Anti-IRS-1 antibody was raised in rabbits as described (31). The following peptide and antibodies were purchased: FLAG (F3165, Sigma, St. Louis, MO, USA), anti-FLAG antibody-conjugated agarose beads (A2220, Sigma), and anti-Myc antibody (9E10, Millipore, Billerica, MA, USA). All other chemicals were of reagent grade and were obtained commercially.

To generate FLAG-tagged or Myc-tagged IRS-1 and IRS-2, the open reading frame of rat IRS-1 and human IRS-2 was subcloned into the pCMV–FLAG-2 vector (Sigma) or pCMV–Myc vector in-frame as described previously (32, 33). Full-length rat PABPC1 was subcloned into the pCMV mammalian expression vector containing an N-terminal FLAG tag (FLAG–PABPC1) as follows. An EcoRI restriction site was introduced at the 5′ end of the PABPC1 open reading frame and a SalI site at the 3′- end by PCR using the oligonucleotides 5′-TTAAGAATTCAAGATGAACCCCAGCGCCCCCAGCTA-3′ and 5′-TTAAGTCGACTTAGACAGTTGGAACACCAGTGG-3′, respectively. The PCR product was digested with EcoRI and SalI and subcloned into pCMV–FLAG cut with EcoRI and SalI. Full-length human 15.5K was subcloned into the pCMV mammalian expression vector containing an N-terminal FLAG tag (FLAG-15.5K) or N-terminal GFP tag (GFP-15.5K) as follows. A BglII restriction site was introduced at the 5′ end of the 15.5K open reading frame and a SalI site at the 3′ end by PCR using the oligonucleotides 5′-TTAAGAATTCAAGATGAACCCCAGCGCCCCCAGCTA-3′ and 5′-TTAAGTCGACTTAGACAGTTGGAACACCAGTGG-3′. The PCR product was digested with BglII and SalI and subcloned into pCMV–FLAG and pEGFP-C1 cut with BglII and SalI. Full-length human coilin was subcloned into the pCMV mammalian expression vector containing an N-terminal GFP tag (GFP–coilin) as follows. A KpnI restriction site was introduced at the 5′ end of the coilin open reading frame and a BamHI site at the 3′ end by PCR using the oligonucleotides 5′-TTAAGGTACCATGGCAGCTTCCGAGACGGTTAGGCTACG-3′ and 5′-TTAAGGATCCTCAGGCAGGTTCTGTACTTGATGTGTTACTTGG-3′. The PCR product was digested with KpnI and BamHI and subcloned into pEGFP-C1 cut with KpnI and BamHI.

Plasmids expressing Myc–IRS-1 fused with SV40 large T antigen-derived Nuclear Localizing Signal (NLS, PKKKRKV) in its C-terminus was constructed by PCR using pCMV–Myc–IRS-1 as template and the oligonucleotides 5′-AAGGCTGTCCTTGGGGGATCC-3′ and 5′-ATCGGGATCCCTATACCTTTCTCTTCTTTTTTGGTTGACGGTCCTCTGGTTG-3′. The PCR product amplified was digested with BamHI and subcloned into pCMV–Myc–IRS-1 cut with the same restriction enzyme.

For the in vitro RNA synthesis, U96A snoRNA sequence was amplified by PCR using the oligonucleotides 5′-TTAATCTAGACCTGGTGATGACAGATGGCATTGTCAG-3′ and 5′-TTAACTGGAGTTCAGAATTGCAGGACATGTCCTCACTCC-3′. The primers were engineered to contain XbaI and XhoI sites to introduce these restriction sites at the 5′ and 3′ end of the PCR product, respectively. The PCR product was digested with XbaI and XhoI and subcloned into the downstream of T7 promoter of pBS-KS(+) cut with XbaI and XhoI, which is referred to as pBS-KS(+)-U96A snoRNA. U6 small nuclear RNA (snRNA) sequence was also amplified by the PCR using the oligonucleotides 5′-TTAATCTAGAGTGCTCGCTTCGGCAGCACATATACTAAAATTGG-3′ and 5′-TTAACTCGAGAAAATATGGAACGCTTCACGAATTTGCGTGTCATCC-3′, and subcloned into the downstream of T7 promoter of pBS-KS(+) cut with XbaI and XhoI.

