Data reporting
Table of Contents
No statistical methods were used to predetermine sample size. The experiments were not randomized and investigators were not blinded to allocation during experiments and outcome assessment.
Cell culture and treatments
HeLa ovarian carcinoma cells from the American Type Culture Collection were used for all experiments unless stated otherwise. They were confirmed to be mycoplasma negative and grown in RPMI medium (Thermo Fisher Scientific) and 10% fetal bovine serum. Knockdown experiments were performed with Lipofectamine RNAiMAX according to the manufacturer’s instructions. The small interfering RNAs (siRNAs) used were from OriGene oligo duplex ATF5 (SR307793) and custom made for LONP1 (sense 5′-GGACGUCCUGGAAGAGACCAAUAUU-3′, anti-sense 5′-AAUAUUGGUCUCUUCCAGGACGUCC). MISSION esiRNA (Sigma) were ATF5 (EHU039491), DNAJA1 (EHU114481), HSF1 (EHU107721), DNAJA2 (EHU005311), DNAJB1 (EHU109151), NRF1 (EHU069871) and PITRM1 (EHU011041). Gene KOs were conducted by CRISPR–Cas9-mediated genome editing. The single guide RNAs (sgRNAs) were cloned into eSpCas9 (1.1; Addgene, catalogue no. 71814). The sgRNA sequences used were 5′-GCAACAGAAAGTCGTCAACA-3′ (HSF1), 5′-TCTCTTAGATGATTACCTGG-3′ (ATF4), TCAGCCAAGCCAGAGAAGCA-3′, 5′-ATTTCCAGGAGGTGAAACAT-3′ (DDIT3), 5′-TGGCTCCCTATGAGGTCCTT-3′ (ATF5_1) and 5′-AGACTATGGGAAACTCCCCC-3′ (ATF5_2). Together with sgRNA-containing plasmid, cells were cotransfected with puromycin-resistant plasmids and selected for 24 h with 1 μg ml−1 puromycin (Invivogen). After the selection, single cells were seeded into 96-well plates and incubated for 2 weeks. Resulting colonies were expanded, and gene KO was confirmed by Sanger sequencing and western blot.
Transient overexpression of MTS-Abeta-GFP was carried out with Lipofectamine 2000 according to the manufacturer’s instruction. Cells were harvested after 24 h.
Acute induction of the UPRmt was performed with 10 µM GTPP (Shanghai Chempartner), 5 µM CDDO (Cayman Chemical) and 40 µM Ucf-101 (Cayman Chemical) for 6 h unless stated otherwise (for early response, a 3 h incubation was used). For the cell viability assay, a toxic concentration of 15 µM GTPP for 16 h was applied (Extended Data Fig. 9d–f). mtROS induction was done by treating cells with 10 µM antimycin A (Sigma) or 2 µM rotenone (Sigma) for 6 h. To scavenge ROS, cells were pretreated with 10 mM NAC (Sigma) or 10 mM GSH (Cayman Chemical) for 1 h or 100 µM MnTBAP (Sigma) overnight that was continued as a cotreatment. For Hyper7 references, 20 µM antimycin A and 1 mM H2O2 (Carl Roth) were used; 4,4′diisothiocyanatostilbene-2,2′-disulfonate (75 µM, Sigma) was used to inhibit VDAC1 for 6 h. General translation was blocked by treatment with 35 µM CHX for 30 min and continued as cotreatment for 6 h. Mitochondrial import inhibition was performed with 5 µM oligomycin A (Sigma) for 6 h. Different mitochondrial stressors were applied by 6 h of treatment with 10 µM carbonyl cyanide m-chlorophenyl hydrazone (CCCP, Abcam), 100 µM deferiprone (DFP, Sigma) or 10 µM Menadione (Sigma). The hypoxic condition was generated by incubating cells in the BD GasPak EZ Pouch system (BD Diagnostics) for 6 h. Staurosporine (Cayman Chemical) was used to induce apoptosis as a control treatment at 1 µM for 3 h or 200 nM overnight for the HSF1 KO cell viability assay (Extended Data Fig. 9d–f).
