Postmenopausal osteoporosis, marked by diminished bone regenerative capacity and elevated fracture risk, urgently requires innovative osteoanabolic strategies. Plant-derived extracellular vesicle (PDEV)-like particles (EVLPs) have emerged as promising therapeutic candidates due to their bioactive cargo and cross-kingdom regulatory potential. Building on the traditional use of Dipsaci Radix (DR) in bone repair, this study investigates DR-derived EVLPs (DREVLPs) as novel bone-forming nanotherapeutics. We successfully isolated and characterized DREVLPs, demonstrating their remarkable osteogenic capacity through bone morphogenetic protein 2 (BMP2) pathway activation. In bone marrow mesenchymal stem cells (BMSCs), DREVLPs significantly upregulated RUNX2 and collagen I while triggering the BMP2/Smads signaling phosphorylation cascade. Oral administration in ovariectomized mice revealed precise skeletal targeting with DREVLPs preferentially accumulating in femurs and BMSCs. Treatment substantially preserved trabecular architecture, increasing bone volume fraction compared with untreated controls. Molecular analyses confirmed pathway activation through elevated BMP2, p-Smad1/5/9, and osteogenic markers in bone tissue—effects comparable to clinical bisphosphonates but through anabolic rather than anti-resorptive mechanisms. These findings establish plant EVLPs as a new category of bone-forming agents, with DREVLPs representing a translatable oral nanotherapy that addresses the critical unmet need for safe anabolic treatments in postmenopausal osteoporosis through targeted activation of endogenous BMP2 signaling.

Liver metastases often respond poorly to immunotherapy because of a “cold” tumor microenvironment characterized by limited lymphocyte infiltration. Napabucasin, a small-molecule naphthoquinone, has been suggested to modulate the tumor microenvironment. This preclinical study investigates whether Napabucasin enhances immunotherapy efficacy and prevents liver metastasis. Pretreatment with Napabucasin before tumor inoculation significantly reduces liver metastases in mice and increases immune cell infiltration within metastatic lesions. Mechanistically, Napabucasin induces secretion of the chemokine C–C motif ligand 21 (CCL21) from hepatocytes through activation of the transcription factor c-Fos, leading to the recruitment of lymphocytes into the liver. This immunomodulatory effect is cancer cell–independent and liver-specific, with no effect on splenic, lung, or subcutaneous tumors and no CCL21 upregulation in other organs. The recruited T lymphocytes co-express immune activators and immune checkpoint molecules, indicating a phenotype suitable for reactivation by immune checkpoint inhibitors. Clinical transcriptomic data further show that a Napabucasin-associated gene signature correlates with improved immunotherapy responses. Consistently, Napabucasin pretreatment sensitizes liver metastases to anti–PD-1/programmed death-ligand 1 (PD-L1) therapy in mice. Overall, these findings demonstrate that early administration of Napabucasin reprograms the hepatic microenvironment, converts potential liver metastases into immunologically “hot” tumors, and enhances the efficacy of immune checkpoint blockade.

Drug-induced liver injury (DILI) remains a major clinical challenge due to the absence of specific biomarkers and dependence on subjective diagnostic criteria. This study presents an interpretable gradient-boosting decision tree model (XGB-D) that uses routine laboratory data to enable early and accurate DILI detection. Developed through a multicenter cohort of 36,199 patients, XGB-D shows superior diagnostic performance (area under the curve [AUC] = 0.971) compared with conventional methods and demonstrates robust generalizability across three independent validation cohorts (AUC = 0.881–0.935). SHAP (SHapley Additive exPlanations) analysis identifies alanine aminotransferase and C-reactive protein as key contributors, revealing mechanistic links between hepatocellular damage and inflammatory responses. In prospective real-world monitoring, XGB-D detected DILI signals 2–4 weeks earlier than expert assessment in 29.4% of cases, supporting timely clinical intervention. By integrating interpretable machine learning with clinical hepatology, this work establishes a scalable and transparent framework for precision toxicology with significant implications for drug safety evaluation and personalized medicine.

