Sustained exosome delivery improves bone formation in p75NTR-deficient mice

Extracellular vesicles released by wild-type ectomesenchymal stem cells restored deficient bone-forming activity in cells lacking the p75 neurotrophin receptor when delivered from a coaxial nanofibrous membrane, according to a preclinical study in mouse cells and a critical-sized cranial defect model. The membrane-supported treatment enhanced osteogenic differentiation and mineralization in p75NTR-knockout cells and promoted formation of mechanically competent bone in knockout mice. The results associate the effect with coordinated activation of PI3K/Akt and ERK1/2 signaling, although the experiments used pharmacological pathway modulators rather than genetic interventions, and the authors call for genetic validation and larger safety studies.

The study addressed an apparent contradiction in the biology of p75NTR, a member of the tumor necrosis factor receptor superfamily that regulates cell death, survival and differentiation in a context-dependent manner. Earlier experiments had associated the receptor with enhanced osteogenesis in some mesenchymal cell systems, while other studies reported inhibitory effects. The researchers examined p75NTR-knockout mice and ectomesenchymal stem cells, or EMSCs, isolated from the maxillofacial tissue of E16.5 embryos.

Knockout animals were smaller than their wild-type littermates and showed delayed skeletal development, with more cartilage and less mineralized bone. Micro-computed tomography also revealed reduced bone volume fraction, bone mineral density and trabecular number in the femur and mandible. The knockout phenotype was not confined to a single differentiation assay.

Flow cytometry showed that wild-type and knockout EMSCs expressed the mesenchymal stem-cell markers CD29, CD90 and Sca-1 at high frequencies while showing little expression of the hematopoietic marker CD45, supporting their characterization as mesenchymal stem-cell-like populations. However, knockout EMSCs displayed weaker multilineage differentiation than wild-type cells. After osteogenic induction, knockout cultures formed fewer calcium deposits, while adipogenic and chondrogenic assays showed fewer lipid droplets and cartilage nodules.

These observations placed the therapeutic work in the context of a broad developmental impairment associated with the loss of p75NTR. The researchers isolated extracellular vesicles by differential ultracentrifugation from cultures of wild-type and knockout EMSCs. Wild-type vesicles had a mean particle diameter of 142.6 ± 64.9 nanometres, and knockout-derived vesicles measured 150.5 ± 95.8 nanometres.

Transmission electron microscopy showed rounded or cup-shaped structures in both preparations. Both vesicle types contained the exosomal markers TSG101 and CD9 but lacked detectable calnexin, an endoplasmic-reticulum marker used as a negative control. Fluorescently labelled wild-type vesicles entered recipient EMSCs and accumulated in the perinuclear region at 12 and 24 hours.

In cell-culture assays, wild-type vesicles increased knockout-cell viability in a concentration-dependent manner; 50 micrograms per millilitre produced a 24% increase in viability over the control group on day five. At that dose, proliferation of treated knockout cells approached the wild-type level by day five, although the authors described the effect as a partial restoration of proliferative capacity. Alkaline phosphatase staining indicated stronger early differentiation, while Alizarin Red staining after 21 days showed more calcium-nodule formation.

At 50 micrograms per millilitre, the treated knockout cells expressed several osteogenic markers at levels similar to those of untreated wild-type cells. Conversely, vesicles derived from knockout EMSCs suppressed collagen I, alkaline phosphatase and Runx2 expression in wild-type cells. The reciprocal pattern shows that vesicle activity differed according to the donor cells’ p75NTR state, but the experiments did not identify which exosomal cargo was responsible.

To improve local retention and duration of treatment, the team encapsulated wild-type vesicles in the aqueous core of coaxial electrospun fibres. The shell combined silk fibroin and polycaprolactone, while the core contained polyvinyl alcohol and the vesicles. After screening several shell compositions and loading concentrations, the authors selected an SF/PCL ratio of 2:5 and a vesicle concentration of 10 milligrams per millilitre.

Scanning electron microscopy showed relatively uniform fibres in the selected formulation, with a mean diameter of 146.8 ± 27.6 nanometres before vesicle loading and 156.9 ± 40.5 nanometres afterward. Dual fluorescence confirmed that the vesicle-containing solution occupied the core of the fibres. Release followed a three-stage pattern, with an initial burst followed by more gradual delivery over 28 days.

By day 28, cumulative release reached 56.88 ± 2.01% for the selected formulation, and released vesicles were taken up by knockout EMSCs during culture. The membrane was biocompatible with knockout EMSCs: live/dead staining detected no dead cells after three days, and the exosome-loaded membrane accelerated proliferation from day three through day five. At seven and 14 days, alkaline-phosphatase activity was higher with the loaded membrane than with the unloaded membrane or control, while the loaded membrane group showed the highest observed activity.