Cloning
For the generation of the construct MTS-Abeta-GFP, pcDNA5/FRT/TO (Thermo) was used as a backbone. The following inserts were amplified by Q5 High-Fidelity DNA Polymerase (NEB) and cloned into the backbone via NEBuilder HiFi DNA Assembly Master Mix (NEB): MTS (2× COX8 presequence in tandem) amplified from pCMV CEPIA2mt (Addgene, catalogue no. 58218), Abeta (Aβ1-42) amplified from HeLa wild-type complementary DNA (cDNA) with primers (5′-TCC ATG CGG GGT TCT GAT GCA GAA TTC CGA CAT GAC TCA GGA TAT G-3′ and 5′-CTC GCC CTT GCT CAC GGA TCC CGC TAT GAC AAC ACC GCC CAC C-3′, containing a GS linker) and enhanced green fluorescent protein (EGFP) amplified from Su9-EGFP (Addgene, catalogue no. 23214).
RNA sequencing
Total RNAs were extracted from cells using the NucleoSpin RNA Plus kit (Macherey-Nagel) following the manufacturer’s instructions and subsequently digested with Turbo DNase (Thermo Fisher Scientific). Library preparation for bulk sequencing of poly(A)-RNA was done as described previously31. Briefly, barcoded cDNA of each sample was generated with a Maxima RT polymerase (Thermo Fisher Scientific) using an oligo-dT primer containing barcodes, unique molecular identifiers (UMIs) and an adaptor. Ends of the cDNAs were extended by a template switch oligo, and full-length cDNA was amplified with primers binding to the template switch oligo site and the adaptor. The NEB UltraII FS kit was used to fragment cDNA. After end repair and A tailing, a TruSeq adaptor was ligated, and 3′-end fragments were finally amplified using primers with Illumina P5 and P7 overhangs. In comparison with Parekh et al.31, the P5 and P7 sites were exchanged to allow sequencing of the cDNA in read1 and barcodes and UMIs in read2 to achieve a better cluster recognition. The library was sequenced on a NextSeq 500 (Illumina) with 63 cycles for the cDNA in read1 and 16 cycles for the barcodes and UMIs in read2.
RNA sequencing analysis
Gencode gene annotations v.35 and the human reference genome GRCh38 were derived from the Gencode homepage (European Molecular Biology’s European Bioinformatics Institute (EMBL-EBI)). Drop-Seq tools (v.1.12)32 were used for mapping raw sequencing data to the reference genome. The resulting UMI filtered count matrix was imported into R (v.4.0.5), and lowly expressed genes were subsequently filtered out. Data were then variance stabilized via the rlog function as implemented in DESeq2 (v.1.18.1)33. For accurate dispersion estimation, the experimental design (treatment at a given time point) was provided to the function. rlog normalized data were used to perform clustering analysis (fuzzy C means) with R package mFuzz (v.2.50.0)34. Transcripts with rlog normalized values of less than three were excluded from the analysis. The number of clusters was set to three. Transcripts were assigned to increased and decreased cluster groups based on cluster membership greater than or equal to 0.8 for each cluster. Gene Ontology (GO) enrichment analysis was performed on each cluster group by using Database for Annotation, Visualization and Integrated Discovery (DAVID). GO enrichments were visualized with the EnrichmentMap (v.3.3.2) plug-in in Cytoscape (v.3.7.1).
Quantitative polymerase chain reaction analysis
Total RNAs were extracted from cells using the NucleoSpin RNA Plus kit (Macherey-Nagel) following the manufacturer’s instructions. cDNA synthesis was performed with the High-capacity cDNA reverse transcription kit (Applied Biosystems). Quantitative polymerase chain reaction (qPCR) analysis was performed with primaQuant SYBRGreen master mix without ROX (Steinbrenner Laborsysteme) according to the manufacturer’s instructions. KiCqStart primers SYBR green from Sigma (Supplementary Table 4) were used to perform qPCR measurement with LightCycler 480 SW (v.1.5) on the LightCycler 480 real-time PCR system (Roche) in 384-well format. ACTB was used as an internal control. Fold changes of the transcript level were calculated using the comparative CtΔΔCt (cycle threshold) method.
FACS measurement
MitoSOX Red (Thermo Fisher Scientific) was used to measure mtROS production according to the manufacturer’s instructions. Cell deaths were measured with a combination of Annexin V conjugated to Alexa Fluor 488 (Thermo Fisher Scientific) and propidium iodide (Thermo Fisher Scientific) according to the manufacturer’s instructions. Fluorescence-activated cell sorting (FACS) was performed with FACSDiva (v.6.1.3) on FACSCanto II and FACSymphony A5 flow cytometry systems (BD) for MitoSOX and Annexin V measurements, respectively. Analysis of FACS data was performed with FlowJo v.10 software.
