Actin remodelling controls proteasome homeostasis upon stress

Abstract

When cells are stressed, bulk translation is often downregulated to reduce energy demands while stress-response proteins are simultaneously upregulated. To promote proteasome assembly and activity and maintain cell viability upon TORC1 inhibition, 19S regulatory-particle assembly chaperones (RPACs) are selectively translated. However, the molecular mechanism for such selective translational upregulation is unclear. Here, using yeast, we discover that remodelling of the actin cytoskeleton is important for RPAC translation following TORC1 inhibition. mRNA of the RPAC ADC17 is associated with actin cables and is enriched at cortical actin patches under stress, dependent upon the early endocytic protein Ede1. ede1∆ cells failed to induce RPACs and proteasome assembly upon TORC1 inhibition. Conversely, artificially tethering ADC17 mRNA to cortical actin patches enhanced its translation upon stress. These findings suggest that actin-dense structures such as cortical actin patches may serve as a translation platform for a subset of stress-induced mRNAs including regulators of proteasome homeostasis.

Fig. 1. Identification of proteins interacting with translating RPAC reporter mRNAs.

a, Cartoon depicting the FGH17 reporter, consisting of tandem reporters expressed under control of ADC17 UTRs and western blot analysis of FGH17 expression in untreated cells or cells treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. Empty vector, EV. b, mRNA levels of endogenous ADC17 and of FGH17 bound to ribosomes after 1.5 h rapamycin treatment compared with untreated cells. Analysis was performed by RiboTag immunoprecipitation (IP) followed by qRT–PCR and normalized to the housekeeping gene ALG9. Ribosome-bound mRNA corresponds to the level of RiboTag IP mRNA normalized to the level of Input mRNA. Data are presented as mean ± s.d., n = 4, unpaired two-tailed Student’s t-test. c,d, Western blot analysis of WT and mutant FGH17 reporters (anti-Flag) and Nas6 in untreated yeast cells or yeast cells treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. e, mRNA levels of endogenous ADC17 and of FGH17-70ntΔ bound to ribosomes after 1.5 h rapamycin treatment compared with untreated cells. Analysis was performed as in b. Data are presented as mean ± s.d., n = 4, unpaired two-tailed Student’s t-test. f, Western blot analysis of Adc17 and Nas6 expression in WT and ADC17-70ntΔ untreated cells or cells treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. g, Cartoon depicting the proteomics experimental design. Step 1, cells were treated with 200 nM rapamycin for 1.5 h or were left untreated; step 2, ribosomes were locked on mRNAs by treating cells with 35 µM CHX; step 3, cells were treated with 1.2 J cm⁻² UV to covalently crosslink proteins to RNA; step 4, translating FGH17 mRNAs were immunoprecipitated; step 5, proteins bound to translating FGH17 mRNAs were recovered by RNase treatment before being subjected to quantitative proteomics (Step 6). h, Volcano plot showing the proteins that were differentially recovered from FGH17 and FGH17-70ntΔ immunoprecipitates. Each dot represents a protein. The red and blue dots are proteins significantly more and less bound to FGH17 mRNA compared with FGH17-70ntΔ mRNA, respectively. n = 5 biologically independent samples per condition; P values were determined by multiple unpaired two-tailed t-test. In a, c, d and f, n = 3 independent biological replicates. Source data

Fig. 2. Ede1 regulates proteasome assembly upon TORC1 inhibition.

a, Cells spotted in a fivefold dilution and grown for 3 days on plates with or without 20 ng ml⁻¹ rapamycin. b, Western blot analysis of RPACs in WT and deletion strains that were untreated or treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as a loading control. Asterisk indicates non-specific band. Data are representative of three independent biological replicates. c, Gradient native polyacrylamide gel electrophoresis (PAGE) (3.8–5%) of yeast extracts from untreated cells or cells treated with 200 nM rapamycin (Rapa) for 3 h, monitored by the fluorogenic substrate Suc-LLVY–AMC (left) and by immunoblots (right). CP, single-capped (RPCP), double-capped (RP₂CP) and Blm10-capped (Blm10-CP) proteasome complexes are indicated. Rpt5 and 20 S antibodies recognize the RP and the CP, respectively. Data are representative of three independent biological replicates. Source data

