Recreation of a 3-D Physiological Model of Conventional Outflow Tract in vitro

Abstract

Purpose: To explore the feasibility of engineering a physiologically relevant, 3D human conventional outflow tract utilizing cellular co-culture and differentiation approaches along with microfabrication techniques. This 3D tissue will provide normal and disease-relevant, in vitro model systems for understanding the trabecular outflow physiology and its impact on glaucoma pathology.

Methods: We established an “Artificial HTM” on microfabricated porous scaffolds, which demonstrated an in vivo-like phenotype of HTM cells and responded to pharmacological agents in a similar fashion to that seen ex vivo. In order to better mirror the biological and physiological characteristics of the trabecular outflow tract in vivo, we co-cultured human Schlemm’s canal (HSC) cells or differentiated HSC-like cells with HTM to mimic the Schlemm’s canal inner wall. The effects of pore size, biomolecule coating, and cultivation period for co-culture were investigated. Samples were assayed for expression of HTM and HSC markers. Scanning electron microscopy (SEM) was used to reveal the morphology of the co-cultured tissue. In parallel, the responsiveness of these models to glucocorticoids (PA), rho kinase inhibitor (Y-27632) and TGF-β2 was investigated by perfusion studies. The “outflow facility” of the bioengineered 3D HTM/HSC construct was determined by the ratio of Δ(flow rate)/Δ(pressure).

Results: We successfully established 3D co-cultured HTM/HSC constructs to mimic the conventional outflow tract. Co-cultured samples expressed the HTM marker of αβ-crystalline on one side of the construct while the other side expressed VE-cadherin, fibulin-2 and PECAM-1, markers of HSCs. Both cell layers expressed myocilin; greater expression was observed on the HTM layer. The co-culture decreased outflow facility compared to HTM cell culture alone. Additionally, PA and TGF-β2 treatments increased myocillin and ECM expression/secretion compared to untreated controls. The outflow facility of the co-culture model decreased after 300nM PA treatment, and to a greater extent after TGF-β2 treatment. Exposure to Y-27632 decreased α-smooth muscle actin expression and disrupted F-actin fibers, along with increasing outflow facility.

Conclusions: This work established a bioengineered HTM/HSC model mimicking the conventional outflow tract and demonstrated physiologically-relevant responses to biological agents.