Establishing induced pluripotent stem cell lines from two dominant optic atrophy patients with distinct OPA1 mutations and clinical pathologies

Dominant optic atrophy (DOA) is an inherited disease that leads to the loss of retinal ganglion cells (RGCs), the projection neurons that relay visual information from the retina to the brain through the optic nerve. The majority of DOA cases can be attributed to mutations in optic atrophy 1 (OPA1), a nuclear gene encoding a mitochondrial-targeted protein that plays important roles in maintaining mitochondrial structure, dynamics, and bioenergetics. Although OPA1 is ubiquitously expressed in all human tissues, RGCs appear to be the primary cell type affected by OPA1 mutations. DOA has not been extensively studied in human RGCs due to the general unavailability of retinal tissues. However, recent advances in stem cell biology have made it possible to produce human RGCs from pluripotent stem cells (PSCs). To aid in establishing DOA disease models based on human PSC-derived RGCs, we have generated iPSC lines from two DOA patients who carry distinct OPA1 mutations and present very different disease symptoms. Studies using these OPA1 mutant RGCs can be correlated with clinical features in the patients to provide insights into DOA disease mechanisms.

Interestingly, although OPA1 is expressed ubiquitously, most DOA patients only exhibit signs and symptoms related to RGC loss causing optic neuropathy.Due to the limited availability of human retinal tissues, non-human cells or human cell types other than RGCs have been used to study OPA1 in relation to DOA (Lodi et al., 2004;Alavi et al., 2007;Olichon et al., 2007;Song et al., 2007;Zanna et al., 2008;Alavi et al., 2009;White et al., 2009;Dayanithi et al., 2010;Heiduschka et al., 2010;Williams et al., 2010;van Bergen et al., 2011;Williams et al., 2012;Rahn et al., 2013;Varanita et al., 2015;Chao de la Barca et al., 2017;Chao de la Barca et al., 2020).Recent advances in stem cell technology have enabled the derivation of human retinal neurons from pluripotent stem cells (PSCs) (Osakada et al., 2009;Ohlemacher et al., 2015;Sluch et al., 2017;Rabesandratana et al., 2020;Zhang X. et al., 2021), permitting in vitro DOA disease models to be established using patients' induced pluripotent stem cells (iPSCs).However, the heterogeneity of OPA1 genotypes and phenotypes present in the DOA patient population, coupled with the incomplete penetrance of the disease, necessitates that a range of PSC lines containing distinct OPA1 mutations be generated to properly model and understand OPA1-driven DOA.To help establishing DOA models that recapitulate the spectrum of the disease, we generated iPSC lines from two DOA patients with heterozygous OPA1 mutations that are distinct from those in previously reported iPSC lines (Chen et al., 2016;Galera-Monge et al., 2016;Zurita-Diaz et al., 2017;Jonikas et al., 2018;Zhang XH. et al., 2021;Sladen et al., 2021;Chan et al., 2022;Sladen et al., 2022;Sun et al., 2022).Importantly, we also describe the clinical presentation of both DOA patients at the time of iPSC derivation, which is important information to consider when characterizing and validating disease models derived from these iPSC lines.

Genetic and clinical presentation of DOA patients
Two DOA patients with previously determined, heterozygous OPA1 mutations were recruited for the study.Both patients are male and of European ancestry.The first patient, referred to as "Patient 1," presented with symptoms of DOA at age 7 and has a single G insertion in one allele of the OPA1 gene exon 19, which is a part of the central dynamin region of the OPA1 protein.This insertional mutation causes a frameshift that leads to an early stop codon, resulting in a predicted protein product of 652 amino acids as opposed to the 1,015 amino acids encoded by the full-length wild type (WT) transcript (Table 1).The second patient, "Patient 2," was relatively asymptomatic and found to have an OPA1 mutation when his son with vision loss was diagnosed with DOA.Patient 2 has a heterozygous, two base pair AT deletion in exon 13, which encodes part of the GTPase domain of the OPA1 protein.This mutation also causes a frameshift and subsequent premature stop codon.The mutant transcript is predicted to encode a protein of 483 amino acids as opposed to the 1,015 amino acids encoded by the full-length WT transcript (Table 1).
At the time of enrollment and iPSC derivation, Patient 1 was 49 years old and Patient 2 was 60 years old.Neither patient was reported to have extraocular neurological disease or additional neural retina disorders.Additionally, both patients tested negative for bloodborne infectious diseases including hepatitis B, hepatitis C, HIV, and syphilis.The donor patients underwent a comprehensive ophthalmological examination in which visual acuity, tonometry, fundus imaging, Humphrey visual field (HVF) testing, and spectral-domain optical coherence tomography (SD-OCT) were performed.
In contrast, Patient 2 has remained largely asymptomatic over his lifetime.Despite fundus images that showed temporal pallor of the optic nerve, visual acuity was 20/25 in both eyes (Figure 1A; Table 1).Visual field assessment detected a small paracentral scotoma (Figure 1B), which made it difficult for Patient 2 to focus on the central target during Humphrey Visual Field (HVF 30-2) testing, resulting in high false positive and negative rates.The smaller field of HVF (10-2) testing allowed the patient to maintain central fixation on the target and detected the edge of the central scotomas, which are located superior temporally on HVF 10-2 (Figure 1B).SD-OCT revealed bilateral inferior RNFL thinning with more advanced superior and temporal thinning of the left eye (Figure 2B).

