Fox-3 and PSF interact to activate neural
cell-specific alternative splicing
Kee K. Kim1, Yong C. Kim2, Robert S. Adelstein1and Sachiyo Kawamoto1,*
1Laboratory of Molecular Cardiology, National Heart, Lung, and Blood Institute and2Cellular Immunology
Section, Laboratory of Immunology, National Institute of Allergy and Infectious Disease,
National Institutes of Health, Bethesda, MD 20892, USA
Received September 9, 2010; Revised November 8, 2010; Accepted November 10, 2010
Fox-1 family (Fox) proteins, which consist of Fox-1
mammals, bind to the RNA element UGCAUG and
regulate alternative pre-mRNA splicing. However
the mechanisms for Fox-regulated splicing are
largely unknown. We analyzed the expression
pattern of the three Fox proteins as well as neural
cell-specific alternative splicing of a cassette exon
N30 of nonmuscle myosin heavy chain (NMHC) II-B
in the mouse central nervous system. Histological
and biochemical analyses following fluorescence-
activated cell sorting demonstrate a positive correl-
ation of N30 inclusion and Fox-3 expression. Further,
we identified polypyrimidine tract binding protein-
associated splicing factor (PSF) as an interact-
ing protein with Fox-3 by affinity-chromatography.
In cultured cells, enhancement of N30 inclusion by
Fox-3 depends on the presence of PSF. PSF
enhances N30 inclusion in a UGCAUG-dependent
manner, although it does not bind directly to this
element. Fox-3 is recruited to the UGCAUG element
downstream of N30 in the endogenous NMHC II-B
transcript in a PSF-dependent manner. This study is
the first to identify PSF as a coactivator of Fox
proteins and provides evidence that the Fox-3 and
PSF interaction is an integral part of the mechanism
by which Fox proteins regulate activation of alterna-
tive exons via a downstream intronic enhancer.
Alternative splicing of pre-mRNA is an important mech-
anism for post-transcriptional regulation of gene expres-
sion and has increasingly been appreciated as a major
mechanism to generate diversity of gene products in
higher eukaryotes. Developmentally regulated, cell type-
or tissue-specific and signal-induced alternative splicing of
throughout their lifetimes. Misregulation or abnormalities
in pre-mRNA splicing can lead to a number of cellular
dysfunctions found in human and animal diseases (1,2).
Using various model systems of regulated alternative
splicing, exonic and intronic enhancers as well as silencers
have been defined in pre-mRNAs. RNA-binding proteins
which can be recruited to these RNA elements have also
technologies have been developed for genome-wide
analysis of alternative splicing. Genome-wide splice
array and computational analysis of whole genome se-
quences have defined a number of potential cis-elements
for splicing regulation (4–6). Combinations of splice
arrays, systemic identification of RNA targets for the
sequencing have been used to make a genome-wide
splice map for Nova, Fox-2 and polypyrimidine tract
binding protein (PTB), which relates the position of the
target element of the RNA-binding protein to the splicing
patterns (7–9). These studies have helped to predict
splicing patterns of given genes. However our understand-
ing of the molecular mechanism by which RNA-binding
proteins regulate the splicing process is limited to SR
proteins and some of the hnRNP proteins. Exonic enhan-
cers and their binding proteins, SR proteins, have been
well studied. For example, the interactions of the SR
proteins with snRNP components and other basic pre-
demonstrated during progression of the splicing steps
(10,11). Studies on intronic and exonic silencers and
their binding proteins such as PTB have provided
several models for splicing repression (12,13). However,
how RNA-binding proteins, which are recruited to
intronic enhancers and often do not belong to SR
proteins and hnRNP proteins, activate splicing of alterna-
tive exons is largely unknown. Only a few studies have
*To whom correspondence should be addressed. Tel: +1 301 435 8034; Fax: +1 301 402 1542; Email: firstname.lastname@example.org
Nucleic Acids Research, 2011, Vol. 39, No. 8Published online 21 December 2010
Published by Oxford University Press 2010.
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.
addressed the question of connecting the RNA enhancer
binding proteins and splicing machinery (14,15).
One of the intronic enhancer elements which is involved
in cell type or tissue-specific regulation of alternative
splicing is the UGCAUG element. The importance of
this element has been demonstrated in a number of cases
of alternative splicing specific to neural cells, muscle cells
and other cell types (16–19). This UGCAUG element does
not necessarily function as an enhancer, but it enhances
inclusion of alternative exons when it resides in an intron
downstream of regulated exons (8). Jin et al. discovered
that a zebrafish homolog of Caenorhabditis elegans Fox-1
could bind to this element in a highly sequence-specific
manner (20). Subsequently a human homolog of Fox-1
has been shown to duplicate this property (21) and the
solution structure of the RNA-binding domain of
human Fox-1 in a complex with UGCAUGU has been
determined (22). Identification of Fox-1 as an UGCAU
G-binding protein has had a major impact on the field of
alternative splicing. Several laboratories have demons-
trated that mammalian homologs of Fox-1 do indeed
regulate alternative splicing via the UGCAUG element
using model systems (20,21,23–27). Further, the unusually
high sequence specificity of the Fox-1 target sequence has
attracted investigators in bioinformatics to conduct
genome-wide analysis for the location of the UGCAUG
element and the splicing patterns of alternative exons
In mammals, three genes are recognized as homologs of
C. elegans Fox-1. Two of them were originally cloned in
contexts other than splicing and the third gene was
recognized after whole genome sequences had been
determined (29–31). Different names were used for one
of these genes in the literature. In this report, we use the
nomenclature of Fox-1, Fox-2 and Fox-3. Fox-1 is also
called A2BP1 and Hrnbp1 and Fox-2 is also called Rbm9,
Fxh, Hrnbp2 and RTA. Fox-3 is called Hrnbp3,
D11Bwg0517e and NeuN. All three Fox-1 family (Fox)
proteins contain a single almost identical RNA recogni-
tion motif (RRM) in the central region of the molecule.
Fox-1 is selectively expressed in brain and striated
muscles, whereas Fox-2 is expressed in various tissues
including brain and muscles (20,29,30). Both the Fox-1
and 2 genes generate tissue-dependent as well as tissue-
independent isoforms by alternative splicing and alterna-
subcellular distributions and have quantitatively different
effects on splicing (23,26,32). Fox-3 is expressed exclusive-
ly in neural tissues (31,33). Recently we have identified
NeuN as the Fox-3 gene product (33). NeuN was origin-
ally defined by its immuno-reactivity with the monoclonal
anti-NeuN antibody (34) and has been considered as
a post-mitotic neuron-specific nuclear marker for over
We have been studying the regulatory mechanisms
of neural cell-specific alternative splicing using human
nonmuscle myosin heavy chain II-B (NMHC II-B,
named MYH10 in the human genome) as a model
system. NMHC II-B mRNA is expressed ubiquitously.
However, an alternative exon, N30, which encodes a
30nt coding sequence, is included in the mRNAs from
some neural cells, but skipped in those from all other
cells in mammals and birds (35,36). We have previously
defined an intronic distal downstream enhancer (IDDE)
which is located 1.5kb downstream of N30. The IDDE
confers neural cell-specificity on N30 inclusion and
contains two copies of the UGCAUG element (17,37).
We have shown that exogenously expressed brain
isoforms of Fox-1 and 2 lead to N30 inclusion in
the mRNAs derived from minigenes as well as from the
endogenous NMHC II-B gene in cultured cells. The
muscle-specific isoforms of Fox-1 and 2 demonstrate less
activity than the brain isoforms on N30 splicing (23).
Exogenously expressed Fox-3 isoforms also enhance N30
In this study, we analyze the expression pattern of the
endogenous Fox proteins and the N30 splicing of
endogenous NMHC II-B at the cellular level in mouse
brain and spinal cord and we address a potential correl-
ation of Fox-3 expression and N30 inclusion. Further, we
identify Fox-3 interacting proteins as a step toward under-
standing the mechanism by which Fox-3 promotes inclu-
sion of the alternative exon via binding to a downstream
MATERIALS AND METHODS
Generation of Fox-1, 2 and 3 antisera and other
The N-terminal amino acids 1–116 and 1–112 of Fox-1
and Fox-2, respectively, fused to GST were expressed in
BL21 bacteria and purified by a GSTrap FF column
(GE Healthcare). GST-fused Fox-1 and Fox-2 proteins
were used to generate polyclonal anti-Fox-1 and anti-
Fox-2 in rabbits (Biosynthesis, Inc). The rabbit polyclonal
anti-Fox-3 raised against the N-terminal amino acid 1–97
residues of Fox-3 was described previously (33). The other
primary antibodies (Abs) used in this study were mouse
monoclonal anti-NeuN (Millipore), mouse monoclonal
anti-myc (Invitrogen), mouse monoclonal anti-GAPDH
splicing factor (PSF; Sigma), rabbit polyclonal anti-PSF
(Santa Cruz Biotechnology), goat polyclonal anti-NonO
rabbit polyclonal anti-NMHCII-B (38).