Immunoblotting
Cells were lysed in RIPA buffer containing Complete Mini EDTA-free protease inhibitor (Roche) and GENIUS nuclease (Santa Cruz Biotechnology). Lysates were prepared in 1× Laemmli buffer and boiled for 10 min at 95 °C. Proteins were separated with SDS–PAGE using the Invitrogen Novex system and transferred to nitrocellulose membrane by using Mini Trans-Blot cell (Bio-Rad). Primary antibodies were added to immunoblots for 1 h at room temperature (RT). Antibodies used for the detection were anti-ACTB (SantaCruz, catalogue no. sc69879, 1:4,000), anti-HSPD1 (Abcam, catalogue no. ab46798, 1:2,000), anti-COX5B (Proteintech, catalogue no. 11418-2-AP, 1:1,000), anti-HSF1 (Cell Signaling, catalogue no. 4356, 1:1,000), anti-HSF1 (Abcam, catalogue no. ab2923, 1:10,000), anti-DNAJA1 (Proteintech, catalogue no. 11713-1-AP, 1:2,000), anti-Hsp70 (Proteintech, catalogue no. 10995-1-AP, 1:2,000), anti-NRF1 (Cell Signaling (D9K6P), catalogue no. 46743, 1:1,000), anti-α-tubulin (Cell Signaling (DM1A), catalogue no. 3873, 1:3,000), anti-histone H3 (Active Motif, catalogue no. 39163, 1:5,000), anti-FLAG (Sigma, catalogue no. F1804, 1:5,000), anti-CHOP (Thermo Fisher Scientific, catalogue no. MA1-250, 1:1,000), anti-ATF4 (Cell Signaling, catalogue no. 11815, 1:1,000), anti-PITRM1 (Novus, catalogue no. H00010531-M03, 1:500), anti-LONP1 (Proteintech, catalogue no. 15440-1-AP, 1:2,000), anti-cleaved PARP1 (Cell Signaling, catalogue no. 5625, 1:2,000) and anti-Caspase3 (Cell Signaling, catalogue no. 9661, 1:1,000). Secondary antibodies used were anti-rabbit IgG (H + L) HRP Conjugate (Promega, catalogue no. W4021, 1:10,000), IRDye 800CW goat anti-rabbit IgG (H + L; Li-Cor, catalogue no. 926–32211, 1:15,000) and IRDye 680RD donkey anti-mouse IgG (H + L; Li-Cor, catalogue no. 926–68072, 1:15,000). Appropriate secondary antibodies were used for imaging with Odyssey DLx (LI-COR) or ChemiDoc MP (Bio-Rad) imaging system. Data were collected with Image Studio (v.5.2) or ImageLab v.6.0.1.
Mitochondrial insoluble fraction analysis
Mitochondrial fractions were prepared as previously described35. Briefly, cells were homogenized by passing them through a 27-gauge needle syringe in buffer containing 10 mM HEPES (pH 7.4), 50 mM sucrose, 0.4 M mannitol, 10 mM KCl and 1 mM EGTA. Mitochondrial enrichment was performed with a two-step differential centrifugation at 1,000g followed by 13,000g for 15 min each at 4 °C. The mitochondria-enriched pellets were resuspended in a buffer containing 20 mM HEPES (pH 7.4), 0.4 M mannitol, 10 mM NaH2PO4 and 0.5 M EGTA. An equal volume of lysis buffer containing 2% (vol/vol) NP40 was added and spun down to separate mitochondrial fractions. The resulting supernatants and pellets were kept as the soluble and insoluble fractions, respectively. Proteins were resolved with SDS–PAGE in 1× Laemmli buffer and visualized with InstantBlue Coomassie stain (Expedeon).
Nuclear and cytosolic fractionation
Cells were fractionated with the REAP method36. Cell fractions were prepared by resuspending cells in PBS containing 0.1% (vol/vol) NP40, followed by five times resuspension with a p1000 micropipette (Gilson). Cells were fractionated with a ‘pop spin’ for 10 s at 4 °C in an Eppendorf tabletop microfuge. Supernatants were collected as the cytosolic fractions. Pellets were washed once with 0.1% (vol/vol) NP40 and collected as the nuclear fractions. Both the cytoplasmic and nuclear fractions were used to perform immunoblotting. The ratio of nuclear to cytosolic HSF1 was calculated as follows:
HSF1 (N/C) = (Nuclear HSF1/Histone H3)/(Cytoplasmic HSF1/Tubulin).
