Multiple NF-Y-binding CCAAT boxes are essential for
transcriptional regulation of the human C7orf24 gene,
a novel tumor-associated gene
Yuji Ohno
1
, Akira Hattori
1
, Masamichi Ueda
2
, Susumu Kageyama
3
, Tatsuhiro Yoshiki
4
and
Hideaki Kakeya
1
1 Department of System Chemotherapy and Molecular Sciences, Division of Bioinformatics and Chemical Genomics, Graduate School of
Pharmaceutical Sciences, Kyoto University, Japan
2 Institute for Virus Research, Kyoto University, Japan
3 Department of Urology, Shiga University of Medical Science, Japan
4 Department of Clinical Oncology, Kyoto Pharmaceutical University, Japan
Keywords
C7orf24; CCAAT box; NF-Y; tumor
biomarker
Correspondence
A. Hattori & H. Kakeya, Department of
System Chemotherapy and Molecular
Sciences, Division of Bioinformatics and
Chemical Genomics, Graduate School of
Pharmaceutical Sciences, Kyoto University,
Sakyo, Kyoto 606-8501, Japan
Fax: +81 75 753 4591
Tel: +81 75 753 9267 & +81 75 753 4524
E-mail: ahattori@pharm.kyoto-u.ac.jp &
scseigyo-hisyo@pharm.kyoto-u.ac.jp
(Received 28 June 2011, revised 12 August
2011, accepted 22 August 2011)
doi:10.1111/j.1742-4658.2011.08314.x
Human chromosome 7 ORF 24 (C7orf24) has been identified as a tumor-
related protein, and shown to be a c-glutamyl cyclotransferase. In the cur-
rent study, we characterized the promoter region of the human C7orf24
gene to explore the transcriptional regulation of the gene. We revealed that
the human C7orf24 promoter is a TATA-less promoter, containing five
CCAAT boxes aligned in a forward orientation. By performing a luciferase
reporter assay with 5¢-deleted and site-directed mutated constructs in HeLa,
MCF-7 and IMR-90 cells, we found that three proximal CCAAT boxes
are important for basal transcription. Electrophoretic mobility gel shift
assay and chromatin immunoprecipitation assay demonstrated that NF-Y
specifically bound to all three CCAAT boxes. In addition, the mRNA and
protein expression levels of C7orf24 were significantly reduced in HeLa
cells depleted of NF-YB, a subunit of NF-Y. These results suggested that
NF-Ys bound to the three proximal CCAAT boxes play a central role in
the transcription of the gene. Furthermore, as in the case of other genes
transcribed under the control of multiple NF-Ys, such as human E2f1 and
cyclin b1, the C7orf24 gene expression profile oscillated during the cell
cycle, implying that C7orf24 is a novel cell cycle-associated gene.
Introduction
By comparison of protein expression profiles between
bladder urothelial carcinoma tissue and normal urothe-
lium tissue obtained from patients, Kageyama et al. [1]
identified 15 cancer-specific spots that could constitute
a diagnostic marker for bladder cancer. They demon-
strated that one of the spots was a hypothetical gene
product of chromosome 7 ORF 24 (C7orf24), the
function of which has not been determined. They also
found that ectopic expression of C7orf24 promoted the
proliferation of murine normal fibroblasts, and silenc-
ing of the gene by small interfering RNA (siRNA)
showed an antiproliferative effect on cancer cell lines,
suggesting a role of C7orf24 in cell proliferation. In
addition, it was reported that C7orf24 was expressed
in a range of cancers, such as prostate and breast can-
cers [2–6].
Abbreviations
ARE, antioxidant-responsive element; C7orf24, chromosome 7 ORF 24; ChIP, chromatin immunoprecipitation; EMSA, electrophoretic
mobility gel shift assay; siRNA, small interfering RNA.
