Automethylation of CARM1 allows coupling of
transcription and mRNA splicing
Peter Kuhn1, Rob Chumanov1, Yidan Wang1, Ying Ge2,
Richard R. Burgess1and Wei Xu1,*
1McArdle Laboratory for Cancer Research and2Human Proteomics Program, School of Medicine and Public
Health, University of Wisconsin, Madison, WI 53706, USA
Received June 23, 2010; Revised October 29, 2010; Accepted November 17, 2010
Coactivator-associated arginine methyltransferase
1 (CARM1), the histone arginine methyltransferase
and coactivator for many transcription factors, is
subject to multiple post-translational modifications
(PTMs). To unbiasedly investigate novel CARM1
mass spectrometry. Surprisingly, mouse CARM1
expressed in insect and mammalian expression
systems was completely dimethylated at a single
site in the C-terminal domain (CTD). We demon-
strate that dimethylation of CARM1 occurs both
invivo andin vitro
automethylation mechanism. To probe function of
automethylation, we mutated arginine 551 to lysine
to create an automethylation-deficient CARM1.
Although mutation of CARM1’s automethylation
site did not affect its enzymatic activity, it did
impair both CARM1-activated transcription and
pre-mRNA splicing. These results strongly imply
that automethylation of CARM1 provides a direct
link to couple transcription and pre-mRNA splicing
in a manner differing from the other steroid receptor
coactivators. Furthermore, our study identifies a
CARM1’s catalytic domain to its CTD.
Protein arginine methyltransferases (PRMTs) are a family
of enzymes that mono-and di-methylate arginine residues
on protein substrates. The PRMTs are generally classified
as either type I or II enzymes. While both types catalyze
the formation of a monomethylated arginine intermediate,
type I PRMTs further catalyze the production of an
asymmetrical dimethylarginine, and type II PRMTs
catalyze the formation of a symmetrical dimethylarginine
(1). PRMTs play multiple roles in cellular function,
including the regulation of transcription, pre-mRNA
splicing, ribosome biogenesis, cytokine signaling and
DNA repair (2).
Coactivator-associated arginine methyltransferase 1
(CARM1), also known as PRMT4, is a type I PRMT
which was originally identified as an associated protein
(GRIP1), the p160 family steroid receptor coactivator
(3). CARM1 is best characterized as a coactivator for a
number of transcription factors, either relying on GRIP1
as a tether or binding directly to these factors (3–5).
CARM1 activates transcription via multiple mechanisms,
including histone H3 methylation at R17 (6,7), methyla-
tion of other key coactivators (8,9), and recruitment of
chromatin remodeling proteins (10). Two functional
domains have been identified in CARM1: a central cata-
lytic core and the C-terminal domain (CTD). The central
catalytic corecontains both
(AdoMet) binding and substrate binding domains, a struc-
tural characteristic of PRMT family proteins (11,12). The
methyltransferase activity of CARM1 plays a pivotal role
in its ability to regulate transcription (3) with one excep-
tion (13). The enzyme-dead CARM1 knockin mice
have defects similar to those seen in their knockout
counterparts (14). These observations suggest that the
enzymatic activity of CARM1 is essential for most
CARM1-regulated processes. In contrast to the catalyt-
ic core,the CTDof CARM1
PRMTs. While it is dispensable for methylation of
known CARM1 substrates (11,12), deletion of the CTD
of CARM1 greatly impairs transcriptional coactivation
by CARM1 (15).
Multiple steroid receptor coactivators regulate alterna-
tive splicing. PGC-1a, CoAA and CAPER harbor RNA-
recognition motifs (RRMs) and/or arginine–serine-rich
*To whom correspondence should be addressed. Tel: 608 265 5540; Fax: 608 262 2824; Email: firstname.lastname@example.org
Published online 7 December 2010 Nucleic Acids Research, 2011, Vol. 39, No. 72717–2726
? The Author(s) 2010. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/
by-nc/2.5), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
regions characteristic of SR splicing factors, consistent
with their involvement in mRNA processing (16–18).
implicating itin splicing.
showed that RNA-binding proteins HuR and HuD
(19,20) and several splicing and transcription elongation
factors (SmB, SAP49, U1C and CA150) (21) are bona fide
substrates for CARM1. CA150, a molecule that links
transcription to splicing, only interacts with SMN, a
spliceosome component, when it is methylated by
CARM1 (21). Furthermore, the enzymatic activity of
CARM1 is required to promote exon skipping on CD44
(21). Thus, CARM10s enzymatic activity appears to
regulate both transcription and pre-mRNA splicing
events, although the mechanism of CARM1’s regulation
of these seemingly coupled events remains unknown.
CARM1 is unique among PRMTs in that various
identified post-translational modifications (PTMs) on
CARM1 are emerging as alternative mechanisms to
regulate its function. Phosphorylation of CARM1 occurs
on at least three serine residues. Phosphorylation of two
CARM1 residues appears to regulate the enzymatic
activity of CARM1 during mitosis (22,23). We previously
observed that phosphorylation of serine 228 blocks
CARM1 dimerization, interrupting proper binding of
the dimerization ‘arm’ and thus inhibiting its methylation
activity (23). Phosphorylation at serine 217 appears to
blocks AdoMet binding to the catalytic site of CARM1
(22). More recently, serine 448 on CARM1 was found to
be phosphorylated by PKA, which facilitates a direct
interaction between CARM1 and estrogen receptor a
modified by O-linked acetylglucosamine (25), however
the location and function of this modification remains to
Top-down mass spectrometry (MS) is a powerful
approach for mapping PTMs on a protein of interest
(26–31). In top-down MS, the intact protein ions are
directly introduced into the gas phase, which allows
highly accurate determination of the molecular mass of
proteins and a global view of all detectable modification
forms. The specific modified forms can be isolated and
subsequently fragmented in the mass spectrometer by
tandem MS (MS/MS) such as collisionally activated dis-
sociation (CAD) and electron capture dissociation (ECD)
for highly reliable PTMs mapping (26–31). Furthermore,
this method allows relative quantitation of each protein
ion population to provide accurate measure of the percent
of protein being modified (32–35).