Affinity chromatography and mass spectrometry
For preparation of anti-Fox-3 immobilized resin, affinity
purified polyclonal anti-Fox-3 was crosslinked to protein
A/G agarose (Santa Cruz Biotechnology) with di-
succinimidyl suberate (Sigma) according to an antibody
crosslinking method (39). Nuclear extracts containing
5mg of protein in 10ml Co-IP buffer (Pierce) were
incubated with the anti-Fox-3 crosslinked protein A/G
agarose overnight at 4?C. The immuno-complexes were
washed in Co-IP buffer. The complexes were solubilized
in SDS sample buffer and subjected to SDS-PAGE.
Following Coomassie blue staining, the protein bands
were digested in the gels with trypsin. The recovered
peptides were analyzed using an Applied Biosystems 4700
Nucleic Acids Research,2011, Vol.39, No. 83065
MALDI-TOF/TOF to acquire tandem MS/MS spectra. A
compiled protein database was searched for the peptide
Brain tissues were dissected and passed through a cell
strainer (pore size, 40mm) to prepare a single-cell suspen-
sion. The cells were washed in PBS containing bovine
serum albumin, an RNase inhibitor (Roche) and a
protease inhibitor cocktail (Sigma) and then permeabilized
in a permeabilization buffer (eBioscience). Nonspecific
binding sites were blocked with 5% goat serum for
30min at 4?C, and then the cells were labeled with
anti-NeuN at a final dilution of 1:100 for 45min at 4?C.
The cellswere further
conjugated goat Abs against mouse IgG (Molecular
Probes) at a 1:500 dilution for 15min at 4?C. The resulting
cells were analyzed and sorted using a MoFlo cell sorter
(Cytomation, Inc). After FACS, RNAs and proteins were
In situ hybridization
N30 included NMHC II-B mRNA was detected using a
biotinylated oligonucleotide probe (Bioneer Inc.) comple-
mentary to the N30 region, 50-CTGGGGTTTCACGGG
CTTAGGCGATTCCTG-3’. Hybridization and detection
were carried out using an IsHyb In Situ Hybridization kit
(Biochain Institute Inc.) and a DNADetector System
(KPL Inc.) according to the manufacturers’ protocols.
The specimens were examined using a Zeiss Axiophot
GST pull-down assay
GST–Fox-3 fusion proteins were expressed in KRX
bacteria cells and purified on glutathione sepharose
beads (Amersham Pharmacia). The myc-tagged PSF and
NonO were synthesized in vitro from the pCS3+MT con-
structs using a TNT Coupled Reticulocyte Lysate System
(Promega). ForGST pull-down
synthesized myc-tagged protein and GST-fusion protein
bound to glutathione beads were incubated in 5–10ml of
PP-300 buffer (20mM Tris–HCl pH 8.0, 300mM NaCl,
0.5mM EDTA, 0.2% NP-40, 0.5mM DTT, protease in-
hibitors) for 1h at 4?C. The protein complexes were re-
covered in the SDS sample buffer. Following SDS-PAGE,
the gels were stained with Coomassie blue or subjected to
RNA–protein complex immunoprecipitation
The Fox-3-RNA complexes were immunoprecipitated
with anti-NeuN using a Magna RIP kit (Millipore) ac-
cording to the manufacturer’s instructions. The RNAs re-
covered from the complexes were subjected to RT-PCR
for the IDDE of the NMHC II-B transcripts. The primers
(forward) and 50-GAAAAATTGAGAACGAGTATTCA
RNA–protein UV crosslink
The PCR products containing the T7 promoter upstream
of the IDDEs were used as templates for RNA transcrip-
tion (23). The RNA probes were synthesized by T7 RNA
polymerase in the presence of [a-32P]UTP using a
MAXIscript kit (Ambion). The myc-tagged Fox-3-L and
PSF proteins were synthesized in vitro from pCS3+MT
constructs using the TNT Coupled Reticulocyte Lysate
System. Binding reactions were carried out in a 25ml
mixture that contains 10mM HEPES (pH 7.9), 2mM
MgCl2, 1mM ATP, 20mM creatine phosphate, 50ng
yeast tRNA, 2mM DTT, 2% polyethylene glycol (M.W.
3,550), 4ml of the reticulocyte lysate reaction mixture and
1?106cpm RNA probe for 20min at 30?C. Reaction
mixtures were irradiated with UV light (Stratalinker,
RNase-A/T1 (Ambion) and immunoprecipitated with
2mg of anti-myc. Samples were subject to SDS-PAGE
followed by autoradiography. The radioactive bands
were quantified using a phosphorimager STORM 860
ice, digested with2ml
Distribution of Fox-1, 2 and 3 in the mouse central
Fox-1, 2 and 3 are expressed in a tissue-dependent manner
and neural tissues are the only tissues where all three Fox
mRNAs are expressed (Supplementary Figure S1A). To
analyze the distribution of the three Fox proteins at a
cellular level in the central nervous system, the brain and
spinal cord of postnatal day 10 mice were stained with
rabbit polyclonal Abs. These Abs were newly produced
in this laboratory against the N-terminal one-third of
each Fox protein. The specificities of the Abs were
verified by reactivity to recombinant Fox proteins ex-
pressed in cultured cells (Supplementary Figure S1C)
(Supplementary Figure S1B). In tissue extracts, 1–3
bands migrating at a molecular mass range of 30–
55kDa are detected using each Ab. Each Ab detects the
proteins in only the tissues where the mRNAs are ex-
pressed, indicating the specificity. We have recently
identified Fox-3 as NeuN, the antigen recognized by the
widely used monoclonal Ab, anti-NeuN (33). Taking ad-
vantage of the fact that anti-NeuN is a mouse monoclonal
Ab, tissue sections were stained with one of the polyclonal
Fox Abs and anti-NeuN. Consistent with the fact that
anti-NeuN staining overlap exactly in every single cell
(Figure 1 E, J, O, and T). Co-staining with anti-Fox-1
and anti-NeuN and co-staining with anti-Fox-2 and
anti-NeuN demonstrate overlapping staining as well as
unique staining for each Ab (Figure 1).
In the hippocampal dentate gyrus, Fox-3 as well as
NeuN expression is limited to the granular cell layer
(GCL, Figure 1E), whereas Fox-2 expression is higher in
the subgranular zone (SGZ) than the GCL (Figure 1C).
Fox-1 is uniformly expressed in both the GCL and SGZ
3066 Nucleic Acids Research, 2011,Vol.39, No. 8
(Figure 1A). These differential expressions of Fox-1, 2 and
3 are further demonstrated by overlay images with NeuN
staining (Figure 1B, D and E). In the cerebellum, Fox-3/
NeuN staining is limited to the internal granular layer
(IGL, Figure 1J). Anti-Fox-1 and anti-Fox-2 also
weakly stain the IGL cells and both strongly stain
Purkinje cell nuclei as well as the nuclei of interneurons
in the molecular layer (ML) and Golgi cells (Figure 1F–I).
In addition, Fox-2 is also detected in the inner layer of
cells in the external germinal layer (EGL, Figure 1I,
bracket) and the migrating granular cells (Figure 1H and
I). This unique staining pattern of anti-Fox-2 suggests that
Fox-2 is expressed in early post-mitotic neurons in
addition to mature neurons. In the brain stem region,
there is a mixture of Fox-3 only expressing cells, Fox-3
and Fox-1 or 2 expressing cells and Fox-1 or 2 expressing
cells (Figure 1K–O). However, the expression pattern of
Fox-1 is difficult to compare with that of Fox-2 in this
region. In the spinal cord, motor neurons in the ventral
horn express all three Fox proteins and other neurons in
this area express them at a much lower level (Figure 1P–
T). In summary, all three Fox proteins are expressed
widely in the central nervous system, but there are differ-
ences in the relative levels of expression of each Fox
protein. Fox-3 expression is absent in some specific
neurons. All Fox proteins are localized predominantly to
neuronal cell nuclei. In large neurons such as motor
neurons in the spinal cord and brain stem, however,
Fox-1 and Fox-3 can also be detected in soma.
