TYROSINASE-Deficient Human Retinal Pigment Epithelium Exhibits Melanosome Maturation Defects

Abstract

Purpose

Oculocutaneous albinism type 1A (OCA1A) is a rare recessive genetic condition caused by mutations in TYROSINASE (TYR) that results in pigmentation defects of the skin, hair and eyes. This study was performed to understand melanosome biogenesis and maturation defects in an OCA1A in vitro model using retinal pigment epithelium (RPE) derived from TYR knockout human induced pluripotent stem cells (iPSC).

Methods

CRISPR-Cas9 was used to knockout the TYR gene in iPSC to generate an isogenic pair. A developmentally guided protocol was used to differentiate the isogenic iPSC pair towards RPE monolayer tissue. Monolayer organization, melanosome formation and maturation were studied using electron microscopy. Loss of TYR protein was studied using Western blot and immuno-fluorescence staining. RPE cellular morphology and junction integrity was studied using immunofluorescence staining and transepithelial resistance measurements.

Result

An isogenic pair comprising of untargeted control and TYR knockout iPSC were successfully differentiated towards RPE monolayer tissue with polygonal cell morphology. TYR knockout RPE exhibited significantly reduced TYR protein, increased presence of immature pre-melanosomes and a complete lack of mature melanosomes. We observed abnormal junctional localization of β-catenin staining pattern, as has been reported previously for albino mouse RPE– and OCA1A patient–derived RPE.

Conclusions

Differentiation of TYR-deficient iPSC toward RPE displayed pigmentation defects and absence of mature melanosomes, whereas melanosome biogenesis was not affected, because pre-melanosomes were still observed. These observations were also similar to what was observed in OCA1A patient–derived RPE monolayer tissue, independently confirming the validity of these previous findings.

Oculocutaneous albinism type 1A (OCA1A) is an autosomal-recessive condition characterized by reduced or absent pigmentation of the hair, skin, and eyes.^1,2 Affected individuals with OCA1A can have varying degrees of functional visual deficits (e.g., decreased best-corrected visual acuity) where factors such as foveal hypoplasia, nystagmus, refractive errors, photoaversion, and abnormal decussation of retinal ganglion cell axons at the optic chiasm likely play a role.^3–7 The Genetic and Rare Diseases Information Center lists OCA1A as a rare genetic disorder, with an estimated prevalence of 1/40,000 worldwide.⁸ Caused by mutations in the TYR gene, it is the most severe form of albinism in North America.¹ The tyrosinase enzyme catalyzes the rate limiting step of the melanin biosynthesis pathway and loss of function variants results in partial or complete loss of melanin.^8,9 Melanin is a biopolymer packaged inside melanosomes, a characteristic organelle of pigmented cell types, including the retinal pigment epithelium (RPE) present in the eye.

The RPE, which plays an important role in eye development, is located adjacent to the light sensitive neural retina and is imperative for the proper function of photoreceptors.¹⁰ In all forms of OCA (OCA1-8), including OCA1A, the pigmentation of the RPE is expected to be reduced or completely absent.^2,3 Another consistent feature of all forms of OCA is abnormal fovea development and increased contralateral targeting of optic nerve fibers, the origins of which are currently not completely understood.^11,12 Prior studies have associated reduced levels of zeaxanthin and lutein pigment to age related macular degeneration (AMD),¹³ but the role of melanin pigment is currently unclear. Interestingly, neither OCA1A nor other forms of OCA are associated with any form of retinal or RPE degeneration. There are very few studies on melanosome defects and their biogenesis in the RPE of humans. Studies on cadaver eye tissues from OCA1A patients are rare^14–16 and not practical for mechanistic and large-scale drug discovery studies. We recently developed an in vitro RPE-based model for OCA1A and OCA2 form of albinism using patient-derived iPSC.¹⁷ The present study was undertaken to develop isogenic pairs of untargeted and TYR gene knockout iPSCs, to investigate abnormalities in the biogenesis of melanosomes due to lack of tyrosinase enzyme activity in human RPE and generate a reliable source of cells for drug discovery and gene therapy studies.

Material and Methods

All procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee of National Eye Institute (NEI-605).