Insulin receptor substrates-1 knockout mice were kindly donated by Dr. Kadowaki and Dr. Ohsugi (Faculty of Medicine, The University of Tokyo). All animal experiments in this study were performed according to procedures approved by the Committee on Laboratory Animal Care, Graduate School of Agriculture and Life Sciences, The University of Tokyo.

Human embryonic kidney 293 (HEK293) cells were kindly provided by Dr. Koichi Suzuki (National Institute of Infectious Diseases, Tokyo, Japan). MCF-7 human breast cancer cells (ATCC No. CRL8305) were a kind gift from Dr. Yoichi Hayakawa (Tokyo University of Science, Tokyo, Japan). HEK293, MCF-7 cells, and HeLa cells were maintained at 37°C in a humidified CO₂-controlled atmosphere in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 0.1% NaHCO₃, 50 IU/ml penicillin, 50 μg/ml streptomycin, 100 μg/ml kanamycin, and 0.5 μg/ml amphotericin B. For the serum-starvation, cells were starved for 16 h in DMEM supplemented with 0.1% bovine serum albumin (BSA).

Primary mouse embryonic fibroblasts (MEFs) were prepared from littermates according to the previous report (34) and kept in culture for 2–3 weeks, respectively. Briefly, MEFs were prepared from E14.5 embryos. Embryos were dissociated by 0.025% trypsin in 0.2% EDTA at 37°C for 10 min and then treated with DNaseI. After filtering fibroblasts with a 70 μm cell strainer, they were cultured in DMEM supplemented with 10% FBS, 100 μg/ml streptomycin, 100 U/ml penicillin, and 2 mM glutamine. MEFs were cultured at 9% CO₂ at 37°C and experiments were performed at first passages (P1–P4).

For the plasmid transfection, the expression vectors were transfected into HEK293 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA), according to manufacturer’s protocol.

Immunoprecipitation and immunoblotting were performed as described elsewhere (30).

Cross-linking and immunoprecipitation (CLIP) was performed using serum-starved HEK293 cell expressing FLAG–IRS-1 or FLAG–IRS-2 according to Ule et al. (35) with minor modification.

Nucleus/cytoplasm fractions were prepared according to previous report (36).

Total RNA was extracted from cells using the TRIzol reagent (Invitrogen) followed by DNase I treatment. For RT-PCR, 1 μg of total RNA was reverse transcribed using random primers (Invitrogen) and SuperScript II reverse transcriptase (Invitrogen). A fraction of the RT reaction products was used in subsequent PCR reactions. Amplification parameters were denaturation at 94°C for 30 s, annealing at 64°C for 30 s, and extension at 72°C for 30 s in a 30-cycle program. Primer sequences were as follows: Rack1 exon1–exon8 (5′-ACTGAGCAGATGACCCTTCGTG 3′ and 5′-GTTGTCCGTGTAGCCAGCAAAC-3′, 912 bp). The simultaneous analysis of serial dilution of the amount of RNA used in the RT-PCR reactions ensured that these reactions were quantitative.

For quantitative RT-PCR, the cDNA prepared was subjected to real-time PCR (ABI StepOne™ Real-Time PCR Systems), using SYBR Green Real-time PCR Master Mix Plus (TOYOBO, Osaka, Japan). Data were expressed as relative mRNA levels normalized to housekeeping gene (Gapdh) expression level in each sample. The primer sequences are as follows: Rack1 exon2–intron2 (5′-GATGGTCAGTTTGCCCTCTC-3′, 5′-CTCAGTTCTGCCCACTTTCC-3′), Rack1 exon3–exon4 (5′-GTCCCGAGACAAGACCATAAAG-3′, 5′-TGATAGGGTTGCTGCTGTTC-3′), Gapdh (5′-GTGTTCCTACCCCCAATGTG-3′, 5′-CCTGCTTCACCACCTTCTTG-3′).

