Cell culture and experimental animals. SMMC7721 cells, HAP1 cells, and HEK-293 T cells were preserved in our laboratory. SMMC7721 cells and HEK-293 T cells were cultured in DMEM basic (1×) (Gibco, C14190500BT) supplemented with 10% fetal bovine serum (Gibco, 10099-141) and 1% penicillin and streptomycin (Gibco, 15140-122) in an incubator (Binder, CB160) containing 5% CO2 at 37°C. HAP1 cells were maintained in IMDM modified medium (HyClone, SH30228-01). Cells were digested using 0.25% Trypsin-EDTA (Gibco, 25200056).
eIF2α S51A heterozygous mice were purchased from Jackson Laboratory, USA, strain: B6; 129-Eif2s1tm1Rjk/J, 017601. The tail tissues of WT mice and offspring of eIF2α S51A heterozygous mice were collected and DNA were extracted, and then were screened and identified by PCR, the primer sequences: Forward: 5'-ACACCCATTCCATGATAGTAAAATG-3', Reverse: GTTGTAGACCCTGACAATGAAGG-3'); MEFs cells were cultured in DMEM basic (1×).
MEFs preparation. Fetal rats of 13 ~ 15 days were anesthetized by intraperitoneally injecting 0.5 mL of 2.5% avertin, then euthanized by cervical dislocation. The fetal rat was placed on a paper towel and sprayed enough 75% ethanol to soak the entire mouse body. Transferred fetal rat into biological safety cabinet, cut skin and tissues to expose uterus, removed embryos with a scalpel in 10 cm dish containing PBS, discarded all other tissues, such as the head and visceral tissues. The remaining part of the embryos was washed with PBS 4 times and placed in a clean 10 cm dish and cut embryos into 2 mm pieces with a new scalpel. Added 0.25% trypsin and incubated in 37°C incubators for 10 min. DMEM containing 10% FBS was added to resuspend cells, inoculated into T75 cell flasks, cultured in an incubator containing 5% CO2, discarded non-adherent cells, and replaced culture medium to continue culturing.
Generation of ATF4-KD and eIF2α kinases-KO cell lines construction. ATF4 knockdown cells (ATF4-KD) were constructed by shRNA, and targeting sequences of ATF4 was available from the website (https://www.sigmaaldrich.cn/CN/zh). Firstly, primers were synthesized by Sangon Biotech (Shanghai, China) and cloned into pLKO.1 vector (Addgene, 10878) by digestion and ligation, shRNA sequences for ATF4 were shown in the following: ATF4: 5'-GCCTAGGTCTCTTAGATGATT-3'.
The eIF2α kinases (PERK, PKR, GCN2, and HRI) knockout cell lines (eIF2α kinases-KO) were manufactured using the CRISPR-Cas9 technique. sgRNA were designed using sgRNA Designer from Zhang's lab. The sgRNAs were synthesized and cloned into the Lenti-CRISPR v2 vector (Addgene, 52961) by ligation. sgRNA sequences of eIF2α kinases as follows: PERK: 5'-GTAATTATCAGCACTTTAGA-3'; PKR-1: 5'-GTACTACTCCCTGCTTCTGA-3', PKR-2: 5'-GATTATGAACAGTGTGCATCG-3'; GCN2-1: 5'-CCTCCGGAGAGCTACCCGCAACG-3', GCN2-2: 5'-TATATGTAAAAGTGGATTTG-3'; HRI-1: 5'-CCGAGGGCCCGGACCCCGAA-3', HRI-2: 5'-AAGGGAAGGTTGGCTGTTGT-3'.
For lentivirus packing and infection, 4 µg recombinant lentiviral plasmid was co-transfected with 2 µg packaging plasmid psPAX2 (Addgene, 12260) and 1 µg envelope plasmid pMD2.G (Addgene, 12259) into Lenti-™ 293 T cells using Lipofectamine™ 3000 Transfection Reagent (Thermo Fisher Scientific, L3000015). Infectious particles were harvested from 48 h to 96 h after transfection, centrifugated at 4000 rpm/min for 30 min, and concentrated with a final concentration of 20% PEG-8000, then concentrated lentiviruses were added into SMMC7721 or HAP1 cell lines, respectively, cells were selected by adding 250 µg/mL hygromycin (Invitrogen, 10687010) or 5 µg/mL puromycin (Invitrogen, A1113803). For KO cell lines, positive clones were isolated through limiting dilution assay, and knockdown and knockout efficiency were determined by Western blot.
