eIF4E is a central node of an RNA regulon that governs cellular proliferation - PubMed (original) (raw)

eIF4E is a central node of an RNA regulon that governs cellular proliferation

Biljana Culjkovic et al. J Cell Biol. 2006.

Abstract

This study demonstrates that the eukaryotic translation initiation factor eIF4E is a critical node in an RNA regulon that impacts nearly every stage of cell cycle progression. Specifically, eIF4E coordinately promotes the messenger RNA (mRNA) export of several genes involved in the cell cycle. A common feature of these mRNAs is a structurally conserved, approximately 50-nucleotide element in the 3' untranslated region denoted as an eIF4E sensitivity element. This element is sufficient for localization of capped mRNAs to eIF4E nuclear bodies, formation of eIF4E-specific ribonucleoproteins in the nucleus, and eIF4E-dependent mRNA export. The roles of eIF4E in translation and mRNA export are distinct, as they rely on different mRNA elements. Furthermore, eIF4E-dependent mRNA export is independent of ongoing RNA or protein synthesis. Unlike the NXF1-mediated export of bulk mRNAs, eIF4E-dependent mRNA export is CRM1 dependent. Finally, the growth-suppressive promyelocytic leukemia protein (PML) inhibits this RNA regulon. These data provide novel perspectives into the proliferative and oncogenic properties of eIF4E.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

Enhanced mRNA export corresponds to elevated protein levels of eIF4E-sensitive targets. (A) Relative fold difference of mRNAs bound to nuclear eIF4E. mRNAs were immunoprecipitated from untreated nuclear lysates or those treated with 50 μM m7GpppG or GpppG. Values represent relative fold ± SD (error bars; normalized against untreated immunoprecipitated IgG, which was set to 1). Calculations of fold were performed using the relative standard curve method (User Bulletin #2 ABI Prism 7700; Applied Biosystems). Relative amounts of the target mRNA = 10[C (t)-b]/a were determined for each PCR reaction. Average values ± SD were calculated for each set of triplicates. Average values obtained for the IPs (i.e., average relative amount of immunoprecipitated target mRNA) were divided by values obtained for 5% nuclear input (i.e., average relative amount of target mRNA present in the 5% of the amount of nuclear extract used for IP). Obtained values ± SD (i.e., average IP/average 5% nuclear) were normalized by setting the untreated IgG IP value to 1. (B and C) eIF4E-enhanced mRNA transport leads to up-regulated protein levels of corresponding mRNAs. Total cell lysates from U937 (B) or NIH3T3 (C) cells transfected as indicated were analyzed for protein content by Western methods. Note that in C, where human PML was overexpressed, the 5E10 mAb PML antibody only recognizes the human PML, not the endogenous mouse PML.

Figure 2.

Figure 2.

A common secondary structure for the 4E-SE that acts as a zip code for eIF4E nuclear bodies. (A) Secondary structure for cyclin D1 4E-SE (c4E-SE) and Pim-1 4E-SE (p4E-SE) as determined by RNase mapping experiments. Conserved set of A and U nucleotides (UX2UX2A) are highlighted in yellow. (B) A sample gel. (C, top) Only one of the two predicted SLPs from Pim-1 3′ UTR (p4E-SE) immunoprecipitates with eIF4E. (bottom) eIF4E promotes the export of lacZ mRNA that contains minimal p4E-SE. Cytoplasmic/nuclear (c/n) values represent relative fold ± SD (error bars) normalized to the lacZ control, which was set to 1. Calculations were performed as described in Table II. (D) Colocalization of lacZ–p4E-SE, lacZ–c4E-SE, or lacZ transcripts with PML and eIF4E protein was examined in U2OS cells transfected with lacZ/lacZ–4E-SE. LacZ mRNA was detected using in situ hybridization with a biotin-labeled nick-translated probe to lacZ (red). Cells were then immunostained using an eIF4E mAb conjugated directly to FITC (green) and PML mAb 5E10 (blue). Importantly, lacZ mRNAs containing the 4E-SE from either cyclin D1 or Pim-1 colocalize to eIF4E nuclear bodies (arrowheads). As we have shown previously for endogenous cyclin D1 mRNA, there are two populations of eIF4E nuclear bodies: those that colocalize with lacZ mRNA and those that colocalize with PML. Magnification was 100× and 3× (for lacZ and c4ESE) or 4× (p4ESE) digital zoom. Bars, 10 μM.

Figure 3.

Figure 3.

The 4E-SE is required for the formation of eIF4E-dependent complexes. (A and B) EMSA analysis indicates that lacZ transcripts, which contain either the cyclin D1 4E-SE (c4E-SE) or the Pim-1 4E-SE (p4E-SE), formed high molecular weight complexes in the presence of nuclear lysates (N). LacZ transcripts (control) without the 4E-SE did not form these complexes. The addition of purified mouse eIF4E with a 6-kD solubility tag (m4E) or untagged human eIF4E (h4E) causes partial shifts relative to shifts observed with nuclear lysate. With nuclear lysates immunodepleted of eIF4E (dpl N), gel shifts were not observed. Complexes could be supershifted by an anti-eIF4E antibody (N + α4E). Arrows indicate protein–RNA complexes. (C) Mutation of Pim-1 4E-SE (p4E-SE) reduces the efficacy of the gel shift. No molecular weight markers are shown, as A–C and E are native gels. (D) UV cross-linking experiments showed the formation of specific complexes in the 75–90-kD mass range (indicated by arrows). These complexes are specifically depleted in the presence of excess m7GpppG cap (cap) or if lysates are immunodepleted of eIF4E (dpl N). Asterisks indicate complexes that are cap and 4E-SE independent. (E) The addition of unlabeled p4E-SE ribooligonucleotide (cold probe) to the p4E-SE nuclear complexes indicates that this element can efficiently compete for complex formation. All transcripts were capped and 3′ end labeled.

