Fundus-on-a-Chip

Though organ-on-a-chip experiments are gaining in popularity in drug discovery, they have rarely been used to investigate ophthalmic disease. But Hirokazu Kaji’s team from Tohoku University, Japan, believes the approach holds promise. “My group has been developing microfluidic cell culture systems for 10 years and, in collaboration with ophthalmologists, we decided to develop a model of the fundus,” says Kaji.

To better understand age-related macular degeneration (AMD) pathology, they built a simplified co-culture model of fundus tissue using a microfluidics platform. The model comprises human retinal pigment epithelial cells (ARPE-19s), fibronectin-coated porous membrane – to represent Bruch’s membrane – and human umbilical vein endothelial cells (HUVECs), playing the role of the choroid (1).

The team first characterized monocultures of ARPE-19s and HUVECs, determining that: i) under hypoxic and hypoglycemic conditions, ARPE-19s remained stable, and VEGF secretion from the cells was significantly increased by 77.1 percent (hypoxic) and 68.1 percent (hypoglycemic) compared with control conditions; and ii) exogenously applied VEGF induced a 21.9 percent increase in HUVEC migration compared with control.

The group then co-cultured both cells under hypoxic and hypoglycemic conditions, and found that HUVECs in the lower chamber exhibited directional growth and traveled through the porous membrane towards the VEGF-secreting ARPE-19 cells in the upper chamber. Compared with control conditions, the growth area of ARPE-19 cells significantly decreased in hypoxic and hypoglycemic co-culture conditions over time (p≤0.023); growth area of HUVECs also decreased over time (non-significant).

The upshot? The authors believe the findings indicate that the “invasion” of HUVECs led to the subsequent detachment of ARPE-19 cells – a process that partially recapitulates neovascularization in wet AMD (2).

The team write that they have “taken the first step in elucidating molecular mechanisms of angiogenesis within the device” (1). Kaji comments, “The current results are very preliminary – one of the things that we have to address is the incorporation of functional microvessels into the model so that we can closely mimic choroidal neovascularization.”

Ultimately, the team hope that the model will be useful in drug discovery, and for extending the anatomical and functional understanding of the blood-retina barrier. “We’re also interested in the connection between retinal and neurodegenerative diseases, and we hope our model might help further research into this area,” adds Kaji.