Construction of HAP1-sgHRI-HRI-TurboID cell line. The pEGFP-N1-HRI-TurboID plasmid was constructed by inserting the CDs sequence of HRI into the 5' edge of the TurboID cDNA sequence. To select positive clones against hygromycin, a homologous arms assay was constructed with a pSMPUW-Hygro vector, and lentiviral packaging was performed. Packaged lentiviruses were added into HAP1-HRI−/− cells and selected by 250 µg/mL hygromycin (Invitrogen, 10687010). The positive clones were determined by Western blot.
Western blot. Cells were washed with cold PBS twice, and ultrasonic protein was extracted on ice in protease and phosphatase inhibitor (Thermo Fisher Scientific, A32959) in RIPA buffer (Beyotime, P0013B) with 1% PMSF (Thermo Fisher, 36978). Lysates were centrifuged at 12000 rpm/min for 10 min. Supernatants were collected and denatured at 95°C for 10 min in a 4 × loading buffer. Equal amounts of protein were resolved on 10% electrophoresis acrylamide gel (Bio-Rad, 1610173), transferred to PVDF membranes (Millipore, ISEQ00010), and blocked in 5% nonfat milk (BD, 232100) in TBS-T (50 mM Tris (pH 7.4), 150 mM NaCl, 0.1% Tween20 (Sigma, P7949-500ML) for 2 h at room temperature (RT). In TBS-T, membranes were incubated with primary antibodies overnight at 4°C in 5% bovine serum albumin (BSA)(Millipore, 9048-46-8). Primary antibodies were used in this study as follows: β-tubulin antibody (66240-1-Ig; 1:5000); EIF2A antibody (11170-1-AP; 1:2000); ATF4 antibody (10835-1-AP; 1:1000); LONP1 antibody (15698-1-AP; 1:1000); CLPP antibody (15440-1-AP; 1:1000); HRI antibody (20499-1-AP; 1:1000); TIM23 antibody (11123-1-AP; 1:1000); GOT1 antibody (14886-1-AP; 1:1000); GOT2 antibody (14800-1-AP; 1:1000) were purchased from Proteintech; PERK antibody (3192s; 1:1000); CHOP antibody (D46F1; 1:1000); HSP60 antibody (D6F1; 1:1000); GAPDH antibody (5174; 1:1000); Histone H3 antibody (4499; 1:5000); BIP antibody (3183S; 1:1000); GCN2 antibody (3302S; 1:1000); COX IV antibody (3E11; 1:1000); COX IV antibody (4844; 1:1000); CS antibody (14309; 1:1000); LAMP1 antibody (9091; 1:1000); Streptavidin-HRP antibody (3999s; 1:1000) were purchased from CST; ATF5 antibody (Ab184923; 1:2000); PKR antibody (Ab32506; 1:1000); Phospho-EIF2A antibody (Ser51) (Ab32157; 1:2000) were purchased from Abcam; HSPE1 antibody (Thermo Fisher, PA5-79415; 1:2000); Flag antibody (Sigma-Aldrich, F1804; 1:1000). Membranes were washed with TBS-T three times, 10 min each time. Membranes were incubated in appropriate HRP-conjugated secondary antibodies (Bio-Rad, 170–6515; 1:5000 and 170–6516; 1:5000) at 1:10,000 dilution in 5% milk in TBS-T for 2 h at RT. Membranes were washed three times with TBS-T, 10 min each time. Membranes were detected using standard chemiluminescence with ECL (Affinity, KF003) and imaged by GE Healthcare Amersham™ Imager 600 in an automatic exposure model to ensure that bands are not saturated. Band intensity was determined using ImageJ software (NIH) and adobe photoshop cc software (Adobe). At least 3 independent experiments were performed.
