Biochem. J. (2010) 427, 489–497 (Printed in Great Britain)
489
doi:10.1042/BJ20091856
The DNA-binding activity of mouse DNA methyltransferase 1 is regulated
by phosphorylation with casein kinase 1δ/ε
Yasunori SUGIYAMA*, Naoya HATANO†, Noriyuki SUEYOSHI*, Isao SUETAKE‡, Shoji TAJIMA‡, Eiji KINOSHITA§,
Emiko KINOSHITA-KIKUTA§, Tohru KOIKE§ and Isamu KAMESHITA*1
*Department of Life Sciences, Faculty of Agriculture, Kagawa University, Kagawa 761-0795, Japan, †Integrated Center for Mass Spectrometry, Graduate School of Medicine, Kobe
University, Kobe 657-8501, Japan, ‡Laboratory of Epigenetics, Institute for Protein Research, Osaka University, Osaka 565-0871, Japan, and §Department of Functional Molecular
Science, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima 734-8553, Japan
Dnmt1 (DNA methyltansferase 1) is an enzyme that recognizes
and methylates hemimethylated DNA during DNA replication
to maintain methylation patterns. The N-terminal region of
Dnmt1 is known to form an independent domain structure that
interacts with various regulatory proteins and DNA. In the
present study, we investigated protein kinases in the mouse brain
that could bind and phosphorylate the N-terminal regulatory
domain of Dnmt1. A protein fraction containing protein kinase
activity for phosphorylation of Dnmt1(1–290) was prepared
using Dnmt1(1–290)-affinity, DNA–cellulose and gel-filtration
columns. When the proteins in this fraction were analysed
by LC-MS/MS (liquid chromatography tandem MS), CK1δ/ε
(casein kinase 1δ/ε) was the only protein kinase identified.
Recombinant CK1δ/ε was found to bind to the N-terminal
domain of Dnmt1 and significantly phosphorylated this domain,
especially in the presence of DNA. Phosphorylation analyses
using various truncation and point mutants of Dnmt1 revealed
that the major priming site phosphorylated by CK1δ/ε was Ser146 ,
and that subsequent phosphorylation at other sites may occur
after phosphorylation of the priming site. When the DNA-binding
activity of phosphorylated Dnmt1 was compared with that of
the non-phosphorylated form, phosphorylation of Dnmt1 was
found to decrease the affinity for DNA. These results suggest
that CK1δ/ε binds to and phosphorylates the N-terminal domain
of Dnmt1 and regulates Dnmt1 function by reducing the DNAbinding activity.
INTRODUCTION
rich sequences in DNA [6], the physiological significance of the
DNA-binding activity of this region is still unclear.
Most cellular events are regulated by protein phosphorylation,
and one-third of cellular proteins are believed to be
phosphorylated. In previous studies, phosphorylation of Dnmt1
at multiple sites, including Ser515 , was reported [11,12]. Although
the regulation of gene expression through phosphorylation of
Dnmt1 by protein kinases is a very important and attractive issue,
there is little available information concerning the protein kinases
involved in the phosphorylation and regulation of Dnmt1. In
previous studies, we produced monoclonal antibodies, designated
Multi-PK antibodies, that can detect a wide variety of protein
kinases [13–17]. Using these antibodies, we detected a protein
kinase of approx. 110 kDa in a mouse brain extract that interacted
with Dnmt1, and identified it as CDKL5 (cyclin-dependent
kinase-like 5) [18]. However, the phosphorylation of Dnmt1 by
CDKL5 was very weak, and the role of this phosphorylation
remained unclear. Therefore, we hypothesized that certain other
protein kinases may phosphorylate Dnmt1.
In the present study, we used a GST (glutathione transferase)–
Dnmt1(1–290) affinity column to isolate protein kinases that
associate with and phosphorylate Dmnt1, since Dnmt1(1–290)
contains the N-terminal regulatory domain for binding various
protein factors. We monitored the DBK [Dnmt1(1–290)-binding
kinase] activity by autoradiography, using Dnmt1(1–290) as a
substrate, and Phos-tag SDS/PAGE, and identified it as CK1δ/ε
In mammals, the fifth position of cytosines in CpG sequences
in genomic DNA is often methylated [1]. DNA methylation is
essential for embryonic development and has been shown to be
important for transcriptional repression of imprinted genes [2,3].
The methylation of DNA at CpG dinucleotides is catalysed by two
classes of enzymes: the de novo type of DNA methyltransferase,
Dnmt3a and Dnmt3b, and the maintenance type, Dnmt1. Dnmt1
has a strong preference for hemimethylated CpG sites, which
are the products of DNA replication, and the enzyme plays an
important role in maintaining the methylation pattern of DNA
[4].
Dnmt1 is composed of a 180 kDa single polypeptide comprising the N-terminal regulatory domain, which covers twothirds of the molecule, and the C-terminal catalytic domain, which
contains essential motifs for the methyltransferase activity. The
N-terminal region of the regulatory domain has a DNA-binding
motif [5] and the binding sites for various protein factors. In
particular, the N-terminal 36 kDa sequence is known to form
an independent domain [6], which serves as a platform for the
recruitment of regulatory proteins, such as PCNA (proliferating
cell nuclear antigen) [7], the transcription factor DMAP1 [8],
Dnmt3a and Dnmt3b [9], and MeCP2 (methylated-CpG-binding
protein 2) [10]. Although the N-terminal region of 119–197 is
known to interact preferentially with the minor groove of AT-
Key words: casein kinase 1 (CK1), DNA-binding activity, DNA
methylation, DNA methyltransferase, domain structure, protein
phosphorylation.
Abbreviations used: CDKL5, cyclin-dependent kinase-like 5; CK1, casein kinase 1; DBK, Dnmt1(1–290)-binding kinase; DMEM, Dulbecco’s modified
Eagle’s medium; Dnmt, DNA methyltransferase; FL, full-length; GST, glutathione transferase; KD, kinase-dead; LC-MS/MS, liquid chromatography tandem
MS; PCNA, proliferating cell nuclear antigen.
1
To whom correspondence should be addressed (email [email protected]).
c The Authors Journal compilation c 2010 Biochemical Society
490
Y. Sugiyama and others
(casein kinase 1δ/ε) by LC-MS/MS (liquid chromatography
tandem MS) analysis. Recombinant CK1 isoforms were found
to associate with Dnmt1 and efficiently phosphorylate it.
Phosphorylation assays using various truncation and point
mutants of Dnmt1 revealed that the major phosphorylation site
in Dnmt1 for CK1δ/ε was Ser146 . In addition, the DNA-binding
affinity of the N-terminal region of Dnmt1 was significantly
reduced after phosphorylation by CK1δ/ε.
MATERIALS AND METHODS
Materials
ATP, BSA, α-casein from bovine milk and mouse and
rabbit anti-FLAG antibodies were purchased from Sigma
Chemicals. An anti-His6 antibody was obtained from Roche
Diagnostics. [γ -32 P]ATP (111 TBq/mmol) was purchased from
PerkinElmer. An anti-GST antibody, glutathione–Sepharose
4B and HiTrap Chelating HP columns were obtained from
GE Healthcare Biosciences. The acrylamide-dependant Phostag ligand was obtained from the Phos-tag consortium
(http://www.phos-tag.com). Dnmt1[FL (full-length); 1–1620],
GST–Dnmt1(1–290), GST–Dnmt1(1–197), GST–Dnmt1(1–146),
GST–Dnmt1(1–118), GST–Dnmt1(119–197) and His6 -tagged
PCNA were prepared as described previously [6]. An antiC-terminal Dnmt1 antibody was raised in rabbit and affinity
purified with antigen-coupled Sepharose CL-4B as described
previously [19]. Dnmt1(N; 291–1620) was expressed and
purified as described previously [20]. A monoclonal antibody
against the mouse Dnmt1(72–86) peptide was produced as
described previously [13].