Derivation, Maintenance, and Characterization of Human iPSC

All human iPSC–related work was approved by institutional review board protocol no. 11-E1-0245 (NCT01432847). Control iPSC line has been described previously¹⁸ and was reprogrammed using Sendai virus-mediated delivery (no. A16518, CytoTune2; Thermo Fisher Scientific, Waltham, MA, USA) of the four Yamanaka factors (c-MYC, KLF4, OCT4, and SOX2), following the manufacturer's recommendations. The iPSCs were cultured in Essential 8 medium (no. A1517001; Thermo Fisher Scientific) for expansion until the start of differentiation. The iPSC lines were characterized for pluripotency by immunofluorescence staining using antibodies for OCT4, SOX2, NANOG, SSEA-4, TRA-1-81 and TRA-1-60 (Supplementary Table S1). Directed differentiation towards three germ layers—ectoderm, mesoderm, and endoderm—was performed using Human Pluripotent Stem Cell Functional Identification Kit (no. SC027B; R&D Systems, Minneapolis, MN, USA) following the manufacturer's recommendations. Antibodies against OTX2, TUJ1, SOX17, AFP, BRACHYURY, and SMA (Supplementary Table S1) were provided with the kit and used for characterization of cells of all three germ layers. In vivo teratoma assay was performed by injecting iPSCs in immunocompromised mice, as described earlier.¹⁹ Karyotyping and DNA fingerprint STR profiling (data not shown) was performed by Cell Line Genetics (Madison, WI, USA).

Immunostaining of iPSC Colonies and RPE Monolayers

The iPSCs culturing and differentiation of iPSCs to three germ layers was performed in four-well chamber slides for immunofluorescence staining. The iPSC-derived RPE (from now onward referred to as iRPE) monolayers on trans-well inserts cultured for six to eight weeks were used for immunofluorescence staining. Cells were washed three times with PBS before fixing in 4% paraformaldehyde for 20 minutes at room temperature. After three additional washes with PBS Tween 20 (PBST) (0.5% Tween 20 in PBS), cells were permeabilized with immunocytochemistry blocking buffer (PBST, 0.5% BSA, 0.5% Tween 20, 0.05% sodium azide, 0.1% Triton X-100) for one hour. Cells were incubated with primary antibodies overnight at 4°C (Supplementary Table S1). Samples were incubated in the dark for one hour on a rocker. After three PBST washes, samples were mounted on a glass slide with Fluoromount-G aqueous mounting medium (catalog no. 0100-01; SouthernBiotech, Birmingham, AL, USA).

Generation of Tyrosinase (TYR) Knockout iPSCs

A 20 nt guide RNA (gRNA) sequence (5’-AGAAGGAATGCTGTCCACCG-3’) targeting the first coding exon (Exon1) of TYR was identified by an in silico program²⁰ and cloned into pCAG-eCas9-GFP-U6-gRNA plasmid vector (Addgene no.79145) to express gRNA, a high-fidelity version of SpCas9 (eCas9), and a green fluorescent protein (GFP) marker. In the morning of transfection day, 600,000 wild-type Control iPSCs were plated in one well of a Matrigel (354277; Corning Inc., Corning, NY, USA)–coated six-well plate containing 2.5 mL E8 medium (Thermo Fisher Scientific) and 1× RevitaCell supplement (Thermo Fisher Scientific). After the cells attached in the afternoon, transfection was accomplished by using 4 µL Lipofectamine Stem Reagent (Thermo Fisher Scientific) in 200 µL OptiMEM (Thermo Fisher Scientific) with 1.8 µg pCAG-eCas9-GFP-U6-gRNA (TYR) plasmid and following manufacturer's protocol. At 48 hours after transfection, GFP-positive iPSCs were sorted onto Matrigel-coated 96-well plates containing 100 µL E8 and 1× CloneR supplement (05889; Stemcell Technologies, Vancouver, BC, Canada) at one cell per well. Wells with surviving clones after 10 to 12 days were expanded to isolate genomic DNA for screening. A genomic fragment spanning the gRNA target sites was amplified using primers 5’- GTTTGATGCTGGAGGTGGGA -3’ and 5’-TGGTCCCCAAAAGCCAAACT-3’ with Hot Start Taq 2X Master Mix (M0496; New England Biolabs, Ipswich, MA, USA) and Sanger sequenced (Eurofins Genomics, Ebersberg, Germany) to identify TYR bi-allelic knockout clones. The TIDE online program²¹ was used to analyze Sanger sequencing results for bi-allelic knockout clones.

