Not4-dependent translational repression is important for cellular protein homeostasis in yeast

Abstract

Translation of aberrant or problematic mRNAs can cause ribosome stalling which leads to the production of truncated or defective proteins. Therefore, cells evolved cotranslational quality control mechanisms that eliminate these transcripts and target arrested nascent polypeptides for proteasomal degradation. Here we show that Not4, which is part of the multifunctional Ccr4-Not complex in yeast, associates with polysomes and contributes to the negative regulation of protein synthesis. Not4 is involved in translational repression of transcripts that cause transient ribosome stalling. The absence of Not4 affected global translational repression upon nutrient withdrawal, enhanced the expression of arrested nascent polypeptides and caused constitutive protein folding stress and aggregation. Similar defects were observed in cells with impaired mRNA decapping protein function and in cells lacking the mRNA decapping activator and translational repressor Dhh1. The results suggest a role for Not4 together with components of the decapping machinery in the regulation of protein expression on the mRNA level and emphasize the importance of translational repression for the maintenance of proteome integrity.

Keywords: Ccr4–Not complex; Not4; protein homeostasis; ribosome stalling; translational repression.

Figure 1

Not4 interacts with polysomes

1. Ribosomes from wild-type (wt) yeast lysates were separated on a 10-40% sucrose gradient. Top: Absorbance profile at 254 nm (A₂₅₄). Bottom: Protein fractions were analysed by Western blotting using antibodies directed against the proteins indicated.

2. Ribosomal particles from an RNase A-treated and untreated control lysate were separated by density gradient centrifugation. Top: A₂₅₄ profiles. Bottom: Western blot analysis.

3. not4Δ cells expressing HA-tagged Not4 (Not4-HA) from a plasmid were grown to an optical density (OD₆₀₀) of 0.8. A lysate was prepared, and one half was treated with puromycin (Puro) to release nascent polypeptides and mRNA, while the other half was treated with cycloheximide (CHX) to stall translation. Samples were layered on top of a 20% sucrose cushion, and ribosomes were sedimented by ultracentrifugation. Ribosomal pellets were resuspended, and equal amounts of ribosomes were applied to Western blot analysis. Not4-HA was detected with antibodies directed against the HA-epitope tag. Rpl25 was detected as a loading control.

Source data are available online for this figure.

Figure 2

Not4 inhibits expression of polybasic translation arrest products

1. A Schematic of mRNA encoding the non-stalling GFP-Flag-His3 (K0) control construct or ribosome-stalling constructs where twelve consecutive lysine residues were inserted between GFP and Flag (GFP-K12-Flag-His3; K12-M) or fused to His3 (GFP-Flag-His3-K12; K12-C).

2. B, C Yeast cells transformed with centromeric plasmids expressing either K0 construct or K12-M (B) or K12-C ribosome-stalling construct (C) were grown in SCD −His to an optical density (OD₆₀₀) of 0.8, and normalized lysates were analysed by Western blotting. Full-length proteins and translation arrest products were detected with GFP-specific (α-GFP) and Flag-specific (α-Flag) antibodies. Rpl25 was detected as a loading control. The asterisk marks degradation products.

3. D Quantification of full-length K0 levels (n = 6, plotted on the left y-axis) as well as K12-M (n = 6) and K12-C (n = 3) arrest product levels (plotted on the right y-axis) from independent experiments as shown in (B) and (C). The values were normalized to the loading control, and arrest product levels are expressed relative to ltn1Δ cells (set to 1). Mean ± SD bars are shown.

4. E Top: Schematic of mRNA encoding the P12-M polyproline ribosome-stalling construct GFP-P12-Flag-His3. Bottom: Same experiment as in (B) performed with P12-M and β-NAC was detected as a loading control. Arrest product levels were quantified as in (D). Shown is mean ± SD (n = 3).

5. F Top: Schematic of mRNA with a HIS3 3′ untranslated region as described in Ito-Harashima et al (2007) encoding GFP-Flag-His3 fusion protein (Stop) or non-stop (NS) protein. Bottom: The experiment was performed as in (B) with Stop and NS constructs. β-NAC was detected as a loading control. Arrest product levels were quantified as in (D). Shown is mean ± SD (n = 5).

6. G Top: Schematic of the Rz-M mRNA containing a self-cleavable hammerhead ribozyme sequence (Rz; red) inserted into the open reading frame. Bottom: The experiment was performed as in (B) with the Rz-M construct. β-NAC served as a loading control. Arrest product levels were quantified as in (D). Shown is mean ± SD (n = 3). Asterisks mark a degradation product of K0.

Source data are available online for this figure.

Figure 3

Not4 acts in translational repression

1. A Schematic of mRNA encoding the reporter constructs GFP-2A-Flag-His3 (2A-K0) and GFP-2A-Flag-His3-K12 (2A-K12) containing an in-frame insertion of the FMDV 2A sequence.

