Diabetic foot ulcers and large-area burns are the two toughest challenges in wound repair: wounds that never close, vessels that won’t grow, inflammation that won’t subside. Mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) have been a rising star in regenerative medicine over the past decade — “nano-couriers” dispatched by stem cells, capable of delivering repair signals precisely into cells. But they carry a fatal weakness: cleared from the body within hours of injection, never reaching the wound.

In 2026, a team at Beijing Jishuitan Hospital published a study in Materials Today Bio: coaxial electrospinning woven the vesicles into core-shell nanofibers — gelatin as the shell, PCL as the core, vesicles tucked into the “sandwich core” for sustained release. In a 10 cm² full-thickness skin defect in rats, wounds were almost fully healed at 28 days; three mechanisms — pro-angiogenic, anti-inflammatory, and metabolic reboot — were for the first time integrated and explained on the same mesh.

*All values quoted from the original paper: size and concentration from Section 3.1; branch count from Section 3.5; wound results from Section 3.7.

PART 01 Why Do Large-Area Wounds Drag on into a “Chronic Tug-of-War”?

Small wounds heal because three events happen in sequence: inflammation resolves, tissue remodels, extracellular matrix reorganizes. But full-thickness defects like diabetic foot ulcers and large-area burns scramble all three steps. The paper explains this “chaos chain” plainly in the introduction: hyperglycemia disrupts mitochondrial fission-fusion balance, oxidative phosphorylation breaks down — energy deficit and excess reactive oxygen species (ROS) appear simultaneously; damaged mitochondria release DNA into the cytoplasm, activating the NLRP3 inflammasome; insufficient ATP pushes cells toward necroptosis.

Ultimately, pyroptosis, apoptosis, and necroptosis — three “death programs” — ignite simultaneously in the same batch of cells. The academic community named this coordinated cell death: PANoptosis. Tissue damage worsens, re-epithelialization delays, and the wound becomes a tug-of-war. Meanwhile, the two key players determining the schedule both fall behind: endothelial cells responsible for laying vessels, and fibroblasts responsible for building the house.

PART 02 Meet the Protagonist: The “Nano-Courier” Dispatched by Stem Cells

Mesenchymal stem cell-derived extracellular vesicles (MSC-EVs) are nanometer-scale lipid-membrane vesicles secreted by stem cells, carrying proteins, miRNAs, and other “repair cargo.” Compared to direct stem cell transplantation, they have lower immunogenicity, no tumorigenicity concerns, yet can replicate the immunomodulatory and reparative functions of stem cells — the leading candidate for “cell-free regeneration.” The paper points out they are one step from clinical translation: cleared too quickly from physiological fluid circulation, insufficient local retention, and inadequate sustained supply to the wound.

The vesicles in this study first passed a strict “quality inspection”: under TEM they show typical cup-shaped morphology with intact lipid bilayers, size 80–150 nm; nanoparticle tracking analysis (NTA) showed a main peak at 110 nm, concentration up to 1.5×10¹⁰ particles/mL; immunoblotting confirmed the three key vesicle markers Alix, TSG101, and HSP70 were all present, while the endoplasmic reticulum marker calnexin was negative — high purity, low contamination. Concentration screening was also performed: among 50, 100, and 200 μg/mL, both 100 and 200 μg/mL performed significantly better, with no significant difference between them — the team ultimately chose 100 μg/mL, sufficient and raw material–saving.

PART 03 Weaving Couriers into Fibers: The Specialty of Core-Shell Electrospinning

How to keep vesicles “stationed” at the wound for weeks? The team used coaxial electrospinning — a fiber-forming technology sharing its lineage with melt electrowriting (MEW) in bio 3D printing: two concentric needles, the outer needle carrying shell solution and the inner needle carrying core solution, with a high-voltage electric field drawing the liquid streams into nanometer-scale fibers, wound onto a rotating drum into a membrane. The core is PCL (strength provider), the shell is gelatin (cell-affinity provider), and vesicles are incorporated into the shell solution at 100 μg/mL.

The design’s cleverness lies in the “sandwich core”: vesicles are enclosed within the fiber structure, not directly contacting HFIP, the organic solvent that dissolves PCL — activity loss during spinning is blocked. In vivo, the gelatin shell dissolves first for rapid release, with the PCL core then slowly degrading for sustained release — cell uptake experiments show fluorescently labeled vesicles appear in the cytoplasm as early as day 1 of co-culture, with stronger signals by day 3, synchronized with the scaffold’s release curve. There was also a happy surprise: after vesicle incorporation, fibers actually became finer and more aligned — the negative charge on vesicle surfaces raised solution conductivity, strengthening the electric-field drawing force; the vesicles also disrupted gelatin chain entanglement and lowered viscosity, making fine fibers easier to draw thinner and more uniform.

PART 04 Three Upgrades for One Mesh: More Hydrophilic, Stronger, More Stable

The old problems of pure PCL dressings are hydrophobicity, brittleness, and poor cell adhesion. After vesicle incorporation, nearly every “health check” indicator of the mesh improved: contact angle dropped from 124.8° to 64.8°, shifting from hydrophobic to hydrophilic; tensile strength rose from 1.87 to 3.27 MPa (about +75%), and elongation at break rose from 93.7% to 121% — both stronger and tougher; thermal decomposition onset temperature also rose from about 305 °C to about 320 °C. The paper explains vesicles as “nanofillers”: lipids and proteins restrict PCL chain segment motion, while hydrogen bonds stabilize molecular conformation.