4088 FEBS Journal 278 (2011) 4088–4099 ª2011 The Authors Journal compilation ª2011 FEBS
On the other hand, C7orf24 was identified as an
apoptosis-induced factor by Masuda et al. [7]. By mon-
itoring cytochrome c-releasing activity in a cell-free
system, they purified C7orf24 from the cytosol fraction
of geranylgeraniol-treated U937 cells. They also
reported that ectopic expression of C7orf24 in HeLa
cells resulted in the release of cytochrome cfrom mito-
chondria, which, in turn, promoted apoptosis. These
results suggested that C7orf24 plays a role in the apop-
totic pathway induced by geranylgeraniol.
Recently, Oakley et al. [8] also identified C7orf24 as
ac-glutamyl cyclotransferase. They then characterized
recombinant C7orf24, and showed that it converted
c-glutamyl cysteine to 5-oxoproline and cysteine,
suggesting that C7orf24 may play an essential role in
glutathione homeostasis. Furthermore, they resolved
the crystal structure of human C7orf24 and showed a
structural similarity between the protein and BtrG, a
c-glutamyl cyclotransferase of Bacillus circulans, sup-
porting their functional relationship as c-glutamyl
cyclotransferases [9].
Although these findings suggested a biological signif-
icance of C7orf24, its definitive role in cellular function
remains unclear. In this study, we investigated the reg-
ulatory mechanism of the human C7orf24 gene to gain
basic insights into the physiological relevance of the
gene product. By performing a luciferase reporter gene
assay, we characterized the promoter region of the
human C7orf24 gene. We found that three CCAAT
boxes located proximally to the initiator element are
functional, and that NF-Y plays an essential role in
basal transcription of the gene through binding to
these CCAAT boxes.
Results
Characterization of the human C7orf24 gene
promoter
To characterize the 5¢-flanking region of the human
C7orf24 gene, we referred to the nucleotide sequence
deposited in GenBank (accession no. NW_001839003).
We obtained the 2127-bp fragment of the 5¢-flanking
region of the gene by genomic PCR, cloned it into the
pGL3-Basic reporter plasmid, and then confirmed
the sequence by DNA sequencing. This region contains
a sequence (CCAGTCC) that perfectly matches the
pyrimidine-rich initiator consensus sequence,
PyPy(A+1)N(T ⁄A)PyPy, in which Py represents a
pyrimidine residue (Fig. 1) [10]. In addition, five
CCAAT boxes were located in this flanking region,
whereas no canonical TATA box was found. It is
notable that all of the CCAAT boxes are aligned in a
forward orientation. Several potential regulatory cis-
elements bound by transcription factors, such as Nrf-2
(antioxidant response factor) and Sp1, were also identi-
fied within this region [11].
To determine the regulatory elements mediating the
activation of C7orf24 gene transcription, we initially
performed a luciferase reporter gene assay with HeLa
cells, which express large amounts of C7orf24. When
the reporter plasmid harboring the 2127 bp of the
C7orf24 promoter was introduced, a marked increase
in luciferase activity was found (> 40-fold increase in
pGL3-Basic-transfected cells) (Fig. 2A, construct a). In
contrast, when the cells were transfected with con-
struct kharboring the full-length 5¢-flanking region in
an inverted orientation, its luciferase activity was com-
parable to that of the promoter-less pGL3-Basic-trans-
fected cells (construct j). These results indicated that
this 2.1-kbp region contains a functional promoter
Fig. 1. The 5¢-flanking region of the human C7orf24 gene. (A) Sche-
matic structure of the human C7orf24 promoter. The thin line and a
closed box denote the 5¢-flanking region and exon 1 of the gene,
respectively. The CCAAT box and ARE-like motif are depicted as
open and hatched boxes, respectively. Nucleotide sequences
around the distal CCAAT motifs (C4 and C5) and ARE-like motif are
shown below. (B) Nucleotide sequence of the proximal promoter
region of the human C7orf24 gene. The exon sequence is shown
in upper-case, and that of the untranscribed region is given in
lower-case. The transcriptional initiation site is shown as +1. Tran-
scription factor-binding sites were searched for with TRANSFAC [37],
and are underlined. Arrowheads indicate the start points of the
constructs used for the luciferase reporter gene assay.