Immunoprecipitation
Crosslinking was performed by incubating cells in PBS containing 0.8 mg ml−1 dithiobis[succinimidyl propionate] (Proteochem) for 30 min at RT37. Crosslinking reactions were quenched with PBS containing 200 µM glycine for 15 min at RT. Cells were lysed in cell lysis buffer (50 mM Tris (pH 8.0), 150 mM NaCl, 1% (vol/vol) NP40) containing protease inhibitor and allowed to incubate for 30 min at 4 °C. Lysates containing 2 mg of total proteins were used to perform immunoprecipitation with 10 µl Dynabeads protein A (Thermo Fischer Scientific) containing 1 µg of appropriate antibodies or 10 µl Anti-FLAG M2 magnetic beads (Sigma) for 2 h at 4 °C. Immunoprecipitated proteins were eluted from beads for immunoblotting or digested for interaction proteomics.
Sample preparation for LC–MS/MS
For redox proteomics, cells were lysed in HES buffer (1 mM EDTA, 0.1% (wt/vol) SDS, 50 mM HEPES (pH 8.0)) supplemented with protease inhibitor and 10% (vol/vol) TCA and incubated for 2 h at 4 °C. Each sample was divided into two fractions: (1) oxidized Cys fraction and (2) total Cys fraction. Proteins were precipitated with a TCA and acetone precipitation. For fraction 2, 100 µg of proteins were resuspended in HES buffer supplemented with 5 mM TCEP and incubated for 1 h at 50 °C to reduce all Cys thiols. For fraction 1, 100 µg of the proteins were resuspended in denaturing buffer (6 M urea, 1% (wt/vol) octyl ß-glucopyranoside, 50 mM HEPES (pH 8.0)) supplemented with protease inhibitor and 200 mM iodoacetamide and incubated for 1 h at 37 °C in the dark to block free Cys thiols. Oxidized Cys thiols were reduced as described previously for fraction 2. Proteins were cleaned up by TCA and acetone precipitation. To label the free Cys thiols, proteins were resuspended in denaturing buffer supplemented with iodoTMT#1 (Thermo Fisher Scientific) for fraction 1 or iodoTMT#2 for fraction 2 and incubated for 1 h at 37 °C in the dark. Labelling reactions were quenched with 20 mM DTT. Labelled proteins were pooled together and cleaned up with TCA and acetone precipitation. Proteins were digested with 1:50 (wt/wt) LysC (Wako Chemicals) and 1:100 (wt/wt) Trypsin (Promega) in 10 mM EPPS (pH 8.2) containing 1 M urea overnight at 37 °C. Peptides were purified with (50-mg) SepPak columns (Waters) and then dried. IodoTMT-labelled peptides were enriched with anti-TMT antibody resin (Thermo Fisher Scientific) according to the manufacturer’s instructions. Enriched pools of labelled peptides were subjected to high-pH reverse-phase fractionation with the High pH RP Fractionation kit (Thermo Fisher Scientific) following the manufacturer’s instructions. Fractionated peptides were concatenated into four separate fractions.
To perform interaction proteomics, after immunoprecipitation steps 25 µl of SDC (2% SDC (wt/vol), 1 mM TCEP, 4 mM chloroacetamide, 50 mM Tris (pH 8.5)) buffer was added to the beads. The mixtures were heated up to 95 °C, and the supernatants were collected. For digestion, 25 µl of 50 mM Tris (pH 8.5) containing 1:50 (wt/wt) LysC (Wako Chemicals) and 1:100 (wt/wt) trypsin (Promega) was added and allowed to incubate overnight at 37 °C. Digestion was stopped by adding 150 µl of isopropanol containing 1% (vol/vol) TFA. Peptide purification was performed with the SDB-RPS disc (Sigma) and then dried.