Differentiation of iPSCs to RPE

The iPSCs were differentiated into RPE (iRPE) using a recently published protocol.²² Briefly, iPSC colonies were seeded on vitronectin (no. A1700; Thermo Fisher Scientific) coated surface at a particular cell density in E8 media with a ROCK Inhibitor (Y-27632, cat. no. 1254; Tocris, Bristol, UK) in six-well plates. After two days, cells were reached the confluency of monolayer and E8 media was switched to the differentiation media (DMEM/F12 [cat. no. 11330032; Thermo Fisher Scientific], N2 supplement [cat. no. A1370701; Thermo Fisher Scientific], B27 [cat. no. 17504044; Thermo Fisher Scientific], KSR [cat. no. 12618013; Thermo Fisher Scientific], 20 ng/mL NOGGIN [cat. no. 6057; R&D Systems, Minneapolis, MN, USA], 5 µM CK1-7 Dihydrochloride [cat. no. C0742; Sigma-Aldrich Corp., St. Louis, MO, USA], 5 µM SB 431542 hydrate [cat. no. S4317; Sigma-Aldrich Corp.], and 5 ng/mL IGF-1 [cat. no. AFL291; R&D Systems], 5 µM PD0325901 [cat. no. PZ0612; Sigma-Aldrich Corp.], 10 mM Nicotinamide [cat. no. N0636; Sigma-Aldrich Corp.], 150 ng/ml ACTIVIN A [cat. no. 338-AC/CF; R&D Systems]). Differentiating iPSC committed to RPE fate were reseeded on new vitronectin coated surface and maintained in RPE maintenance media (MEM + glutamax [cat. no. 32561037; Thermo Fisher Scientific], 5% FBS [cat. no. SH30071.03; Hyclone Laboratories, Logan, UT, USA], taurine [cat. no. T-0625; Sigma-Aldrich Corp.], thyronine [cat. no. T-5516; Sigma-Aldrich Corp.], and hydrocortisone [cat. no. H-0396-10; Sigma-Aldrich Corp.]) for 15 days. The iRPE cells were enriched using negative selection with anti-CD24 (no. 655154; BD Bioscience, Franklin Lakes, NJ, USA) and CD-56 (no. 340723; BD Bioscience) antibodies and were seeded onto vitronectin coated trans-wells (no. 3460; Corning, Inc.) and cultured for six weeks before any assays or experiments.

The transepithelial resistance (TER) measurements of iRPE monolayers were carried out using the STX2 electrode set with the EVOM2 meter (World Precision Instruments, Sarasota, FL, USA). The unit area resistance was calculated by multiplying the measured resistance by the effective membrane area (1.12 cm²). Student's t-test was performed to determine the level significance.

Scanning and Transmission Electron Microscopy

For scanning electron microscopy (SEM), confluent iRPE cell monolayers on trans-wells were fixed in EM fixative (2.5% glutaraldehyde Grade 1 [Sigma-Aldrich Corp.], 4% paraformaldehyde EM grade [Electron Microscopy Sciences, Hatfield, PA, USA], and 10 mM CaCl₂ [Quality Biological Inc., Gaithersburg, MD, USA]) in HEPES buffer overnight and prepared as previously described.²³ Samples were mounted on conductive carbon adhesive stubs and imaged using either standard electron microscopy (S4800 Hitachi electron microscope; Hitachi, Tokyo, Japan) or via Helium ion microscopy HIM, carried out on an Orion helium ion microscope (Carl Zeiss Microscopy, White Plains, NY, USA).

For transmission electron microscopy (TEM), confluent monolayers of iRPE cells grown on trans-wells were rinsed with PBS (3×, five minutes each). Cells/tissues were fixed in 2.5% buffered glutaraldehyde and then processed for TEM according to Ogilvy et al.²³ Briefly, specimens were washed in PBS ×3 10 minutes each, post-fixed in 0.5% osmium tetroxide (OsO₄) for one hour, rinsed with PBS ×3, dehydrated in alcohol then propylene oxide, and then embedded in epoxy resin. Blocks were sectioned at ∼90 nm thickness on a Leica EM UC6 ultramicrotome (Leica, Wetzlar, Germany), double-stained with uranyl acetate and lead citrate, and imaged with JEOL JEM-1010 electron microscope (JEOL, Tokyo, Japan). Images were acquired at 15k and 20k magnification to accommodate an RPE cell, and melanosomes observed in the images (CTRL: N = 6, TYR^−/−: N = 5) were counted manually.

