Tài liệu Báo cáo khoa học: Functional analysis of the basic helix-loop-helix transcription factor DEC1 in circadian regulation ppt

Functional analysis of the basic helix-loop-helix transcription factor
DEC1 in circadian regulation
Interaction with BMAL1
Fuyuki Sato
1
, Takeshi Kawamoto
1
, Katsumi Fujimoto
1
, Mitsuhide Noshiro
1
, Kiyomasa K. Honda
1
,
Sato Honma
2
, Ken-ichi Honma
2
and Yukio Kato
1
1
Department of Dental and Medical Biochemistry, Hiroshima University Graduate School of Biomedical Sciences, Hiroshima, Japan;
2
Department of Physiology, Hokkaido University Graduate School of Medicine, Sapporo, Japan
The basic helix-loop-helix transcription factor D EC1 is
expressed i n a circad ian manner in t he suprachiasmatic
nucleus where it s eems to play a role in regulating the
mammalian circadian rhythm by suppressing the
CLOCK/BMAL1-activated promoter. The interaction of
DEC1 with BMAL1 has been suggested a s one of the
molecular mechanisms of the suppression [Honma, S.,

Kawamoto, T., Takagi, Y., Fujimoto, K., Sato, F.,
Noshiro, M., Kato, Y. & Honma, K. (2002) Nature 419,
841–844]. Deletion analysis of DEC1 demonstrated that
its N-terminal region, which includes the basic helix-loop-
helix domain, was essential for bo th the suppressive
activity a nd t he intera ction w ith B MAL1, as DEC1
lacking the basic region did not show any suppression or
interaction. Furthermore, we found that Arg65 in the
basic region, which is conserved among group B b asic
helix-loop-helix proteins, was responsible for the sup-
pression, for the interaction with BMAL1 and for its
binding to CACG TG E-boxes. However, substitution of
His57 for Ala significantly reduced the E-box binding
activity of DEC1, although it did not affect the inter-
action with BMAL1 or suppression of CLOCK/BM AL1-
induced transcription. On the other hand, the basic
region-deleted DEC1 acted in a dominant-negative
manner for DEC1 activity, indicating that the b asic
region was no t required f or homodimer formation of
DEC1. M oreover, mutant DEC1 also counteracted
DEC2-mediated s uppressive activity in a dominant-neg-
ative manner. The heterodimer f ormation of DEC1 and
DEC2 was confirmed by pull-down assay. These findings
suggest that the basic region of DEC1 participates in the
transcriptional regulation through a protein–protein
interaction with BMAL1 and DNA binding to the
E-box.
Keywords:DEC1;DEC2;BMAL1;circadianrhythm;clock.
Circadian rhythms are regulated by a m olecular clock(s),
which h as an endogenous period of  24 h and synchron-

izes to the 24 h period after light entrainment. In mammals,
the clock genes Clock, Bmal1, Per and Cry, and their protein
products, comprise a molecular feedback loop in which a
CLOCK/BMAL1 h eterodimer b inds to a CACGTG E-box
and activat es transcription o f Per and Cry [1,2]; protein
products of Per and Cry in turn suppress the transactivation
by CLOCK/BMAL1 [3,4]. This core feedback loop appar-
ently g enerates a 24 h period in the molecular oscillator.
Furthermore, another feedback loop has been reported to
control the rhythmic expression of Bmal1: expression o f
Rev-Erba is inducible by the CLOCK/BMAL1 heterodimer,
and its protein product suppresses the expression of Bmal1
[5,6]. These two feedback loops may b e interlocked to
stabilize the circadian core loop system.
DEC1 (bhlhb2) and DEC2 (bhlhb3) are basic helix-loop-
helix (bHLH) transcription factors which bind to CAC-
GTG E-boxes and suppress transcription from target genes
[7–12]. Expression of DEC1 and DEC2 showed circadian
rhythms in most organs, including the suprachiasmatic
nucleus (SCN) [7,13], and Dec1 expression in the SCN was
enhanced by a light pulse in a phase-dependent manner
similar t o Per1. M oreover, DEC1 and D EC2 suppressed
Per1 transactivation by CLOCK/BMAL1 through com-
petition for binding to E-boxes and/or protein–protein
interactions of DECs with BMAL1 [ 7]. Furthermore, w e
recently demonstrated the existence of a novel autofeedback
loop associated with Dec1 transcription, with CLOCK/
BMAL1 a s positive elements and DECs as negative
elements [11]. T hree CACGTG E-boxes in the Dec1
promoter were responsible for the rhythmic expression of

Dec1, and the feedback loop of DEC1 (as well as that of
BMAL1) might be interlocked with the core feedback loop
to constitute a network of the circadian clock system
[11,12,14]. In fact, the circadian rhythms of Dec1 and Dec2
expression have been shown to be completely disrupted in
the SCN a nd some o ther tissues of Clock/Clock mutan t
Correspondence to T. Kawamoto, Department of Dental and Medical
Biochemistry, Hiroshima University Graduate School of Biomedical
Sciences, Hiroshima 734–8553, Japan. Fax: +81 82 257 5629,
Tel.: +81 82 257 5629, E-mail:
Abbreviations: A D, activation domain; bHL H, basic helix-loop-helix;
DNA-BD, DNA-binding domain; GST, glutathione S-transferase;
HDAC, histone deacetylase; SCN, suprachiasmatic nucleus;
TK, thymidine kinase; TSA, trichostatin A.
(Received 18 March 2004, revised 14 September 2004,
accepted 24 September 2004)
Eur. J. Biochem. 271, 4409–4419 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04379.x
mice [11,12,15], whereas overexpression of Dec1 decreases
mRNA levels of Clock-dependent genes such as Per2, Dbp,
Dec1 and Dec2 [11,16]. In a recent study of DEC1 in
knockout mice (Stra13
–/–
), Gre
´
chez-Cassiau et al. [17]
identified significant changes in the expression of liver
genes, including several clock-controlled genes, although n o
change was observed in clock gene expression in the liver.
DEC1 is thus confirmed as one of the regulators o f t he
circadian clock system, at least in some peripheral organs.