Construction of plasmids
For mouse CK1δ, a 5 -upstream primer containing an NheI site
(underlined) (5 -GCTAGCATGGAGCTGAGGGTCGGGAA3 ) and a 3 -downstream primer containing an XhoI site
(bold) (5 -ACTCGAGCTTGCCGTGGTGTTCGAAAGG-3 )
were used for PCR with a mouse brain cDNA library as a
template. Mouse CK1ε was generated by amplification of cDNA
fragments using a 5 -upstream primer containing an NheI site
(underlined)
(5 -GCTAGCATGGAGTTGCGTGTGGGAAA
TAA-3 ) and a 3 -downstream primer containing an
XhoI site (bold) (5 -ACTCGAGTTTCCCAAGATGGTCAAATGGC-3 ). The PCR fragments were digested by NheI and
XhoI, and ligated into pET-23a(+) (Novagen) to generate the
plasmids pET-CK1δ and pET-CK1ε respectively. For FLAG–
CK1 isoforms, the NheI-EcoRI fragments from pET-CK1δ
and pET-CK1ε were ligated into pcFLAG [21] to generate
pcFLAG-CK1δ and pcFLAG-CK1ε respectively.
Plasmids for KD (kinase-dead) CK1δ and CK1ε, in which
Lys38 was replaced with an alanine residue [CK1δ(KD) and
CK1ε(KD)], were constructed by inverse PCR [22] using
pET-CK1δ and pET-CK1ε as templates respectively. The obtained
recombinant plasmids were designated pET-CK1δ(KD) and pETCK1ε(KD) respectively.
Mutant-1 (S146A, T149A, S150A, S152A and S153A),
mutant-2 (T156A, T159A and T160A) and mutant-3 (T163A,
T164A and T166A) constructs of GST–Dnmt1(119–197) [6]
were prepared by site-directed mutagenesis with mutagenic
primers. Mutagenesis of Ser146 , Thr149 , Ser150 , Ser152 and Ser153
was performed by inverse PCR [22] using pGEX-mDnmt1(119–
197) as a template with mutagenic primers. The obtained
recombinant plasmids were designated pGEX-Dnmt1(119–
197)(S146A),
pGEX-Dnmt1(119–197)(T149A),
pGEX
c The Authors Journal compilation c 2010 Biochemical Society
Dnmt1(119–197)(S150A),
pGEX-Dnmt1(119–197)(S152A)
and pGEX-Dnmt1(119–197)(S153A) respectively.
Protein expression and purification
pET-CK1δ, pET-CK1ε, pET-CK1δ(KD) and pET-CK1ε(KD)
were introduced into Escherichia coli strain BL21 (DE3)
(Novagen). The transformed bacteria were grown at 37 ◦C for
3 h, and then isopropyl β-D-thiogalactoside was added to a final
concentration of 0.1 mM. After 20 h at 18 ◦C, the bacteria were
harvested by centrifugation (2300 g for 10 min) and suspended
in buffer A (20 mM Tris/HCl, pH 7.5, containing 150 mM NaCl
and 0.05 % Tween 40). After sonication with a Sonifier model
450 (Branson) for 30 × 30 s with 30 s intervals on ice, cell debris
was removed by centrifugation at 20 000 g for 10 min at 4 ◦C, and
the obtained supernatant was loaded on to a HiTrap Chelating HP
column pre-equilibrated with buffer A. The column was sequentially washed with buffer A, buffer A containing 20 mM imidazole
and buffer A containing 50 mM imidazole, and then eluted with
buffer A containing 200 mM imidazole. For GST–Dnmt1(119–
197) alanine mutants, pGEX-Dnmt1(119–197) alanine mutants
were introduced into E. coli strain BL21(DE3). The transformed
bacteria were grown at 18 ◦C for 16 h, and then isopropyl βD-thiogalactoside was added to a final concentration of 0.5 mM.
After incubation at 18 ◦C for 24 h, the cells were harvested by centrifugation (2300 g for 10 min) and suspended in buffer A. After
sonication as above, cell debris was removed by centrifugation
at 20 000 g for 10 min at 4 ◦C, and the obtained supernatant was
loaded on to a glutathione–Sepharose 4B column pre-equilibrated
with buffer A. The column was washed with buffer A, and
then eluted with buffer A containing 10 mM glutathione. The
purified fractions were pooled, dialysed against buffer B (20 mM
Tris/HCl, pH 7.5, containing 0.05 % Tween 40 and 1 mM 2mercaptoethanol) and stored in aliquots at − 30 ◦C until use.
Isolation and identification of protein kinases that bind to and
phosphorylate Dnmt1(1–290)
Dnmt1(1–290)-binding proteins were prepared as described
previously [18]. The Dnmt1(1–290)-binding proteins (12 ml)
were loaded on to a DNA–cellulose (Deoxyribonucleic AcidCellulose Double-stranded from calf thymus DNA; Sigma)
column prewashed with 1 mg/ml BSA and equilibrated with
buffer B containing 0.15 M NaCl. After washing with 50 ml
of buffer B containing 0.15 M NaCl, proteins including putative
Dnmt1-binding kinases were sequentially eluted with buffer B
containing 0.3 M NaCl and 1 M NaCl. Next, the active fractions
were pooled and loaded on to a HiLoad 16/60 Superdex 200
pg column (GE Healthcare Biosciences) prewashed with buffer
B containing 0.15 M NaCl, 1 mM EDTA and 0.5 mM EGTA.
Chromatography was performed with a flow rate of 1.25 ml/min,
and 1.2 ml fractions (80 fractions in total) were collected after a
50 min run. The separation properties were determined in a first
run with molecular-mass standard proteins (β-amylase: 200 kDa,
transferrin: 81 kDa, ovalbumin: 43 kDa, and carbonic anhydrase:
29 kDa).
Fractions 37–46 (DBKs) were pooled, subjected to SDS/PAGE
and visualized by silver staining. Fourteen gel slices were excised
from each sample lane in the 20–80 kDa range, digested in-gel
with 10 μg/ml trypsin (Promega) and analysed by LC-MS/MS as
described previously [23].
GST-pulldown assay
Approx. 10 μg of CK1 or PCNA in 200 μl of buffer B containing
0.15 M NaCl was added to 20 μl of glutathione–Sepharose 4B
Phosphorylation of Dnmt1 by CK1δ/ε
prebound with 10 μg of GST–Dnmt1(1–290). After incubation
with gentle rotation at 2 ◦C for 2 h, the matrixes were washed five
times with buffer B containing 0.15 M NaCl, and the proteins
were eluted with SDS/PAGE sample buffer.
491
cellulose and rotated at 2 ◦C for 2 h. After the incubation, the
matrixes were washed three times with buffer B containing 0.15 M
NaCl, and the proteins were eluted with SDS/PAGE sample buffer.
