Molecular cartography uncovers evolutionary and microenvironmental dynamics in sporadic colorectal tumors.

Colorectal cancer exhibits dynamic cellular and genetic heterogeneity during progression from precursor lesions toward malignancy. Analysis of spatial multi-omic data from 31 human colorectal specimens enabled phylogeographic mapping of tumor evolution that revealed individualized progression trajectories and accompanying microenvironmental and clonal alterations. Phylogeographic mapping ordered genetic events, classified tumors by their evolutionary dynamics, and placed clonal regions along global pseudotemporal progression trajectories encompassing the chromosomal instability (CIN+) and hypermutated (HM) pathways. Integrated single-cell and spatial transcriptomic data revealed recurring epithelial programs and infiltrating immune states along progression pseudotime. We discovered an immune exclusion signature (IEX), consisting of extracellular matrix regulators DDR1, TGFBI, PAK4, and DPEP1, that charts with CIN+ tumor progression, is associated with reduced cytotoxic cell infiltration, and shows prognostic value in independent cohorts. This spatial multi-omic atlas provides insights into colorectal tumor-microenvironment co-evolution, serving as a resource for stratification and targeted treatments.

Differential pre-malignant programs and microenvironment chart distinct paths to malignancy in human colorectal polyps.

Colorectal cancers (CRCs) arise from precursor polyps whose cellular origins, molecular heterogeneity, and immunogenic potential may reveal diagnostic and therapeutic insights when analyzed at high resolution. We present a single-cell transcriptomic and imaging atlas of the two most common human colorectal polyps, conventional adenomas and serrated polyps, and their resulting CRC counterparts. Integrative analysis of 128 datasets from 62 participants reveals adenomas arise from WNT-driven expansion of stem cells, while serrated polyps derive from differentiated cells through gastric metaplasia. Metaplasia-associated damage is coupled to a cytotoxic immune microenvironment preceding hypermutation, driven partly by antigen-presentation differences associated with tumor cell-differentiation status. Microsatellite unstable CRCs contain distinct non-metaplastic regions where tumor cells acquire stem cell properties and cytotoxic immune cells are depleted. Our multi-omic atlas provides insights into malignant progression of colorectal polyps and their microenvironment, serving as a framework for precision surveillance and prevention of CRC.

Reassessment of Exosome Composition.

The heterogeneity of small extracellular vesicles and presence of non-vesicular extracellular matter have led to debate about contents and functional properties of exosomes. Here, we employ high-resolution density gradient fractionation and direct immunoaffinity capture to precisely characterize the RNA, DNA, and protein constituents of exosomes and other non-vesicle material. Extracellular RNA, RNA-binding proteins, and other cellular proteins are differentially expressed in exosomes and non-vesicle compartments. Argonaute 1-4, glycolytic enzymes, and cytoskeletal proteins were not detected in exosomes. We identify annexin A1 as a specific marker for microvesicles that are shed directly from the plasma membrane. We further show that small extracellular vesicles are not vehicles of active DNA release. Instead, we propose a new model for active secretion of extracellular DNA through an autophagy- and multivesicular-endosome-dependent but exosome-independent mechanism. This study demonstrates the need for a reassessment of exosome composition and offers a framework for a clearer understanding of extracellular vesicle heterogeneity.

Human SMARCA5 is continuously required to maintain nucleosome spacing.

Genetic models suggested that SMARCA5 was required for DNA-templated events including transcription, DNA replication, and DNA repair. We engineered a degron tag into the endogenous alleles of SMARCA5, a catalytic component of the imitation switch complexes in three different human cell lines to define the effects of rapid degradation of this key regulator. Degradation of SMARCA5 was associated with a rapid increase in global nucleosome repeat length, which may allow greater chromatin compaction. However, there were few changes in nascent transcription within the first 6 h of degradation. Nevertheless, we demonstrated a requirement for SMARCA5 to control nucleosome repeat length at G1/S and during the S phase. SMARCA5 co-localized with CTCF and H2A.Z, and we found a rapid loss of CTCF DNA binding and disruption of nucleosomal phasing around CTCF binding sites. This spatiotemporal analysis indicates that SMARCA5 is continuously required for maintaining nucleosomal spacing.

PAX3-FOXO1 coordinates enhancer architecture, eRNA transcription, and RNA polymerase pause release at select gene targets.

