Restoration of impaired lysosomal function mitigates drusen-like deposit formation and cell death in Malattia Leventinese

Age-related macular degeneration (AMD) is a leading cause of blindness, with its prevalence continuing to rise in developed countries. It is classified into 2 main types: neovascular AMD, characterized by macular neovascularization, and atrophic AMD, which lacks neovascularization but causes chorioretinal atrophy (1, 2). Both types are thought to result from age-related changes in retinal pigment epithelial (RPE) cells, particularly those related to lipid metabolism. A hallmark of AMD is the presence of drusen — deposits that accumulate between the RPE and Bruch’s membrane (3). Drusen primarily consist of lipid-rich waste products (4), reflecting age-related changes in the RPE. In addition to lipids, proteins involved in the complement activation pathway have also been identified as drusen components (5). However, the precise mechanisms underlying drusen formation remain unclear.

Malattia Leventinese (MAL) is a disease that causes macular dystrophy with clinical features similar to those of AMD, including drusen appearing in early adolescence, macular degeneration, and neovascularization (6). MAL is caused by the Arg345Trp (R345W) mutation in the EGF-like fibulin extracellular matrix protein 1 (EFEMP1) gene (6, 7), which encodes fibulin-3 — a secreted protein involved in extracellular matrix formation and rich in disulfide groups. In patients with MAL, the R345W mutation causes fibulin-3 misfolding, leading to its intracellular retention (8) and aberrant accumulation on the RPE basement membrane instead of proper extracellular secretion (9). Although the exact function of fibulin-3 remains unclear, it regulates matrix metalloproteinase 2 (MMP2), an enzyme that degrades type IV collagen, as well as metalloproteinase inhibitors 1 and 3 (TIMP1 and TIMP3) that inhibit MMPs (10). Mouse models of MAL exhibit sub-RPE deposits, complement activation, and RPE atrophy — pathological features also observed in AMD (11, 12) — suggesting a potential link between the 2 diseases.

Mechanistic insights into EFEMP1-associated pathology have advanced through in vitro models. Using CRISPR-edited ARPE-19 cells harboring the R345W EFEMP1 mutation, Fernandez-Godino et al. demonstrated that the abnormal extracellular matrix produced by mutant cells promotes basal deposit formation and chronic activation of the alternative complement pathway (13). This study highlighted the critical role of extracellular matrix alterations in driving complement activation and sub-RPE deposit formation; however, it did not address intracellular disease mechanisms within RPE cells. More recently, patient-derived RPE cells differentiated from induced pluripotent stem cells (iPSC-RPE) carrying the EFEMP1 mutation have been reported to impair cholesterol efflux, leading to intracellular lipid accumulation and altered lipid homeostasis (14). This implicates dysregulated lipid metabolism as a key contributor to MAL pathology. Nevertheless, the relationship between impaired lipid handling and intracellular protein degradation pathways, as well as its contribution to drusen-like deposit formation, remains incompletely understood.

The cellular pathways linking lipid metabolic abnormalities to drusen-like deposit formation remain unclear. Notably, untargeted lipidomic analyses in EFEMP1-mutant RPE models have not been fully explored. Since lysosomes play a central role in lipid turnover and intracellular degradation, alterations in lipid composition may reflect underlying lysosomal dysfunction, which could, in turn, promote the accumulation of undegraded proteins and lipids, exacerbating extracellular matrix abnormalities and sub-RPE deposit formation. Therefore, in this study, we derived RPE cells from iPSCs generated from a patient with MAL to further elucidate disease mechanisms beyond extracellular matrix alterations and lipid dysregulation. By first performing untargeted lipidomic analyses and subsequently focusing on lysosomal function and its regulation, we aimed to clarify how the EFEMP1 mutation drives drusen-like deposit formation and RPE degeneration while also exploring potential therapeutic strategies targeting lysosomal dysfunction. Human iPSC-derived models provide a unique platform to investigate human-specific intracellular disease mechanisms that cannot be fully captured by transformed cell lines or animal models, thereby offering valuable insights into both MAL and AMD pathogenesis.

