Sex as a biological variable. Our study examined male and female animals, as well as lung tissue samples from both male and female individuals, and similar findings are reported for both sexes.

Mice and hypoxia study. Eif2ak4–/– (KO) mice (The Jackson Laboratory 008240) were backcrossed with C57BL/6 WT mice (The Jackson Laboratory 000664) for 5 generations to maintain C57BL/6 background. Both male and female mice were used for experiments. At the age of 11 weeks, WT and KO mice were exposed to 10% O2 in a hypoxia chamber (BioSpherix) for 3 weeks and then subjected to hemodynamic measurements.

Overexpression of Edn1 in ECs in mice. Mouse Edn1 plasmid was obtained from GenScript (catalog OMu20519) and subcloned into pGL3 plasmid containing human CDH5 promoter. EndoNP1 nanoparticles (MountView Therapeutics LLC, catalog EndoNP1) and 30 μg CDH5-Edn1 plasmid or control GFP plasmid were mixed following manufacturer’s instructions and incubated at room temperature (RT) for 15 minutes and then administered to 11-week-old WT or KO mice. After overnight, the mice were exposed to chronic hypoxia (10%) for 3 weeks. The mixture of nanoparticles/plasmid DNA was administered weekly for 3 times total.

Hemodynamic measurement. RVSP was measured with 1.4F (for mice) or 3.5F (for rats) pressure transducer catheter (Millar Instruments), which was inserted into the right ventricle through right jugular vein of isoflurane-anesthetized rodents. The tracings were recorded and analyzed by AcqKnowledge software (Biopac Systems Inc.). Mouse/rat hearts were then dissected to calculate the RV/(LV+S) ratio for assessment of RV hypertrophy.

RNA sequencing and data analysis. After hemodynamic measurements, lung samples were collected for total RNA isolation with TRIzol (Invitrogen). RNA was further purified with RNeasy Mini Kit (QIAGEN). Each of the 4 samples from the same group were combined as 1 sample for RNA-sequencing analysis. All samples passed RNA purity before library construction through NanoDrop, agarose gel electrophoresis, and Agilent 2100 check. mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. RNA libraries were prepared by using New England BioLabs library, and the final library was obtained by PCR amplification and purification of PCR products by AMPure XP beads. After library construction, we diluted the library to 1.5 ng/μL with preliminary quantitative result by Qubit 2.0 and detected the insert size by Agilent 2100. Libraries were fed into HiSeq/MiSeq machines after pooling. Raw data were sequenced on Illumina HiSeq 2500/MiSeq platform and were transformed to sequenced reads by CASAVA base recognition. Raw data were stored in FASTQ format files and submitted to National Center for Biotechnology (NCBI) Sequence Read Archive (SRA) database (BioProject PRJNA994888). The sequenced reads containing low-quality reads were filtered to get the clean reads. Reads were mapped to the reference mouse genome by TopHat 21 to accomplish the alignment, and the annotated transcripts were assembled. Read counts were used to estimate gene expression level. Genes were identified as significantly differentially expressed at 2-fold change by using the DESeq R package with adjusted P values < 0.05 by the Benjamini-Hochberg approach for controlling the false discovery rate. The heatmap was generated by using HemI 2.0 (Heat Map Illustrator 2.0, https://hemi.biocuckoo.org/).

MCT-induced PAH rats and compound A-92 treatment. Sprague-Dawley rats (Charles River Laboratories) at the age of 6 weeks (150–170 g) were challenged with MCT subcutaneously at the dose of 33 mg/kg body weight (MedChemExpress, catalog HY-N0750). Fourteen days after MCT, the rats were randomized to receive either GCN2 inhibitor compound A-92 (Axon Medchem, catalog Axon 2720) (0.5 mg/kg/d, i.p.) or vehicle treatment for 14 days. Hemodynamic measurements were then carried out.