Fox-3 expression and neural cell-specific alternative
splicing of N30 in NMHC II-B
Next, we asked whether the differences in expression
levels of Fox proteins among different neuronal cells
have any relation to the splicing patterns of some
pre-mRNAs. Because Fox-3, unlike Fox-1 and 2, is ex-
pressed only in neural tissues and Fox-3 expression seems
to be restricted to certain types of neurons, brain cells
Figure 1. Differential distribution of Fox proteins in the mouse central nervous system. Sagittal sections of the brain and transverse sections of the
spinal cord from postnatal day 10 mice were co-stained with polyclonal Abs for the three Fox proteins (red) as indicated, and the monoclonal
anti-NeuN (green). DAPI was used to stain nuclei (blue). Arrows and arrowheads in G and I point to Purkinje and Golgi cells, respectively. The
bracket in I indicates the inner layer of the external germinal layer. Bars, 50mm (E, J and O) or 200mm (T); DH, dorsal horn; VH, ventral horn; C,
canal; WM, white matter.
Nucleic Acids Research,2011, Vol.39, No. 83067
were dissociated and sorted into Fox-3 positive and
negative cells by FACS using the monoclonal anti-Fox-
3 (anti-NeuN). As shown in Figure 2A, 77.4% of
dissociated cells from the cerebellum are Fox-3 positive.
The majority of these cells shows low side scattering
(SSC, granularity), consistent with being relatively small
internal granular cells. Figure 2B shows that 42.5% of
the cells dissociated from the brain stem and spinal cord
are Fox-3 positive and the mean fluorescence intensity of
these cells is 2- to 3-fold higher than those of cerebellar
cells. Successful sorting by Fox-3 expression was verified
by immunoblot analysis as shown in Figure 2C. This
immunoblot confirms not just the presence or absence
of Fox-3 but also shows that the average expression
level of Fox-3 in the brain stem and spinal cord is
higher than in the cerebellum. Of note is that the
average expression levels of Fox-1 and 2 are similar
between the Fox-3 positive and negative cell groups in
the cerebellum, brain stem and spinal cord.
We have been studying neural cell-specific alternative
splicing of pre-mRNA using NMHC II-B as a model. A
cassette exon, N30, can be included only in the mRNA
from some types of neural cells. We previously showed
that the IDDE containing two UGCAUG elements is
essential for N30 inclusion (17). Therefore, to determine
whether N30 inclusion correlates with Fox-3 expression,
the extent of N30 inclusion in the NMHC II-B mRNA
was compared between Fox-3 positive and negative cell
populations. Strikingly, Fox-3 positive cells from the
brain stem and spinal cord include N30 to a much
greater extent than Fox-3 negative cells (Figure 2D,
lanes 3 and 4). Fox-3 positive cells from the cerebellum,
which express Fox-3 at a lower level than those from the
brain stem and spinal cord, also include N30 to a small
extent (Figure 2D, lane 2). In contrast, Fox-3 negative
cells almost completely skip N30 (Figure 2D, lanes 1
and 3), despite the fact that Fox-1 and 2 in Fox-3
negative cells are expressed at similar levels to Fox-3
positive cells (Figure 2C).
We also analyzed the relation between Fox-3 expression
and N30 inclusion using tissue sections. NMHC II-B is
expressed ubiquitously except in a few cell types such as
hematopoietic cells. In neural tissues, Purkinje cells in
the cerebellum and motor neurons in the spinal cord
express especially high levels of NMHC II-B, as shown
in Figure 3A (red, arrowheads in a and arrows in b).
(Figure 3A-b, green, arrows) but not in Purkinje cells
(Figure 3A-a, arrowheads). Instead, Purkinje cell nuclei
contain Fox-1 and 2 (Figure 1F–I). In situ hybridization
analysis using an antisense oligonucleotide probe comple-
mentary to the N30 sequence demonstrates that N30
expression is robust in motor neurons (Figure 3B-b,
arrows) but is not detected in Purkinje cells. Sense oligo-
nucleotide probe does not give any signal (data not
shown). These biochemical and histological analyses
in motor neurons
Figure 2. Inclusion of N30 in NMHC II-B mRNAs preferentially occurs in Fox-3 expressing cells. (A, B) FACS analysis of the cells dissociated from
mouse cerebellum (A) and brain stem and spinal cord (B) using the monoclonal anti-Fox-3 (anti-NeuN). The numbers inside each panel represent
percentage of cells in each pool. The dotted line represents the mean fluorescence intensity. SSC, side scattering. (C) Immunoblot analysis of Fox-3
positive (+) and negative (?) cells sorted by FACS. Abs used in each blot are indicated on the right. GAPDH serves as a loading control. (D) N30
splicing patterns in Fox-3 positive (+) and negative (?) cells sorted by FACS. Ethidium bromide stained agarose gels of the RT-PCR products are
shown. The upper and lower bands include and exclude N30, respectively. Numbers shown between the agarose gels indicate percentage of the N30
3068Nucleic Acids Research, 2011,Vol.39, No. 8
support a positive correlation between N30 splicing and
We have previously shown that all three Fox proteins
are capable of activating N30 inclusion when they are
over-expressed in cultured cells (23,33). There are always
some concerns about the relevance of over-expression ex-
periments. Therefore, we asked whether the endogenous
level of Fox-3 expression is sufficient for N30 splicing.
Embryonic carcinoma-derived P19 cells express Fox-3
during neural differentiation triggered by retinoic acid
treatment, whereas undifferentiated P19 cells do not
express Fox-3 (33). Immunoblot analysis confirms that
Fox-3 expression is induced and also shows that the
Fox-2 expression level is unchanged before and after dif-
ferentiation (Figure 4A). Fox-1 is barely detected under
these culture conditions. These cells were transfected with
the minigene reporter construct which contains exon N30
and the flanking exons E5 and E6, and introns of the
human NMHC II-B gene (Figure 8A). N30 splicing
patterns were then analyzed. Upon retinoic acid treat-
ment, induction of Fox-3 is accompanied by an increase
in N30 inclusion from 12 to 56% (Figure 4C, lanes 1 and
2). Next T-2 and T-3 shRNAs, which target two different
regions of Fox-3 mRNAs, were used to inhibit Fox-3 ex-
pression (Figure 4B, lanes 2–4). GFP shRNA served as a
negative control. The shRNA (T-2 and T-3)-mediated re-
duction of Fox-3 results in a decrease in N30 inclusion
from 54 to 24% (Figure 4C, lanes 3–5). Rescue experi-
ments shown in Supplementary Figure S2B rule out
off-target effects of the T-2 shRNA. Since differentiated
P19 cells express Fox-2 in addition to Fox-3 (Figure 4A),
whether Fox-2 is required for N30 inclusion was tested.
Knock-down of Fox-2 expression by siRNA does not
affect N30 inclusion (Figure 4D). These results indicate
that the endogenous level of Fox-3 causes a significant
enhancement of N30 splicing, whereas the endogenous
level of Fox-2 is not sufficient and even not required for
N30 inclusion in neurally differentiated P19 cells.
Identification of PSF as an interacting protein with Fox-3
Fox-3-mediated regulation of alternative splicing, we
searched for nuclear factor(s) which interact with Fox-3.
We isolated protein complexes containing Fox-3 from
nuclear extracts of mouse brains by affinity chromatog-
raphy using affinity purified anti-Fox-3 covalently coupled
to a resin. Following SDS-PAGE, mass spectrometry
analysis of the protein bands revealed that the Fox-3
PTB-associated splicing factor (PSF) and non-POU
domain-containing octamer-binding protein (NonO) in
this study. Mass spectrometry data are shown in
Fox-3-containing complexes by co-immunoprecipitation
using nuclear extracts from differentiated P19 cells.
Consistent with the results from affinity chromatography,
co-immunoprecipitation with anti-Fox-3 followed by
immunoblots shows that Fox-3 can associate with PSF
and NonO (Figure 5A). Although PSF and NonO
contain RRMs (40), the interaction of these proteins
with Fox-3 is not mediated by RNA as demonstrated in
Figure 5A (lanes 3 and 4). RNase treatment does not
disrupt the interaction of Fox-3 with PSF and NonO.
On the other hand, the association of hnRNP-A3 with
Fox-3 is disrupted by RNase treatment, indicating that
the Fox-3 and hnRNP-A3 complex is mediated by
anti-PSF or anti-NonO followed by immunoblotting
further establishes that Fox-3, PSF and NonO form a
complex (Figure 5B and C).