Generation of MAL patient–specific iPSCs and iPSC-derived RPE. We established iPSC lines from a patient with MAL (MAL1 and MAL2) and 2 healthy individuals (NOR1 and NOR2), and we then differentiated them into iPSC-RPE cells. The differentiated iPSC-RPE cells exhibited polygonal, cobblestone-like morphology and were cultured until they developed high pigmentation, indicating full functional maturity. Immunocytochemical staining showed strong expression of tight junction protein 1 (ZO-1), a tight junction marker, in both NOR and MAL iPSC-RPE cells (Figure 1A). To further assess epithelial junctional integrity and apical polarity, we performed immunostaining for claudin-16 and claudin-19, tight junction–associated claudins expressed in RPE (15). The results indicated continuous junctional localization with comparable staining patterns in NOR and MAL iPSC-RPE cells (Figure 1B and Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.194102DS1). In addition, phosphorylated ezrin/radixin/moesin (pERM), the active phosphorylated form of ERM proteins involved in apical membrane organization (16), exhibited similar apical localization in both NOR and MAL iPSC-RPE cells (Figure 1C), supporting preserved apical polarity.

Figure 1

Altered fibulin-3 distribution in MAL iPSC-RPE. Induced pluripotent stem cell–derived retinal pigment epithelial (iPSC-RPE) cells were cultured for 4 weeks to ensure adequate differentiation. (A and B) Immunostaining of normal (NOR) and Malattia Leventinese (MAL) iPSC-RPE cells for tight junction protein 1 (ZO-1, green, A) and claudin-16 (red, B), counterstained with DAPI (blue). (C) Immunostaining of the cross-sections of iPSC-RPE for phosphorylated ezrin/radixin/moesin (pERM, red), phalloidin (green), and DAPI (blue). (D) Transmission electron microscopy images of NOR and MAL iPSC-RPE cells in vertical cross-sections. (E and F) Western blot analysis of retinal pigment epithelium-specific 65 kDa protein (RPE65), retinaldehyde binding protein 1 (CRALBP), bestrophin-1 (BEST1), and fibulin-3 in NOR and MAL iPSC-RPE cells. Data are shown as mean ± SD, n = 5 biological replicates. (G) Expression levels of the EGF-like fibulin extracellular matrix protein 1 (EFEMP1) gene in NOR and MAL iPSC-RPE cells were analyzed by RNA-seq. The y axis represents expression levels in log₂ TPM (transcripts per million). (H) Immunostaining of the cross-sections of iPSC-RPE for fibulin-3 (red), phalloidin (green), and DAPI (blue). Scale bar: 20 μm (A, B, C, and H); 2 μm (D).

Transmission electron microscopy (TEM) confirmed that NOR and MAL iPSC-RPE cells formed monolayers of highly polarized cells with apical microvilli and melanosomes (Figure 1D). Key RPE biomarkers, including RPE-specific 65 kDa protein (RPE65), retinaldehyde binding protein 1 (CRALBP), and bestrophin-1 (BEST1), were detected in both NOR and MAL iPSC-RPE cells (Figure 1, E and F). Although EFEMP1 RNA expression levels were comparable between MAL and NOR iPSC-RPE cells (Figure 1G), fibulin-3 protein levels were higher in MAL iPSC-RPE cells (Figure 1, E and F). Immunostaining further revealed enhanced fibulin-3 immunoreactivity on and beneath the basal side of the MAL iPSC-RPE cells compared with NOR controls (Figure 1H).

Cytotoxicity in MAL iPSC-RPE. Because intracellular accumulation of misfolded proteins can induce cellular stress and apoptosis, we next examined whether MAL iPSC-RPE cells exhibit increased cytotoxicity. RPE cells are highly specialized, postmitotic cells that phagocytose large amounts of photoreceptor outer segments (POS) daily in vivo and play a central role in lipid metabolism and waste clearance. Accordingly, POS were introduced in the following experiments to impose a physiologically relevant phagocytic and lysosomal burden and to better model disease-associated cellular stress conditions in vitro. Immunostaining of iPSC-RPE cross-sections revealed an accumulation of heat shock protein family A (HSP70) member 5 (HSPA5) in the MAL iPSC-RPE cells (Figure 2A). Additionally, cleaved caspase-3, the active form of the enzyme involved in apoptosis, was detected in MAL iPSC-RPE cells (Figure 2, B and C). MAL iPSC-RPE cells revealed a significant increase in TdT-mediated dUTP nick-end labeling⁺ (TUNEL)⁺ apoptotic cells (Figure 2, D and E). Furthermore, the expression of growth arrest specific 6 (GAS6) and tyrosine-protein kinase receptor (AXL) — components of the Tyro3/Axl/MerTK (TAM) signaling pathway involved in the phagocytosis of apoptotic cells (17) — was significantly upregulated in MAL iPSC-RPE cells (Figure 2F and Supplemental Figure 2A). These findings indicate that MAL-derived iPSC-RPE cells undergo increased cellular damage and apoptotic cell death.

