Ion channels research in hPSC-RPE cells: bridging benchwork to clinical applications

Abstract

Ion channels in retinal pigment epithelial (RPE) cells are crucial for retinal health and vision functions. Defects in such channels are intricately associated with the development of various retinopathies that cause blindness. Human pluripotent stem cells (hPSC)-derived RPE cells, including those from human-induced pluripotent stem cells (hiPSC) and human embryonic stem cells (hESC), have been used as in vitro models for investigating pathogenic mechanisms and screening potential therapeutic strategies for retinopathies. Therefore, the cellular status of hPSC-RPE cells, including maturity and physiologic functions, have been widely explored. Particularly, research on ion channels in hPSC-RPE cells can lead to the development of more stable models upon which robust investigations and clinical safety assessments can be performed. Moreover, the use of patient-specific hiPSC-RPE cells has significantly accelerated the clinical translation of gene therapy for retinal channelopathies, such as bestrophinopathies. This review consolidates current research on ion channels in hPSC-RPE cells, specifically Kir7.1, Bestrophin-1, CLC-2, and Ca_V1.3, providing a foundation for future research.

Introduction

Retinal pigment epithelial (RPE) cells constitute a polarized monolayer between photoreceptors and Bruch membrane. The RPE cells participate in the maintenance of normal visual function through establishing the outer blood-retinal barrier, phagocytosing shed photoreceptor outer segments (POS), regulating hydro-ionic homeostasis within the subretinal space, supporting the retinoid cycle, secreting growth factors, and absorbing scattered light [1, 2]. Ion channels are crucial for maintaining these functions. For instance, Kir7.1 (potassium channel), Bestrophin-1 (chloride channel), Ca_V1.3 (calcium channel), and Na_V1.4 (sodium channel) are all involved in the regulation of POS phagocytosis [3,4,5,6]. Disruption of certain ion channels in RPE cells can lead to various retinopathies, including bestrophinopathies [7], retinitis pigmentosa (RP) [8], Leber congenital amaurosis (LCA) [9], and age-related macular degeneration (AMD) [10]. Thus, the role of ion channels in RPE cells warrants comprehensive evaluation.

Over decades, various effective methods have been established for generating RPE cells from human embryoid stem cells (hESC) and human induced pluripotent stem cells (hiPSC) [11,12,13]. Notably, hiPSC- and hESC-derived RPE cells differ in their genetic and ethical aspects. hiPSC-RPE cells have a genetic makeup identical to the donor’s somatic cells, and mitigate ethical concerns surrounding the utilization of embryonic cells. However, individualized hiPSC-RPE cells are costly and require longer preparation time. Moreover, the process of reprogramming somatic cells into hiPSC may introduce novel pathogenetic mutations or epigenetic modifications that could potentially affect their functionality and stability [14]. Despite these differences, both hiPSC- and hESC-RPE cells are classified as human pluripotent stem cell-derived RPE (hPSC-RPE) cells due to their nearly indistinguishable morphological and functional characteristics. These cells highly express key RPE-specific markers, including RPE65, CRALBP, MITF, Bestrophin-1, and MERTK, and effectively recapitulate critical RPE functions, such as phagocytosis of POS, secretion of VEGF/PEDF, and participation in the retinoid cycle [15, 16]. Electrophysiological studies have shown that hPSC-RPE cells possess membrane potentials and ionic currents comparable to those observed in freshly isolated RPE cells, indicating a degree of functional maturity that closely resembles primary human RPE cells [1, 2]. Consequently, hPSC-RPE cells have become invaluable tools for modeling retinal diseases, investigating pathogenetic mechanisms, and advancing therapeutic approaches such as gene therapy, cell transplantation, and high-throughput drug screening [17,18,19].

Research on ion channels in hPSC-RPE cells has significantly advanced the clinical translation of therapies for retinopathies. It has deepened our understanding of the pathogenic mechanisms and therapeutic treatments for channelopathies, including LCA16, CLCN2-IRD, and bestrophinopathies [7, 20, 21]. Additionally, it has surpported the development of hPSC-RPE cells transplantation therapy for patients with other retinopathies, such as AMD and Stargardt disease (STGD) [22, 23]. Although research on ion channels in hPSC-RPE cells has surged in recent years, no comprehensive review has been published. This review consolidates and critically evaluates existing studies on ion channels in hPSC-RPE cells, highlighting their potential clinical applications.

Potassium channels

Potassium (K⁺) channels are one of the most common ion channel families in RPE cells. In this study, we explored the spectrum of findings from potassium channel research in hPSC-RPE cells, especially inwardly rectifying potassium channels (Kir) and voltage-gated potassium channels (K_V). The K_V family has 40 members across 12 subfamilies and is crucial for generating membrane potential and regulating cell volume by facilitating potassium efflux in excitable cells [24]. K_V also induces electrical excitability in RPE cells [25, 26]. Recent studies have investigated the expression of various K_V members, such as K_V1.3 (KCNA3), K_V1.4 (KCNA4), K_V4.2 (KCND2), and K_V7.1 to K_V7.5 (KCNQ1-KCNQ5) in hESC-RPE cells (Fig. 1 and Suppl. Table 1) [27]. These channels produce diverse potassium currents, including delayed rectifier currents inhibited by Agitoxin-2 (targeting K_V1.3) and the sustained M-currents blocked by Linopirdine (targeting K_V) [27]. Additionally, fast-activating currents generated by K_V1.4 and K_V4.2 resemble the transient A-type currents reported in native human RPE cells [27, 28]. The large-conductance calcium-activated potassium channels (maxi-K) can be classified as voltage-activated K⁺ channels. A recent study showed that maxi-K is expressed in the apical and basolateral membrane of hiPSC-RPE cells [29]. Blocking the maxi-K significantly reduces the secretion of basolateral VEGF and apical PEDF; however, it does not disrupt the POS internalization in the cells [29].

