Signalling molecules involved in the maintenance and differentiation of human pluripotent stem cells and retinal pigment epithelial cells

Abstract

To study diseases affecting the retina, we rely on appropriate models to investigate causes of degenerative onset and progression. Human pluripotent stem cells (hPSCs), and in particular patient-specific induced pluripotent stem cells (iPSCs), provide an excellent source of material for generating retinal cells that can be used as disease models to explore disease-causing mutations, genotype-phenotype relationships, therapeutic drug screening and potentially cell replacement therapy.

I developed a rapid, one-step differentiation protocol to differentiate hPSCs to the retinal pigment epithelium (RPE), the outermost monolayer of cells in the retina that are damaged in age-related macular degeneration (AMD). Using a simple growth factor treatment regime, functional RPE could be differentiated from hPSCs in 60 days for use in disease modelling and basic fundamental research. Given the protocol’s simplicity, it was easily adapted to an automated platform, enabling the generation of hundreds of patient-specific RPE lines simultaneously for large-scale disease modelling purposes.

Retinal degeneration may be mediated by the loss off the outer blood retinal barrier (BRB) integrity, caused by injury of the retinal pigment epithelium. Lysophosphatidic acid (LPA) and its producing enzyme autotaxin (ATX) have been implicated in inflammation, angiogenesis and apoptosis. Given that ATX is a signature marker of RPE, I hypothesized that it plays an important role in maintaining the microenvironment of the outer retina, including the BRB. Using the differentiation protocol I developed, hPSCs were differentiated into RPE and were shown to express high levels of ATX and LPA receptors (LPARs) by quantitative real time polymerase chain reaction (qRT-PCR). Measurements of endogenous LPA and ATX were determined using liquid chromatography mass spectrometry (LC-MS) and western blot respectively, and indicate that hPSC-RPE cells secrete low basal amounts of LPA (0.1-0.4 nM), but secrete large amounts of functional ATX at the apical surface, towards the photoreceptor layer. It is possible that the low detectible amounts of LPA were a result of the short half-life of LPA in culture or the potential lack of the lipid substrate for LPA synthesis in RPE monocultures. It is more plausible that in vivo, given the high expression of ATX, the amount of local LPA produced is much higher. Addition of exogenous LPA to hPSC-RPE cultures resulted in increased expression of tight junction proteins ZO-1 and Occludin at the cell junction, which was validated at the functional level by increased transepithelial resistance (TER), indicative of increased barrier function and reduced paracellular permeability. This data suggests that RPE-derived LPA acts in an autocrine manner to maintain the tight junction barrier, one of the key function of RPE in vivo. To assess the potential paracrine actions of RPE-derived ATX and LPA, the mouse 661W photoreceptor line and optic cup-derived CD73 photoreceptors were treated with different doses of LPA. At high doses (>10 µM), cytoskeletal changes consistent with cellular retraction were observed, as indicated by myosin light chain (MLC) and its phosphorylated form (p-MLC). These results indicate the RPE-derived ATX and LPA act in a paracrine manner, and dysregulation of this pathway could play a role in photoreceptor dysfunction.

Endothelial cells that line the outside of the BRB may also potentially contribute to its destruction. RPE-ATX was also secreted basally, although at much lower levels, and therefore it’s highly plausible that the amount of local LPA produced in this microenvironment have an effect on BRB maintenance. Treatment of CD31+ iPSC-endothelial cells with LPA indicated that high doses of LPA caused reduced tubule formation. This indicates that in the retina, LPA may function to suppress angiogenesis. Furthermore, I showed that treatment of mouse-derived CD4+ and CD8+ T cell with LPA resulted in reduced activation, again suggesting RPE-derived LPA may protect against loss of BRB by supressing inflammatory mediators.

In conclusion, this Thesis has highlighted a potential novel role for RPE-derived ATX and LPA in the maintenance of the BRB, and that possible dysregulation of this signalling axis may be involved in retinal disease onset and progression where the BRB is compromised.