 0.1 g of pRL-CMV (Promega, Inc., Madison, WI) for each of 3
cm-diameter plate using Lipofectin (Life Technologies, Inc., Gaithersburg, MD). At 12 h
after transfection, cells were washed once with serum-containing medium and then
incubated in DMEM and 10% FBS under 5% CO2 at 37oC for another 24 h prior to
harvesting for luciferase assays. Luciferase assays were performed using a dual
luciferase reporter assay system (Promega, Inc.). The treated cells from each well were
washed with PBS, and then 250 l of lysis buffer was added. The plates were rocked
gently for 10 min using an orbital shaker, and cells were harvested by scraping. The
cells were then pelleted by centrifugation and the supernatant was used for luciferase
assays. The firefly luciferase activities of the test plasmids transfected and the renilla
luciferase activities of the pRL-CMV transfected cells were measured by a luminometer.
The luciferase activity of the test plasmid was normalized to the renilla luciferase
activity of pRL-CMV to account for variations in the efficiency of transfection. Each
transfection and luciferase assay experiment was performed in triplicate.

Preparation of Nuclear Extracts
    Nuclear extracts from neonatal rat hearts were prepared by a modification of the
method of Dignam et al, [20]. Ventricles from 100 neonatal rats were removed, cut, and
washed three times with PBS. The heart tissues were homogenized in 10 ml of buffer A
(10 mM HEPES; pH 7.9, 1.5 mM MgCl2, 10 mM KCl, and 0.5 mM DTT) with a Dounce
homogenizer. The homogenate was centrifuged at 6000 rpm for 15 min, and the pellet
was resuspended in 10 ml of buffer A and rehomogenized. The homogenate was
centrifuged at 6000 rpm for 15 min, and the pellet was resuspended in 5 ml of extraction
buffer (20 mM HEPES; pH 7.9, 25% glycerol, 0.55 M NaCl, 1.5 mM MgCl2, 0.2 mM
EDTA, 0.5 mM DTT, and protease inhibitor complete (Boehringer Mannheim GmbH,
Mannheim, Germany)) and homogenized. The extract was centrifuged in a micro
centrifuge at maximal speed for 30 min, and the supernatant was dialyzed against 1 litter
of dialysis buffer (40 mM KCl, 15 mM HEPES; pH 7.9, 1 mM EDTA, 0.5 mM PMSF,
0.5 mM DTT, and 20% glycerol) for 4 h. The extract was centrifuged in a micro
centrifuge for 10 min, and the supernatant was stored at -80oC prior to use.

Gel Shift Assay
     For gel shift assays with cardiac myocyte nuclear extracts, 6 _g of the nuclear
extracts described above were preincubated at 25oC in a 20 _l reaction mixture (10 mM
Tris-HCl; pH 7.5, 50 mM NaCl, 10 mM MgCl2, 0.5 mM DTT, 10% glycerol, and 0.1
_g/_l of poly(dI-dC)-poly(dI-dC)). After 10 mim, 1 _l of approximately 2* 104 cpm of a
32P-end-labeled nucleotide probe was added and the incubation continued for 20 min.
The mixtures were electrophoresed through a 4% polyacrylamide gel in 1* gel buffer (7
mM Tris-HCl; pH 7.5, 3 mM sodium, acetate and 1 mM EDTA) at 12.5 V/cm of gel at
4oC for 2 h. For competition assays, a large excess of unlabeled double-stranded
nucleotide was incubated together with the nuclear extract prior to adding the
32P-labeled probe. To compare the binding affinity of 32P-labeled probes with
homologous sequences, the oligonucleotides NF-Y [21], EFI [22], CTF/NF1 [23], CP2
[24] and C/EBP [25] were used for competition assays. Sequences of these probes
were:
NF-Y: 5'-ACTTTTAACCAATCAGAAAAATGTTT-3';
EFI: 5'-TACTTCCACCAATCGGCATGCACGGT-3';
CTF/NF1: 5'-TCGAATGGCGCGCAGCCAATGGTAGGCC-3';
CP2: 5'-CGCGCCACCCAATGGGGGTAAGAGCT-3'; and
C/EBP: 5'-GTTCGAATTCGCCAATGACAAGACGCT-3';
For supershift assays with anti-NF-YA, anti-NF-YB [26], or anti-YB-1 antibodies [27],
1 _l of affinity purified antibody was incubated with the probe-nuclear extract mixtures
for another 30 min prior to gel electrophoresis. For gel shift assays with bacterially
expressed proteins, about 0.1 _g of in vitro expressed NF-YA, NF-YB and/or YB-1
were used in the reaction mixture described above.