Fig. 1

Potassium and chloride channels in hPSC-RPE cells. The subcellular localization of chloride channels (Bestrophin-1, CLC-2) and potassium channels (Kir4.1, Kir7.1, K_V1.3, K_V1.4, K_V4.2, K_V7.1-K_V7.5, maxi-K) in hPSC-RPE cells. Although CLIC4 is detectable in hPSC-RPE cells, its subcellular localization has not yet been described; hence, its localization is inferred based on observations from ARPE-19 cells. Abbreviations: sER, smooth endoplasmic reticulum; rER, rough endoplasmic reticulum; GC, Golgi complex; Mit, mitochondrion

Kir channels are essential for maintaining potassium homeostasis within cells. Although eight Kir channels (Kir1.1, Kir2.1, Kir2.2, Kir3.1, Kir3.4, Kir4.2, Kir6.1, and Kir7.1) have been identified in human RPE cells [30, 31], only Kir7.1 has been extensively studied in hPSC-RPE cells. The research progress of Kir7.1 in hPSC-RPE cells is detailed below. The other seven Kir channels show significantly lower transcript levels than Kir7.1 [30]; however, their potential contributions to RPE cell function should not be underestimated. For instance, Kir4.1 is located in the apical membrane, particularly enriched in the microvilli of both hESC- and hiPSC-RPE cells, although its role remains poor understood [27, 29]. In summary, aside from Kir7.1, research on other potassium channels in hPSC-RPE cells is still in the preliminary stages, hilighting the need for more in-depth studies to elucidate their roles and functions.

Kir7.1

Function and unique characteristics of Kir7.1

Kir7.1, encoded by the KCNJ13 on chromosome 2q37, functions as an inwardly rectifying potassium channel. Its activity is modulated by several factors, including ATP, phosphatidylinositol 4,5-bisphosphate (PIP2), G proteins, sodium ions, and protons [32, 33]. Kir7.1 is a highly divergent member of the Kir channel family, sharing approximately 50% protein sequence homology with its closest relatives, Kir4.1 and Kir4.2 [34]. This distinct amino acid sequence may confer specialized functions to Kir7.1 in RPE cells, including the maintenance of potassium ion homeostasis and the regulation of POS phagocytosis [5, 35, 36]. Kir7.1 is highly expressed on the apical membrane of native human RPE cells [27, 37] and hiPSC- or hESC-derived RPE cells [5, 27, 29, 38]. Intriguingly, Kir7.1 has also been identified on the basolateral membrane in hESC-RPE cells [27], likely due to differences in culture conditions and cultivation time. These findings suggest that Kir7.1 could serve as a valuable biomarker to assess the polarity and maturation status of hPSC-RPE cells for clinical transplantation.

Advances in KCNJ13-IRD pathogenesis and treatment strategies

Mutations in KCNJ13 lead to two types of channelopathies, including Snowflake vitreoretinal degeneration (SVD, MIM#193230) [39] and Leber congenital amaurosis 16 (LCA16, MIM#614186) [20, 40]. SVD is a very rare autosomal dominant inherited retinal disease (< 1/1000000), characterized by early-onset cataracts, vitreous degeneration, and peripheral retinal abnormalities (minute crystalline-like deposits called snowflakes) [34]. LCA16 is a congenital autosomal recessive retinal degeneration disease characterized by significant disruption of both RPE and photoreceptors, leading to severe visual impairment in patients from the first year of life [20, 40]. Both channelopathies underscore the indispensable role of Kir7.1 in maintaining normal retinal homeostasis and function. Although homozygous Kcnj13 null mice die within the first-day post-birth [41,42,43], mice with RPE-specific conditional Kcnj13 knockout and those with mosaic Kcnj13 expression in the RPE accurately reproduce the severe and progressive degeneration of photoreceptors seen in patients with KCNJ13 mutations [41, 44]. Notably, photoreceptors adjacent to RPE cells lacking Kir7.1 undergo degeneration, whereas those next to Kir7.1-expressing cells can survive, highlighting the critical role of this channel in maintaining photoreceptor integrity and survival [41]. Further mechanism studies on the Kcnj13 mutant zebrafish (obelix^td15, c.502T>C [p.F168L]) revealed impaired phagocytosis, abnormally distributed melanosome, and increased mitochondrial size and number in RPE cells [36].

hPSC-RPE cells have significantly accelerated research into the pathogenetic mechanisms for LCA16. The first LCA16 patient-specific hiPSC-RPE cells was obtained from a 10-year-old boy with a homozygous nonsense mutation (p.W53X) of KCNJ13 [5, 20]. The cells exhibited normal morphology with significantly decreased Kir7.1 activity and impaired POS phagocytosis [5]. Besides, Ying et al. generated Rab28^−/− mouse and showed that Kir7.1 may facilitate POS phagocytosis in murine RPE cells by combining with Rab28 [45]. Yuki et al. demonstrated a direct link between Kir7.1 and phagocytic function in RPE cells. Their study revealed impaired phagocytosis in KCNJ13^−/− hiPSC-RPE cells accompanied by downregulation of phagocytosis-related genes compared to wild-type controls [38]. In addition, they found that the KCNJ13^−/− hiPSC-RPE cells exhibits enlarged cell size, disrupted cell alignment, broken cell surfaces, and heightened sensitivity to oxidative stress [38, 46]. These findings deepen our understanding of the mechanisms driving photoreceptor degeneration in patients with KCNJ13 mutations.

Notably, preclinical therapeutic research on LCA16 hiPSC-RPE cells has made significant progress. Since LCA16 is an autosomal recessive inherited disease, restoring partial expression of wild-type Kir7.1 through read-through compounds, such as neuroblastoma 84 (NB84), or via virus-mediated gene augmentation has significantly rescued impaired Kir7.1 currents and phagocytic function in hiPSC-RPE cells [5]. Recently, Kabra et al. successfully and efficiently corrected the p.W53X mutation using silica nanocapsules carrying adenine base editor 8e mRNA and sgRNA in LCA16 hiPSC-RPE cells and LCA16 mouse model (Kcnj13^W53X/+ΔR). This treatment preserves normal vision function and retinal structure in treated mice [47]. Collectively, these preclinical studies have demonstrated the strong potential of gene therapy for treating LCA16. However, given the early onset of this disease, timely therapeutic intervention is crucial to achieving optimal clinical outcomes.

Chloride channels

Chloride, the most prevalent permeable anion in vivo, is mediated by various chloride channels, including chloride ligand-gated channels (CLC), chloride intracellular ion channels (CLIC), calcium-activated chloride channels (CaCC), and cystic fibrosis transmembrane conductance regulator (CFTR). Previous studies have demonstrated that Bestrophin-1 [6], anoctamin families [48, 49], CLC-2 [21], CLIC4 [50] and CFTR [51] are expressed in RPE cells. These chloride channels combined with sodium-potassium-chloride cotransporter (NKCC) maintain hydro-ionic homeostasis by facilitating chloride transepithelial transport from the subretinal space into the choroid [52]. Additionally, chloride channels are essential for critical RPE cell functions, such as POS phagocytosis and VEGF/PEDF secretion [6, 21, 52]. Consequently, understanding the expression and function of chloride channels within RPE cells is essential for unraveling their contributions to retinal health and disease. Herein, we review findings from previous research on chloride channels in hPSC-RPE cells, focusing on disease pathogenetic mechanisms and treatments related to channelopathies.