To determine whether the interaction of Fox-3 with
PSF and NonO is direct or indirect and to determine
regions of the proteins responsible for the interaction,
in vitro binding assays using recombinant proteins were
carried out. It is known that the Fox-3 gene generates
multiple isoforms by alternative splicing (33,41). Two
isoforms Fox-3-L and Fox-3-S, which include and
exclude 47 amino acids in the C-terminal region, respect-
ively, were used for these experiments. The bacterially ex-
pressed and purified GST-fusion proteins containing
full-length and various domains of Fox-3 (Figure 5D)
were incubated with myc-tagged PSF or NonO, which
were generated by in vitro transcription–translation in
reticulocyte lysates. As shown in the upper panel of
Figure 5E, full-length Fox-3-L and S and the C-terminal
C-L and S regions interact directly with PSF, but the
N-terminal region and the RRM of Fox-3 do not. We
also generated various domains of PSF (Figure 5F). As
seen in Figure 5G, the N-terminal region including the P/
Q rich domain interacts with the Fox-3 C-terminal region.
Notably, NonO does not interact directly with any region
or with full-length Fox-3 (Figure 5E, middle panel). Since
understandthemechanism responsible for
Figure 3. Expression of Fox-3 and N30 in the mouse cerebellum and
spinal cord. (A) Immuno-staining for total NMHC II-B by polyclonal
anti-NMHC II-B detecting both the N30 included and excluded
isoforms (red) and for Fox-3 by anti-NeuN (green). Arrowheads in
(a) and arrows in (b) point to Purkinje cells and motor neurons, re-
spectively. (B) In situ hybridization for N30 included NMHC II-B
mRNA. Arrows in (b) point to Purkinje cells. Bars, 20mm; PL,
Purkinje cell layer.
Nucleic Acids Research,2011, Vol.39, No. 83069
a number of studies have previously reported the direct
interaction of PSF and NonO (42,43), it is likely that
NonO is recruited to the Fox-3 complex via PSF in
native tissue and cell extracts.
Binding activities of PSF to Fox proteins and the effects
of PSF on the Fox-3 and target RNA interaction
We next tested the binding activity of PSF to each Fox
member. Fox-1 and 2 have a number of tissue-dependent
and tissue-independent isoforms generated by alternative
splicing. We chose the isoforms which are expressed in
brain and contain similar C-terminal sequences to that
of Fox-3. In vitro translated PSF and each of the Fox
proteins, all of which were tagged with the myc epitope,
were incubated and immunoprecipitated with anti-PSF.
Following SDS-PAGE, the blots were immuno-stained
using anti-myc. Quantification of myc signals provides
the ratio of myc-Fox, which was co-immunoprecipitated
with PSF, to myc-PSF, which was immunoprecipitated by
anti-PSF regardless of its association with Fox proteins.
Under given conditions, PSF pulls down Fox-3-L and
Fox-3-S 2.7 fold more efficiently than Fox-2 and 7 fold
more efficiently than Fox-1 (Figure 6A). These results
suggest that PSF might have the highest affinity for
Fox-3 among the Fox proteins.
We also examined whether interaction of Fox-3 with
PSF has any effect on the binding activity of Fox-3 to
its target RNA sequence. Radiolabeled IDDE of the
NMHC II-B gene which contains two copies of the UG
CAUG element (wt-IDDE) or a mutant form of this
element (mc-IDDE, see Figure 8A) was mixed with
myc-tagged Fox-3 or myc-tagged PSF or with both
myc-tagged proteins, crosslinked by UV and subjected
to immunoprecipitation with anti-myc. Both Fox-3 and
PSF could be immunoprecipitated regardless of their
interaction and the radiolabeled RNA crosslinked to the
proteins was visualized (Figure 6B). As expected, Fox-3
binds to wild-type IDDE but not to mutant mc-IDDE.
No regionsof the IDDE
PSF, indicating that PSF does not bind directly to the
IDDE. Of note, however, the quantification of radio-
activity by a phosphorimager shows that Fox-3 binds
to the UGCAUG element in the IDDE more efficiently
(2- to 3-fold) in the presence of PSF, suggesting that
the Fox-3-PSF complex has a higher affinity for the UG
CAUG element compared to Fox-3 alone. Since our
crosslinking conditions do not crosslink Fox-3 and PSF,
a protein complex with the expected larger molecular
mass with radiolabeled RNA is not detected. Similar
experiments were carried out using pre-mRNA which
was composed of exons E5, N30 and E6 and the truncated
Essentially the same results were obtained (data not
are crosslinked with
Figure 4. Endogenous Fox-3 expression is associated with N30 inclusion in P19 cells. (A) Induction of Fox-3 expression in P19 cells during neural
differentiation. The cells were untreated (?) or treated with retinoic acid (RA) for the indicated days. The cell extracts were analyzed by immunoblots
using the indicated Abs. (B) Specific knock-down of Fox-3 expression by the Fox-3 targeting shRNAs. Clonal P19 cell lines expressing the indicated
shRNAs were untreated (?) or treated with RA. The cell extracts were analyzed by immunoblots using the indicated Abs. T-2 and T-3 shRNAs
target Fox-3 mRNA. GFP shRNA is used as a negative control. (C) Endogenous expression of Fox-3 is required for N30 inclusion. The P19 cell
lines expressing the indicated shRNAs were untreated (?) or treated (+) with RA and were transfected with the wild-type NMHC II-B minigene (see
Figure 8A). The N30 splicing patterns of the minigene mRNAs were analyzed by RT-PCR. The upper and lower bands include (+) and exclude (?)
N30, respectively. (D) Fox-2 is not required for N30 inclusion. RA-treated P19 cells were transfected with the minigene and Fox-2 siRNA (+) or
non-targeting control siRNA (?) and further cultured under differentiation conditions. The N30 splicing patterns of the minigene mRNAs were
analyzed by RT-PCR (upper panel). Decreased Fox-2 expression by siRNA was verified by immunoblots (lower panel).
3070 Nucleic Acids Research, 2011,Vol.39, No. 8
PSF is an essential coactivator of Fox-3 for N30 splicing
To understand the functional consequences of the inter-
action of Fox-3 with PSF and NonO, we studied the
effects of exogenous expression of these proteins on N30
splicing of the minigene in SK-N-SH cells. As shown in
Figure 7A, exogenous expression of Fox-3-L or S
enhances N30 inclusion (lanes 2 and 3). Interestingly,
exogenous expression of PSF alone also enhances N30
inclusion whereas NonO does not (Figure 7A, lanes 4
and 5). Expression of both Fox-3 and PSF shows an ap-
parently additive effect on N30 splicing (Figure 7A, lanes
6 and 8, see ‘Discussion’ section). Fox-3 together with
Figure 5. Direct interaction of Fox-3 with PSF. (A–C) Fox-3 associates with PSF and NonO. The nuclear extracts of the neurally differentiated P19
cells were immunoprecipitated with nonspecific immunoglobulin (IgG), anti-Fox-3 (a-Fox-3, A), anti-PSF (a-PSF, B) and anti-NonO (a-NonO, C).
Immunoprecipitates were treated with RNase (+) or untreated (?) and subjected to immunoblot analysis using the Abs indicated to the right (A).
The RNase treatment was omitted in the experiments shown in B and C. 10% of the nuclear extracts used for immunoprecipitation was loaded in
lane 1 (Input). IP, immunoprecipitation. (D) Schematic representation of full-length and deletion constructs of Fox-3 fused to GST. Each construct
includes the region indicated by the boxes. Numbers represent amino acids. (E) Fox-3 interacts directly with PSF via its C-terminal region but not
with NonO. The indicated GST-fused Fox-3 proteins which bound to the glutathione beads were incubated with myc-tagged PSF or NonO. The
proteins associated with the various GST–Fox-3 proteins were analyzed by immunoblots using anti-myc (upper and middle panels). The Input lane
includes 10% of the amount used for the pull down assay. The GST–Fox-3 proteins bound to the glutathione beads were verified by SDS–PAGE
followed by Coomassie blue staining (Cooma., lower panel). (F) Schematic representation of full-length and deletion constructs of PSF tagged with
the myc epitope. P/Q, proline and glutamine rich domain;+/?, basic and acidic amino acid rich domain. (G) PSF directly interacts with Fox-3 via its
N-terminal region. The indicated myc-tagged PSF proteins associated with GST–Fox-3-C-S were pulled down by glutathione beads and subjected to
immunoblot analysis using anti-myc (upper right panel). Myc-tagged PSF proteins (5%) used for the pull down assay is shown in the left panel
(Input). GST–Fox-3-C-S bound to glutathione beads was verified by Coomassie blue staining (Cooma., lower panel).
