Oxygen sufficiency controls TOP mRNA translation via the TSC-Rheb-mTOR pathway in a 4E-BP-independent manner

Abstract

Cells encountering hypoxic stress conserve resources and energy by downregulating the protein synthesis. Here we demonstrate that one mechanism in this response is the translational repression of TOP mRNAs that encode components of the translational apparatus. This mode of regulation involves TSC and Rheb, as knockout of TSC1 or TSC2 or overexpression of Rheb rescued TOP mRNA translation in oxygen-deprived cells. Stress-induced translational repression of these mRNAs closely correlates with the hypophosphorylated state of 4E-BP, a translational repressor. However, a series of 4E-BP loss- and gain-of-function experiments disprove a cause-and-effect relationship between the phosphorylation status of 4E-BP and the translational repression of TOP mRNAs under oxygen or growth factor deprivation. Furthermore, the repressive effect of anoxia is similar to that attained by the very efficient inhibition of mTOR activity by Torin 1, but much more pronounced than raptor or rictor knockout. Likewise, deficiency of raptor or rictor, even though it mildly downregulated basal translation efficiency of TOP mRNAs, failed to suppress the oxygen-mediated translational activation of TOP mRNAs. Finally, co-knockdown of TIA-1 and TIAR, two RNA-binding proteins previously implicated in translational repression of TOP mRNAs in amino acid-starved cells, failed to relieve TOP mRNA translation under other stress conditions. Thus, the nature of the proximal translational regulator of TOP mRNAs remains elusive.

Keywords: 4E-BP; TOP mRNAs; hypoxia; mTOR; translational control.

Figure 1

TOP mRNAs are translationally repressed by oxygen deprivation in an mTOR-dependent fashion. (A) Typical polysomal profiles and its portioning to polysomal and subpolysomal fractions. HEK293 cells were either untreated (control) or 16 h starved for oxygen (anoxia). The cytoplasmic extract was size fractionated by centrifugation through a sucrose gradient. The tube content was collected from the bottom and the absorbance at 260 nm was recorded. The vertical dashed line separates between the polysomal (P) fraction (left) and the subpolysomal (S) fraction (right). 80, 60, and 40 represent 80S monosomes, 60S, and 40S ribosomal subunits, respectively. (B) HEK293 cells were untreated (control); oxygen-deprived for 6 h (−O₂) in the absence or presence of 20 nM rapamycin or 50 nM Torin 1; and oxygen replenished for 2 h (−O₂ → +O₂) in the absence or presence of 20 nM rapamycin or 50 nM Torin 1. Subsequently, cells were harvested and cytoplasmic extracts were prepared and subjected to polysomal analysis using cDNAs corresponding to rpS6 (a TOP mRNA) and actin (a non-TOP mRNA). (C) HEK293 cells were treated as in B, and their cytoplasmic proteins were subjected to western blot analysis using the indicated antibodies.

Figure 2

The deficiency of TSC2 or TSC1 or overexpression of Rheb can rescue TOP mRNAs from translational repression in oxygen-deprived cells. (A) TSC2^+/+ and TSC2^−/− MEFs were either untreated (control) or oxygen-starved for 16 h (−O₂), harvested and their cytoplasmic extracts were subjected to polysomal analysis. (B) Cytoplasmic proteins from MEFs treated as in A were subjected to western blot analysis. (C) HEK293T cells were transiently cotransfected with expression vectors encoding rpS16-GH or an empty vector (EV) pRK7, Myc-RhebL64 or Myc-Rheb-5A. After 32 h, cells were starved for oxygen (16 h), harvested, and subjected to polysomal analysis with probes directed toward endogenous (rpS6) and exogenous (S16-GH) TOP mRNAs, as well as actin mRNA. (D) Cytoplasmic proteins from cells treated as in C were subjected to western blot analysis.

Figure 3

4E-BP deficiency failed to alleviate the translational repression of TOP mRNAs in oxygen-deprived cells. (A and B) 4E-BP WT and 4E-BP DKO MEFs were either untreated (+) or oxygen starved (−) for 12 h (A), or either untreated (+) or serum starved (−) for 48 h (B). Cytoplasmic proteins from these cells were subjected to western blot analysis with the indicated antibodies. (C) Cells treated as described in A and B were harvested and subjected to polysomal analysis with the indicated probes. (D) 4E-BP DKO MEFs were infected with an empty retroviral vector (EV) pBABE-puro or retroviral vector encoding either wild-type 4E-BP (WT) or 4E-BP^4Ala (4Ala). After selection with puromycin, cells were either untreated or treated with 250 nM Torin 1 for 2 h and their cytoplasmic proteins were subjected to western blot analysis with the indicated antibodies. (E) Cells derived as described in D were harvested and subjected to polysomal analysis with the indicated probes.