As Dec2 expression was f ound to increase in the mutant
mice, the disruption of DEC1 might be compensated by
DEC2. Further investigation is required to clarify the
functions of DEC1 and DEC2.
The suppressive activities of DEC1 and DEC2 against
CLOCK/BMAL1-activated promoters are strong com-
pared with the transcriptional suppression by DEC1
without CLOCK/BMAL1 activation ( T. Kawamoto,
unpublished data). I n t his s tudy, w e examined, by using
various DEC1 mutants, the relationship between the
DEC1–BMAL1 interactio n and transcriptional suppression
by DEC1. We also evaluated the E-box-binding activity of
these mutants. Our results showed that the region of DEC1
required for transcriptional suppression by DEC1 coincided
with that for interaction with BMAL1 and binding to the
E-boxes, indicating the importance of t his region in DEC1
for its suppressive activity against CLOCK/BMAL1-
induced transcription.
Materials and methods
Plasmid constructions
To obtain deleted fragments from the 3¢-terminus of human
Dec1 (hDec1)cDNA[18],a5¢-terminus pr imer (5 ¢-AAG
CTTCACCATGGAGCGGATCCCCAGCGCGCAACC
AC-3¢)anda3¢-terminus primer (one of 5 ¢-TCTA
GACTAGGAGCTGATCAGGTCACTGCTAGTGAAA
TGG-3¢,5¢-TCTAGACTACCCACTCGAGTGAGCGA
AAGTCCGCTGG-3¢ or 5¢-TCTAGACTATTGACCTG
TTTCGACATTTCTCCCTGACAGCTC-3¢)wereused
for PCR amplification, with hDec1 cDNA as a template.
For the amplification of deleted fragments from the

5¢-terminus of hDec1 cDNA, a 3¢-terminus primer
(5¢-G CAGCAGGATCCTCTAGAGAGTTTAGTCTT
TG-3¢)anda5¢-terminus primer (one of 5¢-AAGCT
TCACCATGTACCCTGCCCACATGTACCAAGTG
TAC-3¢,5¢-AAGCTTCACCATGCCGCACCGG CTC
ATCGAGAAAAAGAG-3¢,5¢-AAGCTTCACCATG
GCAGTGGTTCTTGAACTTACCTTGAAGC-3¢ or
5¢-AAGCTTCACCATGATTGCCCTGCAGAGTGG
TTTACAAGCTG-3¢) were used. Amplified PCR products
were cloned into pCR-Blunt II-TOPO (Invitrogen). The
cDNAs thus obtained were confirmed by nucleotide
sequencing and then subcloned into the expression vector
pcDNA3.1/Zeo (Invitrogen) or the mammalian two-hybrid
vector pACT (Promega) for expression of the VP16
activation domain (AD) fusion protein.
To construct expression vectors for FLAG-tagged
DEC1:2–412 and BMAL1, two sets of primers ( 5¢-AAGC
TTGAGCGGATCCCCAGCGCGCAACCACC-3¢ and
5¢-GCAGCAGGATCCTCTAGAGAGTTTAGTC
TTTG-3¢ for FLAG-DEC1; and 5 ¢-GAATTCGGCGG
ACCAGAGAATGGACATTTCCTCAACCATC-3¢ and
5¢-TCTAGACTACAGCGGCCATGGCAAGTCACTAA
AGTC-3¢ for FLAG-BMAL1) were used for amplification
by PCR, with hDec1 cDNA and mouse Bmal1 (mBmal1)
cDNA, respectively, as templates. Amplified PCR products
were cloned into pCR-Blunt II-TOPO. After confirmation
of the nucleotide sequences, the cDNAs thus obtained were
subcloned into p3xFLAG-CMV-10 (Sigma) for expression
of FLAG-tagged protein.
To construct expression vectors for the bHLH domain-,

basic region- and Orange domain-deleted DEC1 mutants,
three sets of primers (5¢-ATTGATCAGC AGCAGCA
GAAAATCATTGCC-3¢ and 5¢-CTTGCTGTCCTCG
CTCCGCTTTATTCCC-3 ¢ for DEC1DbHLH and
DEC1:4–232DbHLH; 5¢-GACCGGATTAACGAGTGC
ATCGCCCAG-3¢ and 5 ¢-CTTGCTGTCCTCGCTCCGC
Fig. 1. Suppress ive activity of DEC1 against the CLOCK/BMAL1-activated promoter. (A) Deletion a nalysis of DEC1. Expression vec tors (10 ng
per well) encoding various deletion mutants of DEC1 were cotransfected with the luciferase reporter construct pDEC1-E-ABC-TK (2 ng per well),
together with expression vectors for CLOCK and BMAL1 (each 50 ng p er well), into NIH3T3 cells. After incubation for 48 h, luciferase activities
were measured. The values represent relative luciferase activities of pDEC1-E-ABC-TK (mean ± SEM, n ¼ 15). Structures of the DEC1 mutants
are schematically shown in the left-hand panel, with the basic helix-loop -helix ( bHLH) and Orange domains indicate d. The expression of DEC1
mutants was examined by Western blot analysis by using anti-DEC1 immunoglobulin (lower-right panel). Various DEC1 mutants are indicated by
arrowheads. (B) Dose-dependency of DEC1 suppressive activity. Increasing amoun ts (0.1, 1 and 10 ng) of FLAG-DEC1 expression vector were co-
transfected with pDEC1-E-ABC-TK, together with expression vectors for CL OCK and BMAL1. The total amount of transfected DNA was
adjustedtothesamevalue,ineachexperiment,bytheadditionofanemptyvector (pcDNA3.1/Zeo). Relative luciferase activities of pDEC1-E-
ABC-TK (mean ± SEM, n ¼ 6) are shown in the left-hand panel. Expression of FLAG-DEC1 was examined by Western blot analysis with anti-
FLAG immunoglobulin (right panel). (C) Expression levels of DEC1 and BMAL1 were compared by using FLAG-tagged proteins. Expression
vectors for FLAG -BMAL1 ( 50 ng per w ell) an d C LOCK (50 ng p er well) were cotransfected with a FLAG-DEC1 expression vector (lane 2) o r
with an empty vector (lane 1) (10 ng per well). Expression of FLAG-tagged protein was examined by Western blot analys is with anti-FLAG
immunoglobulin (left-hand panel). The transcriptional activities of FLAG-BMAL1 and FLAG-DEC1 were confirmed by using the luciferase assay
(right-hand panel). The values represent relative luciferase activities of pDEC1-E-ABC-TK (mean ± SEM, n ¼ 10). (D) Effect of the histone
deacetylase (HDAC) inhibitor on DEC1 or DEC2 suppressive activity agaist the CLOCK/BMAL1-activated promoter. The reporter construct,
pDEC1-E-ABC-TK, was co-transfe cted with expression vectors for CLOCK and BMAL1 together with increasing amounts of a DEC1 or a DEC2
expression vector, as indicated. The supp ressive activity of DEC2 was much higher than that of DEC1. The HDAC inhibitor, trichostatin A (TSA)
(300 n
M
), was added 24 h after transfection, and incubation was continued for a further 24 h. The luciferase activity (mean ± SEM, n ¼ 5) of
pDEC1-E-ABC-TK, without the DEC1 or DEC2 expression vector, in t he pre sence or absence of T SA, was given a value of 100. P-values were
calculated by using the Student’s t-test (**P<0.01, *P<0.05).
4410 F. Sato et al. (Eur. J. Biochem. 271) Ó FEBS 2004