Nucleic Acids Research,2011, Vol.39, No. 8 3071
NonO or PSF together with NonO shows essentially the
same effect on N30 inclusion as that of Fox-3 alone or
PSF alone (Figure 7A, compare lanes 7, 9 and 10 to lanes
2 and 4). When all three proteins, Fox-3, PSF and NonO,
are co-expressed, the increase in N30 inclusion is similar to
that found in the absence of NonO (Figure 7A, compare
lanes 11 and 12 to lanes 6 and 8). These results suggest
that the interaction of Fox-3 with PSF (and NonO) results
in the cooperative enhancement of N30 inclusion. This
notion is further supported by experiments shown in
Figure 7B and C. The cooperative enhancement of N30
splicing by PSF (and NonO) with Fox-3 can be observed
when the C-terminal region of Fox-3 is intact. The
N-terminal-RRM of Fox-3 does not show the enhance-
ment effect. These observations are consistent with the
fact that PSF interacts with the C-terminal region of
As shown in Figure 7A, lane 4, exogenous expression of
PSF can activate N30 splicing in the absence of Fox-3. We
asked whether PSF-dependent N30 inclusion depends on
the downstream UGCAUG element. Mutation of either
one or both UGCAUG elements results in a significant
decrease in PSF-dependent N30 inclusion (Figure 8B,
lanes 2–4). The host SK-N-SH cells do not express
endogenous Fox-3 or Fox-1 at a detectable level, but
they endogenously express Fox-2. As already shown in
Figure 6A, PSF can interact with Fox-2 though less effi-
ciently than with Fox-3. Therefore we investigated
whether endogenous Fox-2 participates in PSF-induced
N30 inclusion. Knock-down of Fox-2 expression by
siRNA abolishes PSF-induced activation of N30 splicing
as well as the basal level of N30 splicing (Figure 8C).
Conversely, we also asked whether endogenous PSF
participates in Fox-3-induced activation of N30 splicing.
Knock-down of the endogenous PSF by siRNA also
abolishes Fox-3-induced N30 splicing as well as basal
splicing almost completely (Figure 8D). We also tested
whether or not PSF is required for N30 splicing which is
induced by endogenous Fox-3 expression in P19 cells
during neural differentiation. Knock-down of PSF expres-
sion by siRNA abolishes N30 inclusion in differentiated
P19 cells (Figure 8E). Therefore, both endogenous Fox-3
and PSF are required for N30 splicing in these cells.
Furthermore immunostaining of the mouse spinal cord
shows co-localization of Fox-3 and PSF, especially in
motor neurons in the ventral horn where N30 is robustly
expressed (Supplementary Figure S4). This is consistent
with the notion that N30 splicing requires both Fox-3
and PSF. Together all these results indicate that PSF is
an essential coactivator of Fox-3 for N30 splicing. PSF
and Fox-3/Fox-2 directly interact and cooperatively
activate N30 splicing. Activation of N30 splicing by one
of the two factors (PSF and Fox-3/Fox-2) is dependent on
the presence of the other factor.
Human NMHC II-B pre-mRNA is ?156kb in length
and contains 41 constitutive and 3 alternative exon se-
quences. The IDDE region is located 1.5kb downstream
of N30 in the native pre-mRNA. The endogenous
pre-mRNA is much more complex than pre-mRNAs
from the minigenes. Therefore we examined whether the
above observation obtained using minigenes is compar-
able with N30 splicing of the endogenous NMHC II-B
pre-mRNA in human cells. SK-N-SH cells were transient-
ly transfected with different combinations of expression
constructs for Fox-3, PSF and NonO and the splicing
patterns of N30 were analyzed. As shown in Figure 9A
Figure 6. (A) Preferential interaction of PSF with Fox-3 among Fox proteins. The indicated myc-tagged Fox proteins were incubated with
myc-tagged PSF and co-immunoprecipitated with anti-PSF. Immuno-complexes were analyzed by immunoblots using anti-myc (upper panel).
Numbers between the immunoblots indicate the ratio of myc-Fox vs. myc-PSF. Myc-Fox protein (5%) used for immunoprecipitation was also
analyzed by immunoblots (Input, lower panel). (B) PSF enhances Fox-3 binding to the target RNA. The radiolabeled wild-type (wt)- or mutant (mc)-
IDDE (see Figure 8A) RNA probes were incubated with myc–Fox-3-L, myc–PSF or myc–Fox-3-L plus myc–PSF as indicated, crosslinked with UV
and immunoprecipitated with anti-myc. Immunoprecipitates were subjected to SDS–PAGE followed by autoradiography (upper panel) or immuno-
blotting using anti-myc (lower panel).
3072 Nucleic Acids Research, 2011,Vol.39, No. 8
(lane 4), inclusion of N30 in the endogenous mRNA is
markedly increased by simultaneous exogenous expression
of Fox-3, PSF and NonO. We also analyzed the effects of
PSF on the recruitment of Fox-3 to the IDDE region in
the endogenous pre-mRNA. Since SK-N-SH cells express
endogenous PSF, the cellular level of PSF was either
decreased by siRNA or increased by exogenous expression
of PSF. RNA–protein co-immunoprecipitation using the
monoclonal anti-Fox-3 (anti-NeuN) and cellular extracts
followed by RT-PCR for the IDDE region demonstrates
that the recruitment of exogenously expressed Fox-3
to the IDDE depends on the cellular concentration of
PSF (Figure 9B). The lowest panel of Figure 9B shows
quantification of RT-PCR by real-time PCR. siRNA-
mediated reduction of endogenous PSF reduces the
amounts of the IDDE associated with Fox-3 about
4-fold. Exogenous expression of PSF leads to a 30-fold
increase in Fox-3 binding to the IDDE compared to the
PSF-knockdown. Levels of Fox-3 expression and changes
in the levels of PSF among a set of transfections were
Figure 7. Cooperative activation of N30 splicing by Fox-3 and its interacting proteins PSF and NonO. (A) Effects of Fox-3, PSF and NonO on N30
splicing. The NMHC II-B minigene was co-transfected into SK-N-SH cells with the expression constructs using the different combinations indicated.
The N30 splicing patterns of the minigene mRNAs were analyzed by RT-PCR (upper panel). The exogenously expressed proteins were verified by
immunoblots using anti-myc (lower panel). (B) Schematic representation of full-length and deletion constructs of Fox-3 tagged with the myc epitope.
(C) The C-terminal region of Fox-3 confers cooperativity with PSF and NonO on activation of N30 splicing. The experiment was performed similar
Nucleic Acids Research,2011, Vol.39, No. 83073
verified by immunoblots shown in Figure 9C. Taken
together, the increased Fox-3 binding to the IDDE by
PSF is accompanied by increased N30 splicing of the
endogenous pre-mRNA. Interaction of Fox-3 and PSF
is an integral part of the mechanism responsible for the
Fox-3-dependent activation of N30 splicing.
In this study, we identified PSF as a Fox-3 interacting
protein and demonstrated that this interaction is essential
for Fox-3 to activate neural cell-specific alternative
splicing of the N30 exon of NMHC II-B. We also
compared the expression pattern of Fox-3 to those of
Fox-1 and 2 at the cellular level in the mouse central
nervous system and found a correlation between Fox-3
expression and N30 splicing.
Two laboratories have reported the tissue distribution
and subcellular localization of Fox-1 in the mammalian
nervous system (24,29). Although they agreed that Fox-1
is expressed in neuronal cells, there is a discrepancy as to
whether it localizes to the nucleus or cytoplasm. Our ob-
servations largely agree with that of Black and colleagues
that Fox-1 localizes predominantly to nuclei. However, we
also detected Fox-1 in the somas and nuclei of motor
neurons in the spinal cord. Discrepancies in histological
observations might be due in part to differences in the age
Figure 8. Dependence of N30 splicing on Fox proteins and PSF. (A) Schematic diagrams of the NMHC II-B gene, minigene constructs and IDDEs.
The native gene shows a part of the human NMHC II-B gene surrounding the alternative exon N30. E5 and E6 are constitutive exons. The IDDE is
an intronic region consisting of 201nt. Exon size and the IDDE are not drawn to scale. Minigenes include the NMHC II-B genomic DNA fragments
indicated by brackets, which are flanked by exons E2 and E3 of the rat preproinsulin gene (PPI). The wt and mutant (ma, mb and mc) IDDEs are
inserted as indicated in the minigenes. Transcription of the minigene is driven by the Rous sarcoma virus long terminal repeat (RSVLTR). Arrows
above E5 and PPIE3 indicate the location of the primers used for RT-PCR. (B) UGCAUG element-dependent enhancement of N30 inclusion by
PSF. The minigenes containing the indicated IDDE or without the IDDE (?) were co-transfected with the myc–PSF expression construct into
SK-N-SH cells. The N30 splicing patterns of the minigene mRNAs were analyzed by RT-PCR (upper panel). The myc–PSF expression was verified
by immunoblot using anti-myc (lower panel). (C) PSF-induced enhancement of N30 inclusion requires endogenous Fox-2. The minigene containing
the wild-type IDDE was co-transfected into SK-N-SH cells with different combinations of the myc-tagged protein expression construct and siRNA as
indicated. The N30 splicing patterns of the minigene RNAs were analyzed by RT-PCR. Decreased endogenous protein expression by siRNA and the
exogenous expression of myc-protein were verified by immunoblots using the indicated Abs. (D) Fox-3-induced enhancement of N30 inclusion
requires endogenous PSF. The experiment was performed similar to C. (E) PSF is required for N30 inclusion in neurally differentiated P19 cells.