Primary cultures of HLMVECs. HLMVECs from controls (Lonza, catalog cc2527) were cultured in a T75 flask in endothelial basal medium (Lonza, catalog cc-3156) supplemented with 10% FBS and EGM-2 MV Microvascular Endothelial Cell Growth Medium SingleQuots supplements and growth factors (Lonza, catalog cc-4147). HLMVECs were used between passages 4 and 7. First, to determine the effects of hypoxia on GCN2 activation, HLMVECs were seeded in 6-well plates and grown to 80% confluence in EGM-2 medium, then replaced with hypoxia-equilibrated EGM-2 medium and exposed to hypoxia (1% O2) or replaced with fresh EGM-2 medium under normoxia for culturing. Hypoxic conditions were achieved in an O2 Control In vitro Glove Box (Coy Laboratory Products, Inc), which contains an oxygen sensor monitoring oxygen levels continuously. At various hours after exposure, cells were fixed for immunofluorescence staining to assess GCN2-Thr899 phosphorylation or lysed with RIPA buffer for Western blotting to assess GCN2-Thr899 phosphorylation and EIF2α-Ser51 phosphorylation. Second, to determine the effects of hypoxia on gene expression, HLMVECs were seeded and transfected with siRNA or plasmid in 6-well plates and cultured to reach approximately 70% confluence and then starved overnight in FBS-free endothelial cell basal medium. After starvation, cells were replaced with fresh serum-free basal medium and exposed to hypoxia or normoxia for 48 hours to assess EDN1 expression level.

Transfection with GCN2 siRNA and HIF plasmid DNA. GCN2 siRNA was synthesized from Integrated DNA Technologies (IDT) (sense: 5′-rGrCrArArUrUrCrUrGrUrGrGrUrGrCrArUrArATT-3′, antisense: 5′-rUrUrArUrGrCrArCrCrArCrArGrArArUrUrGrCTT-3′). HIF1A and HIF2A plasmids were obtained from Addgene (catalog 18949 and catalog 18950). Control vector plasmid was generated from enzyme digestion of HIF2A plasmid to remove HIF2A gene. Transfection of siRNA or plasmid DNA was performed using Lipofectamine 3000 transfection reagent (Invitrogen, catalog L3000-008) according to the manufacturer’s protocol. Briefly, 1 μL of 50 μM siRNA or control and 10 ng of each plasmid DNA were mixed with 3.75 μL of Lipofectamine 3000 in 250 μL Opti-MEM (Gibco, Catalog 31985070) and added to each well of a 6-well plate containing 1 mL of complete growth medium for culturing for 8 hours. The cells were then cultured in complete growth medium for 24 hours followed by starvation in basal medium without FBS for hypoxia challenge.

RNA isolation and quantitative RT-PCR analysis. Total RNA was isolated from cultured HLMVECs with RNeasy Mini Kit including DNase I digestion (QIAGEN). Total RNA from frozen mouse lung tissues and human lung tissues was isolated with TRIzol reagents followed by clean-up with RNeasy Mini Kit including DNase I digestion. One microgram of RNA was transcribed into cDNA with high-capacity cDNA reverse transcription kits (Applied Biosystems). Quantitative RT-PCR analysis was performed on Quant Studio 6 Flex system (Life Technologies) with the Fast SYBR Green Master Mix (Applied Biosystems, catalog 4385614). Target mRNA expression was determined by the comparative cycle threshold method of relative quantitation. We used 18S rRNA gene as an internal control for analysis of expression of human genes, while cyclophilin A (Ppia) was used for mouse genes. The primers listed below were synthesized by IDT: mPpia, forward primer 5′-GGCAAATGCTGGACCAAACAC-3′, and reverse primer, 5′-TTCCTGGACCCAAAACGCTC-3′; mGcn2, forward primer 5′-GTTCTCCCGGTACTTCATTGAG-3′; and reverse primer, 5′-TTGCAGGGTTGATAGGGATG-3′; mEdn1, forward primer 5′-TTCCCAATAAGGCCACAGACC-3′, and reverse primer, 5′-TTGGGCCCTGAGT-TCTTTTCC-3′; mPerk, forward primer 5′-CAGGCTTTTCCATCCTCAGC-3′, and reverse primer, 5′-GGCACTCACGGAGTCGTATTT-3′; hRNA18S, forward primer 5′-TTCCGACC-ATAAACGATGCCGA-3′, and reverse primer, 5′-GACTTTGGTTTCCCGGAAGCTG-3′; hGCN2, forward primer 5′-GAAGCTGTCAGCCAGCACTA-3′, and reverse primer, 5′-GGCAAGGGAGGTCTGAAGTC-3′; hHIF1A, forward primer 5′-TTACAGCAGCCAGACG-ATCATG-3′, and reverse primer, 5′-TGGTCAGCTGTGGTAATCCACT-3′; hHIF2A, forward primer 5′-CTGATGGCCATGAACAGCATCT-3′, and reverse primer, 5′-TCCTCGAAGTT-CTGATTCCCGA-3′; hEDN1, forward primer 5′-GTCTACTTCTGCCACCTGGAC-3′, and reverse primer, 5′-TCCAAGGCTCTCTTGGACCTA-3′.