DNA–cellulose dissociation assay
Cell culture, transfection, immunoprecipitation and
co-precipitation
Transfection of pcFLAG-CK1 isoforms into Neuro2a cells was
performed using LipofectamineTM 2000 (Invitrogen) according to
the manufacturer’s instructions. Briefly, Neuro2a cells (5 × 105 )
were plated in 35 mm dishes in 2 ml of DMEM (Dulbecco’s
modified Eagle’s medium; Sigma) containing 10 % fetal calf
serum at 24 h before transfection. Transfection was performed
by incubation of the Neuro2a cells in 1 ml of DMEM containing
5 % fetal calf serum, 6 μl of LipofectamineTM 2000 and 2.5 μg
of plasmid DNA for 24 h. For immunoprecipitation, Protein
G–Sepharose 4 Fast Flow (GE Healthcare Biosciences) was
mixed with a mouse anti-FLAG antibody in 200 μl of buffer C
(20 mM Tris/HCl, pH 7.5, 1 mM EDTA and 0.5 % Triton X-100)
containing 300 mM NaCl at 4 ◦C for 2 h, and washed three times
with buffer C containing 300 mM NaCl. Cell extracts of Neuro2a
cells expressing FLAG–CK1δ/ε were incubated with anti-FLAG
antibody-bound Sepharose at 4 ◦C for 2 h, and washed three times
with buffer C containing 300 mM NaCl and buffer C containing
150 mM NaCl. Then, the mixtures were incubated with 4 μg of
Dnmt1(FL) in 200 μl of buffer C containing 150 mM NaCl, and
washed five times with buffer C containing 150 mM NaCl. The
pulled-down proteins were subjected to Western blotting with an
anti-C-terminal Dnmt1 antibody and a rabbit anti-FLAG antibody.
Protein phosphorylation
Phosphorylation of Dnmt1 and Dnmt1 mutants by CK1 was
carried out essentially as described previously [18]. CK1 (10 ng)
was incubated with 500 ng of Dnmt1(FL) and Dnmt1 mutants in
the presence or absence of mouse genomic DNA (10 μg/ml). The
reactions were carried out in 20 mM Hepes/NaOH buffer (pH 7.4)
containing 10 mM MgCl2 , 1 mM dithiothreitol and 100 μM ATP
or [γ -32 P]ATP at 30 ◦C for 30 min. The reactions were stopped
by adding an equal volume of 2 × SDS/PAGE sample buffer.
The mixtures were electrophoresed on an SDS/polyacrylamide
gel or an SDS/polyacrylamide gel containing Phos-tag [24], and
the phosphorylated proteins were visualized by autoradiography
or subjected to Western blotting.
The stoichiometry of phosphate incorporation into GST–
Dnmt1(1–290) was determined as follows. GST–Dnmt1(1–290)
was phosphorylated by CK1 in an aforementioned reaction
mixture in a final volume of 70 μl. After incubation at 30 ◦C
for 30, 60, 90 and 120 min, a 10 μl aliquot of the mixture was
withdrawn, spotted on to a 2 cm square of the Whatman 3MM
chromatography paper (Whatman), and immediately placed in
75 mM phosphoric acid. The 32 P-phosphate incorporation into
GST–Dnmt1(1–290) was measured essentially according to the
method of Corbin and Reimann [25]. Phosphate incorporated into
GST–Dnmt1(1–290) was calculated as mol phosphate/mol GST–
Dnmt1(1–290) by subtracting the radioactivity into CK1 from that
into CK1 plus GST–Dnmt1(1–290).
DNA–cellulose binding assay
Approx. 5 μg of GST–Dnmt1(1–290), GST–Dnmt1(1–
290)(S146A) or Dnmt1(FL) was phosphorylated by CK1δ
(1 μg) essentially as described above, except that mouse genomic
DNA was removed from the reactions. Samples in 300 μl of
buffer B containing 0.15 M NaCl were mixed with 20 μl of DNA–
Approx. 5 μg of GST–Dnmt1(1–290) or Dnmt1(FL) in 250 μl of
buffer B containing 0.15 M NaCl was mixed with 20 μl of DNA–
cellulose. The mixtures were rotated at 2 ◦C for 30 min, and the
matrixes were washed three times with buffer B containing 0.15 M
NaCl. The matrixes were incubated with CK1δ (2 μg) in the
presence or absence of 100 μM ATP at 30 ◦C for 15, 30 or 45 min
under genomic DNA-free conditions. After the incubations, the
supernatants were added to an equal volume of 2 × SDS/PAGE
sample buffer, and analysed by SDS/PAGE followed by Western
blotting.
Protein determination, SDS/PAGE, Phos-tag SDS/PAGE and
Western blotting
Protein concentrations were determined by the method of
Bensadoun and Weinstein [26] using BSA as a standard.
SDS/PAGE was performed essentially according to the method of
Laemmli [27] on slab gels consisting of a 6 % or 10 % acrylamide
separating gel and a 3 % stacking gel. The resolved proteins
were transferred electrophoretically to nitrocellulose membranes,
and immunoreactive protein bands were detected essentially as
described previously [13]. For Mn2+ -Phos-tag SDS/PAGE, 20 μM
acrylamide-pendant Phos-tag ligand and 40 μM MnCl2 were
added to a 6 % separating gel before polymerization. The Phostag SDS/PAGE was performed essentially as described previously
[24].
DNA methylation assay
DNA methylation activities were determined essentially as
described previously [28]. Briefly, the methylation reaction
mixture contained 80 ng of Dnmt1(FL), 30 nM hemimethylated
oligonucleotide substrate (42-mer, 12CpG) and 5.3 μM
S-[methyl-3 H]adenosyl-L-methionine (15 Ci/mmol; Amersham
Biosciences) in 25 μl of reaction buffer comprising 5 % glycerol,
0.5 mM EDTA, 0.2 mM dithiothreitol, 0.1 mg/ml BSA and
20 mM Tris/HCl (pH 7.4). After 1 h of incubation at 37 ◦C, the
radioactivity was determined using a scintillation counter.
RESULTS
Identification of protein kinases that bind to and phosphorylate
Dnmt1(1–290)
The N-terminal domain of Dnmt1 interacts with various proteins,
indicating that it functions as a platform for the recruitment of
regulatory factors. However, there have been no reports regarding
protein kinases that interact with the N-terminal regulatory
domain of Dnmt1, except our previous report [18]. In our previous
study, we investigated a Dnmt1-binding protein kinase in a mouse
brain extract by Western blotting with Multi-PK antibodies, and
identified it as CDKL5 [18]. Although CDKL5 was found to bind
to and phosphorylate Dnmt1, the phosphorylation of Dnmt1 by
CDKL5 was not sufficiently strong enough to explain the phosphorylation of Dnmt1 by a partially purified kinase preparation.
Therefore, we assumed that there may be other protein kinases
responsible for more significant phosphorylation of Dnmt1.
To identify the protein kinase that phosphorylates Dnmt1, we
prepared a protein fraction that contained the protein kinase
responsible for phosphorylation of Dnmt1(1–290) from the
c The Authors Journal compilation c 2010 Biochemical Society
492
Figure 1
Y. Sugiyama and others
Isolation and identification of a Dnmt1-phosphorylating kinase
(A) Affinity chromatography of Dnmt1(1–290)-binding proteins using DNA–cellulose. The Dnmt1(1–290)-binding protein fraction from GST–Dnmt1(1–290) affinity chromatography was applied
to a DNA–cellulose column (1 ml). The proteins bound to the column were eluted with the indicated concentrations of NaCl. An aliquot (1 μl) of each fraction was incubated with 500 ng
of GST–Dnmt1(1–290) and 100 μM [γ -32 P]ATP (or non-radiolabelled ATP). The samples were subjected to SDS/PAGE, and the phosphorylation of GST–Dnmt1(1–290) was detected by
autoradiography (top panel). The phosphorylation was also analysed by Western blotting using an anti-mouse Dnmt1(72–86) antibody after separation by Phos-tag SDS/PAGE (bottom panel). The
fractions indicated by the horizontal bar were pooled as the active fraction. (B) Elution pattern of Dnmt1(1–290)-phosphorylating kinases from a Superdex 200 column. The active fractions from
DNA–cellulose (as described in A) were loaded on to a Superdex 200 column, and fractions of 1.2 ml (80 fractions in total) were collected after a 50 min run. An aliquot (1 μl) of each fraction
was incubated with 500 ng of GST–Dnmt1(1–290) and 100 μM ATP, and analysed by Phos-tag SDS/PAGE followed by autoradiography. The super-shifted and shifted bands of phosphorylated
Dnmt1(1–290) are indicated by the black arrowhead and arrow respectively. The white arrowheads denote the molecular masses of marker proteins applied in a separate run (β-amylase: 200 kDa;
transferrin: 81 kDa; ovalbumin: 43 kDa; and carbonic anhydrase: 29 kDa). (C) Protein staining pattern of the DBK fraction after SDS/PAGE. The eluted fractions 37–46 from the Superdex 200 column
in (B) were pooled and subjected to SDS/PAGE followed by silver staining. The gel was cut into 14 pieces as indicated and subjected to LC-MS/MS analysis. The asterisk indicates the protein band
that was identified as mouse CK1δ/ε.
