Transplantation of cultured adult human or porcine corneal endothelial cells onto human recipients in vitro. Part II: Evaluation in the scanning electron microscope
M Böhnke,P Eggli,K Engelmann
DOI: https://doi.org/10.1097/00003226-199903000-00011
IF: 3.152
Cornea
Abstract:Purpose: To evaluate the morphology of endothelial monolayers, which have been regrafted onto the denuded Descemet's membrane, with scanning electron microscopy (SEM). Methods: Material derived from each of the experimental groups described in part I of this investigation was evaluated in the current study. Recipient corneas, denuded of their native endothelium by mechanical, chemical, or physical debridement, were examined to assess the effectiveness of each technique in killing and removing cells. Porcine or human donor corneal endothelial cells maintained in monolayer culture for up to 10 passages then were seeded onto the denuded Descemet's membranes of recipients in the absence or presence of fibroblast growth factor (FGF). The monolayers thereby established were examined in the SEM, and the morphologic status of individual cells compared with that manifested in normal human donor corneas maintained for 4 weeks in organ culture (reference control). Isolated and cultured human keratocytes regrafted onto the denuded Descemet's membranes of recipient corneas served as nonendothelial control specimens. Tissue was processed for examination in the SEM according to standard techniques. Results: Each of the three methods used to strip recipient corneas of their native endothelium was effective and elicited no gross structural damage to Descemet's membrane. Some small focal defects within this latter layer were, however, observed, these being encountered at higher frequency after mechanical debridement than after chemical or physical stripping. Porcine or human endothelial cells seeded onto the denuded Descemet's membranes of recipient corneas formed stable monolayers. The morphologic status of regrafted cells corresponded to that manifested in monolayer cultures before seeding, porcine ones always being more differentiated than their human counterparts. Poorly differentiated human endothelial cells had a slender, elongated, fibroblast-like appearance, whereas more highly differentiated ones manifested broad, flat, polygonal profiles. Monolayers covered the entire corneal surface and impinged to a variable degree onto the trabecular meshwork, at which juncture cells always assumed a less well-differentiated morphology. FGF consistently effected an increase in differentiation status, and as this became augmented, the capacity of monolayers to violate the corneal-trabecular meshwork border was correspondingly repressed. Seeded keratocytes formed dense, multilayered sheaths, resembling retrocorneal membranes, across the entire corneal surface, trabecular meshwork, and iris root. The surface characteristics of the constituent cells were quite distinct from those manifested by endothelial cells, even the least well-differentiated ones. Conclusion: Regrafting of human corneal endothelial cells onto the denuded Descemet's membranes of recipients resulted in the formation of stable monolayers. Because the morphologic status of seeded cells closely mimicked that manifested in monolayer cultures before transplantation, it may be anticipated that efforts to refine and optimize culturing conditions would yield improvements in this parameter after regrafting. If these expectations can be realized, then the possibility of successfully establishing a "new" and functional endothelium on recipient corneas destined for clinical grafting may well be brought to fruition in the not-too-distant future.