Fig. 3. Ede1 controls ADC17 mRNA translation upon TORC1 inhibition.

a, Cartoon depicting how single-molecule ADC17 mRNAs are labelled with PCP fused to mKate2 for fluorescence live-cell imaging. PP7 stem loops were introduced into the endogenous ADC17 mRNA, allowing it to be selectively labelled with PCP-mKate2 in cells expressing Ede1-GFPEnvy. b, Montage from time-lapse imaging showing contacts between Ede1-GFPEnvy (green) and ADC17 mRNA (magenta). Scale bars, 1 μm. n = 4 biologically independent experiments. c, Frequency of ADC17 mRNAs (green) co-localizing with Ede1-tdimer2 in cells grown for 3 h with or without 200 nM rapamycin (Rapa). UT, untreated. Data are presented as mean ± s.d. n = 4 biologically independent experiments with 521 ADC17 mRNAs for each condition. Statistical analysis was carried out using unpaired two-tailed Student’s t-test. d, Schematic representation of ADC17-SunTag reporter mRNA used for single-molecule imaging of mRNA translation during stress. PCP-GFP labels ADC17 mRNA, whereas scFv-mCherry labels translating Adc17 protein. e, Representative microscopy images of yeast cells expressing ADC17-SunTag reporter mRNA. Translating ADC17 mRNAs are GFP (green)- and mCherry (magenta)-positive, while non-translating ADC17 mRNAs are only positive for GFP. Translating ADC17 mRNAs are denoted by white arrowheads 1 and 2, while non-translating ADC17 mRNAs are denoted by white arrowheads 3 and 4. Higher magnification is shown at the bottom. Scale bars, 3 μm. n = 5 biologically independent experiments. f, Frequency of ADC17 mRNAs undergoing translation in WT and ede1Δ cells that are untreated or treated with 200 nM rapamycin (Rapa) for 3 h using the SunTag labelling method. Data are presented as mean ± s.d., n = 5 biologically independent experiments with 547 ADC17 mRNAs for each condition. Statistical analysis was carried out using two-way ANOVA t-test (Tukey multiple comparison test). Source data

Fig. 4. Ede1, Sla1 and Vrp1 are important for proteasome assembly and activity.

a, Screen for rapamycin sensitivity with deletion strains covering all non-essential endocytic genes. Left: schematic representation of YEPD plates indicating the position of deletion strains. WT yeast was used as a control in the indicated positions. Right: yeast growth for 3 days on YEPD plate with or without 20 ng ml⁻¹ rapamycin from the indicated strains. Strains that failed to grow on rapamycin are indicated in coloured boxes in both the schematic and plate image. b, Western blot analysis of RPACs in WT and deletion strains that were untreated or treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as a loading control. Asterisk indicates non-specific band. c, Frequency of ADC17 mRNAs undergoing translation in WT, sla1Δ and vrp1Δ cells that were untreated or treated with 200 nM rapamycin (Rapa) for 3 h using the SunTag labelling method. Data are presented as mean ± s.d. n = 4 biologically independent experiments with 511 ADC17 mRNAs for each condition. Statistical analysis was carried out using two-way ANOVA i-test (Tukey multiple comparison test). d, Gradient Native PAGE (3.8–5%) of yeast extracts from cells that were untreated or treated with 200 nM rapamycin (Rapa) for 3 h, monitored by the fluorogenic substrate Suc-LLVY–AMC (left) and by immunoblots (right). CP, RPCP, RP₂CP and Blm10-CP proteasome complexes are indicated. Rpt5 and 20 S antibodies recognize the RP and the CP, respectively. In a, b and d, data are representative of three independent biological replicates. Source data

Fig. 5. ADC17 mRNA associates with actin cables and re-localizes to patches upon stress.