At 21 days, the loaded membrane produced the greatest mineral deposition, and immunofluorescence showed stronger collagen I and osteopontin signals. RNA sequencing of knockout cells cultured under osteogenic conditions identified 186 upregulated and 90 downregulated genes in the loaded-membrane group compared with the unloaded membrane, using thresholds of absolute log2 fold change of at least 1 and adjusted p below 0.05. Pathway analysis highlighted PI3K/Akt signaling, while enriched biological processes included skeletal development, ossification and osteoblast differentiation.

Pharmacological experiments implicated both PI3K/Akt and ERK1/2 in the response. The loaded membrane increased phosphorylation of Akt and ERK1/2 and raised the measured spatial association between the two phosphorylated proteins. The PI3K activator 740Y-P further increased Akt phosphorylation and osteogenic outputs, whereas the PI3K inhibitor LY294002 reduced Akt phosphorylation, alkaline-phosphatase activity, mineralization and expression of osteopontin and osteocalcin.

The ERK inhibitor PD98059 reduced ERK1/2 phosphorylation without significantly changing Akt phosphorylation. Modulating PI3K/Akt produced corresponding changes in ERK1/2 phosphorylation. These results support a directional relationship in which PI3K/Akt influences ERK1/2 activation, but no genetic experiment was performed to establish the molecular link or distinguish direct from indirect effects.

In vivo, 8-week-old knockout mice received 3-millimetre cranial defects and were assigned to loaded-membrane, unloaded-membrane, free-vesicle or no-treatment groups, with wild-type mice serving as controls. Six male mice were included in each group at the four- and eight-week observation points, for 60 mice in the full study. Each loaded membrane contained 100 micrograms of vesicle protein; the free-vesicle group received repeated local injections, with the cumulative dose reaching 100 micrograms by the eighth injection.

Micro-computed tomography at four and eight weeks showed greater mineralized bone in defects treated with the loaded membrane than in the other knockout groups. At eight weeks, its bone surface area was 31.91 square millimetres, compared with 20.45 for free vesicles and 20.03 for the unloaded membrane. Maximum load reached 4.188 ± 0.316 newtons, significantly above the free-vesicle and unloaded-membrane groups and not significantly different from the wild-type value of 4.057 ± 0.357 newtons.

Untreated knockout defects lacked sufficient mineralized tissue for reliable load testing. Histology was consistent with the imaging findings. By eight weeks, defects supported by the loaded membrane showed continuous bridging, thicker and more compact trabeculae, reduced inter-trabecular spaces and areas resembling lamellar bone, whereas the other treated groups retained incomplete bridging and fibrous tissue.

Collagen I and osteopontin staining was strongest after membrane-mediated delivery. Immunohistochemistry indicated increased phosphorylated Akt and ERK1/2 in the loaded-membrane group, together with higher ATF4 and Runx2 expression. Examination of the heart, liver, spleen, lungs and kidneys found no reported inflammation or pathological damage at eight weeks.

The experiments therefore support a preclinical cell-free strategy combining exosome-mediated signaling with local biomaterial delivery, but they do not establish efficacy in people. The small-animal model, male-only cohort, eight-week observation period, absence of long-term and large-animal safety data, and inability of the protein-based release assay to distinguish intact vesicles from degraded exosomal protein remain important limitations.

Subject of Research: Preclinical testing of exosome-loaded nanofibrous membranes for p75NTR-associated bone-regeneration deficiency

Article Title: EMSCs-derived exosomes rescue p75NTR −/− -mediated osteogenesis deficiency via PI3K/Akt/ERK1/2 signaling pathways

Article References: Ye, J., Jin, X., Huang, Y., Khup, P. L., Yu, Y., Cheng, M., Yuan, Q., Xu, D., Nie, X., & Cheng, G. (2026). EMSCs-derived exosomes rescue p75NTR−/− -mediated osteogenesis deficiency via PI3K/Akt/ERK1/2 signaling pathways. Materials Today Bio, 41, Article 103755. https://doi.org/10.1016/j.mtbio.2026.103755

Image Credits: AI Generated

DOI: 10.1016/j.mtbio.2026.103755

Keywords: p75NTR, extracellular vesicles, bone regeneration, osteogenesis, nanofibrous membranes, PI3K/Akt signaling, ERK1/2 signaling, cranial defects, ectomesenchymal stem cells, EMSCs-derived, exosomes, rescue