Using high-resolution top-down combined with middle-
down (limited digestion) MS, we precisely mapped a single
mouse CARM1 automethylation site to R551, which is
conserved among allvertebrate
Mutation of the automethylation site from arginine to
lysine does not alter the enzymatic activity of CARM1.
However, mutation of the automethylation site on
CARM1 impairs both CARM1-activated transcription
and pre-mRNA splicing, strongly implying that regulation
of transcription and splicing events are jointly regulated
by the PTM of CARM1.
MATERIALS AND METHODS
Protein expression and purification
FLAG-CARM1 was expressed from Sf9 cells as described
previously (10). Full-length mouse CARM1 was PCR
cloned into pFC14K Flexi vector (Promega, Madison,
WI, USA). The CARM1 R551K mutant was obtained
by site-directed mutagenesis of CARM1 in the pFC14K
vector (CARM1R551Kforward primer: 50gTCAATC
The CARM1 construct contains a C-terminal HaloTag.
The construct was expressed in transiently transfected
HEK293T cells, affinity purified using HaloLink resin
(Promega), and eluted by TEV protease. The purified
CARM1 does not contain Halo tag, which remains cova-
lently coupled to the resin. The exact details of
CARM1 using the HaloTag are described in a separate
Top-down MS: high-resolution FT-ICR MS analysis of
FLAG-CARM1 from Sf9 and CARM1 derived from
HEK293T was performed under similar conditions as
described previously (35). The intact mass spectra were
collected under broad ion-trap isolation of three charge
states. Fragmentation was performed by narrow isolation
of a single charge state and then dissociated by CAD using
15% collision energy.
endoprotease AspN for FT-ICR MS analysis. Fifty micro-
gram of CARM1 from HEK293T cells was incubated with
250ng of AspN for 20min at 37?C in 2.5mM ZnSO4 and
50mM Tris pH 7.5. The resulting reaction was desalted on
a 10kDa NMWL Amicon Ultra (Millipore). Nano-
electrospray ionization was performed in a 50% MeOH,
1% acetic acid spray solution on a Triversa Nanomate
injector (Advion). Fragmentation was performed by
narrow isolation of a single charge state and then
dissociated by CAD using 12–16% collision energy or
ECD using 3% activation energy for 50ms.
The forward primer for the CT/CGRP mini-gene was
labeled by incubation with polynucleotide kinase (PNK)
in PNK buffer (New England Biolabs) with g-labeled
32P-ATP at a final concentration of 1mM per primer.
Reactions were held at 37?C for 30min. PCR was per-
formed with Taq polymerase (Promega), 2ml of 1:100
diluted cDNA per 10ml reaction and a final primer con-
centration of 600nM. Twenty cycles of PCR were per-
formed, with 20s annealing step at 59?C and 30s
extension step at 72?C. Reactions were run on 5%
non-denaturing TBE-acrylamide gels, vacuum-dried and
stored at ?80?C prior to exposure to film.
2718Nucleic Acids Research, 2011,Vol.39, No. 7
In vitro methylation assays
Enzymes and substrates were incubated in 15ml of 5mM
MgCl2, 20mM Hepes pH 7.9, 1mM EDTA, 1mM DTT,
10% glycerol containing 1ml of3H–S-adenosylmethionine
(78Ci/mmol, GE Healthcare) for 5min–1h. The reaction
mix was then separated by SDS–PAGE and fixed in
methanol: acetic acid (50%:5%) containing Coomassie
Coomassie-free methanol: acetic acid for 1h. Fixed gels
were incubated in ‘Amplify’ (Amersham Biosciences) scin-
tillation fluid for 30min, stored at ?80?C overnight and
exposed to a film.
For mini-gene reporter assays. For mini-gene reporter
assays, HEK293T cells were seeded into 6-well culture
dishes with high glucose DMEM containing 10 % FBS.
Twenty-four hours later each well was transfected with
2mg ERE-CT/CGRP reporter, 5ng CMX-ERa and 1mg
with 8ml of TurboFect (Fermentas).
Media was replaced after 6h and cells were grown in
the presence of 10nM 17b-estradiol (E2) or DMSO
alone for 48h.
qRT-PCR of endogenous gene expression, ER18 cells
(HEK293 cells with stable expression of ERa) or
CARM1 null MEF cells were plated in 6-well culture
glucose-DMEM. Twenty-four hours later, ER18 cells
pSG5-CARM1WTor pSG5-CARM1R551Kwith 2.3ml of
TurboFect for 6h. CARM1-null MEF cells were trans-
fected with 2mg of pSG5-vector, pSG5-CARM1WTor
pSG5-CARM1R551Kand 6ml Fugene HD (Roche). At
24h, media was changed fresh. At 48h, transfected cells
were treated with 10nM E2 or DMSO for 3h before
qRT-PCRof endogenousgene expression. For
strippedFBS in high
1mg of pSG5-vector,
HEK293T cells by cotransfecting 1mg of pSG5-vector,
CMX-ERa (5ng), renilla-luciferase expression vector
UAS-Luc reporter (200ng). Cells were then treated with
DMSO or 10nM E2 for 48h prior to lysis and detection of
luciferase. Firefly luciferase from the UAS reporter was
normalized with internal control renilla luciferase.
assays. These wereperformed in
RNA isolation, reverse transcription and qPCR
RNA was isolated using RNeasy kit (Qiagen) following
the manufacturers’ instructions. DNA contamination was
digested using DNAfree (Ambion). RNA was quantified
and 2mg was added to Superscript II (Invitrogen) reverse
transcription reactions with oligo-d (T) as a primer.
Completed reactions were diluted 100-fold prior to use
as a template in qPCR and32P-labeled reactions. qPCR
(Invitrogen). Four microliters of 1:100 diluted cDNA
Sybr Green MasterMix
was added to each 20ml reaction and four replicates
were performed per condition. Ctlevels were compared
to a relative standard curve of similarly prepared cDNA
from the same cell type.
