Regeneration of corneal endothelium following complete endothelial cell loss in rat keratoplasty

Abstract

Purpose: Corneal endothelial cells (EC) are crucial for maintaining corneal clarity before and after keratoplasty. Since it is thought that corneal graft rejection leads to irreversible EC loss and transplant failure, we quantified immune mediated EC loss in the rat keratoplasty model and analyzed whether the EC layer would then regenerate.

Methods: Rats were subjected to orthotopic penetrating keratoplasty. We compared endothelial responses to immunological EC loss following allogeneic transplantations between Fisher and Lewis rats (group R) to those following mechanical EC removal in a syngeneic setting between Lewis rats (group S). Animals were followed clinically for corneal opacity for up to one year. Bulbi were excised and prepared for histological examination at different time points: ECs were defined and characterized using Alicarin red S/ DAPI staining on corneal flatmounts. Ki-67/ DAPI staining on flatmount preparations served to detect cell proliferation. Immunohistochemical staining of corneal cryosections was used to characterize infiltrating immune cells.

Results: GROUP R: After about two weeks the allografts were completely opaque, which was accompanied by a massive leukocyte infiltration in conjunction with EC destruction, signifying rejection. EC loss without an immune reaction (group S) resulted only in medium opacity levels. In both groups, all grafts regained clarity in the following weeks to months, and a newly-formed endothelial cell layer with irregular and enlarged ECs became apparent on the formerly EC free grafts. Scattered Ki-67 positive cells within the endothelial cell layer were observed during re-endothelialization. In addition to re-endothelialization, the immunological infiltration seen in the allografts at the time of rejection had subsided after one year.

Conclusions: Re-endothelialization following keratoplasty takes place in the rat in vivo and restores graft clarity, following both immunological or surgical destruction of ECs. Following rejection, EC replacement is accompanied by a reduction of immune infiltrates. Peripheral recipient ECs are a sufficient source for graft re-endothelialization, as seen in rats following EC removal. Our results suggest that ECs both proliferate and enlarge during re-endothelialization in the rat keratoplasty model.

Figure 1

Endothelial cell loss on rejected corneal allografts. A: Kaplan–Meier survival analysis: 100% rejection rate of allografts (group R), no rejections for syngeneic controls with endothelium (control group) or following mechanical endothelial removal before syngeneic transplantation (group S). B, C: Group R: Endothelium at the time of rejection (day 15 after surgery). B: Staining with Alicarin red S. C: Same cut, DAPI staining of cell nuclei. Superior; Intact host endothelium; Inferior: Graft with endothelium destroyed following immune reaction. Arrow: Suture in the underlying stroma, indicating the graft border. D, E: Control group: Endothelium of untreated syngeneic controls at the corresponding time point. Superior: Intact host endothelium; Inferior: Intact graft endothelium. Arrow: Suture in the underlying stroma, indicating the graft border.

Figure 2

Corneal graft (border zone) on the day of surgery. A: Group S: Transplantation of endothelium-free transplant after endothelial debridement. The host endothelium is intact, while the graft is devoid of endothelium (see magnification). B: Control group: Transplantation of an untreated graft with endothelium visible on the host cornea and the graft (magnification) directly following surgery. Straight arrow: Suture indicating graft border. Alicarin red S flatmount staining; Magnification: Overview 100×, details 400×.

Figure 3

Corneal opacity following keratoplasty. Group R (left): Allogeneic transplantation. Rejection defined as complete graft opacity (opacity grade 4) occurred on average on day 15 after keratoplasty (n=16). All grafts regained clarity in the following weeks to months (follow-up 60 days: n=8, follow-up 1 year: n=8). Group S (middle): syngeneic transplantation of a transplant with surgically removed endothelium. After initially higher opacity compared to the allogeneic group, opacity never reached the maximum grade of 4, thus no rejection occurred (n=11). Control group (right): syngeneic transplantation without prior endothelial alteration. Initial opacity levels are lower compared to the EC-free syngeneic grafts from group S (n=8). (Shown: median±upper/lower quartile).

Figure 4

Endothelial regeneration. A, B: No endothelial cells visible on the grafts, while the host endothelium is intact. A: First day after corneal transplantation of an endothelial cell-free syngeneic graft of group S. B: Time of rejection in group R. C, D: Six weeks later newly-formed endothelial cells are visible on the grafts. These cells show an irregular shape and are larger than those in the peripheral recipient endothelium (C: Group S; D: Group R). A-D: Superior: Host cornea; Inferior: Graft. Black lines: Sutures indicating the graft border. Magnification 100×. E: Following re-endothelialization in group S, mean endothelial cell density on the graft is markedly lower than on the host cornea (Group S, n=11). F: Ki-67 (MIB-1) immunostains (green) of a corneal flatmount counterstained with DAPI (blue) shows endothelial cell division (arrow) of host endothelium adjacent to the graft (6 days following surgery). Magnification 630×.

Figure 5

Immune infiltration and corresponding clinical picture (group R). A, B: Group R: Rate of immune cells infiltrating the graft at the time of rejection (A; n=6). At this time point, the graft is completely opaque (B). C, D: Group R: Significant reduction in infiltrating immune cells one year after surgery (C; n=6; *p<0.001). Graft clarity is restored (D). A and C: Boxplots. Arrows in B and D mark sutures indicating the graft border.