TTTATTCCC-3¢ for DEC1Dbasic and DEC1:4–232Dba-
sic; and 5¢-CTGCAGGGTGGTACCTCCAGGAAGC
CATC-3¢ and 5¢-CTCTTGACCTGTTTCGACATTTCT
CCCTGAC-3¢ for DEC1DOrange) were used for amplifi-
cation by PCR, with a pCR-Blunt II-TOPO plasmid
carrying hDec1:1–412 or hDec1:1–232 cDNA as a template.
To obtain expression vectors for single amino acid-
substituted DEC1 mutants, three sets of primers
(5¢-CGGCAATTTGTAGGTCTCCTTGCTGTCCTCGC
TC-3¢ and 5¢-
GCCCGGCTCATCGAGAAAAAGAGA
CGTGACCGG-3¢ for DEC1-H57A; 5¢-GATGAGCCG
GTGCGGCAATTTGTAGGTCTCC-3¢ and 5 ¢-GAGAA
AAAGAGA
GCTGACCGGATTAACGAGTGC-3¢ for
DEC1-R65A, FLAG-DEC1-R65A and DEC1:4–232-
R65A; and 5 ¢-GATGAGCCGGTGCGGCAATTTGTA
GGTCTCC-3¢ and 5 ¢-GAGAAAAAGAGA
AAGGACC
Ó FEBS 2004 Interaction of DEC1 with BMAL1 (Eur. J. Biochem. 271) 4411
GGATTAACGAGTGCATC-3¢ for FLAG-DEC1-R65K)
were used (substituted nucleotides are underlined). The
resulting PCR products were ligated to make a circular form
of the plasmids and transformed into Escherichia coli
DH5a. The cDNAs obtained from the transformants were
confirmed by nucleotide sequencing and subcloned into a
pcDNA3.1/Zeo, a pACT or a p3xFLAG-CMV-10 vector.
To obtain deleted fragments from the 5¢-terminus of
mBmal1 cDNA, a 3 ¢-terminus primer (5 ¢-TCTAGACTA
CAGCGGCCATGGCAAGTCACTAAAGTC-3¢)anda

5¢-terminus primer (one of 5¢-GGATCCGTGCGGACCA
GAGAATGGACATTTCCTC-3¢,5¢-GGATCCTCACC
GTGCTAAGGATGGCTGTTCAGCAC-3¢ or 5¢-GGAT
CCCCTCCCGGCTATGCTCTGGAGCC-3¢)wereused
for amplification by PCR, wit h mBmal1 cDNA as a
template. The cDNAs thus obtained were subcloned into
the mammalian two-hybrid vector pBIND (Promega) for
expression of the GAL4 DNA-binding domain (DNA-BD)
fusion protein.
Luciferase reporter assay
Twenty-four hours before transfection, NIH3T3 cells were
seeded at 2 · 10
4
cells per 16 m m well. The luciferase
reporter plasmid pDEC1-E-ABC-TK, carrying three hDec1
CACGTG E-boxes connected to the thymidine kinase (TK)
promoter (2 ng per well), or pDEC1-3620 carrying the hDec1
promoter [19], was co-transfected with expression vectors for
mouse CLOCK and BMAL1 (each 50 ng per well), together
with an expression vector for human DEC1 or DEC2 (10 ng
or the indicated amount per well), by using Trans IT
polyamine (Mirus, Madison, WI , USA), as described
previously [11]. As an internal standard, 0.2 ng of phRL-
TK (Promega) was co-transfected. The total amount of
transfected DNA was a djusted to the same va lue, in each
experiment, by using an empty vector (pcDNA3.1/Zeo). The
cells were incubated for 48 h and then subjected to the
luciferase reporter assay by using the Dual-Luciferase
Reporter Assay System (Promega). Luciferase activities were
normalized relative to internal control a ctivities. The experi-

ments were repeated at least twice and the data thus obtained
were combined to represent the mean ± SEM. For the
trichostatin A (TSA) assay, DNA-transfected NIH3 T3 ce lls
were incubated with TSA (300 n
M
; Sigma) for 24 h.
Mammalian two-hybrid assay
The mammalian two-hybrid vectors, pBIND, encoding
GAL4 DNA-BD fusion protein (100 ng per well), and
pACT, e ncoding VP16 AD fusion protein (100 ng per w ell),
were co-transfected with the l uciferase r eporter plasmid
pG5luc (Promega), ca rrying five GA L4-binding sites
upstream of the TATA box (100 ng per well), and phRL-
TK (0.2 ng per well) into NIH3T3 cells. The cells were
incubated f or 48 h a nd subjected to a luciferase reporter
assay.
Western blot analysis
Rabbit antibodies to DEC1 were produced by immun-
izisation with the synthetic peptide fragment Cys-Lys-
Gly-Asp-Leu-Arg-Ser-Glu-Gln-Pro-Tyr-Phe-Lys-Ser-
Asp-His-Gly-Arg-Arg. The antibodies thus obtained
(anti-DEC1:251–268) were pu rified by affinity column
chromatography. NIH3T3 cells were seeded at 1 · 10
5
cells
per 35 mm well, or at 2 · 10
4
cells per 16 mm well, 24 h
before transfection. Expression plasmids (1.5 lgorthe
indicated amount per well) were transfected i nto t he cells by

using PolyFect Transfection Reagent (Qiagen). Forty-eight
hours after transfection, the cells were harvested and
dissolved in 200 or 30 lL of S DS sample buffer. Equal
volumes of the samples (15 lL) were subjected to SDS/
PAGE and transferred onto a nylon membrane (Immobilon
P; Millipore). DEC1, FLAG-DEC1, FLAG-BMAL1 and
VP16-DEC1 were detected with anti-DEC1:251–268, anti-
FLAG (Sigma) and anti-VP16 (Santa Cruz) immuno-
globulins.
Electrophoretic mobility shift assay
Various DEC1 mutant proteins (including VP16-fused
protein) and luciferase protein, as a control product,
were synthesized by using the TNT Quick Coupled
Transcription/Translation S ystem (Promega). T he expres-
sion levels were confirmed by using Western blot analysis
with anti-D EC1 o r anti-VP16 immunoglobulin. T he
double-stranded oligonucleotides of Dec1 E-box C
(5¢-ctagGTCCAA
CACGTGAGACTCtcga-3¢; E-box is
underlined) were end-labelled by using [
32
P]dCTP[aP]
(Du Pont-New England Nuclear) and DNA poly-
merase I Klenow fragment (TAKARA). Synthesized
protein was incubated with approximately 40 000 c.p.m.
of
32
P-labelled E-box C probe for 15 m in at room
temperature in 10 lLof10m
M

