Cell culture

Neonatal normal human dermal fibroblasts (NHDF, CC-2509; Lonza, Basel, Switzerland) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Nacalai Tesque, Kyoto, Japan) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin, and 0.25 µg/mL amphotericin B at 37 °C in 5% CO2. Human umbilical vein endothelial cells (HUVEC; Lonza) and GFP-HUVEC (Angio-Proteomie, Boston, MA, USA) were cultured in endothelial cell medium (EGM-2MV, Lonza). After reaching 80–90% confluency, the NHDF and HUVEC were subcultured using 0.25% trypsin/EDTA solution (Lonza) and used at passage 5 to construct the 3D skin models. Neonatal human epidermal keratinocytes (HEKn, Gibco, Carlsbad, CA, USA) were cultured in EpiLife® medium with 60 μM calcium (Invitrogen, Carlsbad, CA, USA) supplemented with HuMedia-KG growth factor (KURABO, Osaka, Japan). Media were changed every other day.

Preparation of vascularized 3D skin substitutes

The 3D skin containing the vascular network was constructed via an LbL technique that enabled the coating of individual cells with FN-G ECM films on NHDF surfaces without damage. Briefly, NHDF were alternately incubated for 1 min with 0.04 mg/mL fibronectin (FN) or gelatin (G) solutions in phosphate-buffered saline (PBS) and rinsed with PBS between each coating step. The coating procedures were performed a total of 9 times (FN: 5 times; G: 4 times). Then, to construct the dermal layer, FN-G-coated NHDF (viability >97%, 1 × 107 cells/insert) and HUVEC (0.1 × 105, 0.2 × 105, or 1 × 105 cells/insert) or FN-G-coated NHDF alone were seeded onto 12-well culture inserts (1.12-cm2 membrane growth area, 0.4-μm pores; Corning Life Science, Tewksbury, MA, USA) pre-coated with FN; the inserts were cultured in DMEM containing 5% FBS for 24 hours. In other experiments, GFP-HUVEC were seeded to facilitate the monitoring of vessel formation during the cultures. After 24 hours coculture of FN-G-coated NHDF and HUVEC, type IV collagen solution (from human placenta, 0.2 mg/mL in PBS; Sigma-Aldrich, St. Louis, MO, USA) was dropped onto the upper surface of the dermis, and the constructs were incubated for 30 min. Subsequently, HEKn (passage 4, 1 × 106 cells/insert) were seeded onto the collagen type IV-coated dermal layer, then cultured for 24 hours in growth medium (5% FBS/DMEM and EpiLife® at a ratio of 1:1), and air–liquid interface-cultured for 7 days to drive epidermal differentiation, stratification, and cornification in growth media supplemented with 25 μg/mL ascorbic acid. The medium in the lower chamber was replaced every day; flooding of the top chamber was avoided. For in vivo experiments, we seeded 0.2 × 105 HUVEC per construct.

Cell viability assay in 3D skin substitutes

We determined cell viability in the 3D skin substitutes using the LIVE/DEAD Viability/Cytotoxicity kit (L-3224; Molecular Probes, Eugene, OR, USA) according to the manufacturer’s protocol. Briefly, at 2, 4, and 7 days after organotypic coculture, we rinsed the 3D skin substitutes (n = 3 for each time point) 3 times with PBS, then stained with 2 μM calcein AM (which stained live cells green) and 0.5 μM ethidium homodimer (which stained dead cells red) in PBS for 30 min at room temperature. After incubation, the samples were washed twice with PBS and imaged immediately using a Biozero BZ-X700 fluorescence microscope (Keyence, Osaka, Japan).

We acquired 3 images of different portions of each triplicate sample for each time point. Still images of the middle section of the 3D skin substitute (~100 to 120 µm from the culture insert membrane) were captured and serial micrographs were later combined for z-stacked compilation images. The experiment was performed twice.

