Stalled translation by mitochondrial stress upregulates a CNOT4-ZNF598 ribosomal quality control pathway important for tissue homeostasis

Abstract

Translational control exerts immediate effect on the composition, abundance, and integrity of the proteome. Ribosome-associated quality control (RQC) handles ribosomes stalled at the elongation and termination steps of translation, with ZNF598 in mammals and Hel2 in yeast serving as key sensors of translation stalling and coordinators of downstream resolution of collided ribosomes, termination of stalled translation, and removal of faulty translation products. The physiological regulation of RQC in general and ZNF598 in particular in multicellular settings is underexplored. Here we show that ZNF598 undergoes regulatory K63-linked ubiquitination in a CNOT4-dependent manner and is upregulated upon mitochondrial stresses in mammalian cells and Drosophila. ZNF598 promotes resolution of stalled ribosomes and protects against mitochondrial stress in a ubiquitination-dependent fashion. In Drosophila models of neurodegenerative diseases and patient cells, ZNF598 overexpression aborts stalled translation of mitochondrial outer membrane-associated mRNAs, removes faulty translation products causal of disease, and improves mitochondrial and tissue health. These results shed lights on the regulation of ZNF598 and its functional role in mitochondrial and tissue homeostasis.

Fig. 1. ZNF598 is upregulated under mitochondrial stress conditions in vitro and in vivo.

a Western blot analysis and quantification of ZNF598 protein level in HeLa cells treated with EBSS (16 h), Torin1 (0.5 μM, 24 h), and CCCP (10 μM, 24 h). b Western blot analysis and quantification of ZNF598 protein level in HeLa cells treated with rotenone (5 μM, 24 h) or CCCP (10 μM, 24 h). Western blot analysis of HA-tagged ZNF598 (c), Pelo (d), Rack1 (e), and ABCE1 (f) proteins in the muscle tissue of flies treated with rotenone (250 μM) or CCCP (100 μM) for 7 days. Values under the blot show normalized protein levels relative to untreated sample in this figure and all subsequent ones. g Western blot analysis of the indicated proteins in total and mitochondrial fractions isolated from striatum tissues of normal and MPTP-indued mice. Data are representative of at least three biologically independent experiments (mean ± SD), n = 3 in (a–c), and (f), n = 4 in (g). Representative blot is from three independent experiments with similar results in (a–f). p values in (a–c), and (f) were calculated by one-way ANOVA (Tukey’s test). p values in (g) were calculated by unpaired two-tailed t test. Source data are provided as a Source Data file.

Fig. 2. ZNF598 regulates quality control of stalled translation of mRNAs encoding mitochondria-targeted proteins.

a Cell viability assay of control and ZNF598 KO HEK293T cells treated with rotenone, CCCP, TG and BFA for 24 h. b Immunoblots of GFP-P2A-Flag-K20-P2A-mCherry stall reporter expression in control (pLK0.1), ZNF598 KO, and ASCC3 KD HEK293T cells. c Flow cytometry analysis of GFP and RFP in normal, ZNF598 KO, and ASCC3 KD HEK293T cells expressing Dox-inducible GFP-P2A-Flag-K20-P2A-mKate2 after 1 μg/ml Dox incubation for 24 h. d Immunoblots showing effect of GFP-ZNF598 OE on GFP-P2A-Flag-K20-P2A-mCherry stall reporter expression in control (pLK0.1) and ASCC3 KD cells. e Immunoblots showing effect of GFP-ZNF598 and FLAG-ASCC3 OE on GFP-P2A-Flag-K20-P2A-mCherry stall reporter expression in control and ZNF598 KO cells. Asterisk indicates FLAG-ASCC3, and arrow indicates FLAG-Cas9 protein. f Immunoblots showing expression of GFP-P2A-Flag-K20-P2A-mCherry cytosol stall reporter and MTSGFP-K20-P2A-mKate2 mitochondrial stall reporter in control and ZNF598 KO cells. The ratio of arrested protein (AP) to full-length protein (FP) and mKate2 to actin were shown. g Immunoblots of MTS-GFP-K20-P2A-mKate2 stall reporter expression in ZNF598 KO, and ASCC3 KD HEK293T cells. For anti-GFP blot, the ratio of arrested GFP-K20 protein (AP) to full length protein (FP) is shown. h Immunostaining of GFP and mKate2 in normal and ZNF598 KO U2OS cells after transfecting with MTSGFP-K20-P2A-mKate2 for 24 h. i Immunoblots of GFP and mKate2 in mitochondria isolated from normal and ZNF598 KO HEK293T cells after transfecting with MTSGFP-K20-P2A-mKate2 for 24 h. j Effect of Flag-ZNF598 OE on MTS-GFP-K20 stall reporter expression in HeLa cells with or without CCCP treatment (6 hr). k Effect of ZNF598 KO on MTS-GFP-K20-P2A-mKate2 stall reporter expression in HeLa cell with or without CCCP or CCCP + ATP treatment. Data are representative of at least three biologically independent experiments (mean ± SD), n = 5 in (a), n = 3 in (c). Representative blot or image is from at least three independent experiments with similar results. p values in a were calculated by two-way ANOVA (Sidak’s test). p values in (c) were calculated by one-way ANOVA (Tukey’s test). Source data are provided as a Source Data file.

