Friend of Prmt1, FOP is a Novel Component of the Nuclear SMN Complex Isolated Using Biotin Affinity Purification

1Department of Applied Biological Science, United Graduate School of Agriculture, Tokyo University of Agriculture and Technology, 3-5-8 Saiwai-cho, Fuchu, Tokyo 1838509, Japan 2Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST), Sanbancho 5, Chiyoda-ku, Tokyo 102-0075, Japan 3Biomolecular Characterization Team, RIKEN Advanced Science Institute, 2-1 Hirosawa, Wako, Saitama 351-0198 Japan 4Department of Chemistry, Graduate School of Science, Tokyo Metropolitan University, 1-1 Minami-ohsawa, Hachioji, Tokyo 192-0397, Japan 5Department of Cell Biology, Erasmus MC, 3015 GE Rotterdam, The Netherlands


Introduction
SMN is linked to one of the most common inheritable causes of childhood mortality, spinal muscular atrophy (SMA) [1], a neuromuscular disorder characterized by loss of motor neurons, which results from low levels of, or loss-of-function mutations in the SMN protein [2,3]. The SMN gene is duplicated as an inverted repeat on chromosome 5. Only mutations in the telomeric copy, SMN1, give rise to SMA [4]; another centromeric copy, SMN2, expresses a functionally defective form of the protein that lacks the carboxyl terminus encoded by exon 7, but its mutations are not related to SMA.
Human SMN protein is a 294-amino-acid polypeptide that forms complexes with at least eight proteins: Gemin2 (SIP1), Gemin3 (DEAD box RNA helicase 20, DDX20), Gemin4, Gemin5 (a WDrepeat protein), Gemin6, Gemin7, Gemin8 and unrip (unr interacting protein, STRAP). The SMN complexes promote assembly of UsnRNP core complex consisting seven different Sm proteins, and U1, U2, U4, U5, or U7 snRNAs [5]; i. e., the SMN complexes assist the ordered assembly of the seven Sm proteins as three sub-complexes, B/B'-D3, D1-D2 and E-F-G Sm onto the Sm sequence of UsnRNAs [6]. Protein arginine methyl transferase (Prmt) 5 promotes this assembly, in which it catalyzes symmetrical arginine methylation of some of the Sm core proteins [7]. Indeed, SMN interacts with a dimethylarginine residue on some of the Sm core proteins present in the UsnRNP core complex [8]. Gemin5 and unrip also participate in this UsnRNP assembly, and are detached from the SMN complexes-UsnRNPs at later stages of the biogenesis in the nucleus [9][10][11]. The assembled cytoplasmic UsnRNPs are transported into the nucleus, and are targeted to Cajal/ Gems (splicing snRNPs) or to the histone locus body (U7 snRNP) that constitutes a part of Cajal/Gems. Coilin p80 (coilin) is required for the localization of SMN complexes in the Cajal/Gems through direct interaction of SMN with a symmetrically dimethylated arginine in a RG domain of coilin [12,13]. Coilin also interacts with Sm proteins in UsnRNPs through its C-terminal domain, and the binding is not affected by dimethylation [14]. Sm proteins and UsnRNPs are probably released from the SMN complexes upon the binding with coilin in Cajal/Gems [15], and are then transferred into splicing speckles [16]. Dissociation of Gemin5 may also promote the release of UsnRNP from SMN complexes in Cajal/Gems, since Gemin5 is required for the initial binding of snRNA to the SMN complexes in the cytoplasm [17]. The UsnRNPs that are released from the Cajal/Gems then receive additional proteins from the nuclear speckles. The final assembly of the spliceosome from its component UsnRNPs (and other proteins) takes place in situ on nascent pre-mRNAs at the site of transcription that occurs at the outer edge of the nuclear speckle, where all or almost all introns are removed in conjunction with or soon after the transcription [18]. The nuclear speckles are rich in spliced mRNAs and various pre-mRNA metabolic factors, such as spliceosome assembly factor SC35, and thus, are believed to be the places that complete mRNA maturation, release for mRNA export with some recycling factors, and recycle splicing complex including UsnRNPs [19]. Although it has been well documented that improper process of mRNAs impedes transit of the mRNAs through the nuclear speckle [20,21], the mechanism how the mRNAs are retained in recycle splicing complex and impeded the transit remains undetermined. The splicing complexes including UsnRNPs that are recycled after each round of splicing into the cytoplasm, must pass again through the Cajal/Gems for re-assembly before they can take part in a further round of splicing [22][23][24].
In this study, we isolated SMN complexes present in the nucleus and identified Friend of Prmt1, FOP as a novel component of the nuclear SMN complexes. As noted by alternative names of FOP such as SRAG (small protein rich in arginine and glycine) [25] and CHTOP (Chromatin target of Prmt1) [26], human FOP has a central glycine-ariginine-rich (GAR) domain containing 26 RG/GR repeats that is recognized and methylated with Prmt1, localized mainly in the nuclear speckles and tightly associated with facultative chromatin in interphase cells [26]. The C-terminus of FOP also harbors a duplication of the sequence LDXXLDAYM (where "X" is any amino acid) [26]. FOP is expressed widely among tissues and cell lines [25], and is highly conserved in all vertebrates, while no orthologs was identified so far in yeast, worm, and fly. FOP has a critical role in specific mRNA transcriptional regulation during fetal globin expression [27], and is also involved in mRNA export as a component of TREX complex [28].

