Cell isolation and culture
Healthy skin tissue and keratinocytes were isolated from anonymised human abdominal tissue discarded during routine surgery at Queen Victoria Hospital, UK (NHS Trust) with consent from patients and full ethical approval from the National Regional Ethics Service (REC:06/Q1907/81). While no exclusion criteria were used, all donors (n = 6) were female aged 44–66 years (average age of 53). To isolate keratinocytes, full-thickness dermatomed skin slices were first incubated (37 °C, 5% CO2) in dispase II (2U/ml) for 30 min, the epidermis isolated and incubated (10 min) in 0.5% trypsin to free keratinocytes. Isolated keratinocytes were collected by centrifugation (500G, 5 min) in 10 ml DMEM (10% FCS, 1% penicillin/streptomycin) and cultured in CnT-O7S KC defined medium (CalTag, UK) supplemented with 1% penicillin/streptomycin at a seeding density of 2 × 106 per T75 flask. The culture medium was replaced every three days and cells were used at 85% confluence and between passages 1 and 3. All enzymes used were prepared in Hank's Balanced Salt Solution (HBSS) supplemented with 10% penicillin/streptomycin (Gibco Life Technologies, UK).
Keratinocyte seeding and culture on skin explants
Punch biopsies (1 cm diameter) of split-thickness skin shaved to a depth of 300 (for papillary dermis explant) and 600 µm (for reticular dermis explant) (Fig. 1a, b) were placed, with epidermis still attached, in ThinSert® membraned cell culture inserts (Greiner Bio-One Ltd., Stonehouse, UK) upside down (dermal-side up) and cultured in 6-well plates for 7–21 days. Autologous keratinocytes were then seeded (1 × 103 cell per biopsy) on the exposed dermal side (Fig. 1c) and the cultures supplemented with 2 ml culture media allowing for an air–liquid interface to be created at the surface. Three technical replicates were performed. Culture conditions were 37 °C and 5% CO2 with culture medium changes every 2 days.
After 0, 3, 7, 14, and 21 days in culture, skin explants were fixed with 4% neutral-buffered formaldehyde for 10 min, processed in paraffin, and wax embedded. Tissue sections (8 µm thick) were stained with haematoxylin and eosin (H&E) for morphological evaluation and with Movat-Russel’s Modified Pentachrome  to resolve different ECM features. Between stains, slides were rinsed in alcoholic and acidic solutions.
Immunostaining for cytokeratin 14 (CK-14), collagen VII (Col VII), laminin, and connexin 26 and 43 (cx26/43) was performed with antibodies specific for each protein (list below). For antigen retrieval, sections were incubated (20 min, 37 °C) with proteinase K (Abcam, Cambridge, UK) and permeabilized with 0.1% Triton X100 (Sigma-Aldrich, St. Louis, USA) in PBS with Tween 20 and 1% normal goat serum (PBST) (30 min, 30 °C). Non-specific binding was then blocked with 10% normal goat serum in PBST (Vector Laboratories Inc., Burlingame, USA) (all samples) and 1% Bovine Serum Albumin (BSA) in PBS (col VII section) (2 h, 20 °C). Sections were then incubated with primary antibodies for either 2 h, ≈ 20 °C (laminin and, cx 26 and anti 43) and overnight (≈ 12 h) (CK-14 and col vii), followed by secondary antibodies for 30 min (anti-rabbit) and 2 h (anti-mouse) at room temperature.
Primary mouse monoclonal antibodies used: anti-CK14 (LL002, dilution 1:1000, Abcam, Cambridge, UK), anti-col vii (LH7.2, dilution 1:1000, in PBST, Abcam, Cambridge, UK), and anti-Cx26 (CX-1E8, dilution 1:2000 in PBS, Invitrogen, Carlsbad, USA). Primary rabbit polyclonal antibodies used: anti-laminin (AB30320, dilution 1:200 in PBST, Abcam, Cambridge, UK) and anti-Cx43 (SAB4501173, dilution of 1:2000 in PBST, Sigma-Aldrich, St. Louis, USA). The secondary antibodies used: monoclonal DyLight®-488 (Green) anti-mouse IgG and polyclonal DyLight®-549 (Red) anti-rabbit Ig, both at a dilution of 1:200 in PBS (Vector Laboratories Inc., Burlingame, USA).
All sections were re-mounted in Vectashield + DAPI antifade mounting medium (Vector Laboratories Inc., Burlingame, USA) and imaged on a Zeiss Axioscope A.1 microscope equipped with AxioVision v4.8.2 software (Carl Zeiss AG, Oberkochen, Germany). Positive immunostaining for laminin and the number of cx 26 and 43 positive cells per mm2 of tissue within a depth of 200 µm from the exposed surface in two sections of each sample were quantified using ImageJ 1.52C software. Cells and positive stained features in the epidermis and basement membrane within 100 µm distance from the basement membrane into the dermis were excluded from the analysis. All laminin positive stain in blood vessels was excluded wherever it was detected in tissue.