Y. Ohno et al. Transcriptional regulation of human C7orf24
FEBS Journal 278 (2011) 4088–4099 ª2011 The Authors Journal compilation ª2011 FEBS 4089
that regulates C7orf24 gene expression. Subsequently,
we generated sequential 5¢-deletion constructs and
examined their promoter activities to identify the prox-
imal promoter region (constructs b–i). Although sub-
stantial promoter activity was detected when chimeric
constructs containing the 5¢-flanking sequence
upstream from position )135 (constructs b–d) were
transfected, it gradually decreased in proportion to the
insert size (constructs e–h). In cells transfected with
construct i, no significant increase in luciferase activity
was observed. These results suggested that a minimal
promoter region is located 58 bp distal from the ini-
tiator.
We next examined the promoter activity of the
C7orf24 gene 5¢-flanking region in human breast carci-
noma MCF-7 cells (which express large amounts of
C7orf24) and human diploid fibroblast IMR-90 cells
(which barely express C7orf24) (Figs 2A and S1).
When construct aor construct bwas transfected into
MCF-7 cells, a significant increase in luciferase activity
was observed. Elimination of the distal part of the
5¢-flanking region resulted in a gradual reduction in
C7orf24 promoter activity, and the activity was com-
pletely abolished in construct h. A similar promoter
activity profile was obtained in IMR-90 cells, although
the maximum promoter activity was apparently lower
than that in HeLa and MCF-7 cells. It was noted that
constructs aand bshowed comparable promoter activ-
ity in all cells tested, and similar promoter activity pro-
files were obtained for all cells, despite different
expression levels of C7orf24 protein. Collectively, these
results indicated that the minimal promoter essential
for basal transcription of the C7orf24 gene is located
between positions )58 and +14 from the initiation
site, and that the proximal upstream region may be
involved in enhancement of human C7orf24 gene
expression.
As shown in Fig. 1A, the human C7orf24 promoter
contains a putative antioxidant-responsive element
(ARE) [12]. In addition, it has been suggested that
C7orf24 is involved in glutathione homeostasis though
c-Glu-Cys hydrolysis [8]. We thus investigated the role
Fig. 2. Significance of the CCAAT boxes in the human C7orf24 promoter. (A) Reporter plasmids harboring sequentially deleted fragments of
the C7orf24 promoter were transfected into HeLa, MCF-7 and IMR-90 cells. Firefly luciferase activity was measured as described in Experi-
mental procedures, and normalized to the Renilla luciferase activity of a cotransfected internal control plasmid, pCMV-Renilla. (B) The lucifer-
ase reporter assay was conducted with the plasmids harboring a mutation at the CCAAT motif of the C7orf24 promoter in HeLa, MCF-7 and
IMR-90 cells. A schematic representation of the reporter constructs is shown on the left. Open boxes indicate intact CCAAT boxes, and
crossed boxes indicate mutated CCAAT boxes (CCAAA sequence). Relative promoter activity is expressed as fold increase as compared
with that of the promoter-less pGL3-Basic vector. Error bars are the standard errors of quadruplicate results. Similar results were obtained in
five separate experiments. LUC, luciferase.
Transcriptional regulation of human C7orf24 Y. Ohno et al.
4090 FEBS Journal 278 (2011) 4088–4099 ª2011 The Authors Journal compilation ª2011 FEBS
of the element in C7orf24 gene expression. As it is well
known that a transcription factor, Nrf-2, binds to the
ARE and plays an essential role in the antioxidant
response [13,14], we explored the effect of Nrf-2 activa-
tion on C7orf24 promoter activity. When HeLa cells
transfected with construct awere exposed to t-butyl
hydroquinone, an Nrf-2 activator [15], no difference
was observed in luciferase activity between Nrf-2-acti-
vated cells and control cells (Fig. S2A). Furthermore,
although the mRNA level of a well-known antioxidant
response gene, heme oxygenase-1 (HMOX1), was mark-
edly elevated by t-butyl hydroquinone treatment [16],
no change was observed in C7orf24 mRNA levels, indi-
cating that oxidative stress did not affect C7orf24 gene
expression (Fig. S2B).
