2026年 · 第09卷第03期

Nogo-A is well known as an inhibitor of neurite outgrowth. Accumulating evidence suggests that it may also participate in the pathogenesis of non-small cell lung cancer (NSCLC), an aggressive malignancy with a 5-year survival rate of only 16.6%. However, the role of Nogo-A in NSCLC remains poorly understood, partly due to the potential cross-reactivity of antibodies targeting different Nogo isoforms, which may lead to ambiguous results. Here, we employed bioinformatics analyses to identify candidate binding proteins of Nogo-A, followed by coimmunoprecipitation (Co-IP) for experimental validation. A 2-step protein-protein docking procedure combined with molecular dynamics (MD) simulations was subsequently performed to explore the possible binding mechanisms between Nogo-A and CERT1. In addition, a protein-protein interaction network was constructed to explore the possible functional role of Nogo-A in NSCLC. The results revealed that Nogo-A expression was elevated in NSCLC tissues compared with normal tissues. CERT1, SNX15, and SYT16 were identified as candidate Nogo-A-binding proteins, among which CERT1 was experimentally confirmed to bind with Nogo-A in H1299 NSCLC cells overexpressing Nogo-A. MD simulations further suggested that Nogo-A might bind to CERT1 via its reticulon homology domain (RHD), with the interaction involving both the PH and START domains of CERT1. Network enrichment analysis indicated that Nogo-A might be involved in apoptotic processes through the modulation of CERT1-mediated ceramide binding and transport. Taken together, our findings identify CERT1 as a novel putative interacting partner of Nogo-A in NSCLC. The predicted binding interface and its potential functional association with apoptosis provide a basis for future mechanistic studies.

Background: Osteosarcoma is a relatively rare malignant tumor characterized by high aggressiveness and poor prognosis. Traditional diagnostic approaches have several limitations, particularly in early detection and disease monitoring. In recent years, exosome-derived RNAs have attracted increasing attention in cancer research because of their potential relevance to intercellular communication and tumor-related molecular alterations. However, the molecular characteristics associated with plasma exosomal RNAs in osteosarcoma remain poorly understood. Methods: In this study, RNA expression in plasma exosomes from osteosarcoma patients was investigated using differential expression, enrichment, immune infiltration, and competing endogenous RNA (ceRNA) network analyses, followed by in vitro assessment of both RNA and protein expression levels. Results: ceRNA network analysis revealed 7 candidate core molecules and 2 putative ceRNA axes—hsa_circ_0015475/hsa-miR-4753-3p/retinoic acid receptor-related orphan receptor alpha (RORA) and growth arrest-specific 5 (GAS5)/hsa-miR-10a-5p/RORA—which converge on the common messenger RNA RORA. In vitro assessment revealed that RORA messenger RNA and protein expression levels were reduced in osteosarcoma cell lines. A pan-cancer analysis further revealed that RORA was differentially expressed across multiple cancer types. Conclusion: This study revealed 2 potential RORA-related ceRNA axes, hsa_circ_0015475/hsa-miR-4753-3p/RORA and GAS5/hsa-miR-10a-5p/RORA, that may be associated with osteosarcoma-related molecular alterations. The reduced expression of RORA in osteosarcoma cell lines suggests its potential relevance to osteosarcoma biology.

