Limitations and Promise of Retinal Tissue From Human Pluripotent Stem Cells for Developing Therapies of Blindness

Abstract

The self-formation of retinal tissue from pluripotent stem cells generated a tremendous promise for developing new therapies of retinal degenerative diseases, which previously seemed unattainable. Together with use of induced pluripotent stem cells or/and CRISPR-based recombineering the retinal organoid technology provided an avenue for developing models of human retinal degenerative diseases "in a dish" for studying the pathology, delineating the mechanisms and also establishing a platform for large-scale drug screening. At the same time, retinal organoids, highly resembling developing human fetal retinal tissue, are viewed as source of multipotential retinal progenitors, young photoreceptors and just the whole retinal tissue, which may be transplanted into the subretinal space with a goal of replacing patient's degenerated retina with a new retinal "patch." Both approaches (transplantation and modeling/drug screening) were projected when Yoshiki Sasai demonstrated the feasibility of deriving mammalian retinal tissue from pluripotent stem cells, and generated a lot of excitement. With further work and testing of both approaches in vitro and in vivo, a major implicit limitation has become apparent pretty quickly: the absence of the uniform layer of Retinal Pigment Epithelium (RPE) cells, which is normally present in mammalian retina, surrounds photoreceptor layer and develops and matures first. The RPE layer polarize into apical and basal sides during development and establish microvilli on the apical side, interacting with photoreceptors, nurturing photoreceptor outer segments and participating in the visual cycle by recycling 11-trans retinal (bleached pigment) back to 11-cis retinal. Retinal organoids, however, either do not have RPE layer or carry patches of RPE mostly on one side, thus directly exposing most photoreceptors in the developing organoids to neural medium. Recreation of the critical retinal niche between the apical RPE and photoreceptors, where many retinal disease mechanisms originate, is so far unattainable, imposes clear limitations on both modeling/drug screening and transplantation approaches and is a focus of investigation in many labs. Here we dissect different retinal degenerative diseases and analyze how and where retinal organoid technology can contribute the most to developing therapies even with a current limitation and absence of long and functional outer segments, supported by RPE.

Keywords: assembloids; disease modeling; drug screening; photoreceptors; pluripotent stem cells; retinal degeneration; retinal organoids; retinal pigment epithelium.

FIGURE 1

Comparing retinogenesis between human pluripotent stem cell derived retinal organoid growing in a dish and in human fetal retina. Both retinal organoids and fetal retina develop neural retina (NR), but only fetal retina (not organoids) develop continuous layer of RPE surrounding NR (Stage1). Both fetal retina and retinal organoids undergo lamination (Stage 2), and develop outer and inner neuroblast layers (ONBL, where progenitors, including photoreceptor progenitor and young photoreceptors localize, and INBL, where RGCs and 2^nd order neurons migrate). However, as tissue maturation proceeds in retinal organoids, the RGCs gradually die (no connectivity) and then a layer of 2^nd order neurons becomes progressively thinner. And, while cilia and inner segments in photoreceptors develop, outer segments never elongate full-length, compared to that in mature mammalian retina (Stage 3). In fetal retina (2^nd−>3^rd trimester) photoreceptor layer undergoes maturation, cilia and inner segments are formed and elongation of outer segments takes place (which continues after birth). Organoids at stage3 preserve the features of immature human retina (IPL-ONL-IS-cc region, shown). This diagram is designed based on the following data: https://embryology.med.unsw.edu.au/embryology/index.php/Vision_-_Retina_Development, https://embryology.med.unsw.edu.au/embryology/index.php/Carnegie_Stage_Comparison, https:// embryology.med.unsw.edu.au/embryology/index.php/Carnegie_stage_table (O’Brien et al., 2004; Hendrickson et al., 2008; Jukic et al., 2013; Hendrickson, 2016; Hoshino et al., 2017).

FIGURE 2

Presence of LGR5 in retinal organoids. Immunostaining of human pluripotent stem cell derived retinal organoid with anti-LGR5 antibody.

FIGURE 3

Localization of SFRP1 in human fetal retina. Immunostaining of human fetal retina (wk10) with anti-SFRP1 and anti-BRN3B antibody shows SFRP1 presence in the apical side of ONBL and distal side of INBL. Anti BRN3B colocalized with SFRP1 in the INBL.

FIGURE 4

Summary of generating three-dimensional retinal tissue from human pluripotent stem cells and its application in disease modeling, drug screening and cell therapy or tissue replacement.

FIGURE 5

Anatomy of human eye and retinal circuits. (A) Schematic drawing of a cross-section through human eye. Light enters the eye through the cornea, passes through the pupil, lens and strikes the retina. Retina is the light-sensitive tissue lining the inner surface of the eye. Visual information from retina transmits to the brain through optic nerve fiber. In the middle of the retina small depression is called the fovea and is responsible for high resolution vision. Region surrounding the fovea is called as macula and are rich in only cones. Retinal pigment epithelium (RPE) is a pigmented layer and separates the choroidal blood supply from the photoreceptors. Choroid is a vascular layer of the eye. The sclera is a tough white sheath around the outside of the eyeball. (B) Schematic diagram of normal retina circuits. Mammalian retina consists of six major types of neuronal cells – rod (RC) and cone (CC) photoreceptors also horizontal (HC), bipolar (BC), amacrine (AC) and retinal ganglion cell (RGC). The Muller cell are the glial cells that span across the retina and their somata. RPE provides metabolic and transport functions essential for homeostasis of the neural retina. Bruch’s membrane (BM) is a highly specialized and multi-laminar structure in our retinas that forms the basis for mediating interactions between the retinal pigment epithelium and blood flow from the choroid. Choroidal capillaries (CC) are the blood capillaries present in choroid that supply oxygen and nourishment to the outer layer of the retina. Retinal blood vessels are present in OPL, IPL and RGC layers.

FIGURE 6

Localization of PMEL17 in the retinal organoid. Immunostaining the retinal organoid (day 70) with pigmented RPE marker PMEL17 show patches of retinal organoids were pigmented. HNu stains the human nuclei. The insets in panel a are high magnification of area marked with asterisk (*). DAPI counter stains nuclei.

FIGURE 7

Schematic diagram showing important functions of RPE and its interaction with rod and cone photoreceptor outer segments. The RPE microvili interacts with photoreceptor OS and RPE cells are involved in visual cycle, phagocytosis of outer segments disc, nutrient uptake and paracrine secretion of PEDF, VEGF.

FIGURE 8

Major challenges of human pluripotent stem cell derived retinal organoid approach growing in dish. In retinal organoid there islack of RPE interaction with neural retina and lack of RPE interaction with choroid. Also, there is lack of retina brain connectivity. However, early retinal development can be studied using retinal organoid (cell fate acquisition). Once the advanced systems of co-culturing will be established (retinal-brain organoids =assembloids, photoreceptor sheets in organoids -RPE sheets, and vascularized organoids) one will be able to design better models of human retina for both drug screening and therapeutic applications.