In forensic cases, the accurate prediction of time since deposition (TsD) for body fluids plays a critical role in evaluating the relevance of biological evidence to criminal cases and reconstructing the timelines of criminal events. While transcriptomics offers avenues for TsD analysis, the environmental sensitivity of mRNA limits its practical utility. In contrast, miRNAs demonstrate superior potential as biomarkers due to their short sequences, high stability, and environmental resistance; however, their forensic application for TsD estimation remains underexplored. This study applied small RNA sequencing to analyze miRNA expression in semen samples from 10 donors across seven TsD intervals (0–48 h). Time-dependent miRNA expression modules were identified through Mfuzz clustering and weighted gene co-expression network analysis. We implemented a multi-stage feature selection pipeline, commencing with least absolute shrinkage and selection operator regression and random forest (RF) that selected 261 candidate miRNAs for model development, followed by recursive feature elimination with ElasticNet to refine the set to 12 miRNAs, and concluding with XGBoost-based multicollinearity reduction and exhaustive optimization to yield a minimal set of 7 miRNAs. The selected miRNA candidates were subsequently validated using reverse transcription-quantitative polymerase chain reaction on an independent sample set. Machine learning models constructed with the initial 261 miRNAs demonstrated that RF achieved optimal performance in the binary classification of early (0–12 h) versus late (24–48 h) TsD, with an accuracy of 0.76, F1-score of 0.75, and area under the curve of 0.82. In regression analysis, an ensemble model integrating partial least squares, ElasticNet, support vector machine, and Ridge attained a test mean absolute error of 6.76 h and an R² of 0.72. This research establishes a novel miRNA-based prediction framework for TsD estimation of semen, integrating dynamic expression patterns with machine learning for the advancement of forensic body fluid analysis.

Accurate detection and risk assessment are crucial for the prognosis of patients with pulmonary hypertension (PH). This study aimed to develop a detection model by employing ultra-high-sensitivity magnetocardiography (MCG) combined with machine learning while exploring the potential of MCG features in enhancing prognostic evaluation. PH patients were allocated into the exploratory and external validation cohort according to enrollment period in a single-center. The control group comprises patients with symptoms including dyspnea, chest tightness, chest pain, or fatigue in the absence of PH. Outpatient follow-up is performed for PH patients in the exploratory cohort. Seven machine learning algorithms and 15 out of 40 MCG features were employed to develop PH detection models. Cox regression is used to construct a risk assessment model. The exploratory cohort includes 312 subjects (156 PH patients), whereas the external validation cohort includes 162 subjects (81 PH patients). The Random Forest algorithm demonstrates superior performance, achieving sensitivities of 79.6% and 84.0% and specificities of 92.4% and 91.4% in different datasets, respectively. During the average follow-up period of 33 weeks, 14 patients experienced clinical deterioration events. Compared to the baseline model developed using three parameters from the guideline-recommended model, the combined model incorporating 2 MCG features performs better in this small cohort to detect deterioration events. In conclusion, the Random Forest-based MCG model demonstrated robust detection accuracy for PH. The integration of 2 MCG features with clinical baseline parameters may improve short-term risk assessment.