Figure 2

Apoptosis and extracellular matrix degradation in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by an additional 4 weeks of incubation in the presence or absence of photoreceptor outer segments (POS). (A) Immunostaining of cross-sections of NOR and MAL iPSC-RPE, showing heat shock protein family A (Hsp70) member 5 (HSPA5, red), phalloidin (green), and DAPI (blue). (B and C) Western blot analysis of total and cleaved caspase-3 (t-Casp3 and c-Casp3) in NOR and MAL iPSC-RPE cells treated with POS. (D and E) TdT-mediated dUTP nick-end labeling (TUNEL) staining (green) of cross-sections of NOR and MAL iPSC-RPE cells. (F and G) The expression levels of growth arrest specific 6 (GAS6), tyrosine-protein kinase receptor (AXL), matrix metalloproteinase-2 (MMP2), and metalloproteinase inhibitors 1 and 3 (TIMP1 and TIMP3) in NOR and MAL iPSC-RPE cells were analyzed by RNA-seq. The y axis represents expression levels in log₂ transcripts per million (TPM). Corresponding analyses, including both POS (–) and POS (+) conditions, are shown in Supplemental Figure 2. (H and I) Western blot analysis of MMP2, TIMP1, and TIMP3 expression in NOR and MAL iPSC-RPE cells. (J) Measurement of MMP2 activity in medium supernatants of NOR and MAL iPSC-RPE using gelatin zymography. Scale bar: 20 μm (A and D). *P < 0.05, ***P < 0.005. Data in C, E, I, and J were analyzed using a linear mixed-effects model followed by Tukey’s HSD test. Data in F and G were analyzed using Student’s t test; n = 3 (C, F, G, and J), n = 5 (I), n = 10 (E). Data are shown as mean ± SD.

Activation of matrix metalloproteinase in the extracellular milieu of MAL iPSC-RPE. Given the cellular damage observed in MAL iPSC-RPE cells during prolonged culture, and considering that fibulin-3 regulates extracellular matrix homeostasis by modulating matrix metalloproteinases and their inhibitors (10), we focused on MMP2 and its endogenous inhibitors, TIMP1 and TIMP3. Gene expression analysis showed no significant difference in MMP2 levels between NOR and MAL iPSC-RPE cells (Figure 2G and Supplemental Figure 2B). However, the expression of TIMP1 and TIMP3 was significantly reduced in MAL iPSC-RPE cells (Figure 2G). In contrast, Western blot analysis revealed that MMP2 and TIMP1 protein levels were higher in MAL iPSC-RPE cells than in NOR iPSC-RPE cells (Figure 2, H and I), while TIMP3 protein levels remained similar between the 2 groups. Additionally, gelatin zymography of culture supernatants demonstrated that secreted and active MMP2 levels were elevated in MAL iPSC-RPE cells compared with NOR iPSC-RPE cells (Figure 2J). These findings suggest that MMP2 activation in the extracellular milieu of MAL iPSC-RPE cells may contribute to enhanced extracellular matrix degradation, potentially playing a role in disease progression.