Bestrophin-1

Overview of BEST1 and bestrophinopathies

BEST1, also known as vitelliform macular dystrophy 2 (VMD2), encodes Bestrophin-1, an integral membrane protein predominantly expressed in the basolateral membranes of RPE cells [53]. Functionally, Bestrophin-1 forms a homopentameric protein complex and is the primary calcium-activated chloride channel (CaCC) in RPE cells [54, 55]. Bestrophin-1 regulates the POS phagocytosis and transepithelial Cl^- transport [56]. Its mutations are related to various bestrophinopathies, including adult-onset vitelliform macular dystrophy (AVMD), best vitelliform macular dystrophy (BVMD), autosomal recessive bestrophinopathy (ARB), autosomal dominant vitreoretinochoroidopathy (ADVIRC) and retinitis pigmentosa 50 (RP50) [57, 58]. However, the diagnosis of RP50 remains controversial, as patients often present with additional pathogenic mutations in genes other than BEST1, which may also contribute to the RP phenotype [59, 60]. Additionally, ADVIRC can easily be misdiagnosed as RP since both conditions share a subset of similar clinical features, such as peripheral circumferential hyperpigmentation [60, 61]. Although several bestrophinopathies cases exhibit typical clinical characteristics, including the decreasing light peak of electrooculogram and specific egg yolk-like lesions, it remains difficult to differentiate all forms of bestrophinopathies from other types of IRD based solely on clinical features, thus necessitating genetic diagnosis.

Advances in disease model and pathogenetic mechanism of bestrophinopathies

Developing more suitable disease models is crucial for advancing our understanding of the diagnosis, treatment, and pathogenesis of bestrophinopathies. Canine multifocal retinopathy (cmr) mirrors bestrophinopathies both genetically and clinically. To date, three naturally occurring BEST1 biallelic mutated canine models have been identified: cmr1 with a stop mutation (p.R25X), cmr2 with a missense mutation (p.G161D), and cmr3 with a frameshift mutation (p.P463fs) [62,63,64]. These models display recessive inheritance patterns, while most bestrophinopathies cases exhibit dominant inheritance, restricting their applicability. In addition to canine models, mice have also been employed to model bestrophinopathies. However, two independently created Best1 KO mice did not exhibit significant retinal abnormalities, except for the reduced light peak of electrooculogram [65, 66]. Although the Best1 p.W93C knock-in mouse displayed pronounced retinal detachments and subretinal debris deposition. The mouse maintained normal chloride currents in RPE cells, an unexpected finding considering the impairment of Bestrophin-1 [67]. Additionally, another knock-in mouse carrying the Best1 p.Y227N mutation displayed no significant ocular abnormalities even up to 2 years of age [68]. These findings suggest that mouse models face challenges in recapitulating the phenotypes of bestrophinopathies in humans.

The development and application of hPSC-RPE cells can significantly alleviate the limitations of traditional animal models, providing a more flexible platform for disease research. Various bestrophinopathies patient-specific hiPSC-RPE cells have been successfully generated. These cells showed the disruption of RPE specific functions, including defects in POS phagocytosis, accumulation of lipofuscin, and reduced transepithelial fluid transport [66, 69,70,71,72,73,74,75,76]. More specifically, Singh et al. found that the BEST1 mutations (p.A146K and p.N296H) impaired POS phagocytosis, possibly due to disturbances in protein degradation pathways, including decreased free-ubiquitin levels, increased protein oxidation, and disturbed exosome secretion rates in patient hiPSC-RPE cells [74, 75]. Lee et al. revealed that the ARB hiPSC-RPE cells carrying p.L40P and p.A915V mutations exhibited decreased transcellular fluid transport and significant differences in epithelial-mesenchymal transition gene profiles and tumor necrosis factor alpha (TNF-α) signaling compared to normal control or BVMD hiPSC-RPE cells with p.G96A mutation [76]. Furthermore, Volume-regulated anion channel (VRAC)-mediated currents were significantly reduced in hiPSC-RPE cells with p.A243V or p.Q238R mutations, indicating the crucial role of Bestrophin-1 in intracellular volume regulation [66].

Mutational effects on subcellular location and function of Bestrophin-1

Different mutations in Bestrophin-1 result in diverse functional outcomes, including loss of function (LOF) [72, 73, 77], gain of function (GOF) [78, 79], and dominant negative (DN) effects [79, 80]. Stephen et al. generated numerous hiPSC-RPE cells from BVMD patients with BEST1 mutations (p.A10T, p.R218H, p.L234P, p.A243T, p.Q293K, and p.D302A) and from ARB patients with p.I201T and p.P274R mutations. These mutations caused LOF by impairing calcium-dependent chloride currents [72, 73, 77]. Notably, p.366fsX18 mutation significantly enhanced chloride currents in hiPSC-RPE cells derived from ARB patients [78]. Similarly, hiPSC-RPE cells from an ADVIRC patient with p.V86M mutation exhibited increased chloride currents [71]. These GOF effects have also been observed in HEK293 cells overexpressing various mutant Bestrophin-1, including p.D203A, p.I205T, and p.Y236C [81].

Some mutations also affect the subcellular localization of Bestrophin-1 [82]. Herein, we focus on the change in hPSC-RPE cells. The p.V235A mutant Bestrophin-1 is abnormally expressed at the apical and basolateral membranes of hiPSC-RPE cells from ADVIRC patients [69]. Cordes et al. demonstrated that transfecting normal hiPSC-RPEs with the p.T6P mutant BEST1 could significantly decrease the membrane localization of both Bestrophin-1 and Ca_V1.3 [83]. Moreover, mutations, such as p.R141H or p.A195V, significantly decreased Bestrophin-1 expression compared with the controls [6]. Interestingly, some heterozygous mutations (p.A10T, p.R218H, p.L234P, p.A243T, p.Q293K, p.D302A) caused allelic expression imbalance (AEI) in patient-specific hiPSC-RPE cells, displaying a significantly higher mRNA expression of the mutated allele than the wild-type BEST1 allele, indicating a DN effect at the transcriptional level [79, 80]. The abnormalities in the expression, subcellular localization, and channel activity of Bestrophin-1 reflect the varied inheritance patterns and the diverse clinical phenotypes in bestrophinopathies. These complexities also indicate the challenges in developing an univeral, effective treatment for all disease subtypes.