To detect small RNAs, total RNA was separated on a 10% denaturing acrylamide gel and transferred to Hybond-N⁺ membrane (GE Healthcare Biosciences, Pittsburgh, PA, USA) using a Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad, Hercules, CA, USA). For the detection of U96A snoRNA, DNA probes were synthesized with a Megaprime DNA labeling Kit (GE healthcare) and [α⁻³²P]dCTP using U96A snoRNA cDNA as a template. The probes were hybridized using prehybridization buffer (50% deionized formamide, 5 × SSPE, 5 × Denhardt, 10% dextran sulfate, 0.1% SDS, 20 μg/ml salmon sperm) at 50°C. Densitometric analysis of RNA bands was performed using ImageJ gel analysis software (http://rsbweb.nih.gov/ij/).

HEK293 cells expressing FLAG-tagged proteins were serum-starved and washed twice with ice-cold PBS and irradiated with 4000 J/m² ultraviolet (UV) in ice-cold PBS. The cells were collected in 1.5 ml microtubes and pelleted by centrifugation. The cell pellets were resuspended with 1 ml of NET-2 [50 mM Tris–HCl, pH 7.4, 300 mM NaCl, 0.05% NP-40, 500 μM Na₃VO₄, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 20 μg/ml phenylmethylsulfonyl fluoride (PMSF); 100 KIU/ml aprotinin] containing 100 U of RNase-inhibitor (Ambion, Austin, TX, USA) and were incubated for 10 min on ice. Cells were lysed by 15 (3 × 5) bursts using a Branson Sonifier. Homogenates were centrifuged at 14,000 g for 10 min at 4°C, cleared by incubation with 20 μl Protein G-Sepharose beads (GE Healthcare) for 60 min, and incubated with a specific antibody for 1.5 h at 4°C, after which 30 μl Dynabeads protein G paramagnetic beads (Invitrogen) were added with gentle mixing for another 30 min at 4°C. The beads were washed five times with high-salt wash buffer [50 mM HEPES–KOH, pH 7.5, 500 mM KCl, 0.05% (v/v) NP-40, 0.5 mM DTT] and proteins on beads were eluted by incubation with lysis buffer containing 150 ng/ml FLAG peptide. Nine-tenth of the eluates was treated with Proteinase K, and the eluted RNAs were purified with the TRIzol Reagent and analyzed by Northern blotting. The remainder of the eluates was fractionated by SDS-PAGE and analyzed by immunoblotting.

pBS-KS(+)-U96A snoRNA 0.5 μg linearized with XhoI and gel purified was used as a template for in vitro transcription with Riboprobe in vitro Transcription kit (Promega, Madison, WI, USA). The RNA products were treated with DNase I and purified with phenol/chloroform/isoamyl alcohol mixture and ethanol precipitation. The amount of transcribed RNA was quantitated by absorbance at 260 nm and the product size and purity were verified by 10% denaturing acrylamide gel.

To obtain FLAG-tagged proteins, whole-cell lysates from HEK293T cells were prepared from transfected cells with lysis buffer (20 mM HEPES, pH7.5, 150 mM NaCl, 1% Triton X-100, 1 mM DTT, 1 mM EDTA, 10% glycerol, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 20 μg/ml PMSF, 100 KIU/ml aprotinin). FLAG-tagged proteins were purified from whole-cell lysates by using anti-FLAG-M2 affinity resin, washed with high-salt buffer (20 mM HEPES, pH7.5, 400 mM NaCl, 1% Triton X-100, 1 mM DTT, 1 mM EDTA, 10% glycerol, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 20 μg/ml PMSF, and 100 KIU/ml aprotinin), and eluted with FLAG peptide. Purity of the eluted proteins was verified by SDS-PAGE, followed by Coomassie Brilliant Blue staining.

Gel shift assay was conducted using ~50,000 cpm in vitro transcribed, [³²P]-UTP-labeled U96A snoRNA (~100 nt), U95 snoRNA (~100 nt), or U6 snRNA probe (~120 nt). The probe was heat denatured for 5 min by heating at 65°C and renatured prior to being added to binding reactions containing FLAG–GFP, FLAG–IRS-1, FLAG–15.5K proteins (100 ng). The binding reaction was performed at 4°C for 45 min in 20 μl binding buffer (20 mM HEPES-KOH, pH7.5, 6 mM MgCl₂, 60 mM KCl, 7.5% glycerol, 4 mM DTT, 125 ng/μl yeast tRNA, 50 ng/μl BSA, and 5 U RNase-inhibitor) was added. Bound complexes were resolved on native 5% polyacrylamide gels in 0.25 × TAE buffer. Autoradiography was performed using a FLA-5000 imaging system (Fujifilm Life Sciences, Tokyo, Japan).