PCR Cloning of NF-YA, NF-YB and YB-1 cDNAs
    In order to obtain cDNAs encoding NF-YA, NF-YB, and YB-1, a reverse
transcription PCR cloning method was used. Full-length mouse NF-YA and NF-YB
cDNA were cloned by PCR using mRNA isolated from mouse liver. PCR primers were
designed according to published sequences with an XbaI site at the 5'-end to facilitate
cloning [28]. A fragment of the rat YB-1 cDNA was cloned by PCR using mRNA
isolated from rat heart. Using this 32P-labeled fragment as a probe, the full-length rat
YB-1 cDNA was obtained by screening from a rat heart cDNA library in the _ZAPII
vector (Stratagene). The cloned cDNAs were verified by nucleotide sequencing [28]
[29]. The cDNAs of NF-YA, NF-YB, and YB-1 were ligated into the pET-28a(+)
bacterial expression vector (Novagen, Madison, WI) encoding a His Tag and T7 Tag
epitopes in the N-terminus. Finally, the cDNAs of NF-YA and NF-YB were ligated into
a pcDNA3 mammalian expression vector (Invitrogen, San Diego, CA) in both the sense
and antisense orientation.

In vitro Protein Expression of NF-YA, NF-YB and YB-1
    pET-28a(+) (Novagen) constructs harboring each of NF-YA, NF-YB, and YB-1
cDNAs were transformed into Escherichia coli BL21(DE3)pLysS, and the cells were
grown to an optical density of 1.0 in LB-chloramphenicol-kanamycin medium and
diluted 1:100 in LB-chloramphenicol-kanamycin medium. When the optical density
reached 0.6 to 1.0, IPTG was added to a final concentration of 1 mM, and the culture
was shaken for another 3 h at 30oC for induction of the fusion proteins. Expressed
proteins carrying the His Tag and T7 Tag epitopes were affinity purified on His Bind
columns as outlined in the protocol for the pET system (Novagen). Purified proteins
were concentrated and equilibrated with nuclear extraction buffer at concentration of 0.1
_g/_l by using Centricon 10 columns (Amicon, Beverly, MA).

Results

Characterization of a cis-acting element that controls the PLN gene expression in
cardiac myocytes

    To investigate the regions responsible for the transcriptional regulation of the PLN
gene, primary cultured neonatal cardiac myocytes were transiently transfected with
luciferase constructs containing different lengths of the 5'-flanking region of the PLN
gene obtained by progressive deletion of the 5'-end. The transcriptional levels of these
constructs in transfected cardiac myocytes were analyzed by measuring luciferase
activities in cell extracts. To compare the luciferase activities of constructs, the activity
of each construct is presented as relative value to the activity of control pRL-CMV (Fig.
2). Significant luciferase activity was detected in the cardiac myocytes transfected with
the luciferase constructs containing the 5'-flanking region up to -3.1 k bp. Deletions of
this region from -3.1 k bp to -96 bp had no significant effect on the luciferase activities
of cells transfected with these constructs compared with the full-length construct.
However, further deletion from -96 to -78 bp resulted in a marked loss of luciferase
activity. These results indicated that transcription of the PLN gene is enhanced by the
region from -96 to -78 bp.
    Nucleotide sequence analysis was performed to identify the consensus cis-elements
in the region from -96 to 78 bp (Fig. 3). A CCAAT motif, which has been identified to
be involved in the binding of a small group of transcription factors such as
CCAAT-binding factor, is located at -93 bp. To further determine whether the region
from -96 to -78 bp is responsible for transcription of the PLN gene, six internal
deletions were generated as shown in Fig. 4. The pLuc-375b and pLuc-375c, which
were devoid of the CCAAT sequence, showed an almost total loss of luciferase activity,
while pLuc-375e and pLuc-375f, lacking the TATA-like box, showed same activity as
pLuc-375. Other internal deletion constructs, pLuc-375a and pLuc-375d, lacking the
flanking sequence of the CCAAT sequence showed an approximately 50% loss of the
activity. These results indicated that the region from -96 to -78 bp including the CCAAT
core sequence, but not the TATA-like box, is responsible for the positive regulation of
the PLN gene transcription, with flanking sequences of the CCAAT sequence being less
effective in cardiac myocytes.
 