Tris/HCl (pH 8.0),
0.5 m
M
dithiothreitol, 10% (v/v) glycerol, 1 lgof
poly(dI-dC), 50 m
M
NaCl and 5 m
M
MgCl
2
,afterwhich
the mixtures were subjected to PAGE (5% gel) in
electrophoresis buffer ( 12 m
M
Tris/HCl, 125 m
M
glycine,
1m
M
EDTA) at 4 °C, and visualized by using auto-
radiography.
Pull-down assay
Glutathione S-transferase (GST)-mouse DEC2 [20] fusion
protein (GST-DEC2) was expressed i n E. coli BL21 and
purified as described p reviously [21].
35
S-labelled m ouse
DEC1 was synthesized by using the TNT Quick Coupled
Transcription/Translation S ystem (Promega) and incubated
with 2 lg of GST or GST-DEC2 on glutathione-agarose

beads in binding buffer (20 m
M
Tris/HCl, pH 8.0, 200 m
M
NaCl, 1 m
M
EDTA, 0.5% (v/v) Nonidet P-40 and
5mgÆmL
)1
of bovine serum albumin) for 2 h at 4 °C. The
beads were washed three times with binding buffer and the
bound proteins were analyzed by SDS/PAGE and a uto-
radiography.
Results
Deletion analysis of DEC1 on its suppressive activity
in the presence of CLOCK and BMAL1
Using deletion analyses, two regions in the protein
product of Dec1 have been identified as important
domains for i ts suppressive activity of transcription from
4412 F. Sato et al. (Eur. J. Biochem. 271) Ó FEBS 2004
some genes, including Dec2 and PPARc2,intheabsence
of other transcription facto rs, such as CLOCK and
BMAL1 [9,10,22,23]: an N-terminal region between amino
acids 1 and 141, and another region between amino acids
147 and 354, were reported to be essential f or the
suppression of the target g enes. To d etermine which
region in DEC1 is required for the suppression of
CLOCK/BMAL1-induced transcription, truncated forms
of DEC1 were expressed in NIH3T3 cells together with
CLOCK and BMAL1, and their transcriptional activities

were examined by using a lucife rase assay with a reporter
construct containing three CACGTG E-boxes of human
Dec1 connected to the TK promoter (pDEC1-E-ABC-
TK). As shown i n Fig. 1A, the promoter activity o f
pDEC1-E-ABC-TK was enhanced by CLOCK/BMAL1,
and the increased activity was reduced by full-length
DEC1 (DEC 1:1–412), a s d escribed previously [11]. Dele-
tion of 27 residues from t he N-terminal region of DEC1
(DEC1:28–412) did not diminish the suppressive activity,
whereas deletion of 55 residues (DEC1:56–412) decreased
the DEC1 activity, and deletion of 9 3 residues (DEC1:
94–412) or more than 93 residues ( DEC1:119–412)
abolished the suppression (Fig. 1A). On the other h and,
deletion of up to 273 residues from the C-terminal region
of DEC1 (DEC1:1–309, DEC1:1–232 and DEC1:1–139)
had little effect on the suppressive activity of DEC1.
Expression of these truncated DEC1 mutants w as con-
firmed b y Western blot analysis, a lthough the expression
levels of DEC1 mutants varied among clones. To examine
the dose-dependency of t ranscriptional suppression by
DEC1, 0.1, 1 or 10 ng of FLAG-tagged DEC1 expression
vector was co-transfected with expression vectors for
CLOCK a nd BMAL1 ( 50 ng of each). Transfection of
0.1 ng of the DEC1 expression vector did not significantly
decrease the transcriptional activity of CLOCK/BMAL1
(Fig. 1B). Transfection o f 1 ng of the expression vector
resulted in a d ecreased e xpre ssion of t he DEC 1 pro tein,
causing significant but much lower levels of suppressive
activity than transfection with 10 ng of the expression
vector. To compare protein l evels of DEC1 a nd BMAL1,

FLAG-tagged proteins were expressed in NIH3T3 c ells
and subjected to Western blot analysis by using anti-
FLAG immunoglobulin. As shown in Fig. 1C, transfec-
tion o f 10 ng of FLAG-DEC1 expression vector and
50 ng of FLAG-BMAL1 expression vector resulted in
adequate levels of expression. The transcriptional activity
of FLAG-BMAL1, and the suppressive activity of FLAG-
DEC1, were s imilar to those of intact BMAL1 and
DEC1, respectively (Fig. 1A,C). These findings indicate
that the region between amino acids 28 and 139 of DEC1
(including the bHLH domain) is sufficient for the
suppressive activity of DEC1. The C-terminal region
(amino acids 140–412), including the Orange domain
[18], i s not required for suppression in the presence of
CLOCK/BMAL1.
Involvement of histone deacetylase (HDAC) in DEC1
or DEC2 suppression of CLOCK/BMAL1-induced gene
expression
DEC1 and GAL4 DNA-BD-fused DEC2 bo und to HDAC
and suppressed t ranscription from, r espectively, the Dec1
(Stra13) promoter and the GAL4 response promoter
[24,25]; however, suppression of c-myc expression by
DEC1 did not require HDAC [24]. To examine whether
CLOCK/BMAL1-induced gene expression is suppressed by
DEC1 or DEC2 via an HDAC-dependent pathway, we
added TSA (a specific inhibitor of HDAC) to the cell
cultures 24 h after transfection of reporter plasmids. The
addition of TSA r eversed the suppression b y DEC1 o r
DEC2, a s s hown i n Fig. 1D, indicating that an HDAC-
co-repressor complex(es) may be involved, at l east partly, in

the suppression of CLOCK/BMAL1-induced transcription
by DEC1 or DEC2.
Fig. 2. Effect of an internal deletion of DEC1
on its suppressive activity. Expression ve ctor s
for DEC1 carrying various deletion mutations
were co-transfected with pDEC1-3620 or
pDEC1-E-ABC-TK, together with expression
vectors for CL OCK and B MAL1, into
NIH3T3 cells. Relative luciferase activities of
pDEC1-3620 (mean ± SEM, n ¼ 7) or
pDEC1-E-ABC-TK (mean ± SEM, n ¼ 11)
are presented. **P<0.01 (Studen t’s t-test).
The structures of the DEC1 mutants expressed
are shown in the upper-left panel. Expression
of DEC1 mutants was examined by Western
blot analysis with anti-DEC1 immunoglobulin
(lower-left panel).
Ó FEBS 2004 Interaction of DEC1 with BMAL1 (Eur. J. Biochem. 271) 4413
The basic region of DEC1 is essential for its suppressive
activity
To further n arrow down the region required for the
suppressive activity of DEC1, we generated several con-
structs with deletions in internal regions. Deletion of the basic
region (DEC1 Dbasic) or bHLH domain (DEC1 DbHLH)
disrupted DEC1 suppressive activit y against CLOCK/
BMAL1-induced transcription from the DEC1 promoter
or the TK promoter connected to the DEC1 E-boxes
(Fig. 2), while deletion of the Orange domain ( DEC1DOr-
ange) had no effect on the suppressive activity, indicating
the importance of t he bHLH region of DEC1.