Innovative therapy strategies for bestrophinopathies

Developing effective treatments for bestrophinopathies remains a compelling target, because they are one of the most prevalent form of retinal channelopathies. Moreover, bestrophinopathies progress slowly, enabling many patients to retain basic visual functions into their later decades, thus providing a substantial window for therapeutic intervention [80, 84]. Additionally, this disorder is mainly caused by the disruption of RPE cells, which are strategically positioned as a single layer between the neural retina and the choroidal capillaries. This anatomical arrangement facilitates drug delivery via intravitreal, subretinal, or choroidal routes.

Small molecule drugs are considered promising and cost-effective options for treating bestrophinopathies. Specifically, bafilomycin-A1 slows down POS degradation in BVMD hiPSC-RPE cells, whereas valproic acid accelerates POS degradation [75]. Lin et al. found that curcumin could significantly enhance the expression of ZO-1 and Bestrophin-1 compared with other candidate drugs (vitamin C, β-carotene, retinoic acid, forskolin, isoproterenol, Q10), thus improving intracellular reactive oxygen species and promoting phagocytosis functions in BVMD hiPSC-RPE cells [85]. Liu et al. demonstrated that 4PBA (sodium phenylbutyrate), an FDA-approved drug for urea cycle disorders, and its analogue 2-NOAA (2-naphthoxyacetic acid), could rescue the expression, localization, and channel activity of Bestrophin-1 in BVMD or ARB patient-specific hiPSC-RPE cells [86]. However, the effective therapeutic concentrations of 4PBA and 2-NOAA are extremely high (2.5 mM) for practical treatment [86, 87]. Recent studies have discovered that 25 µM of tadalafil, an alternative to 4PBA, can effectively restore chloride current in HEK293T cells overexpressing p.M325T mutation of Bestrophin-1 [87]. However, tadalafil is ineffective for the p.R141H and p.L234V mutations, highlighting the complexity of bestrophinopathies [87].

Compared with small molecule drugs, gene therapy appears to have greater potential for treating bestrophinopathies. Gene augmentation therapy works by expressing an exogenous normal protein corresponding to the mutated protein within cells to rescue cellular function. BEST1 encodes only 585 amino acids [53], indicating that it can easily be encapsulated into adeno-associated virus (AAV), making it highly suitable for this treatment approach. Studies have shown that various BEST1 mutations, which lead to LOF or DN effects, can be effectively addressed by expressing exogenous normal Bestrophin-1. This method restores chloride current and POS phagocytosis in hiPSC-RPE cells, regardless of whether the inheritance pattern is AD or AR [6, 73, 82, 88]. However, some autosomal dominant genetic bestrophinopathies are resistant to gene augmentation, possibly due to the heightened toxicity associated with BEST1 mutations, where the expression of exogenous normal copies cannot sufficiently counteract the negative effects [6, 79]. Gene editing is used in such cases to silence the mutated BEST1 allele. For instance, CRISPR/Cas9 gene editing, specifically targeting the p.A146K mutation in BEST1, can restore chloride current and improve POS digestion function in patient-specific hiPSC-RPE cells [6]. Gene editing can also be used for GOF mutations in bestrophinopathies. Moreover, non-selective CRISPR/Cas9-mediated gene silencing combined with gene augmentation can be used to treat bestrophinopathies. Recent studies have demonstrated that silencing both alleles of BEST1 in hiPSC-RPE cells and expressing exogenous normal BEST1 can effectively rescue the Bestrophin-1-mediated chloride current [79]. Notably, this approach non-discriminately silences both alleles of BEST1, theoretically making it applicable to nearly all bestrophinopathies patients. Nonetheless, the efficiency and potential off-target effects of gene editing should be further evaluated.

CLC-2

CLC-2 and retinal degeneration

Chloride voltage-gated channel 2 (CLC-2), a member of the CLC family, is a two-pore homodimeric voltage-gated chloride channel. It was first cloned from the rat heart and brain in 1992 and was subsequently found to be widely expressed in mammalian cells, especially the epithelial cells [89, 90]. CLC-2 participates in the regulation of cell volume, stabilizing membrane potential, facilitating transepithelial chloride transport, and aiding in signal transduction [90, 91].

Mutations in CLCN2 can lead to various disorders, including leukoencephalopathy with ataxia (LKPAT; MIM#615651) [92, 93], epilepsy (though this remains controversial) [94, 95], familial hyperaldosteronism type 2 (HALD2; MIM#605635) [96, 97], infertility [98] and retinal degeneration [21, 92]. In previous studies, ocular defects were reported in 6 of 15 LKPAT patients carrying homozygous CLCN2 mutations. More specifically, Depienne et al. initially observed low vision and visual field defects in LKPAT patients carrying CLCN2 mutation (p.W570X or p.L144_I145del) without detailed phenotypic descriptions or objective examinations [92]. Bella et al. also detected abnormal ERG and VEP responses in a patient with p.G503R mutation. However, the patient maintained a normal fundus and visual field, suggesting the retinal disorder is mild [98]. Another study detected presbyopia and normal far-sight in a patient with p.H590P mutation. However, the nearly normal fundus appearance did not corroborate the CLCN2 mutation as the cause of these ocular abnormalities [99]. Our studies identified an LKPAT patient with homozygous p.R753X nonsense mutation [21, 100]. This patient showed severe bilateral retinal degeneration involving loss of photoreceptors and RPE cells in the central macula, diffuse chorioretinal atrophy, peripheral retinal bone spicule pigmentation, and bilateral microvascular leakage [21]. Most recently, Cheng et al. reported another LKPAT patient with a novel CLCN2 mutation (p.P461Lfs*13). The patient displayed bilateral blurred vision, posterior capsular cataracts, macular retinal atrophy, peripheral retinal pigmentation, and a novel sign of vacuole-like vitreoretinopathy [101]. Collectively, these cases suggest that retinal degeneration associated with CLCN2 mutations should be classified as a novel autosomal recessive IRD (CLCN2-IRD), which may aid in the identification of similar cases.