DNA content in cells was determined using PicoGreen dsDNA Quantitation Reagent (Invitrogen) according to manufacturer’s procedure.

For the analysis of cell cycle distribution, the cells were starved for 48 h in DMEM plus 0.1% FBS. MEFs were trypsinized and thoroughly dissociated into single cells in FACS buffer [0.1% FBS, 1mM EDTA, 1 mM NaCl in PBS(−)] and passed through a 40 μm cell strainer (BD Falcon). The collected cells were fixed in 70% ice-cold ethanol overnight at −20°C. Fixed cells were subsequently stained with 20 g/ml Propidium Iodide diluted in PBS containing 0.2 mg/ml RNAse A for 30 min at 37°C. Cells were sorted and counted using BD FACSCalibur (BD Biosciences, San Jose, CA, USA) according to the manufacturer’s instruction. All the analyses were performed at least three times on different genetic backgrounds.

HeLa cells and MCF-7 cells grown on coverslips were serum-starved for 16 h. The cells were fixed in 4% paraformaldehyde/PBS for 15 min at room temperature, permeabilized with PBS containing 0.25% Triton X-100 for 5 min at room temperature, blocked with blocking buffer (3% BSA and 0.025% NaN3 in PBS) for 1 h at room temperature. Then, primary antibody against Myc epitope was added and incubated overnight at 4°C. After the samples were washed with PBS, secondary antibody incubation was done for 1 h at room temperature using anti-mouse Alexa Fluor 594. The slides were mounted with Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA). Images were captured using FV500 confocal microscope (Olympus, Tokyo, Japan) and analyzed with Fluoview version 1.4 and Photoshop CS5.

Data are expressed as means ± standard error of mean (SEM). Comparisons between two groups were performed using Student’s t-test.

Our previous studies showed that IRS-1 forms high molecular mass complexes with mRNPs (30). Importantly, IRS-1 but not IRS-2 was incorporated into mRNPs (30). Prior to identifying RNAs in IRS-1 complex, we first examined whether IRS-1 interacts with RNAs in close proximity or not. To this end, we performed an UV CLIP analysis (35, 37). HEK293 cells expressing FLAG–IRS-1 or FLAG–IRS-2 were irradiated with UV to cross-link between proteins and RNAs in living cells. Cells were lysed, and RNAs were partially cleaved by the treatment with low-concentrations of RNase, providing short RNA tags covalently cross-linked to target proteins. FLAG–IRSs were then immunoprecipitated and 5′ ends of cleaved RNAs in the precipitates were radiolabeled with polynucleotide kinase (PNK) and [γ⁻³²P]ATP. The immunoprecipitates were then separated by SDS-PAGE, transferred onto nitrocellulose membrane, and detected by autoradiography and immunoblotting (Figures 1A,B). In samples without UV irradiation, we observed radioactive signals at 180 kDa around the band of IRS-1, indicating non-specific phosphorylation of IRS-1 by PNK in vitro. However, UV irradiation greatly increased the radioactive signals at 180 kDa, indicating that IRS-1 covalently interacted with RNAs. In the CLIP analysis, a portion of radioactive signals of the protein–RNA complexes showing slower electrophoretic mobility are often observed because of covalent interaction with longer RNAs (35). Indeed, we actually observed weak but specific radioactive signals with slower electrophoretic mobility than that of IRS-1 itself (Figure 1A). In contrast to IRS-1, UV irradiation did not increase the radioactive signals around the band of IRS-2 (Figure 1A) suggesting that IRS-2 does not interact with RNAs. These results clearly indicated that IRS-1 is in direct contact with RNAs in cells.

We next sought to identify RNAs that are in direct contact with IRS-1 in cells. For this purpose, we extracted RNAs from the radioactive region around IRS-1–RNA complexes (Figure 1, arrowhead) and amplified them by RT-PCR, followed by sequence analysis. We excluded tags with imperfect (<80%) matches to genomic sequences, and accordingly identified 33 IRS-1-bound RNAs from three experiments, which had an average length of 103 nucleotides (Tables A1 and A2 in Appendix). The resulting set of tags included not only tags mapped to 25 messenger-RNA-encoding genes (Table A1 in Appendix) but also tags mapped to 8 non-coding RNAs including snoRNAs and snRNAs (Table A2 in Appendix). RNA sequences that we determined were mapped not only to exons but also to introns or 3′UTRs. As negative control experiments, we repeated CLIP analysis using cells expressing FLAG, but could not amplify the tags.