Cellular fractionation and preparation of nuclear extract
Sub-confluent 293TRex cells expressing HB-SMN-F or HB-FOP stably, 293EBNA cells, or 293T cells were collected, lysed with buffer A (16.7 mM Tris-HCl (pH7.4), 50 mM NaCl, 1.67 mM MgCl 2 , 1 mM PMSF, 2 µg/ml aprotinin, 2 µg/ml pepstatin A and 2 µg/ml leupeptin) containing 0.1% Triton X-100, incubated for 5 min on ice, and centrifuged at 1,000×g for 5 min. The supernatant was collected as the cytoplasmic extract (Cyto). To fully remove the cytoplasmic constituents, the pellet was suspended again with buffer A and centrifuged at 1,000×g for 5min. The supernatant was collected as the cytoplasmic wash fraction (Wash), and the pellet was collected as the nuclear pellet. To prepare the nuclear extract under low salt condition, the nuclear pellet was lysed with buffer B (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM MgCl 2 , 1 mM PMSF, 2 µg/ ml aprotinin, 2 µg/ml pepstatin A and 2 µg/ml leupeptin) containing 0.5% IGEPAL CA-630 for 10 minat 4°C and centrifuged for 20 min at 20,000 ×g. The supernatant was used as the nuclear extract under the low salt condition (Nuc: Low salt). To prepare the nuclear extract under high salt condition, the nuclear pellet was lysed with buffer C (50 mM Tris-HCl pH 7.4, 400 mM NaCl, 5 mM MgCl 2 , 1 mM PMSF, 2 µg/ ml aprotinin, 2 µg/ml pepstatin A and 2 µg/ml leupeptin) containing 0.5% IGEPAL CA-630 for 10 min at 4°C and centrifuged for 20 min at 20,000×g. The supernatant was used as the nuclear extract under the high salt condition (Nuc: high salt). To prepare total cell extract, the collected cells were lysed with 1 x SDS sample buffer, sonicated twice for 30 sec using a Bioruptor (Cosmo Bio, Tokyo, Japan) at its highest setting, and then centrifuged at 20,000×g for 30 min. The supernatant was used as total cell extract (TCE). An equal aliquot of each extract was subjected to western blot analysis. The nuclear extract under the high salt condition was subjected to the following biotin affinity pull down or immune precipitation.

Western blot analysis
Proteins were separated by SDS-PAGE and electrophoretically transferred to a PVDF membrane (Millipore, Billerica, MA). The membrane was blocked with 5% non-fat dried skim milk in TBS for 1 h at 25°C, and incubated with the appropriate primary antibody at 4°C overnight. After washing with TBS containing 0.1% (w/v) Tween 20 (TBST) for 10min three times, the membranes were incubated with a secondary antibody conjugated with horseradish peroxidase (HRP) for 1 h, and then washed with TBST for 10 min three times. To detect biotin-tag of HB-SMN-F or HB-FOP, HRP-conjugated streptavidin (Thermo Scientific) was used. Chemiluminescent-stained proteins were detected by using LAS4000 image analyzer (Fujifilm).