Significance of the three proximal CCAAT boxes
in the human C7orf24 gene promoter
The proximal promoter sequence from )205 to )35
contains three CCAAT sequences; therefore, we next
explored the significance of these motifs in C7orf24
gene transcription. We made a mutation in the proxi-
mal CCAAT boxes (C1, C2, and C3) to generate a
CCAAA sequence on the reporter plasmid a, and mea-
sured promoter activity in HeLa and MCF-7 cells
(Fig. 2B). Although a mutation at a single CCAAT
site did not show any effect on luciferase activity in
either type of cell, when mutations were introduced
into two CCAAT sequences, C7orf24 promoter activity
was significantly decreased, except for the combination
of C1 and C3 (construct a-mcm). Strikingly, mutations
in both C1 and C2 (construct a-cmm) caused a marked
decrease in C7orf24 promoter activity in HeLa cells.
When all three CCAAT motifs were replaced with
CCAAA sequences, a further decrease in promoter
activity was observed in both types of cell. Collec-
tively, these findings suggest the significance of all
three CCAAT boxes in C7orf24 gene expression. The
importance of all three proximal CCAAT boxes was
further demonstrated in IMR-90 cells. It was noted
that construct a-cmc, harboring a mutation only in C2,
showed reduced promoter activity in the cells.
Binding of NF-Y to CCAAT boxes in the proximal
promoter of the human C7orf24 gene
We next investigated the interaction of trans-acting
factor(s) on proximal CCAAT boxes in the C7orf24
gene promoter by electrophoretic mobility gel shift
assay (EMSA). As shown in Fig. 3A, when the oli-
goduplex probe for each CCAAT box was incubated
with nuclear extracts of HeLa cells, a single retarded
band was observed in each probe. Subsequently, its
binding specificity was examined by employing an
excess amount of unlabeled oligoprobe. A 100-fold
molar excess amount of unlabeled probe completely
inhibited the formation of complexes between nuclear
extracts and radiolabeled probes. When EMSA was
carried out with mutant oligonucleotide probes,
replacing the CCAAT motifs with the CCAAA
sequences, no retarded band was detected in any
mutant probes (Fig. 3B). Taken together, these
results indicated the sequence-specific formation of a
CCAAT motif-protein complex.
NF-Y is one of the major CCAAT-binding proteins,
and is also referred as CCAAT-binding factor [17]. It
is also known that NF-Y is not able to bind to the
CCAAA sequence, because the protein is the only
CCAAT-binding factor that absolutely requires the
pentanucleotide sequence CCAAT for DNA sequence
recognition [18,19]. Hence, we examined the binding of
NF-Y to the three CCAAT boxes (C1–C3) in the
proximal promoter of the gene. When a mAb against
subunit A of NF-Y (NF-YA) was employed, a super-
shift of the probe–protein complex was detected in
each probe (Fig. 3A), suggesting the binding of NF-Y
to each CCAAT box. Specific binding of NF-Y to
each CCAAT box of the C7orf24 proximal promoter
was also observed in IMR-90 cells (Fig. 3A,B).
We further evaluated direct interactions between
NF-Y and CCAAT boxes of the C7orf24 gene pro-
moter in vivo with a chromatin immunoprecipitation
(ChIP) assay. As shown in Fig. 3C, the proximal
C7orf24 promoter fragment containing three CCAAT
boxes (C1–C3) ()179 to )19) was specifically amplified
from the immunoprecipitate with antibody against
NF-YB, but not control IgG. Furthermore, no signifi-
cant amplification of the distal promoter of the gene
either with C4 and C5 CCAAT boxes ()1218 to
)1088) or without any CCAAT box ()1577 to )1467)
was observed in either immunoprecipitate. These data
strongly indicated the in vivo recruitment of NF-Y to
the proximal promoter region of the gene.