Interstitial fibrosis is the best indicator of irreversible or ongoing renal injury after kidney transplantation and faces considerable diagnostic challenges. Owing to the direct connection between urine and kidney, small urinary extracellular vesicles (suEVs) are promising candidates for developing non-invasive and highly sensitive diagnostic biomarkers for allograft fibrosis. Herein, we established an optimized method for separating suEVs from renal transplant recipients with high yield and purity. Through unbiased proteomic and in-depth bioinformatic analyses, we delineated an extensive protein landscape of suEVs and identified sorting nexin 3 (SNX3), vacuolar protein sorting-associated protein 4B (VPS4B) and smoothened (SMO) proteins as potential biomarkers for accurate and consistent diagnosis of fibrosis. Extensive validation across three independent cohorts demonstrated their excellent diagnostic performance in both transplant recipients and chronic kidney disease patients, achieving an outstanding AUC value of 0.9909 and accuracy of 90.6%, respectively. Moreover, this model demonstrated a prognostic value in a 3-month follow-up assessment of renal allograft recipients. Mechanically, we indicated that SNX3 promotes fibroblast activation through the regulation of Wnt secretion. Our study is the first to report the suEV protein biomarkers for the diagnosis of allograft fibrosis, offering a non-invasive alternative to renal biopsy and enabling improved risk stratification for transplant patients, which aids in better management of transplant recipients to improve long-term allograft survival.

As a natural polysaccharide, chitosan has gained increasing attention as a versatile matrix for developing carriers in drug delivery due to its biocompatibility, biodegradability, and modifiable functional groups. Stimuli-responsive strategies have been extensively investigated to achieve controlled and on-demand drug release by exploiting pathological cues, exhibiting significant potential for precision medicine. This review summarizes recent advances in chitosan-based drug delivery carriers with diverse responsive mechanisms, including pH-, redox-, enzyme-, and reactive oxygen species-responsive systems, as well as multi-stimuli responsive platforms. The design principles, underlying mechanisms, and therapeutic outcomes of these systems are discussed. Finally, the review highlights the challenges that hinder clinical translation and outlines future directions to optimize chitosan-based delivery systems for precision medicine.

Acute anterior uveitis (AAU) is the most common extra-articular manifestation of ankylosing spondylitis, leading to recurrent inflammation and irreversible visual impairment. Long-term corticosteroid therapy is associated with substantial adverse effects, and effective strategies to prevent relapse are lacking. Single-cell RNA sequencing (scRNA-seq) was performed on peripheral blood mononuclear cells from seven patients with ankylosing spondylitis-acute anterior uveitis (AS-AAU) and six healthy controls, and aqueous humor samples were analyzed together with genome-wide association study (GWAS) data. Bioinformatic pipelines were used to infer cell–cell communication, metabolic programs, extracellular vesicle (EV)-related signals, and disease-relevant cell types, followed by validation in clinically derived specimens. A monocyte subcluster with upregulated cytokine and chemokine transcripts and signatures of trained immunity, termed Mono-TI, is identified. EVs potentially transport inflammatory mediators associated with Mono-TI, thereby contributing to immune dysregulation. Pseudotime analysis indicates that Mono-TI may differentiate into macrophages within the ocular microenvironment that share overlapping transcriptional features. Cell–cell interaction analysis positions Mono-TI as a highly connected node, receiving IFN-γ signals from CD8⁺ T cells and natural killer (NK) cells while actively regulating neutrophils. Integrated GWAS analysis further implicates Mono-TI, together with CD8⁺ T cells and NK cells, as key contributors to disease pathogenesis. This study delineates the immune landscape of AS-AAU and highlights a monocyte subset with trained immunity–like transcriptional signatures and related immune cell populations as central mediators of systemic and ocular inflammation and as promising therapeutic targets.

Inflammation is a precursor to vascular diseases, including atherosclerosis, and is modulated by the local biomechanical environment. There is a need for in vitro models to advance understanding and test new therapeutics. This study describes the development and characterisation of a human coronary artery organ-chip model of vascular inflammation with physiological biomechanical stimulation. Human coronary artery endothelial cells and smooth muscle cells were cultured on appropriate extracellular matrices in the two adjoining channels of the Chip-S1^® (Emulate Inc). Both endothelial and smooth muscle cells demonstrated characteristic phenotypic identity, as shown by expression of CD31 and α-SMA, respectively. Application of pulsatile tensile strain induced alignment of both cell types, perpendicular to strain direction, as seen in vivo. Addition of TNF-α to the vascular channel drove an inflammatory response in both cell types, as shown by upregulation of ICAM-1 and P65, and attachment and invasion of circulating THP-1 monocytes. Analysis revealed spatial variation in strain with 12% in the centre of the chip, and 5% towards the ends. Pulsatile tensile strain reduced the inflammatory response to TNF-α, with a greater inflammatory response in areas of lower strain, further replicating in vivo behaviour. In conclusion, we present a fully characterised, tri-culture model of the human coronary artery with endothelial and smooth muscle cells and circulating immune cells. This model successfully recapitulates the physiological effects of pulsatile vessel dilation on cell morphology and localised inflammatory susceptibility. Our model was developed upon a commercially available, organ-chip platform, allowing for rapid adoption for therapeutic testing, and fundamental discovery science.