Transplantation of tissue-engineered 3D skin substitutes

All surgical procedures were conducted according to protocols approved by the National Defense Medical College Animal Care and Use Committee (Permit number: 17012). Pre-vascularized 3D skin substitutes were transplanted to excisional wounds in C.B-17 SCID mice. Eight-week-old male mice (body weight, 24.2 ± 3.7 g; Charles River Laboratories, Kanagawa, Japan) were randomly divided into 3 groups (n = 15 animals/group). The excisional wound splinting model was employed as previously described30. In brief, mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg), then their dorsal surfaces were shaved with electric clippers and treated with a depilatory agent to remove remaining hair. Under sterile conditions, a circular full-thickness excisional wound (8-mm diameter) was created on the middle dorsum of each mouse using a biopsy punch. A donut-shaped silicone splint (0.5-mm thick, 12-mm inner diameter, 20-mm outer diameter) was placed around the perimeter of the wound and fixed to the skin with cyanoacrylate glue (Aron Alpha, Toagosei, Tokyo, Japan) and interrupted 6-0 nylon sutures (Fig. 1). The splint was used to minimize wound contraction. The skin substitutes were cut according to the dimensions of the full-thickness wound, and each wound was treated with a different substitute: pre-vascularized 3D skin, non-vascularized 3D skin (negative control), or acellular synthetic bilayer skin, which is a collagen-glycosaminoglycan sponge with a temporary silicone rubber epidermal barrier. The grafts were covered with a non-adherent silicone gel dressing (SI-AID, ALCARE, Tokyo, Japan) and wrapped with a semipermeable adhesive film (OPSITE, Smith & Nephew, Largo, FL, USA) to protect the wound site and prevent desiccation. The animals were housed individually. After 7 (n = 7/group) or 14 (n = 8/group) days, the mice were euthanized and tissue biopsies of the grafted areas were collected.

Transillumination stereomicroscopy

To visualize the vascular network by transillumination, the implants and surrounding skin were excised at day 7 after grafting; specimens were then quickly placed on a Petri dish and observed under an inverted microscope.

Histology and immunostaining

To observe the vessel network formed by HUVEC in vitro after 7 days of culture, we stained whole-mount 3D skin substitutes with a fluorescently labeled anti-CD31 antibody. The constructs (n = 3) were fixed in 4% paraformaldehyde for 15 min, then permeabilized in PBS with 0.2% Triton™ X-100 for 15 min. The tissues were blocked with 1% bovine serum albumin for 1 h, then incubated with a mouse anti-human CD31 antibody (1:100; M0823, clone JC70A; Dako, Carpinteria, CA, USA). The tissues were incubated with Alexa Fluor® 546-conjugated goat anti-mouse IgG (1:500; Invitrogen) for 1 h. Immunofluorescence images of the stained vessels were taken with a BZ-X700 microscope. Vascular profiles were characterized by positive staining for human CD31 in structures with an identifiable vascular lumen (>5 μm). The vascular area density and number of branching points were analyzed with AngioTool software (National Cancer Institute, Rockville, MD, USA). All analyses were carried out manually.

The grafted sites were fixed in 10% neutral buffered formalin overnight, then dehydrated and embedded in paraffin. Hematoxylin and eosin, trichrome, and immunohistochemical/immunofluorescent staining were performed on 4-μm-thick tissue sections. Trichrome staining was carried out using the Modified Gomori’s One-Step Trichrome Staining Kit (Biocare Medical, Concord, CA, USA) and immunohistochemical staining was performed with the Vector ImmPRESS™ peroxidase polymer Anti-mouse or Anti-rabbit IgG reagent kits (Vector Labs, Burlingame, CA, USA) according to the manufacturers’ instructions. For immunofluorescent staining, we incubated with Alexa Fluor® 488-conjugated (Abcam, Cambridge, MA, USA) and Alexa Fluor® 594-conjugated (Molecular Probes, Eugene OR, USA) secondary antibodies, then performed nuclear staining with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich, St. Louis, MO, USA). The following primary antibodies were used for immunostaining of: a human–specific CD31 (1:200, NBP2-15202, clone C31.3; Novus Biologicals, Centennial, CO, USA), a mouse–specific CD31 (1:100, 14-0311-82, clone 390; Invitrogen), HLA-Class I ABC (1:2,500, ab70328, clone: EMR8-5; Abcam), and laminin 5 (1:200; ab14509, Abcam). The MOM kit (Vector Labs) was used for mouse derived monoclonal antibodies.

Statistical analysis

The data are expressed as the mean ± standard deviation (SD). Statistical analyses were performed by unpaired Student’s t-test or one-way analysis of variance (ANOVA) with Tukey’s multiple comparison post hoc test using GraphPad Prism 7.0 software (La Jolla, CA, USA). P-values < 0.05 were considered significant.