 

A role of a cis-acting element on the PLN gene expression in other cell lines

    Previous studies have demonstrated that the PLN mRNA was expressed in both
cardiac muscles and slow-twitch skeletal muscles, but not in fast-twitch skeletal muscles
[9] [30] [31]. To investigate whether the region from -96 to -78 bp mediates higher
transcriptional activity of the PLN gene in cardiac myocytes, transcriptional levels of the
PLN-luciferase constructs in other cell lines lacking endogenous PLN expression were
measured. C2C12 and Sol8 skeletal muscle cell lines and C6 glioma cell lines were used
for the transient expression assays, as these cell lines did not express the PLN mRNA
endogenously, in contrast to the cardiac myocytes (Fig. 5A). The PLN-luciferase
constructs were transfected into these cell lines and then luciferase activities of the
constructs were measured. When the luciferase activity of the construct was normalized
by that of pLuc-2080 in each cell type, all types of cells showed similar pattern of
luciferase activities; the increased activities of pLuc-375 and pLuc-96 and the decreased
activities of pLuc-78 (Fig. 5B). However, when the luciferase activity of pLuc-96 was
normalized by that of the TK-promoter or the CMV-promoter in each cell type, the
normalized luciferase activity of pLuc-96 in cardiac myocytes was approximately 2.5 to
20-fold higher than that in C2C12, Sol8, and C6 cell lines (Fig. 5C). Thus, although the
transcriptional activity of the region from -96 to -78 bp was not the cardiac-specific, this
region may be related to the enhancing activity of the PLN gene expression in cardiac
myocytes.
 
 

NF-Y binds to a cis-acting element of the PLN gene

    To examine the binding of transcriptional factors to cis-acting elements of the PLN
gene, gel shift assays were performed (Fig. 6). When 32P-labeled fragment A from -131
to +81 bp and the nuclear extracts were incubated, a single DNA-protein complex was
identified. Addition of 100-fold molar excesses of unlabeled fragment A probe inhibited
the appearance of the complex, indicating that this complex is specific to fragment A.
32P-labeled fragment B from -131 to -78 bp also showed binding to the nuclear extracts,
but fragment C from -78 to +81 bp lacking the CCAAT motif, showed no binding. The
32P-labeled 26-mer oligonucleotide (OligoI) spanning -101 to -76 bp, also showed
complex formation, as is indicated by arrowhead in Fig. 7. The indicated band, but not
other bands, was competitively abolished by excess of unlabeled self oligonucleotides.
These results indicated that the region from -101 to -76 bp represents a
sequence-specific binding site for some transcriptional factor(s). Since the sequence of
the region from -96 to -78 bp has a high homology to several transcriptional
factor-binding motifs containing the CCAAT core sequence (Table 2), we compared the
binding affinity of the OligoI probe to the nuclear extracts in the presence of competitive
oligonucleotides homologous to the region from -96 to -78 bp. An excess of the
unlabeled NF-Y oligonucleotide carrying the inverted CCAAT sequence of the major
histocompatibility complex class II E_ gene or the EFI oligonucleotide (YB-1 binding
oligonucleotide) carrying the inverted CCAAT sequence of the RSV LTR strongly
reduced the binding of OligoI probe to the nuclear extracts. Excesses of unlabeled CP2,
C/EBP and NF-I oligonucleotides did not significantly reduce the binding activities (Fig.
7). These results indicated that the region from -96 to -78 bp displays sequence bound to
nuclear proteins NF-Y and/or YB-1.
    To determine whether nuclear proteins NF-Y and/or YB-1 participate in formation of
the OligoI-nuclear protein complex, antibody inhibition assays were performed with
antibodies against NF-YA, NF-YB, and YB-1 (Fig. 8). The formation of the
OligoI-nuclear protein complex was inhibited by antibodies against NF-YA and NF-YB
but was not inhibited by antibodies against YB-1. When the antibodies against NF-YB
were added to the reaction mixture, a super-shifted band appeared, as indicated by open
arrow head (Fig. 8 lane 3). These results indicated that NF-YA and NF-YB are
components of the OligoI-nuclear protein complex but YB-1 is not.
    In order to confirm this result, gel shift assays using in vitro bacterially expressed
NF-YA, NF-YB, and YB-1 were performed (Fig. 9). The combination of NF-YA and
NF-YB bound to OligoI with high affinity (open arrowhead), but any other combinations
among two of YB-1, NF-YA, and NF-YB could not bind to OligoI. Neither recombinant
NF-YA, NF-YB, nor YB-1 alone could bind to OligoI. Thus, the recombinant NF-YA
and NF-YB could bind to the region from -101 to -76 bp by forming heterodimer. This
finding is in agreement with the previous result that the dimer formation of NF-YA and
NF-YB is necessary for their binding to the target sequence [28] [32]. However, the
mobility of the OligoI-recombinant NF-YA and NF-YB complex (open arrowhead)
appeared to be faster than that of the OligoI-nuclear protein complex (closed
arrowhead), suggesting that the third factor may be involved in OligoI-nuclear protein
complex in gel shift assay.
 