Extracellular vesicles (EVs) have emerged as promising nanotherapeutics for kidney diseases due to their innate biocompatibility, barrier penetration, and regenerative cargo delivery. However, native EVs face critical limitations including low drug-loading efficiency, poor renal targeting, and batch heterogeneity, hampering their clinical utility. Recent advances in EV bioengineering—encompassing cargo loading (electroporation, transfection), membrane modification (ligand conjugation, biomimetic vesicles), and donor cell preconditioning (hypoxia, pharmacological priming)—have significantly enhanced therapeutic precision and efficacy. Integration with biomaterials (e.g., responsive hydrogels, scaffolds) further enables sustained release and targeted delivery, improving outcomes in kidney injury models. This review systematically analyzes these innovative strategies, highlighting mechanistic insights, comparative advantages, and unresolved challenges. We critically evaluate ongoing clinical trials and propose scalable manufacturing solutions and regulatory frameworks to accelerate translation. Engineered EVs represent a next-generation platform for personalized renal nanomedicine, poised to bridge the regenerative potential with clinical reality.

The tumor vascular barrier restricts the penetration of antibody-conjugated nanomedicine, severely limiting its therapeutic efficacy. Selective delivery of nitric oxide (NO) to tumors can modulate vascular permeability, thereby enhancing the penetration of antibody-conjugated nanomedicine. Herein, we developed an ultrasound-responsive NO-releasing material, named HA-SNO, based on a hyaluronic acid backbone, which releases NO upon exposure to ultrasound waves. After systemic administration of HA-SNO, tumor-localized release of NO was achieved using a focused ultrasound system operating at a low intensity (1 MHz, 2.0 W/cm², 50% duty cycle). The released NO induced vasodilation and disrupted the tumor vascular barrier, thereby promoting extravasation and more homogeneous intratumoral distribution of HER2-targeted antibody–conjugated nanomedicine (HER2–SN38). This strategy exhibited significantly enhanced anti-tumor efficacy, with a tumor inhibition rate of 97.67%. These findings highlight the potential clinical utility of ultrasound-triggered tumor-selective delivery of NO in overcoming vascular barriers to drug delivery.

Exosomes have emerged as highly promising drug delivery vehicles due to their intrinsic biocompatibility, low immunogenicity, and natural ability to traverse biological barriers. However, their clinical translation is hindered by several inherent limitations, including restricted loading efficiency for therapeutic proteins and nucleic acids, limited cargo specificity, and variability in delivery performance. To address these challenges, increasing attention has been directed toward exploiting the endogenous molecular pathways that naturally govern cargo selection and packaging within exosomes. Advances in this area have enabled the development of engineering strategies that emulate or amplify these intrinsic mechanisms to achieve more precise and efficient cargo loading while preserving vesicle structure and biological function. Such endogenous engineering approaches have shown superior reproducibility and are more compatible with large-scale manufacturing compared with exogenous manipulation. Notably, most exosome therapeutics currently in phase II or III clinical evaluation rely on natural or endogenously modified vesicles, underscoring the translational promise of this direction. This review provides a systematic overview of the endogenous pathways involved in protein and RNA sorting, summarizes engineering strategies derived from these mechanisms, and highlights representative applications. Together, these insights aim to support the rational design of next-generation exosome-based therapeutics and accelerate their path toward clinical use.