LC–MS/MS
Peptides were resuspended in a 2% (vol/vol) acetonitrile/1% (vol/vol) formic acid solution and separated on an Easy nLC 1200 (Thermo Fisher Scientific) and a 35-cm-long, 75-μm-inner-diameter fused-silica column, which had been packed in house with 1.9-μm C18 particles (ReproSil-Pur, Dr. Maisch) and kept at 50 °C using an integrated column oven (Sonation). For redox proteome, peptides were eluted by a nonlinear gradient from 4 to 36% (vol/vol) acetonitrile over 90 min and directly sprayed into a QExactive HF mass spectrometer equipped with a nanoFlex ion source (Thermo Fisher Scientific) at a spray voltage of 2.3 kV. Full-scan MS spectra (350–1,400 m/z) were acquired at a resolution of 120,000 at m/z 200, a maximum injection time of 25 ms and an automatic gain control (AGC) target value of 3 × 106. Up to 20 of the most intense peptides per full scan were isolated using a 1-Th window and fragmented using higher-energy collisional dissociation (normalized collision energy of 35). MS/MS spectra were acquired with a resolution of 45,000 at m/z 200, a maximum injection time of 86 ms and an AGC target value of 1 × 105. Ions with charge states of one, five to eight and more than eight as well as ions with unassigned charge states were not considered for fragmentation. Dynamic exclusion was set to 20 s to minimize repeated sequencing of already acquired precursors.
For interaction proteomics, peptides were eluted by a nonlinear gradient from 3.2 to 32% acetonitrile over 60 min followed by a stepwise increase to 95% B in 6 min, which was kept for another 9 min and sprayed into an Orbitrap Fusion Lumos Tribrid Mass Spectrometer (Thermo Fisher Scientific) at a spray voltage of 2.3 kV. Full-scan MS spectra (350–1,500 m/z) were acquired at a resolution of 60,000 at m/z 200, a maximum injection time of 50 ms and an AGC target value of 4 × 105. The most intense precursors with a charge state between two and six per full scan were selected for fragmentation (‘Top Speed’ with a cycle time of 1.5 s) and fragmented using higher-energy collisional dissociation (normalized collision energy of 30). MS/MS spectra were acquired with a resolution of 15,000 at m/z 200, a maximum injection time of 22 ms and an AGC target value of 1 × 105. Ions with charge states of one and more than six as well as ions with unassigned charge states were not considered for fragmentation. Dynamic exclusion was set to 45 s to minimize repeated sequencing of already acquired precursors.
LC–MS/MS data analysis
For analysis of redox proteomics data, raw files were analysed using Proteome Discoverer 2.4 software (Thermo Fisher Scientific). Spectra were selected using default settings and database searches performed using the SequestHT node in Proteome Discoverer. Database searches were performed against a trypsin-digested Homo sapiens SwissProt database and FASTA files of common contaminants (‘contaminants.fasta’ provided with MaxQuant) for quality control. Dynamic modifications were set as methionine oxidation (C, +15.995 Da), iodoTMT6plex (C, +329.227 Da) and carbamidomethyl (C, +57.021 Da) at cysteine residues. One search node was set up to search with Met loss + acetyl (M, −89.030 Da) as dynamic modifications at the N terminus. Searches were performed using Sequest HT. After each search, posterior error probabilities were calculated, and peptide spectrum matches were filtered using Percolator with default settings. Consensus workflow for reporter ion quantification was performed with default settings, except that the minimal signal-to-noise ratio was set to 10. Results were then exported to Excel files for further processing. Non-normalized abundances were used for quantification. The percentage of cysteine oxidation for each peptide was calculated as follows:
Percentage of oxidized Cys = (abundance of fraction 1/abundance of fraction 2) × 100%.
For peptides with several different Cys modifications, fold changes of the percentage of oxidized Cys from each different combination were considered.
For DNAJA1 interaction proteomics, MS raw data processing was performed with MaxQuant (v.1.6.17.0) and its in-build label-free quantification algorithm MaxLFQ applying default parameters38. Acquired spectra were searched against the human reference proteome (Taxonomy identification 9606) downloaded from UniProt (12-03-2020; ‘One sequence per gene’, 20,531 sequences) and a collection of common contaminants (244 entries) using the Andromeda search engine integrated in MaxQuant39. Identifications were filtered to obtain false discovery rates below 1% for both peptide spectrum matches (minimum length of seven amino acids) and proteins using a target-decoy strategy40. Results were then exported to Excel files for further processing. Abundance of interactors was normalized to the abundance of DNAJA1 from each sample. Fold changes were calculated from normalized data. GO enrichment analysis of DNAJA1 interactome was performed by using DAVID. GO enrichments were visualized with the EnrichmentMap (v.3.3.2) plug-in in Cytoscape (v.3.7.1). Subcellular locations of increased interactors upon GTPP treatment were manually curated from UniProt.