Figure 4

Translation of cyclin D3 mRNA, but not those encoding rpL32 and actin, is 4E-BP dependent. (A) Wild-type (WT) and 4E-BP DKO (DKO) MEFs were maintained in 0.5% FBS for 14 h and stimulated with 10% FBS in the absence or presence of 250 nM Torin 1 for 2 h. Levels and the phosphorylation status of the indicated proteins were determined by western blotting. (B−D) WT and DKO MEFs were treated as in A and the cytoplasmic extract was size fractionated by sucrose gradient centrifugation (Supplementary Figure S5 for polysomal profiles). Distribution of rpL32 (B), cyclin D3 (C), and β-actin (D) mRNAs among heavy (H, ≥4 ribosomes) and light (L, 2 to 3 ribosomes) polysomal fractions as well as subpolysomal fraction (S) was monitored by reverse transcriptase-quantitative PCR (RT–qPCR). Values are expressed as a percentage of total mRNA loaded onto the sucrose gradient (input). Data are presented as mean ± SD (two independent biological replicates). Each biological replicate was carried out in three technical replicates.

Figure 5

Hyperphosphorylation of 4E-BP failed to derepress TOP mRNA translation. (A and B) Dicer^+/+ and Dicer^−/− hemangiosarcoma cells were either untreated (+), oxygen starved (−) for 12 h (A) or treated with 50 nM Torin 1 (+) for 3 h (B) and cytoplasmic proteins were subjected to western blot analysis with the indicated antibodies. (C) Cells treated as described in A and B as well as cells deprived of oxygen for 12 h and then resupplied with oxygen for 3 h were harvested and subjected to polysomal analysis. (D and E) Untreated L2 lymphoblastoids were subjected to western blot (D) and polysomal (E) analyses.

Figure 6

Raptor and rictor are dispensable for translational activation of TOP mRNAs by oxygen. (A) iRapKO or iRicKO cells were either untreated (−) or treated (+) with 4HT for 4 days and cytoplasmic proteins were subjected to western blot analysis with the indicated antibodies. The relative abundance of raptor, rictor, and phospho-Akt(Ser473) was normalized to that of actin, whereas phospho-rpS6(Ser240/244) was normalized to rpS6. The results are numerically presented relative to those obtained without 4HT and were arbitrarily set at 1. (B) iRapKO cells were either treated (for 4 days) or untreated with 4HT, and then kept under normoxia (control) in the absence or presence of 20 nM rapamycin or 50 nM Torin 1 for 3 h; deprived of oxygen for 10 h (−O₂) in the absence or presence of 20 nM rapamycin or 50 nM Torin 1 for 3 h; or deprived of oxygen for 10 h and then subjected to oxygen supply for 3 h (−O₂ → +O₂). (C) iRicKO cells were either treated (for 4 days) or untreated with 4HT, and then kept under normoxia (control) deprived of oxygen for 10 h (−O₂), or deprived of oxygen for 10 h and then subjected to oxygen supply for 3 h (−O₂ → +O₂). Cells treated as in B and C were harvested and subjected to polysomal analysis.

Figure 7

Translational repression of TOP mRNAs by anoxia does not rely on TIA-1, TIAR, or phospho-eIF2α. (A) HEK293 cells were infected with viruses expressing HcRed (red) shRNA or an shRNA that can co-target both TIA-1 and TIAR (TIA-1/R). After 48 h, cells were subjected to selection by puromycin and harvested at 48 h later. The abundance of TIA-1 and TIAR was monitored by western blot analysis of cytoplasmic proteins with the indicated antibodies. (B) HEK293 cells were infected and selected as described in A and then either kept untreated (control), amino acid starved for 8 h (–AA), amino acid starved during the last 3 h of 24 h serum starvation (–Ser/AA), or deprived of oxygen (−O₂) for 16 h. Cells were harvested and their cytoplasmic extract were subjected to polysomal analysis. (C) eIF2α^S/S and eIF2α^A/A MEFs were untreated (+), Oxygen (O₂) deprived for 16 h (−), amino acid (AA) starved for 16 h (−), or serum starved for 48 h (−). Cytoplasmic proteins were subjected to western blot analysis. (D) Cytoplasmic extracts from eIF2α^S/S and eIF2α^A/A MEFs, treated as described in C, were subjected to polysomal analysis.