Kubacka D et al. (2013), Investigating the consequences of eIF4E2 (4EHP)...

XB-ART-49901

PLoS One 2013 Jan 01;88:e72761. doi: 10.1371/journal.pone.0072761.

Show Gene links Show Anatomy links

Investigating the consequences of eIF4E2 (4EHP) interaction with 4E-transporter on its cellular distribution in HeLa cells.

Kubacka D , Kamenska A , Broomhead H , Minshall N , Darzynkiewicz E , Standart N .

???displayArticle.abstract???
In addition to the canonical eIF4E cap-binding protein, eukaryotes have evolved sequence-related variants with distinct features, some of which have been shown to negatively regulate translation of particular mRNAs, but which remain poorly characterised. Mammalian eIF4E proteins have been divided into three classes, with class I representing the canonical cap-binding protein eIF4E1. eIF4E1 binds eIF4G to initiate translation, and other eIF4E-binding proteins such as 4E-BPs and 4E-T prevent this interaction by binding eIF4E1 with the same consensus sequence YX 4Lϕ. We investigate here the interaction of human eIF4E2 (4EHP), a class II eIF4E protein, which binds the cap weakly, with eIF4E-transporter protein, 4E-T. We first show that ratios of eIF4E1:4E-T range from 50:1 to 15:1 in HeLa and HEK293 cells respectively, while those of eIF4E2:4E-T vary from 6:1 to 3:1. We next provide evidence that eIF4E2 binds 4E-T in the yeast two hybrid assay, as well as in pull-down assays and by recruitment to P-bodies in mammalian cells. We also show that while both eIF4E1 and eIF4E2 bind 4E-T via the canonical YX 4Lϕ sequence, nearby downstream sequences also influence eIF4E:4E-T interactions. Indirect immunofluorescence was used to demonstrate that eIF4E2, normally homogeneously localised in the cytoplasm, does not redistribute to stress granules in arsenite-treated cells, nor to P-bodies in Actinomycin D-treated cells, in contrast to eIF4E1. Moreover, eIF4E2 shuttles through nuclei in a Crm1-dependent manner, but in an 4E-T-independent manner, also unlike eIF4E1. Altogether we conclude that while both cap-binding proteins interact with 4E-T, and can be recruited by 4E-T to P-bodies, eIF4E2 functions are likely to be distinct from those of eIF4E1, both in the cytoplasm and nucleus, further extending our understanding of mammalian class I and II cap-binding proteins.

???displayArticle.pubmedLink??? 23991149
???displayArticle.pmcLink??? PMC3749138
???displayArticle.link??? PLoS One
???displayArticle.grants??? [+]

Species referenced: Xenopus
Genes referenced: eif4e eif4e2 eif4g1 gigyf2 inhbc.1 MGC89871 patl1 pknox1 tial1 xpo1

???attribute.lit??? ???displayArticles.show???