Confocal and Brightfield Microscopy

For routine observation of iPSC and iRPE cell cultures and bright field imaging, Zeiss Vert.A1 (Carl Zeiss, Inc., White Plains, NY, USA) inverted microscope was used. Confocal microscopy was performed using Zeiss LSM 880 and 700 microscopes.

Western Blotting

For Western blotting, cells were lysed with RIPA lysis buffer (Sigma-Aldrich Corp.) containing protease inhibitor cocktail and Halt phosphatase inhibitor cocktail (Pierce Biotechnology, Rockford, IL, USA) and spun in a centrifuge at 14,000g, and supernatant was collected. Total protein concentration was determined by BCA protein assay kit (Pierce Biotechnology). Samples (40 µg/each well) were SDS-PAGE electrophoresed using Any-kD pre-casted gels (Bio-Rad Life Science, Hercules, CA, USA), blotted onto polyvinylidene fluoride membranes (Bio-Rad Life Science) and immunoreacted with antibody to TYR and β-ACTIN for normalization. For TYR Western blot, RPE monolayers were treated with bovine photoreceptor outer segments, prior to protein isolation. The membranes were then incubated with secondary antibodies (goat anti-mouse IRDye 800CW [LI-COR Biosciences, Lincoln, NE, USA] or donkey anti-rabbit IRDye 680RD, 1:20,000 [LI-COR Biosciences]) for 45 minutes at room temperature followed by three PBST washings and scanned using the Odyssey infrared scanner and analyzed using Image Studio Lite Ver.4.0 (LI-COR Biosciences).

Results

TYR^−/−-iRPE Exhibit Lack of Pigmentation In Vitro

Control (CTRL) iPSCs line were derived from an unaffected individual and have been reported previously.¹⁸ CRISPR-Cas9 system was used to generate bi-allelic insertions/deletions in exon 1 of TYR gene and six clones demonstrating bi-allelic knockout by Sanger sequencing (Fig. 1A) were expanded and cryopreserved before characterization for pluripotency. We chose a clone which has 1bp insertion and 8 bp deletion on the two alleles, respectively. The TYR knockout iPSC (from now on referred to as TYR^−/−-iPSC) described in this study exhibited normal human karyotype (Fig. 1B). Immuno-stained colonies from TYR^−/−-iPSC were positive for the pluripotency associated transcription factors OCT-4, NANOG and SOX-2, and the cell surface markers TRA-1-81, TRA-1-60, and SSEA-4 (Fig. 1C). Pluripotency was further confirmed using in vitro directed differentiation (Fig. 1D) and in vivo teratoma formation assays (Fig. 1F). TYR^−/−-iPSC lines differentiated in vitro expressed markers for ectodermal (OTX-2 and β-III tubulin), mesodermal (Brachyury and smooth muscle actin) and endodermal (SOX-17 and alpha fetoprotein) lineages (Fig. 1D, left panel). Upon injection into immuno-compromised mice, TYR^−/−-iPSC line generated teratoma in vivo, exhibiting derivatives of the three germ layers (Fig. 1F, right panel).

Figure 1.

(A) Wild type TYR sequence with highlighted CRISPR target and PAM site. Sanger sequencing chromatograms of six independent iPSC clones with insertions/deletions in TYR, depicted in red font. (B) Karyotype of TYR^−/−-iPSCs clone derived after introduction of indels. (C) Confocal microscopy microphotographs of pluripotency markers (NANOG, TRA-1-81, OCT-4, SSEA-4, SOX-2 and TRA-1-60) immunofluorescence on iPSC colonies. Scale bar, 100 µm. (D, left panel), Confocal microscopy microphotographs of ectodermal (OTX-2 and TUJ-1), mesodermal (Brachyury/TBXT and Smooth Muscle Actin/SMA) and endodermal (SOX-17 and α Feto Protein/AFP) markers immunofluorescence on iPSCs post directed differentiation. Scale bar, 100 µm. (E, right panel), Brightfield microscopy microphotographs of H&E-stained teratoma sections showing derivatives of the three germ layers. Scale bar, 500 µm.

