Developing improved tissue-engineered buccal mucosa grafts for urethral reconstruction

Introduction: We aimed to compare alternative synthetic scaffolds suitable for future implantation and to examine the use of an inhibitor of lysyl oxidase (beta-amino-propionitrile [β-APN]) to reduce contraction in these implants.

Methods: Three synthetic scaffolds were compared to natural dermis as substrates for the production of tissue-engineered skin. For natural dermis, Euroskin was used to provide a cell-free cadaveric dermis. Synthetic scaffolds consisted of microfibrous poly-L-lactic acid (PLA), nanofibrous poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV), and a micro-/nanofibrous trilayer of PLA-PHBV-PLA. The latter were all electrospun and then all four scaffolds (three synthetic, one natural) were placed in six well plates. A culture well was formed on the scaffold using a 1 cm diameter stainless steel ring and 1.5x105 oral fibroblasts were seeded one side; after two days of culture, the ring was placed on the other side of the scaffolds and 3x105 oral keratinocytes were seeded on to the scaffolds and cultured with keratinocytes uppermost. After a further two days of culture, scaffolds were cut to 1 cm2 and raised to an air-liquid interface on stainless steel grids; some were treated with 200 μg/ mL β-APN throughout the culture period (28 days). Contraction in vitro was assessed by serial digital photography of cell-seeded scaffolds and cell-free scaffolds three times a week for 28 days. All cell-seeded scaffolds were assessed for cell metabolic activity, mechanical properties, histology, and morphology by scanning electron microscopy (SEM).

Results: The mean fibre diameters and pore sizes of PLA and PHBV scaffolds were 2.4±0.77, 0.85±0.21 μm (p<0.001), and 10.8±2.3, 4.3±1.1 μm (p<0.001), respectively. Oral fibroblasts and keratinocytes were tightly adhered and grew well on both surfaces of trilayer. The ultimate tensile strength (UTS) and Young’s modulus (YM) of PLA samples were significantly lower than Euroskin (p<0.001 and p<0.05, respectively); only the UTS of the trilayer samples was slightly significantly lower (p<0.05). Metabolic activity was significantly increased for cells on all scaffolds, without significant differences between them from Day 0 to Day 28. There were no adverse effects of β-APN on cell viability. With respect to contraction, cells on trilayer and PHBV monolayers did not undergo any significant contraction; however, cells on PLA monolayer and Euroskin contracted 25.3% and 56.4%, respectively, over 28 days. The addition of 200 μg/ml β-APN significantly reduced contraction of Euroskin compared with the control (p<0.01); however, β-APN did not affect PLA contraction during this culture period (p>0.05).

Conclusions: This study shows that a trilayer micro-nano-3D porous synthetic scaffold is suitable for oral keratinocyte and fibroblast growth with good cell viability and minimal contraction. This material also has good mechanical properties and histological analyses showed its ability to mimic normal human oral mucosal morphology. Furthermore, synthetic trilayer scaffolds have advantages over biological scaffolds — there is no risk of disease transmission or immunological rejection and they appear resistant to contraction. We suggest they present a good alternative to allodermis for future use in urethral reconstruction.