Background: Renal cell carcinoma (RCC) is a common malignant tumor of the urinary system. Angiogenesis is a main contributing factor for tumorigenesis. E74-like transcription factor 5 (ELF5) has been verified to participate in the progression of different cancers and can regulate angiogenesis. This study was aimed to explore the functions of ELF5 in RCC. Methods: Bioinformatics tools were used to predict the expression of ELF5 in RCC. RT-qPCR was applied for testing ELF5 expression in RCC cells. Cell behaviors were evaluated by colony formation, CCK-8, and transwell assays. The tube formation assay was used for determining angiogenesis. Methylation-specific PCR (MSP) was utilized for measuring the methylation level of ELF5 in RCC cells. ChIP and luciferase reporter assays were applied for assessing the binding of ELF5 and ubiquitin-specific protease 3 (USP3). Co-IP and GST pull-down were utilized for detecting the interaction of WD40 and tetratricopeptide repeats 1 (WDTC1) and USP3. Ubiquitination level of WDTC1 was determined by ubiquitination assay. Results: ELF5 was lowly expressed in RCC cells and tissues. High expression of ELF5 expression notably suppressed RCC cell proliferative, migratory, and invasive capabilities, and inhibited angiogenesis. The tumor growth in mice was inhibited by ELF5 overexpression. ELF5 was highly methylated in RCC samples, and DNA methyltransferases (DNMTs) can promote hypermethylation level of ELF5 in RCC cells. ELF5 was further proved to transcriptionally activate USP3 in RCC. Moreover, USP3 inhibited WDTC1 ubiquitination. ELF5 can promote USP3-mediated WDTC1 stabilization. Additionally, WDTC1 silencing reversed the functions of ELF5 overexpression on RCC progression. Conclusion: Downregulation of ELF5 due to DNA hypermethylation inhibits RCC development though the USP3/WDTC1axis in RCC.

Tumor microenvironment (TME)-induced nanocatalytic therapy is a trending strategy for tumor-targeting therapy, but the low catalytic efficiency remains to limit its therapeutic effect. The single-atom catalysts (SACs) appear as a novel type of nanozymes that possesses incredible catalytic activity. Here, we developed PEGylated manganese/iron-based SACs (Mn/Fe PSACs) by coordinating single-atom Mn/Fe to nitrogen atoms in hollow zeolitic imidazolate frameworks (ZIFs). Mn/Fe PSACs catalyze cellular hydrogen peroxide (H₂O₂) converting to hydroxyl radical (•OH) through a Fenton-like reaction; it also enhances the decomposition of H₂O₂ to O₂ that continuously converts to cytotoxic superoxide ion (•O₂⁻) via oxidase-like activity. Mn/Fe PSACs can reduce the depletion of reactive oxygen species (ROS) by consuming glutathione (GSH). Here, we demonstrated the Mn/Fe PSACs-mediated synergistic antitumor efficacy among in vitro and in vivo experiments. This study proposes new promising single-atom nanozymes with highly efficient biocatalytic sites and synergistic therapeutic effects, which will give birth to abundant inspirations in ROS-related biological applications in broad biomedical fields.

Rationale: Accurate diagnosis is crucial for predicting interstitial lung disease (ILD) progression and guiding treatment strategies. Surgical lung biopsy (SLB) has a high complication risk. Models for predicting post-SLB complications are lacking.Objectives: This multicenter study investigated high post-surgical lung biopsy complication risk factors and constructed a machine-learning model for predicting such events in patients with ILD.Methods: We retrospectively analyzed data from 302 patients who underwent SLB at five Chinese centers between January 2012 and July 2024. For feature selection, we combined multivariate analysis, least absolute shrinkage selection operator regression, and Random Forest out-of-bag error evaluation. Eight machine-learning algorithms were used for modeling. GridSearchCV was used for five-fold cross-validation for parameter optimization. Model performance was evaluated based on accuracy, precision, recall, F1-score, and area under the ROC curve (AUC) with 95%.Measurements and Main Results: Final ILD subtypes included nonspecific interstitial pneumonia, idiopathic pulmonary fibrosis, hypersensitivity pneumonitis, connective tissue disease–related ILD, sarcoidosis, unclassifiable ILD, and others; diagnoses were assigned by multidisciplinary discussion integrating clinical, HRCT, and histopathology. Of 302 included patients, 32.5% experienced serious adverse events. The cohort was split into a training set (n=127) and a test set (n=85); an external validation cohort (n=90) was drawn from two additional centers. Increased age, high smoking index, elevated Charlson Comorbidity Index, decreased forced vital capacity and diffusing capacity of the lungs for carbon monoxide, higher Quantitative Lung Fibrosis score, and long-term immunosuppressant use were independent predictors. The eXtreme Gradient Boosting algorithm performed best (accuracy: 0.92, test AUC: 0.88 [95% CI: 0.782-0.978]). External validation AUC was 0.86 [95% CI: 0.762-0.958], revealing model stability.Conclusion: We identified predictors of serious post-SLB complications in patients with ILD. The eXtreme Gradient Boosting model was accurate and reliable, as verified by decision curve analysis and external validation. This multicenter setting and external validation support potential clinical utility for pre-operative risk assessment