Mouse CARM1 is methylated at arginine 551 in vitro
and in vivo
To map and determine the relative quantity of mouse
CARM1 PTMs, we employed high-resolution top-down
MS. We first attempted to analyze PTMs on full-length
FLAG-CARM1 expressed in Sf9 insect cells using a
Linear Trap Quadrupole (LTQ) Fourier Transform-Ion
Cyclotron Resonance (FT-ICR) mass spectrometer. No
ion peak was found at the predicted unmodified mass of
FLAG-CARM1 (calculated mass 67 935.44, Figure 1A). In
contrast, the major peak was 42 Da larger than the un-
modified mass (Figure 1A). The experimentally determined
mass was 67977.30, which is very close to calculated mass
67977.48 if CARM1 is trimethylated. To map the location
of this additional mass, the 55+charge state of the 67977.3
Da mass peak was isolated in the LTQ and fragmented by
CAD. MS/MS of FLAG-CARM1 identified an additional
mass of 14 Da (consistent with monomethylation) on all
N-terminal fragments of FLAG-CARM1. The minimal
N-terminal sequence that is modified was primarily the
FLAG-tag and linker sequence, which encompasses
commonly methylated residues (lysines and arginines).
Further site-mapping was not pursued because this modi-
fication would not occur on endogenous CARM1. An add-
itional mass of 28 Da (consistent with dimethylation) was
observed on C-terminal fragments that included the
sequence from amino acids 537–582 (amino acids 555–
600 of the FLAG-CARM1 fusion protein). This region
only contains one likely dimethylation site, arginine (R)
551 (Figure 1B).
The presence of monomethylation on the FLAG-tag
potentially complicated the analysis of the role of methy-
lation in CARM1 function, thus we switched to a mam-
malian expression system to overexpress CARM1 fused to
a cleavable HaloTag?
Halo-CARM1 was expressed in HEK293T cells, purified
on Halo-link resin, and cleaved with TEV protease to
obtain full-length CARM1. Surprisingly, this approach
produced a very pure protein (Figure 1C) and a signifi-
cantly higher yield than FLAG-CARM1 purified from Sf9
cells. Top-down MS analysis of CARM1 from HEK293T
matched the calculated mass of full-length CARM1 with a
dimethyl modification (calculated mass is 67317.29 and
experimentally determined mass is 67317.47), while un-
modified CARM1 was undetectable (Figure 1D).
MS/MS of full-length CARM1 exhibited limited frag-
mentation data due to the poor stability of CARM1
during electrospray ionization.
middle-down MS approach
CARM1 with the endoproteinase AspN prior to analysis
on the LTQ/FT-ICRto
dimethylation. Multiple charge states (M7+–M10+) in the
FT-ICR fragmentation spectrum corresponding to an
determine the site(s)of
Nucleic Acids Research,2011, Vol.39, No. 72719
Asp-N proteolytic peptide (D521–E617) were detected to
be dimethylated, with no observable peak matching the
mass of the unmodified peptide (Supplementary Figure
S1). Isolation of the M9+charge state of the dimethylated
D521–E617 peptide and subsequent CAD or ECD frag-
mentation unequivocally localized R551 as the single
dimethylation site (Figure 1E).
We created an arginine (R) to lysine (K) mutation in
the HaloTag expression vector to preclude CARM1
purified CARM1R551Kfrom HEK293T cells. To monitor
CARM1R551Kwas incubated with a radiolabeled methyl
Comparison of the two proteins in the absence of add-
itional substrate indicated that neither protein was signifi-
cantly methylated in vitro; however, extended exposure
indicates in vitro methylation of co-eluted proteins
(Figure 2A). To determine whether this lack of in vitro
methylation was due to saturating levels of methylation
in vivo, hypomethylatedproteins
pre-treating Halo-CARM1 transfected HEK293T cells
with a methylation inhibitor, adenosine dialdehyde
(Adox). Adox treatment for 48h at 20mM in the culture
media significantly reduced basal level of endogenous
asymmetric, di-methylated proteins as detected by a
di-methyl H3R17 antibody (Supplementary Figure S2)
as this antibody was shown to recognize histone
H3R17me2 and other CARM1 substrates (37). When
CARM1 proteins from these Adox-treated cells were
compared in vitro, CARM1WTexhibited strong incorpor-
ation of the3H-methyl groups, while CARM1R551Kdid
not (Figure 2A). These biochemical data corroborate
our MS data and indicate that R551 is the primary methy-
lation site on CARM1 both in vivo and in vitro.
The strong invitro
implies an automethylation mechanism. However, the
presence of minor co-eluted proteins as indicated in the
in vitro methylation assay (Figure 2A) raised the possibil-
ity that one of these co-eluted proteins may methylate
CARM1. To rule out this possibility, we incubated
hypomethylated CARM1WTwith increasing amounts of
the CARM1 specific substrate poly(A) binding protein 1
(PABP1) (38). The presence of PABP1 effectively inhibits
methylation of CARM1WTin a dose-dependent manner,
and this inhibition is highly effective even at low concen-
trations of PABP1 (Figure 2B). This result strongly
automethylation, as other PRMTs do not methylate
expressed in HEK293T
of CARM1is in fact
Figure 1. High-resolution
dimethylation at R551 of the C-terminal domain (CTD) of CARM1.
(A) Mass of the 55+ precursor ion (intact protein) of an N-terminally
tagged FLAG-CARM1 from Sf9 cells. The isolated spectra represents a
single population of FLAG-CARM1 at isotopic resolution. LTQ/
FT-ICR MS was used to determine the intact mass of recombinant
CARM1 protein and map the site of methylation. The calculated
(Calc’d) mass represents the mass predicted based on amino acid com-
position and putative PTMs. The theoretical peak distribution is rep-
resented graphically by a series of open (methylated) circles. The
experimental (Expt’l) mass represents the mass determined from the
MS spectra. (B) CAD fragmentation of full-length FLAG-CARM1
maps dimethylation site to the sequence between amino acids 555
and 600. Green half circles represent a fragment that is ?14 Da
larger than predicted (indicating monomethylation), while a red circle
represents a fragment that is ?28 Da larger than the predicted fragment
top- andmiddle-down MS reveal
(indicating dimethylation). Sequence map shows the coverage and
placement of CAD fragments of FLAG-CARM1. (C) Coomassie
stainingof CARM1(with Halo
HEK293T cells. (D) CARM1 from HEK293T cells is dimethylated.