Biotin affinity purification method
The nuclear extract under the high salt condition was prepared as described in Cellular fractionation and preparation of nuclear extract. 4 mg of the nuclear extract was incubated with 30 µl of Neutr Avidin Agarose Resins (Thermo Scientific) for 4 h at 4°C and washed five times with 1 ml of buffer C. HB-tagged protein complexes were eluted with 150 µl of Protein-RNA extraction buffer (7 M Urea, 350 mM NaCl, 1% SDS, 10 mM Tris-HCl(pH8.0), 10 mM EDTA, 2% 2-mercaptoethanol) for 5 min. The eluted protein complexes containing RNAs were separated from Neutr Avidin Agarose Resins by centrifugation at 1,000×g for 5 min at 4°C, and subjected to Phenol-chloroform extraction; namely, 150 µl of the eluates were added to 150 µl of Phenol-chloroform (pH8.0) (Ambion) and 15 µl of 3 M sodium acetate (pH5.2), and after vigorous mixing for 20 min at 25°C, the organic phase containing proteins and the aqueous phase containing RNAs were separated by centrifugation at 20,000×g for 20 min at 4°C. Proteins from organic phase were precipitated with 700 µl of isopropanol and following centrifugation at 20,000×g for 10 min at 4°C. RNAs from aqueous phase were collected by isopropanol precipitation. The precipitated proteins or RNAs were washed independently with 75% ethanol, dried up, and subjected to SDS-PAGE or denaturing urea PAGE, respectively. For liquid chromatography (LC)-tandem mass spectrometry (MS/MS) analysis of the HB-FOP complex, the precipitated proteins were dissolved with 5 µl of dissolution buffer (8M Urea, 50 mM Tris-HCl (pH 8.0), and 1% 2-mercaptoethanol), mixed well, diluted to ten times with 50 mM of ammonium hydrogen carbonate and 5 mM of calcium chloride, and subjected to the digestion with 200 ng of trypsin for at 37°C for 6 hr. The resulting trypsin digest solution containing peptides adding 0.1% formic acid was analyzed using a nanosccale LC-MS/MS system.

Protein identification by LC-MS/MS
Pulled-down proteins were separated on a gel by SDS-PAGE; the gel was cut into some pieces, each of which was subjected to in-gel digestion with trypsin [30], and the resulting peptides were analyzed using a nanoflow LC-MS/MS system with qua-drupole-time-of-flight 2 hybrid mass spectrometer (Q-Tof2, Micromass, Wythenshawe, UK) as described [31,32]. The peptide mixture was applied to a Mightysil-RP-18 (3 µm particle, Kanto Chemical, Osaka, Japan) fritless column (45 mm×0.150 mm i.d.) and separated using a 0 -40% gradient of acetonitrile containing 0.1% formic acid over 80 min at a flow rate of 50 nl/min. Eluted peptides were sprayed directly into Q-Tof2. The peptides were detected in the MS mode to select a set of precursor ions for a data-dependent, collision-induced dissociation (CID) MS/MS analysis, and every 4 s the four ions having the greatest signal intensity were subjected to the MS/MS analysis. The MS/MS signals were acquired by MassLynx version 3.5 (Micromass, UK) and converted to text files by ProteinLynx software (Micromass). The database search was performed by MASCOT version 2.3.2 (Matrix Science Ltd., London, UK) against the UniProt human databases version SwissProt_2013_08, 540732 sequences, with the following parameters: fixed modification; carbamidemethyl (Cys), variable modifications: oxidation (Met), N-acetylation, pyroglutamine; maximum missed cleavages: 2; peptide mass tolerance: 200 ppm; MS/MS tolerance: 0.5 Da. The criteria were based on the vendor's definitions (Matrix Science, Ltd.). Furthermore, we set more strict criteria for protein assignment: more than two peptides with p value<0.05 were considered as 'hit'.

Denaturing urea PAGE of RNAs
The dried up RNA from the pulled-down protein complexes was resolved well in 100% formamide, and heated at 65°C for 5 min and cooled immediately into ice. Then one-fourth of volume of 50 mM EDTA (pH8.0) containing loading dye was added to RNAs in formamide and used as a RNA sample. RNA was separated on 8% (w/v) polyacrylamide gels containing 8 M urea and 0.5×TBE (45 mM Tris, 32.3 mM boric acid, 1.25 mM EDTA, pH 8.3) with 0.5×TBE as a running buffer, stained with SYBR Gold (Invitrogen, Carlsbad, CA) for 10 min, and visualized with LAS4000 image analyzer.

