Page 38 - Read Online

Page 51 Plössl et al. J Transl Genet Genom 2022;6:46-62 https://dx.doi.org/10.20517/jtgg.2021.39

RESULTS
Recruitment of probands and generation of an iPSC-RPE repository
To investigate the molecular mechanisms contributing to AMD disease, our aim was to generate iPSC-RPE
cell lines from donors with both a very low and a very high genetic risk for the disease. We included 161
patients recruited at the Eye Clinic of the University Hospital Regensburg, Germany. Of these, 77 were
categorized as having early (AREDS grades 1-4; n = 16) or late stage [mostly neovascular (NV); n = 61]
AMD. Seventy-four individuals were free of any signs of early or late AMD pathology and served as
controls. The probands were genotyped for 13 SNPs at 8 chromosomal loci known to be sufficient to
calculate an individual AMD-associated GRS [5,31] [Table 1].

The genetic AMD risk profiles were calculated applying the model reported by Grassmann et al. (2012).
[20]
According to this model, a risk score in Category 1 represents a very low genetic risk for AMD, whereas a
risk score in Category 5 confers a very high genetic AMD risk. Study participants were normally distributed
over the five risk categories, with the majority classifying as Categories 2-4. Only 10% of the participants
were assigned to the extreme GRS Categories 1 and 5 [Figure 1A]. When participants were categorized in
AMD patients and asymptomatic controls a shift in the distribution over the risk categories becomes
evident [Figure 1B].

As we wanted to generate an iPSC-RPE repository that would allow us to perform experiments in cell lines
with maximally different genetic risks for AMD, four probands from Category 1 (LR) who showed no AMD
phenotype and four persons from Category 5, all with late stage NV-AMD (HR) [Table 2], were selected as
donors for the subsequent generation of iPSC-RPE cells. PBMCs or fibroblasts were used to reprogram the
cells to iPSCs and subsequently differentiate them into iPSC-RPE. We did not observe any differences in
iPSC-RPE originating from fibroblasts or PBMCs with regard to their cell characteristics. Information on
the donors of the biomaterials is given in Table 2. Male and female individuals were included in both risk
groups and the mean age of HR donors was 64.8 years (± 1.8 years), whereas the LR donors were slightly
older with a mean age of 80 years (± 8.3 years). GRS of all donors are given in Table 2 traceable via SNP IDs
shown in Table 1. While the HR donors are assigned to GRS Category 5 and the LR donors to GRS
Category 1, there are still some differences in individual genotypes within the HR and LR cell lines. For
example, donor HR3 is homozygous for 11 out of 13 AMD-risk altering alleles, whereas donor LR3 is
homozygous for only three of the AMD-associated SNPs, clearly showing the genetic differences between
the HR and LR cell lines. Specifically, all high-risk individuals were homozygous for CHF p.Y402H and
ARMS2 p.A69S, while the low-risk individuals harbored no risk altering variant at these two loci.

Prior to analysis, iPSC-RPE from HR and LR donors were matured on Transwell inserts for a duration of
six weeks and the discrete TEER was measured at weekly intervals. TEER values increased until week 2 and
then remained constant over the course of time, with a mean value including all cell lines reaching
230 Ω*cm [Figure 2A]. Measurements for individual cell lines are shown in Supplementary Figure 1A. After
2
maturation, RPE-specific proteins BEST1 (bestrophin 1) and RPE65 (retinoid isomerohydrolase 65) were
robustly expressed, as shown by Western blot analysis [Figure 2B]. Immunocytochemistry revealed a
characteristic cobblestone-like RPE morphology and typical BEST1 and ZO-1 staining patterns [Figure 2C
and Supplementary Figure 1B]. iPSC-RPE polarity was further confirmed by directed secretion of VEGF,
which was significantly higher towards the basal Transwell compartment (P < 0.05, Mann-Whitney U-test)
[Figure 2D]. A POS phagocytosis assay was conducted to confirm RPE characteristics on a functional level
[Figure 2E and Supplementary Figure 1C]. All cell lines efficiently internalized porcine POS, which were
subsequently degraded with time, as indicated by a decreasing rhodopsin signal in Western blots after 2 and
4 h of incubation. Quantification of the rhodopsin signals a clearly showed a decrease of the signal
intensities in both HR and LR cell lines, but there was no difference to be observed in POS degradation
efficiency with regard to the AMD GRS [Figure 2D]. Together, these results demonstrate the integrity and