Steroid-induced osteonecrosis of the femoral head (SONFH) is a rapidly progressing and disabling complication of long-term glucocorticoid therapy, lacking effective early-stage intervention mechanisms. Its early manifestation involves a fate shift in bone marrow mesenchymal stem cells (BMSCs) characterized by decreased osteogenic differentiation (OGD) and increased adipogenic differentiation (AGD), yet the upstream regulatory mechanisms remain unclear. Herein, we integrated AGD-related microRNA (miRNA) microarray data with exosomal miRNA sequencing data and identified miR-199a-3p as a crucial candidate driver of this lineage imbalance. Our results revealed the up-regulation of miR-199a-3p in SONFH tissues and in glucocorticoid-treated cellular models, and indicated that its overexpression suppresses the OGD of BMSCs while markedly promoting the AGD. Further integrating mRNA-sequencing profiling during AGD with target prediction, protein–protein interaction network analysis, and dual-luciferase reporter assays, we confirmed integrin β8 (ITGB8) as a direct target of miR-199a-3p, which is consistently decreased in SONFH tissues and during adipogenic induction. We further revealed that miR-199a-3p suppressed the OGD of BMSCs by repressing ITGB8 expression, thereby inactivating the focal adhesion kinase (FAK)–extracellular signal-regulated kinase (ERK)–runt-related transcription factor 2 (RUNX2) signaling cascade. Conversely, silencing miR-199a-3p restores ITGB8 levels, reactivates this pathway, and corrects the OGD/AGD bias. In vivo, local administration of antagomiR-199a-3p in a SONFH rat model markedly improved trabecular bone architecture, increased bone mass, and up-regulated RUNX2 expression. These findings reveal for the first time that the miR-199a-3p/ITGB8–FAK–ERK–RUNX2 axis represents an unrecognized pathogenic pathway in SONFH, and support the local suppression of miR-199a-3p as a translatable early intervention strategy.

Parity–time and anti-parity–time symmetries, originally formulated in the context of non-Hermitian quantum mechanics, have now been experimentally realized across diverse physical platforms, including photonic, wave-guiding, acoustic, and mechanical systems, as well as systems operating near exceptional points. While the active manipulation of these symmetries has been well explored in wave-based systems, their implementations in diffusive physics have long remained predominantly confined to static phases, where systems are locked in a fixed symmetric or symmetry-broken configuration due to their invariant structural parameters. Bridging this gap, recent work has introduced and verified the concept of temporal anti-parity–time symmetry, demonstrating that energy transport in diffusive media can be actively programmed in the time domain.

Lanthanide nanocrystals hold exceptional promise for electroluminescence applications due to their unique optical properties. However, their intrinsic insulating character and localized 4f orbitals severely restrict carrier injection, thereby hindering direct electrical excitation. In a recent study published in Nature, Tan and colleagues circumvented this fundamental bottleneck via molecular engineering of the nanocrystal surface. They developed a series of functionalized ligands (e.g., carbazole-phosphine oxide) to establish an electroactive interface, facilitating efficient transfer of electro-generated triplet excitons to lanthanide ions. Notably, a wide-ranging multicolor electroluminescence from lanthanide nanocrystals was achieved for the first time, exhibiting high power efficiency and external quantum efficiency. These findings provide new opportunities for electrically driven luminescence in lanthanide nanocrystals or other insulating systems.

As a fundamental imaging task, All-in-One Image Restoration (AiOIR) aims to achieve image restoration caused by multiple degradation patterns via a single model with unified parameters. However, existing methods typically rely on sample-wise supervision, which tends to entangle degradation features with image content. Furthermore, the inevitable distribution shift between training data (source domain) and real-world samples (target domain) weakens degradation awareness, severely limiting the generalization capability of models in real-world scenarios. To address this, a Unified Domain-Adaptive Image Restoration (UDAIR) computer vision model is proposed by achieving the transition from learning unstable local features to learning robust universal prototypes. To decouple degradation from content, a Cross-Sample Contrastive Learning mechanism is implemented by a codebook-based module. By contrasting samples with shared degradations but diverse content, the proposed model learns discrete embeddings as degradation prototypes. Furthermore, to actively bridge the distribution gap during inference, a correlation alignment-based test-time adaptation mechanism is designed to dynamically pull drifting target features toward their corresponding degradation cluster centers to effectively eliminate residual alignment discrepancies. Experimental results on 10 open-source datasets demonstrate that UDAIR achieves new state-of-the-art performance for the AiOIR task, in which each technical module contributes to the desired performance improvement. Most importantly, the feature cluster validates the degradation identification under multiple degradation patterns, and qualitative comparisons showcase robust generalization to real-world scenarios.

Wearable haptic systems enable natural and intuitive human–machine interaction, with growing relevance in immersive entertainment, physical rehabilitation, social communication, and personalized therapy. By delivering real-time tactile feedback via electrical, thermal, and mechanical stimulation, individually or in combination, these systems enhance user engagement and enable more dynamic and responsive digital interactions. This review presents a structured overview of the design principles, actuation mechanisms, and material architectures that define current wearable haptic technologies. We evaluate the strengths and limitations of major feedback modalities and highlight how system-level designs are tailored to specific use cases, from immersive gaming and training simulations to postoperative rehabilitation and emotion-sensitive communication. We also identify key challenges in miniaturization, multimodal integration, user comfort, and long-term wearability. The insights presented here aim to inform the development of next-generation haptic interfaces that are adaptive, scalable, and deployable in real-world applications.