The isogenic iPSC pair was then differentiated towards RPE using a previously published protocol.²² After differentiation, iRPE cells were cultured in a trans-well culture system (Fig. 2A) for six to eight weeks to help the formation of a polarized monolayered epithelial tissue.²⁴ Visual examination and brightfield microscopy revealed the presence of pigmentation in the CTRL-iRPE monolayer, whereas pigmentation in TYR^−/−-iRPE was absent (Figs. 2A, 2B). At least two different clones from TYR^−/−-iPSC line were differentiated to iRPE and the pigmentation defects were consistent across both clones (Supplementary Fig. S1). Long term culturing (>6 months) did not result in pigment accumulation suggesting complete lack of melanin accumulation in the TYR^−/−-iRPE monolayers (Supplementary Fig. S2). This observation is consistent with OCA1A patient derived RPE¹⁷ and ophthalmic findings where no age-associated increase in pigmentation has been observed in the eyes.²⁵

Figure 2.

(A) CTRL and TYR^−/−-iRPE monolayers on 2-D trans-well culture system showing different degrees of pigmentation. (B) Representative brightfield microscopy images of CTRL- and TYR^-/--iRPE monolayers displaying variable amounts of melanin. Scale bar: 500 µm. (C) Transmission electron microscopy (TEM) microphotographs of CTRL-, TYR^−/−-iRPE monolayers on trans-well membranes. At least two independent clones were characterized. Arrows indicate location of apical processes, arrowheads indicate location of basal infoldings, and position of nuclei are denoted by N. Scale bar: 2 µm.

Transmission electron microscopy of CTRL and TYR^−/−-iRPE displayed normal RPE morphology consisting of a monolayer of cells atop the trans-well membrane (Fig. 2C), apical microvilli processes (Fig. 2C, arrow), basal infoldings (Fig. 2C, arrowhead) and a basally located nuclei (Fig. 2C, N). The CTRL-iRPE cells were packed with melanosomes, whereas the TYR^−/−-iRPE cells lacked mature melanosomes. No other remarkable differences were observed in the CTRL- and TYR^−/−-iRPE cells at the ultrastructure level (Fig. 2C).

TYR^−/−-iRPE Lacks Mature Melanosome

Studies have shown that melanosomes progress through four developmentally categorized stages (I-IV).²⁶ Stage I pre-melanosomes resemble late endosomes and contain internal membranous vesicles and irregular fibrous structures. Stage II pre-melanosomes, can be identified by the presence of intralumenal fibers that form regular parallel or concentric arrays which serve as substratum for melanin deposition (stage III), resulting in their darkening. Stage IV mature melanosomes are densely packed with melanin, appearing homogeneously black.^26,27 We performed extensive ultrastructure studies using TEM (N = 6 images per group) to further investigate the defects in melanosome biogenesis in TYR^−/−-iRPE monolayers. Melanosomes appeared densely packed in CTRL-iRPE cells (64.3 ± 20.5/cell) and were rarely observed in TYR^−/−-iRPE (10.0 ± 3.0/cell; Figs. 3A, 3B). Graph 3B denotes the total number of melanosome (pre+mature melanosomes) like structures observed approximately per section in the TEM images. In CTRL-iRPE cells, we observed heavily pigmented mature (stage IV) melanosomes (59.3 ± 18.6/cell,) that were circular and elliptical in shape (arrowheads in Figs. 3A and 3B and solid black circles in Fig. 3C), whereas pre-melanosomes were rarely observed (5.0 ± 2.5/cell, arrows in Fig. 3A and open circles in Fig. 3C). In contrast, in TYR^−/−-iRPE, we observed only pre-melanosomes (10.0 ± 3.0/cell, stage I and II, Fig. 3A, arrows) and an absence of mature melanosomes (stage III and IV). The percentage of mature melanosomes (Fig. 3C, solid circles) was significantly higher than the pre-melanosomes (Fig. 3C, open circles) in the CTRL-iRPE, whereas in TYR^−/−-iRPE mature melanosomes were absent (Fig. 3C).

Figure 3.

(A) Higher magnification TEM microphotographs of control CTRL- and TYR^−/−-iRPE monolayers on trans-well membranes used for quantification, arrows point at pre-melanosomes and arrowheads at mature melanosomes. (B) Quantification of the number of melanosomes/cell in CTRL- (N = 6) and TYR^-/--iRPE (N = 5). (C) Quantification of number of pre-melanosome (open circles) and mature-melanosome (solid circles) per cell in CTRL- and TYR^−/−-iRPE. *P ≤ 0.05.