Degradation and safety also passed: after 14 days in 37 °C water, core-shell fibers lost 24±4.2% mass (gelatin shell dissolving first), while pure PCL lost less than 5% — a “fast-then-slow” controlled release rhythm; hemolysis rates of both membrane groups were below the 5% safety threshold. Cell-level performance was consistent: L929 fibroblasts proliferated significantly more on PCL-EVs, with SEM showing cells spread into plump spindle shapes, pseudopod networks tightly interlocking with fibers; mitochondrial ROS levels significantly decreased — oxidative stress was suppressed.

PART 05 Cells Sign for the Delivery: Expanding Mitochondria, Restoring Energy

Fibroblasts are the “construction crew” of the wound, and their energy budget determines the schedule. The construction crew in diabetic wounds has long been in a “power outage”: mitochondria on strike, ATP bottomed out. After 72 hours of PCL-EVs culture, the TEM images showed the crew’s “power plants” visibly multiplying — single-cell mitochondrial counts increased significantly, total ATP production significantly rose, which the paper interprets as direct evidence of vesicles promoting mitochondrial regeneration.

Untargeted metabolomics gave a more panoramic view: a total of 389 metabolites were detected, with as many as 138 differential metabolites in the PCL-EVs group (16 in the control group, 23 in the PCL group). Three lipid metabolites — nonanoic acid, oleic acid, palmitic acid — were significantly elevated, pointing to a comprehensive acceleration of fatty acid β-oxidation, the “burn fat for fuel” pathway; lipid toxicity markers such as ceramides did not appear, indicating lipid metabolism was “accelerated but not overloaded.” The paper speculates this is related to enzymes delivered by vesicles, such as carnitine palmitoyltransferase 1, which transports fatty acids into mitochondria.

PART 06 One Hand Promotes Vessels, the Other Holds Down “Collective Death”

The exam results on the vascular side were highly intuitive: in scratch assays, endothelial cells in the PCL-EVs group almost completely closed the wound within 8 hours, while the PCL group clearly lagged; in Matrigel tube formation, branch points were 3.5× those of the PCL group, with higher vascular network density. Gene and protein levels echoed this: VEGF expression upregulated 4.2×, vWF 3.8×, HIF-1α 2.9×, MMP-9 3.1× (all P<0.01); ELISA-measured secreted VEGF reached 450±35 pg/mL, 2.3× the PCL group (195±20). Mechanistically, the paper points to synergistic activation of the HIF-1α-VEGF and cGMP-PKG cascades.

The inflammatory side findings were even rarer. Transcriptome sequencing of rat skin flaps: 1,507 differential genes out of 28,450 transcripts (667 up, 840 down), with the most prominent KEGG enrichment being the cGMP-PKG pathway. Measured cGMP content and PKG expression were significantly elevated; pro-apoptotic p53 and PANoptosis marker Caspase-1 significantly decreased, anti-apoptotic Bcl-2 upregulated; PANoptosis-related genes Nlrp3 and Card10 overall downregulated; macrophages polarized toward the repair-type M2. The paper also calls out the comparative advantage of the core-shell design: VEGF-loaded collagen scaffolds promote vascularization but often exacerbate inflammation (uncontrolled growth factor release); PLGA-EV systems have good sustained release but lack structural flexibility — core-shell fibers pack both “mechanical support” and “biological signals” into the same mesh.

PART 07 Rat Study: The 28-Day Test on a 10 cm² Wound

The animal experiment design has an easily overlooked detail: rat wounds heal mainly through skin contraction, while human wounds heal through new tissue formation — directly extrapolating results to humans would be inaccurate. The team therefore sutured an anti-contraction splint onto each rat’s wound, physically locking contraction and forcing the wound to “regrow flesh” like a human wound. Twenty-four male SD rats (250–280 g) with 3.3×3.3 cm full-thickness skin removed from the back (~10 cm², 40–50% of back skin area), divided into control (gauze), PCL, and PCL-EVs groups, photographed and quantified on days 0/7/14/21/28.

The differences were clear: the control group still had large unhealed wounds at day 28; the PCL group showed improvement but with obvious residue; the PCL-EVs group had nearly complete healing at 28 days, with the lowest residual wound area among all groups. Notably, even the aligned PCL fiber membrane without vesicles healed faster than gauze — the paper considers aligned topography itself an “active physical signal module” capable of guiding cells to align and migrate along fiber direction. Histology refined the timeline: at day 14 the PCL-EVs group already showed neovascular sprouting, while the control and PCL groups waited until day 28; at week 2 CD31 fluorescence intensity was about 1.5× the other two groups; collagen density was higher, epidermal hyperplasia lighter; Sirius Red red/green ratio decreased — the I:III collagen ratio approached the fetal “less scarring” matrix.

PART 08 Sober View: Four Hurdles from Rats to the Sickbed

The paper does not avoid its limitations:

Closing Thoughts

📄 Paper Information

· END ·