A model of giant vacuole dynamics in human Schlemm's canal endothelial cells

Abstract

Aqueous humour transport across the inner wall endothelium of Schlemm's canal likely involves flow through giant vacuoles and pores, but the mechanics of how these structures form and how they influence the regulation of intraocular pressure (IOP) are not well understood. In this study, we developed an in vitro model of giant vacuole formation in human Schlemm's canal endothelial cells (HSCECs) perfused in the basal-to-apical direction (i.e., the direction that flow crosses the inner wall in vivo) under controlled pressure drops (2 or 6 mmHg). The system was mounted on a confocal microscope for time-lapse en face imaging, and cells were stained with calcein, a fluorescent vital dye. At the onset of perfusion, elliptical void regions appeared within an otherwise uniformly stained cytoplasm, and 3-dimensional reconstructions revealed that these voids were dome-like outpouchings of the cell to form giant vacuole-like structures or GVLs that reproduced the classic "signet ring" appearance of true giant vacuoles. Increasing pressure drop from 2 to 6 mmHg increased GVL height (14 ± 4 vs. 21 ± 7 μm, p < 0.0001) and endothelial hydraulic conductivity (1.15 ± 0.04 vs. 2.11 ± 0.49 μl min⁻¹ mmHg⁻¹ cm⁻²; p < 0.001), but there was significant variability in the GVL response to pressure between cell lines isolated from different donors. During perfusion, GVLs were observed "migrating" and agglomerating about the cell layer and often collapsed despite maintaining the same pressure drop. GVL formation was also observed in human umbilical vein and porcine aortic endothelial cells, suggesting that giant vacuole formation is not a unique property of Schlemm's canal cells. However, in these other cell types, GVLs were rarely observed "migrating" or contracting during perfusion, suggesting that Schlemm's canal endothelial cells may be better adapted to withstand basal-to-apical directed pressure gradients. In conclusion, we have established an in vitro model system to study giant vacuole dynamics, and we have demonstrated that this system reproduces key aspects of giant vacuole morphology and behaviour. This model offers promising opportunities to investigate the role of endothelial cell biomechanics in the regulation of intraocular pressure in normal and glaucomatous eyes.

Figure 1

A transmission electron micrograph showing giant vacuoles (GVs) along the inner wall endothelium of Schlemm’s canal (SC). The classic “signet ring” appearance is particularly well exhibited by the second giant vacuole from the left where the cell appears as a thin, continuous lining around the giant vacuole cavity with the nucleus (n) bulging to one side. The first and third giant vacuoles from the left have basal openings or “meshwork pores” (asterisks) where, presumably, aqueous humour enters the giant vacuole cavity from the underlying juxtacanalicular tissue (JCT). Note that the endothelial cells on the outer wall of Schlemm’s canal, which is not typically involved in aqueous humour filtration, are flat.

Figure 2

Diagram of the perfusion system (A) and arrangement of cells on the filter membrane with respect to the microscope objective (B). The flow rate, Q, of perfusion medium across the cell layer is set by a computer-controlled syringe pump that adjusts Q to maintain a user-defined pressure drop, ΔP. The cells are cultured on the bottom-facing surface of a filter membrane with flow (red arrows in panel A) crossing the cell layer in the basal-to-apical direction. The distance, h, between the cells and cover glass can be adjusted using the membrane insert adapter to position the cells within the working distance of the microscope objective. Panels C and D show representative pressure (solid curve; left axis) and flow (dashed curve; right axis) tracings for perfusions at a constant pressure of 2 and 6 mmHg. Note that the perfusion system reaches steady state within 5 minutes.

Figure 3

Elliptical void regions (arrows) appear within the otherwise uniformly stained cytoplasm of living cells during basal-to-apical directed perfusion of HSCECs (top row), PAECs (middle row), or HUVECs (bottom row). The left column of images shows cells prior to perfusion, while the right column shows the same cells after approximately 20 minutes of perfusion at a pressure drop of 6 mmHg in the basal-to-apical direction. Cells were stained with calcein-AM. Note that HSCECs are intrinsically larger and are therefore displayed at a lower magnification to show approximately the same number of cells per field. The HSCEC cells shown in the top row of images are the same as those shown in the top row of Figure 8.