a, Cartoon depicting the role of Ede1, Sla1 and Vrp1. (1) Ede1 is recruited to nascent endocytic sites. (2) Sla1 recruits the NPF Las17 aided by the presence of Ede1. (3) Vrp1 is recruited to the endocytic site by Las17. (4) Vrp1 helps recruit Myo3 and Myo5. (5) The NPFs Las17, Myo3 and Myo5 recruit and activate the actin nucleator complex Arp2/3. b, Representative microscopy images (maximum-intensity Z-projection) of yeast containing the PCP-GFP-labelled ADC17 mRNA (cyan) and stained for actin (red). Z1, Z2 and Z3 areas are shown at higher magnifications. White, yellow and green arrowheads indicate ADC17 mRNAs bound to actin cable, cortical actin patch and not associated to actin, respectively. Scale bars, 3 μm. n = 4 biologically independent experiments. c, Frequency of ADC17 mRNAs bound to actin cable, cortical actin patch and not associated to actin structures. Data are presented as mean ± s.d., n = 4 biologically independent experiments (n = 232 ADC17 mRNAs per condition). d, Representative microscopy images showing ADC17 mRNA (cyan) interaction with actin cable (Abp140-mKate2, red). Scale bars, 3 μm. n = 3 biologically independent experiments. e, Representative microscopy images (maximum-intensity Z-projection) of yeast that was untreated or treated with 200 nM rapamycin for 1 h and stained with rhodamine phalloidin to visualize actin (hot red LUT). Scale bars, 3 μm. n = 4 biologically independent experiments. f, Frequency of polarized cells following rapamycin (200 nM) treatment for the indicated time. Data are presented as mean ± s.d. n = 4 biologically independent experiments (n = 399, 361, 337, 329 and 252 cells for conditions 0H, 1H, 2H, 3H and 4H, respectively). g, Western blot analysis of RPACs in WT cells that were untreated or treated with 200 nM rapamycin (Rapa) for the indicated time. Ponceau S staining was used as loading control. n = 3 biologically independent experiments. h, Frequency of ADC17 mRNA bound to actin cable, cortical actin patch or not associated to actin in WT cells that were untreated or treated with 200 nM rapamycin for the indicated time. Data are presented as mean ± s.d. n = 4 biologically independent experiments (n = 220 ADC17 mRNAs per condition). In f and h, one-way ANOVA t-test (Dunnett multiple comparison test) was used. Source data

Fig. 6. Lat-B induces RPACs and proteasome assembly.

a, Representative microscopy images (maximum-intensity Z-projection) of yeast cells containing the PCP-GFP-labelled ADC17 mRNA (cyan), with or without 25 μM Lat-B for 1 h and stained with rhodamine phalloidin to visualize actin (red). Z1 and Z2 areas are shown at higher magnifications. Scale bars, 2 μm. n = 4 biologically independent experiments. b, Frequency of ADC17 mRNA bound to actin cable, cortical actin patch or not associated to actin in WT cells that are untreated or treated with 25 μM Lat-B for the indicated time. Data are presented as mean ± s.d. n = 4 biologically independent experiments (n = 206 ADC17 mRNAs for each condition). Statistical analysis was carried out using one-way ANOVA t-test (Dunnett multiple comparison test). c, Western blot analysis of RPACs and Mpk1 kinase in WT cells that are untreated or treated with 25 μM Lat-B for the indicated time. Ponceau S staining was used as a loading control. d, Gradient native PAGE (3.8–5%) of yeast extracts from cells that are untreated or treated with 200 nM rapamycin (Rapa) or 25 μM Lat-B for 3 h, monitored by the fluorogenic substrate Suc-LLVY–AMC (left) and by immunoblots (right). CP, RPCP, RP₂CP and Blm10-CP proteasome complexes are indicated. Rpt5 and 20S antibodies recognize the RP and the CP, respectively. In c and d, data are representative of three independent biological replicates. Source data

Fig. 7. Tethering of ADC17 mRNA to actin patches enhances its translation upon stress.