Quantitative Real-Time PCR (RT-qPCR). Total RNA was extracted using RNeasy Plus Universal Mini Kit (50) (QIAGEN, 73404). According to the manufacturer's directions, RNA was converted to cDNA using the ReverTra Ace qPCR RT Master Mix (TOYOBO, FSQ-301). Each reaction includes 2 µL 4×DN Master Mix ( with gDNA Remover), RNA template 0.2 µg, and nuclease-free water 5 µL, 37°C incubated 5 min, and then added 5×RT Master Mix II 2 µL, 37°C 15 min, 50°C 5 min, 98°C 5 min. RT-qPCR was performed using Blaze TaqTM SYBR Green qPCR Mix2.0 kit (Gene Copoeia, QP043). Each reaction of qPCR contains 2×SYBR Green PCR Master Mix 10 µL, forward primer 0.7 µM, reverse primer 0.7 µM, RNase-free water 5.2 µL, template cDNA 2 µL. PCR initial heat activation 95°C 2 min, 95°C, denaturation 5 s, 60°C, extension 10 s, 40 cycles. The primer used in qPCR as following: ACTIN (Human) Forward: 5'-GGACCTGACTGACTACCTCAT-3', Reverse: CGTAGCACAGCTTCTCCTTAAT-3'); LONP1 (Human) (Forward: 5'-TCAATGTCACCCGCAACTAC-3', Reverse: 5'-GAACTCCAGGATGCGTTTCT-3'); CLPP (Human) (Forward: 5'-CTCTTCCTGCAATCCGAGAG-3', Reverse: 5'-GGATGTACTGCATCGTGTCG-3'); HSP60 (Human) (Forward: 5'-TCCAGGGTTTGGTGACAATAG-3', Reverse: 5'-GATTCAGGGTCAATCCCTCTTC-3'); HSPE1 (Human) (Forward: 5'-GCTGCTGAAACTGTAACCAAAG-3', Reverse: 5'-CTCCACCCTTTCCTTTAGAACC-3'); CHOP (Human) (Forward: 5'-GTCTAAGGCACTGAGCGTATC-3', Reverse: 5'-CAGGTGTGGTGATGTATGAAGA-3'); ATF5 (Human) (Forward: 5'-CCCACCTGACCTGGAAGC-3', Reverse: 5'-CCTCGTTGCGGCAGTAGAT-3'); ATF4 (Human) (Forward: 5'-GGAGATAGGAAGCCAGACTACA-3', Reverse: 5'-GGCTCATACAGATGCCACTATC-3'); BIP (Human) (Forward: 5'-AACCATCCCGTGGCATAAA-3', Reverse: 5' -GGACATACATCAAGCAGTACCA-3'’). The sample was analyzed in three biological replicates, and the relative expression values were obtained using the 2−ΔΔCt method .
ROS detection. ROS was detected using DCFDA cellular ROS detection assay kit (Abcam, ab113851). The protocol was as follows: HAP1 cells were harvested and seeded in a dark 96-well assay plate (black with clear flat bottom), allowing cells to be attached overnight. Cells were washed once with 1×Buffer. Cells were stained with 25 µM DCFDA in 1×Buffer for 45 min at 37°C, washed once in 1 × Buffer, and treated with GTPP (10 µM), ISRIB (200 nM), and GTPP + ISRIB, DMSO as a negative control, 55 µM TBHP (Tert-Butyl Hydrogen peroxide) as the positive control. Read signals at Ex/Em: 485/535 nm by PerkinElmer VICTOR Nivo multimode plate reader.
Transmission electron microscopy. HAP1 cells were seeded into a 60 mm dish and treated with GTPP (10 µM) for 2 and 4 h when cells were confluence about 80 ~ 90%, with DMSO as the negative control. Method for ultrastructure analysis of mitochondria was performed according to the protocol: fixation: cells were collected and fixed with 2.5% glutaraldehyde at 4°C for 2 h, rinsed with 0.1 mol/l phosphoric acid rinsing solution three times, 15 min each time, fixed with 1% osmium acid for 2 h, rinsed with 0.1mol/L phosphoric acid rinsing solution 3 times, 15 min. Dehydration: dehydrated with 50% of ethanol dehydration for 15 min, dehydrated with different concentrations of ethanol dehydration for 15 min, 70, 90, and 95%, respectively, the above septs were conducted in a fume cupboard, dehydrated with 100% acetone 3 times at RT, 15 min each time. Embedding: cells were embedded with pure acetone along with an embedding solution (2:1) at 37°C for 3 h, embedded with pure acetone along with an embedding solution (1:2) at 37°C for 3 h, and placed in pure embedding solution at 45°C for more than 2 h, and then converted to 60°C. Solidify: cells were placed in a drying oven at 45°C for more than 3 h and at 60°C for 24 h. Sliced by Lycra UC7 ultra-thin microtome to 50–60 nm. Double stained with 3% uranyl acetate-lead citrate. Cellular ultrastructure was observed and filmed with a JEOL JEM-1400 transmission electron microscope.