To confirm the loss of tyrosinase protein we performed Western blot and observed significant reduction in tyrosinase protein band intensity (Fig. 4A). We then performed immunofluorescence staining for tyrosinase which is known to be present on maturing and mature stage III and IV melanosomes^27,28 and observed a punctate staining pattern in CTRL- but not in TYR^−/−-iRPE -, which lacks functional tyrosinase protein. Some tyrosinase staining was observed in the TYR^−/−-iRPE but it was not in the form of large and oval puncta as observed in the CTRL iRPE (Fig. 4B). Higher magnification TEM imaging of CTRL-iRPE revealed spherical and characteristic oval shaped melanosomes that were densely packed with melanin whereas pigmented melanosomes were never observed in TYR^−/−-iRPE cells (Fig. 4C). Next, we immunostained the isogenic iRPE monolayers with TYRP2, which is known to localize on mature melanosomes^27,28 and observed both a bright punctate fluorescent signal on melanosomes and a diffuse cytoplasmic signal in CTRL-iRPE. In TYR^−/−-iRPE we did not observe the punctate signal, but the diffuse cytoplasmic staining was relatively preserved (Fig. 4D; Supplementary Fig. S3). As expected, loss of tyrosinase in TYR^−/−-iRPE cells results in complete loss of mature melanosomes, which was also observed in OCA1A patient–derived iRPE.

Figure 4.

(A) Western blot of TYR protein in iRPE monolayers. Bar graphs depicting relative protein levels after normalization with β-Actin and averaging three samples. *P ≤ 0.05. (B) Confocal microscopy microphotographs of TYR immunofluorescence staining in iRPE monolayers. Scale bar: 5 µm. (C) TEM microphotograph of mature-melanosomes observed in CTRL-iRPE and pre-melanosomes in TYR^−/−-iRPE. Scale bar: 500 nm. (D) Confocal microscopy microphotographs of TYRP2 and ZO-1 immunofluorescence staining in iRPE monolayers. Scale bar: 10 µm.

TYR^−/−-iRPE Exhibits Normal Melanosome Biogenesis

Higher magnification TEM imaging of control (CTRL) and TYR^−/−-iRPE pre-melanosomes revealed the presence of intralumenal fibrillar striations (Fig. 5A; black arrows), which are thought to include the structural protein PMEL and serve as the matrix for melanin deposition.^26–28 Some stage II melanosomes had an amorphous core surrounded by concentrically organized intralumenal fibrillar striations but no evidence of melanin deposition inside the melanosomes (Supplementary Fig. S4A, middle panel, arrowhead). In TYR^−/−-iRPE we routinely observed abnormal pre-melanosomes where organization of the intralumenal fibrillar striations were disorganized and the amorphous region was not centrally located (Supplementary Fig. S4, right panel, arrowhead), and these abnormal melanosomes were never observed in CTRL iPRE cells. We also observed what appear to be degenerating melanosomes (Fig. 5A; asterisk) in TYR^−/−-iRPE as were reported in patient derived OCA1A iRPE cells. To further confirm the presence of pre-melanosomes in TYR^−/−-iRPE cells, we studied the pMEL17 localization pattern. Immunofluorescence staining with PMEL-17 antibody displayed punctate staining pattern in both isogenic RPE but was much denser in TYR^−/−-iRPE compared to CTRL-iRPE (Figs. 5B, 5C). This is in accordance with the presence of a higher number of pre-melanosomes in OCA-iRPE compared to CTRL-iRPE by morphologic criteria (Fig. 3C) or the lack of fluorescence quenching from melanin. These observations suggest that early melanosome biogenesis is not significantly affected, and pre-melanosomes are generated in TYR^−/−-iRPE but they do not mature to reach stages III and IV (Supplementary Fig. S4B). In TYR^−/− iRPE, the observed significant increase in pMEL17-positive puncta likely reflects an accumulation of stage II melanosomes that fail to mature into stage III and IV.

Figure 5.

(A) TEM microphotograph of different stages of melanosome biogenesis observed in CTRL and TYR^−/−-iRPE monolayer. Arrow indicates intralumenal fibers of a pre-melanosome. Scale bar: 500 nm. (B) Confocal microscopy microphotographs of PMEL-17 immunofluorescence staining in iRPE monolayers. Scale bar: 2 µm. (C) Graph depicting quantification of pMEL17 puncta staining observed in Figure 5B.