RA-treated P19 cells were transfected with the wild-type minigene and PSF siRNA (+) or non-targeting control siRNA (?), and further cultured
under differentiation conditions. The N30 splicing patterns of the minigene mRNAs were analyzed by RT-PCR (upper panel). Decreased PSF
expression by siRNA was verified by immunoblots (lower panel).
3074Nucleic Acids Research, 2011,Vol.39, No. 8
and species of the animals, differences in methods for
tissue fixation and antigen retrieval and differences in
the epitopes recognized by the Abs. Of interest, it has
been reported that the depolarization-induced change in
the Fox-1 isoformresults
nucleo-cytoplasmic ratio of Fox-1 (32). We have also pre-
viously reported that different isoforms of Fox-1 show
different subcellular localizations when they are exogen-
ously expressed in cultured cells (23). Therefore, it may be
necessary to determine which isoform of Fox-1 is
dominant in each particular system.
Our co-immunostaining analysis and FACS followed by
immunoblotting demonstrated that some neuronal cells
expressing Fox-1 or 2 do not express Fox-3. Of note is
that the expression levels of Fox-3, but not Fox-1 or 2,
correlate with the N30 splicing patterns in the cerebellum,
brain stem and spinal cord. Most of the previous studies
on brain specific alternative splicing have used whole
brain or anatomically dissectible parts of the brain. We
made use of FACS to separate cells according to the
expression levels of Fox-3. A combination of cell sorting
and biochemical analysis led us to find different splicing
patterns between the Fox-3 positive and negative cell
populations. In spite of our previous observation that
exogenous expression of Fox-1 or 2 enhances N30 inclu-
sion in cultured cells (23), brain cells which express Fox-1
and/or Fox-2 but not Fox-3 do not activate N30 splicing.
This could be explained in part by the following. Since
the average expression level of Fox-1 or 2 is similar
between Fox-3 positive and negative cells in the cerebel-
lum, brain stem and spinal cord, additional expression of
Fox-3 simply increases the total amounts of all three Fox
proteins. Another reason might be related to the isoforms
of Fox-1 and 2. Fox-1 and 2 genes generate multiple
ina change in the
alternatively spliced isoforms. Differences in amino acid
sequences at the C-terminal region result in significant
differences in the splicing activity toward N30 (23).
Currently our study does not distinguish differences in
the C-terminalamino acids,
generated against the N-terminal region common to
Fox-1 isoforms and the same is true for anti-Fox-2.
Therefore, cells negative for Fox-3 might express Fox-1
and/or Fox-2 isoforms with lower splicing activities. A
third reason might be a difference in affinity for interact-
ing protein(s). A number of proteins including ataxin-1
and 2, atrophin-1, quaking, Fyn tyrosine kinase and
estrogen receptor-a have been reported to interact with
Fox-1 or Fox-2 in mammals (29,44–46). However, most
of these proteins have not been functionally characterized
in the context of pre-mRNA splicing. Of particular
interest is a report showing the interaction of an U1
snRNP component, U1C protein, with Fox-1 and 2 in
yeast two-hybrid screening (47), although the functional
outcome of this interaction has not been studied. In this
study, we identified PSF as an essential interacting protein
of Fox-3 for activation of N30 splicing. Even though we
compared the isoforms of Fox-1 and 2 which show the
highest sequence similarity to Fox-3 and the highest
activity for N30 splicing among the known isoforms,
PSF binds to Fox-3 more efficiently compared to Fox-1
and 2. Therefore this could explain why Fox-3 is more
active in N30 inclusion. This affinity difference also
suggests that Fox-3 might play a role in the determination
of neural specificity of alternative splicing. However
differences in affinity of PSF for Fox proteins still need
to be studied in greater detail biochemically and evaluated
more precisely in the cellular context.
Figure 9. Fox-3 and PSF cooperatively enhance N30 inclusion in endogenous NMHC II-B mRNAs by recruitment of Fox-3 to the IDDE.
(A) Enhancement of N30 inclusion in endogenous NMHC II-B mRNA by Fox-3 and its interacting proteins PSF and NonO. SK-N-SH cells
were transiently transfected with the expression constructs for myc-tagged Fox-3-L, PSF and NonO with the different combinations indicated. The
N30 splicing patterns of the endogenous NMHC II-B mRNAs were analyzed by RT-PCR. The exogenously expressed proteins were verified by
immunoblot using anti-myc. (B, C) PSF-dependent recruitment of Fox-3 to the IDDE of the endogenous NMHC II-B transcript. SK-N-SH cells
were transfected with the expression constructs and siRNA in different combinations as indicated. An empty vector and a non-targeting siRNA were
used as negative controls (?). The cell extracts were subject to RNA-protein co-immunoprecipitation using the monoclonal anti-Fox-3 (anti-NeuN).
The IDDE region of the NMHC II-B transcripts associated with Fox-3 was analyzed by RT-PCR. The end products of RT-PCR are shown
following agarose gel electrophoresis (B, middle panel, lanes 6–9). Quantification by real-time PCR is shown in the lowest panel. The presence
of equal amounts of the IDDE transcript among a set of the transfected cells was verified using cell extracts without immunoprecipitation
(B, lanes 1–4). Immunoblot analysis of the input samples verifies changes in amounts of the expressed proteins (C). ND, not detectable;
Endo-PSF, endogenous PSF.
Nucleic Acids Research,2011, Vol.39, No. 83075
Although we found a good correlation between the level
of Fox-3 expression and the extent of N30 splicing, this
study does not exclude a possible contribution of Fox-1
and 2 to N30 splicing. Fox-1 and 2 have been reported to
interact with each other using a yeast two-hybrid system
(46). Our lab also detected the interaction of Fox-2 with
Fox-3 using the yeast two-hybrid system as well as the
interaction of Fox-3 with Fox-1 and with Fox-2 by
co-immunoprecipitation (unpublished observations). In
fact, we rarely observed neuronal cells which express
only Fox-3. It was unexpected to find that Fox-1 localized
predominantly to nuclei in intact brain, since all Fox-1
isoforms which we have analyzed were diffusely distributed
in both the nuclei and cytoplasm or predominantly
localized to the cytoplasm when they were exogenously
expressed in cultured cells (23). In contrast, exogenously
expressed Fox-3 isoforms localize almost exclusively
to nuclei in cultured cells (Supplementary Figure S1D).
The Fox-2 isoforms which we have analyzed localize
predominantly to nuclei (23). These observations raise the
possibility that Fox-1 might dimerize with Fox-3 or Fox-2
and localize to nuclei in vivo. The Fox proteins might
function as heterodimers or heteromultimers, especially
when the pre-mRNA contains multiple UGCAUG
elements. Whether there is a difference in splicing activity
among heterodimers and homodimers still needs to be
Fox proteins can function as activators or repressors
depending on their binding location on pre-mRNAs
relative to the regulated exons. Recent genome-wide
studies together with earlier studies using model systems
have proposed a general rule for Fox proteins to influence
the choice of exons (8,20). When Fox proteins bind to the
intron downstream of the alternative exon, exon inclusion
occurs. On the other hand, when Fox proteins bind to the
intron upstream of the alternative exon, exon skipping
occurs. Recently a few reports have begun to address the
mechanism by which Fox proteins repress or activate the
usage of alternative exons. Using F1g pre-mRNA, Fox-1
which is recruited to the upstream intron has been shown
to block formation of the early pre-spliceosome complex
on the intron downstream of the regulated exon (48). In
the case of calcitonin/CGRP pre-mRNA, Fox-2 which is
bound to the upstream intron inhibits the recruitment
of SF1 at the branch site, and further, Fox-2 which is
bound to the alternative exon inhibits the recruitment
of U2AF at the 30splice site (27). The interaction of
sequence-specific RNA-binding protein SUP-12, which is
expressed specifically in muscle, has been reported in
C. elegans. The FOX-1 (or ASD-1) and SUP-12 inter-
action enhances their binding to their adjacent target
RNA elements on the egl-15 pre-mRNA and leads to in-
clusion of a muscle specific mutually exclusive exon. This
report provides a mechanism for the strict tissue specificity
of Fox-regulated alternative splicing (49).