Transcriptional control is a highly dynamic process that changes rapidly in response to various cellular and extracellular cues, making it difficult to define the mechanism of transcription factor function using slow genetic methods. We used a chemical-genetic approach to rapidly degrade a canonical transcriptional activator, PAX3-FOXO1, to define the mechanism by which it regulates gene expression programs. By coupling rapid protein degradation with the analysis of nascent transcription over short time courses and integrating CUT&RUN, ATAC-seq, and eRNA analysis with deep proteomic analysis, we defined PAX3-FOXO1 function at a small network of direct transcriptional targets. PAX3-FOXO1 degradation impaired RNA polymerase pause release and transcription elongation at most regulated gene targets. Moreover, the activity of PAX3-FOXO1 at enhancers controlling this core network was surprisingly selective, affecting single elements in super-enhancers. This combinatorial analysis indicated that PAX3-FOXO1 was continuously required to maintain chromatin accessibility and enhancer architecture at regulated enhancers.

A pan-cancer transcriptome analysis of exitron splicing identifies novel cancer driver genes and neoepitopes.

Exitron splicing (EIS) creates a cryptic intron (called an exitron) within a protein-coding exon to increase proteome diversity. EIS is poorly characterized, but emerging evidence suggests a role for EIS in cancer. Through a systematic investigation of EIS across 33 cancers from 9,599 tumor transcriptomes, we discovered that EIS affected 63% of human coding genes and that 95% of those events were tumor specific. Notably, we observed a mutually exclusive pattern between EIS and somatic mutations in their affected genes. Functionally, we discovered that EIS altered known and novel cancer driver genes for causing gain- or loss-of-function, which promotes tumor progression. Importantly, we identified EIS-derived neoepitopes that bind to major histocompatibility complex (MHC) class I or II. Analysis of clinical data from a clear cell renal cell carcinoma cohort revealed an association between EIS-derived neoantigen load and checkpoint inhibitor response. Our findings establish the importance of considering EIS alterations when nominating cancer driver events and neoantigens.

METTL1-mediated m7G modification of Arg-TCT tRNA drives oncogenic transformation.

The emerging "epitranscriptomics" field is providing insights into the biological and pathological roles of different RNA modifications. The RNA methyltransferase METTL1 catalyzes N7-methylguanosine (m7G) modification of tRNAs. Here we find METTL1 is frequently amplified and overexpressed in cancers and is associated with poor patient survival. METTL1 depletion causes decreased abundance of m7G-modified tRNAs and altered cell cycle and inhibits oncogenicity. Conversely, METTL1 overexpression induces oncogenic cell transformation and cancer. Mechanistically, we find increased abundance of m7G-modified tRNAs, in particular Arg-TCT-4-1, and increased translation of mRNAs, including cell cycle regulators that are enriched in the corresponding AGA codon. Accordingly, Arg-TCT expression is elevated in many tumor types and is associated with patient survival, and strikingly, overexpression of this individual tRNA induces oncogenic transformation. Thus, METTL1-mediated tRNA modification drives oncogenic transformation through a remodeling of the mRNA "translatome" to increase expression of growth-promoting proteins and represents a promising anti-cancer target.

In vivo single-molecule analysis reveals COOLAIR RNA structural diversity.

Cellular RNAs are heterogeneous with respect to their alternative processing and secondary structures, but the functional importance of this complexity is still poorly understood. A set of alternatively processed antisense non-coding transcripts, which are collectively called COOLAIR, are generated at the Arabidopsis floral-repressor locus FLOWERING LOCUS C (FLC)1. Different isoforms of COOLAIR influence FLC transcriptional output in warm and cold conditions2-7. Here, to further investigate the function of COOLAIR, we developed an RNA structure-profiling method to determine the in vivo structure of single RNA molecules rather than the RNA population average. This revealed that individual isoforms of the COOLAIR transcript adopt multiple structures with different conformational dynamics. The major distally polyadenylated COOLAIR isoform in warm conditions adopts three predominant structural conformations, the proportions and conformations of which change after cold exposure. An alternatively spliced, strongly cold-upregulated distal COOLAIR isoform6 shows high structural diversity, in contrast to proximally polyadenylated COOLAIR. A hyper-variable COOLAIR structural element was identified that was complementary to the FLC transcription start site. Mutations altering the structure of this region changed FLC expression and flowering time, consistent with an important regulatory role of the COOLAIR structure in FLC transcription. Our work demonstrates that isoforms of non-coding RNA transcripts adopt multiple distinct and functionally relevant structural conformations, which change in abundance and shape in response to external conditions.