mouse brain. First, the mouse brain extract was incubated with
GST–Dnmt1(1–290) bound to glutathione–Sepharose 4B, washed
with an equilibration buffer and eluted with a buffer containing
0.3 M NaCl. The Dnmt1(1–290)-bound fractions were loaded on
to a DNA–cellulose column, washed with a buffer containing
0.15 M NaCl, and then eluted with a buffer containing 0.3 M
or 1 M NaCl. Protein kinase activity toward Dnmt1(1–290) was
measured by incubating each fraction with GST–Dnmt1(1–290) in
the presence of [γ -32 P]ATP and analysed by SDS/PAGE followed
by autoradiography. Protein kinase activities that phosphorylated
Dnmt1(1–290) were detected in both the flow-through fraction
and the 0.3 M NaCl fractions (Figure 1A, upper panel).
The above fractions were also analysed by Phos-tag SDS/PAGE
[24], which detects differently phosphorylated proteins as
differently upward-shifted protein bands. When GST–Dnmt1(1–
290) was incubated with the brain extract (input) fraction
and analysed by Phos-tag SDS/PAGE followed by Western
blotting, two differently shifted bands, a significantly shifted band
(designated a super-shifted band) and a slightly shifted band,
were observed (Figure 1A, lower panel). The slightly shifted band
was observed in the pass, wash and eluted fractions, whereas the
super-shifted band was only detected in the 0.3 M NaCl fractions
(Figure 1A, lower panel). In the present study, we attempted to
identify the protein kinase that caused this super-shift of GST–
Dnmt1(1–290) upon phosphorylation.
For further purification of the protein kinase responsible for
causing the super-shifted product, a 0.3 M NaCl fraction was
loaded on to a Superdex 200 column and eluted fractions of 1.2 ml
were collected. Each fraction was incubated with GST–Dnmt1(1–
290) in the presence of [γ -32 P]ATP and analysed by Phos-tag
SDS/PAGE followed by autoradiography. Two peaks of Dnmt1(1–
c The Authors Journal compilation c 2010 Biochemical Society
290)-phosphorylating activities were detected, comprising a weak
phosphorylation activity at approx. 150 kDa (fractions 13–19) and
a significant phosphorylation activity corresponding to approx.
38 kDa (fractions 37–46) that produced a super-shifted band of
GST–Dnmt1(1–290) (Figure 1B). These results suggested that
fractions 37–46 contained the protein kinase activity responsible
for the significant phosphorylation of Dnmt1(1–290), and we
therefore designated this fraction the DBK fraction.
To identify the protein kinase that phosphorylates Dnmt1, the
DBK fraction was subjected to SDS/PAGE and 14 gel slices
in the region between 20 and 80 kDa were subjected to LCMS/MS analysis (Figure 1C). Among the gel slices evaluated,
the CK1 isoforms CK1δ and CK1ε were identified when the
gel slice containing the 45-kDa protein band was analysed
(Figure 1C and Table 1). Molecular masses of CK1δ and CK1ε are
calculated to be 46 789 and 47 122 respectively (Table 1). Since
no other protein kinases were detected in the other slices, we
concluded that the protein kinase in the DBK fraction responsible
for the significant phosphorylation of Dnmt1(1–290) was
CK1δ/ε.
When the fractions from the Superdex 200 column were
analysed by Western blotting using antibodies against CK1δ/ε,
the immunoreactive band was detected in fractions 37–46 (results
not shown). Although the elution position of CK1 from the gel
filtration column was somewhat slower than that of the predicted molecular size, it was confirmed with three independent
experiments. On the other hand, CDKL5 was detected in
fractions 13–19 by Western blotting with an anti-CDKL5 antibody
(results not shown). These results suggest that Dnmt1(1–290)
was phosphorylated by both CK1δ/ε and CDKL5, but more
significantly phosphorylated by CK1δ/ε.
Phosphorylation of Dnmt1 by CK1δ/ε
Table 1 List of identified proteins from the gel containing the 45-kDa band
that phosphorylated Dnmt1(1–290)
Protein name
NCBI ID
Mass (Da)
Mascot
score
CKIε
Pleckstrin- and Sec7-domain containing 3 isoform3
CK1δ1
CK1δ2
Microtubule-associated protein 2 (MAP 2)
gi|7798595
gi|163644296
gi|20544149
gi|20544147
gi|126741
47 122
42 274
47 286
46 789
198 860
282
145
131
131
90
493
290) by the CK1 isoforms was stimulated by the addition of
DNA (Figure 3B). When phosphorylated GST–Dnmt1(1–290)
was separated by Phos-tag SDS/PAGE, the phosphorylated bands
showed essentially the same shift patterns when CK1 isoforms
and DBK were used for phosphorylation of Dnmt1(1–290)
(Figure 3C). When the phosphate incorporation into Dnmt1(1–
290) was examined, at least two moles of phosphate were
incorporated into one mole of GST–Dnmt1(1–290). Furthermore,
the phosphate incorporation into Dnmt1 increased as the amount
of the CK1 isoforms in the reaction mixture increased (results not
shown). These results suggest that multiple phosphorylation sites
for CK1δ/ε are present in the N-terminal region of Dnmt1.
Binding of CK1δ/ε to Dnmt1(1–290)
In the present study, the protein kinase responsible for
phosphorylation of Dnmt1(1–290) was identified as CK1δ/ε.
Therefore we produced CK1δ and CK1ε using an E. coli
expression system. The recombinant CK1 isoforms were expressed as soluble forms in E. coli, and purified using HiTrap Chelating
columns.
Binding of CK1δ and CK1ε to Dnmt1 was examined by
GST-pulldown assays using GST–Dnmt1(1–290) (Figure 2A).
The CK1 isoforms were able to bind to GST–Dnmt1(1–290)
(Figure 2A, top panel). In contrast, no CK1 was coprecipitated
when GST alone was used instead of GST–Dnmt1(1–290)
(Figure 2A, middle panel), suggesting that the binding of CK1δ/ε
to Dnmt1(1–290) was specific.
Next, we examined whether or not CK1 isoforms could bind
to Dnmt1(FL). Cell extracts were prepared from Neuro2a cells
transfected with or without FLAG–CK1δ/ε. When Dnmt1(FL)
was incubated with the expressed CK1 which had been bound
to anti-FLAG antibody, Dnmt1(FL) was coprecipitated with CK1
isoforms (Figure 2B). These results suggest that CK1 isoforms are
able to bind not only to Dnmt1(1–290) but also to Dnmt1(FL).