Figure 4.

Figure 4.

Export of 4E-SE–containing mRNAs is independent of ongoing RNA and protein synthesis, and the pathway is saturated by excess 4E-SE. (A) Quantitative real-time PCR analysis of mRNA export of lacZ–c4E-SE and lacZ in eIF4E-overexpressing cells. Cytoplasmic/nuclear (c/n) ratios represent relative fold ± SD (error bars) normalized to the lacZ untreated control, which was set to 1. Calculations were performed as described in Table II. Treatments were 10 μg/ml actinomycin D for 1 h and 100 μg/ml cycloheximide (CHX) for 1 h. (B and C) LacZ mRNA export was monitored as a function of both time and expression of lacZ transcripts ± 4E-SE induced with doxycycline. Expression as a function of time is shown. In parallel, the extent of export was monitored as the ratio of cytoplasmic/nuclear mRNA for each case. Solid lines represent trends in cells expressing lacZ–c4E-SE; dotted lines represent cells expressing lacZ–p4E-SE. Endogenous mRNAs from the same samples were also examined. Cyclin D1 mRNA export was reduced in cells expressing either lacZ–c4E-SE or lacZ–p4E-SE. Importantly, VEGF, which does not contain a 4E-SE, did not have its export affected in either case. Clearly, as the amount of 4E-SE–containing mRNAs increases in the cell (C), the ability to export these is reduced presumably because 4E-SE–dependent export becomes saturated (B). Cytoplasmic/nuclear values represent relative fold ± SD normalized to lacZ only for each time point as described in Table II. For total RNAs, values represent relative fold ± SD normalized to the first time point of induction for each transcript (4 h), which was set to 1. Average values of lacZ mRNA obtained for each time point were normalized by GAPDH mRNA values obtained for the same sample.

Figure 5.

Figure 5.

eIF4E-dependent export is NXF1 independent and CRM1 dependent. (A) Comparison of lacZ mRNA in the NXF1 IP fractions. Cells were cotransfected with FlagNXF1/Flagp15 and lacZ or lacZ–c4E-SE. IPs were performed with anti-Flag antibody. LacZ/lacZ–c4E-SE mRNA was monitored by real-time PCR and normalized to IgG controls (as described in Fig. 1 A). (B) NXF1 siRNA treatment (72 h) inhibits the export of lacZ but not lacZ–c4E-SE–containing mRNAs. The cytoplasmic/nuclear (C/N) ratios of lacZ or lacZ–c4E-SE mRNAs in cells overexpressing eIF4E as a function of siRNA treatment are shown. LacZ mRNA levels were normalized to 18S rRNA, whose cytoplasmic/nuclear ratio is unaffected by NXF1 siRNA. (C) Western blot (WB) analysis indicates that NXF1 protein levels are reduced by siRNA treatment but not by scrambled controls (DS (-control)). A Western blot for eIF4G is shown as a negative control. LacZ protein levels correspond to alterations in mRNA export shown in B. (D) Dependence of c4E-SE export on leptomycin B (LMB). The cytoplasmic/nuclear ratio of lacZ–c4E-SE mRNA in eIF4E-overexpressing U2OS cells indicated that 4E-SE export was sensitive to LMB (10 ng/ml for 4 h), whereas lacZ was not sensitive. 18S rRNA export was inhibited by LMB as expected, and β-actin mRNA export was consistently not affected. All RNAs were normalized to GAPDH mRNA (as described in Table II). (B and D) Cytoplasmic/nuclear lysate values represent relative fold ± SD (error bars) normalized to the lacZ untreated control, which was set to 1.

Figure 6.

Figure 6.

Schematic representation of mechanisms for the export of different classes of RNA. Overviews of characteristic features delineating the export of mRNAs via CRM1 or NXF1/p15 pathways are shown together with features of the eIF4E-mediated export of mRNAs.

Similar articles

Cited by

References

    1. Bachmann, M., and T. Moroy. 2005. The serine/threonine kinase Pim-1. Int. J. Biochem. Cell Biol. 37:726–730. - PubMed
    1. Borden, K.L. 2002. Pondering the promyelocytic leukemia protein (PML) puzzle: possible functions for PML nuclear bodies. Mol. Cell. Biol. 22:5259–5269. - PMC - PubMed
    1. Cao, Q., and J.D. Richter. 2002. Dissolution of the maskin-eIF4E complex by cytoplasmic polyadenylation and poly(A)-binding protein controls cyclin B1 mRNA translation and oocyte maturation. EMBO J. 21:3852–3862. - PMC - PubMed
    1. Chen, Y.C., Y.N. Su, P.C. Chou, W.C. Chiang, M.C. Chang, L.S. Wang, S.C. Teng, and K.J. Wu. 2005. Overexpression of NBS1 contributes to transformation through the activation of phosphatidylinositol 3-kinase/Akt. J. Biol. Chem. 280:32505–32511. - PubMed
    1. Clemens, M.J., and U.A. Bommer. 1999. Translational control: the cancer connection. Int. J. Biochem. Cell Biol. 31:1–23. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources