Recent evidence suggests a potential role of immune cell activity in the development of sarcopenia, but the causal relationship remains unclear. To address this, we conducted a bidirectional 2-sample Mendelian randomization (MR) analysis to evaluate whether immune cell traits causally influence sarcopenia. Genetic instruments for 731 immune cell traits were derived from a cohort of 3757 individuals of European ancestry, and sarcopenia-related data (including hand grip strength, appendicular lean mass, and walking pace—) were obtained from large-scale genome-wide association studies of European populations. Causal effects were estimated using inverse-variance weighted and weighted median MR, with MR-Egger regression applied for sensitivity analyses. Our results identified several immune cell traits with significant causal effects on sarcopenia. Increased expression of human leukocyte antigen-DR on plasmacytoid and total conventional dendritic cells, as well as human leukocyte antigen-DR on cluster of differentiation 33 ⁻ myeloid cells, was associated with lower appendicular lean mass, slower walking pace, and higher risk of low hand grip strength. Conversely, certain lymphoid cell traits, including higher numbers of specific T and B cell subsets, were associated with protective effects against sarcopenia. Reverse MR analyses indicated no strong causal effects of sarcopenia traits on immune cell profiles. These findings provide robust genetic evidence supporting a causal link between immune cell composition and sarcopenia, highlighting myeloid cells as potential contributors to sarcopenia development, while lymphoid cells may exert protective effects. This insight advances our understanding of sarcopenia pathophysiology and may inform future preventive and therapeutic strategies.

Drug resistance in acute promyelocytic leukemia (APL) remains clinically challenging despite the success of all-trans retinoic acid (ATRA) and arsenic trioxide (ATO) therapy. Existing explanations, including promyelocytic leukemia protein–retinoic acid receptor alpha (PML–RARα) mutations and alterations in individual death pathways, do not fully account for the heterogeneity of resistant phenotypes. This review proposes regulated cell death network (RCDN) remodeling as a testable framework for APL resistance. RCDN remodeling is defined as coordinated changes in network coupling, stress routing, and cellular death thresholds. We hypothesize that resistance emerges from coordinated adaptation across redox, survival signaling, and mitochondrial gating, rather than from defects in any single pathway. This predicts coupled shifts across multiple death modules and non-additive responses to combinatorial perturbations, distinguishing network-level remodeling from pathway-specific resistance. We further define experimental criteria for distinguishing network-level remodeling from pathway-specific resistance and discuss Tanshinone IIA (Tan IIA) as a preclinical multi-node perturbation probe.

The antioxidant N-(1,3-dimethylbutyl)-N'-phenyl-p-phenylenediamine (6PPD), widely used in tire rubber production, and its oxidative transformation product 6PPD-quinone (6PPD-Q), have become emerging urban environmental pollutants distributed worldwide, which can cause respiratory exposure in humans through atmospheric particulate matter and aerosols, potentially damaging the respiratory system; moreover, studies have suggested that such exposure is associated with the occurrence and development of lung cancer. In this study, a multidimensional computational research framework integrating network toxicology, weighted gene co-expression network analysis (WGCNA), machine learning, molecular docking, and molecular dynamics simulation was constructed to systematically decipher the dynamic molecular mechanisms by which 6PPD regulates lung cancer progression. Through cross-screening of multiple databases, 35 common interaction targets between 6PPD and lung cancer were identified, and candidate hub genes were screened out. GO and KEGG analyses showed that the targets were enriched in cancer-related pathways such as chemical carcinogenesis-reactive oxygen species and VEGF, suggesting that oxidative stress may be the core regulatory direction. The Ridge machine learning prediction model achieved an accuracy of 1.000 (AUC = 0.985) in the training set, and also exhibited good generalization ability in two independent validation datasets, and further screened out CA4, CDC25C, and NQO1 as core biomarkers; all three were significantly overexpressed in lung cancer tissues and exhibited good diagnostic efficacy for lung cancer. Molecular docking and 100 ns molecular dynamics simulation showed that 6PPD possessed strong spontaneous binding ability toward NOX4 (−8.0 kcal/mol), NQO1 (−7.9 kcal/mol), KDR, CA4, and CDC25C, and the complexes formed with the core targets remained generally stable throughout the simulations, suggesting that 6PPD can persistently bind to target proteins through stable intermolecular interactions. Trajectory hierarchical clustering analysis further showed that the conformational dynamic characteristics of each complex differed; among them, the KDR conformation was the most stable, whereas NQO1 and CA4 showed higher conformational diversity. In summary, the results of this study suggest that 6PPD may participate in oxidative stress pathway activation, DNA damage, cell cycle regulation, and tumor angiogenesis through multi-target synergistic effects, and be associated with lung cancer progression; this provides a computational pathway for deciphering the carcinogenic molecular mechanisms of environmental pollutants and assessing respiratory carcinogenic risk, and the core targets may provide candidate molecular markers for early warning and intervention of lung cancer.

Background: Obesity-associated cognitive vulnerability involves peripheral metabolic organs, the circulation, and the neurovascular unit/blood–brain barrier (NVU/BBB). Whether adipose and hepatic dysfunction contribute through compartment-specific redox remodelling remains unclear. Methods: We searched PubMed/MEDLINE, Web of Science Core Collection, and Scopus for human, rodent, and cellular studies published from 1 January 2017 to 13 September 2026. Ninety-seven publications met the inclusion criteria. No meta-analysis or formal risk-of-bias assessment was undertaken. Results: Human studies associated adipose or hepatic dysfunction and circulating redox markers with brain and cognitive outcomes but rarely established tissue origin, temporal ordering, or mediation. Animal and cellular studies supported adipose mitochondrial and NOX-derived ROS, MASLD/MASH-related hepatic redox dysfunction, liver-derived signals, NVU/BBB impairment, and brain mitochondrial, inflammatory, synaptic, and cognitive changes. Conclusions: The adipose–liver–brain network is biologically plausible but remains unproven as a causal model. Future longitudinal and intervention studies should establish compartment specificity, temporal ordering, mediator reproducibility, reversibility, and domain-specific cognitive effects.