Differential Regulation of the Three Eukaryotic mRNA Translation Initiation Factor (eIF) 4Gs by the Proteasome

Abstract

The 4G family of eukaryotic mRNA translation initiation factors is composed of three members (eIF4GI, eIF4GII, and DAP5). Their specific roles in translation initiation are under intense investigations, but how their respective intracellular amounts are controlled remains poorly understood. Here we show that eIF4GI and eIF4GII exhibit much shorter half-lives than that of DAP5. Both eIF4GI and eIF4GII proteins, but not DAP5, contain computer-predicted PEST motifs in their N-termini conserved across the animal kingdom. They are both sensitive to degradation by the proteasome. Under normal conditions, eIF4GI and eIF4GII are protected from proteasomal destruction through binding to the detoxifying enzyme NQO1 [NAD(P)H:quinone oxidoreductase]. However, when cells are exposed to oxidative stress both eIF4GI and eIF4GII, but not DAP5, are degraded by the proteasome in an N-terminal-dependent manner, and cell viability is more compromised upon silencing of DAP5. These findings indicate that the three eIF4G proteins are differentially regulated by the proteasome and that persistent DAP5 plays a role in cell survival upon oxidative stress.

Keywords: DAP5; NQO1; NRF2; PEST; eIF4G; mRNA translation; oxidative stress; proteasome.

FIGURE 1

Differential half-lives of eIF4GI, eIF4GII and DAP5. (A) NIH-3T3 cells were treated at different times with 50 μg mL^-1 cycloheximide (CHX) and incubated with 10 μg mL^-1 puromycin (puro) 10 min before lysis. Proteins were visualized by western-blotting with the indicated antibodies (left and middle) and signals were quantified by densitometric analysis (right). Data are the means ± SD of three separate experiments. (B) NIH-3T3 cells were treated at different times with 20 μM MG-132 and proteins visualized by western-blotting as indicated. (C) NIH-3T3 cells were co-treated at different times with 50 μg mL^-1 CHX and 20 mM MG-132 and proteins visualized by western-blotting as indicated.

FIGURE 2

Identification of PEST regulatory motifs in eIF4GI and eIF4GII N-termini. (A) Schematic representation of eIF4G protein family members highlighting the binding domains with their main partners (colored boxes) and N-terminal third and C-terminal two thirds of eIF4GI and eIF4GII proteins (brackets) (top). Schematic representation of HA-tagged (black boxes) full-length and deletion fragments used for transient transfections. The computer-predicted (ePEST-find, EMBOSS) PEST motifs detected in eIF4GI and eIF4GII are numbered 1–5 (middle). Scores and locations of the five PESTs found in eIF4GI and eIF4GII polypeptides (bottom). (B) Following transfection with HA-tagged full-length (left) or N-terminal (right) cDNAs, NIH-3T3 cells were untreated or treated at different times with 50 μg mL^-1 CHX or MG-132 and proteins visualized by western-blotting as indicated. (C) Following transfection with HA-eIF4GII (top), HA-DAP5 (middle), or HA-eIF4GII-C (bottom) cDNAs, NIH-3T3 cells were untreated or treated at different times with 50 μg mL^-1 CHX and proteins visualized by western-blotting as indicated.

FIGURE 3

eIF4GI and eIF4GII, but not DAP5, are degraded under oxidative stress. (A) NIH-3T3 cells were untreated or treated with increasing concentration of H₂O₂ in the presence or absence of lactacystin (described in Alard et al., 2009), and protein extracts were subjected to western-blotting as indicated. (B) NIH-3T3 cell extracts were subjected to western-blotting with the indicated antibodies either directly (input) or after immunoprecipitation (IP) with either eIF4GI or eIF4GII antibodies (left). NIH-3T3 extracts of cells either untransfected of transfected with NQO1 cDNA were subjected to western-blotting with the indicated antibodies either directly (input) or after immunoprecipitation (IP) with NQO1 antibodies (right). (C) NIH-3T3 cells were untreated or treated with 300 μM dicumarol (Dic) at different times and proteins were visualized by western-blotting as indicated. (D) Following transfection with HA-tagged, full-length or N-terminal cDNAs, NIH-3T3 cells were untreated or treated with 300 μM dicumarol for 8 h and proteins visualized by western-blotting as indicated.

FIGURE 4

DAP5 is involved in cell survival under oxidative stress. (A) NIH-3T3 cells stably transfected with non-specific (NS) or DAP5-specific (sh1-DAP5 or sh2-DAP5) inducible shRNAs were untreated or treated with doxycycline for 48 h and protein extracts were subjected to western-blotting with the indicated antibodies. (B) The survival of stably transfected NIH-3T3 cells in the absence of Dox but treated for 24 h with H₂O₂ was monitored as a function of H₂O₂ concentration. Data are the means ± SD of three separate experiments and normalized (%) to the values obtained with H₂O₂-untreated cells. (C) The survival of stably transfected NIH-3T3 cells incubated with Dox for 48 h and treated for 16 or 24 h with H₂O₂ was monitored as a function of H₂O₂ concentration. Data are the means ± SD of five separate experiment and normalized (%) to the values obtained with H₂O₂-untreated cells.

FIGURE 5

Induction of NRF2 and NQO1 proteins under oxidative stress is independent of DAP5. Protein extracts of stably transfected NIH-3T3 cells grown in the absence or presence of doxycycline (Dox) for 48 h and untreated or treated with 1 mM H₂O₂ for 4 h were subjected to western-blotting with the indicated antibodies. The bottom-to-top α–β–γ symbols denote hypo- to hyperphosphorylated 4E-BP1 isoforms.

FIGURE 6

Conservation of N-terminal eIF4Gs’ PEST motifs among different branches of the animal kingdom. The ePESTfind software was run out for eIF4GI, eIF4GII, or DAP5 proteins from representative species of main branches of the animal kingdom. The color code is as in Figure 1 except that PEST motifs are not numbered. Note that for convenience, the colored boxes corresponding to the binding domains of PAPB, eIF4E, eIF4A, eIF3, and MNK1/2 validated in the mammalian protein sequences have been copy/pasted in protein sequences of other species even if they have not been always experimentally confirmed.