2. B Northern blot analysis of 2A-K12 and 2A-K0 mRNA levels in yeast cells. The membrane was stained with methylene blue (MB) to visualize the 18S ribosomal RNA (rRNA) as a loading control. The reporter mRNA signals were quantified and normalized to the loading control. Shown is mean ± SD (n = 4 for 2A-K12 and n = 5 for 2A-K0).

3. C, D Lysates of yeast cells expressing 2A-K12 (C) or 2A-K0 (D) were analysed by Western blotting with antibodies against GFP (α-GFP) and β-NAC. Bar graph: Western blot signals of full-length proteins and GFP-2A of three independent experiments were quantified, normalized to the loading control and expressed relative to values in wild-type (wt) cells. Shown is mean ± SD (n = 3).

4. E Yeast cells were transformed with a plasmid expressing the ribosome-stalling construct 2A-K12. Cells were grown in SCD −His medium to the mid-log phase and treated with MG132. Lysates were prepared and the fusion proteins were immunoprecipitated. Samples of the lysates and the precipitated proteins were analysed by Western blotting. Proteins were detected with Flag-specific antibodies, and ubiquitination was detected with ubiquitin-specific (α-Ub) antibodies. Similar results were obtained in at least two separate experiments.

Source data are available online for this figure.

Figure 4

Altered mRNA levels have a minor influence on expression of arrested proteins

1. Northern blot analysis of K12-M or K0 mRNA levels in yeast cells. The membrane was stained with methylene blue (MB) to detect the 18S ribosomal RNA (rRNA) as a loading control. Bar graph: The mRNA signals were quantified, normalized to the loading control and expressed relative to wild-type (wt). Shown is mean ± SD (n = 4 for K12-M and n = 3 for K0).

2. Northern blot analysis as in (A) of K12-M and K0 mRNA levels in ccr4–not mutants. Shown is mean ± SD (n = 4 for K12-M and n = 3 for K0).

3. Parallel analysis of K12-M mRNA levels (top) and K12-M protein levels (bottom). Northern blot analysis was performed as in (A). GFP- (α-GFP) and Flag-specific (α-Flag) antibodies were used to detect reporter proteins by Western blotting. Arrest product levels were normalized to the β-NAC control signals. The asterisk marks non-specific bands. Similar results were obtained in three separate experiments.

Source data are available online for this figure.

Figure 5

Not4 is required for fast translational repression in response to nutrient withdrawal

1. A Polysome profiling with wild-type (wt) or mutant yeast cells. Absorbance traces at 254 nm (A₂₅₄) are shown. Cells were grown to an optical density (OD₆₀₀) of 0.5 in YPD, pelleted, resuspended in YP with or without 2% glucose and incubated for 10 min. Translation was stopped by the addition of cycloheximide, and cells were collected for polysome profiling on 15–45% sucrose gradients.

2. B ³⁵S-methionine incorporation into proteins after glucose depletion. Cells were grown in SCD medium to OD₆₀₀ 0.5 and transferred to SC labelling medium without glucose containing radioactive ³⁵S-methionine. Cells were incubated for 10 min and samples were taken. TCA-precipitable radioactivity was measured by liquid scintillation counting. Translation activity is given as incorporated radioactivity relative to t = 0. Best-fit trendlines are shown in grey.

3. C, D Polysome profiling of wt and mutant cells as in (A). Cells were grown in SCD medium to OD₆₀₀ 0.5 and transferred to SCD or yeast nitrogen base (YNB) containing 2% glucose without amino acids. Cells were incubated for 10 min prior to polysome analysis. Quantitative analysis of individual ribosome species is shown in (D) with mean values ± SD (n = 3).

Source data are available online for this figure.

Figure 6

Not4 and Dhh1 act in transcript-specific translational repression

1. A Spot assay to monitor growth defects of mutant yeast cells. Cells were adjusted to an optical density (OD₆₀₀) of 0.5, and 5-fold serial dilutions were spotted onto YPD plates. The plates were incubated as indicated.

2. B–D The ribosome-stalling K12-M or the non-stalling K0 control construct was expressed in wild-type (wt) and mutant yeast cells. Normalized lysates were applied to Western blot analysis. Full-length proteins and translation arrest products were detected with GFP-specific (α-GFP) antibodies. β-NAC was detected as a loading control. The asterisk indicates unspecific bands. Bar graph: Arrest product levels of three experiments were quantified, normalized to the loading control and expressed relative to ltn1Δ. Shown is mean ± SD (n = 3 in B; n = 6 in C and D).

3. E Northern blot analysis of K12-M levels. The membrane was stained with methylene blue (MB) to visualize the 18S ribosomal RNA (rRNA) as a loading control. The reporter mRNA signals from independent experiments were quantified and normalized to the loading control. Shown is mean ± SD (n = 3).

