Regenerative medicine has been steadily moving away from transplanting living stem cells toward cell-free therapies that capture the healing power of cells without the safety baggage. Exosomes, the nanoscale vesicles that cells release into their surroundings, can carry regenerative proteins and genetic instructions to damaged tissue while sidestepping the tumor risk, immune rejection, and genomic instability that complicate stem cell transplants. The catch has always been manufacturing: producing enough exosomes, with enough therapeutic punch, to matter clinically. A new study published in Materials Today Bio reports a surprisingly simple materials-based fix. By adding graphene oxide to the culture medium, researchers in South Korea coaxed induced pluripotent stem cells to form multiple small spheroids instead of one large aggregate per well, nearly doubling exosome output and sharpening the vesicles’ ability to drive blood vessel growth.
The team, led by Woochan Kim and Jangho Kim, began by thoroughly characterizing their graphene oxide nanosheets using transmission electron microscopy, zeta potential analysis, infrared and Raman spectroscopy, X-ray photoelectron spectroscopy, and atomic force microscopy. The sheets measured roughly 300 to 400 nanometers across and about 1.7 nanometers thick, with the negatively charged, oxygen-rich surface chemistry that gives graphene oxide its affinity for proteins and extracellular matrix molecules. When human induced pluripotent stem cells were cultured with graphene oxide at concentrations up to 20 micrograms per milliliter, viability remained high in both flat monolayer and three-dimensional spheroid formats, with no evident cytotoxicity at any tested dose.
The structural effect was dramatic. Conventional cultures in ultra-low-attachment plates typically produced a single dominant spheroid per well, but graphene oxide-treated cultures generated multiple smaller aggregates that persisted throughout five days of culture. The researchers quantified this shift by counting spheroids per hundred wells and measuring diameters at days one, three, and five, finding substantially higher spheroid yields in the engineered condition. Live-dead staining confirmed that the smaller spheroids stayed healthy, and F-actin staining showed preserved cytoskeletal organization. Gene expression analysis revealed altered levels of CDH1, ITGB1, and YAP1, markers tied to cell adhesion and mechanotransduction, along with reduced HIF-1alpha signaling, suggesting the smaller aggregates suffered less of the hypoxic stress that plagues the nutrient-starved cores of large spheroids.
Crucially, the architectural change translated into a manufacturing advantage. When exosome production was normalized to the number of cells originally seeded, graphene oxide-engineered spheroids yielded 2.75 billion particles per million cells, compared with 1.4 billion from conventional spheroids, nearly a twofold increase. The isolated vesicles carried the canonical exosomal markers CD9, TSG101, and Syntenin-1, while the endoplasmic reticulum protein Calnexin, a marker of cellular contamination, was absent. Raman spectroscopy detected no graphene oxide signature in the final exosome preparations, indicating the nanomaterial itself did not contaminate the therapeutic product. MicroRNA sequencing then revealed that the engineered exosomes carried a distinct cargo profile, with altered levels of angiogenesis-associated microRNAs including miR-126-3p, miR-130a-3p, and miR-210-5p, and pathway enrichment pointing toward VEGF, MAPK, Wnt, and TGF-beta signaling.
Functional testing followed in human umbilical vein endothelial cells. Exosomes from the engineered spheroids, dubbed GO-M-Exo, elevated expression of angiogenic factors including ANG-1, ANG-2, HGF, bFGF, and VEGF more strongly than exosomes from conventional spheroids. In scratch wound assays, GO-M-Exo drove the greatest wound closure over 24 hours, and in Matrigel tube formation assays it significantly increased total tube length, branching length, node count, and loop count. When the researchers added SU5416, a pharmacological inhibitor of the VEGF receptor VEGFR2, the enhanced network formation collapsed, providing mechanistic evidence that the engineered exosomes act through VEGFR2-dependent signaling rather than through nonspecific effects of increased vesicle number.
The team then pushed the platform into living animals. In a rat model of full-thickness skin wounds, local administration of GO-M-Exo at days zero, three, seven, and ten produced near-complete wound closure of roughly 97 percent by day 14, compared with about 91 percent in untreated controls. Histology told a consistent story: thicker granulation tissue, more complete re-epithelialization, greater collagen deposition on Masson’s trichrome staining, and a higher density of CD31-positive microvessels in the engineered exosome group. Repeated dosing produced no obvious inflammatory infiltration, necrosis, or abnormal tissue reactions at the endpoint, an encouraging early safety signal for a nanomaterial-derived therapy.