Mass of the 57+ precursor ion of full-length CARM1 purified from
HEK293T cells is shown, no peak is observed at the expected mass of
unmodified CARM1. (E) CAD and ECD fragmentation map of an
AspN proteolytic peptide (D520–E616) of CARM1. Blue marks repre-
sent ECD fragments while red marks represent CAD fragments as
indicated in the figure graphic at the bottom.
tag removed)purified from
2720Nucleic Acids Research, 2011,Vol.39, No. 7
PABP1 (38,39). Sequence comparison of the CTD of
CARM1 in other annotated species indicates a high
degree of conservation among vertebrate species near the
CARM1 automethylation site (Figure 2C).
The main cellular functions of CARM1 are currently
attributed to its methyltransferase activity. To determine
the effect of mutation of the automethylation site on
CARM10s enzymatic activity, we performed in vitro
methylation assays with the CARM1 specific substrate
PABP1. Both CARM1WTand CARM1R551Kwere able
to methylate PABP1 to a similar degree (Figure 3A and
B), and exhibited similar activity and specificity when
incubated with nuclear or cytoplasmic extracts from
MCF7 in a dose-dependent manner (Figure 3C and D).
Separation of CARM1 substrates in two-dimentional
SDS–PAGE also revealed no quantitative differences
profiles (Supplementary Figure S3). Thus, we conclude
that mutation of the automethylation site of CARM1
has no effect on its methyltransferase activity.
CARM1 is best characterized as a transcriptional
coactivator. Furthermore, the CTD of CARM1 is
required to maintain its full transcriptional activity (15).
To assess whether mutation of the automethylation site of
CARM1 affects its transcriptional activity, we examined
expression of several endogenous ERa target gene by
overexpressed in ER18 cells, a HEK293T cell line stably
expressing ERa (40), and treated with or without
17-b-estradiol (E2) for 3h. mRNA levels of CARM1WT
and CARM1R551Kwere present at similar levels in these
transfected cells (Supplementary Figure S4). Known
CARM1-responsive, ERa-target genes were tested for ex-
pression by qRT-PCR. pS2, IGFPB4 and pTGES ex-
overexpressed CARM1WTin a ligand-dependent manner
CARM1R551Ktransfected cells was identical to that of
cells transfected with vector control, suggesting that
mutation of the automethylation site impaired CARM10s
ability to activate transcription. The mRNA levels of
cathepsin D, an E2-responsive but CARM1-independent
gene, showed no enhancement of transcription by trans-
fecting either CARM1WTor CARM1R551K. A similar
pattern was observed for Ebag9 and Stc2 in CARM1?/?
MEF cells, using cathepsin D as a negative control
(Figure 4B). Ebag9and
CARM1?/?MEFs because those genes were shown to
be CARM1-dependent in MEF cells (23,37) while pS2,
IGFPB4 and pTGES were found not expressed in
MEFs. To test whether the lack of transcriptional
activity by the automethylation deficient mutant is
specific to ERa target genes, we determined transcription
of two other transcription factors known to be activated
by CARM1 in a Gal4 reporter assay. The plasmid
co-transfected with vector, CARM1WTor CARM1R551K
along with a UAS-reporter, into HEK293T cells. With
automethylation-deficient CARM1R551Kalso exhibited
compromised activity in
in the presenceof
Stc2were analyzed in
or Gal4-NFkBIB was
Figure 2. The CTD of CARM1 is automethylated in vivo and in vitro.
(A) CARM1WTand CARM1R551Kexpressed in the absence or presence
of a methylation inhibitor (Adenosine dialdehyde, Adox) prior to
in vitro methylation without any additional substrates. In vitro methy-
lation reactions were performed in the presence of
separated by SDS–PAGE, stained with coomassie and exposed to
film. Autoradiograph shown represents a long exposure to detect
low-level methyation. CS=coomassie staining, AR=autoradiograph.
(B) Poly (A) binding Protein 1 (PABP1) can efficiently compete with
Hypomethlated (Adox treated) CARM1WT(1mg) was incubated alone
or in the presence of increasing amounts of PABP1 (0–800ng). After
exposure to a film, bands were excised and quantitiated for3H incorp-
oration. (C) Partial CLUSTALW alignment of known and putative
CARM1 sequences indicates high conservation of the automethylation
site within vertebrates (purple).
Nucleic Acids Research,2011, Vol.39, No. 72721
Figure 3. CARM1 enzymatic activity and substrate specificity are not affected by mutation of automethylation site R551! K. (A) In vitro methy-
lation of PABP1 by CARM1WT(WT) or CARM1R551K(Mut) (n=2). AR represents autoradiograph, CS represents coomassie staining. (B) In vitro
methylation of PABP1 by CARM1WT(WT) or CARM1R551K(Mut) for 5min, followed by excision of the PABP1 bands and quantification by
scintillation counting (n=3). Error bars represent standard deviations from the mean. (C) In vitro methylation of 7.5mg nuclear lysates from
CARM1-silenced MCF7 cells by 0.25 and 1.25mg of recombinant CARM1WT, CARM1R551Kor GST-PRMT1. (D) In vitro methylation of
control buffer (Auto), 15mg nuclear lysates (Nuc) or 15mg cytoplasmic lysates (Cyto) from CARM1?/?MEF cells by 1mg CARM1WTor
Figure 4. Loss of automethylation impairs transcriptional activation activity of CARM1. (A and B) ER18 cells (A) are HEK293T cells stably
expressing ERa, and CARM1 null MEF cells (B) were cotransfected with pSG5-vector: pSG5-CARM1WTor pSG5-CARM1R551K. After 48h
transfection, cells were treated with DMSO or 10nM E2 for 3h, followed by RNA isolation and qRT-PCR analysis of endogenous gene
products. A mean (±SD, n=3) quantification is graphically displayed. (C) CARM1R551Kdisplayed low transcriptional activity as compared
with CARM1WTto activate Sertad and NFkBIB in Gal4-reporter assays. A mean (±SD, n=3) quantification is graphically displayed.
2722Nucleic Acids Research, 2011,Vol.39, No. 7
(Figure 4C), suggesting that automethylation-dependent
CARM1 activation of transcription is broadly significant.
