Sensitivity of global translation to mTOR inhibition in REN cells depends on the equilibrium between eIF4E and 4E-BP1

Abstract

Initiation is the rate-limiting phase of protein synthesis, controlled by signaling pathways regulating the phosphorylation of translation factors. Initiation has three steps, 43S, 48S and 80S formation. 43S formation is repressed by eIF2α phosphorylation. The subsequent steps, 48S and 80S formation are enabled by growth factors. 48S relies on eIF4E-mediated assembly of eIF4F complex; 4E-BPs competitively displace eIF4E from eIF4F. Two pathways control eIF4F: 1) mTORc1 phosphorylates and inactivates 4E-BPs, leading to eIF4F formation; 2) the Ras-Mnk cascade phosphorylates eIF4E. We show that REN and NCI-H28 mesothelioma cells have constitutive activation of both pathways and maximal translation rate, in the absence of exogenous growth factors. Translation is rapidly abrogated by phosphorylation of eIF2α. Surprisingly, pharmacological inhibition of mTORc1 leads to the complete dephosphorylation of downstream targets, without changes in methionine incorporation. In addition, the combined administration of mTORc1 and MAPK/Mnk inhibitors has no additive effect. The inhibition of both mTORc1 and mTORc2 does not affect the metabolic rate. In spite of this, mTORc1 inhibition reduces eIF4F complex formation, and depresses translocation of TOP mRNAs on polysomes. Downregulation of eIF4E and overexpression of 4E-BP1 induce rapamycin sensitivity, suggesting that disruption of eIF4F complex, due to eIF4E modulation, competes with its recycling to ribosomes. These data suggest the existence of a dynamic equilibrium in which eIF4F is not essential for all mRNAs and is not displaced from translated mRNAs, before recycling to the next.

Figure 1. Translation rate is insensitive to exogenous growth factors but sensitive to ER stress.

(A) REN cells were starved overnight of growth factors. Cells were treated with 50 nM insulin (INS), with 50 µg/ml cycloheximide (CHX) or left untreated and pulsed with ³⁵S-methionine for 45 minutes. Methionine incorporation in newly translated proteins was measured in triplicate. Insulin does not increase translation in REN cells. (B) REN cells were starved overnight of growth factors. Cells were treated with 100 µM arsenite (As), 10 mM dithiothreitol (DTT), 2 µM thapsigargin (Tg) or 10 µg/ml brefeldin A (BrefA) for 45 minutes and total proteins were extracted. Total extracts were analyzed in WB to test eIF2α phosphorylation. Densitometric analysis of phospo-eIF2α normalized to total eIF2α is reported. (C) REN cells were starved overnight of growth factors. Cells were treated as indicated and pulsed with ³⁵S-methionine. Methionine incorporation in newly translated proteins was measured in triplicate. Arsenite, dithiothreitol and thapsigargin block translation in REN cells.

Figure 2. Translation rate is largely insensitive to the efficient inhibition of mTORc1.

(A) Scheme of major linear pathways regulating translation. (B) REN cells were treated for 45 minutes with 50 nM rapamycin (RAPA), 10 µM U0126, 5 µM 4-Amino-5-(4-fluoroanilino)-pyrazolo[3,4-d]pyrimidine (Mnk1 inhib.) as indicated and total proteins were extracted. Total extracts were analyzed by WB to test activity of mTORc1, ERK1/2 and Mnk1 signaling. Rapamycin blocks mTORc1 activity, preventing 4E-BP1 and rpS6 phosphorylation. U0126 inhibits MEK, preventing phosphorylation of downstream targets ERK1/2, Mnk1 and eIF4E. Mnk1 inhibitor 4-Amino-5-(4-fluoroanilino)-pyrazolo[3,4-d]pyrimidine blocks eIF4E phosphorylation. (C) REN cells were treated as indicated and pulsed with ³⁵S-methionine. Methionine incorporation in newly translated proteins was measured in triplicate. mTOR, ERK1/2, or Mnk1 inactivation minimally reduce translation. Even combination of multiple drugs, in order to switch off multiple signaling pathways, has no additive effect on reducing protein synthesis.

Figure 3. PP242 does not affect translation rate.

(A) REN cells were treated with rapamycin or with PP242 and total proteins were extracted. Total extracts were analyzed in WB to test activity of mTORc1 and mTORc2 signaling. PP242 blocks both rpS6 and 4E-BP1 phosphorylation by mTORc1 and Akt phosphorylation by mTORc2. (B) REN cells were treated as indicated and pulsed with ³⁵S-methionine. Global translation was measured in triplicate. mTOR kinase inactivation does not reduce translation.

Figure 4. The cap complex formation is inhibited by mTOR inactivation.