In gel RNase digestion for LC-MS/MS analysis
In gel RNase digestion was done principality by the method described [33]. Briefly, gel pieces containing RNA bands were excised from gels, cut into small pieces, and dried under vacuum. The gel pieces were digested with 15 µl of 2 ng/µl RNase T1, with incubation at 37°C for 1 hr. The nucleolytic fragments were extracted from the gel using 100 µl of RNase-free water, passed through a centrifugal filter unit with a polyvinylidene fluoride membrane (Ultrafree-MC, Millipore, Billerica, MA), and then 5 µl of 2 M triethylammonium acetate (pH 7.0) was added before LC-MS analysis.

LC-MS/MS Apparatus for RNA Analysis
The LC system used was essentially as described [34], consisting of a nanoflow pump (LC Assist, Tokyo, Japan) that delivers solvent to the fritless spray tip electrospray ionization column and a ReNCon gradient device. The column was prepared with a fused-silica capillary (150 µm i.d.×375 µmo.d.) using a laser puller (Sutter Instruments Co., Novato, CA) and was slurry-packed with reversed-phase material (Develosil C30-UG-3, particle size 3 µm, Nomura Chemical, Aichi, Japan) to a length of 50 mm. High voltage for ionization in negative mode (−1.4 kV) was applied to the metal union, and the eluate from RPLC was Affinity Proteomics J Proteomics Bioinform ISSN: 0974-276X JPB, an open access journal sprayed on-line to an LTQ-Orbitrap hybrid mass spectrometer (model XL, Thermo Fisher Scientific, San Jose, CA). Reverse phase separation of oligoribonucleotides was performed at flow rate of 100 nl/min using a 30-min linear gradient from 10 to 40% methanol in 10 mM triethylammonium acetate (pH 7.0).The mass spectrometer was operated in a data-dependent mode to automatically switch between Orbitrap-MS and linear ion trap-MS/MS acquisition as described [35]. Survey full scan MS spectra (from m/z 500 to 1,500) were acquired in the Orbitrap with resolution R=30,000 (after accumulation to a target value of 500,000 ions in the linear ion trap). The most intense ions (up to four, depending on signal intensity) were sequentially isolated for fragmentation in the linear ion trap using CID at a target value of 30,000 ions. An MS scan was accumulated for 2 s and an MS/MS scan for 3 s. The resulting fragment ions were recorded in the linear ion trap with a high-scan rate 'Enhanced' mode. Target ions selected for MS/ MS were dynamically excluded for 60 s. General mass spectrometric conditions were as follows: electrospray voltage, 1.4 kV; no lock mass option; with sheath and auxiliary gas flow; normalized collision energy, 35% for MS/MS. Ion selection threshold was 10,000 counts for MS/MS. An activation q-value of 0.25 and activation time of 30 ms were applied for MS/MS acquisitions. LC was performed at a flow rate of 100 nl/ min using a 30-min linear gradient from 10% to 40% methanol in 10 mM triethylammonium acetate (pH 7.0). The mass spectrometer was operated in a mode to automatically switch between Orbitrap-MS and linear ion trap-MS/MS acquisition as described [35].

Database search and interpretation of MS/MS Spectra of RNA
We used the software, Ariadne [36], for database searches for RNA identification. Ariadne searched DNA/RNA sequence database using MS/MS data of an RNase T1 digest of an RNA sample. Databases used were an in-house small RNA database, which was constructed from the cDNA of less than 1000 bases in NCBI GenBank (http://www. ncbi.nlm.nih.gov/genbank/), and EMBL sequence database (http:// www.ebi.ac.uk/embl/), which contained more than 1,500 entries, and the genome database of Homo Sapiens (reference assembly GRCh37 obtained from ftp://ftp.ncbi.nlm.nih.gov/genomes/H_sapiens/). The following search parameters were used: the maximum number of missed cleavages was set at 1; the maximum number of variable modifications for a nucleotide was 1; The precursor mass tolerance of ± 20 ppm and MS/MS tolerance of ± 750 ppm were allowed; The significance thresholds for both the nucleotide and RNA identifications were set at 5%; The nucleotide lengths considered by the software were from 4 to 19. The original MS/MS spectra for identified nucleotides by Ariadne searches were carefully inspected by at least two MS specialists; Nucleotides with less than 50% sequence coverage were eliminated from the identifications; the length of an identified RNA estimated by PAGE and that from the database was used to confirm RNA identification. After the RNA identification, unidentified MS/MS spectra were manually inspected to identify unidentified nucleotides of the RNA such as those having lengths shorter than 3 nt or longer than 20 nt. The manual identification criteria were as follows: mass differences for nucleotides were within 20 ppm; product ions should cover more than half of the sequence of a candidate nucleotide.