TYR^−/−-iRPE Cells Exhibits Abnormal Apical Junction Organization

The characteristic features of RPE monolayers in vivo and in vitro include polygonal cell packing and apical-basal polarization.²⁴ We performed immunostaining for ZO-1 and β-catenin that are located on cell junctions and marks the cell/cell border.²⁹ Both iRPE lines in the isogenic pair exhibited characteristic polygonal cell packing (Fig. 6A). It has been suggested that albino mouse RPE cells might exhibit adhesion defects, which was based on observations of junctional protein P-cadherin immunolocalization pattern.³⁰ We analyzed the apical tight junction properties of the isogenic iRPE pair by immunofluorescence staining and TER measurements. ZO-1 and β-catenin are junctions protein localized on cell borders for RPE.²⁹ We observed ZO-1 localization predominantly at the cell borders only in both CTRL- and TYR^−/−-iRPE (Fig. 6A). In the CTRL-iRPE we observed co-localization of β-catenin with ZO-1 at the RPE apical cell borders, whereas TYR^−/−-iRPE, not only displayed β-catenin staining localized with ZO-1 but staining was also observed in multiple z-planes (Fig. 6A and S5, white and green arrow), similar to the observations of Iwai‐Takekoshi et al.³⁰ for P-cadherin. We observed no differences in the staining pattern of ZO-1 protein between CTRL- and TYR^-/--iRPE (Fig. 6A).

Figure 6.

(A) Confocal microscopy microphotographs of ZO-1 and β-catenin immunofluorescence on control (CTRL)- and TYR^−/−-iRPE monolayers (Left panel). Scale bar: 10 µm. Higher magnification of ZO-1 and β-catenin immunofluorescence staining on CTRL- and TYR^−/−-iRPE monolayers shown in panels 2–4. Scale bar: 5 µm. (B) Trans-epithelial resistance of iRPE monolayers cultured on trans-well membranes. N = 10 for each group. (C) Confocal microphotographs of ezrin and collagen-IV immunofluorescence in orthogonal view (top panels) in CTRL- and TYR^−/−-iRPE. Scale bar: 10 µm. (D) SEM (top view) microphotographs of apical processes present on the apical side of iRPE cells. Scale bar: 2 µm.

To check whether junctional integrity was preserved in TYR^−/−-iRPE, we measured TER of the RPE monolayers. The TER is a characteristic feature of RPE and other epithelial monolayers and is an indirect measure of junctional integrity of the cells within the monolayer.²⁴ We detected no differences between the TER of CTRL- and TYR^−/−-iRPE, suggesting that the overall junctional integrity of OCA1A-iRPE was not compromised in our in vitro cultures (Fig. 6B).

TYR^−/−-iRPE Monolayer Exhibits Proper Apical Basal Organization

The RPE monolayer displays apical-basal polarity and abuts the photoreceptors on the apical side and the choroid on the basal side, and attainment of polarity is necessary for the RPE monolayer to perform its function in the eye.³¹ The polarized organization of CTRL- and TYR^−/−-iRPE monolayers was analyzed using immunofluorescence staining for ezrin and collagen-IV which are present on the apical and basal side, respectively. Ezrin staining was observed specifically on the apical side (Fig. 6C, red fluorescence) and collagen-IV on the basal side (Fig. 6C, green fluorescence) of CTRL- and TYR^−/−-iRPE monolayers. A dense cover of apical processes, which serve to increase the surface area interacting with photoreceptor outer segments were also observed using SEM in CTRL- and TYR^−/−-iRPE suggesting proper morphogenesis of the apical surface of all iRPE monolayers (Fig. 6D). These data suggest that apical-basal polarization in TYR^−/−-iRPE is successfully established.