Involvement of NF-Y in C7orf24 gene expression
For further evaluation of the involvement of NF-Y in
C7orf24 gene transcription, we examined promoter
activity in cells depleted of NF-YB. By gene silencing
through an RNA interference mechanism, the NF-yb
mRNA level was decreased to nearly 15% (48 h after
transfection) in NF-yb-specific siRNA (NF-YB-1 and
NF-YB-2)-transfected HeLa cells (Fig. 4A). In NF-yb-
eliminated cells, the mRNA level of C7orf24 declined
to nearly 61% of that of mock-transfected cells. With
Y. Ohno et al. Transcriptional regulation of human C7orf24
FEBS Journal 278 (2011) 4088–4099 ª2011 The Authors Journal compilation ª2011 FEBS 4091
regard to the E2f1 gene, which is known to be tran-
scribed in an NF-Y-dependent manner [20], elimina-
tion of NF-YB reduced E2f1 mRNA expression to a
similar degree as observed for the C7orf24 mRNA
level (68% of control siRNA-transfected cells). In
accordance with the parallel reduction in Nf-yb and
C7orf24 mRNAs, C7orf24 protein expression was
shown to be significantly decreased in Nf-yb-eliminated
HeLa cells (Fig. 4B). These results strongly suggest
that NF-Y plays an important role in transcriptional
regulation of the human C7orf24 gene.
Oscillation of C7orf24 gene expression during cell
cycle progression
It has been reported that the transcription of some cell
cycle-associated genes, such as E2f1 and cyclin b1
(CCNB1), is mediated through the promoter harboring
multiple CCAAT boxes that are bound by NF-Y [20–
22]. Therefore, we analyzed the gene expression profile
of human C7orf24 with M phase-synchronized HeLa
cells. After release from thymidine ⁄nocodazole block-
ade, the C7orf24 mRNA level gradually increased from
the late stage of the G
1
phase and peaked in the mid-
S phase (9–12 h after release) (Figs 5A and S3A).
Although the expression level of each gene varied, this
cell cycle-dependent profile was similar to that of the
E2f1 gene, a typical S phase gene, but not of cyclin b1,
aG
2
⁄M phase gene [20,21]. Moreover, dynamic
recruitment of NF-Y to the C7orf24 promoter was
observed at late G
1
⁄early S phase, followed by an ele-
vation of the C7orf24 mRNA level (Fig. 5B). Fig-
ures 5C and S3B also show the cell cycle-dependent
oscillation of C7orf24 mRNA expression in IMR-90
Fig. 3. NF-Y binds to all proximal CCAAT boxes of the human C7orf24 promoter. Nuclear extracts prepared from HeLa and IMR-90 cells
were incubated with radiolabeled probes. Sequence-specific interactions between nuclear extracts and radiolabeled probes were analyzed in
a 4% native polyacrylamide gel. (A) An excess amount (100-fold) of the unlabeled probe was employed for competition. For the supershift
assay, antibody against NF-YA was added to the reaction mixture and incubated for 2 h at 4 C before addition of labeled probes. (B) EMSA
was carried out with mutated oligonucleotide probes. W, probe harboring an intact CCAAT sequence; M, invalid probe harboring a CCAAA
sequence. (C) Cross-linked chromatin complexes were immunoprecipitated with antibody against NF-YB or normal rabbit IgG. The amounts
of precipitated DNA fragments derived from the C7orf24 promoter, i.e. )179 to )19 containing C1–C3 (closed bar), )1218 to )1088 contain-
ing C4 and C5 (gray bar), and )1577 to )1467 lacking a CCAAT box (open bar), were measured by quantitative PCR. Data are expressed as
a percentage of the input DNA with the standard errors of quadruplicate results. Similar results were obtained in three separate experi-
ments. Ab, antibody.
Transcriptional regulation of human C7orf24 Y. Ohno et al.
4092 FEBS Journal 278 (2011) 4088–4099 ª2011 The Authors Journal compilation ª2011 FEBS