Formation of drusen-like deposits beneath MAL iPSC-RPE. Because extracellular matrix dysregulation and impaired waste clearance by RPE are central to drusen formation in AMD and MAL, we next examined whether MAL iPSC-RPE cells form drusen-like deposits. When MAL iPSC-RPE cells were cultured for 4 weeks and subsequently exposed to POS, whitish drusen-like structures began to appear beneath the cells after approximately 3 weeks of POS treatment. POS exposure was associated with a trend toward increased size and prominence of these deposits (Supplemental Figure 3, A and B). To further characterize these structures, we performed immunostaining of both cross-sections and en face preparations of iPSC-RPE cultures. In NOR iPSC-RPE cells, apolipoprotein E (ApoE) was detected mainly as fine punctate signals within the membrane, with little intracellular accumulation, consistent with physiological secretion (Figure 3A). In contrast, MAL iPSC-RPE cells exhibited intracellular accumulation of ApoE in the form of enlarged punctate structures (Figure 3, A and B, and Supplemental Figure 3C), as well as ApoE⁺ deposits located between the cells and the culture membrane (Figure 3A and Supplemental Figure 3C), distinct from the fine membrane-associated signals seen in NOR cultures. Consistent with these findings, en face immunostaining further demonstrated prominent intracellular ApoE⁺ puncta in MAL iPSC-RPE cells (Figure 3D). Quantitative analysis confirmed that, compared with NOR controls, MAL iPSC-RPE cells exhibited a significantly higher proportion of cells exhibiting enlarged intracellular ApoE⁺ puncta, together with increased ApoE immunoreactivity within intracellular regions and in the subcellular space adjacent to the culture membrane (Figure 3C). Further analysis revealed that fibulin-3, complement 3 (C3), and collagen IV immunoreactivity were predominantly associated with the drusen-like structures located in the subcellular space beneath the RPE layers and adjacent to the culture membrane in MAL iPSC-RPE (Figure 3, E–J, and Supplemental Figure 3, D–F). Consistent with these observations, C3⁺ puncta were significantly more abundant in MAL iPSC-RPE (Figure 3H), and fluorescence intensity measurements revealed significantly higher immunoreactivity for fibulin-3 and collagen IV in MAL iPSC-RPE cells relative to NOR iPSC-RPE cells (Figure 3, F and J). Notably, the gene expression levels of ApoE and C3 were slightly higher in MAL iPSC-RPE than in NOR iPSC-RPE, suggesting that these proteins were likely produced and secreted by RPE cells (Supplemental Figure 3G). Additionally, scattered C5b-9 aggregates were observed (Supplemental Figure 3H).

Figure 3

Formation of drusen-like accumulation beneath MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by the addition of POS for an additional 4 weeks. (A–C) Immunostaining of cross-sections of NOR and MAL iPSC-RPE for apolipoprotein E (ApoE, red), phalloidin (green), and DAPI (blue). Arrows highlight intracellular punctate immunoreactivity, while arrowheads denote extracellular deposits beneath the cells. (B) Representative high-magnification image showing punctate ApoE staining in MAL iPSC-RPE cells. (C) Quantification of intracellular ApoE⁺ puncta-containing cells, intracellular ApoE (ApoE [intra]) fluorescence intensity, and basal ApoE fluorescence intensity in NOR and MAL iPSC-RPE cells. (D) En face image and z stack reconstructions of NOR and MAL iPSC-RPE stained for ApoE (red), phalloidin (green), and DAPI (blue). (E–J) Immunostaining of NOR and MAL iPSC-RPE cross-sections for fibulin-3 (red, E), complement C3 (C3, red, G), or collagen type IV (red, I), along with phalloidin (green) and DAPI (blue). (F, H, and J) Quantification of fibulin-3 (F) and collagen type IV (J); fluorescence intensity and number of C3⁺ puncta per section (H) in NOR and MAL iPSC-RPE cells. (K and L) Transmission electron microscopy images. (K) Vertical sections of the iPSC-RPE cells. (L) Horizontal section of the basal side (indicated by a white dotted line in the vertical section in K) of iPSC-RPE cells. Arrowheads indicate aggregated fibers. The squares indicate lipid granules. (M) Focused ion beam scanning electron microscopy analysis of drusen-like structures beneath MAL iPSC-RPE cells. Arrowheads indicate lipid granules. A color-coded figure is shown in Supplemental Figure 4. *P < 0.05, **P < 0.01, ***P < 0.005. P values in C, F, H, and J were calculated using a linear mixed-effects model followed by Tukey’s HSD test. n = 10. Scale bars: 20 μm (A, B, D, E, G, and I), 10 μm (M), and 2 μm (K and L).