	Figure 2. eIF4E2 is recruited to P-bodies by ectopic 4E-T.A. Cellular eIF4E2 distribution in HeLa cells was assessed by indirect immunofluorescence in the absence and presence of GFP-4E-T, and similarly for HA-tagged eIF4E2 co-transfected or not with GFP-4E-T. B. Wild-type but not mutant GFP-4E-T recruits eIF4E2 to P-bodies. Mutations in the 4E-T eIF4E-binding site residues Y30X4LL to alanine are indicated. Cells were also stained with DAPI. Scale bar, 10 µm.
	Figure 3. The eIF4E-binding site in 4E-T.A. Top. The sequences of the eIF4E2-binding site in Dm Bicoid [17], Mm Prep1 [18], Hs GIGYF2 [19] and Hs 4E-T ( [20]. Bottom. Alignments of five indicated vertebrate 4E-T sequences using ClustalW2. In bold are indicated the eIF4E-binding site residues, note the upstream tyrosine in Dr 4ET, bold and underlined. Underlined are the residues of a potential conserved second eIF4E-binding site. B. HEK293 cells were transfected with wild type or mutant GFP-4E-T as indicated and lysates were analysed by GFP-Trap and western blotting with eIF4E1, eIF4E2 and 4E-T antibodies. C. mRNAs encoding FLAG-MS2 and FLAG-MS2-X4E-T proteins were injected into Xenopus oocytes. After 16 hours, lysates were prepared and immunoprecipitated with FLAG antibodies, and the FLAG peptide eluates were analysed by western blotting with indicated antibodies. L lane indicates uninjected oocyte lysate. The mutations in human and Xenopus 4E-T sequences are indicated in B and C.
	Figure 4. eIF4E1 and eIF4E2 distribution in arsenite-treated cells.Control HeLa cells (- Ars) and cells treated with arsenite (+ Ars) to induce stress granules were immunostained with eIF4E1 (A), eIF4E2 (B) and 4E-T (C) antibodies, and counter-stained with antibodies against p54/rck (P-body marker), and TIAR or G3BP (stress granule markers) as indicated. Scale bar, 10 µm.
	Figure 5. eIF4E1 and eIF4E2 distribution in Actinomycin D-treated cells.Control HeLa cells (- ActD) and cells treated with Actinomycin D (+ ActD) to inhibit RNA synthesis were immunostained with eIF4E1 (A) and eIF4E2 (B) antibodies, and counter-stained with Pat1b antibodies. Note that the apparent nuclear staining in the case of Pat1b antibody was previously shown to be insensitive to Pat1b siRNA [49]. Cells were also stained with DAPI. Scale bar, 10 µm.
	Figure 6. eIF4E2 is a nucleocytoplasmic shuttling protein.Hela cells were treated with Leptomycin B for 5 hrs (+ LMB) or with methanol as control (- LMB), without (A and C) and with ectopic GFP-4E-T (B and D), and immunostained with eIF4E1 (A and B), eIF4E2 (C and D) and 4E-T antibodies, while GFP-4E-T fluorescence was detected directly in transfected cells. Scale bar, 10 µm.
	Figure 7. eIF4E2 shuttling in the presence of LMB is 4E-T independent.Untransfected Hela cells (no siRNA), or cells transfected with 4E-T siRNA or control β-globin siRNA were treated with LMB and immunostained with eIF4E2 and 4E-T antibodies as indicated. Cells were also stained with DAPI. Scale bar, 10 µm.
	Figure 8. The leucine-rich C-terminus of eIF4E2 does not act as a NES.A. Sequence alignment of human eIF4E1 and eIF4E2 performed with ClustalW2. Secondary structural elements of α-helices (H0-H4) and β-strands (S1-S7) were shown above the alignment according to the crystal structure of human eIF4E2 in complex with m7 GTP (PDB id: 2JGB) [15]. Identities are shaded in grey. With blue and red are marked hydrophobic residues that typically form the NES. The leucine-rich C-terminus of eIF4E2 is underlined in red. B. ClustalW alignment of the C-terminal regions of indicated eIF4E2 proteins. C. HeLa cells transfected with HA-eIF4E2 full-length protein or HA-eIF4E2ΔC (truncated at residue 222) were treated with LMB or methanol (-) for 5 hrs and immunostained with HA and 4E-T antibodies as indicated. Cells were also stained with DAPI. Scale bar, 10 µm.
	Figure 1. Quantitation of eIF4E1, eIF4E2 and 4E-T levels. A. Verification of chicken eIF4EL3/eIF4E2 antibody using untransfected Hela cells (1), and cells overexpressing GFP-eIF4E2 (2) and HA-eIF4E2 (3). Top panel developed with eIF4E2 antibody, bottom panels with eIF4E1 and 4E-T antibodies. Molecular weight standards are indicated in kDa. B. Comparison of eIF4E1, eIF4E2 and 4E-T levels in HeLa and HEK293 cells analysed by western blotting with indicated antibodies. Increasing amounts of lysate from HeLa cells (20, 40 and 80 µg total protein) and HEK293 cells (10, 20 and 40 µg) approximating to 5.7, 11.4, 17 and 23 x 104 cells. C. Table comparing eIF4E1, eIF4E2 and 4E-T levels (protein copies/per cell) in HeLa and HEK293 determined by western blotting (top), to those reported in genome-wide mass spectrometry studies in HeLa [35], U20S [34] and NIH3T3 [36] cells.

References [+] :

Aitken, A mechanistic overview of translation initiation in eukaryotes. 2012, Pubmed