The HXXXXXXXR sequence in the bas ic r egion h as
been rep orted to be conserved among group B bHLH
proteins [26,27]. We t herefore examined whether these
amino acid residues are required for DEC1 activity.
Substitution of Arg65 for Ala(DEC1-R65A) severely
reduced the suppressive activity of DEC1, whereas substi-
tution of His57 for Ala(DEC1-H57A) did not alter the
activity (Fig. 3A). As Western blot analysis showed that the
expression levels of the R65A mutant were lower than that
of full-length DEC1, we next g enerated expression
constructs for FLAG-fused R65A mutant and R65K
mutant DEC1. The expression levels of R65A and R65K
mutants were comparable to those of FLAG-fused DEC1,
but they did not have any significant suppressive activity
against C LOCK/BMAL1-induced transcription. From
these findings, we conclude that the basic region, p artic-
ularly the conserved A rg65, but not His57, is essential for
the suppressive activity of DEC1.
Determination of the binding domain of DEC1 to BMAL1
The interaction of DEC1 and BMAL1 was previously
demonstrated by a yeast two-hybrid assay [ 7], and this
interaction may be involved in the suppressive activity of
DEC1 against CLOCK/BMAL1-induced transcription.
To confirm that the binding of DEC1 and BMAL1
actually occurs in mammalian cells, w e performed a
mammalian t wo-hybrid assay by using various DEC1
mutant constructs. The N-terminal region of DEC1
(DEC1:4–232 or DEC1:4–139) interacted with BMAL1
(Fig. 4). However , deletion of the basic region or of the
bHLH domain completely abrogated the DEC1–BMAL1

interaction, and substitution of Arg65 for Ala(DEC1:4–
232-R65A) also abolished the in teraction, indicating that
the basic region is essential for the i nteraction. However,
Fig. 3. Effect of a single amino a cid substitu-
tion in the basic region of DEC1 o n its sup-
pressive activity. (A) Expression vectors for
DEC1 carrying variou s point m utations w ere
co-transfected with pDEC1-3620 or pD EC1-
E-ABC-TK, tog ether w ith expre ssion v ectors
for CLOCK an d BMAL 1. The sub stituted
amino acids of DEC1 are sh own in the upper-
left panel. Relative luciferase activities of
pDEC1-3620 (mean ± SEM, n ¼ 7) or
pDEC1-E-ABC-TK (me an ± SEM, n ¼ 11)
are presented. **P<0.01 (Student’s t-test).
Expression of DEC1 mutants was examined
by Western b lotting by using anti-DEC1
immunoglobulin. (B) Exp ressio n vectors f or
FLAG-fused DEC1 w ere co-transfected w ith
pDEC1-E-ABC-TK, together with expression
vectors for CLOCK a nd BMAL1. Luciferase
activities of pDEC1-E-ABC-TK (mean ±
SEM, n ¼ 6) were examined. Expression of
DEC1 mutants was analyzed by Western
blotting with anti-F LAG immun oglobulin.
4414 F. Sato et al. (Eur. J. Biochem. 271) Ó FEBS 2004
substitution of His57 for Ala(DEC1:4–232-H57A) did not
affect the interaction, which coincided with the strong
suppressive activity of the H57A mutant against the
CLOCK/BMAL1 heterodimer. Similar levels of expres-

sion of VP16-fused DEC1 mutants were confirmed by
Western blot analysis. Taken together with the results
shown in Figs 1–3, these findings indicate that the region
required for the suppressive activity of DEC1 is also
required for the interaction with BMAL1.
Binding of DEC1 mutants to a CACGTG E-box
To e xamine the binding ability o f DEC1 mutants to the
CACGTG E-box in the Dec1 promoter, an electrophoretic
mobility shift assay was performed. Shifte d bands were
observed by using full-length DEC1 (DEC1:1–412) (Fig. 5,
lanes 1 and 5 ) or O range domain-deleted DEC1 (DEC1-
DOrange) (lane 4), whereas no bands were detected by using
basic region-deleted D EC1 ( DEC1Dbasic and DEC1:4–
232Dbasic) (lanes 2 and 14) or bHLH domain-deleted
DEC1 (DEC1DbHLH and D EC1:4–232DbHLH) (lanes 3
and 15). His57-substituted DEC1 (DEC1-H57A) showed a
very low binding ab ility for the E-box (lane 7), and
substitution of Arg65 for Ala(DEC1-R65A and DEC1:4–
2332-R65A) abolished DEC1 binding to the E-box (lanes 6
and 16). On t he other hand, deletion of up to 273 residues
from the C-terminal region of DEC1 (VP16-DEC1:4–232
and VP16-DEC1:4–139) did not diminish the b inding
activity (lanes 9, 10 and 13), whereas deletion of 296
residues (VP16-DEC1:4–116) decreas ed the binding (lane
11), and deletion of 312 residues (VP16-DEC1:4–100)
abolished it (lane 12). Expression levels of DEC1 mutants
and VP16-tagged DEC1 mutants synthesized by in vitro
transcription/translation were confirmed by W estern blot
analysis (Fig. 5 ). These r esults indicate that the bHLH
region, including His57 and Arg65, is responsible for t he

E-box binding.
Determination of the region in BMAL1 for binding
to DEC1
To identify the region in BMAL1 required for the binding to
DEC1, w e constructed e xpression plasmids for truncated
BMAL1 (Fig. 6). D eletion of 111 amino acids f rom the
N-terminal region (BMAL1:112–626) did not affect the
binding of BMAL1 to DEC1; deletion of 235 amino acids
(BMAL1:236–626) slightly diminished the binding,
although a strong binding ability of BMAL1:236–626 with
DEC1 still existed. As similar results were obtained in the
yeast two-hybrid a ssay (F. Sato, unpublished data), i t i s
likely that the C-terminal region, including the PAS-B
domain of BMAL1, is required for the binding to DEC1,
Fig. 4. Identification of the domain of D EC1
that interacts with BMAL 1. Interactions of
various DEC1 mutants with BMAL1 i n
NIH3T3 cells were examined. The mamma-
lian two-hybrid vector, pACT, encoding var-
ious deletion or point mutants of DEC1 f or
expression of th e VP16 act ivation d omain
(AD) fusion prote in, was u sed. Th e p ACT
vector carrying mutant Dec1 cDNA (100 ng
per well) was co-transfected w ith pBIN D
carrying Bmal1:2–626 cDNA (100 ng per
well) for expression of the GAL4 DNA-
binding domain ( DNA-BD) f usio n p rotein ,
together with the l uciferase repo rter plasmid
pG5luc carrying five GAL4-binding sites up-
stream of the TAT A box (100 ng per well). An

emptypACTorpBINDvectorwasusedasa
control. The cells were inc ubated for 48 h and
subjectedtotheluciferase reporter assay.
Relative luciferase activities of pG5luc
(mean ± SEM, n ¼ 11) were normalized
relative to internal control activities.
**P<0.01 (Student’s t-test). Interactions
between the VP16 AD- DEC1 f usion prote in
and the GAL 4 DNA-BD-BMAL1 fusion
proteinresultedinanincreaseinexpression
of the luc iferase g ene. E xpression o f the
VP16-DEC1 fusion protein was examined by
Western blot analysis with an a nti-VP16
immunoglobulin.
Ó FEBS 2004 Interaction of DEC1 with BMAL1 (Eur. J. Biochem. 271) 4415
whereas the N-terminal region, containing the bHLH
domain, is not essential for the interaction.
Dominant-negative DEC1 counteracts the suppression
of CLOCK/BMAL1-induced transcription in the presence
of full-length DEC1
As DEC1 lacking the basic region interfered with full-length
DEC1 for the transcriptional suppression in the absence of
CLOCK/BMAL1 [9], we examined whether DEC1Dbasic is
a dominant-negative competitor in t he presen ce of CLOCK
and BMAL1. Co-transfection with an expression plasmid
for DEC1Dbasic diminished the suppressive activity of
full-length DEC1 in a dose-dependent manner (Fig. 7A).
In addition, DEC1 carrying the substitution Arg65 for
Ala (DEC1-R65A) showed a similar ability (Fig. 7B).
However, neither DEC1Dbasic nor DEC1-R65A alone

had any significant suppressive activities (Figs 2 and 3).
These findings suggest that DEC1 forms homodimers
through the helix-loop-helix region, but not through the
basic region, when it s uppresses the CLOCK/BMAL1-
induced transcription. Accordingly, DEC1 lacking t he
bHLH domain (DEC1DbHLH) did not decrease the
activity of co-expressed full-length DEC1 (Fig. 7C).
Heterodimer formation of DEC1 and DEC2
Although DEC1 functioned as a homodimer, interactions
between DEC1 an d DEC2 h ad not previously been
demonstrated. To investigate whether DEC1 and DEC2
could form heterodimers, an expression plasmid encoding
DEC1Dbasic w as co-transfected with a DEC2 expression
vector, together with expression vectors for CLOCK and
BMAL1: DEC1Dbasic counteracted DEC2 activity in a
dose-dependent manner ( Fig. 8A). The heterodimer f or-
mation of DEC1 and DEC2 w as confirmed by a pull-
down assay: in this assay,
35
S-labelled DEC1 bound to
GST-DEC2 fusion protein, bu t n ot to GST protein
(Fig. 8 B).
Discussion
In the present study, we f ound that the N-terminal region
(1–139) of DEC1 was essential for DEC1 suppressive
activity against CLOCK/BMAL1-induced transcription. In
addition, the N-terminal region of DEC1, including the
bHLH domain, interacted with the C-terminal r egion o f
BMAL1 in a mammalian two-hybrid assay. Accordingly, a
Fig. 5. Analysis of DEC1 mutants for binding

to the CACGTG E-box in the Dec1 promoter.
Various DEC1 mutants and VP1 6-tagged
DEC1 mutants were synthesized by using an
in vitro transcription/translation system. The
32
P-labelled Dec1 E-box C probe was incuba-
ted with DEC1 mutants (lanes 1–7), V P16-
tagged DEC1 mutants (lanes 9–16) or luci-
ferase protein synthesized as a control product
(cont.) (lanes 8 and 17). Shifted bands of
radiolabelled E-box C a nd m utant DEC1
complexes are indicated by asterisks. Expres-
sion levels of DEC1 mutants and VP16-DEC1
mutants were con firmed by W estern b lot
analysis with anti-DEC1 and anti-VP16
immunoglobulin (lower p anels). Faint, non-
specific bands were also observed, even when
the probe was i ncubated with a control
protein.
Fig. 6. Identification of t he domain of B MAL1 that inte racts with
DEC1. Interactions of various truncated BMAL1 mutants with DEC1
in NIH3T3 cells were examined. pBIND, carrying various lengths of
Bmal1 cDNA, w as co-transfected with pACT carrying Dec1 cD NA
together with pG 5luc . Luciferase activities of pG5 luc (mean ± SEM,
n ¼ 11) were normalized by internal control activities. **P<0.01
(Student’s t-test).TheregionsofBMAL1expressedasaGAL4fusion
protein are schematically shown in the upper panel. The basic helix-
loop-helix (bHLH), PAS-A and PAS-B domains [32] are i ndicated.
Interactions between GAL4 DNA-BD-BMAL1 fusion protein and
VP16 AD-DEC1 fusion protein resulted in an increase in luciferase

gene expression.
4416 F. Sato et al. (Eur. J. Biochem. 271) Ó FEBS 2004
recent w ork [28] demonstrated the binding of human DEC1
to BMAL1 by using a c o-immunoprecipitation assay. Our
mutation analysis showed that the region comprising amino
acids 1–139 in DEC1, essential for its suppressive activity,
was identical to the region required for the interaction with
BMAL1. In addition, the N-terminal region of DEC1
(DEC1:4–139 or DEC1:4–116) boun d to t he Dec1-CAC-
GTG E-box, which was recognized by the CLOCK/
BMAL1 heterodimer. The basic region ( amino acids
51–65) of DEC1 was essential for both the interaction of
Fig. 7. DEC1 mutants that act in a dominant-negative manner.
Increasing amounts of an expression v ector for DEC1 Dbasic ( A),
DEC1-R65A (B) or DEC1DbHLH (C) w er e co-transfected with
pDEC1-E-ABC-TK, together with expression vectors for CLOCK,
BMAL1 and full-length DEC1 into NIH3T3 cells. The amounts o f
transfected plasmid DNA (ng per well) are indicated. The total amount
of transfected DNA was adjusted to the same value, in each experi-
ment, by using an empty vector. After incubation of the cells for
48 h, the luciferase activities of pDEC1-E-ABC-TK were determined
(mean ± SEM, n ¼ 5).
Fig. 8. He terodimer formation of DEC1 and DEC2. (A) Increasing
amounts o f an expression vector f or DEC1Dbasic w ere co-transfected
with pDEC1-E-ABC-TK, together with expression vectors for
CLOCK, BMAL1 and DEC2 into NIH3T3 cells. The amounts of
transfected p lasmid DNA ( ng per well) are i ndicated. The values
represent relative l uciferas e activities of pDEC1-E-ABC-TK
(mean ± SEM, n ¼ 6). (B) The binding of DEC1 and DEC2 was
examined by using a pull-down assay.