This study presents a systematic evaluation of classical machine learning, deep learning, and hybrid models for heart disease prediction using the University of California, Irvine dataset. A standardized framework with consistent preprocessing and 5-fold cross-validation was applied to ensure fair comparison. The results indicate that the hybrid models achieved competitive performance across multiple metrics, with the proposed convolutional neural network (CNN)-Attention-eXtreme Gradient Boosting (XGBoost) ensemble obtaining an accuracy of 82.07% ± 2.68% and an F1 score of 81.78% ± 2.68%. However, CNN + Support Vector Machine (SVM) achieved the highest accuracy (83.37% ± 2.11%) among all models, whereas Random Forest achieved the highest area under the curve (89.36% ± 1.42%) and showed strong calibration performance. In terms of probabilistic calibration, Brier Skill Score (BSS) analysis showed that all models outperformed the baseline (BSS > 0), with CNN + SVM (0.477), SVM (0.474), and Random Forest (0.465) achieving the highest calibration skill. Statistical analysis using the Friedman test and post hoc Wilcoxon signed-rank tests with Bonferroni correction revealed no statistically significant pairwise differences among the models, suggesting that the observed performance differences were marginal despite numerical variations. Computational analysis highlighted a trade-off between performance and efficiency, as hybrid models required longer training times (up to 7.59 s per fold) than lightweight tree-based models (<0.3 s). A cross-country evaluation further revealed reduced generalizability under domain shift.

This study proposes a hip implant designed with 6 layers of functionally graded polymers in graduated volumetric ratios, namely, polylactic acid (PLA) and ultra-high-molecular-weight polyethylene (UHMWPE). In addition, polyvinyl alcohol (PVA) was introduced as a porous PLA/UHMWPE/PVA inner layer.This design aims to promote osseointegration and reduce stress variation with age resulting from patient differences. The effective modulus of elasticity and density of each layer were assessed using the blending rule, while the properties of the layers were estimated using the Gibson-Ashby foam model. The force-law function was used to describe the continuous change in the modulus of elasticity across thickness. Changes in body weight, contact forces at the hip joint, and activity level were derived from previous studies. Hip implants made from a functionally graded material exhibited a median stress lifetime (approximately 29% to 34%) and significantly lower standard deviation than those made from homogeneous PLA. One-way analysis of variance (ANOVA) confirmed that the differences in the modulus of elasticity, density, stiffness, and abrasion parameters across the graded layers were statistically significant (P < 0.05). The effect size was very large (Cohen’s coefficient d = 4.60), indicating a substantial mechanical improvement owing to the stepped structure, with a 95% confidence interval between 3.35 and 5.85. These findings confirm that the proposed functionally graded porous structure offers improved mechanical compatibility with human bone.

Background: Colorectal cancer (CRC) progression involves dynamic transitions between proliferative and migratory tumor states, often described as a proliferation-migration trade-off. However, tumor cells may adopt hybrid phenotypes reflecting phenotypic plasticity. The RNA binding protein Ewing sarcoma breakpoint region 1 (EWSR1) regulates transcription and RNA processing, but its role in coordinating these phenotypes in fusion-negative CRC remains unclear. Methods: An integrative transcriptomic analysis of 988 CRC tumors from 3 cohorts (The Cancer Genome Atlas, Clinical Proteomic Tumor Analysis Consortium, and Sidra-Leiden University Medical Center) was performed. Proliferation, epithelial-mesenchymal transition (EMT), and stromal programs were quantified using single-sample gene set enrichment analysis following ComBat batch correction. EMT scores were residualized against stromal signatures to derive tumor-intrinsic migration signals. Associations were assessed using correlation, multivariable regression, and interaction modeling. Results: Raw EMT scores were strongly correlated with stromal signatures (r = 0.964, P < 0.001). After adjustment, a clear inverse relationship between proliferation and migration emerged (r = -0.226, P < 0.001). EWSR1 expression was strongly associated with proliferation (r = 0.341, P < 0.001) and modestly associated with migration (r = 0.076, P = 0.017) but not with stromal content. In multivariable models, proliferation negatively predicted migration (β = -0.286, P < 0.001), whereas EWSR1 positively predicted migration (β = 0.174, P < 0.001). The interaction was not significant (P = 0.630), although stratified analyses revealed attenuation of the trade-off in EWSR1-high tumors. Conclusion: EWSR1 expression was associated with migration-related activity without coupling proliferation and migration, instead the relationships among ESWR1 expression, migration and proliferation suggest a role for EWSR1 in phenotypic plasticity through partial relaxation of the proliferation-migration trade-off in CRC.