Phosphorylation of Dnmt1(1–290) by CK1δ/ε
Next, we examined the phosphorylation of Dnmt1(1–290) by
CK1δ/ε. When GST–Dnmt1(1–290) was incubated with CK1δ,
CK1ε or DBK in the presence of [γ -32 P]ATP, Dnmt1(1–290)
was significantly phosphorylated. GST–Dnmt1(1–290) served
as an effective substrate for CK1δ/ε, simlar to the case for
casein, which is known to be the most efficient substrate for
CK1 (Figure 3A). Furthermore, phosphorylation of Dnmt1(1–
Figure 2
Determination of the phosphorylation sites of Dnmt1 by CK1δ/ε
Next, the phosphorylation sites in Dnmt1 for CK1δ/ε
were investigated. For this experiment, we prepared purified
preparations of Dnmt1(FL) and various deletion mutants of
Dnmt1 (Figure 4A and B). Dnmt1(FL) and Dnmt1(N; 291–
1620), an N-terminal truncation mutant of Dnmt1, were incubated
with CK1δ and CK1ε in the presence of [γ -32 P]ATP and analysed
by SDS/PAGE followed by autoradiography. Dnmt1(FL) and
Dnmt1(1–290) were significantly phosphorylated by the CK1
isoforms, whereas Dnmt1(N) was not (Figure 4C), suggesting
that the phosphorylation sites are located in the N-terminal region. Therefore various truncation mutants of the N-terminal
region of Dnmt1, such as GST–Dnmt1(1–197), GST–Dnmt1(1–
146) and GST–Dnmt1(1–119), were prepared and compared
with GST–Dnmt1(1–290). Among these mutants, Dnmt1(1–290)
and Dnmt1(1–197) were efficiently phosphorylated by CK1.
Although Dnmt1(1–146) was also significantly phosphorylated
by CK1, the phosphate incorporation into Dnmt1(1–146) was
reduced to 63 % of that into Dnmt1(1–290) (Figure 4D).
In contrast, Dnmt1(1–118) was not phosphorylated at all.
These results suggest that the major phosphorylation sites in
Dnmt1 are located in the N-terminal region between amino
acids 119 and 146. To determine further the phosphorylation
sites in this region, three mutants of Dnmt1(119–197) whose
serine/threonine clusters were replaced by alanine residues
were produced as follows: mutant-1 (S146A, T149A, S150A,
S152A and S153A); mutant-2 (T156A, T159A and T160A); and
mutant-3 (T163A, T164A and T166A). Mutant-2 and mutant3 were phosphorylated by the CK1 isoforms to essentially
the same extent as wild-type Dnmt1(119–197), whereas the
Binding of CK1δ/ε to Dnmt1
(A) Binding assay of CK1 isoforms to Dnmt1(1–290) by GST pulldown. Approx. 10 μg of recombinant CK1δ and CK1ε were mixed with 20 μl of GST–Dnmt1(1–290)-bound (top) or GST-bound
(middle) glutathione–Sepharose 4B, and pulled down by centrifugation. The pulled-down CK1 isoforms were resolved by SDS/PAGE and detected by Western blotting with an anti-His6 antibody. (B)
Immunoprecipitation of CK1 isoforms with Dnmt1(FL). Cell extracts of Neuro2a cells expressing FLAG–CK1δ/ε were immunoprecipitated with mouse anti-FLAG antibody. FLAG–CK1δ/ε–Sepharose
was incubated with 4 μg of Dnmt1(FL). The pulled-down proteins were subjected to Western blotting with an anti-(C-terminal Dnmt1) antibody and a rabbit anti-FLAG antibody.
c The Authors Journal compilation c 2010 Biochemical Society
494
Y. Sugiyama and others
phosphorylating conditions in the presence of ATP, whereas that
of the nonphosphorylatable mutant Dnmt1(1–290)(S146A) was
not (Figure 5A, left panel). The DNA-binding activity of Dnmt1
was also examined using Dnmt1(FL) instead of Dnmt1(1–290).
In this experiment, we found that the DNA-binding activity of
Dnmt1(FL) was also decreased when it was phosphorylated by
CK1 (Figure 5A, right panel).
In the next experiment, we examined whether Dnmt1(1–
290) and Dnmt1(FL) were dissociated from DNA–cellulose
upon phosphorylation by CK1δ. When GST–Dnmt1(1–290) or
Dnmt1(FL) bound to DNA–cellulose was incubated with CK1δ
in the presence or absence of ATP, Dnmt1(1–290) and Dnmt1(FL)
were gradually dissociated from DNA–cellulose when incubated
under phosphorylating conditions (Figure 5B). These results
suggest that the DNA-binding activity of the N-terminal domain
of Dnmt1 is decreased by phosphorylation with CK1δ and that
the phosphorylation at the N-terminal domain of Dnmt1 induces
dissociation of Dnmt1 from DNA.
Effects of phosphorylation by CK1δ/ε on the function of Dnmt1
Figure 3
Phosphorylation of Dnmt1 by CK1δ/ε
(A) Phosphorylation of Dnmt1(1–290) and casein by CK1 isoforms. The DBK fraction
(1 μl) or CK1 isoforms (10 ng) were incubated with 500 ng of GST–Dnmt1(1–290) or
casein in the presence of 100 μM [γ -32 P]ATP. Phosphorylation of the proteins was
analysed by SDS/PAGE followed by autoradiography. (B) Effects of genomic DNA on the
phosphorylation of Dnmt1(1–290). Approx. 10 ng of CK1 isoforms was incubated with 500 ng
of GST–Dnmt1(1–290) in the presence or absence of mouse genomic DNA (10 μg/ml).
Phosphorylation of the proteins was analysed by SDS/PAGE followed by autoradiography.
The asterisks indicate autophosphorylation of CK1. (C) Phos-tag SDS/PAGE showing the
shifted bands of phosphorylated Dnmt1(1–290). The phosphorylation of GST–Dnmt1(1–290)
(500 ng) by DBK (1 μl) or CK1 isoforms (10 ng) was analysed by Phos-tag SDS/PAGE followed
by autoradiography. The arrowhead and arrow indicate the super-shifted and shifted bands of
phosphorylated GST–Dnmt1(1–290) respectively.
phosphorylation of mutant-1 was almost completely abolished
(Figure 4E). To identify further the phosphorylation sites in
this sequence, five alanine residue substitutions of Dnmt1(119–
197), namely S146A, T149A, S150A, S152A and S153A, were
constructed and analysed for their phosphorylation by the CK1
isoforms. The phosphorylation of Dnmt1(119–197)(S146A) was
almost completely abolished compared with that of wild-type
Dnmt1(119–197) and the other mutants (Figure 4F). These results
suggest that Ser146 of Dnmt1 is the major phosphorylation site for
CK1δ/ε, and that phosphorylation at this site may play a potential
role as the priming site for phosphorylation.
Regulation of the DNA-binding activity of Dnmt1 by
phosphorylation with CK1δ
The N-terminal region of Dnmt1 has been reported to recognize
and bind to AT-rich sequences in DNA [6]. To investigate the effect
of Dnmt1 phosphorylation by CK1δ on its DNA-binding activity,
Dnmt1(1–290) and Dnmt1(1–290)(S146A) were incubated with
CK1δ in the presence or absence of ATP and their DNAbinding activities were compared. Briefly, these proteins were
incubated with DNA–cellulose, and their DNA-binding abilities were quantified by Western blotting. The DNA-binding
activity of Dnmt1(1–290) was decreased after incubation under
c The Authors Journal compilation c 2010 Biochemical Society
In the present study, we have demonstrated that CK1δ/ε-mediated
phosphorylation of the N-terminal domain of Dnmt1 reduces the
affinity of Dnmt1 for DNA. The next question to be resolved
was the effect of phosphorylation on the function of Dnmt1. To
investigate the effects of phosphorylation on the DNA methylation
activity of Dnmt1, Dnmt1 was incubated with wild-type or KD
forms of CK1δ and CK1ε under phosphorylating conditions
and the DNA methylation activities were measured. The DNA
methylation activity of phosphorylated Dnmt1 did not differ
from that of non-phosphorylated Dnmt1 (Figure 6A), indicating
that the catalytic activity of Dnmt1 remained unchanged upon
phosphorylation by CK1δ/ε under the conditions used.
