The Splicing Factor SF2/ASF Regulates Translation Initiation by Enhancing Phosphorylation of 4E-BP1  Gracjan Michlewski, Jeremy R. Sanford, Javier F. Cáceres  Molecular Cell  Volume 30, Issue 2, Pages 179-189 (April 2008) DOI: 10.1016/j.molcel.2008.03.013 Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 1 SF2/ASF Stimulates Translation Initiation (A and B) RRLs were programmed with radiolabeled transcripts harboring or lacking six binding sites for SF2/ASF (mini-EDA and mini-MT, respectively) in the presence of CHX. Addition of exogenous m7G-cap analog (specific blocker of initiation step) serves as a negative control of translation. The translation reactions were loaded on 5%–25% linear sucrose gradients, and complexes were resolved by centrifugation. The position of ribosomal complexes and the mRNP fraction is indicated. (C) Reactions were performed as in (B) but with addition of GMP-PNP, a GTP analog that inhibits the joining of the large ribosomal subunit (60S) to the 48S preinitiation complex positioned at the initiation codon. The position of the 48S complex and the mRNP fraction is indicated. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 2 Recombinant SF2/ASF Does Not Stimulate Translation of CSFV or PV Reporters but Stimulates an HAV Reporter Harboring SF2/ASF-Binding Sites (A) Schematic diagram of the luciferase reporters with IRES elements inserted in the 5′UTR region. On the right-hand side, the initiation factors that are dispensable for translation initiation mediated by the respective IRES element are listed. (B) Reporter mRNAs (200 ng) were incubated in a HeLa cell translation extract. Following incubation at 37°C, luciferase assays were performed. The fold activation of luciferase activity mediated by addition of recombinant SF2/ASF (black bars) was plotted relative to reactions without additional SF2/ASF (white bars) whose values were set to one. Mean values and standard deviations of at least three independent experiments are shown. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 3 SF2/ASF-Mediated Stimulation of Translation Is Cap Dependent (A) Schematic diagram of the luciferase reporters. Six SF2/ASF-binding sites (6xEDA) or their mutant counterparts (6xEDAmt) were inserted in the 3′UTR of ApG or m7G-capped mRNAs. (B) Reporter mRNAs (200 ng) 6xEDAmt (white bars) or 6xEDA (gray bars) capped with ApG or m7G, respectively, were incubated in a HeLa cell translation extract. Following incubation at 37°C, luciferase assays were performed. The fold activation of luciferase activity derived from transcripts with 6xEDA relative to 6xEDAmt was plotted with 6xEDAmt values set to one. (C) Recombinant SF2/ASF stimulates expression of the m7G-capped reporter mRNA harboring SF2/ASF-binding sites. Translation reactions were performed in the presence (black bars) or absence (white bars) of recombinant SF2/ASF protein (200 ng), as described above. Mean values and standard deviations of at least three independent experiments are shown. (D) Recombinant SF2/ASF can restore translation inhibited by addition of an exogenous m7G-cap analog. The fold activation or inhibition of luciferase activity derived from m7G 6xEDA transcript (white bar) relative to reactions carried out with addition of exogenous ApG-cap (diagonal stripped bars), m7G-cap (vertical stripped bars), or in the presence of increasing amounts of recombinant SF2/ASF (gray and black bars) was plotted, with the control 6xEDA transcript reaction values set to one. Mean values and standard deviations of at least three independent experiments are shown. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 4 Addition of SF2/ASF to HeLa Cell Extracts Decreases the Rate of 4E-BP Dephosphorylation (A) Status of 4E-BP phosphorylation in HeLa cell extracts was analyzed at different time points with or without addition of recombinant SF2/ASF or hnRNP A1 (as a control) proteins. Following 60 min incubation of cell extracts alone or with addition of hnRNP A1, 4E-BP1 was greatly dephosphorylated. Upon addition of SF2/ASF, the rate of 4E-BP1 dephosphorylation was substantially reduced, whereas the levels of 4E-BP1 total protein remained unchanged. (B) A similar experiment but with the addition of 40 nM rapamycin, a specific inhibitor of mTOR kinase. (C) Reporter mRNAs (200 ng) harboring (6xEDA, black