Mutant CLC-2 disrupts critical functions of RPE cells

Genetically engineered mouse models can recapitulate the clinical features of CLCN2 mutation-mediated retinal disorders. Mice with LOF mutations in CLC-2, including Clcn2^nmf240 [102], Clcn2^nmf289 [103], Clcn2^tm1Mlv [104], and Clcn2^tm1Tjj [105], exhibit progressive retinal degeneration, leukodystrophy and azoospermia. The retinal disorders persist even when Clcn2 is conditionally knocked out in murine RPE cells [106]. Furthermore, targeted expression of CLC-2 in RPE cells can rescue photoreceptor degeneration in Clcn2 KO mice, indicating the critical role of RPE cells as the primary site of the retinal degeneration [107]. Nevertheless, the localization of CLC-2 in RPE cells remains contentious. Ectopic HA-tagged CLC-2 is specifically localized in the apical membrane of RPE cells of Clcn2 KO mice, aligning with subcellular locations observed in human fetal RPE (hfRPE) and ARPE-19 cells [107]. Recent studies have found the same subcellular location of CLC-2 in hiPSC-RPE cells [21]. However, Mamaeva et al. showed distinct subcellular locations of CLC-2 in hiPSC-RPE cells (basolateral membrane and apical microvilli). They also showed that CLC-2 was highly expressed at the apical pole of fluid-filled domes [29]. CLC-2 has also been reported in both cell center and intercellular junction of hESC-RPE cells [108]. The variability in the expression and localization of CLC-2 may be influenced by the state of RPE cells and culture conditions used (conventional multi-well plates vs. Transwell inserts). In Transwell inserts, fluid does not accumulate beneath RPE cells, while a subset of RPE cells with active apical-basal fluid transport in culture plates forms fluid-filled domes.

CLCN2 mutations affect essential RPE cell functions, particularly POS phagocytosis. Mamaeva et al. showed that CLC-2 inhibitor (GaTx2) could impede POS phagocytosis in hiPSC-RPE cells [29]. Similarly, CLCN2 p.R753X mutation impairs POS phagocytosis in patient-specific hiPSC-RPE cells [21]. POS is ingested by the RPE cells forming a motile organelle, phagosome, which is digested by interacting with endosomes and lysosomes [109]. In this process, CD36 plays a crucial role in mediating POS ingestion [110]. Microtubule-based motor transport is essential for the timely degradation and clearance of POS [109]. However, both processes can be disrupted in patient hiPSC-RPE cells, indicating that defects in phagocytosis promote the initiation and progression of CLCN2-IRD [21]. Besides POS phagocytosis of RPE in retina, the neuroglia (oligodendrocytes and astrocytes) in the brain and Sertoli cells in the testis play critical roles in the phagocytosis of myelin debris [111, 112] and apoptotic germ cells [113], respectively. Targeted deletion of CLCN2 in these cells promote leukodystrophy and testicular degeneration [106], suggesting that impaired phagocytosis may be a common underlying mechanism in the varied clinical manifestations observed in patients with CLCN2 mutations.

CLC-2 also plays an important role in maintaining the integrity of the tight junction (TJ) barrier, the length and arrangement of apical microvilli, and the secretion of VEGF/PEDF in RRE cells. Studies have shown that transdermal current and voltage are decreased in the RPE cells of Clcn2 KO mice as the tight junction (TJ) protein ZO-1 is accumulated at branch points [105, 107]. CLC-2 overexpression enhances the TJ barrier and increases the expression of occludin in human intestinal Caco-2 cells [114]. However, the hiPSC-RPE cells in patient with CLCN2 p.R753X mutation maintain the normal transepithelial resistance, permeability of FITC-dextran, and expression of occludin, as observed in the epithelia of the small intestine and colon of Clcn2 KO mice [21, 115]. Besides the disruption of TJ barrier, highly disarranged and elongated microvilli have been observed in RPE cells of Clcn2 KO mice. However, it can be rescued by exclusive expression of normal CLC-2 in these cells [102, 107]. Nonetheless, the patient hiPSC-RPE cells have similar distributions, morphologies, and lengths for their apical microvilli compared with normal control [21]. These may be due to the fact that the p.R753X mutant CLC-2 is still inserted into the cell membrane and retains a weak chloride current [21]. As for the secretion of VEGF and PEDF, it is significantly decreased by the CLC-2 inhibitor GaTx2 in hiPSC-RPE cells, suggesting a potential impact on the choroid and neural retina [29].

Collectively, these findings partially support the hypothesis that the disruption of RPE cells may the primary cause of retinal degeneration in CLCN2-IRD. Additionally, it is noteworthy that retinal disorder has not been observed in HALD2 patients or mice carrying CLC-2 GOF mutations [90, 91, 109].

Feasible treatment approaches for CLCN2-IRD

Previous study shows that the photoreceptor degeneration in Clcn2 KO mice can be effectively rescued by targeted expression of normal CLC-2 in RPE cells [106]. In our study, two isogenic hiPSC-RPE cell lines were developed using CRISPR/Cas9 technology to correct one allele of the CLCN2 p.R753X mutation. Both corrected hiPSC-RPE cells showed normal chloride currents and POS phagocytosis [21], illustrating that partial restoration of CLC-2 function can maintain the overall functionality of these cells. These findings suggest that gene therapy is a promising treatment strategy for CLCN2-IRD.

Other chloride channels

Despite the presence of other chloride channels, such as CFTR, CLIC4, Anoctamin-1 and Anoctamin-2 in RPE cells [48,49,50, 116, 117], research on these channels remains sparse. Moreover, Xu et al. analyzed anion conductance in murine RPE cells, hinting at the existence of additional, yet-to-be-identified chloride channels [118]. The hPSC-RPE cells offer a promising platform to investigate the expression, subcellular localization, and function of these chloride channels in RPE cells.

CFTR is a 1480 amino acid-long membrane protein and anion channel belonging to the ABC transporter superfamily. It is expressed in various tissues, including the lung, liver, intestine, sweat gland, reproductive tract, and retina [116, 119]. Therefore, mutations in CFTR can lead to multisystem diseases, such as cystic fibrosis (MIM#219700), congenital bilateral absence of vas deferens (MIM#277180) and hereditary pancreatitis (MIM#167800). Apart from modulating chloride conductance, CFTR regulates the activity of other channels, fluid transport, and lipid metabolism [119]. In hfRPE cells, CFTR is found to be expressed on the apical and basolateral membranes [52, 116]. In hESC-RPE cells, CFTR is concentrated in ring-like structures, but its subcellular localization is not well known [120]. Previous research has demonstrated that CFTR regulates fluid transport and reacidification of alkalinized lysosomes [51, 121]. Furthermore, Cftr KO mice show decreased response in RPE-driven ERG components [122]. However, retinal disorders have not been reported in patients carrying CFTR mutations. Over the last decade, few studies have explored the CFTR in RPE cells and future investigations are warranted to uncover its roles and mechanisms.