CYCLOIDEA-like genes control floral symmetry, floral orientation, and nectar guide patterning.

Actinomorphic flowers usually orient vertically (relative to the horizon) and possess symmetric nectar guides, while zygomorphic flowers often face horizontally and have asymmetric nectar guides, indicating that floral symmetry, floral orientation, and nectar guide patterning are correlated. The origin of floral zygomorphy is dependent on the dorsoventrally asymmetric expression of CYCLOIDEA (CYC)-like genes. However, how horizontal orientation and asymmetric nectar guides are achieved remains poorly understood. Here, we selected Chirita pumila (Gesneriaceae) as a model plant to explore the molecular bases for these traits. By analyzing gene expression patterns, protein-DNA and protein-protein interactions, and encoded protein functions, we identified multiple roles and functional divergence of 2 CYC-like genes, i.e. CpCYC1 and CpCYC2, in controlling floral symmetry, floral orientation, and nectar guide patterning. CpCYC1 positively regulates its own expression, whereas CpCYC2 does not regulate itself. In addition, CpCYC2 upregulates CpCYC1, while CpCYC1 downregulates CpCYC2. This asymmetric auto-regulation and cross-regulation mechanism might explain the high expression levels of only 1 of these genes. We show that CpCYC1 and CpCYC2 determine asymmetric nectar guide formation, likely by directly repressing the flavonoid synthesis-related gene CpF3'5'H. We further suggest that CYC-like genes play multiple conserved roles in Gesneriaceae. These findings shed light on the repeated origins of zygomorphic flowers in angiosperms.

Clinical targeting of HIV capsid protein with a long-acting small molecule.

Oral antiretroviral agents provide life-saving treatments for millions of people living with HIV, and can prevent new infections via pre-exposure prophylaxis1-5. However, some people living with HIV who are heavily treatment-experienced have limited or no treatment options, owing to multidrug resistance6. In addition, suboptimal adherence to oral daily regimens can negatively affect the outcome of treatment-which contributes to virologic failure, resistance generation and viral transmission-as well as of pre-exposure prophylaxis, leading to new infections1,2,4,7-9. Long-acting agents from new antiretroviral classes can provide much-needed treatment options for people living with HIV who are heavily treatment-experienced, and additionally can improve adherence10. Here we describe GS-6207, a small molecule that disrupts the functions of HIV capsid protein and is amenable to long-acting therapy owing to its high potency, low in vivo systemic clearance and slow release kinetics from the subcutaneous injection site. Drawing on X-ray crystallographic information, we designed GS-6207 to bind tightly at a conserved interface between capsid protein monomers, where it interferes with capsid-protein-mediated interactions between proteins that are essential for multiple phases of the viral replication cycle. GS-6207 exhibits antiviral activity at picomolar concentrations against all subtypes of HIV-1 that we tested, and shows high synergy and no cross-resistance with approved antiretroviral drugs. In phase-1 clinical studies, monotherapy with a single subcutaneous dose of GS-6207 (450 mg) resulted in a mean log10-transformed reduction of plasma viral load of 2.2 after 9 days, and showed sustained plasma exposure at antivirally active concentrations for more than 6 months. These results provide clinical validation for therapies that target the functions of HIV capsid protein, and demonstrate the potential of GS-6207 as a long-acting agent to treat or prevent infection with HIV.

Mettl1/Wdr4-Mediated m7G tRNA Methylome Is Required for Normal mRNA Translation and Embryonic Stem Cell Self-Renewal and Differentiation.

tRNAs are subject to numerous modifications, including methylation. Mutations in the human N7-methylguanosine (m7G) methyltransferase complex METTL1/WDR4 cause primordial dwarfism and brain malformation, yet the molecular and cellular function in mammals is not well understood. We developed m7G methylated tRNA immunoprecipitation sequencing (MeRIP-seq) and tRNA reduction and cleavage sequencing (TRAC-seq) to reveal the m7G tRNA methylome in mouse embryonic stem cells (mESCs). A subset of 22 tRNAs is modified at a "RAGGU" motif within the variable loop. We observe increased ribosome occupancy at the corresponding codons in Mettl1 knockout mESCs, implying widespread effects on tRNA function, ribosome pausing, and mRNA translation. Translation of cell cycle genes and those associated with brain abnormalities is particularly affected. Mettl1 or Wdr4 knockout mESCs display defective self-renewal and neural differentiation. Our study uncovers the complexity of the mammalian m7G tRNA methylome and highlights its essential role in ESCs with links to human disease.