 

Co-expression of sense and antisense cDNAs encoding NF-YA and NF-YB modulates
transcriptional activity of the PLN gene

    Biochemical assays demonstrated that NF-YA and NF-YB are components of the
DNA-nuclear extract complex. In order to confirm the functional role of NF-Y on
expression of the PLN gene, we constructed pcDNA3 vectors containing NF-YA or
NF-YB cDNAs in the sense or antisense orientation and co-transfected them with the
luciferase construct pLuc-375 into cardiac myocytes. To compare the luciferase
activities of pLuc-375 among experimental groups, the activity of each group is
presented as relative value to the activity of control pRL-CMV (Fig. 10). When the
pcDNA3 vectors containing the sense NF-YA or NF-YB cDNA were co-transfected
with pLuc-375 into cardiac myocytes, the luciferase activity was the same level as that
observed when pcDNA3 alone was co-transfected with pLuc-375. By contrast, when the
pcDNA3 vectors containing antisense NF-YA or NF-YB cDNAs were co-transfected
with pLuc-375, the luciferase activity of pLuc-375 was reduced to approximately 50%
of the activity observed after co-transfection with the pcDNA3 vector alone (Fig. 10).
These results indicated that the NF-YA and NF-YB functionally participate in
transcriptional regulation of the PLN gene.