CLIC4 is a multifunctional protein belonging to the CLIC family. It exerts chloride channel activity in synthetic lipid bilayers and modulates several intracellular processes such as cell proliferation, apoptosis, apical membrane biogenesis, cytoskeletal remodeling, and endosomal sorting [123, 124]. In rat RPE cells, CLIC4 is highly expressed in the apical microvilli. Silencing it impairs apical microvilli, basolateral infoldings of RPE cells, and reduces retina-RPE adhesion [125]. In human RPE cells, CLIC4 co-localizes with matrix metalloproteinase 14 (MMP14) in the late endosomes of ARPE19 cells. Downregulation of CLIC4 decreases the secretion of MMP14 and MMP2, thereby blocking the degradation of the extracellular matrix (ECM) [50]. Although CLIC4 is also expressed in hESC-RPE cells, its subcellular localization has not been reported [50]. Knockdown of CLIC4 disrupts the cell morphology and decreases the apical secretion of MMP2 in polarized hESC-RPE cells [50]. It has been reported that MMP2 and MMP14 regulate the ECM and might be related to AMD [126, 127]. A novel mouse model with RPE-specific Clic4 KO exhibited major functional and pathological hallmarks of dry AMD [117]. This conditional KO reduced the expression of MMP2 and MMP3 in RPE cells, while no changes were observed in MMP14 [117]. Alterations in MMPs expression may contribute to Bruch’s membrane remodeling, potentially promoting the development of drusen, a hallmark of dry AMD. Given the multifaceted role of CLIC4, current research on its effects in RPE cells remains insufficient. Future studies using both animal models and hPSC-RPE cells are essential for comprehensively elucidating the role of CLIC4 in RPE cells and its involvement in AMD pathogenesis.

Sodium channels

Sodium channels include two major classes: the epithelial sodium channel (ENaC) and the voltage-gated sodium channel (VGSC or Na_V). ENaCs are heterotrimeric proteins predominantly found in polarized epithelial cells. They are key regulators of cellular ion and water homeostasis. Certain ENaC subunits have been detected in human donor RPE cells, indicating their potential roles in sodium transepithelial transport [52, 128]. Currently, the expression and function of ENaC in hPSC-RPE cells has not yet been investigated. In contrast, recent studies have revealed the expression and function of Na_V in hESC-RPE cells [4]. This section primarily focuses on the exploration of Na_V channels in hPSC-RPE cells.

Voltage-gated sodium channel (Na_V)

Na_V consists of an alpha subunit and one or two beta subunits, with the alpha subunit alone being sufficient for channel activity [129]. Further, the alpha subunit contains ten subtypes, ranging from Na_V1.1 to Na_V1.9, and Na_X [130]. They are considered as the hall marks of various excitable cells like neurons, glial cells, and muscles, contributing to the formation of an action potential [131]. RPE cells are classified as non-excitable ones, leading to ongoing debate regarding the expression and function of Na_V in these cells. Electrophysiological studies show that sodium currents are undetectable in mature RPE cells and are only observed in cultured RPE cells. This discrepancy suggests that the presence of these currents may be result from neuroepithelial differentiation associated with in vitro culture conditions [52, 132]. Lidgerwood et al. found no detectable sodium currents in hESC- or hiPSC-derived RPE cells using an innovative one-step differentiation protocol [133]. However, recent research has challenged this view, with findings that the Na_V expression, subcellular localization, and function are detected in both hESC-RPE cells and freshly isolated murine RPE cells [4]. Immunostaining revealed the presence of Na_V1.1 and Na_V1.3 to Na_V1.9 on the apical membrane and/or cell-cell junctions of hESC-RPE cells (Fig. 2 and Suppl. Table 1). During the first nine days of maturation, Na_V became more specifically localized to cell-cell junctions (Na_V1.4) and the apical membrane (Na_V1.5 and Na_V1.8), indicating their potential as indicators of RPE cell maturation and polarization [4].

Fig. 2

Sodium and calcium channels in hPSC-RPE cells. The subcellular localization of sodium channels (Na_V1.1, Na_V1.3 to Na_V1.9) and calcium channels (Ca_V1.1 to Ca_V1.3, Ca_V3.1 to Ca_V3.3) in hPSC-RPE cells. Despite functional studies of TRPML1 in hPSC-RPE cells reported, its localizations in hPSC-RPE cells have not been studied; thus, the subcellular localization is deduced from its locations in ARPE19 cells. Abbreviations: sER, smooth endoplasmic reticulum; rER, rough endoplasmic reticulum; GC, Golgi complex; Mit, mitochondrion

The currents of various Na_V have been recorded in hESC-RPE cells by whole-cell patch-clamp. These currents were confirmed through the application of sodium channel blockers (TTX and QX-314), as well as selective blockers of Na_V subtypes (4,9-anhydro-TTX for Na_V1.6, A-803467 for Na_V1.8 and μ-Conotoxin GIIB for Na_V1.4) [4]. Furthermore, inhibiting these Na_V by blockers or shRNA suppresses the POS phagocytosis, particularly during internalization and processing [4]. Recent study has demonstrated that sodium based voltage spikes in hESC-RPE cells can spread laterally in the monolayer through gap junctions, revealing a novel potential role of RPE cells [134]. Although these findings are compelling, the absence of independent validation from other researchers highlights the need for further studies to draw definitive conclusions.

Calcium channels

Calcium ions, as critical second messengers, regulate various cellular processes in RPE cells [52]. Multiple calcium channels modulate intracellular calcium level. Mutations in these channels have been linked to several retinal diseases, including AMD and bestrophinopathies [56, 83, 135]. However, the current understanding of calcium channels in RPE cells remains incomplete. Inconsistencies among existing findings likely stem from the intricate and sensitive nature of calcium signaling in these cells [3, 29]. In this context, we aim to provide a comprehensive review of current research on calcium channels in hPSC-RPE cells, highlighting the advances made to date and the critical gaps that warrant further investigation.

Voltage-dependent calcium channels (VDCCs), also known as voltage-gated calcium channels (VGCCs), selectively allow the passage of calcium ions when activated by depolarizing membrane potentials. These channels comprise at least 10 different subtypes, including L-type channels (Ca_V1.1-1.4), P/Q-type channel (Ca_V2.1), N-type channel (Ca_V2.2), R-type channel (Ca_V2.3), and T-type channels (Ca_V3.1-3.3) [52].