Perhaps the most forward-looking part of the study is the extension of the platform to lineage-specific production. The researchers differentiated their graphene oxide-engineered spheroids into vascular endothelial cells through a stepwise 20-day protocol spanning mesoderm induction, endothelial specification, and maturation. The engineered endothelial spheroids, called GO-MV, again formed multiple smaller aggregates, expressed higher levels of the endothelial markers VWF and CD31, produced more intracellular nitric oxide, and yielded 2.82 billion exosome particles per million cells versus 1.6 billion from conventional endothelial spheroids. Their exosomes were enriched in microRNAs tied to angiogenesis, migration, and differentiation, including miR-133a-3p, miR-199a-3p, and miR-214-3p, and again showed VEGFR2-dependent enhancement of endothelial tube formation.
In the wound model, the endothelial lineage exosomes outperformed everything else. Wounds treated with GO-MV-Exo reached approximately 98 percent closure by day 14, ahead of 95 percent for conventional endothelial exosomes and 94 percent for untreated controls, and showed the highest microvessel density, collagen deposition, and granulation tissue thickness of any group. The authors argue this demonstrates that exosome quality is determined not only by the donor cell source but by how that population is structurally organized and developmentally programmed. In other words, lineage commitment and microenvironmental nanoengineering appear to act synergistically, a conceptual point they suggest should reshape how exosome manufacturing platforms are designed.
The comparison with existing spheroid technologies underscores the practical appeal of the approach. Hanging-drop and ultra-low-attachment methods are simple but favor single spheroids, while microwell and microfluidic platforms offer control at the cost of specialized equipment, and bioreactor systems face challenges from shear stress and aggregate fusion. Graphene oxide achieves its effect within a conventional culture format by modulating early cell-cell and cell-material interactions, likely through protein adsorption and redistribution of interfacial forces during aggregation nucleation. That makes it a scalable upstream design parameter for exosome production, rather than another downstream concentration step.
Significant hurdles remain before the platform reaches the clinic. The authors are candid that short-term cytocompatibility data cannot rule out long-term graphene oxide toxicity, tissue accumulation, or delayed inflammatory responses, even though the nanomaterial was used only as a culture cue and the final therapeutic product showed no detectable graphene oxide by Raman spectroscopy or electron microscopy. Precipitation-based isolation can co-isolate contaminants below detection limits, so orthogonal purification methods such as size-exclusion chromatography and tangential-flow filtration, along with standardized release criteria and long-term biodistribution studies, will be essential. Still, the study establishes a compelling principle: that a cheap, well-characterized nanomaterial, applied at the right moment in cell culture, can simultaneously boost the quantity and the therapeutic quality of regenerative exosomes, pointing toward next-generation cell-free treatments for chronic wounds, ischemic tissue, and vascular disease.
Subject of Research: Graphene oxide-engineered iPSC spheroid platforms for enhanced production of angiogenic exosomes in tissue regeneration
Article Title: Graphene oxide-induced multiple iPSC spheroid formation for enhanced production of angiogenic exosomes accelerating tissue regeneration
Article References: Kim, W., Sharma, H., Kim, D., Lee, S., Bhakuni, S., & Kim, J. (2026). Graphene oxide-induced multiple iPSC spheroid formation for enhanced production of angiogenic exosomes accelerating tissue regeneration. Materials Today Bio, 41, Article 103723. https://doi.org/10.1016/j.mtbio.2026.103723
Image Credits: AI Generated
DOI: 10.1016/j.mtbio.2026.103723
Keywords: graphene oxide, exosomes, induced pluripotent stem cells, spheroids, angiogenesis, wound healing, regenerative medicine, VEGFR2 signaling, vascular endothelial cells, microRNA, cell-free therapy, nanobiomaterials
News Source: Drew Townsend. (October 13, 2026). Graphene oxide turns stem cell spheroids into exosome factories that heal wounds faster. Scienmag.
