Engineered matrices with aligned microarchitectures are pivotal in regulating the fibroblast-to-myofibroblast transition, a critical process for wound healing and scar reduction. However, developing a three-dimensional (3D) aligned matrix capable of effectively controlling this transition remains challenging. Herein, we developed a cell-adaptive hydrogel with highly oriented microporous structures, fabricated through bioprinting of thermo/ion/photo-crosslinked gelatin methacrylate/sodium alginate (GelMA/SA) incorporating shear-oriented polyethylene oxide (PEO) filler. The synergistic interactions among GelMA, PEO, and SA yield a homogeneous mixture conducive to the printing of biomimetic 3D constructs with anisotropic micropores. These anisotropic micropores, along with the biochemical cues provided by the GelMA/PEO/SA scaffolds, enhance the oriented spreading and organization of fibroblasts. The resultant spread and aligned cellular morphologies promote the transition of fibroblasts into myofibroblasts. By co-culturing human keratinocytes on the engineered dermal layer, we successfully create a bilayer skin construct, wherein the keratinocytes establish tight junctions accompanied by elevated expression of cytokeratin-14, while the fibroblasts display a highly spread morphology with increased fibronectin expression. The printed hydrogels accelerate full-thickness wound closure by establishing a bioactive microenvironment that mitigate inflammation and stimulate angiogenesis, myofibroblast transition, and extracellular matrix remodeling. This anisotropic hydrogel demonstrates substantial promise for applications in skin tissue engineering.
Diabetic ulcers are among the most common and challenging complications of diabetes mellitus, and effective therapeutic strategies remain elusive. While stem cell secretome (SCS)-based therapy has attracted considerable attention due to its regenerative potential, its direct application is hindered by low bioavailability and rapid diffusion at the wound site. To address these limitations, we designed a bilayer bacterial cellulose–gelatin (Bi-BCG) scaffold inspired by the hierarchical structure of native skin. This scaffold features a compact bacterial cellulose (BC) upper layer with nanoscale porosity and a porous BCG lower layer with pore sizes of ~52 μm, optimized for SCS delivery. The Bi-BCG scaffold demonstrated a water vapor transmission rate of 2384 g/(m2·24 h) and exhibited significantly improved SCS retention capacity while maintaining high fluid absorption, outperforming monolayer BCG scaffolds. Functionally, human umbilical cord-derived mesenchymal stem cell (hUCMSCs)-derived secretomes significantly enhanced the proliferation (by up to 70.7%) and migration of skin fibroblasts under high-glucose conditions, promoted vascular endothelial cell proliferation (increasing Ki-67+ cells from 25.87% to 46.89%) and angiogenic network formation, and effectively suppressed macrophage-mediated inflammatory responses and oxidative stress. In vivo, the combination of SCSs with the Bi-BCG scaffold exhibited a clear synergistic effect, achieving a wound closure rate of 78.8% by day 10 and promoting superior structural restoration with well-organized collagen deposition, outperforming either treatment alone. These findings underscore the potential of the Bi-BCG scaffold combined with SCSs as an effective strategy for enhancing diabetic wound healing.
In recent years, large language models (LLMs) have achieved remarkable progress in natural language processing and demonstrated potential applications in medicine. However, their professional capabilities in specific medical subfields, such as immunology, still require systematic evaluation. This study systematically evaluated 11 representative LLMs, including DeepSeek, GPT, Llama, Gemma, and Qwen series, based on the Chinese National Health Professional Qualification Examination in Rheumatology and Clinical Immunology. The evaluation covered four dimensions: basic medical knowledge, related medical knowledge, immunology knowledge, and professional practice ability. Results show significant differences among LLMs. DeepSeek-R1 and Qwen3 achieve the best performance, with accuracy exceeding 90%. However, performance on professional practice ability tasks remained relatively low, highlighting limitations in complex clinical applications.