Corneal wounds usually heal quickly; but diabetic patients have more fragile corneas and experience delayed and painful healing. In the present study, we compared the healing capacity of corneal epithelial cells (CECs) between normal and diabetic conditions and the potential mechanisms. Primary murine CEC derived from wild-type and diabetic (db/db) mice, as well as primary human CEC were prepared. Human CEC were exposed to high glucose (30 mM) to mimic diabetic conditions. Cell migration and proliferation were assessed using Scratch test and MTT assays, respectively. Reactive oxygen species (ROS) production in the cells was measured using dichlorofluorescein reagent. Western blot was used to evaluate the expression levels of Akt. Transepithelial electrical resistance (TEER) and zonula occludens-1 (ZO-1) expression were used to determine tight junction integrity. We found that the diabetic CEC displayed significantly slower cell proliferation and migration compared with the normal CEC from both mice and humans. Furthermore, ROS production was markedly increased in CEC grown under diabetic conditions. Treatment with an antioxidant N-acetyl cysteine (NAC, 100 μM) significantly decreased ROS production and increased wound healing in diabetic CEC. Barrier function was significantly reduced in both diabetic mouse and human CEC, while NAC treatment mitigated these effects. We further showed that Akt signaling was impaired in diabetic CEC, which was partially improved by NAC treatment. These results show that diabetic conditions lead to delayed wound-healing capacity of CEC and impaired tight junction formation in both mice and human. Increased ROS production and inhibited Akt signaling may contribute to this outcome, implicating these as potential targets for treating corneal wounds in diabetic patients.

No abstract

Basement membrane is composed of ECM proteins that have viscoelastic properties. When the viscoelasticity is mimicked in vitro, epithelial cells coalesce by ''dragging'' the ECM protein through the PDMS substrate. This mechanosensing of viscoelasticity is achieved through the translocation of vinculin from the focal adhesions to the cell-cell junctions and is sensitive to the level of vinculin in the cell. Apart from the composition of cell-matrix and cell-cell adhesion complexes within the cell, we find that other biophysical and biochemical cues from environment affect the cell response on a viscoelastic substrate. By varying the interfacial force between ECM protein, fibronectin, and the PDMS substrate through physisorption or covalent linkage, we found that increasing the adhesion force hinders the coalescence of cells on a viscoelastic substrate as if on a soft-elastic substrate, suggesting the role of ECM-substrate interaction for in vitro models. Also, stronger cell-cell adhesions cause coalescence when coated with either fibronectin or collagen-1 alone but not when coated with Matrigel, consisting of collagen-IV, laminin, and other ECM proteins. To gain further insights into this phenomenon, we are using quantitative super-resolution microscopy to investigate how the difference in ECM anchoring is sensed by the cell at the molecular level and traction force microscopy to quantify the ECM remodeling and substrate deformation. These results on the viscoelastic substrate would provide new insights into in vivo basement membrane and cell-cell dynamics in general, and help to have better in vitro model, mimicking in vivo ECM arrangement and viscoelasticity, where these are crucial.