35
S-labelled DEC1 was incu-
bated with glutathione S-transferase (GST) or GST-DEC2 on gluta-
thione-agarose beads for 2 h at 4 °C. The b eads were washed three
times, subjected to S DS/PAGE with 10% of input
35
S-labelled DEC1,
and visualized by autoradiography.
Ó FEBS 2004 Interaction of DEC1 with BMAL1 (Eur. J. Biochem. 271) 4417
DEC1 with BMAL1 and the bin ding to the E-box.
Substitution of Arg65 for Ala abolished the interaction of
DEC1 with BMAL1 and the binding of DEC1 to the E-box.
However, substitution of His57 for Ala did not affect the
interaction of D EC1 w ith BMAL1, nor its suppressive
activity, although it strongly decreased the DEC1 binding
activity to the CACGTG E-box. DEC1 therefore appears to
suppress the CLOCK/BMAL1-induced transcription, at
least in part, by interacting with BMAL1.
The amino acid residue Arg65 in the basic region of
DEC1 is conserved a mong the g roup B bHLH proteins
(such as USF, c-Myc, MAX and MAD) that can bind
to the CACGTG E-box [26], along with some other
transcription factors [29,30]. This amino acid residue was
also important for the interaction between DEC1 and
BMAL1, as shown in this study, and, moreover, the Arg
residue might be crucial for t he activities of the o ther
group B bHLH proteins. In addition to BMAL1, DEC1
can bind to various transcription factors such as USF2
[29], MASH1 [23] and E47 [31], or t o co-repressors such
as H DAC1, mSin3A and NcoR [ 24]. USF2-DEC1

heterodimer f ormation inhibited USF2 f rom binding to
a CACGTG E-box in the M4-Luc promoter [ 29], even
though USF2 or DEC1 alone could bind to the element.
These findings suggest that the g roup B b HLH proteins,
including DEC1, work t hrough t wo mechanisms: inter-
action with s ome other t ranscription factor(s); and
binding to an E-box.
DEC1 has been r eported t o act as a homodimer t o
suppress transcription from the reporter gene carrying three
CACGTG e lements [9]. Here we showed that dimer
formation w ould a lso be required f or the suppression of
CLOCK/BMAL1-induced transcription. The basic region
of DEC1 is not required for homodimer formation, as the
basic region-deleted or Arg65-substituted DEC1 acted as a
dominant-negative competitor. We also found that DEC1
could form heterodimers with DEC2, which had a strong
similarity (90%) to DEC1 in the region between amino
acids 48–124 of DEC1. This region, including the bHLH
domain, coincided w ith t he r egion required for bo th the
suppressive activity of DEC1 and the interaction with
BMAL1. DEC2 can a lso i nteract with BMAL1, b ind to
CACGTG E-boxes and has a s uppressive activity similar to
that of DEC1 [7,11]. Hence, DEC1 and DEC2 may be
interchangeable, playing roles in transcriptional regulation
in a situation-dependent manner.
Acknowledgements
This work was supported by grants-in-aid for science from the Ministry
of Education, Culture, Sport, Science and Technology of Japan.
References
1. Darlington, T.K., Wager-Smith, K., Ceriani, M.F., Staknis, D.,

Gekakis, N., Steeves, T.D., Weitz, C.J., Takahashi, J.S. & Kay,
S.A. (1998) Closing the circadian loop: CLOCK-induced tran-
scription of its own inhibitors per and tim. Science 280, 1599–1603.
2. Gekakis, N., Staknis, D., Nguyen, H.B., Davis, F.C., Wilsbacher,
L.D., King, D.P., Takahashi, J.S. & Weitz, C.J. (1998) Role of the
CLOCK protein in the mammalian circadian mechanism. Science
280, 1564–1569.
3. Kume, K., Zylka, M.J., Sriram, S., Shearman, L.P., Weaver, D.R.,
Jin, X., Maywood, E.S., Hastings, M.H. & Reppert, S.M. (1999)
mCRY1 and mCRY2 are e ssential components of the negative
limb of the circadian clo ck fe edback loop. Cell 98, 193–205.
4. Yagita, K., Yamaguchi, S., Tamanini, F., van der Horst, G.T.,
Hoeijmakers, J.H., Yasui, A., Loros, J.J., Dunlap , J.C. &
Okamura, H. (2000) Dimerization and nuclear entry of mPER
proteins in mam malian cells. Genes Dev. 14, 1353–1363.
5. Preitner, N., Damiola, F., Lopez-Molina, L., Zakany, J.,
Duboule, D. , Albrecht, U. & Schibler, U. ( 2002) The orphan
nuclear re ceptor R EV-ERBalp ha c ontro ls circad ian transcription
within the p ositive limb of the mammalian circadian oscillator.
Cell 110, 251–260.
6. Ueda, H.R., Chen, W ., Adachi, A., W akamatsu, H., Hayashi, S.,
Takasugi, T., Nagano, M., Nakahama, K., Suzuki, Y., Sugano,
S., Iino, M., Shigeyoshi, Y. & Hashimoto, S. (2002) A transcrip-
tion factor response element for gene expression during circadian
night. Nature 418, 534–539.
7. Honma, S., Kawamoto, T., Takagi, Y., Fujimoto, K., Sato, F.,
Noshiro, M., Kato, Y. & Honma, K. (2002) Dec1 and Dec2 are
regulators of the mammalian molecular clock. Nature 419, 841–
844.
8. Zawel,L.,Yu,J.,Torrance,C.J.,Markowitz,S.,Kinzler,K.W.,