a, Representative microscopy images (maximum-intensity Z-projection) of WT and ede1Δ cells treated with 25 μM Lat-B for 1 h and stained for actin (hot red LUT). Scale bars, 3 μm. n = 3 biologically independent experiments. b, Western blot analysis of RPACs in WT and ede1Δ cells that are untreated or treated with 25 μM Lat-B for 3 h. Ponceau S staining was used as loading control. n = 3 biologically independent experiments. c, Frequency of ADC17 mRNA bound to actin cable, cortical actin patch or not associated to actin in WT and ede1Δ cells that are untreated or treated with 200 nM rapamycin (Rapa) for 1 h. Data are presented as mean ± s.d. n = 5 biologically independent experiments (n = 212 ADC17 mRNAs for each condition). d, Schematic representation of the system used to artificially tether ADC17 mRNA to Ede1. aGFP, nanobody against GFP. e, Representative microscopy images of yeast cells containing PCP-GFP-labelled ADC17 mRNA (green) and expressing either WT Ede1 or Ede1 tagged with a nanobody against GFP (Ede1-aGFP). Scale bars, 3 μm. n = 3 biologically independent experiments. f, Western blot analysis of RPACs in cells shown in e that are untreated or treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. n = 4 biologically independent experiments. g, Quantification of Adc17 protein level from experiments represented in f. Data are presented as mean ± s.d. n = 4 biologically independent experiments. h, Western blot analysis of RPACs in the indicated cells that are untreated or treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. n = 5 biologically independent experiments. i, Quantification of Adc17 protein level from experiments represented in h. Data are presented as mean ± s.d. n = 5 biologically independent experiments. j, Western blot analysis of RPACs in the indicated cells that are untreated or treated with 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as loading control. n = 5 biologically independent experiments. k, Quantification of Adc17 protein level from experiments represented in j. Data are presented as mean ± s.d. n = 5 biologically independent experiments. In c, g, i and k, two-way ANOVA t-test (Tukey multiple comparison test). Source data

Extended Data Fig. 1. Characterisation of the region important for FGH17 regulation.

a, Schematic showing the introduction of the ADC17 5’UTR Kozak sequence into the FGH17-70ntΔ vector. Western blot analysis showing the impact on FGH17 levels in cells treated ± 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as a loading control. b, FGH17 vectors with the full 5’UTR, the 5’UTR lacking the 70 nucleotides upstream of the start codon (FGH17-70ntΔ) or containing only the 70 nucleotides upstream of the start codon (FGH17-70nt only). Western blot analysis showing the impact on FGH17 levels in cells treated ± 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as a loading control. c, Relative abundance of FGH17 and FGH17-70ntΔ mRNA (mRNA of interest normalised to ALG9 housekeeping mRNA) in yeast cells. Data are presented as mean ± s.d., n = 4 biologically independent experiments. Statistical analysis was carried out using unpaired two-tailed Student’s t-test. a, b, Data are representative of three independent biological replicates. Source data

Extended Data Fig. 2. Ede1 regulates RPAC levels downstream of TORC1 inhibition.

a, Cells spotted in a fivefold dilution and grown for 3 days on plates ± 20 ng/ml rapamycin. b, Western blot analysis of RPACs in WT and ede1Δ cells treated ± 200 nM rapamycin (Rapa) for 4 h. Ponceau S staining was used as a loading control. c, Western blot analysis of RPACs and P-Rps6 in WT and ede1Δ cells treated ± 200 nM rapamycin (Rapa) for the indicated time. Ponceau S staining was used as a loading control. a-c, Data are representative of three independent biological replicates. Source data

Extended Data Fig. 3. ADC17 mRNA interacts with cortical actin patch proteins.