Recently CARM1 was shown to methylate a number of
splicing factors and regulate alternative pre-mRNA
automethylation in its regulation of alternative splicing,
we performed a splicing assay using a mini-gene
reporter. HEK293T cells were transfected with a CT/
CGRP reporter gene under the control of an ERE
promoter (17) (illustrated in Figure 5A). Cells were add-
itionally transfected with an ERa expression vector and
CARM1R551K, CARM1WTor vector control prior to
treatment with E2 or DMSO. The mRNA levels of each
alternative exon were quantified by32P-labeled RT-PCR
(Figure 5B) and Sybr Green qRT-PCR using splicing
form-specific primers (Figure 5D–F). The ratio of inclu-
sion of the CGRP exon over inclusion of the CT exon
increased with E2 treatment, and this increase was
enhanced by the overexpression of CARM1WT, but not
CARM1R551K(Figure 5B and D). We confirmed that
the protein levels (Figure 5C) of CARM1WT
CARM1R551Kare similar. CARM1 appears to mediate
change in the mini-gene splicing ratio by increasing ex-
pression of the CGRP exon without affecting the expres-
sion of the CT exon (Figure 5E). This observation is
consistent with a model where CARM1 co-regulates tran-
scription and mRNA splicing in an interdependent
manner. To further examine this possibility, the total
mRNA expressed from the mini-gene reporter was
normalized to b-actin and compared between the three
transfection conditions. Interestingly, the total mRNA
of bothalternative transcripts
overexpression of CARM1WTbut not CARM1R551Kin
a ligand-dependent manner (Figure 5F). Thus, the
impaired activity in mRNA splicing as well as transcrip-
tional activation of the same transcriptional target.
automethylation site at the CTD of mouse CARM1,
high-resolution MS,we identifieda single
Figure 5. Loss of automethylation affects CARM1-mediated alterna-
tive splicing. (A) Cartoon representation of the ERE-CT/CGRP mini-
gene reporter with primers indicated; non-specific forward primer (open
triangle), CT exon-specific reverse primer (gray triangle) and CGRP
exon-specific reverse primer (black triangle). (B) Overexpression of
20-cycle PCR of cDNA from mini-gene reporter cDNAs, separated
by non-denaturing PAGE and exposed to a film. (C) CARM1WTor
CARM1R551Kwere expressed at a similar level as determined by
western blot. Whole cell lysates from HEK293T cells transfected with
control, CARM1WTor CARM1R551Kexpression vector were probed
witha rabbit polyclonalCARM1
exon-skipping of mini-gene splicing reporter (A) by CARM1WTbut
not CARM1R551K. Two splicing products was quantitated using
qRT-PCR and presented as CGRP/CT ratio. qRT-PCR was performed
on cDNA from mini-gene reporter assays. (E) The expression level of
each alternative exon from the splicing reporter was normalized to
b-actin. (F) CARM1WTincreases total mRNA of CT/CGRP exons as
compared with CARM1R551K. The total mRNA levels of CT and
CGRP exons from the mini-gene reporter were normalized to endogen-
ous b-actin levels to determine total mRNA expression from the
32P-labeled forward primer was used in
Nucleic Acids Research,2011, Vol.39, No. 72723
when it is expressed in either insect or mammalian expres-
sion systems. Mutation of the automethylation site of
CARM1 impairs both its regulation of transcription and
alternative mRNA splicing, highlighting an interesting
paradigm that transcription and splicing are co-regulated
by PTM of CARM1. Our study also enhances our under-
standing of the poorly defined CARM1 CTD function in
transcriptional regulation and presents a self-regulatory
mechanism to signal from CARM1’s catalytic domain to
Mapping CARM1 PTMs using top- and middle-down MS
Here our high-resolution top-down MS/MS data unam-
biguously revealed a 28 Da increase in the mass of
C-terminal fragments containing the R551 site, which
indicated that CARM1 is modified near its C-terminus.
electrospray ionization compromised the MS/MS frag-
mentation process and produced a limited number of frag-
mentation ions. Therefore, we used a combination of top-
and middle-down (limited digestion) MS to map the site of
dimethylation to a single amino acid at R551. All detect-
able CARM1 was found already automethylated. The
absence of unmethylated CARM1 explains previous ob-
servations of extremely weak automethylation of purified
CARM1 in vitro (8,41) because CARM1 automethylation
is nearly complete in vivo (Figures 1A and D, and 2A).
Among the dimethylated CARM1 population from 293T
cells, ?50% was also found to have an O-linked
acetylglucosamine (O-GlcNAc) modification (data not
shown). We are currently mapping the site of the O-
GlcNAc modification and are investigating its function.
As expected, we did not observe phosphorylation of
CARM1 at the characterized locations because phosphor-
ylation on these three sites only occurs during mitosis
(22,23) or under specific stimuli (24). The cells we used
to purify CARM1 were unsynchronized, thus the abun-
dance of phosphorylated CARM1 would have been very
low. This low abundance would make detection of a
modified intact protein ion difficult by FT-ICR MS.
of CARM1 during
One-way, self-regulatory mechanism for two CARM1
CARM1 has two functional domains that are essential for
its full transcriptional activity (15). The enzymatic activity
located in the central catalytic domain is required for most
CARM1 mediated-transcription (5,8,42) and for main-
taining the in vivo function of CARM1 (14). CARM1
differs from other PRMTs by carrying an extended
CTD. Multiple lines of evidence suggest that CARM10s
CTD is important for CARM1’s function (15,43,44).
Here we present evidence that CARM1 undergoes
automethylation in the CTD, and that mutation of this
automethylation site inhibits CARM1’s ability to regulate
pre-mRNA splicing and transcription. Not only have our
results highlighted the importance of CARM1’s CTD but
also revealed a one-way self-regulatory scheme from the
CARM1 catalytic domain
automethylation by mutation of R551 to lysine on
CARM1 does not appear to affect substrate specificity
to theCTD. Lossof
or the enzymatic activity of CARM1 in vitro (Figure 3
and Supplementary Figure S3). This result is consistent
with the previous observation that the CTD of CARM1
is not required for substrate methylation (15). In contrast,
the central catalytic domain of CARM1 can methylate
other protein substrates as well as itself, at the CTD,
auto-regulation. Since methylation of the CTD is also de-
pendent on CARM1’s enzymatic activity, the functions of
CARM1’s CTD may have been masked in previous
studies when CARM1’s enzymatic activity is eliminated
in vivo and in vitro.