PRMD: an integrated database for plant RNA modifications.

The scope and function of RNA modifications in model plant systems have been extensively studied, resulting in the identification of an increasing number of novel RNA modifications in recent years. Researchers have gradually revealed that RNA modifications, especially N6-methyladenosine (m6A), which is one of the most abundant and commonly studied RNA modifications in plants, have important roles in physiological and pathological processes. These modifications alter the structure of RNA, which affects its molecular complementarity and binding to specific proteins, thereby resulting in various of physiological effects. The increasing interest in plant RNA modifications has necessitated research into RNA modifications and associated datasets. However, there is a lack of a convenient and integrated database with comprehensive annotations and intuitive visualization of plant RNA modifications. Here, we developed the Plant RNA Modification Database (PRMD; http://bioinformatics.sc.cn/PRMD and http://rnainformatics.org.cn/PRMD) to facilitate RNA modification research. This database contains information regarding 20 plant species and provides an intuitive interface for displaying information. Moreover, PRMD offers multiple tools, including RMlevelDiff, RMplantVar, RNAmodNet and Blast (for functional analyses), and mRNAbrowse, RNAlollipop, JBrowse and Integrative Genomics Viewer (for displaying data). Furthermore, PRMD is freely available, making it useful for the rapid development and promotion of research on plant RNA modifications.

Structure-based discovery of potent WD repeat domain 5 inhibitors that demonstrate efficacy and safety in preclinical animal models.

WD repeat domain 5 (WDR5) is a core scaffolding component of many multiprotein complexes that perform a variety of critical chromatin-centric processes in the nucleus. WDR5 is a component of the mixed lineage leukemia MLL/SET complex and localizes MYC to chromatin at tumor-critical target genes. As a part of these complexes, WDR5 plays a role in sustaining oncogenesis in a variety of human cancers that are often associated with poor prognoses. Thus, WDR5 has been recognized as an attractive therapeutic target for treating both solid and hematological tumors. Previously, small-molecule inhibitors of the WDR5-interaction (WIN) site and WDR5 degraders have demonstrated robust in vitro cellular efficacy in cancer cell lines and established the therapeutic potential of WDR5. However, these agents have not demonstrated significant in vivo efficacy at pharmacologically relevant doses by oral administration in animal disease models. We have discovered WDR5 WIN-site inhibitors that feature bicyclic heteroaryl P7 units through structure-based design and address the limitations of our previous series of small-molecule inhibitors. Importantly, our lead compounds exhibit enhanced on-target potency, excellent oral pharmacokinetic (PK) profiles, and potent dose-dependent in vivo efficacy in a mouse MV4:11 subcutaneous xenograft model by oral dosing. Furthermore, these in vivo probes show excellent tolerability under a repeated high-dose regimen in rodents to demonstrate the safety of the WDR5 WIN-site inhibition mechanism. Collectively, our results provide strong support for WDR5 WIN-site inhibitors to be utilized as potential anticancer therapeutics.

H2B oncohistones cause homologous recombination defect and genomic instability through reducing H2B monoubiquitination in Schizosaccharomyces pombe.

Canonical oncohistones are histone H3 mutations in the N-terminal tail associated with tumors and affect gene expression by altering H3 post-translational modifications (PTMs) and the epigenetic landscape. Noncanonical oncohistone mutations occur in both tails and globular domains of all four core histones and alter gene expression by perturbing chromatin remodeling. However, the effects and mechanisms of noncanonical oncohistones remain largely unknown. Here we characterized 16 noncanonical H2B oncohistones in the fission yeast Schizosaccharomyces pombe. We found that seven of them exhibited temperature sensitivities and 11 exhibited genotoxic sensitivities. A detailed study of two of these onco-mutants H2BG52D and H2BP102L revealed that they were defective in homologous recombination (HR) repair with compromised histone eviction and Rad51 recruitment. Interestingly, their genotoxic sensitivities and HR defects were rescued by the inactivation of the H2BK119 deubiquitination function of Ubp8 in the Spt-Ada-Gcn5-Acetyltransferase (SAGA) complex. The levels of H2BK119 monoubiquitination (H2Bub) in the H2BG52D and H2BP102L mutants are reduced in global genome and at local DNA break sites presumably due to enhanced recruitment of Ubp8 onto nucleosomes and are recovered upon loss of H2B deubiquitination function of the SAGA complex. Moreover, H2BG52D and H2BP102L heterozygotes exhibit genotoxic sensitivities and reduced H2Bub in cis. We therefore conclude that H2BG52D and H2BP102L oncohistones affect HR repair and genome stability via the reduction of H2Bub and propose that other noncanonical oncohistones may also affect histone PTMs to cause diseases.

