Scientists Build Functional Human Blood Vessels in Just 5 Days Using Stem Cells

Blood vessels are indispensable to virtually every tissue, yet recreating functional vascular networks from human stem cells has remained slow, technically demanding, and difficult to control. Existing vascular organoid protocols typically take three weeks or more, rely on spontaneous mural cell emergence, and require extracellular matrix embedding from the outset—limiting scalability and therapeutic translation.

The research team engineered two distinct iPSC lines, each carrying a doxycycline-inducible copy of either ETV2 (an endothelial fate determinant) or NKX3.1 (a mural cell fate determinant). Cells from both lines were first primed toward mesoderm over two days, then combined 1:1 and aggregated on an orbital shaker. Three days of doxycycline exposure simultaneously activated both transcription factors, producing thousands of uniformly sized vascular organoids (~250 μm diameter) in just five days. Approximately 50% of cells became CD31+/VE-Cadherin+ endothelial cells, while the remainder adopted mural identities expressing α-SMA and MYH11. Crucially, no exogenous ECM was required during this initial phase.

Bulk RNA sequencing and qPCR confirmed that three-dimensional co-differentiation enhanced maturation compared to standard 2D monolayer protocols: endothelial cells upregulated CDH5, VWF, ERG, TEK, and KLF2, while mural cells showed higher expression of contractile markers ACTA2, TAGLN, and MYH11, as well as TGF-β/BMP signaling components. When VOs were subsequently embedded in collagen/Matrigel hydrogels for five additional days, vascular networks expanded dramatically—reaching ~1,000 μm in diameter—with clear lumens, basement membrane deposition, and an arterial shift in endothelial gene expression. This maturation was attributable to ECM exposure itself, not simply extended culture time.

To eliminate genomic integration concerns, the authors also demonstrated successful VO generation using modified mRNA (modRNA) delivery of ETV2 and NKX3.1—a genetic footprint-free alternative compatible with clinical translation. Single-cell RNA sequencing of mature VOs revealed diverse vascular subpopulations including arterial, venous, tip-cell, and proliferating endothelial subtypes, alongside smooth muscle cells and pericytes, recapitulating in vivo vascular heterogeneity. Temporal modulation of transcription factor expression allowed tuning of arterial versus angiogenic endothelial phenotypes.

In vivo transplantation into immunodeficient NSG mice confirmed functional integration: VOs engrafted under the renal capsule formed perfused human vessels anastomosed with mouse circulation within 14 days. In hindlimb ischemia models, VO transplantation significantly improved blood flow recovery. In a pancreatic islet co-transplantation model, VOs enhanced islet engraftment and function, underscoring therapeutic versatility. Taken together, this platform offers a rapid, scalable, and controllable approach to vascular organoid production with strong potential for regenerative medicine, disease modeling, and organ engineering.