CARM1 automethylation provides a hub for transcription
and splicing co-regulation
Significant attention has been focused on the role of the
methyltransferase activity in CARM1-regulated cellular
functions since its discovery (3). While non-enzymatic
interactions have been observed (10), the majority of
CARM1 interacting proteins are also substrates (45,46),
and CARM1’s ability to methylate histone H3 is con-
sidered to be an integral aspect of its coactivation mech-
anism (6). Our results show that the mutation of the
automethylation site to lysine affects neither the localiza-
tion of CARM1 (Supplementary Figure S5) nor the sub-
strate specificity or enzymatic activity in vitro (Figure 3
and Supplementary Figure S3). However, CARM1
automethylation in cells appears to play an important
role in regulating multiple functions of CARM1.
Truncation of CARM1 into various domains identified
the CTD as an independently important region in tran-
scription (15); however, the mechanism for functional sig-
nificance of CARM1 CTD
unexplored. The only protein that was reported to
interact with CTD is TIF1a, where TIF1a’s characterized
role in coactivation is limited to the stabilization of inter-
action between CARM1 and GRIP1 (47). We surmise that
the CARM1 CTD likely facilitates protein–protein inter-
actions that are necessary for transcription, and the
dimethylated arginine residue on the CTD might serve
as a platform to recruit effector proteins. Our observation
that CARM1 lacking the automethylation site has reduced
ability to activate transcription of a GAL4 reporter
(Figure 4C) as well as endogenous ERa-target genes
(Figure 4B) support this notion. In fact, automethylation
of a histone lysine methyltransferase G9a was found to be
necessary and sufficient for mediating in vivo interaction
with heterochromatin protein 1 (HP1) (48). PRMT8, a
CARM1 family member, was shown to be automethylated
at two sites at its N-terminus. However, the effect of the
automethylation reactions on the activity of PRMT8 was
not determined (41). Thus, CARM1 appears to be the first
PRMT whose automethylation is demonstrated to play an
important regulatory role.
Evidence of CARM1’s role in regulating alternative
splicing is emerging but the mechanism is still poorly
defined. The current view for CARM1’s regulation of
pre-mRNAsplicing is through
of splicing factors including CA150 (21). Methylation of
CA150 by CARM1 is required for its interaction with
2724 Nucleic Acids Research, 2011,Vol.39, No. 7
SMN, a primary regulator of spliceosome assembly. Our
findings that an automethylation-defective mutant has
impaired activityin regulating
(Figure 5) renders an alternative model that CARM1 is
directly involved in splicing regulation through its CTD.
Two possible models may explain how CARM1’s CTD
might be engaged in the regulation of alternative
splicing. First, CARM1 may depend on its association
pre-mRNA splicing, a commonly accepted theme of
transcription-coupled mRNA splicing (49). Our observa-
tion that the CARM1-dependent alternative splicing of
the CT/CGRP mini-gene occurs when CARM1WTalso
specifically activates transcription (Figure 5D and E) is
consistent with a model where CARM1 co-regulates tran-
scriptionand splicing. Chromatin
recognized as a scaffold for pre-mRNA splicing and the
chromatin remodeling machineries have been found to
be involved in splicing decisions (50). CARM1 was previ-
ously shown to associate with the SWI/SNF complex
(10), a chromatin remodeling complex involved in regula-
tion of alternative splicing
automethylation may regulate the direct interaction
between CARM1 and the splicing factors, thus CARM1
may be an integral part of the spliceosome. Mechanistic
studies are ongoing and may provide insights into the
mechanism by which automethylation regulates CARM1
(51). Second, CARM1
Supplementary Data are available at NAR Online.
We thank Mark Bedford for providing the CARM1
MEF?/?cell line, Elaine Alarid for the ER18 cell line,
Jiandie Lin for Gal4-Sertad and Gal4-NFkBIB plasmids
and Bert O’Malley for the ERE-CT/CGRP plasmid. We
would like to especially acknowledge UW-Madison
Human Proteomics Program for support in acquiring
mass spectrometry data. We would like to thank
Promega Corporation for providing HaloTag reagents in-
strumental for publication of CARM1. We would like to
thank Emery Bresnick, David Wassarman and Chih-hao
Lee for critical reading of the manuscript. P.K., R.C.,
Y.G., R.B. and W.X. designed experiments. P.K., R.C.
and Y.W. performed experiments. P.K., R.C. and W.X.
analyzed data. P.K. Y.G. and W.X. wrote the manuscript.
RO3MH089442 to W.X., T32 CA009135 to P.K.);
Wisconsin Partnership Fund for a Healthy Future,
UW-Madison Human Proteomics Program; Promega
Funding for open access charge: National Institutes of
Institutesof Health (RO1CA125387,
Conflict of interest statement. R.B. is required by the
University of Wisconsin-Madison Conflict of Interest
Committee to disclose that he has a financial interest in
Promega Corp, which markets the HaloTag technology
used to purify CARM1.
1. Bedford,M.T. and Richard,S. (2005) Arginine methylation an
emerging regulator of protein function. Mol. Cell, 18, 263–272.
2. Kuhn,P. and Xu,W. (2009) Chapter 9 protein arginine
methyltransferases nuclear receptor coregulators and beyond.
Prog. Mol. Biol. Transl. Sci., 87, 299–342.
3. Chen,D., Ma,H., Hong,H., Koh,S.S., Huang,S.M., Schurter,B.T.,
Aswad,D.W. and Stallcup,M.R. (1999) Regulation of
transcription by a protein methyltransferase. Science, 284,
4. Hassa,P.O., Covic,M., Bedford,M.T. and Hottiger,M.O. (2008)
Protein arginine methyltransferase 1 coactivates
NF-kappaB-dependent gene expression synergistically with
CARM1 and PARP1. J. Mol. Biol., 377, 668–678.
5. An,W., Kim,J. and Roeder,R.G. (2004) Ordered cooperative
functions of PRMT1, p300, and CARM1 in transcriptional
activation by p53. Cell, 117, 735–748.
6. Ma,H., Baumann,C.T., Li,H., Strahl,B.D., Rice,R., Jelinek,M.A.,
Aswad,D.W., Allis,C.D., Hager,G.L. and Stallcup,M.R. (2001)
Hormone-dependent, CARM1-directed, arginine-specific
methylation of histone H3 on a steroid-regulated promoter.