The N-terminal domain of Dnmt1 functions as a platform for
the binding of various modulator proteins. PCNA is known to
bind to this region and Dnmt1–PCNA complexes are recruited
to replication forks at S-phase during the cell cycle [6,29].
Furthermore, since the DNA-binding sequence in this region
partly overlaps with the PCNA-binding motif [6], we examined
the effects of phosphorylation by CK1δ/ε on the PCNA-binding
activity of Dnmt1. GST–Dnmt1(1–290) or GST was incubated
with CK1δ in the presence or absence of ATP and then
coupled to glutathione–Sepharose 4B. After washing of CK1
with 0.3 M NaCl, the GST- and GST–Dnmt1(1–290)-bound
glutathione–Sepharose 4B preparations were mixed with PCNA
and pulled down by centrifugation. PCNA was able to bind
to GST–Dnmt1(1–290), but not to GST alone, indicating that
PCNA specifically associated with the N-terminal domain of
Dnmt1. However, the PCNA-binding ability to Dnmt1 remained
unchanged under both phosphorylating and non-phosphorylating
conditions (Figure 6B), suggesting that the association of PCNA
with Dnmt1 is not regulated by phosphorylation with CK1δ/ε.
DISCUSSION
Dnmt1 was reported to be phosphorylated in vivo and its
function may be regulated through protein phosphorylation
[11,12]. However, the protein kinases responsible for the
phosphorylation of Dnmt1 and the physiological significance of
its phosphorylation remain to be elucidated. It is known that the
N-terminal region of Dnmt1 forms a structurally independent
domain from the other parts of the enzyme, including the catalytic
domain [6] and that the N-terminal domain serves as a platform
for the recruitment of various regulatory proteins [7–10]. In our
previous study, we identified CDKL5 as a protein kinase that
Phosphorylation of Dnmt1 by CK1δ/ε
Figure 4
495
Determination of the phosphorylation site in Dnmt1
(A) Schematic representations of the Dnmt1 truncation mutants used in the present study. The amino acid numbers of the truncation mutants are indicated on the left. The relative strengths of the
phosphorylation of Dnmt1 and its mutants by CK1δ/ε are shown on the right. (B) Protein staining pattern of recombinant Dnmt. Approx. 0.5 μg of Dnmt1(FL), Dnmt1(N), GST–Dnmt1(1–290),
GST–Dnmt1(1–197), GST–Dnmt1(1–146) and GST–Dnmt1(1–118) were subjected to SDS/PAGE 6 % (left panel) or 10 % (right panel) gels and stained with Coomassie Brilliant Blue.
(C) Phosphorylation of Dnmt1(FL), the N-terminal deletion mutant Dnmt1(N) and Dnmt1(1–290) by CK1δ/ε. GST–Dnmt1(1–290), Dnmt1(N) and Dnmt1(FL) (500 ng of each) were phosphorylated
by DBK (1 μl) or CK1 isoforms (10 ng), and analysed by SDS/PAGE followed by autoradiography. (D) Phosphorylation of various Dnmt1 truncation mutants by CK1 isoforms. GST and truncation
mutants of GST–Dnmt1, comprising GST–Dnmt1(1–290), GST–Dnmt1(1–197), GST–Dnmt1(1–146) and GST–Dnmt1(1–118) (500 ng of each) were phosphorylated by CK1δ (10 ng) and analysed
by SDS/PAGE followed by autoradiography. [32 P]Phosphate incorporation into Dnmt1 deletion mutants was quantified and the relative values (%) indicated below the panel. (E) Determination of the
phosphorylation locus using alanine-substituted mutants of Dnmt1(119–197). Alanine residue mutants of GST–Dnmt1(119–197), comprising mutant-1 (S146A, T149A, S150A, S152A and S153A),
mutant-2 (T156A, T159A and T160A) and mutant-3 (T163A, T164A and T166A), (500 ng of each) were incubated with DBK (1 μl) or CK1 isoforms (10 ng) in the presence of [γ -32 P]ATP, and
analysed by SDS/PAGE followed by autoradiography. (F) Determination of the phosphorylation site in Dnmt1 by CK1δ/ε. Point mutants of GST–Dnmt1(119–197)(S146A), namely T149A, S150A,
S152A and S153A (500 ng each) were phosphorylated with DBK (1 μl) or CK1 isoforms (10 ng), and analysed by SDS/PAGE followed by autoradiography. Relative phosphorylation level of point
mutants of GST–Dnmt1(119–197) were quantified by Scion Image and indicated as percentage values below the panel. Autorad., autoradiation; CBB, Coomassie Brilliant Blue.
binds to the N-terminal region of Dnmt1 [18]. CDKL5 was found
to bind to and phosphorylate Dnmt1 at its N-terminal domain,
although the physiological significance of the phosphorylation by
CDKL5 is still unclear. In the present study, we demonstrated
that CK1δ/ε also binds to the N-terminal regulatory domain of
Dnmt1 and phosphorylates Dnmt1 much more significantly than
CDKL5. Futhermore, we identified the major phosphorylation site
in the N-terminal region of Dnmt1 as Ser146 . We also demonstrated
that the DNA-binding affinity of the N-terminal region of Dnmt1
was significantly reduced by phosphorylation with CK1δ. These
results indicate that CK1δ/ε binds to and phosphorylates the
N-terminal region of Dnmt1, thereby regulating the functions of
Dnmt1 through a reduction in its DNA-binding activity.
The final test would be to demonstrate that the CK1
phosphorylation site Ser146 , in the N-terminal domain of Dnmt1,
is phosphorylated in vivo. All our attempts, however, to detect
phosphorylation at this site by MS analysis were unsuccessful
(results not shown). Difficulty in detecting phosphorylation at
Ser146 of Dnmt1 in vivo might be due to the occurrence of
multiple phosphorylation sites in the serine/threonine cluster
region adjacent to Ser146 . It is known that the consensus
sequences of CK1 are E/D-X1−2 -S/T and S/T(P)-X1−2 -S/T (where
c The Authors Journal compilation c 2010 Biochemical Society
496
Y. Sugiyama and others
Figure 6
Figure 5
Effects of phosphorylation on the DNA-binding activity of Dnmt1
(A) DNA-binding activities of the phosphorylated and nonphosphorylated forms of Dnmt1.
Approx. 5 μg of GST–Dnmt1(1–290), GST–Dnmt1(1–290)(S146A) or Dnmt1(FL) was incubated
with CK1δ (1 μg) in the presence or absence of 100 μM ATP as indicated, and the mixtures
were then rotated with 20 μl of DNA–cellulose. After washing, the DNA–cellulose-bound
GST–Dnmt1(1–290) or Dnmt1(FL) were eluted with SDS sample buffer and analysed by Western
blotting with an anti-GST antibody or an anti-(C-terminal Dnmt1) antibody respectively. Relative
densities of immunoreactive bands of GST–Dnmt1(1–290), GST–Dnmt1(1–290)(S146A)
and Dnmt1(FL) were quantitated by Scion Image and are indicated as percentage values
below the panel. (B) Dissociation of Dnmt1 from DNA–cellulose by phosphorylation.
GST–Dnmt1(1–290) or Dnmt1(FL) (5 μg) was mixed with 20 μl of DNA–cellulose. After
washing, the DNA–cellulose-bound GST–Dnmt1(1–290) or Dnmt1(FL) was incubated with CK1δ
(2 μg) in the presence or absence of 100 μM ATP. At the indicated times, GST–Dnmt1(1–290)
and Dnmt1(FL) released from the DNA–cellulose were monitored by Western blotting with an
anti-GST antibody or an anti-(C-terminal Dnmt1) antibody respectively.
X is any amino acid) [30,31]. Therefore a substrate including
a serine/threonine cluster is often phosphorylated at multiple
residues after initiation from a priming phosphorylation site.
The N-terminal domain of Dnmt1 has a serine/threonine cluster,
including the consensus sequences for CK1, and this domain was
actually phosphorylated at multiple sites by CK1δ/ε. In addition,
Dnmt1(119–197)(S146A) was not phosphorylated by CK1δ/ε at
all. These results suggest that the multiple sites in the N-terminal
region of Dnmt1 may be sequentially phosphorylated by CK1δ/ε
after phosphorylation of Ser146 as the priming site.
Although the N-terminal region of Dnmt1 is known to interact
with AT-rich sequences in DNA [6], the physiological meaning of
the DNA-binding activity of this domain in Dnmt1 is unknown. In
the present study, we found that the DNA-binding activity of the
N-terminal region of Dnmt1 was decreased by phosphorylation
with CK1δ. However, the DNA methylation activity of Dnmt1
was not changed by phosphorylation with CK1δ/ε under the
present conditions. These results agree with previous reports that
the N-terminal domain of Dnmt1 is not required for the DNA
methylation activity [19,32]. Therefore the physiological role of
the DNA-binding activity of the N-terminal domain of Dnmt1 is
the next issue to be studied.
The N-terminal region of Dnmt1 is also known to act as a
platform for interactions with various proteins. The N-terminal
sequence of amino acids 119–197 is necessary for the DNAbinding activity, and this region also contains a PCNA-binding
c The Authors Journal compilation c 2010 Biochemical Society
Effects of phosphorylation on the functions of Dnmt1
(A) Effects of phosphorylation by CK1δ/ε on the DNA methylation activity of Dnmt1. Dnmt1(FL)
(400 ng) was incubated with wild-type or KD CK1δ (100 ng) in the presence of 100 μM ATP.
After incubation at 30 ◦C for 30 min, 80 ng of Dnmt1(FL) was assayed for its DNA methylation
activity by incubation with S -[methyl-3 H]adenosyl-L-methionine at 37 ◦C for 1 h. Incorporation
of 3 H into a hemimethylated oligonucleotide substrate (42-mer, 12CpG) was determined using
a liquid scintillation counter. The relative DNA methylation activity was determined as the fold
increase by comparing the active CK1 with the inactive CK1(KD). Each bar represents the
mean +
− S.E.M. of three separate experiments. (B) Effects of phosphorylation of Dnmt1 by CK1
on the PCNA-binding activity of Dnmt1(1–290). GST–Dnmt1(1–290) (10 μg) was rotated with
20 μl of glutathione–Sepharose 4B. After washing, the mixtures were incubated with CK1δ
(2 μg) in the presence or absence of 100 μM ATP at 30 ◦C for 1 h, and CK1 was then washed
out with 0.3 M NaCl. The mixtures were subsequently incubated with PCNA (10 μg), and pulled
down by centrifugation. The pulled-down PCNA was detected by Western blotting with an
anti-His6 antibody.
motif at amino acids 160–172 [6]. Therefore, we examined the
effects of phosphorylation by CK1δ/ε on the PCNA-binding
activity of Dnmt1. However, the PCNA-binding activity of Dnmt1
remained unchanged after phosphorylation with CK1δ. The
N-terminal region of Dnmt1(1–290) has an isoelectric point of
8.55, and this isoelectric point of the N-terminal region would be
significantly changed by multiple phosphorylations in this region.
Therefore we need to investigate the effects of phosphorylation
by CK1δ/ε on the binding activities of the various proteins that
associate with the N-terminal regulatory domain.
It is known that CK1 is mainly localized in the cytoplasm
and that the activity of cytoplasmic CK1 is repressed by
autophosphorylation at the C-terminal autoinhibitory domain
[33]. The kinase activity of CK1 is important for its localization
and the KD form of CK1 is localized in the nucleus. In addition,
CK1 has putative nuclear localization sequences and a significant
proportion of this protein is often localized in the nucleus,
including the central body and spindle poles during mitosis [34].
On the other hand, Dnmt1 is known to be a typical nuclear protein
[4], although the majority of Dnmt1 in neurons is localized in the
cytoplasm [35]. These findings suggest that both Dnmt1 and CK1
may co-localize and interact with each other during the cell cycle
or in cells such as neurons. Further investigations concerning the
relationship between Dnmt1 and CK1 in vivo are necessary to
elucidate the mechanisms of the regulation of gene expression
through DNA methylation.
AUTHOR CONTRIBUTION
Yasunori Sugiyama performed most experiments, summarized data and contributed to
manuscript preparation. Naoya Hatano performed the LC-MS/MS analysis. Isao Suetake
and Shoji Tajima contributed to studies involving the DNA methylation assay. Eiji Kinoshita,
Emiko Kinoshita-Kikuta and Tohru Koike contributed to studies involving Phos-tag
SDS/PAGE. Noriyuki Sueyoshi and Isamu Kameshita designed experiments, analysed
data and were involving in manuscript preparation.
Phosphorylation of Dnmt1 by CK1δ/ε
FUNDING
This work was supported in part by grants-in-aid for Scientific Research from the Ministry
of Education, Culture, Sports, Science and Technology of Japan [grant numbers 21510226
(to I.K.), 21770144 and 19770110 (to N.S.)], and Research Fellowships of the Japan Society
for the Promotion of Science for Young Scientists [grant number 19-1637 (to Y.S.)].
REFERENCES
1 Ehrlich, M., Gama-Sosa, M. A., Huang, L. H., Midgett, R. M., Kuo, K. C., McCune, R. A.
and Gehrke, C. (1982) Amount and distribution of 5-methylcytosine in human DNA from
different types of tissues of cells. Nucleic Acids Res. 10, 2709–2721
2 Okano, M., Bell, D. W., Haber, D. A. and Li, E. (1999) DNA methyltransferases Dnmt3a
and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99,
247–257
3 Kaneko-Ishino, T., Kohda, T. and Ishino, F. (2003) The regulation and biological
significance of genomic imprinting in mammals. J. Biochem. (Tokyo) 133, 699–711
4 Bestor, T. H. (2000) The DNA methyltransferases of mammals. Hum. Mol. Genet. 9,
2395–2402
5 Chuang, L. S, Ng, H. H., Chia, J. N. and Li, B. F. (1996) Characterisation of independent
DNA and multiple Zn-binding domains at the N-terminus of human DNA-(cytosine-5)
methyltransferase: modulating the property of a DNA-binding domain by contiguous
Zn-binding motifs. J. Mol. Biol. 257, 935–948
6 Suetake, I., Hayata, D. and Tajima, S. (2006) The amino-terminus of mouse DNA
methyltransferase 1 forms an independent domain and binds to DNA with the sequence
involving PCNA binding motif. J. Biochem. (Tokyo) 140, 763–776
7 Chuang, L. S., Ian, H. I., Koh, T. W., Ng, H. H., Xu, G. and Li, B. F. (1997) Human
DNA-(cytosine-5) methyltransferase-PCNA complex as a target for p21WAF1. Science
277, 1996–2000
8 Rountree, M. R., Bachman, K. E. and Baylin, S. B. (2000) DNMT1 binds HDAC2 and a
new corepressor, DMAP1, to form a complex at replication foci. Nat. Genet. 25, 269–277
9 Kim, G. D., Ni, J., Kelesoglu, N., Roberts, R. J. and Pradhan, S. (2002) Co-operation and
communication between the human maintenance and de novo DNA (cytosine-5)
methyltransferases. EMBO J. 21, 4183–4195
10 Kimura, H. and Shiota, K. (2003) Methyl-CpG-binding protein, MeCP2, is a target
molecule for maintenance DNA methyltransferase, Dnmt1. J. Biol. Chem. 278,
4806–4812
11 Goyal, R., Rathert, P., Laser, H., Gowher, H. and Jeltsch, A. (2007) Phosphorylation of
Serine-515 activates the mammalian maintenance methyltransferase Dnmt1. Epigenetics
2, 155–160
12 Glickman, J. F., Pavlovich, J. G. and Reich, N. O. (1997) Peptide mapping of the murine
DNA methyltransferase reveals a major phosphorylation site and the start of translation.
