Assessment of Corneal Endothelial Barrier Function Based on "Y-Junctions": A Finite Element Analysis

Abstract

Purpose: Corneal endothelial cell dysfunction is a major contributor to corneal edema, opacity, and, in severe cases, corneal blindness. Currently, no direct and reliable clinical indicator is available for evaluating the function of corneal endothelial cells. This study aimed to identify new noninvasive indicators for the clinical assessment of corneal endothelial barrier function.

Methods: This study established a finite element simulation model of a monolayer full-corneal endothelium to screen for sensitive indicators that reflect the barrier function of the corneal endothelium. Scanning electron microscopy (SEM) was used to observe the morphology of endothelial junctions, and immunofluorescence was used to examine the expression of fluorescent particles.

Results: The "Y-junctions" area was identified as the parameter most sensitive to changes in intraocular pressure when considering the different analytical indices obtained from the finite element model. SEM of Corneal endothelial dysfunction models in rabbits and mice further confirmed a substantial increase in the "Y-junctions" relative to control groups. Additionally, functional in vitro experiments provided further evidence of a positive relationship between larger "Y-junctions" and enhanced permeability to fluorescent particles. Finally, clinical analysis of measurements related to "Y-junctions" in patients suffering from various corneal endothelial disorders consistently revealed that these junctions were significantly larger compared to those observed in healthy control subjects.

Conclusions: The "Y-junctions" area serves as a potentially sensitive indicator for assessing endothelial barrier integrity. Consistent observation of alterations in this area may facilitate the early identification of dysfunction in circulating endothelial cells.

Figure 1.

Workflow for finite element modeling of CEC monolayers. Original CEC images were processed for geometric modeling and finite element analysis. Fine meshes were applied to regions of interest, with gradual coarsening in transition areas and coarse meshes in noncritical regions. Tight junction proteins were modeled using beam elements to simulate their biomechanical behavior under varying conditions.

Figure 2.

Extraction and comparative analysis of biomechanical parameters in the corneal endothelium. (A) FEA were used to systematically extract biomechanical parameters, including displacement, stress, and “Y-junctions.” (B) In the normal group, parameters such as displacement, edge stress, “Y-junctions,” F/L, angle stress, and edge beam values showed statistically significant variations under elevated intraocular pressure (P values Bonferroni-corrected). Error bars, representing the standard error of the mean, were sometimes indistinguishable because of low data variability. (C) Under controlled intraocular pressure, “Y-junctions” demonstrated the most pronounced and statistically significant changes among the analyzed parameters.

Figure 3.

Experimental validation of “Y-junctions” and CEC barrier function. (A) Compared to the normal group (white arrow), SEM revealed a higher prevalence of large Y-junctions (black arrow) in the corneal endothelium of rabbits seven days after cryotherapy. (B) SEM of the corneal endothelium in col8a2^Q455K mice (black arrow) revealed similar findings compared to the normal group (white arrow). (C) Minimal 40 kD fluorescent particles were detected in the corneal stroma after anterior chamber injection in the normal group. (D) Cryotherapy significantly increased stromal 40 kD fluorescent particles. (E) Quantitative fluorescence analysis confirmed elevated particle levels at D7 versus the normal group. (Z-axis set to 80 µm; scale bar: 100 µm).

Figure 4.

FEA revealed significant biomechanical differences between normal individuals and patients with CED. (A) Finite element modeling was performed to compare biomechanical parameters between normal individuals and patients with CED. (B) The “Y-junctions” area in the CED group was significantly larger compared to those in the normal group, with statistical significance (P < 0.0001). (C) F/L values observed in the CED group were markedly elevated relative to the normal group, demonstrating statistical significance (P < 0.0001). (D) Angle stress values in the CED group were significantly higher than those in the normal group, with strong statistical evidence (P < 0.01). (E) A strong positive correlation was observed between the “Y-junctions” area and corneal thickness (R = 0.8464). (F) The “Y-junctions” exhibited a robust negative correlation with ECD (R = 0.8340).

Figure 5.

Workflow for evaluating CEC barrier function using FEA system.