Immediately following any sort of epithelial injury, stromal keratocytes underlying the epithelial injury undergo rapid keratocyte apoptosis (Wilson, et al., 1996
). It is possible that other stromal cells such as Langerhans’ cells, nerves and a few resident and circulating inflammatory cells could also be caught up in the wave of apoptosis, but there is no conclusive evidence one way or the other. In species with thin corneas, such as the mouse, one occasionally notes corneal endothelial cells that also appear to undergo apoptosis in response to extensive corneal epithelial injury, such as scrape (Wilson, et al., 1996
The injury precipitating programmed keratocyte death can be produced by epithelial scrape (), incisional injuries from scalpel blades or microkeratomes, or even significant pressure of a contact lens on the epithelial surface (Wilson, et al., 1996
; Wilson, 1998
; Helena, et al., 1998
; Mohan, et al., 2003
). Most commonly, apoptosis is detected using the TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling) assay, and with this assay the labeling peaks at approximately four hours after epithelial injury (Wilson, et al., 1996
). However, using transmission electron microscopy, it can be noted that chromatin condensation, cell shrinkage, and budding of apoptotic bodies begins immediately after epithelial injury (Wilson, et al., 1996
). It cannot be emphasized enough how important it is to confirm TUNEL assay results using another method when studying a new system if at all possible. Unfortunately, under some circumstances, the TUNEL assay also labels cells undergoing necrosis where there is random degradation of deoxyribonucleic acid (DNA). There are two important examples of this that have been noted in our laboratory. The first is when TUNEL labeling of cells in the anterior stroma after epithelial scrape injury or epithelial scrape with photorefractive keratectomy continue to be noted for a week or more after the injury even though transmission electron microscopy shows that by a few days after injury almost all the cells that are continuing to die are undergoing necrosis (Mohan, et al., 2003
). The second example is when the femtosecond laser is used to make a lamellar cut in the stroma, without injury to the epithelium, keratocytes surrounding the cut label with the TUNEL assay and are found to be dying only by necrosis when the tissue is studied with transmission electron microscopy (Netto, et al., in press
). Thus, confirmation of TUNEL results in a particular system is critical. The gold standard for this remains transmission electron microscopy. There is some hope that other methods, such as immunocytochemical detection of activated components of the apoptosis cascade, such as activated caspase-3, will supplant the laborious transmission electron microscopy method. However, to date, we have not been convinced of the reliability of these newer methods and, therefore, do not depend on them when working with a new model in which apoptosis has not been previously verified by characteristic cellular morphology detected using transmission electron microscopy.
Fig. 1 Superficial keratocyte apoptosis (red label indicated by arrows) at four hours after epithelial scrape and −9 diopter photorefractive keratectomy injury in a rabbit cornea detected with a fluorescent TUNEL assay. Intact nuclei of residual keratocytes (more ...)
What is the mechanism(s) that triggers keratocyte apoptosis following corneal epithelial injury? Any mechanism proposed must account for interesting observations that have been made regarding the process. First, it is not necessary to remove the epithelium to trigger keratocyte apoptosis. Pressure on a contact lens can trigger underlying anterior stromal keratocyte apoptosis without removal of the epithelium (Wilson, 1998
). Second, a lamellar cut through the peripheral epithelium and into the central corneal stroma, such as the incision produced by a microkeratome, triggers keratocyte apoptosis anterior and posterior to the lamellar cut across the diameter of the cornea, far removed from the site of epithelial injury, even if the resulting flap is not lifted (Helena, et al., 1998
). We believe that these observations suggest that keratocyte apoptosis is triggered by soluble mediators released from injured corneal epithelial cells themselves. A large body of work has suggested that interleukin-1 (Wilson, et al. 1996
) and tumor necrosis factor α (Mohan, et al., 2000b
) released from injured epithelial cells are involved in mediating keratocyte apoptosis, probably via more specific cell death pathways such as those that involve Fas and Fas ligand (Wilson, et al., 1996
; Mohan, et al., 1997
There has been work suggesting that the tear film contains mediators that can trigger keratocyte apoptosis (Zhao, Nagasaki, and Maurice, 2001; Zhao and Nagasaki, 2003
). It makes sense that modulators that can induce keratocyte apoptosis would be present in tears since the tears continuously bathe the corneal epithelium. However, the rabbit or mouse eye can be removed from the animal and bathed extensively before epithelial scrape and anterior stromal keratocyte apoptosis still occurs in response to the injury (Mohan R.R., Mohan R.R., Ambrósio R. Jr., Wilson S.E. Activation of keratocyte apoptosis in response to epithelial scrape injury does not require tears. Association for Research in Vision and Ophthalmology (ARVO) Annual Meeting, Program No. 1679, May, 2002) and, therefore, tears are not required for induction of keratocyte apoptosis.
What is the function of such the elaborate early keratocyte apoptosis system that has been identified in the corneas of all species examined, including mouse, rabbit, pig, and human (Ambrosio R. Jr., Wilson S.E., unpublished data, 2002)? We hypothesized that the early apoptosis response to epithelial injury serves as an initial defense mechanism against posterior extension of infectious organisms, such as herpes simplex and adenovirus, that initially infect the epithelium, but have the capacity to extend to the keratocytes, corneal endothelium, retina, and even into the central nervous system (Wilson, et al., 1997
). Once a cornea infection begins, it takes several hours for immune cells to mobilize into the cornea to fight the infectious organism. Rapid death of underlying keratocytes in response to viral injury to the epithelium sets up a firebreak of sorts to retard extension of the virus into the deeper cornea where it gains access to the endothelium, and the retina and central nervous system beyond. Consistent with this hypothesis, studies in stat 1 null mice defective in apoptosis (Mohan, et al., 2000a
) found that mutant mice, but not control normal mice, were susceptible to virus extension into the central nervous system when the cornea was infected with herpes simplex virus (Hill J.M. and Wilson S.E., unpublished data, 2001). Such an infection would represent a major threat to the survival of the organism and there would likely be a selective advantage to having an effective early response.