Discussion

    Being an intrinsic regulatory factor of SERCA2 in cardiac muscle SR, the
structure-function relationship of PLN and its genomic structure have been studied
extensively [2][3][4]. Although a great deal is known about the mRNA and
post-translational regulation of PLN [16][17][18], little is known about mechanisms
regulating the PLN gene transcription. In this study, using different experimental
approaches, we identified the 5'-flanking region from -96 to -78 bp which can confer the
transcriptional activity of the PLN gene, and characterized two nuclear factors, NF-YA
and NF-YB, which preferentially bind to this region and govern the PLN gene
expression in cardiac myocytes.
    By measuring the luciferase activities of a series of deletion constructs of the PLN
gene in cardiac myocytes, we found that the region from -96 to -78 bp is essential for its
transcriptional activity and that this region causes significantly higher level of luciferase
activity in cardiac myocytes than other types of cells. By nucleotide sequence analysis,
the region from -96 to -78 bp contains a sequence highly homologous to transcriptional
factor-binding motifs containing the CCAAT core sequence in a reverse orientation
(Table 2). Although the CCAAT sequence is known to be a crucial component of several
promoters [33], this motif may have a more complicated function, including modulation
of transcriptional levels to achieve tissue-specific expression of several genes during
differentiation [34] [35] [32] [36]. Together with the result showing that the deletion
construct lacking the TATA-like sequence has the wild-type transcriptional activity, we
conclude that this region containing the CCAAT sequence functions as a promoter
element for the PLN gene expression.
    Gel shift assays showed that the region from -101 to -76 bp specifically binds to
nuclear proteins extracted from rat hearts. When oligonucleotides corresponding to the
family members of the CCAAT-related motif and antibodies against several CCAAT
binding factors were used in the gel shift assay, NF-YA and NF-YB were found to be
transcriptional factors bound to the region from -101 to -76 bp. In the presence of
antisense cDNA encoding NF-YA and NF-YB in cardiac myocytes, the transcriptional
activity of the coexpressed PLN gene construct was reduced to approximately 50% of
that in the presence of sense cDNAs. We therefore conclude that the region from -96 to
-78 bp of the PLN gene associates with NF-Y, one of the CCAAT-binding factors,
thereby regulating transcription of the PLN gene.
    NF-Y was identified as a protein in the widespread of the tissues which binds to the
CCAAT-related sequence of the major histocompatibility complex class II gene [37],
and found to be identical to CBF for the _2(I)collagen gene [24], and CP1 for the
_-globin gene [28]. NF-Y consists of 40kDa and 32kDa subunits (the A and B subunits,
respectively) and functions as a dimer to regulate the basal transcriptional activity of
several genes [28][32]. In agreement with this notion, the region from -96 to -78 bp
showed the enhancement of the PLN gene expression by associating with both of NF-YA
and NF-YB. However, it remains unclear as to how the association between the
CCAAT-motif and ubiquitous NF-Ys regulates the tissue-specific PLN gene expression.
One possibility is that the specific cis-elements, located in the region other than that we
analyzed in this study, may be required. Indeed, in vivo analysis indicated that a large
fragment of the PLN gene covering 7 kb of 5'-flanking region, the first exon, and the first
intron is capable of the cardiac specific PLN gene expression in transgenic mice [38].
Alternatively, the third factor may be involved in the transcriptional activity of the
region from -96 to -78 bp in cardiac myocytes, in addition to NF-Y. Observed
differences between mobility of OligoI-recombinant NF-YA and NF-YB complex and
that of OligoI-nuclear protein complex suggest that the third unknown factor(s) is
involved in the OligoI-nuclear protein complex. Indeed, the interaction of the ubiquitous
transcriptional factor with the third factor has been found to regulate the tissue-specific
transcriptional regulation of several genes [39][40][41]. The Pit-1 POU domain factor
can synergistically interact with Oct-1 to control the expression of pituitary specific
genes, such as growth hormone gene [39]. The hepatocyte-specific factor HNF-1 can
interact with Oct-1 to regulate the activity of the hapatitis B viral promoter gene [40].
Likewise, the YB-1 can interact with cardiac-specific co-factor CARP to regulate the
expression of cardiac-specific gene, such as myosin light chain-2v gene [41]. Although
the identity of the third factor is not determined in this study, it will become of future
interest to study the functional role of this factor on the tissue-specific distribution of
PLN.
    In conclusion, we demonstrated that the cis-acting element involved in the
transcriptional activity of the PLN gene is the NF-Y binding motif in the promoter region
from -96 to -78 bp and the NF-Y is involved in this regulatory mechanism.

Acknowledgments

We are particularly grateful to Drs. D. Mathis and R. Mantovani, CNRS INSERM
Universit_ Louis Pasteur, for their generous gifts of the antibodies against NF-YA and
NF-YB. We also thank Dr. K. R. Chien, University of California, San Diego, School of
Medicine, for gift of the antibody against YB-1. This study was supported partly by the
International Human Frontier Science Program Organization and Grants-in-Aid for
Scientific Research from the Ministry of Education, Science and Culture of Japan.

References
 
[1] Tada, M., Kirchberger, M.A. and Katz, A.M. (1975) Phosphorylation of a
22,000-dalton component of the cardiac sarcoplasmic reticulum by adenosine
3':5'-monophosphate-dependent protein kinase, J Biol Chem 250, 2640-7. [2] Tada, M.
and Katz, A.M. (1982) Phosphorylation of the sarcoplasmic reticulum and sarcolemma,
Annu Rev Physiol 44, 401-23. [3] Tada, M. (1992) Molecular structure and function of
phospholamban in regulating the calcium pump from sarcoplasmic reticulum, Ann N Y
Acad Sci 671, 92-102; discussion 102-3. [4] MacLennan, D.H., Toyofuku, T. and
Kimura, Y. (1997) Sites of regulatory interaction between calcium ATPases and
phospholamban, Basic Res Cardiol 92 Suppl 1, 11-5. [5] M. Tada and Y. Toyofuku
(1998) Molecular regulation of phospholamban function and expression, Trends in
Cardiovascular Medicine, 8, 330-340. [6] Kimura, Y., Kurzydlowski, K., Tada, M. and
MacLennan, D.H. (1997) Phospholamban inhibitory function is activated by
depolymerization, J Biol Chem 272, 15061-4. [7] James, P., Inui, M., Tada, M., Chiesi,
M. and Carafoli, E. (1989) Nature and site of phospholamban regulation of the Ca2+
pump of sarcoplasmic reticulum, Nature 342, 90-2. [8] Toyofuku, T., Kurzydlowski, K.,
Tada, M. and M