Western blotting analysis. Cultured HLMVECs were lysed with RIPA buffer (Santa Cruz Biotechnology, catalog sc-24948) supplemented with PMSF, protease inhibitor, and phosphatase inhibitor cocktail (Santa Cruz Biotechnology, catalog sc-45044). Protein samples were subjected to SDS-PAGE. After electrophoresis, separated proteins were transferred onto nitrocellulose membranes. The membranes were blocked in 5% nonfat milk for 1 hour and subsequently incubated with diluted primary antibodies at 4°C overnight. After primary antibody incubation, membranes were rinsed with TBS containing 0.1% Tween 20 for 3 times and then incubated with peroxidase-conjugated secondary antibodies for 1 hour at RT. After 3 washes, the membrane was incubated with ECL for 5 minutes, and the protein bands were detected by Molecular Imager ChemiDoc XRS + with Image Lab Software (Bio-Rad Laboratories). The intensities of the protein bands were quantified by ImageJ (NIH). The following primary antibodies were used in this study: anti-GCN2 (Abcam, catalog ab134053, 1:500), anti-GCN2 (Cell Signaling Technology, catalog 3302, 1:500), anti–phospho-Thr899-GCN2 (Invitrogen, catalog PA5-105886, 1:500), anti–phospho-Ser51-EIF2α (Abcam, catalog ab32157, 1:1,000), anti-EIF2α (Cell Signaling Technology, catalog 9722, 1:1,000), anti–endothelin-1 (Abcam, catalog ab117757, 1:1,000), and anti–HIF-2α (Novus Biologicals, Bio-Techne, catalog NB100-122, 1:500). Anti–β-actin (MilliporeSigma, catalog A2228, 1:10,000) or anti–α-tubulin (MilliporeSigma, catalog T5168, 1:10,000) was used as a loading control.

ELISA analysis. HLMVECs were seeded and transfected with siGCN2 or control siRNA in 6-well plates and cultured to reach approximately 70% confluence and then starved overnight in FBS-free endothelial cell basal medium. After starvation, cells were replaced with fresh serum-free endothelial cell basal medium and exposed to hypoxia or normoxia for 48 hours. Culture media were collected for assessment of secreted EDN1 protein levels with Endothelin-1 ELISA Kit (R&D Systems, Bio-Techne, catalog DET100) according to the manufacturer’s instructions.

HLMVECs’ regulation of PASMC proliferation. As described above in the ELISA analysis section, culture media collected from HLMVECs under normoxia or 48-hour hypoxia were added to serum-starved PASMCs under normoxia. BrdU was added at the same time for assessing cell proliferation. At 24 hours later, PASMCs were fixed for anti-BrdU immunostaining (BD Biosciences, catalog 347580) according to the manufacturer’s instructions.

Histological assessment. Mouse lung tissues were fixed in situ with 10% PBS-buffered formalin for 5 minutes with 20 cm H2O pressure followed by overnight in 10% formalin at RT and then processed for paraffin embedding and sectioning. Semithin lung sections (5 μm) were deparaffinized, dehydrated, and subjected to H&E staining or Russell-Movat pentachrome staining (American MasterTech, catalog KTRMP) according to the manufacturer’s instructions. Sections were imaged with ZEISS Axio microscope. For assessment of pulmonary arterial wall thickness, pulmonary arteries from 50 images at 20× original magnification were quantified by ImageJ. Wall thickness was calculated by the distance between internal wall and external wall divided by the distance between external wall and the center of lumen.