Curr. Biol., 11, 1981–1985.
7. Schurter,B.T., Koh,S.S., Chen,D., Bunick,G.J., Harp,J.M.,
Hanson,B.L., Henschen-Edman,A., Mackay,D.R., Stallcup,M.R.
and Aswad,D.W. (2001) Methylation of histone H3 by
coactivator-associated arginine methyltransferase 1. Biochemistry,
8. Xu,W., Chen,H., Du,K., Asahara,H., Tini,M., Emerson,B.M.,
Montminy,M. and Evans,R.M. (2001) A transcriptional switch
mediated by cofactor methylation. Science, 294, 2507–2511.
9. Feng,Q., Yi,P., Wong,J. and O’Malley,B.W. (2006) Signaling
within a coactivator complex: methylation of SRC-3/AIB1 is a
molecular switch for complex disassembly. Mol. Cell Biol., 26,
10. Xu,W., Cho,H., Kadam,S., Banayo,E.M., Anderson,S.,
Yates,J.R. III, Emerson,B.M. and Evans,R.M. (2004) A
methylation-mediator complex in hormone signaling. Genes Dev.,
11. Troffer-Charlier,N., Cura,V., Hassenboehler,P., Moras,D. and
Cavarelli,J. (2007) Functional insights from structures of
coactivator-associated arginine methyltransferase 1 domains.
EMBO J., 26, 4391–4401.
12. Yue,W.W., Hassler,M., Roe,S.M., Thompson-Vale,V. and
Pearl,L.H. (2007) Insights into histone code syntax from
structural and biochemical studies of CARM1 methyltransferase.
EMBO J., 26, 4402–4412.
13. Jayne,S., Rothgiesser,K.M. and Hottiger,M.O. (2009) CARM1
but not its enzymatic activity is required for transcriptional
coactivation of NF-kappaB-dependent gene expression.
J. Mol. Biol., 394, 485–495.
14. Kim,D., Lee,J., Cheng,D., Li,J., Carter,C., Richie,E. and
Bedford,M.T. (2010) Enzymatic activity is required for the in vivo
functions of CARM1. J. Biol. Chem., 285, 1147–1152.
15. Teyssier,C., Chen,D. and Stallcup,M.R. (2002) Requirement for
multiple domains of the protein arginine methyltransferase
CARM1 in its transcriptional coactivator function. J. Biol.
Chem., 277, 46066–46072.
16. Auboeuf,D., Dowhan,D.H., Dutertre,M., Martin,N., Berget,S.M.
and O’Malley,B.W. (2005) A subset of nuclear receptor
coregulators act as coupling proteins during synthesis and
maturation of RNA transcripts. Mol. Cell Biol., 25, 5307–5316.
17. Auboeuf,D., Honig,A., Berget,S.M. and O’Malley,B.W. (2002)
Coordinate regulation of transcription and splicing by steroid
receptor coregulators. Science, 298, 416–419.
Nucleic Acids Research,2011, Vol.39, No. 72725
18. Monsalve,M., Wu,Z., Adelmant,G., Puigserver,P., Fan,M. and Download full-text
Spiegelman,B.M. (2000) Direct coupling of transcription and
mRNA processing through the thermogenic coactivator PGC-1.
Mol. Cell, 6, 307–316.
19. Fujiwara,T., Mori,Y., Chu,D.L., Koyama,Y., Miyata,S.,
Tanaka,H., Yachi,K., Kubo,T., Yoshikawa,H. and Tohyama,M.
(2006) CARM1 regulates proliferation of PC12 cells by
methylating HuD. Mol. Cell Biol., 26, 2273–2285.
20. Li,H., Park,S., Kilburn,B., Jelinek,M.A., Henschen-Edman,A.,
Aswad,D.W., Stallcup,M.R. and Laird-Offringa,I.A. (2002)
Lipopolysaccharide-induced methylation of HuR, an
mRNA-stabilizing protein, by CARM1. Coactivator-associated
arginine methyltransferase. J. Biol. Chem., 277, 44623–44630.
21. Cheng,D., Cote,J., Shaaban,S. and Bedford,M.T. (2007) The
arginine methyltransferase CARM1 regulates the coupling of
transcription and mRNA processing. Mol. Cell, 25, 71–83.
22. Feng,Q., He,B., Jung,S.Y., Song,Y., Qin,J., Tsai,S.Y., Tsai,M.J.
and O’Malley,B.W. (2009) Biochemical control of CARM1
enzymatic activity by phosphorylation. J. Biol. Chem., 284,
23. Higashimoto,K., Kuhn,P., Desai,D., Cheng,X. and Xu,W. (2007)
Phosphorylation-mediated inactivation of coactivator-associated
arginine methyltransferase 1. Proc. Natl Acad. Sci. USA, 104,
24. Carascossa,S., Dudek,P., Cenni,B., Briand,P.A. and Picard,D.
(2010) CARM1 mediates the ligand-independent and
tamoxifen-resistant activation of the estrogen receptor alpha by
cAMP. Genes Dev., 24, 708–719.
25. Cheung,W.D., Sakabe,K., Housley,M.P., Dias,W.B. and
Hart,G.W. (2008) O-linked beta-N-acetylglucosaminyltransferase
substrate specificity is regulated by myosin phosphatase targeting
and other interacting proteins. J. Biol. Chem., 283, 33935–33941.
26. Siuti,N. and Kelleher,N.L. (2007) Decoding protein modifications
using top-down mass spectrometry. Nat. Methods, 4, 817–821.
27. Sze,S.K., Ge,Y., Oh,H. and McLafferty,F.W. (2002) Top-down
mass spectrometry of a 29-kDa protein for characterization
of any posttranslational modification to within one residue.
Proc. Natl Acad. Sci. USA, 99, 1774–1779.
28. Kuhn,P., Xu,Q.G., Cline,E., Zhang,D., Ge,Y. and Xu,W. (2009)
Delineating Anopheles gambiae coactivator associated arginine
methyltransferase 1 automethylation using top-down high
resolution tandem mass spectrometry. Protein Sci., 18, 1272–1280.
29. Ge,Y., Rybakova,I.N., Xu,Q.G. and Moss,R.L. (2009) Top-down
high-resolution mass spectrometry of cardiac myosin binding
protein C revealed that truncation alters protein phosphorylation
state. Proc. Natl Acad. Sci. USA, 106, 12658–12663.