Discussion

Here we show for the first time, derivation of TYR knockout iPSC lines and their successful differentiation towards RPE monolayer tissue. The TYR^−/−-RPE monolayer exhibited complete lack of pigmentation, a characteristic feature observed in OCA1A patients’ RPE.¹⁷ OCA1A is due to bi-allelic variants in TYR and is the most severe form of OCA in North America, with no treatments currently available. The iPSC lines generated provide a renewable source of human tissue that until now was derived from cadaver eyes. The iPSC lines can also be differentiated not just towards RPE but toward many different cell types including melanocytes³² for drug discovery experiments and organoids to understand the molecular and cellular basis of developmental vision defects associated with OCA1A. Our in vitro disease model is useful in understanding pigmentation defects at a cellular and organelle level. Using the TYR^−/−-iRPE monolayer tissue, we have shown that the early-stage pre-melanosomes (stage-I and stage-II) are formed in the absence of tyrosinase enzyme but do not subsequently mature, suggesting that it is dispensable for early melanosome formation. We also show that TYR^−/−-iRPE cells completely lack stage-III and stage-IV melanosomes and any evidence of melanin deposition. Similar observations were made in the Tyrosinase mutant mouse (c^2-j) model³⁰ and iRPE derived from OCA1A patients,¹⁷ both of which lacked mature stage III and IV melanosomes. In the previous study,¹⁷ patient-derived OCA1A iRPE lines were compared to unrelated control iRPE, leaving open the possibility that genetic variation could have contributed to some of the observed phenotypes. By generating the isogenic TYR^−/− iRPE lines from the same parental control cells, we potentially eliminate the genetic background differences as a confounding factor. This allows us to directly attribute the observed phenotype to TYR loss alone. The isogenic approach not only strengthens the causal link between genotype and phenotype but also suggests that the structural defects observed in patient lines are faithfully recapitulated in the CRISPR engineered lines, thereby validating and extending our earlier findings.

Etiology of foveal hypoplasia and misrouting of optic nerve fibers associated with different form of albinism is not understood but are probably developmental in origin. Interestingly, most forms of albinism are not associated with any form of retinal or RPE degeneration or age related-macular degeneration, bringing in question the role of melanosomes and melanin in RPE and choroid. It also suggests that RPE morphology and functionality should be relatively normal and our preliminary studies suggests this.¹⁷ Although we did observe some abnormally organized apical processes in patient-derived OCA1A-iRPE,¹⁷ these were absent from TYR^−/− iRPE suggesting these were possibly in vitro culture artifacts.

A consistent observation that was made in albino mice RPE and OCA1A-iRPE were the presence of abnormal P-cadherin staining³⁰ and pan-cadherin¹⁷ and catenin mislocalization, respectively. These defects were consistently replicated in the TYR^−/− iRPE, but the electro-physiological manifestation of these altered staining pattern is unclear because we did not observe significant changes to the electrical resistance across the OCA1A and TYR^−/− iRPE monolayer tissue. Also, there are no reports suggesting that albino mice or OCA1A patients’ RPE tissue is susceptible to degeneration or breakdown of the outer blood-retinal barrier. We speculate that the aberrant localization of junctional proteins observed in TYR^−/− iRPE might be mechanistically linked to the lack of mature melanosomes. RPE cells exhibit a well-defined apical-basal polarity, with melanosomes specifically localized to the apical region. This apical localization depends on a specialized cytoskeletal architecture that supports organelle positioning and cell polarity. In the absence of melanosomes, as in TYR^−/− iRPE, this cytoskeletal organization might be disrupted, leading to loss of proper spatial cues necessary for the maintenance of junction complexes. Therefore the lack of melanosomes could indirectly cause mislocalization of junctional proteins by disturbing cytoskeletal organization.

The animal models of OCA1A are well characterized and recapitulate pigmentation defects robustly^33,34; however, these studies focus on pigmentation defects in the skin melanocytes with very few reports involving ocular pigmentation defects.³⁵ Although albino rodent retinas may exhibit retinal degeneration with advanced age under certain light exposure paradigms likely due to the absence of melanin and increased phototoxicity, this does not translate into a higher incidence of spontaneous retinal degeneration under standard lighting conditions, where albino and pigmented strains generally maintain comparable retinal structure over time.³⁶ Also patients with OCA1A generally have stable retinal function throughout their life.⁶ Animal models are ideal for studying retinal developmental in the presence of pigmented and the albino eyes, the only drawback being the lack of fovea and differences in the retino-tectal targeting in the rodents. This is especially important as prior research suggests that RPE monolayer is a developmental regulator of retinogenesis.^37,38 Animal models for studying OCA-1A, although valuable for developmental studies, are not suitable for high-throughput drug discovery studies, and currently, there are no available human cell lines for studying OCA1A related pigmentation defects. We have shown that our in vitro model faithfully recapitulates the in vivo human disease biology at the cellular and the organelle level. The TYR^−/−-iPSC we have generated will provide a consistent source of target cell/tissue type for studying pigmentation defects and identifying drugs for rescuing these defects.