(A) REN cells were treated as indicated and total proteins were incubated with 7-Methyl GTP-Sepharose beads. Input is 10% of the purification. Cap binding proteins were analyzed by WB with anti-4E-BP1 and with anti-eIF4G. eIF4E shows equal amount of purified proteins. Rapamycin treatment induces 4E-BP1 binding to eIF4E and eIF4G is displaced from the complex. Densitometric analysis of eIF4G levels normalized to eIF4E are reported. (B) Densitometric analysis of eIF4G levels normalized to eIF4E, from 5 independent experiments are reported. (C) REN cells were treated as indicated and total proteins were incubated with 7-Methyl GTP-Sepharose beads. Input is 5% of the purification. Cap binding proteins were analyzed by WB with anti-4E-BP1 and with anti-eIF4G. eIF4E shows equal amount of purified proteins.

Figure 5. TOP mRNA translation is inhibited by mTOR inactivation.

REN cells were treated as indicated and total extracts were loaded on sucrose gradient. Polysomal profiles were performed, fractions from the gradient were collected and RNA was precipitated. Northern Blot analysis of rpL7 (TOP) and β-Actin mRNAs were performed. Polysomal profiles confirm that blocking mTOR and MAPK signaling has no effect on global translation. Rapamycin and PP242 displace TOP messengers from polysomes. PMP indicates the percentage of rpL7 mRNA associated to polysomes.

Figure 6. 4E-BP1 overexpression or eIF4E reduction restore rapamycin sensitivity.

(A) REN cells were infected with myc tagged-4E-BP1 WT, LM/AA (unable to bind eIF4E) or with control adenovirus with the indicated Multiple of Infection (m.o.i.). WT and LM/AA HA-4E-BP1 expression was confirmed by WB analysis of the myc tag. Exogenous 4E-BP1 is expressed as the same level as the endogenous protein. Serine 65 phosphorilation of 4E-BP1 is reduced by rapamycin treatment, as expected. (B) Infected cells were treated with rapamycin or left untreated and pulsed with ³⁵S-methionine. Methionine incorporation in newly translated proteins was measured in triplicate. WT 4E-BP1 overexpression restores the rapamycin sensitivity in REN cells. (C) REN cells were transfected with eIF4E siRNA or control. After 48 hrs, total protein were analyzed by WB in order to measure eIF4E downregulation. Densitometric analysis of eIF4E level normalized to β-Actin is reported. (D) REN cells were transfected with eIF4E siRNA or control. Cells were treated with rapamycin and the rate of translation was measured. Rapamycin impairs translation where eIF4E level is reduced.

Figure 7. Rapamycin treatment has no effect on global protein synthesis in two epithelial malignant mesothelioma cell lines.

REN (epithelial malignant mesothelioma), NCI-H28 (epithelial malignant mesothelioma), mesMM98 (sarcomatoid malignant mesothelioma) and ME-180 (cervical carcinoma) cells were treated with rapamycin and pulsed with ³⁵S-methionine. Global translation was measured in triplicate, normalized on total protein content. The level of translation of untreated cells is set to 100%. Rapamycin reduces translation in mesMM98 and ME-180 cell lines, but it is ineffective on REN and NCI-H28.

Figure 8. In malignant mesothelioma cell line NCI-H28, mTOR inhibition blocks the formation of the cap complex, without affecting the ongoing translation.

(A) NCI-H28 cells were treated with 2 µM thapsigargin (Tg) and total proteins were extracted. Extracts were analyzed by WB to test eIF2α phosphorylation. (B) NCI-H28 cells were treated as indicated and for 45 minutes and pulsed with ³⁵S-methionine. Methionine incorporation in newly translated proteins was measured in triplicate. Thapsigargin blocks translation in NCI-H28 cells (C) Cells were treated for 45 minutes with 50 nM rapamycin (RAPA), 10 µM U0126, 5 µM 4-Amino-5-(4-fluoroanilino)-pyrazolo[3,4-d]pyrimidine (Mnk1 inhib.) as indicated and total proteins were extracted. Total extracts were analyzed by WB to test activity of mTORc1, ERK1/2 and Mnk1 signaling. Rapamycin blocks mTORc1 activity. (D) NCI-H28 cells were treated as indicated and total proteins were incubated with 7-Methyl GTP-Sepharose beads. Input is 10% of the purification. Cap binding proteins were analyzed by WB with anti-4E-BP1 and with anti-eIF4G. eIF4E shows equal amount of purified proteins. (E) Cells were treated as indicated and pulsed with ³⁵S-methionine. Methionine incorporation in newly translated proteins was measured in triplicate. mTORc1 inactivation does not reduce translation. (F) Cells were treated with rapamycin or with PP242 and total proteins were extracted. Extracts were analyzed by WB to test activity of mTORc1 and mTORc2 signaling. PP242 blocks both rpS6 and 4E-BP1 phosphorylation by mTORc1 and Akt phosphorylation by mTORc2. (G) Cells were treated as indicated and pulsed with ³⁵S-methionine. Global translation was measured in triplicate. mTOR kinase inactivation reduces translation only at 1 µM concentration.