Immunoprecipitation of endogenous SMN complexes
Mouse monoclonal anti-SMN antibody bound to Dynabeads were prepared by incubating 3 µg of the antibody with 15 µl of protein G Dynabeads (Invitrogen, Carlsbad, CA) in PBS containing 0.01% Triton X-100 for 30 min at 25°C, and were washed with PBS containing 0.01% Triton X-100 three times. The antibody-bound Dynabeads were incubated with 2 mg of the nuclear extract from 293EBNA cells for 4 h at 4°C and washed five times with 1 ml of buffer C. SMN binding proteins were eluted by 1X SDS sample buffer, and subjected to the following western blot analysis. For SILAC analysis, anti-SMN bound beads was made using Dynabeads Co-Immunoprecipitation Kit (Invitrogen) according to the manufacturer's instructions. Briefly, 20 µg of anti-SMN antibody was bound to 5 mg of Dynabeads, Epoxy M-270. FLAG-tagged protein was pulled down using 15 µl of anti-FLAG M2 agarose beads (SIGMA). The cytoplasmic extract or nuclear extract was incubated with FLAG-M2 affinity Agarose Resins (SIGMA) for 2 h at 4°C. After washing five times with buffer B or buffer C, FLAGprotein complexes were eluted with 500 µg/ml of FLAG peptide or 1X SDS sample buffer, and subjected to the following SDS-PAGE and silver staining or western blot analysis.

Ribonuclease treatment of the pulled-down protein complexes
The pulled-down complexes bound to the agarose resins described above were incubated with 50 mM Tris-HCl pH 8.0, 150 mM NaCl, 5 mM MgCl 2 containing 10 µg/ml RNase A for 30 min at 37C, washed three times with 1 ml of buffer C and eluted with Protein-RNA extraction buffer. The eluted proteins and RNAs were separated as described above, and subjected to the following western blot analysis and northern blot analysis, respectively.

Immunofluorescence staining
The FLAG-FOP inducibly expressed cells were grown on collagen-coated culture slides and FLAG-FOP was expressed by the treatment with 100 ng/ml of doxycycline for 24 hr. To visualize nuclei, the cells were washed with PBS followed by permeabilization with PBS containing 0.1% (w/v) Triton X-100 for 10 min at room temperature. After washing twice with PBST, the cells were fixed with 4% paraformaldehyde in PBS for 10 min at room temperature and washed twice with PBST. The cells were then blocked with 3% (w/v) nonfat dried milk in PBS for 1 h, and incubated with anti-FLAG rabbit polyclonal antibody (SIGMA) and anti-SMN antibody (2B1) as the primary antibodies for more than 2 h. The cells were washed three times in PBST for 10 min and then incubated with FITC-conjugated anti-rabbit IgG goat antibody (American Qualex) and Cy-conjugated anti-mouse IgG sheep antibody (SIGMA) as the secondary antibodies for 1 h at room temperature. The cells were washed three times in PBST for 10 min and then mounted with VECTASHIELD (Vector Laboratories) containing 4',6-diamidino-2-phenylindole (DAPI). The resulting cells were examined with an Axiovert 200 M microscope (Carl Zeiss, Germany).