bars) or lacking (6xEDAmt, white bars) SF2/ASF-binding sites and capped with m7G cap were incubated in HeLa cell translation extracts. Where indicated, rapamycin (40 nM) and recombinant SF2/ASF protein (200 ng) were added to the reaction. Following incubation at 37°C, luciferase assays were performed. Mean values and standard deviations of at least three independent experiments are shown. (D) Recombinant, mTOR kinase stimulates translation of the m7G-capped reporter mRNA harboring SF2/ASF-binding sites to a far greater extent than the control reporter mRNA lacking SF2/ASF-binding sites. Translation reactions were performed in the presence (black bars) or absence (white bars) of the C-terminal fragment of recombinant mTOR kinase (200 ng). The fold activation of luciferase activity in the presence of mTOR (black bars) was plotted relative to reactions without mTOR (white bars), whose values were set to one. Mean values and standard deviations of at least three independent experiments are shown. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 5 Depletion of 4E-BP1 in HeLa Cells Abrogates the SF2/ASF Stimulatory Effect on Translation Regulation (A) Schematic diagram of the pLCS reporter harboring one SF2/ASF-binding site or a mutant version inserted in-frame and upstream of the firefly luciferase open reading frame. (B) HeLa cells were transfected with a smart pool of siRNAs against 4E-BP1 (50 nM), 4E-BP2 (50 nM), or both. A pool of siRNAs against GAPDH (50 nM) or fluorescently labeled siRNA (50 nM) served as RNAi controls. Forty-eight hours after transfection, the levels of the corresponding proteins were assayed by western blots. (C) Downregulation of 4E-BP1, but not of 4E-BP2, abrogates the SF2/ASF stimulatory effect on translation. To maintain low levels of 4E-BP proteins, cells were retransfected with the corresponding siRNAs, together with the luciferase reporters and a TK renilla luciferase expression vector, serving as an internal control. Additionally, cells were transfected with plasmid carrying T7-epitope-tagged SF2/ASF or empty plasmid. Mean values and standard deviations of at least three independent experiments are shown. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 6 SF2/ASF Binds Both mTOR Kinase and the Catalytic Subunit of PP2A Phophatase through Interactions Mediated by Its Second RRM (A) Extracts prepared from 293T cells were incubated with anti-PP2Ac antibody bound to Sepharose beads (Novagen). The bound proteins were separated on a 10% SDS-polyacrylamide gel and analyzed by western blotting with anti-SF2/ASF antibody. Alternatively, the immunoprecipitate was treated with RNases A/T1 before loading on the gel. Lane 1 was loaded with 2% of the amount of extract used for each immunoprecipitation. (B) Extracts prepared from 293T cells were incubated with anti-SF2/ASF antibody or preimmune serum bound to Sepharose beads and analyzed as described in (A). The blot was probed with anti-mTOR antibody. Alternatively, the immunoprecipitate was treated with RNases A/T1 before loading on the gel (lane 3). Lane 1 was loaded with 2% of the amount of extract used for each immunoprecipitation. (C) T7 epitope-tagged SF2/ASF and the ΔRS mutant efficiently bind both mTOR and PP2Ac proteins, whereas the AAA (WDK→AAA in RRM2) mutant fails to interact with these proteins. CoIPs with anti T7-agarose were done using extracts from 293T cell that were transfected with T7-tagged control pCG plasmid, wild-type SF2/ASF, and AAA and ΔRS mutants. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions

Figure 7 Two Non-Mutually Exclusive Models for the Activation of Translation Initiation of a Subset of mRNAs Bound by the Shuttling Protein SF2/ASF (A) By recruiting the mTOR kinase to the mRNP complex, SF2/ASF could facilitate phosphorylation of 4E-BP proteins leading to release of eIF4E, therefore allowing more efficient translation initiation. (B) Alternatively, SF2/ASF could inhibit the function of the PP2A phosphatase in the vicinity of mRNP complexes, most likely through an indirect interaction, shifting the kinetics of 4E-BP phosphorylation to a more hyperphosphorylated state and thus activating translation of mRNAs bound by SF2/ASF. Molecular Cell 2008 30, 179-189DOI: (10.1016/j.molcel.2008.03.013) Copyright © 2008 Elsevier Inc. Terms and Conditions