a, Frequency of ADC17 mRNAs colocalizing with Sla1-mKate2 in cells grown for 3 h ± 200 nM rapamycin (Rapa). Untreated (UT). Data are presented as mean ± s.d., n = 4 biologically independent experiments. (n = 529 ADC17 mRNAs per condition). Statistical analysis was carried out using unpaired two-tailed Student’s t-test. b, Frequency of ADC17 mRNAs colocalizing with Vrp1-mKate2 in cells grown for 3 h ± 200 nM rapamycin (Rapa). Untreated (UT). Data are presented as mean ± s.d., n = 4 biologically independent experiments. (n = 408 ADC17 mRNAs per condition). Statistical analysis was carried out using unpaired two-tailed Student’s t-test. c, Representative single frames from time-lapse imaging showing contacts between Abp1-mKate2 (red) and PCP-GFP-labelled ADC17 mRNA (cyan). Scale bars, 1 μm. n = 3 biologically independent experiments. Source data

Extended Data Fig. 4. Genetic and chemical disruptions of actin induces RPAC levels.

Representative microscopy images (maximum intensity Z-projection) of WT (ACT1) and act1-101 cells either grown at the permissive temperature (25 °C) or shifted to the non-permissive temperature (37 °C) for 4 h and stained with Rhodamine phalloidin to visualise actin (hot red LUT). Scale bars, 3 μm. n = 3 biologically independent experiments. b, Western blot analysis of RPACs in WT and act1-101 cells either grown at the permissive temperature (25 °C) or shifted to the non-permissive temperature (37 °C) for 4 h. Ponceau S staining was used as a loading control. n = 3 biologically independent experiments. c, Representative microscopy images (maximum intensity Z-projection) of WT cells treated or not with either 25 μM Latrunculin-B or 12.5 μM Latrunculin-A for 3 h and stained for actin (hot red LUT). Scale bars, 3 mm. n = 3 biologically independent experiments. d, Frequency of ADC17 mRNAs colocalising with Ede1-tdimer2 in cells treated with either 25 μM Latrunculin-B (Lat-B) or 12.5 μM Latrunculin-A (Lat-A) for 3 h. Data are presented as mean ± s.d., n = 4 biologically independent experiments (>500 ADC17 mRNAs per condition). Statistical analysis was carried out using unpaired two-tailed Student’s t-test. ns not significant. e, Western blot analysis of RPACs and Mpk1 kinase in WT cells treated or not with either 25 μM Latrunculin-B (Lat-B) or 12.5 μM Latrunculin-A (Lat-A) for 3 h. Ponceau S staining was used as a loading control. n = 3 biologically independent experiments. Source data

Extended Data Fig. 5. Validation of ADC17 mRNA tethering to Ede1.

a, Schematic representation of PCP-GFP recruitment to Ede1 fused with tdimer2 and aGFP (nanobody against GFP). b, Representative microscopy images of yeast cells containing PCP-GFP (green) and Ede1-tdimer2 (magenta) tagged with a nanobody against GFP (Ede1-tdimer2-aGFP). Scale bars, 3 μm. n = 3 biologically independent experiments. c, Schematic representation of doubly tagged ADC17 mRNA (magenta) recruitment to Ede1-aGFP/PCP-GFP complex. d, Representative microscopy images of yeast cells expressing PCP-GFP (green), MCP-mCherry (magenta), Ede1 tagged with a nanobody against GFP (Ede1-aGFP) and ADC17 mRNA containing PP7-stem-loops (PP7SL) and MS2-stem-loops (MS2SL) which are recognised by PCP-GFP and MCP-mCherry, respectively. Scale bars, 3 μm. n = 3 biologically independent experiments. e, Quantification of the number of ADC17 mRNAs per cell in adc17Δ Ede1 WT and Ede1-aGFP cells expressing ADC17 mRNA containing MS2 and PCP stem loops, PP7-GFP and MCP-mCherry grown for 2 h ± 200 nM rapamycin (Rapa). Untreated (UT). mRNAs were quantified using MCP-mCherry as a marker. Data are presented as mean ± s.d., n = 4 biologically independent experiments. (n = 261 ADC17 mRNAs per condition). Statistical analysis was carried out using two-way ANOVA t-test (Tukey multiple comparison test). ns, not significant. Source data