Hydrogels are 3-dimensional (3D), hydrophilic polymeric networks capable of absorbing and retaining substantial amounts of water while preserving their structural integrity. Due to their high biocompatibility, tunable physicochemical properties, and structural resemblance to the natural extracellular matrix, hydrogels have emerged as promising candidates for a wide spectrum of biomedical applications. This review presents a comprehensive examination of recent advances in hydrogel design, synthesis, and functionalization strategies, including the development of smart hydrogels with stimuli-responsive, self-healing, and conductive properties. Cutting-edge fabrication techniques, such as 3D bioprinting, microfluidic templating, and photopatterning, are discussed in relation to enhancing hydrogel functionality for targeted clinical use. A special emphasis is placed on their applications in drug and gene delivery, tissue engineering, and regenerative medicine, with a focus on challenging medical conditions such as bone tissue engineering, cancer therapy, liver regeneration, myocardial infarct repair, diabetic wound healing, and lung tissue reconstruction. Furthermore, this review explores the translational landscape, including ongoing clinical trials, regulatory hurdles, and commercialization efforts. Emerging directions such as 4D dynamic hydrogels, artificial-intelligence-guided hydrogel optimization, and integration with wearable and implantable biomedical devices are also highlighted. By connecting fundamental materials science with critical medical needs, this review highlights the transformative potential of hydrogels in shaping the future of personalized and regenerative healthcare.

Rheumatoid arthritis (RA) is a recurring and chronic autoimmune disorder that causes joint pain, stiffness, and inflammation because the immune system attacks the joints. RA is a disease that affects the quality of life of nearly 1% of the entire population worldwide. The gastrointestinal tract (GIT) is also affected by the side effects associated with conventional treatments, oral and injectable medications, poor adherence to treatment, and low long-term treatment responses. A widespread unofficially recognized use of salicylic acid is as a skin analgesic and anti-inflammatory agent in the form of methyl salicylate, an ester form of salicylic acid that inhibits the enzyme cyclooxygenase and the synthesis of prostaglandins. It, however, has restrictions in usage, as although its toxicity is low, it has poor skin penetration. Newer systems of drug delivery to the skin, such as nanogels, can increase methyl salicylate bioavailability, solubility, and skin retention and achieve controlled release to minimize systemic toxicity. This paper discusses the epidemiology and pathogenesis of RA, the pharmacology of salic/meprobamate drugs, nanogel formulation procedures, mechanisms of skin interactions, drug release, and drug tests. New patents and commercial preparations are addressed, and the use of methyl salicylate-impregnated nanogels as new transdermal drugs is discussed. Overall, this treatment method may revolutionize RA management since there are no side effects, the treatment is effective, and this solution is more friendly than conventional methods are.

Quantum-scale photomolecular genetic editing is emerging as a precise, noncontact strategy for modulating endothelial function through controlled light-matter interactions. This paradigm integrates photonics, quantum-scale energy transfer, and molecular genetics to regulate intracellular signaling, gene expression, and barrier dynamics with high spatial and temporal fidelity. Red and near-infrared photons interact with chromophores, delocalized electron clouds, and mitochondrial respiratory complexes to generate quantized electronic excitations that propagate toward the transcriptional and epigenetic machinery. In endothelial cells, these processes influence oxidative stress, calcium flux, cytoskeletal restructuring, and tight junction integrity—the core determinants of vascular homeostasis and blood-tissue interface stability. Quantum-scale photomolecular editing enables reversible, tunable, and dose-dependent genetic modulation without exogenous vectors or permanent genomic alteration. This review synthesizes advances in photobiomodulation, quantum biology, and endothelial mechanobiology to outline mechanistic frameworks that connect photon absorption to molecular genetic outcomes. We discuss implications for neurovascular disorders, inflammation, and regenerative repair, and we highlight challenges in standardization, restructuring, and translational deployment.