In this study we identified PSF and the PSF–NonO
complex as proteins interacting with Fox-3. The human
ortholog of NonO is often called p54nrb. PSF and NonO
are structurally related proteins that contain two RRMs
and can bind to RNA as well as DNA (40). PSF and
NonO interact with each other to form a heterodimer
and the participation of PSF or the PSF–NonO complex
has been reported in many aspects of nuclear function
including transcription, transcriptional
pre-mRNA splicing, 30end processing of mRNA and
RNA retention (40,50,51). PSF and NonO are expressed
in various tissues and cell types. We demonstrated that the
N-terminal region of PSF, and Fox-3 and NonO interact
indirectly via PSF. PSF enhances the binding of Fox-3 to
the target UGCAUG element in an in vitro crosslinking
assay. Moreover the presence of PSF enhances the recruit-
ment of Fox-3 to the IDDE of the NMHC II-B transcript,
which contains the UGCAUG elements, in intact cells.
The effect of PSF on Fox-3 binding to the target RNA
element in intact cells shows an apparently greater degree
of enhancement (>30-fold) than in vitro (2- to 3-fold). This
difference could be due to the different efficiencies of the
Fox-3 and PSF complex formation between in vitro and
intact cells. Another possibility is that PSF and Fox-3
might be co-transcriptionally recruited to the Fox-3
target sites of pre-mRNAs. PSF and NonO have been
reported to be associated with phosphorylated RNA poly-
merase II in a large complex containing transcriptional
elongation factors (52,53). If Fox-3 was included in that
complex via PSF, Fox-3 could be more efficiently re-
cruited to target sites of pre-mRNAs during transcription-
al elongation, compared to simple diffusion in the nucleus.
Although we showed that Fox-3 binding to the target
RNA element was enhanced by PSF, the role of PSF in
Fox-3-dependent activation of alternative splicing may
not be limited to this effect. PSF was originally found as
a spliceosome associated protein (54). A number of studies
pre-spliceosome and in the spliceosome at different
stages and in complexes containing snRNPs (55). PSF
has also been reported to interact directly with U5
snRNA and with the 50splice site under splicing condi-
tions (42,53). Therefore PSF may function as a mediator
between Fox-3-bound pre-mRNA and the splicing ma-
chinery. This notion is supported by our observation
that PSF does not bind directly to the IDDE or to the
pre-mRNA from the NMHC II-B minigene in the absence
of other nuclear factors. Additionally, both Fox-3 (or
other Fox proteins) and PSF are essential for the UGC
AUG-dependent activation of N30 splicing. In the
presence of endogenous PSF and Fox-2, exogenous
Fox-3 activates N30 splicing to some extent as does ex-
ogenous PSF. These effects of exogenous Fox-3 and PSF
are apparently additive. When endogenous Fox-2 or PSF
is eliminated, however, neither exogenous Fox-3 nor
PSF can activate N30 splicing. The enhancing effect of
PSF on N30 splicing is dependent on the UGCAUG
element, although it does not bind to this element, and
is absolutely dependent on the presence of Fox-3 (or
other Fox proteins). The enhancing effect of Fox-3 on
N30 splicing is also absolutely dependent on the
presence of PSF. Thus, the effects of the two proteins
are actually not additive, but rather cooperative. PSF
seems to function as a coactivator or mediator of Fox-3
during the splicing process. The simplest model to
3076 Nucleic Acids Research, 2011,Vol.39, No. 8
accommodate all the data is that PSF or the PSF-
containing complex bridges Fox-3 and the splicing
machinery. Although the detailed molecular mechanism
of N30 exon recognition following the Fox-3 and PSF
interaction remains to be elucidated, this interaction has
now been shown to be an integral part of the mechanism
responsible for Fox protein regulated activation of alter-
native exon inclusion via a downstream intronic enhancer.
Supplementary Data are available at NAR Online.
The authors thank Xuefei Ma, Mary Anne Conti and
Jong H. Kim for reagents and helpful discussions. They
thank Christian A. Combs (Light Microscope Core
Facility, NHLBI) and J. Philip McCoy Jr. (Flow
Cytometry Core Facility, NHLBI) for professional skills
and advice. They also thank Antoine F. Smith for tech-
nical assistance and Mary Anne Conti for critical reading
of the manuscript.
Funding for open access charge: Division of Intramural
Research, National Heart, Lung, and Blood Institute,
National Institutes of Health, USA.
Conflict of interest statement. None declared.
1. Wang,G.S. and Cooper,T.A. (2007) Splicing in disease: disruption
of the splicing code and the decoding machinery. Nat. Rev.
Genet., 8, 749–761.
2. Tazi,J., Bakkour,N. and Stamm,S. (2009) Alternative splicing and
disease. Biochim. Biophys. Acta, 1792, 14–26.
3. 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.
4. Voelker,R.B. and Berglund,J.A. (2007) A comprehensive
computational characterization of conserved mammalian intronic
sequences reveals conserved motifs associated with constitutive
and alternative splicing. Genome Res., 17, 1023–1033.
5. Yeo,G.W., Van Nostrand,E.L. and Liang,T.Y. (2007) Discovery
and analysis of evolutionarily conserved intronic splicing
regulatory elements. PLoS Genet., 3, e85.
6. Castle,J.C., Zhang,C., Shah,J.K., Kulkarni,A.V., Kalsotra,A.,
Cooper,T.A. and Johnson,J.M. (2008) Expression of 24,426
human alternative splicing events and predicted cis regulation in
48 tissues and cell lines. Nat. Genet., 40, 1416–1425.
7. Ule,J., Stefani,G., Mele,A., Ruggiu,M., Wang,X., Taneri,B.,
Gaasterland,T., Blencowe,B.J. and Darnell,R.B. (2006) An RNA
map predicting Nova-dependent splicing regulation. Nature, 444,
8. Yeo,G.W., Coufal,N.G., Liang,T.Y., Peng,G.E., Fu,X.D. and
Gage,F.H. (2009) An RNA code for the FOX2 splicing regulator
revealed by mapping RNA–protein interactions in stem cells.
Nat. Struct. Mol. Biol., 16, 130–137.
9. Xue,Y., Zhou,Y., Wu,T., Zhu,T., Ji,X., Kwon,Y.S., Zhang,C.,
Yeo,G., Black,D.L., Sun,H. et al. (2009) Genome-wide analysis of
PTB-RNA interactions reveals a strategy used by the general
splicing repressor to modulate exon inclusion or skipping.
Mol. Cell, 36, 996–1006.
10. Long,J.C. and Caceres,J.F. (2009) The SR protein family of
splicing factors: master regulators of gene expression. Biochem. J.,
11. Shepard,P.J. and Hertel,K.J. (2009) The SR protein family.
Genome Biol., 10, 242.
12. Spellman,R. and Smith,C.W. (2006) Novel modes of splicing
repression by PTB. Trends Biochem. Sci., 31, 73–76.
13. House,A.E. and Lynch,K.W. (2008) Regulation of alternative
splicing: more than just the ABCs. J. Biol. Chem., 283,
14. Forch,P., Puig,O., Martinez,C., Seraphin,B. and Valcarcel,J.
(2002) The splicing regulator TIA-1 interacts with U1-C to
promote U1 snRNP recruitment to 50splice sites. EMBO J., 21,
15. Goo,Y.H. and Cooper,T.A. (2009) CUGBP2 directly interacts
with U2 17S snRNP components and promotes U2 snRNA
binding to cardiac troponin T pre-mRNA. Nucleic Acids Res., 37,
16. Huh,G.S. and Hynes,R.O. (1994) Regulation of alternative
pre-mRNA splicing by a novel repeated hexanucleotide element.
Genes Dev., 8, 1561–1574.
17. Kawamoto,S. (1996) Neuron-specific alternative splicing of
nonmuscle myosin II heavy chain-B pre-mRNA requires a
cis-acting intron sequence. J. Biol. Chem., 271, 17613–17616.
18. Del Gatto,F., Plet,A., Gesnel,M.C., Fort,C. and Breathnach,R.
(1997) Multiple interdependent sequence elements control splicing
of a fibroblast growth factor receptor 2 alternative exon. Mol.
Cell Biol., 17, 5106–5116.
19. Modafferi,E.F. and Black,D.L. (1997) A complex intronic splicing
enhancer from the c-src pre-mRNA activates inclusion of a
heterologous exon. Mol. Cell Biol., 17, 6537–6545.
20. Jin,Y., Suzuki,H., Maegawa,S., Endo,H., Sugano,S.,
Hashimoto,K., Yasuda,K. and Inoue,K. (2003) A vertebrate
RNA-binding protein Fox-1 regulates tissue-specific splicing via
the pentanucleotide GCAUG. EMBO J., 22, 905–912.
