Hyperosmolar-induced ocular surface cell death is a key mitochondria-mediated event in inflammatory eye diseases. Transglutaminase (TGM)-2, a cross-linking enzyme, is purported to mediate cell death, but its link to mitochondria is unclear. In the cornea, the integrity of the epithelial cells is important for maintaining transparency of the cornea and therefore functional vision. We evaluated the role of TGM-2 and its involvement in hyperosmolarity-stimulated mitochondrial cell death in human corneal epithelial (HCE-T) cells. HCE-T cell lines stably expressing either shRNA targeting TGM-2 (shTG) or scrambled shRNA (shRNA) were constructed. Hyperosmolar conditions reduced viability and increased mitochondrial depolarization in shRNA cells. However, hyperosmolarity failed to induce mitochondrial depolarization to the same extent in shTG cells. Transient overexpression of TGM-2 resulted in very high levels of TGM-2 expression in shTG and shRNA cells. In the case of shTG cells after overexpression of TGM-2, hyperosmolarity induced the same extent of mitochondrial depolarization as similarly treated shRNA cells. Overexpression of TGM-2 also elevated transamidase activity and reduced viability. It also induced mitochondrial depolarization, increased caspase-3/7 and -9 activity, and these increases were partially suppressed by pan-caspase inhibitor Z-VAD-FMK. Corneal epithelial apoptosis via mitochondrial dysfunction after hyperosmolar stimulation is partially dependent on TGM-2. This TGM-2-dependent mechanism occurs in part via caspase-3/7 and -9. Protection against mitochondrial stress in the ocular surface targeting TGM-2 may have important implications in the survival of cells in hyperosmolar stress.

No abstract

We recently demonstrated that chemical proteasome inhibition induced inner retinal degeneration, supporting the pivotal roles of the ubiquitin–proteasome system in retinal structural integrity maintenance. In this study, using beclin1-heterozygous (Becn1-Het) mice with autophagic dysfunction, we tested our hypothesis that autophagy could be a compensatory retinal protective mechanism for proteasomal impairment. Despite the reduced number of autophagosome, the ocular tissue morphology and intraocular pressure were normal. Surprisingly, Becn1-Het mice experienced the same extent of retinal degeneration as was observed in wild-type mice, following an intravitreal injection of a chemical proteasome inhibitor. Similarly, these mice equally responded to other chemical insults, including endoplasmic reticulum stress inducer, N-methyl-D-aspartate, and lipopolysaccharide. Interestingly, in cultured neuroblastoma cells, we found that the mammalian target of rapamycin-independent autophagy activators, lithium chloride and rilmenidine, rescued these cells against proteasome inhibition-induced death. These results suggest that Becn1-mediated autophagy is not an effective intrinsic protective mechanism for retinal damage induced by insults, including impaired proteasomal activity; furthermore, autophagic activation beyond normal levels is required to alleviate the cytotoxic effect of proteasomal inhibition. Further studies are underway to delineate the precise roles of different forms of autophagy, and investigate the effects of their activation in rescuing retinal neurons under various pathological conditions.