SUMOylation is a dynamic and reversible post-translational modification that has evolved from a regulator of fundamental cellular processes to a key driver of tumorigenesis and progression. By centering on three core pathological dimensions, namely aberrant cell death, metabolic reprogramming, and immune microenvironment dysfunction, this review systematically elucidates how SUMOylation orchestrates malignant tumor progression through these interconnected axes. It provides a detailed overview of the SUMOylation enzymatic cascade, its regulatory features including dynamics, substrate specificity, and responsiveness to cellular stress, and its complex crosstalk with other modifications such as ubiquitination and phosphorylation, which together enable precise and context-dependent functional outcomes in various cancers. A distinctive aspect of this review is its integration of mechanistic insights with emerging therapeutic strategies, evaluating the translational potential of targeting the SUMO pathway, highlighting opportunities to overcome drug resistance and modulate antitumor immunity, while also addressing current challenges in clinical development. In contrast to most existing reviews that follow a trajectory from molecular mechanism to tumor type specific roles, this review presents a paradigm shift by organizing the regulatory network of SUMOylation in cancer around three pathological axes aberrant cell death, metabolic reprogramming, and immune dysfunction, a framework that aligns more closely with the core pathological drivers of tumor progression and transcends conventional classifications based on tumor type or individual signaling pathways. In parallel, the review emphasizes key nodes of crosstalk with ubiquitination and identifies translational bottlenecks, offering targeted insights for the development of personalized therapeutic strategies directed at the SUMO pathway while establishing a theoretical foundation for SUMOylation as a novel therapeutic axis in oncology.

Vulvovaginal candidiasis (VVC), a common mucosal infection caused by Candida spp., affects approximately 75% of reproductive-aged women worldwide. Nevertheless, this significant health concern remains understudied. Consequently, novel therapeutic interventions are urgently required. Probiotics have emerged as a viable alternative for VVC management owing to their favorable safety profile and proven efficacy. However, evidence supporting their antifungal properties remains inconclusive, necessitating further elucidation of the underlying mechanisms. Numerous studies have demonstrated that probiotics and their derivatives can effectively counteract Candida by reducing Candida’s virulence factors, modulating immune responses, promoting vaginal barrier, remodeling the metabolic environment and restoring microbiota homeostasis. Providing an updated summary of VVC pathogenesis and its relationship with vaginal microbiota, this review aims to comprehensively elucidate the roles and mechanisms of probiotics and their derivatives in VVC. Additionally, it discusses the existing evidence for probiotic-based strategies, including probiotics, probiotics derivatives and vaginal microbiome transplants, and provides a better theoretical foundation for interventions targeting VVC based on probiotics.

Microelectromechanical systems (MEMS)-based devices, characterized by miniaturization, high sensing performance, integration, and enhanced comfort, have been increasingly employed in diagnosis, treatment, and monitoring in ophthalmology. MEMS-based devices have been widely used to design sensors that enable real-time monitoring of physiological and chemical markers in vitro and tear biomarkers in vivo. MEMS-based technology improves resolution, imaging depth, speed, and precision in optical coherence tomography and microscopy. MEMS-based devices overcome biological barriers, enable targeted drug delivery, and allow feedback-controlled release for high efficacy and low side effects. MEMS-based retinal prostheses restore visual perception by electrically stimulating retinal neurons. MEMS-based surgical tools enhance surgical precision, minimizes infection risks, and improves patient outcomes. MEMS-based wearable products monitor blink frequency and eye tracking for improvement of the life quality. Thus, we systematically summarized advances in MEMS-based devices for accurate diagnosis, effective treatment, and daily monitoring in ophthalmology. The challenges and potential development directions of MEMS-based devices applied to ophthalmological diseases were discussed. This review highlights the current applications and future potential of MEMS-based devices in revolutionizing eye health.