Microscopy analysis
For HyPer7 measurements, cells were transfected with different constructs of HyPer7 (ref. 13) (Addgene, catalogue nos. 136466, 136469 and 136470) with Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer’s instructions. Measurements were performed 24 h after transfection in a 96-well plate format at 37 °C. Time-series live cell imaging was done with a CQ1 confocal imaging cytometer (Yokogawa). HyPer7 was excited sequentially with 405- and 488-nm laser beams. Emission was collected using a 525/50-bandpass emission filter. After five images were acquired, 10 µM GTPP was added to each group of cells expressing different constructs. Image analysis was performed using ImageJ (v.1.53). Fluorescence was calculated for regions of interests inside the imaged cell. The ratiometric signal of HyPer7 was calculated by dividing the intensity of the emission signals excited by 488/405 nm.
Monitoring of DNAJA1 localization was performed using SP8 Confocal (Leica). Cells were incubated in media containing 150 nM MitoTracker deep red FM (Thermo Fisher Scientific) for 30 min at 37 °C in the dark. After subsequent washes, the cells were fixed with 4% (vol/vol) formaldehyde in PBS and permeabilized with 0.01% (vol/vol) TritonX100. Cells were blocked with a PBS buffer containing 1% (wt/vol) bovine serum albumin (BSA), 300 mM glycine and 0.1% (vol/vol) Tween20 for 30 min at RT. Cells were incubated overnight at 4 °C with 1:100 dilutions of anti-DNAJA1 antibody (11713-1-AP, Proteintech) in PBS containing 1% (wt/vol) BSA and 0.1% (vol/vol) Tween20. After 3× washes, 1:1,000 dilution of Alexa Fluor 488 anti-rabbit IgG in 1% BSA PBS was used to incubate the cells for 1 h at RT as a secondary antibody. A drop of ProLong diamond antifade mountant containing DAPI (Thermo Fisher Scientific) was used to mount the cells. Data were collected with Leica Application Suite X. Image analysis was performed using ImageJ (v.1.53). Pearson’s and Mander’s colocalization coefficients were calculated with the JACoP plug-in41. Calculation from the independent images was reported.
For monitoring mitochondrial import, MTS-EGFP (Addgene, catalogue no. 23214) was transiently transfected together with 10 µM GTPP treatment. After 6 h of incubation, cells were stained with 50 nM Mitotracker Deep Red FM (Thermo Fisher Scientific) for 15 min and transferred to RPMI 10% fetal calf serum for live cell imaging. CQ1 (Yokogawa) with 40× magnification was used with the following laser settings: 488-nm excitation and 525/50-nm emission for EGFP and 640-nm excitation and 685/40-nm emission for Mitotracker Deep Red FM. Three wells with a total of 100 cells per condition were manually characterized into one of five categories.
For Halo-tagged reporter assay, T-Rex-HeLa cells stably expressing Halo-tagged ATP5A1 and GREPL1 were used. Blocking of previously synthesized Halo-tagged proteins was done by incubating cells in media containing 5 µM empty HaloTag ligand (Promega) overnight. Treatments were started on the following day. Newly synthesized Halo-tagged proteins were labelled with 5 µM HaloTag TMR ligand (Promega) in the last hour of the treatments. Mitochondria were stained with 50 nM Mitotracker Deep Red FM (Thermo Fisher Scientific). CQ1 (Yokogawa) with 40× magnification was used with following laser settings: 488-nm excitation and 525/50-nm emission for TMR and 561-nm excitation and 617/73-nm emission for Mitotracker Deep Red FM. Several images were collected from three independent replicates, and image analysis was performed using ImageJ (v.1.53). Pearson’s and Mander’s colocalization coefficients were calculated with the JACoP plug-in41. Calculation from three independent replicates was reported.
Cell viability assay
Cell viability was measured with Cell Counting Kit-8 (Dojindo) according to the manufacturer’s instructions. Cell viability was evaluated 16 h (overnight) after treatment with the different chemicals used in the experiment.
Statistics and plots
A general statistical analysis was performed with a two-tailed Student’s t test (considered significant for P ≤ 0.05), unless it was stated otherwise. All plots were created using the R packages ggplot2 (v.3.3.3), gplots (v.3.1.1) and RColorBrewer (v.1.1-2). Visualization of the final figures was done with Adobe Illustrator CS5.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.