Source data are available online for this figure.

Figure 7

Deletion of NOT4 affects cellular protein homeostasis

1. Wild-type (wt) and not4Δ cells with a complementation plasmid (pNOT4) or empty vector (pEV) were adjusted to an optical density (OD₆₀₀) of 0.5, and 5-fold serial dilutions were spotted onto SCD −Ura plates. Plates were incubated as indicated.

2. Cells were grown at 30°C to the mid-log phase, and Hsp104 levels were analysed in normalized lysates by immunoblotting. Glucose-6-phosphate dehydrogenase (G6PDH) was detected as a loading control. Hsp104 signals were quantified and normalized to the loading control. Shown is mean ± SD (n = 4).

3. Wt and not4Δ cells were transformed with plasmids encoding GFP–Flag. Expression was controlled by either one of three different heat-shock responsive promoters (P) derived from the HSP104, RPN4 and HSP12 genes (P_HSP104, P_RPN4 and P_HSP12). Expression was analysed by Western blotting using Flag-specific antibodies. Immunodetection of Zuo1 served as a loading control. The signals were quantified and normalized to the loading control. Left: As a control, wt cells were grown at 25°C to OD₆₀₀ 0.8 and samples were taken before and 40 min after heat-shock (hs) at 38°C. Right: Wt and not4Δ cells expressing the reporter constructs were grown at 30°C to OD₆₀₀ 0.8 and samples were taken.

4. Analysis of protein aggregation in wt and mutant yeast cells. Cells were grown in YPD to OD₆₀₀ 0.8, and protein aggregates were isolated from equal volumes of normalized lysates. The insoluble proteins and samples of the normalized lysates were separated by SDS–PAGE and visualized by CBB staining. Bottom: Parallel Western blot analysis of the total and aggregate fractions. G6PDH and the chaperones Hsp104, Ssb1, Ssa1 and Sse1 were detected.

5. not4Δ cells were grown at 22°C or 30°C to OD₆₀₀ 0.8 and aggregates were analysed as in (D).

6. Eno2-Flag-mCherry was expressed in wt and not4Δ cells at 22°C or 30°C and analysed by fluorescence microscopy. Numbers give the percentage of cells that showed discrete mCherry foci. Scale bars, 5 μm.

7. not4Δ mutants were grown at 22°C to OD₆₀₀ 0.8 and shifted to 30°C with or without cycloheximide (CHX). Samples were taken at the indicated time intervals and aggregates were isolated for quantification. Mean values and SD bars of three experiments (n = 3) are shown.

8. Leucine auxotrophic not4Δ mutants were grown at 22°C to OD₆₀₀ 0.8 and shifted to 30°C with or without leucine. Samples were taken and analysed as in (G). Shown is mean ± SD (n = 3).

9. Expression of the K12-M reporter at 22°C and 30°C was analysed by Western blotting with antibodies against GFP (α-GFP) and β-NAC. The asterisk marks a non-specific band.

Source data are available online for this figure.

Figure 8

Decapping factors are required for the maintenance of cellular protein homeostasis

1. Protein aggregation was analysed in wild-type (wt) and mutant yeast cells. Cells were grown in YPD to an optical density (OD₆₀₀) of 0.8. Aggregated proteins were isolated from equal volumes of normalized lysates. Insoluble proteins and samples of the normalized lysates were separated by SDS–PAGE and visualized by CBB staining.

2. Cells were grown to the mid-log phase at 30°C, and Hsp104 levels were analysed in normalized lysates by Western blotting. Glucose-6-phosphate dehydrogenase (G6PDH) was detected as a loading control, and Hsp104 signals were quantified and normalized to the loading control. Shown is mean ± SD (n = 4).

Source data are available online for this figure.

Figure 9

Translational repression of ribosome-stalling mRNAs involves Not4, Dhh1 and the decapping factors Dcp1–Dcp2 Not4, together with the Ccr4–Not complex, associates with polysomes (grey) that likely contain stalled (red cross) and jammed ribosomes. Transient ribosome stalling on mRNAs within open reading frames (ORF) leads to Not4-dependent translational repression. The decapping activator Dhh1 and the decapping proteins Dcp1–Dcp2, which remove the 7-methylguanosine (m⁷Gppp) cap structure from the 5′ end of mRNAs, are also required for translational repression of ribosome-stalling mRNAs, suggesting that the Ccr4–Not complex and the decapping machinery act together in this process. Potential repression mechanisms include modulation of transcript-specific decapping or direct inhibition of translation initiation. This prevents further ribosome jamming and synthesis of arrested proteins. Upon disassembly of stalled ribosomes, the arrested nascent chains (NC) are ubiquitinated by Ltn1 to initiate their degradation and mRNAs may become eliminated.