Generation of OPA1 heterozygous-mutant iPSCs from DOA patients
To generate iPSC lines from both DOA patients, peripheral blood mononuclear cells (PBMCs) were isolated from patients' blood samples and reprogrammed by episomally expressing the pluripotency factors OCT3/4, SOX2, KLF4, L-Myc, shp53, Lin28, and SV40LT (Okita et al., 2013;Toombs et al., 2020).Two independent iPSC lines (n1a and n1b) were established from each patient.Lines generated from Patient 1 and Patient 2 were named 1iDOAn1a/1iDOAn1b, and 2iDOAn1a/2iDOAn1b, respectively.Although all four iPSC lines were authenticated, data in this report depicts the n1a lines only.
Both 1iDOA and 2iDOA displayed typical pluripotent stem cell morphology and stained positive for alkaline phosphatase (AP) activity (Figure 3A).Immunocytochemistry was performed to further verify the expression of pluripotency markers by the  iPSCs in comparison to an established human embryonic stem cell (ESC) line, UCLA1.All lines were shown to similarly co-express the PSC markers SOX2, OCT3/4, and NANOG (Figure 3B).Furthermore, 1iDOA and 2iDOA iPSCs displayed a normal male karyotype with 22-pair autosomes and XY chromosomes (Figure 3C).The derived iPSC lines tested negative for mycoplasma and were characterized by the short tandem repeat (STR) analysis to establish their genetic identities (not shown).To confirm that the iPSC lines carried the DOA patients' specific OPA1 mutations, genomic DNA was extracted from 1iDOA and 2iDOA iPSCs and used as templates to obtain PCR products spanning the expected mutation sites in exons 19 and 13, respectively.DNA sequencing analysis of the PCR products validated that each patient's existing heterozygous OPA1 mutation was present in the given iPSC lines (Figure 3D).

Detection of ROS production in PSC lines
Since OPA1 mutations may result in elevation of cellular reactive oxygen species (ROS), we performed live cell imaging to monitor mitochondria and detect ROS in DOA patients' iPSC lines and a control H9 ESC line.A similar distribution of mitochondria was observed in 1iDOA, 2iDOA, and the control H9 using MitoTracker Red (Figure 4).However, elevated ROS signals were detected in 2iDOA cells compared to 1iDOA or H9 cells (Figures 4A, C, E).When the cells were treated with a ROS elevating reagent menadione, all PSC lines showed increased ROS levels (Figures 4B, D, F).Further detailed analysis will be necessary to determine the mechanism of the differential ROS production and whether mitochondrial function in PSC-derived neurons are affected.