Immunofluorescence staining. For immunofluorescence staining of anti-Thr899 phospho-GCN2, formalin-fixed lung sections from IPAH patients and control donors, mice, and rats were deparaffinized and rehydrated according to standard protocols. Antigen retrieval was performed by boiling the slides in target retrieval buffer (Advanced Cell Diagnostics, Bio-Techne) for 15 minutes, followed by Proteinase Plus (Advanced Cell Diagnostics, Bio-Techne) treatment at 40°C for 30 minutes. After blocking with 0.1% Triton X-100 and 5% normal goat serum in PBS for 1 hour at RT, the sections were incubated with anti-Thr899 phospho-GCN2 (Invitrogen, catalog PA5-105886, 1:50) and anti-vWF (Invitrogen, catalog MA5-14029, 1:5) for human lung tissue, anti-CD31 (BD Biosciences, catalog 550274, 1:25) for mouse lung tissue, or anti-CD31 (Bio-Rad, catalog MCA1334G, 1:10) for rat lung tissue at 4°C overnight. After 3 rinsings with PBS, the sections were incubated with Alexa Fluor 594–conjugated and Alexa Fluor 488–conjugated secondary antibodies (Life Technologies) at RT for 1 hour. The slides were mounted with antifade mounting medium containing DAPI (Vector Laboratories, catalog H-1200).

Mouse lung tissues were perfused with PBS and embedded in OCT for cryosectioning. For immunofluorescence staining of proliferative SMCs and ECs, lung cryosections (5 μm) were fixed with 4% paraformaldehyde for 20 minutes at RT and then blocked with 0.1% Triton X-100 and 5% BSA in PBS at RT for 1 hour. After 3 rinsings with PBS, slides were incubated with anti-Ki67 antibody (Invitrogen, catalog MA5-14520, 1:50) and anti–α-SMA (Abcam, catalog ab8211, 1:300) or anti-CD31 (R&D Systems, Bio-Techne, catalog AF3628, 1:100) at 4°C overnight, then incubated with Alexa Fluor 594–conjugated anti-rabbit IgG (Life Technologies, catalog A-11012, 1:300) and Alexa Fluor 488–conjugated anti-goat IgG (Life Technologies, catalog A-11055, 1:300) at RT for 1 hour. The nuclei were counterstained with DAPI contained in antifade mounting medium. Images were taken with the ZEISS confocal microscope (LSM880) equipped with a plan-Apochromat 20×/0.8 M27 and a plan-Apochromat 63×/1.4 Oil DIC M27 objective lens.

To quantify muscularization of distal pulmonary vessels, lung cryosections were immunostained with anti–α-SMA (Abcam, catalog ab8211, 1:300) and mounted with antifade mounting medium with DAPI. Images were taken with the ZEISS confocal microscope (LSM880) equipped with a plan-Apochromat 20×/0.8 M27 objective lens. Twenty pictures of 20× fields were taken for each section (5–6 sections for each group). The total number of α-SMA–positive distal pulmonary vessels (d ≤ 50 μm) for each section was used for each mouse.

Statistics. Prism 8 (GraphPad Software, Inc.) was used for statistical analysis. Two-group comparisons were analyzed by the unpaired 2-tailed t test for normal distribution or Mann-Whitney U test for non-normal distribution. Multiple-group comparisons were analyzed by 1-way ANOVA followed by Holm-Šídák, Tukey’s, or Dunnett’s multiple comparisons test or 2-way ANOVA followed by Tukey’s multiple comparisons test. P value less than 0.05 denoted the presence of a statistically significant difference. All dot plot figures represent means + SD.

Study approval. All animals were handled according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (National Academies Press, 2011), and the Northwestern University Institutional Animal Care and Use Committee approved protocols. Archived human lung tissue samples from patients with IPAH and control individuals were obtained from the Pulmonary Hypertension Breakthrough Initiative. The acquirement and handling of these samples were approved by the Institutional Review Board of Ann & Robert H. Lurie Children’s Hospital of Chicago.

Data availability. RNA-sequencing data were deposited at NCBI SRA database (SRA accession PRJNA994888). The data points shown in graphs are listed in an associated Supporting Data Values spreadsheet.