Deep Learning for Echo Analysis, Tracking, and Evaluation of Mitral Regurgitation (DELINEATE-MR).

BACKGROUND: Artificial intelligence, particularly deep learning (DL), has immense potential to improve the interpretation of transthoracic echocardiography (TTE). Mitral regurgitation (MR) is the most common valvular heart disease and presents unique challenges for DL, including the integration of multiple video-level assessments into a final study-level classification. METHODS: A novel DL system was developed to intake complete TTEs, identify color MR Doppler videos, and determine MR severity on a 4-step ordinal scale (none/trace, mild, moderate, and severe) using the reading cardiologist as a reference standard. This DL system was tested in internal and external test sets with performance assessed by agreement with the reading cardiologist, weighted κ, and area under the receiver-operating characteristic curve for binary classification of both moderate or greater and severe MR. In addition to the primary 4-step model, a 6-step MR assessment model was studied with the addition of the intermediate MR classes of mild-moderate and moderate-severe with performance assessed by both exact agreement and ±1 step agreement with the clinical MR interpretation. RESULTS: A total of 61 689 TTEs were split into train (n=43 811), validation (n=8891), and internal test (n=8987) sets with an additional external test set of 8208 TTEs. The model had high performance in MR classification in internal (exact accuracy, 82%; κ=0.84; area under the receiver-operating characteristic curve, 0.98 for moderate/severe MR) and external test sets (exact accuracy, 79%; κ=0.80; area under the receiver-operating characteristic curve, 0.98 for moderate or greater MR). Most (63% internal and 66% external) misclassification disagreements were between none/trace and mild MR. MR classification accuracy was slightly higher using multiple TTE views (accuracy, 82%) than with only apical 4-chamber views (accuracy, 80%). In subset analyses, the model was accurate in the classification of both primary and secondary MR with slightly lower performance in cases of eccentric MR. In the analysis of the 6-step classification system, the exact accuracy was 80% and 76% with a ±1 step agreement of 99% and 98% in the internal and external test set, respectively. CONCLUSIONS: This end-to-end DL system can intake entire echocardiogram studies to accurately classify MR severity and may be useful in helping clinicians refine MR assessments.

BRD9-SMAD2/3 Orchestrates Stemness and Tumorigenesis in Pancreatic Ductal Adenocarcinoma.

BACKGROUND & AIMS: The dismal prognosis of pancreatic ductal adenocarcinoma (PDAC) is linked to the presence of pancreatic cancer stem-like cells (CSCs) that respond poorly to current chemotherapy regimens. The epigenetic mechanisms regulating CSCs are currently insufficiently understood, which hampers the development of novel strategies for eliminating CSCs. METHODS: By small molecule compound screening targeting 142 epigenetic enzymes, we identified that bromodomain-containing protein BRD9, a component of the BAF histone remodeling complex, is a key chromatin regulator to orchestrate the stemness of pancreatic CSCs via cooperating with the TGFß/Activin-SMAD2/3 signaling pathway. RESULTS: Inhibition and genetic ablation of BRD9 block the self-renewal, cell cycle entry into G0 phase and invasiveness of CSCs, and improve the sensitivity of CSCs to gemcitabine treatment. In addition, pharmacological inhibition of BRD9 significantly reduced the tumorigenesis in patient-derived xenografts mouse models and eliminated CSCs in tumors from pancreatic cancer patients. Mechanistically, inhibition of BRD9 disrupts enhancer-promoter looping and transcription of stemness genes in CSCs. CONCLUSIONS: Collectively, the data suggest BRD9 as a novel therapeutic target for PDAC treatment via modulation of CSC stemness.

Scalable and model-free detection of spatial patterns and colocalization.

The expeditious growth in spatial omics technologies enables the profiling of genome-wide molecular events at molecular and single-cell resolution, highlighting a need for fast and reliable methods to characterize spatial patterns. We developed SpaGene, a model-free method to discover spatial patterns rapidly in large-scale spatial omics studies. Analyzing simulation and a variety of spatially resolved transcriptomics data showed that SpaGene is more powerful and scalable than existing methods. Spatial expression patterns identified by SpaGene reconstruct unobserved tissue structures. SpaGene also successfully discovers ligand-receptor interactions through their colocalization.