Vogelstein, B . & Zhou, S. (2002) DEC1 is a downstream target of
TGF-beta with sequence-specific transcriptional repressor activ-
ities. Proc. Natl Acad. Sci. USA 99, 2848–2853.
9. St-Pierre, B., Flock, G., Zacksenhaus, E. & Egan, S.E. (2002)
Stra13 homodimers repress transcription through class B E-box
elements. J. Biol. Chem. 277, 46544–46551.
10. Li, Y., Xie, M., Song, X., Gragen, S., Sachdeva, K., Wan, Y. &
Yan, B. (2003) DEC1 negatively regulates the expression of DEC2
through binding to the E-box in the proximal promoter. J. Biol.
Chem. 278, 16899–16907.
11. Kawamoto,T.,Noshiro,M.,Sato,F.,Maemura,K.,Takeda,N.,
Nagai, R., I wata, T., Fujimoto, K ., Furukawa, M., Miyazaki, K.,
Honma, S., Honma, K. & Kato, Y. (2004) A novel autofeedback
loop of Dec1 transcription involved in circadian rhythm regula-
tion. Biochem. Biophys. Res. Commun. 313, 117–124.
12. Hamaguchi, H., Fujimoto, K., Kawamoto, T., Noshiro, M.,
Maemura, K., Takeda, N., Nagai, R., Furukawa, M., Honma, S.,
Honma, K., Kurihara, H. & K ato, Y. (2004) Expression of the
gene for Dec2, a basic hel ix-loop-helix transcription factor, is
regulated by a molecular c lock system. Biochem. J. 382, 43–50.
13. Noshiro, M., Kawamoto, T., F urukawa, M., Fujimoto, K.,
Yoshida,Y.,Sasabe,E.,Tsutsumi,S.,Hamada,T.,Honma,S.,
Honma, K. & Kato, Y. (2004) Rhythmic expression of DEC1 and
DEC2 in peripheral t issues: DEC2 is a potent suppressor for
hepatic cytochrome P450s opposing DBP. Genes Cells 9, 317–329.
14. Roenneberg, T. & Merrow, M. (2003) The n etwork of time:
understanding t h e molecular c ircadian syst em. Cur r. B iol. 13,
R198–R207.
15. Butler, M.P., Honma, S., Fukumoto, T., Kawamoto, T., Fuji-
moto, K., Noshiro, M., Kato, Y. & Honma, K. (2004) Dec1 a nd

Dec2 expression is disrupted in th e suprachiasmatic nuclei o f
Clock mutant mice. J. Biol. Rhythms 19, 126–134.
16. Shen, M., Yoshida, E., Yan, W., Kawamoto, T., Suardita, K.,
Koyano, Y., Fujimoto, K., Noshiro, M. & Kato, Y. (2002) Basic
helix-loop-helix protein DE C1 promot es chondrocyte d ifferentia-
tion at the early and terminal stages. J. Biol. Chem. 277, 50112–
50120.
17. Grechez-Cassiau, A., Panda, S., Lacoche, S., Teboul, M., Azmi,
S.,Laudet,V.,Hogenesch,J.B.,Taneja,R.&Delaunay,F.(2004)
The transcriptional repressor STRA13regulatesasubsetofper-
ipheral circadian outputs. J. Biol. Chem. 279, 1141–1150.
18. Shen, M., Kawamoto, T., Yan, W., Nakamasu, K., Tamagami,
M.,Koyano,Y.,Noshiro,M.&Kato,Y.(1997)Molecular
4418 F. Sato et al. (Eur. J. Biochem. 271) Ó FEBS 2004
characterization of the n ovel basic helix-loop -helix protein DEC1
expressedindifferentiatedhumanembryochondrocytes.Biochem.
Biophys. Res. Commun. 236, 294–298.
19. Miyazaki, K., Kawamoto, T., Tanimoto, K ., Nishiyama, M.,
Honda,H.&Kato,Y.(2002)Identificationoffunctionalhypoxia
response elements in the promoterregionoftheDEC1andDEC2
genes. J. Biol. Chem. 277, 47014–47021.
20. Fujimoto, K., Shen, M., Noshiro, M ., Matsubara, K., S hingu, S.,
Honda,K.,Yoshida,E.,Suardita,K.,Matsuda,Y.&Kato,Y.
(2001) Molecular cloning and characterization of DEC2, a new
member of basic helix-loop-helix proteins. Biochem. Biophys. Res.
Commun. 280, 164–171.
21. Guan, K.L. & Dixon, J.E. (1991) Eukaryotic proteins expressed in
Escherichia coli: an improved thrombin cleavage and purification
procedure of fusion proteins with g lutathione S-transferase. Anal.
Biochem. 192, 262–267.

22. Yun, Z., Maecker, H.L., Johnson, R.S. & Giaccia, A.J. (2002)
Inhibition of PPAR gamma 2 gene expression by t he HIF-1-
regulated gene DEC1/Stra13: a mechanism for regulation of
adipogenesis by hypo xia. De v. Cell 2, 331–341.
23. Boudjelal, M., Taneja, R., Mats ubara, S., Bouillet, P ., Dolle, P. &
Chambon, P. (1997) O verexpression of S tra13, a novel retinoic
acid-inducible gen e of the basic helix-loop-helix family, i nhibits
mesodermal and promotes neuronal d iff erentiatio n of P19 cells.
Genes Dev. 11, 2052–2065.
24. Sun, H. & Taneja, R. (2000) Stra13 expression is associated with
growth arrest and represses transcription through histone deace-
tylase (HDAC) -depend ent and HDAC-independent mechanisms.
Proc.NatlAcad.Sci.USA97, 4058–4063.
25. Garriga-Canut, M., Roopra, A. & Buckley, N.J. ( 2001) The basic
helix-loop-helix protein, sharp-1, represses transcription by a his-
tone deacetylase-dependent and histone deacetylase-independent
mechanism. J. Biol. Chem. 276, 14821–14828.
26. Atchley, W.R. & Fitch, W.M. (1997) A natural classification of the
basic helix-loop-helix class o f transcription factors. Proc. N atl
Acad. Sci. USA 94, 5172–5176.
27. Robinson, K.A. & Lope s, J. M . (2000) SURVEY and SUM-
MARY: Saccharomyces cerevisiae basic helix-loop-helix proteins
regulate diverse biological pro cesses. Nucleic A cids Res. 28 ,
1499–1505.
28. Li, Y., Song, X., Ma, Y., Liu, J., Yang, D. & Yan, B. (2004) DNA
binding but not interactio n with Bmal1 is re sponsible for DEC1-
mediated transcription regulation of the circadian gene mPer1.
Biochem. J. 382, 895–904.
29.Dhar,M.&Taneja,R.(2001)Cross-regulatoryinteraction
between Stra13 and USF results in functional antagonism.

Oncogene 20, 4750–4756.
30. Grandori, C., Cowley, S.M., James, L.P. & Eisenman, R.N. (2000)
The Myc/Max/Mad n etwork and the transcriptional control o f
cell behavior. Annu.Rev.CellDev.Biol.16, 653–699.
31. Dear, T.N., Hainzl, T., Follo, M., Nehls, M., Wilmore, H.,
Matena, K. & Boehm, T. (1997 ) I dentification of interaction
partners for the basic-helix-loop-helix protein E47. Oncogene 14,
891–898.
32. Yu, W., Ikeda, M., Abe, H., Honma, S., Ebisawa, T., Yamauchi,
T., Honma, K. & Nomura, M. (1999) Characterization of three
splice variants and genomic o rganization of the mouse BMAL1
gene. Biochem. Biophys. Res. Commun. 260, 760–767.
Ó FEBS 2004 Interaction of DEC1 with BMAL1 (Eur. J. Biochem. 271) 4419