Discussion
Although OPA1 function has been extensively studied, it remains an enigma why retinal projection neuron RGCs are particularly prone to degenerate when OPA1 function is compromised, given that OPA1 is a ubiquitously expressed mitochondrial protein.One significant roadblock in the past has been the scarcity of human RGCs.Over the last decade, human PSCs have emerged as replenishable sources to produce different somatic tissue and cell types.Since DOA has not been studied in-depth using human RGCs, stem cell-based models provide unprecedented opportunities to investigate DOA disease mechanisms.However, given the molecular and clinical heterogeneity of the DOA patient population and incomplete penetrance of OPA1 mutations, it is unlikely that an in vitro model from PSCs containing one specific OPA1 mutation will be able to recapitulate the full spectrum of the disease.The iPSC lines from the two DOA patients reported here carry OPA1 mutations distinct from those in previously reported iPSC lines (Chen et al., 2016;Galera-Monge et al., 2016;Zurita-Diaz et al., 2017;Jonikas et al., 2018;Zhang XH. et al., 2021;Chan et al., 2022;Sladen et al., 2022;Sun et al., 2022), thus providing additional tools to study DOA using PSC-based human RGC models.Furthermore, the clinical records and DOA-associated ophthalmological symptoms for each DOA patient at the time of iPSC derivation will be useful information when comparing in vitro cellular level findings with clinical manifestations.The two DOA patients' heterozygous OPA1 mutations do not affect the mitochondrial targeting or alternative splicing of OPA1, but are predicted to produce truncated proteins.Ophthalmic examinations showed that Patient 1, whose mutation falls within the central dynamin region of OPA1 (p.Glu650Glyfs*4), has more severe and extensive DOA symptoms.Previous studies have established iPSC lines from DOA+ patients with mutations in the central dynamin region (Galera-Monge et al., 2016;Zurita-Diaz et al., 2017), suggesting a correlation of this regions with a more severe disease pathology.Interestingly, Patient 2 has a frameshift mutation (p.Ile473Phefs*12) that leads to a stop codon within the GTPase domain.Missense mutations in the GTPase domain have been shown to be more likely to cause DOA+ (Hoyt, 1980;Treft et al., 1984;Amati-Bonneau et al., 2008;Yu-Wai-Man et al., 2010b;Maeda-Katahira et al., 2019).However, Patient 2 has remained largely asymptomatic.This suggests that OPA1 proteins with a mutated GTPase domain may act differently from OPA1 proteins with a partially truncated GTPase domain.The mutation carried by 2iDOA iPSC lines is different from the other iPSC lines previously generated containing mutations in the GTPase domain (Sladen et al., 2021;Chan et al., 2022;Sun et al., 2022).To our knowledge, this also is the first report of iPSC line generated from OPA1-DOA patient with a mild symptom.Intriguingly, the initial characterization of 2iDOA PSCs has revealed higher cellular ROS compared to 1iDOA PSCs or the control H9 ESCs.This suggests that under in vitro culture conditions, effects of OPA1 mutations may be more readily detected.
OPA1 protein is known to undergo proteolytic processing upon entry into mitochondria and to form protein complexes (Ishihara et al., 2006;Song et al., 2007;Anand et al., 2014;Del Dotto et al., 2017).Despite the autosomal dominant inheritance pattern of DOA, the disease pathogenicity can occur via dominant negative or haploinsufficiency mechanisms, depending on the type and location of the OPA1 mutations (Delettre et al., 2001).The availability of OPA1 mutant iPSC lines can benefit the further characterization of OPA1 mutant iPSC-derived RGCs and shed light on DOA disease mechanisms, especially regarding the suitability of different types of molecular therapies.In the case of haploinsufficiency, gene supplement therapy providing wild type OPA1 is expected to rescue neuronal deficiencies.On the other hand, for dominantly acting OPA1 mutations, alternative approaches such as RNAi knock-down or CRISPR gene editing can be designed to dampen or eliminate the dominant interfering effects of a given mutant.Therefore, establishing in vitro OPA1-DOA disease models is important for developing molecular therapies and enabling personalized medicine.
Increasing evidence indicates that mitochondrial dysfunction plays an important role in neurodegenerative and aging-related diseases (Carelli et al., 2017;Iannielli et al., 2018).Findings from OPA1-mutant iPSC-derived disease models can advance the understanding of the pathological mechanisms underlying DOA as well as provide important insights into other neurodegenerative diseases that share common metabolic deficiencies with DOA.

Patient enrollment and ophthalmic examinations
This research was conducted at the University of California, Los Angeles under the institutional review board (IRB) protocol

Generation of DOA patient iPSCs
Peripheral blood was collected from DOA patients the same day the ophthalmologic tests were performed.Peripheral blood mononuclear cells (PBMCs) were reprogrammed into iPSCs at the Cedars Sinai Medical Center iPSC core using a nonintegration method as previously described (Toombs et al., 2020).Two independent iPSC lines were established for each DOA patient.The iPSC lines were named as 1iDOAn1a and 1iDOAn1b for Patient 1, and 2iDOAn1a and 2iDOAn1b for Patient 2, respectively.The characterization included mycoplasma testing, alkaline phosphatase staining, G-band karyotyping, and short-tandem repeat (STR) cell line authentication.Lines are referred to as 1iDOA and 2iDOA throughout this manuscript, but experiments described primarily used n1a clones.1iDOA carries the heterozygous OPA1 mutation NC_000003.12:g.193648807dupand 2iDOA carries the heterozygous OPA1 mutation NC_000003.12:g.193643567_193643568del.

Human PSC cell cultures
Human ESCs and iPSCs were maintained in mTeSR TM Plus medium (Stemcell Technologies) supplemented with 1% Antibiotic Antimycotic (Gibco/ThermoFisher Scientific) on Matrigel (Corning) coated plates at 37 °C with 5% CO 2 .PSCs were passaged by dissociating monolayer cells into a single cell suspension with Accutase (Stemcell Technologies).Single cells were plated in mTeSR TM Plus supplemented with 1% Antibiotic Antimycotic and 10 μM of Y-27632 (Stemcell Technologies) for 24 h, after which the medium was replaced with mTeSR TM Plus with 1% Antibiotic Antimycotic.The medium was changed no less frequently than every other day.