Fig. 3. ZNF598 aborts stalled translation of mitochondrial outer membrane-associated C-I30 mRNA and rescues mitochondrial and neuromuscular defects of PINK1 mutant flies.

a Immunostaining of ZNF598 and TOM20 or GFP in U2OS cells under normal, CCCP treatment, and MTSGFP-nonstop OE conditions. b Immunoblots of ZNF598 in isolated mitochondria from normal and CCCP treatment cells, followed by proteinase K + Triton X-100 treatment. VDAC as out membrane, CHCHD3 as inner mitochondrial membrane, and ACTB as cytosolic protein markers. c Localization of HA-tagged early RQC factors ZNF598 and Rack1 to mitochondria of adult DA neurons. DA neuron mitochondria are marked with TH-Gal4-driven mito-GFP expression. d Western blot analysis showing effect of ZNF598 and Rack1 OE on CAT-tailed C-I30-u formation in PINK1^B9 fly muscle. e, f Effect of ZNF598 OE (by UAS-ZNF598-HA or ZNF598-EP) or RNAi on abnormal wing posture in PINK1^B9 flies. g Effect of ZNF598 OE on mitochondrial morphology in PINK1^B9 fly muscle. Red arrowheads mark aggregated mitochondria. Mito-GFP labels mitochondria. Bar graph shows quantification of damaged mitochondria. Effect of ZNF598 OE on DA neuron number in the PPL1 cluster of PINK1^B9 adult brains (h). Bar graph shows data quantification (i). Data are representative of at least three biologically independent experiments (mean ± SD), n = 4 and 5 in (e), n = 3 in (f, g, i). Representative blot or image is from at least three independent experiments with similar results. p values in (e) and (g) were calculated by unpaired two-tailed t-test. p-values in (f) were calculated by one-way ANOVA (Tukey’s test). p values in (i) were calculated by two-way ANOVA (Sidak’s test). Source data are provided as a Source Data file.

Fig. 4. ZNF598 regulates the quality control of stalled translation of C9ALS-associated poly(GR) and rescues poly(GR)-induced toxicity in Drosophila and patient cell models.

a Immunoblots showing the effect of ZNF598-HA OE on the level of Flag-tagged GR80 expressed in fly muscle. b Effect of ZNF598-HA OE on wing posture in Mhc>GR80 flies. c Immunoblots showing effect of ZNF598 RNAi on the level of FLAG-tagged GR80 in Mhc>GR80 flies. d Effect of ZNF598-RNAi on wing posture in Mhc>GR80 flies. e Effect of ZNF598 OE on mitochondrial morphology in Mhc > GR80 fly muscle, which normally exhibits vacuolated and disconnected mitochondria. f Dot blots showing effect of ZNF598 OE or RNAi on poly(GR) protein level in C9ALS patient fibroblasts. g Immunoblots showing effect of ASCC3 or ZNF598 silencing, or ZNF598 OE, on Flag-GR80 protein expression in HELA cells. Effect of ZNF598 on mitochondrial calcium stained by Rhod-2AM (h) and MMP stained by TMRM (i) in C9ALS fibroblasts. GFP was co-transfected with ZNF598 at 1:3 ratio so that virtually all ZNF598 transfected cells would be marked with GFP. Bar graphs show data quantification. j Effect of ZNF598 siRNA on MMP stained by TMRM in C9ALS fibroblasts. Bar graph shows data quantification. Data are representative of at least three biologically independent experiments (mean ± SD), n = 3. Representative blot or image is from at least three independent experiments with similar results. p-values were calculated by unpaired two-tailed t-test. Source data are provided as a Source Data file.

Fig. 5. GR80 interacts with RQC factors and induces ribosomal stalling.

a Western blot analysis of indicated proteins in HEK293T cells overexpressing FLAG-GR80. b Immunoprecipitation analysis testing potential interaction between FLAG-GR80 and RQC factors (ASCC3, ZNF598, RPS10, EDF1, 4EHP, GIGYF2, CNOT4). Calmodulin serves as a negative control. c Immunoblots showing effect of GR80 on translational readthrough of the GFP-P2A-FLAG-K20-P2A-mCherrry stall reporter in normal and ZNF598 KO HEK293T cells. d Cell extracts from control (left panels) or FLAG-GR80-overexpressing HEK293T cells (right panels) were fractionated on 10-50% sucrose gradients. A total of 11 fractions were analyzed by SDS-PAGE and immunoblotted with the indicated antibodies. e Cell extracts from control (left panels) or FLAG-GR80-overexpressing HEK293T cells (right panels) were incubated with RNase A (10 μg/ml) for 30 min and fractionated on 10-50% sucrose gradients. Polysome profiles were derived from RNase-digested lysates of HEK293T cells and GR80 overexpressing HEK293T cells. 11 fractions were analyzed by SDS-PAGE and immunoblotted with the indicated antibodies. Representative blot is from at least three independent experiments with similar results. Source data are provided as a Source Data file.