30. Ge,Y., Lawhorn,B.G., ElNaggar,M., Strauss,E., Park,J.H.,
Begley,T.P. and McLafferty,F.W. (2002) Top down
characterization of larger proteins (45kDa) by electron capture
dissociation mass spectrometry. J. Am. Chem. Soc., 124, 672–678.
31. Ayaz-Guner,S., Zhang,J., Li,L., Walker,J.W. and Ge,Y. (2009)
In vivo phosphorylation site mapping in mouse cardiac troponin
I by high resolution top-down electron capture dissociation mass
spectrometry: Ser22/23 are the only sites basally phosphorylated.
Biochemistry, 48, 8161–8170.
32. Pesavento,J.J., Mizzen,C.A. and Kelleher,N.L. (2006) Quantitative
analysis of modified proteins and their positional isomers by
tandem mass spectrometry: human histone H4. Anal. Chem., 78,
33. Zabrouskov,V., Han,X., Welker,E., Zhai,H., Lin,C., van
Wijk,K.J., Scheraga,H.A. and McLafferty,F.W. (2006) Stepwise
deamidation of ribonuclease A at five sites determined by top
down mass spectrometry. Biochemistry, 45, 987–992.
34. Zabrouskov,V., Ge,Y., Schwartz,J. and Walker,J.W. (2008)
Unraveling molecular complexity of phosphorylated human
cardiac troponin I by top down electron capture dissociation/
electron transfer dissociation mass spectrometry. Mol. Cell
Proteomics, 7, 1838–1849.
35. Kuhn,P., Xu,Q., Cline,E., Zhang,D., Ge,Y. and Xu,W. (2009)
Delineating Anopheles gambiae coactivator associated arginine
methyltransferase 1 automethylation using top-down high
resolution tandem mass spectrometry. Protein Sci., 18, 1272–1280.
36. Chumanov,R. et al. (2011) Expression and purification of full-
length mouse CARM1 from transiently transfected 60 HEK293
cells using HaloTag technology. Protein Expression and
Purification, in press.
37. Yadav,N., Lee,J., Kim,J., Shen,J., Hu,M.C., Aldaz,C.M. and
Bedford,M.T. (2003) Specific protein methylation defects and gene
expression perturbations in coactivator-associated arginine
methyltransferase 1-deficient mice. Proc. Natl Acad. Sci. USA,
38. Lee,J. and Bedford,M.T. (2002) PABP1 identified as an arginine
methyltransferase substrate using high-density protein arrays.
EMBO Rep., 3, 268–273.
39. Swiercz,R., Person,M.D. and Bedford,M.T. (2005) Ribosomal
protein S2 is a substrate for mammalian PRMT3 (protein
arginine methyltransferase 3). Biochem. J., 386, 85–91.
40. Valley,C.C., Metivier,R., Solodin,N.M., Fowler,A.M.,
Mashek,M.T., Hill,L. and Alarid,E.T. (2005) Differential
regulation of estrogen-inducible proteolysis and transcription by
the estrogen receptor alpha N terminus. Mol. Cell Biol., 25,
41. Sayegh,J., Webb,K., Cheng,D., Bedford,M.T. and Clarke,S.G.
(2007) Regulation of protein arginine methyltransferase 8
(PRMT8) activity by its N-terminal domain. J. Biol. Chem., 282,
42. Lee,D.Y., Northrop,J.P., Kuo,M.H. and Stallcup,M.R. (2006)
Histone H3 lysine 9 methyltransferase G9a is a transcriptional
coactivator for nuclear receptors. J. Biol. Chem., 281, 8476–8485.
43. Ohkura,N., Takahashi,M., Yaguchi,H., Nagamura,Y. and
Tsukada,T. (2005) Coactivator-associated arginine
methyltransferase 1, CARM1, affects pre-mRNA splicing in an
isoform-specific manner. J. Biol. Chem., 280, 28927–28935.
44. Matsuda,H., Paul,B.D., Choi,C.Y. and Shi,Y.B. (2007)
Contrasting effects of two alternative splicing forms of
coactivator-associated arginine methyltransferase 1 on thyroid
hormone receptor-mediated transcription in Xenopus laevis.
Mol. Endocrinol., 21, 1082–1094.
45. Chevillard-Briet,M., Trouche,D. and Vandel,L. (2002) Control of
CBP co-activating activity by arginine methylation. EMBO J., 21,
46. Lee,Y.H., Coonrod,S.A., Kraus,W.L., Jelinek,M.A. and
Stallcup,M.R. (2005) Regulation of coactivator complex assembly
and function by protein arginine methylation and
demethylimination. Proc. Natl Acad. Sci. USA, 102, 3611–3616.
47. Teyssier,C., Ou,C.Y., Khetchoumian,K., Losson,R. and
Stallcup,M.R. (2006) Transcriptional intermediary factor 1alpha
mediates physical interaction and functional synergy between the
coactivator-associated arginine methyltransferase 1 and
glucocorticoid receptor-interacting protein 1 nuclear receptor
coactivators. Mol. Endocrinol., 20, 1276–1286.
48. Sampath,S.C., Marazzi,I., Yap,K.L., Krutchinsky,A.N.,
Mecklenbrauker,I., Viale,A., Rudensky,E., Zhou,M.M., Chait,B.T.
and Tarakhovsky,A. (2007) Methylation of a histone mimic
within the histone methyltransferase G9a regulates protein
complex assembly. Mol. Cell, 27, 596–608.
49. Chen,M. and Manley,J.L. (2009) Mechanisms of alternative
splicing regulation: insights from molecular and genomics
approaches. Nat. Rev. Mol. Cell Biol., 10, 741–754.
50. Allemand,E., Batsche,E. and Muchardt,C. (2008) Splicing,
transcription, and chromatin: a menage a trois.
Curr. Opin. Genet. Dev., 18, 145–151.
51. Batsche,E., Yaniv,M. and Muchardt,C. (2006) The human
SWI/SNF subunit Brm is a regulator of alternative splicing.
Nat. Struct. Mol. Biol., 13, 22–29.
2726Nucleic Acids Research, 2011,Vol.39, No. 7