Quantitative MS with SILAC
293T cells were grown in SILAC DMEM (Thermo Scientific) supplemented with either "light" L-12 C 6 lysine-2HCl (L-Lys) and L-12 C 6 , 14 N 4 arginine-HCl (L-Arg) or "heavy" L-13 C 6 lysine-2HCl (H-Lys) and L-13 C 6 , 15 N 4 arginine-HCl (H-Arg). For full incorporation of the "light" or "heavy" amino acids, cells were grown for at least six passages. Subsequently, 2×10 6 cells cultured in DMEM with L-Lys and L-Arg, or H-Lys and H-Arg were transfected with FOP scRNA and siRNAas described above, respectively. After being cultured in 96 hour, each of the si/scRNA transfected cells was harvested and immunoprecipitated with anti-SMN antibody as described above. Their immunoprecipitants were eluted with 1x SDS sample buffer, respectively, combined and subjected to SDS-PAGE followed by staining with colloidal blue staining kit (Invitrogen). For quantitative MS, the gel stained with colloidal blue was cut into 16 pieces, each of which was subjected to in-gel digestion with trypsin and the resulting peptides were analyzed using a nanoflow LC-MS/MS system with LTQ-Orbitrap hybrid mass spectrometer (model XL, Thermo Fisher Scientific, CA, USA). The mass spectrometer was operated in a data-dependent mode to automatically switch between Orbitrap-MS and linear ion trap-MS/ MS acquisition. Survey full scan mass spectra (from m/z 450 to 1500) acquired in the Orbitrap with resolution r=15,000 (an AGC target value of 500,000 counts). The most intense ions (top ten, depending on signal intensity) were sequentially isolated for fragmentation. The resulting fragment ions were recorded in the linear ion trap with a normal mode (Scan Rate, 16666Da/sec). General mass spectrometric conditions were as follows: electrospray voltage, 1.6 kV; normalized collision energy, 35% for MS/MS. Ion selection threshold was 10,000 counts for MS/MS. The resulting raw files were processed and analyzed using Proteome Discoverer version 1.3 (Thermo Fisher Scientific) with default parameters, except for the labels where lysine 6 and arginine 10 were selected. Searches were conducted within MASCOT version 2.3.2 against UniProt human database version SwissProt_2013_08, 540732 sequences and a reverse database, with the following parameters: fixed modification; carbamidemethyl (Cys), variable modifications: oxidation (Met), N-acetylation, pyroglutamine, K+6 and R+10; maximum missed cleavages: 2; peptide mass tolerance 15 ppm; MS/MS tolerance: 0.8 Da. We selected the candidate peptides with probabilitybased Mowse scores that exceed its threshold indicating a significant homology (p < 0.05), and referred to them as "hits." The criteria were based on the vendor's definitions. Furthermore, we set more strict criteria for protein assignment: more than two peptides with p value<0.05 were considered as 'hit'. Protein ratios were calculated based on vendor's default definition. Only protein groups with at least three SILAC counts were considered for further analysis.

The nuclear SMN associates with Friend of Prmt1 (FOP)
Although the roles and the components of SMN complexes present in the cytoplasm have been well established, those in the nucleus are unknown except that coilin and fibrillarin were associated with SMN in the nucleus [12,13,37]. We, therefore, tried to isolate the SMN complexes present in the nucleus, and first constructed 293TRex cell line, which has a single-copy transgene encoding HB-SMN-F [29] ( Figure 1A). The cytomegalovirus promoter was used to drive stable expression of the transgene, which was integrated at a common locus in the 293TRex genome [38]. As most SMN complex population existed in the cytoplasm, we first separated the HB-SMN-F expressing 293TRex cells into the cytoplasmic and the nuclear fractions by cell fractionation method with the lysis buffer containing 150 mM NaCl (low salt buffer). Under this low salt condition, about 80% of the SMN population in the total cell extract was recovered as the SMN complexes from the cytoplasmic fraction by a pull down method using anti-FLAG M2 antibody beads (Figures 1B and 1C) [30][31][32], while we detected only a small amount of the nuclear HB-SMN-F ( Figure 1B). To improve the yield of the nuclear SMN, we used lysis buffer containing 400 mM NaCl (high salt buffer), and found that this high salt buffer increased the yield of the nuclear HB-SMN-F about three folds of that extracted with low salt buffer ( Figure 1B). Our initial attempts to extract the nuclear SMN complex from anti-FLAG M2 antibody beads with the high salt buffer were unsuccessful (data not shown), probably because the interaction of anti-FLAG M2 antibody and the FLAG epitope was inhibited under high salt condition. Therefore, the biotin affinity purification procedure presented here is essential for efficient recoveries of the nuclear HB-SMN-F and its associated proteins ( Figures 1B and 1C).The subsequent MS-based protein identification showed that the isolated nuclear SMN complexes contained Gemins 2, 3 (DDX20), 4, and 8, endogenous SMN, and fibrillarin that is reported to associate with SMN ( Figure  1C and Supplementary Table 1) [39]. In addition, the MS-based method identified Friend of Prmt1, FOP, with the protein score 113 and matched sequence 2, as a potential component of the nuclear SMN complexes ( Figure 1D and supplementary Table 1).

