A fluorescence-based cell-viability assay was used to test endothelial cell viability in ex vivo corneas. Control, untransfected corneas and corneas incubated for 3 hours with either the pIRES2-EGFP or pIRES2-E2F2/EGFP plasmid were incubated for 48 hours in M199 and 0.1% FBS before assay. The 48-hour time point was chosen, since this was the time determined to yield the greatest number of E2F2-positive cells.
3 presents representative fluorescence micrographs of the results. The endothelium of untransfected corneas
3 shows mainly green nuclear fluorescence (calcein-AM), indicating live cells and only a few red nuclei (ethidium-D), indicating dead cells. Similar staining patterns were observed in the endothelium of ex vivo corneas transfected with either the pIRES2-EGFP
3 or pIRES2-E2F2/EGFP plasmid
3 . Cell-viability and TUNEL assays were conducted on fully confluent cultures of rabbit corneal endothelial cells to evaluate further the effect of lipid-based transfection and E2F2 overexpression on viability. As indicated in the Materials and Methods section, culture conditions included the following: (1) incubation of confluent cells without lipid reagent (LipofectAMINE-Plus; Invitrogen) or plasmid, (2) incubation of confluent cells with lipid reagent, but without plasmid, (3) transfection with the pIRES2-EGFP control plasmid, or (4) transfection with pIRES2-E2F2/EGFP plasmid. As with the ex vivo corneas, assays were performed 48 hours after transfection. Representative micrographs of the cell-viability assay are presented in
4 .
1 presents the quantitative data. For this assay, methanol-fixed cultures acted as positive controls for cell death. Fluorescence microscopy of methanol-fixed cultures revealed red-stained nuclei throughout the culture
4 . A small number of dead cells were observed in control cultures incubated without transfection reagent or plasmid
4 . These cells were interspersed between the live cells of the endothelial monolayer. Cultures incubated in the reagent but without plasmid
4 contained a larger number of red-stained nuclei. Of note was the presence of cellular debris associated with red-stained nuclei in the lipid-treated cultures. This debris was not as evident in control cultures incubated in the absence of the reagent, but was found in cultures transfected with either the pIRES2-EGFP
4 or pIRES2-E2F2/EGFP plasmid
4 4 , using a lipid-based transfection reagent. As shown in
1 , cell death increased from 0.657% to 27.16% when cultures were incubated in lipid-based reagent alone. Little or no additional cell death was observed in cultures incubated in this reagent plus either pIRES2-E2F2/EGFP or the control plasmid. The same incubation conditions were used to test for apoptosis using the TUNEL assay. Representative micrographs are presented in
5 and quantification of the results is given in
2 . For this assay, hydrogen peroxide (H
2O
2) treatment was used as a positive control for apoptosis. As shown in
5 , 24 hours after H
2O
2 treatment most of the cells had died and lifted from the chamber slide. All remaining attached cells were TUNEL positive. No TUNEL-positive apoptotic cells were observed in cultures incubated in the absence of transfection reagent and plasmid
5 . In cultures incubated with reagent but without plasmid
5 , very few cells (0.324%) were TUNEL positive. No TUNEL-positive cells were observed in cultures transfected with pIRES2-EGFP
5 and only 0.059% of cells transfected with the pIRES2-E2F2/EGFP plasmid were TUNEL positive
5 5 .