21. Ponthier,J.L., Schluepen,C., Chen,W., Lersch,R.A., Gee,S.L.,
Hou,V.C., Lo,A.J., Short,S.A., Chasis,J.A., Winkelmann,J.C.
et al. (2006) Fox-2 splicing factor binds to a conserved intron
motif to promote inclusion of protein 4.1R alternative exon 16.
J. Biol. Chem., 281, 12468–12474.
22. Auweter,S.D., Fasan,R., Reymond,L., Underwood,J.G.,
Black,D.L., Pitsch,S. and Allain,F.H. (2006) Molecular basis of
RNA recognition by the human alternative splicing factor Fox-1.
EMBO J., 25, 163–173.
23. Nakahata,S. and Kawamoto,S. (2005) Tissue-dependent isoforms
of mammalian Fox-1 homologs are associated with tissue-specific
splicing activities. Nucleic Acids Res., 33, 2078–2089.
24. Underwood,J.G., Boutz,P.L., Dougherty,J.D., Stoilov,P. and
Black,D.L. (2005) Homologues of the Caenorhabditis elegans
Fox-1 protein are neuronal splicing regulators in mammals.
Mol. Cell Biol., 25, 10005–10016.
25. Baraniak,A.P., Chen,J.R. and Garcia-Blanco,M.A. (2006) Fox-2
mediates epithelial cell-specific fibroblast growth factor receptor 2
exon choice. Mol. Cell Biol., 26, 1209–1222.
26. Yang,G., Huang,S.C., Wu,J.Y. and Benz,E.J. Jr (2008) Regulated
Fox-2 isoform expression mediates protein 4.1R splicing during
erythroid differentiation. Blood, 111, 392–401.
27. Zhou,H.L. and Lou,H. (2008) Repression of prespliceosome
complex formation at two distinct steps by Fox-1/Fox-2 proteins.
Mol. Cell Biol., 28, 5507–5516.
28. Zhang,C., Zhang,Z., Castle,J., Sun,S., Johnson,J., Krainer,A.R.
and Zhang,M.Q. (2008) Defining the regulatory network of the
tissue-specific splicing factors Fox-1 and Fox-2. Genes Dev., 22,
29. Shibata,H., Huynh,D.P. and Pulst,S.M. (2000) A novel protein
with RNA-binding motifs interacts with ataxin-2. Hum. Mol.
Genet., 9, 1303–1313.
30. Lieberman,A.P., Friedlich,D.L., Harmison,G., Howell,B.W.,
Jordan,C.L., Breedlove,S.M. and Fischbeck,K.H. (2001)
Androgens regulate the mammalian homologues of invertebrate
sex determination genes tra-2 and fox-1. Biochem. Biophys. Res.
Commun., 282, 499–506.
Nucleic Acids Research,2011, Vol.39, No. 83077
31. McKee,A.E., Minet,E., Stern,C., Riahi,S., Stiles,C.D. and Download full-text
Silver,P.A. (2005) A genome-wide in situ hybridization map of
RNA-binding proteins reveals anatomically restricted expression
in the developing mouse brain. BMC Dev. Biol., 5, 14.
32. Lee,J.A., Tang,Z.Z. and Black,D.L. (2009) An inducible change
in Fox-1/A2BP1 splicing modulates the alternative splicing of
downstream neuronal target exons. Genes Dev., 23, 2284–2293.
33. Kim,K.K., Adelstein,R.S. and Kawamoto,S. (2009) Identification
of neuronal nuclei (NeuN) as Fox-3, a new member of the Fox-1
gene family of splicing factors. J. Biol. Chem., 284, 31052–31061.
34. Mullen,R.J., Buck,C.R. and Smith,A.M. (1992) NeuN, a neuronal
specific nuclear protein in vertebrates. Development, 116, 201–211.
35. Takahashi,M., Kawamoto,S. and Adelstein,R.S. (1992) Evidence
for inserted sequences in the head region of nonmuscle myosin
specific to the nervous system. Cloning of the cDNA encoding
the myosin heavy chain-B isoform of vertebrate nonmuscle
myosin. J. Biol. Chem., 267, 17864–17871.
36. Ma,X., Kawamoto,S., Uribe,J. and Adelstein,R.S. (2006)
Function of the neuron-specific alternatively spliced isoforms
of nonmuscle myosin II-B during mouse brain development.
Mol. Biol. Cell, 17, 2138–2149.
37. Guo,N. and Kawamoto,S. (2000) An intronic downstream
enhancer promotes 30splice site usage of a neural cell-specific
exon. J. Biol. Chem., 275, 33641–33649.
38. Phillips,C.L., Yamakawa,K. and Adelstein,R.S. (1995) Cloning
of the cDNA encoding human nonmuscle myosin heavy chain-B
and analysis of human tissues with isoform-specific antibodies.
J. Muscle Res. Cell Motil., 16, 379–389.
39. Qoronfleh,M.W., Ren,L., Emery,D., Perr,M. and Kaboord,B.
(2003) Use of immunomatrix methods to improve protein–protein
interaction detection. J. Biomed. Biotechnol., 2003, 291–298.
40. Shav-Tal,Y. and Zipori,D. (2002) PSF and p54(nrb)/NonO—
multi-functional nuclear proteins. FEBS Lett., 531, 109–114.
41. Damianov,A. and Black,D.L. (2010) Autoregulation of Fox
protein expression to produce dominant negative splicing factors.
RNA, 16, 405–416.
42. Peng,R., Dye,B.T., Perez,I., Barnard,D.C., Thompson,A.B. and
Patton,J.G. (2002) PSF and p54nrb bind a conserved stem in U5
snRNA. RNA, 8, 1334–1347.
43. Dong,X., Sweet,J., Challis,J.R., Brown,T. and Lye,S.J. (2007)
Transcriptional activity of androgen receptor is modulated by two
RNA splicing factors, PSF and p54nrb. Mol. Cell Biol., 27,
44. Kai,N., Mishina,M. and Yagi,T. (1997) Molecular cloning of
Fyn-associated molecules in the mouse central nervous system.
J. Neurosci. Res., 48, 407–424.
45. Norris,J.D., Fan,D., Sherk,A. and McDonnell,D.P. (2002) A
negative coregulator for the human ER. Mol. Endocrinol., 16,
46. Lim,J., Hao,T., Shaw,C., Patel,A.J., Szabo,G., Rual,J.F.,
Fisk,C.J., Li,N., Smolyar,A., Hill,D.E. et al. (2006) A
protein–protein interaction network for human inherited
ataxias and disorders of Purkinje cell degeneration. Cell, 125,
47. 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,
48. Fukumura,K., Kato,A., Jin,Y., Ideue,T., Hirose,T., Kataoka,N.,
Fujiwara,T., Sakamoto,H. and Inoue,K. (2007) Tissue-specific
splicing regulator Fox-1 induces exon skipping by interfering E
complex formation on the downstream intron of human
F1gamma gene. Nucleic Acids Res., 35, 5303–5311.
49. Kuroyanagi,H., Ohno,G., Mitani,S. and Hagiwara,M. (2007)
The Fox-1 family and SUP-12 coordinately regulate tissue-specific
alternative splicing in vivo. Mol. Cell Biol., 27, 8612–8621.
50. Kaneko,S., Rozenblatt-Rosen,O., Meyerson,M. and Manley,J.L.
(2007) The multifunctional protein p54nrb/PSF recruits the
exonuclease XRN2 to facilitate pre-mRNA 30processing and
transcription termination. Genes Dev., 21, 1779–1789.
51. Bond,C.S. and Fox,A.H. (2009) Paraspeckles: nuclear bodies built
on long noncoding RNA. J. Cell Biol., 186, 637–644.
52. Emili,A., Shales,M., McCracken,S., Xie,W., Tucker,P.W.,
Kobayashi,R., Blencowe,B.J. and Ingles,C.J. (2002) Splicing and
transcription-associated proteins PSF and p54nrb/nonO bind to
the RNA polymerase II CTD. RNA, 8, 1102–1111.
53. Kameoka,S., Duque,P. and Konarska,M.M. (2004) p54(nrb)
associates with the 50splice site within large transcription/splicing
complexes. EMBO J., 23, 1782–1791.
54. Patton,J.G., Porro,E.B., Galceran,J., Tempst,P. and
Nadal-Ginard,B. (1993) Cloning and characterization
of PSF, a novel pre-mRNA splicing factor. Genes Dev., 7,
55. Jurica,M.S. and Moore,M.J. (2003) Pre-mRNA splicing: awash
in a sea of proteins. Mol. Cell, 12, 5–14.
3078Nucleic Acids Research, 2011,Vol.39, No. 8