RNA-Seq. Total RNA was isolated and purified using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's procedure. The RNA amount and purity of each sample were quantified using NanoDrop ND-1000 (NanoDrop, Wilmington, DE, USA). The RNA integrity was assessed by Bioanalyzer 2100 (Agilent, CA, USA) with a RIN number > 7.0. The final cDNA libraries were constructed using TruSeq® Stranded mRNA Library Prep Kit (Illumina, San Diego, CA, USA). The 2×150 bp paired-end sequencing (PE150) was performed on an Illumina Novaseq™ 6000 (LC-Biotechnology CO., Ltd., Hangzhou, China) following the vendor's recommended protocol.
Analysis of RNA data. Fastp software (https://github.com/OpenGene/fastp) was used to remove the reads that contained adaptor contamination, low-quality bases (Q ≤ 20, more than 50%), and undetermined bases (more than 10%). Then, sequence quality was also verified using fastp. HISAT2 (https://ccb.jhu.edu/software/hisat2) was used to map reads to the reference genome of Homo sapiens Ensembl_release101. The mapped reads of each sample were assembled using StringTie (https://ccb.jhu.edu/software/stringtie) with default parameters. Then, all transcriptomes from all samples were merged to reconstruct a comprehensive transcriptome using gffcompare (https://github.com/gpertea/gffcompare/). RSEM was used to perform gene expression levels by calculating fragments per kilobase million (FPKM). The differentially expressed genes were selected with |log2 Fold Change| > 1 and FDR < 0.05 by R package DESeq2 (http://www.bioconductor.org/packages/release/bioc/html/DESeq2.html).
Ribo-Seq. Cycloheximide was added to the cell medium to a final 100 µg/ml concentration to block translation. The resuspended extracts in lysis buffer were transferred into a clean microtube, pipetted several times, and incubated on ice for 10 min. Then cells were triturated 10 times through a 26-G needle. The lysate was centrifuged at 20000 g/min for 10 min at 4°C, and the supernatant was collected. To prepare ribosome footprints (RFs), 7.5 µL of RNase I and 5 µL of DNase I was added to 300 µL of lysate to incubate for 45 min at RT with gentle mixing on a nutator mixer. Nuclease digestion was stopped by adding 10 µL of SUPERase·In RNase inhibitor. Size exclusion columns (Illustra MicroSpin S-400 HR Columns; GE Healthcare; catalog no. 27-5140-01) were equilibrated with 3 ml of polysome buffer by gravity flow and centrifuged at 600 g/min for 4 min at RT. 100 µL of digested RFs were added to the column and centrifuged at 600 g/min for 2 min. Next, 10 µL 10% (wt/vol) SDS was added to the elution, and RFs with a size greater than 17nt were isolated according to the RNA Clean and Concentrator-25 kit (Zymo Research; R1017). rRNA was removed using DNA probes complementary to rRNA sequences. Then RNase H and DNase I were used to digest the probes. RFs were purified using magnet beads (Vazyme). Ribo-seq libraries were constructed using NEBNext® Multiple Small RNA Library Prep Set for Illumina® (catalog no. E7300S, E7300L) after obtaining the ribosome footprints above. PCR products were sequenced using Illumina HiSeq™ 2500 (LC-Biotechnology CO., Ltd., Hangzhou, China).
Analysis of Ribo-seq data. Raw reads containing over 50% of low-quality bases or over 10% of N bases were removed. Adapter sequences were trimmed. Reads with lengths between 10 ~ 50 bp were retained for subsequent analysis. Bowtie2 was used for mapping reads to the ribosome RNA (rRNA) database. The rRNA removed reads of each sample were mapped to the reference genome by Bowtie2, allowing no mismatches. Reads number in the open reading frame of coding genes was calculated by the software RiboTaper, and the gene expression level was normalized using the FPKM method, the same as the RNA-Seq. To identify differentially translated genes across sample groups, the edgeR package (http://www.rproject.org/) was used. Genes with |log2 Fold Change| > 1 and FDR < 0.05 in comparison were considered as significant DTGs. DTGs were then subjected to enrichment analysis of GO functions and KEGG pathways.
Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. Cells were treated with G-TPP (10 µM) for 2 and 4 h, respectively, equal volume DMSO was added as the control, and 500 mmol/L Biotin (Sigma, V900418-1G) was added for 30 min before finishing treatment time. The samples were performed by Immunoprecipitation. IP samples were separated by SDS-PAGE, and a gel lane was cut for LC/MS/MS. Briefly, the samples were digested, desalted, and analyzed by Q Exactive high-resolution mass spectrometer (Thermo Scientific) accompanied by Easy-nLC 1200 (Thermo Scientific). Proteome Discover 2.5 software was used to map reads to the reference genome of UniProt-Human and analyze the data.
Immunoprecipitation. To investigate the relationship between HRI and MAVS, the pEGFP-N1-MAVS-TurboID plasmid was constructed by inserting the CD sequence of MAVS into the 5' edge of the TurboID cDNA sequence. The pEGFP-N1-MAVS-TurboID plasmid was transfected into HEK 293T cells using Lipofectamine 3000 reagent for 36 h, and cells were treated with G-TPP for 4 h, equal volume DMSO as the control, and 500 mmol/L Biotin (Sigma, V900418-1G) was added for 30 min before finishing treatment time. Rinsed with pre-cooled PBS once, added 400 µL RAPA lysis solution, placed on ice for 30 min to absolutely dissolve, sonicated with cell sonicator for 1 min, 12000 rpm/min, centrifuged for 10 min, and 80 µL of lysate were transferred into a clean 1.5 mL centrifuge tube, and 5×loading buffer was added, heated at 98°C, 10 min, as Input samples. 50 µL of Hydrophilic Streptavidin Magnetic beads (NEB, S1421S) were washed with 500 µL lysate buffer, discarded supernatant, and beads were added into remaining cell lysate, 18 rpm/min, 4°C overnight. The cell lysate was washed 3 times, 500 µL each time and 80 µL of lysate buffer were added into 5×loading buffer, 98°C heated for 10 min. Input samples and IP samples were performed with Western blot.
Cytoplasmic and nuclear extraction test. HAP1 cells were seeded into a 60 mm dish, and cells were treated with G-TPP (10 µM) for 2 h when cellular confluence was up to 80% and an equal volume of DMSO as a control. Cells were washed twice with pre-cooled PBS after treatment and discarded PBS. The extraction assay was performed according to MinuteTM Cytoplasmic and Nuclear Extraction Kit (Invent Biotechnologies, Inc., SC-003). 500 µL of cytoplasmic extraction buffer were added, placed on ice for 5 min, then transferred into a pre-cooled 1.5 mL microtube, vortexed vigorously for 15 s, and centrifuged for 5 min at top speed in a microcentrifuge at 4°C. The supernatant (cytosol fraction) was transferred into a fresh pre-chilled 1.5 mL microtube. Appropriate amounts of nuclear extraction buffer were added to the pellets, vortexed vigorously for 15 s, incubated microtube on ice for 1 min, and repeated this step 4 times. Immediately transferred the nuclear extract into a pre-chilled filter cartridge with a collection microtube and centrifuged at top speed (14,000 ~ 16,000 rpm/min) in a microcentrifuge for 30 s at 4°C. Discarded filter cartridge, Added 5×loading buffer in the cytoplasm and the nucleoplasm components, heated at 95°C, 10 min to denature protein, and then performed by Western blot.
Drugs treatment. G-TPP (synthesized from Chengdu Ruizhi Chemical Research Co. LTD) is dissolved in DMSO, and the final concentration is 10 µM. ISRIB (Med Chem Express) dissolves in DMSO, and the final concentration is 200 nM.
Statistical analysis. Data were expressed as mean ± standard deviation (SD). The significance of the variability between different treatment groups was analyzed by a two-way analysis of variance (ANOVA) test via GraphPad Prism software (version 8.0.2.). P < 0.05 was considered statistically significant. The pictures of Western blot were imaged by both Image J software and photoshop cc software.