• corneal endothelial death
• organ culture

We read with interest the remarks of Crowston et al¹ on our article.² We showed that the TUNEL technique revealed a far higher percentage of endothelial cells (ECs) irreversibly engaged in a cell death process than that obtained by trypan blue staining.

The two techniques were performed sequentially: after observation of trypan blue staining, corneas were immediately fixed in formaldehyde for TUNEL. Crowston et al suggest that the trypan blue itself and/or the time spent outside the organ culture medium before fixing in formaldehyde may have caused an artefactual increase in the percentage of TUNEL positive ECs. Two arguments counter this remark.

1. The trypan blue staining procedure is identical to that used, during endothelial examination(s) of grafts, in all European cornea banks that use organ culture during endothelial examination(s) of grafts. Neither the low concentration of trypan blue (0.4%) nor the short exposure time (about 1 minute) nor the short incubation in the presence of 0.9% NaCl has ever been incriminated in the over-mortality of ECs in routine practice.³ Moreover, the innocuity of injections of trypan blue into the anterior chamber, a common feature during cataract surgery, has been well demonstrated.⁴

2. The time spent outside the organ culture medium before fixing in formaldehyde, a period required for vital staining and microscopic examination of the endothelium, lasts only a few minutes. The cornea remains under the microscope for about 1 minute only, the time needed for image acquisition. Such rapidity is possible by using a prototype automatic analyser of the endothelium, which we developed and have recently published.⁵ This is very probably insufficient time for DNA fragmentation to occur at the level we observed. Moreover, the fixing of the endothelial layer in 10% formaldehyde is immediate, and prevents any continuation of fragmentation phenomena. On balance, it is highly unlikely that the succession of markings is responsible for the discrepancy between the positivity percentages of the two techniques. In addition, we chose not to perform the two techniques simultaneously on paired corneas or on the halves of one cornea because we wanted to superimpose the two stains on the same cornea and thus obtain a double cell staining.

The second remark by Crowston et al is particularly interesting. We too were surprised by the high percentage of TUNEL positive ECs (mean 12.7%, SD 16.4). This may imply that all the cells died within 8 days, which was evidently not the case. We believe this apparent contradiction can be explained by the following theory. The TUNEL staining is positive during a relatively long window (24–48 hours⁶). The TUNEL index, measured at a given moment, provides a global view of all the cells with fragmented DNA. However, the DNA fragmentation may be at different stages, and the cells very likely spread according to a Gaussian distribution. Therefore the cells, which are TUNEL positive at a given moment, will not all die instantaneously and simultaneously. Only the cells furthest to the right on the curve will die in the very short term, and it is probably these that are liable to be revealed by trypan blue. If it were possible to perform TUNEL on two consecutive days, the percentage of positive cells revealed would probably be very similar, but a large majority of the positive cells recorded on the second day would have already been counted on day one. It is, however, undeniable that the cells that are TUNEL positive at a given moment will all die eventually. In other words, we believe that, at the end of storage, corneas contain a number of ECs engaged in an irreversible cell death process far more extensive than the highly unreliable trypan blue staining technique suggests.