Investigation of two pore types in Schlemm's canal inner wall endothelium in flow-type areas using serial block-face scanning electron microscopy.

With the introduction of automated serial block-face imaging, we can now more accurately investigate the characteristics of pores in the inner wall endothelium of Schlemm's canal (SC). Previous studies analyzing pores using two-dimensional (2D) scanning or transmission electron microscopy were limited to either en face or random cross-sectional imaging and may have overlooked some of the pores. Additionally, the pores' morphological characteristics in different segmental flow areas (i.e. high-, low-, and non-flow) have not been fully explored. This study aimed to further characterize the two types of pores in different flow areas of human eyes using serial block-face scanning electron microscopy (SBF-SEM). Two normal human eyes were perfused at 15 mmHg with fluorescent tracers to label high-, low-, and non-flow outflow areas before being dissected, sectioned serially (thickness = 0.13 μm), and imaged. A total of 10,832 serial images were examined to identify all intracellular pores (I-pores) and intercellular or border pores (B-pores). Pore densities per unit inner wall area and per inner wall cell nucleus were measured, and pore sizes were determined. The locations of I-pores on giant vacuoles (GVs), whether on the top or sides of GVs, were classified. Pores that would be potentially missed using 2D scanning electron microscopy and I-pores on the edges of inner wall cells that could be mistaken for B-pores were identified. We observed 729 pores, of which 656 (90.0 %) were GV-associated I-pores, 25 (3.4 %) were non-GV-associated I-pores, and 48 (6.6 %) were B-pores. Overall, there were significantly more I-pores (93.4 %) than B-pores (6.6 %). Of all I-pores, the majority (96.3 %) were associated with GVs. A significantly greater number of GV-associated I-pores were located on the side of GVs (413/656, 63.0 %), compared to the top of GVs (243/656, 37.0 %; P ≤ 0.01), which was observed similarly in all flow-type areas. I-pore density was higher in high-flow areas compared to low- or non-flow areas in both eyes, whereas B-pore densities did not appear to differ between flow-type areas. From an en face view via traditional SEM, 6.6 % of I-pores were on the sides of GVs and/or near the edge of cells and could have been mistaken as B-pores. Additionally, 32.0 % of all pores could have been missed by 2D-SEM due to obstruction by neighboring cells/GVs (31.4 % of I-pores and 39.6 % of B-pores). Median major and minor axes of I-pores were significantly smaller in high-flow areas, compared to non-flow areas (P = 0.024 and P = 0.048, respectively). B-pores were not significantly different in size among flow-type areas. SBF-SEM and three-dimensional (3D) reconstruction provided an improved method to identify the pores of SC inner wall, as the cells could be rotated freely to any viewpoint. This study applied a novel method to investigate the morphological characteristics of pores and found that pores in high-flow areas were more numerous and smaller than in non-flow areas, suggesting that pore morphology and density may play a role in the regulation of segmental aqueous humor outflow.