Development of a three-dimensional microphysiological Retina-on-a-Chip system

Abstract

The human retina is a complex neurosensory system that features multiple layers of different retinal neurons. Those neurons are arranged in a unique architecture and function to transmit a signal to the human brain that is interpreted as visual perception. Vision impairment is affecting millions of people worldwide while at the same time, for many disorders, pharmacological treatment options are not available or can only ameliorate the symptoms. To be able to investigate underlying disease mechanisms and to find new pharmacologic treatment options, new retina models are urgently required. Up to now, there are several different retinal model systems available, ranging from animal models to in silico as well as in vitro cell culture models. These systems differ considerably in their advantages and applicability. However, the limitations of each system lead to the consequence that a new and physiological accurate model system is necessary that is able to represent the human retina biology with all of its cell types as precisely as possible. Retinal organoids (ROs) as miniature “retina in a dish” have the potential to serve as new in vitro model system. They feature all retinal layers, can be generated from healthy human cells but also from patient material. Here especially, they can serve as disease model and allow to test potential treatment options. However, standard dish culture of these organoids leads to several limitations since the tissues’ natural environment is not considered. This thesis substantially contributed to the development of a new microfluidic retina-on-a-chip (RoC) system. For this purpose, we combined RO-technology with organ-on-a-chip technology (OoC). OoC technology uses microfluidic devices for cell-culture to simulate an organ-like physiology. We used ROs as well as retinal pigment epithelium (RPE) cells derived from human induced pluripotent stem cells by retinal differentiation to integrate them into a microfluidic chip system. By first establishing individual culture chips for monoculture of RPE or ROs alone, we verified that both tissues are viable and can be cultured in the chip environment. Using immunohistochemistry and qRT-PCR we showed that characteristic markers expression is not affected and using electron microscopy that the typical morphology is preserved. The chips were then combined into a co-culture RoC system, enabling the cultivation of ROs in close contact with RPE cells. We verified that it was possible to bring both tissues into a physiological and close contact by analyzing the distance between RPE and RO inside the chip using live-cell imaging and immunohistochemistry. Further, we found that the setup inside the RoC leads to improved segment formation in the photoreceptors of the ROs. This was shown in a qualitative fashion using immunohistochemistry and also in a quantitative fashion, using electron microscopic comparisons between dish-cultured and chip-cultured ROs. In this context, we also observed a positive impact of the presence of RPE inside the chip regarding photoreceptor segment formation. As another functionality test to show a physiological setup, we analyzed the phagocytotic ability of the RPE cells for digestions of shed photoreceptors segments inside the RoC. Using live-cell imaging, immunohistochemistry and electron microscopy, we were able to confirm phagocytosis inside the RPE layer within the RoC. Lastly, as a proof-of-principle study, we showed that the RoC is suitable as an in vitro drug-testing device for analysis of retinal toxicity. The known retinopathic effect of two different drugs, chloroquine and gentamicin, was verified by analyzing cell death with live-cell imaging of treated RoCs and subsequent quantitative comparison to non-treated RoCs. In the case of chloroquine, also the known lysosomotropic effect was verified using immunohistochemistry. In summary, we have generated a new and physiological microfluidic retina-on-a-chip system that helps to improve RO generation and maturation. This system represents a new retinal model system and is suitable not only for testing of candidate or established drugs regarding retinal toxicity, but it has the outmost potential to serve as a disease model to identify new pharmacological treatment options as well as underlying disease mechanisms.