Ca_V1.3

Subcellular location of Ca_V1.3

Three L-type calcium channels (Ca_V1.1, Ca_V1.2, and Ca_V1.3) are expressed in RPE cells, among which Ca_V1.3 is considered to play a primary role [52, 56, 132, 136, 137]. The activity of Ca_V1.3 is regulated through phosphorylation, which shifts its voltage-dependent activation toward more negative voltage, enhancing its activity at the resting membrane potential of RPE cells [56]. In vivo, Ca_V1.3 is localized at the basolateral membrane of murine RPE cells [138]. In contrast, in vitro studies show that it is present on both the apical and basolateral membranes of cultured murine RPE cells [3]. In hESC-RPE cells, the expression pattern of Ca_V1.3 changes as the cells mature. Initially, on the first day post-confluence, Ca_V1.3 is distributed broadly across the cell. As the cells adopt a more epithelioid and eventually cobblestone morphology, Ca_V1.3 becomes increasingly concentrated on both the apical and basolateral sides. This localization includes distinct apical puncta that co-localize with pericentrin. However, as the cells continue to mature and develop pigmentation, detection of Ca_V1.3 at the basolateral membrane becomes challenging, while the apical localization becomes more homogeneous and pronounced [3]. This dynamic subcellular localization indicates that Ca_V1.3 may serve as a biomarker for the maturity and quality of hPSC-RPE cells.

Role of Ca_V1.3 in retinal diseases

Ca_V1.3 has been shown to participate in the pathological mechanism of various retinal diseases, such as AMD and bestrophinopathies [56, 83, 135]. Research on hPSC-RPE cells has significantly advanced our understanding of the role of Ca_V1.3 in these disorders. In AMD patient-specific hiPSC-RPE cells, Ca_V1.3 exhibits a greater shift toward intracellular compartments compared to control cells, which may alter calcium signaling and indirectly disrupt key molecular pathways related to inflammasome activation and autophagy in AMD pathology [135]. Choroidal neovascularization (CNV) represents the most visually threatening clinical phenotype for AMD. Abnormal secretion of VEGF and PEDF is vital for the development of CNV. VEGF promotes the formation of new blood vessels, while PEDF inhibits neovascularization. Research on hiPSC-RPE cells has shown that blocking L-type calcium channels with nifedipine significantly decreases basolateral secretion of VEGF [29]. Further research has confirmed the effects of nifedipine and demonstrated that treating hESC-RPE cells with the L-type channels activator (-) BayK8644 increases the secretion of VEGF [3]. In addition, nifedipine also significantly decreased PEDF secretion on the basolateral side of hiPSC-RPE cells [29]. In AMD patient-specific hiPSC-RPE cells, apical PEDF was significantly reduced than the control, although the basolateral secretion was unchanged [135]. These observations highlighting the pivotal role of Ca_V1.3 in the regulation of VEGF and PEDF secretion and development of wet AMD [139].

Ca_V1.3 also contributes to the development of bestrophinopathies. The studies indicate that POS phagocytosis is commonly disrupted in bestrophinopathies. In Ca_V1.3 KO mice, it was found that Ca_V1.3 mediated the circadian regulation of POS phagocytosis, exhibiting electrophysiological characteristics similar to those observed in bestrophinopathies patients [56, 140]. Manipulating L-type channels activity in hESC-RPE cells impaired POS phagocytosis, regardless of inhibiting or activating these channels [3]. Furthermore, Cordes et al. demonstrated that mutant Bestrophin-1 (p.D302A) decreased its interaction with Ca_V1.3, thereby decreasing the calcium channel activity and altering membrane localization [83]. This disruption affects calcium signaling in hiPSC-RPE cells, highlighting the essential role of Ca_V1.3 in the development of bestrophinopathies [83].

Other L-type calcium channels

In addition to Ca_V1.3, Ca_V1.1 and Ca_V1.2 have been described in hPSC-RPE cells. Ca_V1.2 is expressed on the basolateral membrane of hESC-RPE cells, while Ca_V1.1 has not been detected [3]. Mamaeva et al. found that Ca_V1.1 and Ca_V1.3 were located in the basolateral membrane in hiPSC-RPE cells [29]. In contrast, Karema-Jokinen et al. reported that all three L-type channels were detectable in hiPSC-RPE cells [135]. The differences in expression and subcellular localization of these channels might be influenced by cell type and culture conditions. Recent studies have uncovered a novel function of L-type calcium channels in RPE cells: downregulation of these channels with nifedipine enhances gap junctional conductance in hESC-RPE cells [134].

T-type calcium channels

Although seldom discussed, T-type channels are also expressed in RPE cells [3, 136]. In primary human RPE cells, the mRNA level of Ca_V3.1 and Ca_V3.3 was detected by RT-PCR [136]. In hESC-RPE cells, fast transient T-type resembling currents are detectable [3]. The Ca_V3.1 and Ca_V3.2 are detected in murine RPE cells and hESC-RPE cells, whereas Ca_V3.3 is not detectable by immunofluorescent staining [3]. It has been shown that Ca_V3.1 is specifically localized on the apical membrane of hESC-RPE cells and in the entire cell membrane of murine RPE cells. Ca_V3.2 is expressed on both the basolateral membrane and cell-cell junctions of these cells [3]. In hiPSC-RPE cells, Ca_V3.1 exhibits a similar distribution, while Ca_V3.3 is also detectable on the apical membrane [29]. These channels regulate the POS phagocytosis, and inhibition of the T-type channels with ML218 in hESC-RPE cells enhances the binding and internalization of POS [3], whereas another T-type channel inhibitor, TTA-A2, reduces the internalized number of POS in hiPSC-RPE cells [29].

Overall, VDCCs are essential for hPSC-RPE cells to perform their functions. However, the regulatory mechanisms of these ion channels in RPE cells are complex and multifaceted. For example, both activators and inhibitors of L-type channels have been shown to impair POS phagocytosis [3]. This intricate regulation may involve negative feedback mechanisms and the interplay of other ion channels, including maxi-K, Kir7.1, Bestrophin-1, and CLC-2 [5, 21, 29, 83, 140]. Furthermore, since hPSC-RPE cells express multiple L-type and T-type calcium channels, the functional loss of one calcium channel could potentially be compensated by others within the same family, as observed in mouse basal forebrain neurons [141]. Altogether, the complex regulatory mechanisms of VDCCs in hPSC-RPE cells necessitate further investigation.

Transient receptor potential (TRP) channels

Transient receptor potential (TRP) channels, comprising 28 members, are classified into seven subfamilies including TRPA, TRPC, TRPM, TRPML, TRPN, TRPP, and TRPV. These channels regulate calcium signaling and various functions of RPE cells. A recent comprehensive study by David et al. summarized the recent advances for these channels in retina [142]. In this section, we focus on the roles of TRPML1 in RPE cells.