Figure 4

Three-dimensional confocal reconstructions of a giant vacuole-like structure (GVL) in a HSCEC layer that was perfusion fixed at a pressure drop of 6 mmHg in the basal-to-apical direction (cell line SC52, passage 5). The surface rendering is shown in panel A, with the GVL indicated by an asterisk. A computer-generated vertical cross-section through the GVL (B) reveals a dome-like elliptical cavity with a cellular lining about its entire circumference, including along its base, consistent with the classic “signet ring” appearance attributed to giant vacuoles (cf. Figure 1). A horizontal cross-section through the GVL reveals that the base of the GVL cavity contains a small opening or “meshwork pore” (~3.5 µm diameter) that is presumably the site where perfusion fluid enters the GVL cavity from below. Cells were glutaraldehyde-fixed to induce auto-fluorescence and stained using phalloidin to label F-actin.

Figure 5

Three-dimensional confocal reconstructions of giant vacuole-like structures (GVLs) in HSCEC layers that were perfusion fixed at a pressure drop of 2 mmHg (panels A, B) or 6 mmHg (panels C, D). Surface renderings are shown in panels A and C, with GVLs indicated by asterisks. Vertical cross-sections through the same GVLs (panels B and D) reveal the classic “signet ring” appearance, with a continuous cellular lining about the entire GVL cavity (cf. Figure 1). Note that the insets in panels B and D are presented at the same magnification to demonstrate that GVL size increases with increasing pressure drop, with obvious thinning of the cellular lining with increasing pressure. GVLs were obtained from cell line SC52, passage 4 (panels C, D) and 5 (panels A, B), and the cells were glutaraldehyde-fixed to induce auto-fluorescence and stained with phalloidin to label F-actin.

Figure 6

The size and number of giant vacuole-like structures (GVLs; arrows) increases with increasing pressure drop, but there is a large variability in the pressure response between individual cell lines. Cell line SC52 (71 year-old donor, top row) shows several GVLs at both 2 mmHg (left column) and 6 mmHg (right column) after approximately 20 minutes of perfusion. However, SC58 (34 year-old donor, bottom row) shows very few GVLs at 2 mmHg, and those at 6 mmHg are smaller and less numerous than those observed in SC52. All panels are presented at the same magnification. Images of SC52 were obtained from different experiments.

Figure 7

Representative tracings (panel A, cell line SC58) and aggregate values (panel B, cell lines SC52 and SC58; mean ± SD) of hydraulic conductivity, L_p (µl·min⁻¹·mmHg⁻¹·cm⁻²), within HSCEC layers at a pressure drop of 2 or 6 mmHg. L_p increased almost two-fold between 2 and 6 mmHg (p<0.001; N = 7 or 8). Note that L_p values account for the hydraulic resistance of the filter membrane. Tracings shown in panel A are taken from the same experiments used to produce the pressure and flow tracings shown in Figure 2C and 2D.

Figure 8

Time-lapse images of giant vacuole-like structures (GVLs; arrows) forming during basal-to-apical directed perfusion of HSCECs (SC52; top row), PAECs (middle row) and HUVECs (bottom row) at a pressure drop of 6 mmHg. In all cell types, GVL size tended to increase throughout the perfusion, however in HSCECs, GVLs were often observed “migrating” about the cell layer or collapsing despite maintaining the same pressure drop. In contrast, GVLs in PAECs and HUVECs tended to be stationary without contracting during perfusion. The region circled in the second HSCEC cell image identifies a cluster of smaller GVLs that were highly dynamic and can be seen throughout the cell layer (cf. Supplemental Movies). The time of image acquisition (in minutes) after the start of perfusion is indicated at the lower right corner of each frame. Note that GVL formation occurs despite relatively large discontinuities in the cell layer (e.g., the gap indicated by the asterisk in HUVEC images), which is a beneficial consequence of using track-etch filter membranes, as described in the Discussion. Cells were stained with calcein-AM, and HSCECs are presented at a lower magnification on account of their larger cell size.