J. Biol. Chem. 272, 17851–17857
13 Kameshita, I., Tsuge, T., Kinashi, T., Kinoshita, S., Sueyoshi, N., Ishida, A., Taketani, S.,
Shigeri, Y., Tatsu, Y., Yumoto, N. and Okazaki, K. (2003) A new approach for the detection
of multiple protein kinases using monoclonal antibodies directed to the highly conserved
region of protein kinases. Anal. Biochem. 322, 215–224
14 Sugiyama, Y., Sueyoshi, N., Shigeri, Y., Tatsu, Y., Yumoto, N., Ishida, A., Taniguchi, T. and
Kameshita, I. (2005) Generation and application of a monoclonal antibody that detects a
wide variety of protein tyrosine kinases. Anal. Biochem. 347, 112–120
15 Kinoshita, S., Sueyoshi, N., Shoju, H., Suetake, I., Nakamura, M., Tajima, S. and
Kameshita, I. (2004) Cloning and characterization of a novel Ca2+ /calmodulin-dependent
protein kinase I homologue in Xenopus laevis . J. Biochem. 135, 619–630
16 Kameshita, I., Nishida, T., Nakamura, S., Sugiyama, Y., Sueyoshi, N., Umehara, Y.,
Nomura, M. and Tajima, S. (2005) Expression cloning of a variety of novel protein kinases
in Lotus japonicas . J. Biochem. (Tokyo) 137, 33–39
497
17 Kameshita, I., Yamada, Y., Nishida, T., Sugiyama, Y., Sueyoshi, N., Watanabe, A. and
Asada, Y. (2007) Involvement of Ca2+ /calmodulin-dependent protein kinase in mycelia
growth of the basidiomycetous mushroom, Coprinus cinereus . Biochim. Biophys. Acta
1770, 1395–1403
18 Kameshita, I., Sekiguchi, M., Hamasaki, D., Sugiyama, Y., Hatano, N., Suetake, I.,
Tajima, S. and Sueyoshi, N. (2008) Cyclin-dependent kinase-like 5 binds and
phosphorylates DNA methyltransferase 1. Biochem. Biophys. Res. Commun. 377,
1162–1167
19 Suetake, I., Shi, L., Watanabe, D., Nakamura, M. and Tajima, S. (2001) Proliferation
stage-dependent expression of DNA methyltransferase (Dnmt1) in mouse small intestine.
Cell Struct. Funct. 26, 79–86
20 Vilkaitis, G., Suetake, I., Klimasauskas, S. and Tajima, S. (2005) Proessive methylation of
hemimethylated CpG sites by mouse Dnmt1 DNA methyltransferase. J. Biol. Chem. 280,
64–72
21 Shimomura, S., Nagamine, T., Nimura, T., Sueyoshi, N., Shigeri, Y. and Kameshita, I.
(2007) Expression, characterization, and gene knockdown of zebrafish doublecortin-like
protein kinase. Arch. Biochem. Biophys. 463, 218–230
22 Imai, Y., Matsushima, Y., Sugimura, T. and Terada, M. (1991) A simple and rapid method
for generating a deletion by PCR. Nucleic Acids Res. 19, 2785
23 Fujimoto, T., Yurimoto, S., Hatano, N., Nozaki, N., Sueyoshi, N., Kameshita, I., Mizutani,
A., Mikoshiba, K., Kobayashi, R. and Tokumitsu, H. (2008) Activation of SAD kinase by
Ca2+ /calmodulin-dependent protein kinase kinase. Biochemistry 47, 4151–4159
24 Yamada, S., Nakamura, H., Kinoshita, E., Kinoshita-Kikuta, E., Kioke, T. and Shiro, Y.
(2007) Separation of a phosphorylated histidine protein using phosphate affinity
polyacrylamide gel electrophoresis. Anal. Biochem. 360, 160–162
25 Corbin, J. D. and Reimann, E. M. (1974) Assay of cyclic AMP dependent protein kinases.
Methods Enzymol. 38, 287–290
26 Bensadoun, A. and Weinstein, D. (1976) Assay of proteins in the presence of interfering
materials. Anal. Biochem. 70, 241–250
27 Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature 227, 680–685
28 Suetake, I., Miyazaki, J., Murakami, C., Takeshima, H. and Tajima, S. (2003) Distinct
enzymatic properties of recombinant mouse DNA methyltransferases Dnmt3a and
Dnmt3b. J. Biochem. (Tokyo) 133, 737–744
29 Mortusewicz, O., Schermelleh, L., Walter, J., Cardoso, M. C. and Leonhardt, H. (2005)
Recruitment of DNA methyltransferase I to DNA repair sites. Proc. Natl. Acad. Sci. U.S.A.
102, 8905–8909
30 Songyang, Z., Lu, K. P., Kwon, Y. T., Tsai, L. H., Filhol, O., Cochet, C., Brickey, D. A.,
Soderling, T. R., Bartleson, C., Graves, D. J. et al. (1996) A structural basis for substrate
specificities of protein Ser/Thr kinases: primary sequence preference of casein kinases I
and II, NIMA, phosphorylase kinase, calmodulin-dependent kinase II, CDK5, and Erk1.
Mol. Cell. Biol. 16, 6486–6493
31 Kreegipuu, A., Blom, N., Brunak, S. and Jarv, J. (1998) Statistical analysis of protein
kinase specificity determinants. FEBS Lett. 430, 45–50
32 Zimmermann, C., Guhl, E. and Graessmann, A. (1997) Mouse DNA methyltransferase
(MTase) deletion mutants that retain the catalytic domain display neither de novo nor
maintenance methylation activity in vivo . Biol. Chem. 378, 393–405
33 Knippschild, U., Gocht, A., Wolff, S., Huber, N., Lohler, J. and Stoter, M. (2005) The
casein kinase 1 family: participation in multiple cellular processes in eukaryotes.
Cell. Signaling 17, 675–689
34 Milne, D. M., Looby, P. and Meek, D. W. (2001) Catalytic activity of protein kinase CK1δ
(casein kinase 1δ) is essential for its normal subcellular localization. Exp. Cell Res. 263,
43–54
35 Inano, K., Suetake, I., Ueda, T., Miyake, Y., Nakamura, M., Okada, M. and Tajima, S.
(2000) Maintenance-type DNA methyltransferase is highly expressed in post-mitotic
neurons and localized in the cytoplasm compartment. J. Biochem. (Tokyo) 128, 315–321
Received 7 December 2009/24 February 2010; accepted 1 March 2010
Published as BJ Immediate Publication 1 March 2010, doi:10.1042/BJ20091856
c The Authors Journal compilation c 2010 Biochemical Society