Immunofluorescent staining and imaging
PSCs grown on Matrigel-coated NUNC ™ Thermanox ™ 13 mm plastic coverslips (ThermoFisher Scientific) were fixed in 4% paraformaldehyde in PBS for 2 min and then incubated in blocking solution (0.1% TritonX-100, 2% donkey serum, 10% FBS in DMEM).Coverslips were incubated with primary antibodies, followed by secondary antibodies and DAPI diluted in blocking solution (Supplementary Table S1).All incubations were for 1 h at room temperature, and staining periods were followed by three 5-min washes in PBS with 0.1% Tween 20.Coverslips were mounted on glass slides (Fisher Scientific) with mounting media (Fluro-Gel, Electron Microscopy Sciences) and imaged using an Olympus BX61 scanning laser confocal microscope with Plan-APO objectives.

Sequencing of genomic PCR products
Prior to enrolling in the study, each DOA patient's specific OPA1 mutation was classified/diagnosed via exon sequencing.To confirm the presence of these mutations in the DOA patients' iPSC lines, genomic DNA was isolated from iPSCs using the Purelink genomic DNA Mini Kit (Invitrogen).200 ng of DNA was used as templates in PRCs using Hot Star Taq DNA Polymerase (Qiagen) and primers flanking the respective OPA1 mutation site (Supplementary Table S2).Thermocycler conditions were as follows: 15 min at 95 °C; 35 cycles of 30 s at 94 °C, 30 s at 59 °C and 1 min at 72 °C; and 10 min at 72 °C.PCR products were sequenced by Sanger method using the primers listed in Supplementary Table S2.

Cellular ROS and mitochondria imaging
PSCs grown as a monolayer were washed once with PBS, followed by incubating in mTeSR TM Plus medium containing 5 μM CellROX Green (Invitrogen), 250 nM MitoTracker Red (Invitrogen), and 5 μg/mL Hoechst 33342 at 37 °C for 30 min.A different set of PSC cultures were incubated with 100 μM menadione at 37 °C for 1 h prior to detection using CellROX and MitoTracker.After washing with PBS and HBSS without phenol red, cells were immediately imaged using an Olympus IX81 scanning laser confocal microscope with Plan-APO objectives.All images were captured under identical conditions.used in the study.AS and JC interpreted DOA patient clinical data.KP and XZ performed experiments described in the manuscript.KP and X-JY wrote the manuscript, AS and JC contributed to the manuscript editing.All authors contributed to the article and approved the submitted version.

Funding
This work was in part supported by NIH grant 2R01EY026319 and CIRM grant DISC2-13475 awarded to X-JY, NIH grant F31EY033242 awarded to KP, NIH core grant P30EY000331, and an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology at University of California, Los Angeles.NIH training grant 5T32EY007026.

FIGURE 3
FIGURE 3 Characterization of OPA1-mutant iPSCs from DOA patients.(A) Brightfield images of iPSCs reprogrammed from DOA Patient 1 (1iDOA) and Patient 2 (2iDOA) display normal pluripotent stem cell morphology (top row) and stain positive for alkaline phosphatase (bottom row).Scale bars: 200 μm (top row), 500 μm (bottom row).(B) 1iDOA and 2iDOA iPSCs express the pluripotent stem cell markers SOX2, OCT3/4, and NANOG similarly to the wild type ESC line, UCLA1.Scale bar: 40 μm.(C) Both iPSC lines display a 46, XY normal male karyotype.(D) DNA sequences of iPSC lines 1iDOA and 2iDOA (bottom row) aligned to the corresponding regions of OPA1 in the wild type (WT) reference genome (top row).Sequences confirm that 1iDOA contains a heterozygous, single base pair (G) insertion in exon 19 and that 2iDOA has a heterozygous, 2 base pair (AT) deletion in exon 13.Peaks reflect both WT and Mutant allele bases at a given position.Allele sequence identities are clarified above peaks.Boxed regions indicate bases inserted (1iDOA, mutant allele) or deleted (2iDOA, WT allele).

FIGURE 4
FIGURE 4 Detection of ROS production in PSC lines.Detection of mitochondria and cellular ROS in pluripotent stem cell (PSC) lines were performed using live cells.(A, C, E) show representative signals of ROS (green), MitoTracker (red) and merged images for 1iDOA, 2iDOA, and the control ESC H9 lines.(B, D, F) show merged ROS and MitoTracker signals after the perspective PSC lines were treated with menadione to increase ROS.Scale bar, 20 μm for all.

TABLE 1
Summary of patients' OPA1 mutations and clinical symptoms.