Statistics from Altmetric.com
We read with interest the paper by Gain et al,1 which assessed two distinct techniques to quantify corneal endothelial cell death in donor corneas. A significantly higher rate of cell death was observed with the TUNEL assay, which labels nuclei with fragmented DNA, compared to the trypan blue exclusion method, which detects cells with disrupted cell membranes. The authors conclude that TUNEL analysis is more accurate than trypan blue exclusion as a means of assessing the impact of different corneal storage methods on endothelial viability.
Our experience of using a number of cell death assays to investigate fibroblast apoptosis together with the findings of others2,3 supports the notion that trypan blue exclusion is not a good method for detecting apoptosis in vitro. However, we would like to propose that the sequential analysis of the same corneal tissue in this study might account for some of the disparity observed between the two methods. Following initial incubation in trypan blue, buttons were subjected to image analysis as well as cell density measurements after further incubation in 0.9% sodium chloride. After this the buttons were fixed overnight in 10% formaldehyde in preparation for TUNEL. It is possible that the higher rates of death observed by TUNEL reflect the known toxicity of trypan blue, or are a consequence of subsequent manipulation in image analysis and cell density measurement. The low rates of cell death observed by both techniques in non-stored corneas do not negate this possibility since healthier corneas may be more resistant to the effects of trypan blue and subsequent analysis. Randomisation of the sequence of analysis between the techniques compared would not have been possible, but the authors could have divided the corneas before storage or used paired eyes as separate matched specimens.
The authors argue that the disparity between endothelial cell loss and observed cell death is greater for trypan blue exclusion because loss of membrane integrity occurs relatively late, giving a shorter observational window in which to detect dying cells than TUNEL analysis, which detects apoptosis earlier. But the relatively high percentage of apoptotic cells (12.7%) observed by TUNEL analysis may be an overestimate. Although the time span for apoptosis varies greatly depending on the cell type and nature of the apoptotic trigger, many estimates suggest that the processes is completed in less than 24 hours.4 If 12.7% of cells undergo apoptosis at any given time it can be predicted that complete endothelial cell death would occur within 8 days. The actual loss observed over the 22 day incubation period in this study was however only around 14%.
We agree with the authors regarding the need for accurate methods for determining endothelial cell death. No individual assay per se is ideal for both quantifying and determining the mode of cell death and combinations of assays should give a clearer picture of the impact of variations in corneal storage on endothelial viability.