Developmental wave of programmed ganglion cell death in human retinal organoids

Abstract

The delicate and complex structure of the neural retina that enables proper visual function is achieved during embryonic development through a precise balance of proliferation, differentiation, and cell death. Retinal ganglion cells (RGC), the only output neurons of the retina, show a steady increase in numbers during development except for two waves of cell death that are highly conserved in vertebrates. However, the mechanisms responsible for these phenomena and their conservation in the human retina are incompletely understood. In this work we took advantage of human induced pluripotent stem cell (hiPSC)-derived retinal organoids to explore these questions. Using different markers and quantitative techniques in three different hiPSC lines, we found a consistent decrease in RGC numbers at week 8 of differentiation, a developmental stage that is equivalent to that of the first wave of RGC death in other species. This decrease coincided with a peak in caspase 3 activation and TUNEL(+) staining, suggesting an apoptotic mechanism. Notably, this was accompanied by a decrease in the BAX/BCL2 ratio and a lack of caspase 9 activation. However, we observed a marked increase in caspase 8 activation at this stage, suggesting the involvement of the extrinsic apoptotic pathway. Together, these results show for the first time the intrinsic ability of the human retina to regulate RGC numbers through programmed cell death mechanisms, which could lead to new insights regarding congenital retinal abnormalities. Moreover, this work has implications for experimental design in basic and translational research using human stem cell-derived retinal organoid models.

Figure 1.. Retinal ganglion cell quantification reveals a developmental trough in human retinal organoids.

a-e) Representative confocal images of retinal organoid cross sections at 20x taken on the five different differentiation time points. Sections were immunostained for the RGC marker HuC/D (red) and counterstained with DAPI (blue). f) Percentage of HuC/D(+) area/ total retinal area measured by DAPI staining. Bar graph represents mean ± SEM. N=5 biological replicates per time point as plotted. ****p<0.0001. Scale bar: 100 μm.

Figure 2.. Microglia are not present in human retinal organoids.

a-b) Representative confocal images of adult human post-mortem retina at 20x: DAPI (blue), IBA1 immunostaining for microglia (green). White arrows indicate positively stained microglia. c-d) Representative confocal images of retinal organoids at 10x: DAPI (blue), IBA1 immunostaining (green). e) PCR for microglia markers (CD11B and CX3CR1) on three independent human retinal organoid samples and a human retina sample (including RT(−) control) shows the absence of microglia in human retinal organoids. Scale bar: 100 μm.

Figure 3.. Caspase-mediated programmed cell death is highest at the time of RGC loss.

a-e) Representative confocal images of retinal organoid cross sections at 20x taken on the five different collection time points post-differentiation: DAPI (blue), cleaved caspase3 (ClC3; green). f, Percentage of ClC3(+) area/ total area measured by DAPI at each collection time point. g) Percentage of TUNEL(+) area/ total area measured by DAPI at each collection time point. N=5 biological replicates per time point as plotted. Bar graph represents mean ± SEM. ****p<0.0001. Scale bar: 100 μm.

Figure 4.. Analysis of apoptotic pathway components in retinal organoids.

a) Representative Western blot images show expression of cleaved caspase-3 (ClC3), cleaved caspase-8 (ClC8), cleaved caspase-9 (ClC9), BAX, and BCL2, in retinal organoids at weeks 7 and 8 of differentiation. Notice the lack of ClC9 detection. b) Western blot quantification shows a decrease in the BAX/BCL2 ratio at 8 weeks of differentiation- the time of RGC death. c) Western blot quantification also shows an increase in ClC8 at 8 weeks of differentiation. N=3–8 biological replicates per time point, as plotted. Bar graph represents mean ± SEM. *p<0.05; **p<0.01. Scale bar: 100 μm.