Transient receptor potential mucolipin 1 (TRPML1), encoded by the MCOLN1, is expressed on the endo-lysosomal membrane across almost all mammal cells, where it transduces calcium signals and lysosomal homeostasis [143, 144]. It is involved in diverse cellular processes such as lysosomal exocytosis, lipid trafficking, and autophagy [145]. Mutations in TRPML1 cause mucolipidosis type IV (MLIV), a severe neurological lysosomal storage disorder characterized by psychomotor retardation and ophthalmologic abnormalities [146]. In Mcoln1 KO mouse, photoreceptor degeneration and optic nerve pathology have been detected [147]. Recent studies found that treatment of hESC-RPE cells, ARPE19 cells, and murine RPE cells with the TRPML agonist ML-SA1 significantly increased cytoplasmic calcium levels. The increase originates from lysosomal stores rather than the endoplasmic reticulum or extracellular space. Moreover, this TRPML1-mediated response can be diminished by accumulation of lysosomal lipid [148, 149]. These findings shed light on the pathogenetic mechanism underlying MLIV and highlight the critical role of TRPML1 in RPE cells. Besides TRPML1, other TRP channels such as TRPV2, TRPV4, TRPV5, TRPV6, TRPM1, TRPM3, TRPM7, TRPC1, and TRPC4 have not comprehensively been investigated in hPSC-RPE cells [142]. Understanding their functions may provide new insights into retinal disease mechanisms and potential therapeutic targets.

General summary

In this review, we summarize the current research on ion channels in hPSC-RPE cells (Figs. 1 and 2 and Suppl. Table 1). These studies have significantly advanced the clinical applications of hPSC-RPE cells. On one hand, they have directly facilitated research on the pathogenic mechanisms and therapies for channelopathies. On the other hand, they have contributed to the development of more stable and mature RPE cells for preclinical and clinical studies in RPE cell transplantation.

Dysfunction of bestrophin-1, CLC-2, and Kir7.1 channels in RPE cells can lead to bestrophinopathies, CLCN2-IRD, and LCA16, respectively [20, 21, 40, 57, 58, 92]. Patient-specific hiPSC-RPE cells with these channelopathies exhibit impaired RPE functions, characterized by reduced ion channel conductivity and diminished POS phagocytosis. Gene therapies, including gene augmentation and gene editing, have demonstrated promising therapeutic effects in these hiPSC-RPE cells [5, 6, 21]. Moreover, the blood-retinal barrier reduces the immune response to AAV delivery. RPE cells, as a monolayer, are particularly amenable to AAV infection via intravitreal or subretinal injection [150]. Based on these findings, we believe that gene therapies for these channelopathy patients are safe and hold significant promise. However, clinical trials targeting RPE channelopathies are currently lacking, possibly due to the mild and slow progression of vision loss in many patients, which may diminish the need for such trials. Additionally, the rarity of channelopathies such as CLCN2-IRD may limit the commercial viability of developing gene therapy [21]. To facilitate the clinical translation of treatments for these channelopathies, it is essential to optimize current gene therapy strategies to enhance safety, efficacy, and cost-effectiveness.

Since the initial use of hESC-RPE cells in treating AMD and STGD, both hESC- and hiPSC-derived RPE cells have been widely employed in clinical trials for retinal degenerative diseases [22, 23, 151, 152]. While delivery methods vary, using either RPE cell suspensions [22, 151] or cell patches [23, 153], studies consistently demonstrate the safety of hPSC-RPE cell transplantation. Notably, some patients have reported slight improvements in vision following hPSC-RPE transplantation, although surgical procedures and immunomodulation may also contribute to these positive outcomes [12, 153,154,155]. The physiological condition of hPSC-RPE cells is a major factor influencing the success of cell transplantation therapies. Treatments for AMD and STGD specifically target macular degeneration. Research indicates that RPE cells in the macula exhibit distinct cellular morphology and gene expression profiles compared to those in surrounding regions [156, 157]. Furthermore, the macula has a heightened demand for RPE functions, such as POS phagocytosis, N-retinylidene-N-retinylethanolamine (A2E) degradation, vitamin A recycling, and fluid regulation—all of which are closely linked to ion channel activity [3,4,5,6, 158]. By assessing ion channel expression and localization in hPSC-RPE cells, researchers may identify key biomarkers that indicate cell maturity and quality, thus improving the selection of cells for more effective cell transplantation therapy. For example, Ca_V1.3 may serve as a useful biomarker, as its subcellular localization is closely associated with the maturation of hPSC-RPE cells [3].

The hPSC-RPE cells provide a valuable platform for investigating ion channel activity, subcellular localization, expression, and function. However, several challenges persist. Firstly, the differentiation methods used to obtain hPSC-RPE cells from hESC or hiPSC vary widely, leading to inconsistencies in cell purity, maturity, and characteristics [159,160,161]. Furthermore, culture conditions—including media composition (e.g., FBS, B27, NEAA), materials (e.g., plates, chamber inserts, coverslips, amniotic membranes, Parylene C, Poly L-lactic-co-glycolic Acid), and coatings (e.g., Matrigel, laminin, poly-D-lysine) —significantly impact cell growth and polarization [162]. Additionally, as hPSC-RPE cells mature, they progressively acquire pigmentation, establish tight junctions, and attain robust polarity—features essential for POS phagocytosis, transepithelial transport and VEGF/PEDF secretion. However, these characteristics rapidly diminish with passaging, indicating that variations in the maturation stages of hPSC-RPE cells can result in differing experimental outcomes. This may explain discrepancies in the subcellular localization of VDCCs [3, 29, 135] and the expression of Na_V in RPE cells [4, 133]. To ensure reliable results, it is crucial to maintain stringent hPSC-RPE cell culture conditions in future research.

Research on hPSC-RPE cells has greatly enhanced our understanding of ion channels in RPE cells, particularly Bestrophin-1, CLC-2, Kir7.1, and Ca_V1.3. Recent discoveries of Na_V, K_V, and maxi-K channels, along with evidence of electrical excitability and lateral signal propagation in hPSC-RPE cells, suggest novel functions beyond the traditional roles attributed to these cells [27, 29, 134]. Additionally, POS phagocytosis in hPSC-RPE cells is highly sensitive to the activity of various ion channels, including potassium, sodium, chloride, and calcium channels [3,4,5,6]. These findings underscore the urgent need for further in-depth research to elucidate the complex roles of ion channels in hPSC-RPE cells.