FOP is a component of the nuclear SMN complexes that lackSmB/B' and SmD1
We validated the results of MS-based analyses by western blotting; the method clearly detected Gemins 2, 3, 4, and 8, endogenous SMN, fibrillarin, and FOP (Figure 2A). The method also detected Gemin6 and unrip, the known components of SMN complexes, as well as coilin, which is a marker of Cajal/Gems in the nuclear HB-SMN-F complexes ( Figure 2A). Interestingly, SmB/B' and SmD were absent in the isolated nuclear SMN complexes (Figure 2A). We considered that FOP is a novel de facto component of the SMN complexes present specifically in the nucleus for the additional evidences: (1) the endogenous nuclear SMN complexes prepared by immunoprecipitation using anti-SMN antibody contained FOP, as well as the several known components, Gemins 2, 3, 4, and endogenous SMN ( Figure 2B); (2) the SMN complexes isolated from the nuclear fraction prepared by centrifugal cell fractionation technique contained FOP, whereas the equivalent complexes isolated from the cytoplasmic fraction did not (data not shown); (3) the pulleddown protein mixture prepared from the nuclear fraction using HB-SMN-F as affinity bait contained FOP, whereas the equivant mixture prepared from the cytoplasmic fraction did not ( Figure 2C).
We also recovered several small RNA components from the nuclear SMN complexes isolated under the high salt condition ( Figure  2D), and identified the RNAs by the LC-MS/MS system equipped with a genome-oriented database searching engine Ariadne [33,35,36]. This analysis identified a number of known RNA components of the SMN complexes, U1, U3, and U4 snRNAs ( Figure 2D, supplementary Table  2), showing that our preparation of the nuclear SMN complexes in fact contained UsnRNAs. However, the result suggest that the UsnRNAs are present in only small population of the isolated nuclear SMN complexes because the nuclear SMN complexes lack SmB/B' and SmD (Figure 2A).

The FOP-associating SMN complexes do not contain U1 and U2 snRNAs
To confirm the association of FOP with the nuclear SMN complexes further, we performed a reverse pulldown-assay using FOP associating protein(s) as bait(s). We prepared the high salt nuclear extract of 293TRex cells expressing HB-FOP, where the DNA fragment-encoding FOP fused to HB-tag towards amino-terminus was integrated into the 293TRex genome ( Figure 3A). Like the nuclear SMN, the expressed HB-FOP was extracted more efficiently under the high salt condition than the low salt condition ( Figure 3B). This is consistent with the report of van Dijk et al. [26] which demonstrated that FOP was associated with chromatin, localized mainly in facultative heterochromatin region, and thus was extracted mostly under high salt conditions. Biotin affinity purification and MS-based analyses showed that HB-FOP was associated with SMN, Gemins 2, 3, 4, and 8, and unrip ( Figure 3C and Table 1 and  Supplementary Table 3). The analyses also identified Fragile X mental retardation 1 protein (FMR1), which was reported to associate with SMN [40] ( Table 1, and supplementary Table 3). The other proteins identified include protein arginine N-methyltransferases 1 (Prmt1) and 5 (Prmt5) both of which were reported to be the FOP binding proteins (26) Table 1 and Supplementary Table 3). Western blot analysis using available antibodies confirmed minimally the presence of SMN, Gemins 2, 3, 4, 6, and 8, unrip, p32, Prmt1, and FMR1 in the HB-FOP associating complexes ( Figure 3D). The isolated HB-FOP associating complexes contained neither Sm B/B' and Sm D, nor Gemin5 and coilin ( Figure 3D). HB-FOP was associated with U4, U5, and U6 snRNAs but not with U1, U2, and U3 snRNAs as revealed by northern blot analysis ( Figure 3E). These data suggest that FOP is associated with SMN complexes that were separated from the major U1 and U2 snRNPs containing Sm proteins in Cajal/Gems. We treated the HB-FOP associating complexes with RNase A and found that SMN and FOP were associated with each other independently of RNA, whereas the RNase treatment reduced U6 snRNA in the complexes ( Figure 3F). In addition, we examined whether SMN and FOP were co-localized in the nucleus by immunocytochemical analysis. Because SMN mainly localizes in the cytoplasm, we stained SMN and FOP in the nucleus with their corresponding antibodies after the removal of the cytoplasmic proteins by permeabilization with 0.1% Triton, and showed that the two proteins were co-localized in several foci in the nucleus ( Figure 3G), supporting their specific in vivo association.