Among IRS-1 bound RNAs, we focused on U96A snoRNA (U96A), a member of the Box C/D type snoRNAs, which are well established to function in post-transcriptional chemical modification of rRNA (19). U96A is encoded in the second intron of the Rack1 gene (Figure 2A) (38). We first examined the expression of U96A in MCF-7 cells by Northern-blot analysis and the result showed that U96A was expressed and localized in the nucleus (Figure 2B), which is the same subcellular distribution as canonical box C/D snoRNAs. The interaction of IRS-1 with U96A was confirmed by co-immunoprecipitation assay, followed by Northern-blot analysis using labeled U96A as a probe (Figure 2C). We neither detected U96A in the immunoprecipitates of GFP, IRS-2 nor poly-A binding protein (PABP). In a positive control experiment, U96A was detected in the RNAs co-purified with 15.5K, which is known to primarily interact with Box C/D type snoRNAs as a component of snoRNP (19, 22, 39). Interestingly, U95 snoRNA, which is encoded in the first intron of Rack1 pre-mRNA was not co-immunoprecipitated with IRS-1. The direct interaction between IRS-1 and U96A was also confirmed by in vitro binding assay (Figure 2D).

Because snoRNAs were generally excised from an intron of protein-coding mRNAs coupled with splicing (22, 25), we examined the possibility that IRS-1 also interacts with Rack1 pre-mRNA. RNAs co-immunoprecipitated IRS-1 was subjected to RT-PCR using Rack1 specific primers designed within the first and last exon. As a result, we detected single band with a size around 1400 bases (Figure 2E). By sequencing analysis, The PCR product was proved to be Rack1 pre-mRNA that retains only intron 2. A similar size DNA was amplified using RNAs co-immunoprecipitated with CBP80, a protein that can bind pre-mRNAs (40, 41).

To gain insight into the biological significance of the association of IRS-1 with Rack1 pre-mRNA and U96A, we examined the amount of U96A in embryonic fibroblasts derived from IRS-1-deficient mice (IRS-1^−/− MEFs) and wild-type MEFs (IRS-1^+/+ MEFs). Embryonic fibroblasts prepared from littermate IRS-1^+/+ MEFs and IRS-1^−/− MEFs were cultured under serum-free conditions and total RNAs were normalized to DNA content since cell cycle distribution was the same in IRS-1^+/+ and IRS-1^−/− MEFs (Figure 3A). Northern-blot analysis showed that the U96A levels in IRS-1^−/− MEFs were less than those in WT MEFs (Figure 3B). These results demonstrated that IRS-1 positively regulates the abundance of U96A. It is also important to note that the amounts of both Rack1 mature mRNA and Rack1 pre-mRNA retaining only intron 2 were unchanged, suggesting that the effect of IRS-1 depletion on the reduction of U96A in IRS-1^−/− MEFs is not due to altered Rack1 transcription and splicing (Figures 3C,D).

Small nucleolar RNAs are generally produced concomitantly with transcription/splicing of pre-mRNA in nucleoplasm and translocated to the nucleolus, which is the site of rRNA transcription and processing. Several snoRNAs are also reported to be detected in Cajal bodies, where they mature (20, 42). On the other hand, a part of IRS-1 is reported to be localized in nucleus (43–49). To investigate the possibility that nuclear IRS-1 is linked to snoRNA biogenesis, we performed immunofluorescent staining of NLS-fused Myc–IRS-1 (Myc–IRS-1–NLS) using HeLa cells and MCF-7 cells expressing either GFP–coilin or GFP-15.5K, a molecular component of Cajal bodies and nucleolar. As shown in Figure 3E, Myc–IRS-1–NLS certainly localizes to nucleus and exhibits a punctate staining pattern as well as general staining of the nucleoplasm. We found that Myc–IRS-1–NLS significantly colocalizes with GFP–coilin, suggesting that a function of IRS-1 in the nucleus might be related to snoRNA biogenesis.