Bioengineering HA-specific FVIII-secreting vascular networks in hemophilic mice. (A) Schematic of the microvascular graft model. Grafts were prepared by combining either HA-iECs (n = 5) or HA-FLF8-iECs (n = 10) with MSCs in hydrogels followed by subcutaneous injection into immunodeficient hemophilic mice (SCID-f8ko). Images are macroscopic views of the explanted grafts at day 7. (B) Hematoxylin-and-eosin (H&E) staining of a representative explanted HA-FLF8-iEC graft. Underneath, representative pictures are taken from 4 separate regions of the graft. Perfused microvessels (yellow arrows) identified as lumenal structures containing erythrocytes. (C) Comparison of microvessel density (perfused vessels per mm²) between HA-iEC and HA-F8FL-iEC implants. Bars represent mean ± SD. (D) Immunofluorescence staining of explanted grafts after 7 days in vivo. Human lumens stained by h-CD31. Perivascular coverage stained by α-SMA. Nuclei stained by DAPI. (E) Percentage of human lumens with α-SMA⁺ perivascular coverage in explanted grafts on day 7. Bars represent mean ± SD. (F) Immunofluorescence staining of FVIII in explanted grafts after 7 days in vivo. Human lumens stained by h-CD31. Nuclei stained by DAPI. Human vessels in grafts formed with HA-FLF8-iECs overexpressed FVIII, whereas grafts formed by unedited HA-iECs had virtually undetectable levels of FVIII. Scale bars, 5 mm (A), 500 µm (B, gross), and 100 μm (B,D,F).