In solid tumors, the immunosuppressive tumor microenvironment and antigenic heterogeneity pose important challenges for effective immunotherapy, often leading to limited T-cell infiltration and inadequate immune activation. To overcome these barriers in solid tumors, the development of next-generation vaccines capable of eliciting robust and durable anti-tumor immunity has constituted a major focus in the clinic. A promising strategy involves a bioinspired approach that functionalizes synthetic nanocarriers with native cell membranes. These engineered platforms are designed to preserve the surface properties of native cell membranes (immune cells, nonimmune cells, and hybrid cell membranes) to enhance antigen presentation, prolong systemic circulation, and improve biocompatibility. This study systematically examines the design principles and mechanisms of bioinspired vaccines with native cell membranes, highlighting their capability to integrate multiple antigenic and adjuvant signals for superior antigen presentation and T-cell activation. We further explore the synergistic therapeutic effects of the next-generation vaccines when combined with common anti-tumor therapies. Moreover, the primary challenges for their clinical translation in solid tumors are critically discussed. These bioinspired nanoplatforms represent a transformative direction for developing more effective and personalized immunotherapies for solid tumors.

Atmospheric water harvesting (AWH) has garnered widespread attention for the alleviation of freshwater scarcity. However, AWH still presents challenges in simultaneously achieving rapid adsorption–desorption, low-energy regeneration, and environmentally sustainable development. Herein, a biomimetic all-biomass double-layer sponge is prepared to spatially separate moisture capture from solar-driven desorption while integrating hydrovoltaic power generation (HG) for the co-production of water and electricity. Inspired by tree transpiration, the carboxylated wood scaffold loaded with LiCl mimics root-like moisture uptake, while the recycled demethylated lignin coated on the upper layer serves as a photothermal evaporation layer that drives efficient solar-induced desorption analogous to leaf-mediated sunlight absorption. The biomimetic double-layer structure enables efficient moisture absorption and desorption, with vertical humidity gradients guiding water through directional fiber microchannels and generating sustained flow potentials. Under 60% relative humidity (RH), 25 °C, and 1-sun illumination, the open-circuit voltage (V_oc) of the double-layer wood sponge is ~430 mV, with a short-circuit current (I_sc) of ~6.3 μA and water uptake capacity of ~1.45 g g⁻¹, maintaining stable properties over multiple adsorption–desorption cycles. Under simulated conditions outdoors, freshwater can be effectively harvested, while the generated electricity can power lighting or be stored in capacitors to supply commercial electronic devices. In this strategy, the natural cellulose–lignin structure within wood is restructured and repurposed into a low-cost, sustainable platform that supplies both freshwater and environmental energy, providing a scalable pathway for efficient collection of water resources in resource-constrained regions.

Halide perovskites are novel photovoltaic materials with soft ionic lattice that are continuously exposed to sunlight during operation. High-energy ultraviolet (UV) irradiation will cause halide ion oxidation within the perovskite layer, accelerating halide ion migration and component loss, which results in a marked degradation in the device’s photovoltaic performance. Here, we introduce a 2,3-bis(2,4,5-trimethyl-3-thienyl) maleimide (BTTM) molecule capable of interconversion between UV and visible light into the perovskite structure, intensifying the light stability of the perovskite layer via ion anchoring. Meanwhile, the incorporation of BTTM promotes the growth of perovskite crystals and efficiently passivates defects within the perovskite film, significantly boosting the open-circuit voltage of the perovskite solar cells. This results in a power conversion efficiency increase from 22.07% to 24.71%. Under UV irradiation (365 nm), BTTM molecules mitigate the degradation of perovskite by suppressing the ion migration of iodide ions. After cumulative exposure to 5 kWh/m² of continuous UV irradiation, BTTM-based devices retain over 90% of their initial power conversion efficiency, demonstrating significantly enhanced UV stability. This study offers a straightforward approach to UV protection, providing novel insights for the environmental application of perovskite solar cells under intense UV irradiation.

The long-standing trade-off between illumination performance (such as uniformity) and optical system complexity has hindered the development of compact lithographic tools. Here, we overcome this limitation by introducing a metasurface-enabled lensless illumination system that seamlessly combines light homogenization and monitoring capabilities. Using an asymmetric diffusion metasurface integrated with intelligent optimization algorithms, we achieve precise light field control and self-adaptive feedback without relying on conventional lens assemblies. When deployed on a self-developed large-field lithography platform, the system demonstrates uniform ultraviolet illumination with nonuniformity below 3% and successfully exposes patterns across a range of feature sizes. A gray-level to power-level mapping algorithm enables real-time assessment of light-emitting diode performance, providing diagnostic alerts and adaptively reconfigurable exposure configurations to circumvent defective emitters. This synergy between metasurface optics and algorithmic intelligence establishes a new paradigm for compact, low-cost, and scalable lithographic illumination, offering a pathway toward next-generation intelligent manufacturing systems.