FOP is required for the association of the nuclear SMN complexes with hnRNPs, histones and RNA-binding proteins
To gain insight into the functional relationship between FOP and the nuclear SMN complexes, we examined the effect of FOP knockdown on protein components associated with the nuclear SMN by SILAC MS analysis. We cultured scRNA-and siRNA-treated cells with light and heavy amino acids, and fractionated into the cytoplasmic and nuclear extracts by cell fractionation method, respectively. Each of the nuclear extracts was then subjected to immunoprecipitaion with anti-SMN antibody to isolate the SMN-associating complexes ( Figure  4A). The 96 h-siRNA treatment reduced FOP by about 10% relative to GAPDH when compared with scRNA treatment ( Figure 4B). The immunoprecipitates from scRNA-and siRNA-treated cells were combined, separated by SDS-PAGE, stained with colloidal blue ( Figure  4B), and subjected to in-gel protease digestion. Relative ratios of the protein components pulled down with endogenous SMN with and without FOP knockdown were obtained by LC-MS/MS analysis of the digest as the ratios of the mass intensities of the corresponding peptides with light and heavy amino acids [heavy (siRNA)/light (scRNA)] ( Figures 4B and 4C

Discussion
We summarized the strategy of the present study in Figure 5. In this study, we successfully isolated the nuclear SMN complexes by a combination of high salt extraction and biotin affinity purification methods, and found that FOP is a novel component of the nuclear SMN complexes (Step 1 in Figure 5). Our data suggest that the nuclear SMN complexes isolated using HB-SMN-F as affinity bait contain at least two distinct SMN complexes; one contains coilin in addition to Gemins and unrip (Figurer 2A), and another contains FOP but does not contain coilin, U1 and U2 snRNAs ( Figure 3D and 3E). The former nuclear SMN complex probably corresponds to that present in Cajal/Gems, because the complex contains coilin, which is presumably required to dissociate the SMN complex from U1 and U2 snRNPs in the Cajal/ Gems [12][13][14][15]17]. Our results also suggest that the use of HB-FOP as affinity bait isolates preferentially the latter SMN complex(FOP-SMN complex) (Step 2 in Figure 5). In addition, the quantitative analysis of the components of the endogenous nuclear SMN complexes before and after the knockdown of FOP suggests that the FOP-SMN complex is retained on chromatin [26] possibly via the association with histones (Step 3 in Figure 5). FOP is also required for the association of SMN complexes with a number of hnRNPs and RNA binding proteins (Step 3 in Figure 5). Based on these data and the previous reports [26], we propose that FOP-SMN complex has a role in retaining spliced and/or unspliced mRNAs in near the transcription sites on chromosome. This does not conflict to the previous notion that FOP has a critical role in specific mRNA transcriptional regulation and mRNA export [27,28]. We do not exclude, however, other roles of FOP-SMN complex, such that it may prevent improperly processed mRNAs from transit through the nuclear speckle [20,21], or that FOP may keep the nuclear SMN complexes near the transcription sites on chromosome until it is used for recycling UsnRNPsafter splicing [22][23][24].
Thus, for the first time in our knowledge, our present study identifies the FOP-SMN complexes that are separated from the mature U1 and U2 snRNPs in the nucleus, and provides its potential role in post-transcriptional gene regulation; i.e., FOP bridges between the nuclear SMN complexes and the region near the transcription sites on chromosome via the association with histones, and possibly retains spliced and/or unspliced mRNAs via the association with a number of hnRNPs and RNA-binding proteins. It has been reported that loss of Cajal/Gems is a cellular hallmark of SMA and amyotrophic lateral sclerosis (ALS) [41,42]. Since the formation of FOP-SMN complex is probably related to or coupled with the events taken place in Cajal/ Gems, the detailed analysis of FOP-SMN complex may open up a novel molecular mechanism by which a common biochemical pathway leads to SMA and ALS.
Step 1 Step 2 Step