Multi-task bioassay pre-training for protein-ligand binding affinity prediction.

Protein-ligand binding affinity (PLBA) prediction is the fundamental task in drug discovery. Recently, various deep learning-based models predict binding affinity by incorporating the three-dimensional (3D) structure of protein-ligand complexes as input and achieving astounding progress. However, due to the scarcity of high-quality training data, the generalization ability of current models is still limited. Although there is a vast amount of affinity data available in large-scale databases such as ChEMBL, issues such as inconsistent affinity measurement labels (i.e. IC50, Ki, Kd), different experimental conditions, and the lack of available 3D binding structures complicate the development of high-precision affinity prediction models using these data. To address these issues, we (i) propose Multi-task Bioassay Pre-training (MBP), a pre-training framework for structure-based PLBA prediction; (ii) construct a pre-training dataset called ChEMBL-Dock with more than 300k experimentally measured affinity labels and about 2.8M docked 3D structures. By introducing multi-task pre-training to treat the prediction of different affinity labels as different tasks and classifying relative rankings between samples from the same bioassay, MBP learns robust and transferrable structural knowledge from our new ChEMBL-Dock dataset with varied and noisy labels. Experiments substantiate the capability of MBP on the structure-based PLBA prediction task. To the best of our knowledge, MBP is the first affinity pre-training model and shows great potential for future development. MBP web-server is now available for free at: https://huggingface.co/spaces/jiaxianustc/mbp.

scKWARN: Kernel-weighted-average robust normalization for single-cell RNA-seq data.

MOTIVATION: Single-cell RNA-seq normalization is an essential step to correct unwanted biases caused by sequencing depth, capture efficiency, dropout, and other technical factors. Existing normalization methods primarily reduce biases arising from sequencing depth by modeling count-depth relationship and/or assuming a specific distribution for read counts. However, these methods may lead to over or under-correction due to presence of technical biases beyond sequencing depth and the restrictive assumption on models and distributions. RESULTS: We present scKWARN, a Kernel Weighted Average Robust Normalization designed to correct known or hidden technical confounders without assuming specific data distributions or count-depth relationships. scKWARN generates a pseudo expression profile for each cell by borrowing information from its fuzzy technical neighbors through a kernel smoother. It then compares this profile against the reference derived from cells with the same bimodality patterns to determine the normalization factor. As demonstrated in both simulated and real datasets, scKWARN outperforms existing methods in removing a variety of technical biases while preserving true biological heterogeneity. AVAILABILITY AND IMPLEMENTATION: scKWARN is freely available at https://github.com/cyhsuTN/scKWARN.

Ferroptosis Contributes to Microvascular Dysfunction in Diabetic Retinopathy.

Ferroptosis is a new form of cell death characterized by iron-dependent lipid peroxidation. Whether ferroptosis is involved in retinal microvascular dysfunction under diabetic condition is not known. Herein, the expression of ferroptosis-related genes in patients with proliferative diabetic retinopathy and in diabetic mice was determined with quantitative RT-PCR. Reactive oxygen species, iron content, lipid peroxidation products, and ferroptosis-associated proteins in the cultured human retinal microvascular endothelial cells (HRMECs) and in the retina of diabetic mice were examined. The association of ferroptosis with the functions of endothelial cells in vitro was evaluated. After administration of ferroptosis-specific inhibitor, Fer-1, the retinal microvasculature in diabetic mice was assessed. Characteristic changes of ferroptosis-associated markers, including glutathione peroxidase 4, ferritin heavy chain 1, long-chain acyl-CoA synthetase 4, transferrin receptor protein 1, and cyclooxygenase-2, were detected in the retinal fibrovascular membrane of patients with proliferative diabetic retinopathy, cultured HRMECs, and the retina of diabetic mice. Elevated levels of reactive oxygen species, lipid peroxidation, and iron content were found in the retina of diabetic mice and in cultured HRMECs. Ferroptosis was found to be associated with HRMEC dysfunction under high-glucose condition. Inhibition of ferroptosis with specific inhibitor Fer-1 in diabetic mice significantly reduced the severity of retinal microvasculopathy. Ferroptosis contributes to microvascular dysfunction in diabetic retinopathy, and inhibition of ferroptosis might be a promising strategy for the therapy of early-stage diabetic retinopathy.