Fig. 6. ZNF598 undergoes regulatory K63-linked ubiquitination under mitochondrial stress.

a Immunoblots showing ubiquitinated ZNF598 in HEK293T cells co-transfected with FLAG-ZNF598 and HA-Ub. Cells were treated with 2 μM Rotenone or 10 μM CCCP for 24 h followed by denaturing FLAG-IP. b Immunoprecipitation analysis of self-assembly of ZNF598 in cells co-transfected with GFP-ZNF598 and ZNF598-FLAG. c Effect of wild type ZNF598 and catalytically inactive mutant ZNF598 (ZNF598 CS) on GFP-P2A-Flag-K20-P2A-mKate2 stall reporter expression in HEK293T cells. d Effect of wild type ZNF598 and catalytically inactive ZNF598 (ZNF598 CS) on FLAG-GR80 expression in transfected HeLa cells. e Immunoblots showing ubiquitination of ZNF598 WT and inactive ZNF598 mutant (ZNF598 CS) in HEK293T cells co-transfected with FLAG-ZNF598 and HA-Ub. Cells were treated with 10 μM CCCP for 24 h followed by denaturing FLAG-IP. f Immunoblots showing effect of CCCP treatment (10 μM, 24 h) on ZNF598 ubiquitination in HEK293T cells co-transfected with Flag-ZNF598 and HA-Ub followed by denaturing IP and detection of Ub. The signals in the Flag-IP represent ubiquitinated ZNF598. g Immunoblots showing effect of CCCP treatment (10 μM, 24 h) on ZNF598 ubiquitination in HEK293T cells co-transfected with Flag-ZNF598 and HA-Ub-K63 or HA-Ub-K63R followed by denaturing IP and detection of Ub. All data are representative of at least three independent experiments. h Immunoblots showing effect of HA-Ub-K63 and HA-Ub-K63R OE on GFP-FLAG-K20-mCherry reporter expression in ZNF598 overexpressing HEK293T cells. i Cell viability measurement using CCK8 assay in HEK293T cells transfected with plasmids overexpressing ZNF598, Ub-K63, or Ub-K63R for 48 h followed by treatment with 20 μM CCCP (24 h). Data are representative of at least three biologically independent experiments (mean ± SD). Representative blot is from at least three independent experiments with similar results. p values were calculated by one-way ANOVA (Tukey’s test). Source data are provided as a Source Data file.

Fig. 7. A CNOT4-ZNF598 axis regulates quality control of stalled translation and stress response.

a Immunoblots showing levels of ZNF598 in HEK293T cells with or without CNOT4 transfection. b, c Analysis of ubiquitination status of endogenous ZNF598 immunoprecipitated under denaturing condition from CNOT4-deficient HEK293T cells transfected with HA-Ub-K63 (b) or CNOT4 overexpressing HEK293T cells transfected with HA-Ub-K63 or HA-Ub-K63R (c). d Immunoblots showing effect of CNOT4 OE and CNOT4 knockdown on the level of FLAG-tagged GR80 expressed in fly muscle. e Immunoblots showing effect of CNOT4 OE on GFP-FLAG-K20-mCherry reporter expression. f Immunoblots showing effect of CNOT4 OE on GFP-FLAG-K20-mCherry reporter expression in ZNF598 KO HEK293T cells. g Immunoblots showing effect of CNOT4 knockdown on GFP-FLAG-K20-mCherry reporter expression in ZNF598 overexpressing HEK293T cells. h Immunoblots showing effect of CNOT4 OE on GFP-FLAG-K20-mCherry reporter expression in ZNF598 overexpressing HEK293T cells. i Immunoblots showing reduced ZNF598-KR ubiquitination by Ub-K63 when compared with ZNF598 WT. j Immunoblots showing effect of CNOT4 OE on GFP-FLAG-K20-mCherry reporter expression in HEK293T cell overexpressing ZNF598 WT or ZNF598-KR. Cell viability measurement using CCK8 assay in HEK293T cells transfected with siRNA targeting ZNF598 and plasmid overexpressing CNOT4 for 48 h (k) or plasmid overexpressing ZNF598 or CNOT4 for 24 h (l). Cells were treated with 20 μM CCCP (24 h). m Immunoblots showing effect of ZNF598 WT and ZNF598 KR on GFP-FLAG-K20-mCherry reporter expression in CNOT4 overexpressing cells with endogenous ZNF598 knocked down. n Effect of ZNF598 WT and ZNF598 KR on cell viability in CNOT4 overexpressing cells treated with CCCP and with endogenous ZNF598 knocked down. Data are representative of at least three biologically independent experiments (mean ± SD), n = 4 in (k, l), n = 3 in (n). Representative blot is from at least three independent experiments with similar results. p-values were calculated by one-way ANOVA (Tukey’s test). ns no significance. Source data are provided as a Source Data file.