TEM images of horizontal sections along the basal side — corresponding to the dotted line in the vertical section (Figure 3K) — revealed numerous nonuniform short fibers beneath NOR iPSC-RPE cells. Similar fibers were present in MAL iPSC-RPE cells; however, thick and elongated aggregated fibers were also detected (Figure 3L). Additionally, lipid granules were found beneath MAL iPSC-RPE cells (Figure 3L). To further analyze the drusen-like structures beneath MAL iPSC-RPE cells, we employed focused ion beam scanning electron microscope (FIB-SEM) tomography (Figure 3M and Supplemental Figure 4) and reconstructed a 3D image from serial FIB-SEM section images (Supplemental Videos 1 and 2). The reconstructed images revealed stacked degenerated cells, a widely distributed extracellular matrix on the basal side, and interspersed lipid granules.

These findings suggest that the whitish structures beneath the RPE cells consist of collagen and fibulin-3 fibers. Within these structures, lipid granules, ApoE, and complement components accumulate, resembling drusen observed in patients (4, 18).

Difference in lipid content between NOR and MAL iPSC-RPE. Because lipid accumulation is a major component of drusen and was evident ultrastructurally beneath MAL iPSC-RPE cells, we performed an untargeted lipid analysis to characterize intracellular lipid alterations. The analysis revealed significantly higher levels of hexosylceramide and bis-monoacylglycerophosphate (BMP) in MAL iPSC-RPE cells compared with those in NOR iPSC-RPE cells (Figure 4, A–D). Notably, when comparing NOR and MAL iPSC-RPE cells treated with POS, MAL iPSC-RPE cells exhibited a significant increase in BMP species containing 22:6-22:6 side chains (22 carbon atoms and 6 double bonds in each fatty acyl chain; DHA-DHA) (Figure 4C). Since POS is rich in long-chain fatty acids, particularly 22:6 (19), these findings suggest that phagocytosed POS components accumulate in MAL iPSC-RPE cells. Furthermore, immunostaining of iPSC-RPE cross-sections using an antilysobisphosphatidic acid (LBPA) antibody — LBPA being a specific configuration of BMP — revealed that LBPA formed aggregates and accumulated within MAL iPSC-RPE cells (Figure 4E). The increased expression of phospholipase D family member 3 (PLD3, Figure 4F), the enzyme responsible for BMP synthesis (20) — along with the decreased expression of α/β hydrolase domain containing 6 (ABHD6, Figure 4G), the enzyme responsible for BMP degradation (21) — suggests that these alterations contribute to elevated BMP levels in MAL iPSC-RPE cells.

Figure 4

Increased hexosylceramide and bis-monoacylglycerophosphate levels in MAL iPSC-RPE cells revealed by lipidomic analyses. iPSC-RPE cells were cultured for 4 weeks, followed by an additional 4 weeks of incubation in the presence or absence of POS. (A–D) Quantitative analysis of lipid species intensity and total lipid abundance using untargeted liquid chromatography–mass spectrometry and statistical analysis of NOR and MAL iPSC-RPE cells. (A and B) Hexosylceramide. (C and D) Bis-monoacylglycerophosphate (BMP). (E) Immunostaining of cross-sections of NOR and MAL iPSC-RPE cells treated with POS for lysobisphosphatidic acid (LBPA, red), phalloidin (green), and DAPI (blue). Scale bar: 20 μm. (F and G) Expression levels of phospholipase D family member 3 (PLD3, F) and α/β hydrolase domain containing 6 (ABHD6, G). The y axis represents expression levels in log₂ TPM (transcripts per million). (H) Comparison of intracellular free cholesterol levels. *P < 0.05, **P < 0.01, ***P < 0.005; 1-way ANOVA followed by the Tukey-Kramer test (A–D, F, and G), a linear mixed-effects model followed by Tukey’s HSD test in (H), n = 3. Data are shown as mean ± SD.

The elevated hexosylceramide levels observed in MAL iPSC-RPE cells suggest a possible connection to the pathophysiology of Gaucher disease, a lysosomal storage disorder characterized by the intracellular accumulation of glucosylceramide, a type of hexosylceramide (22). Given that α-synuclein aggregation is a known feature of Gaucher disease (22), we examined α-synuclein expression in MAL iPSC-RPE cells. While overall α-synuclein expression levels were comparable between NOR and MAL iPSC-RPE cells, immunofluorescence staining revealed α-synuclein aggregation specifically in MAL iPSC-RPE cells (Supplemental Figure 5, A–C). Additionally, we observed a downregulation of carboxylesterase 1 (CES1; Supplemental Figure 5D), an enzyme involved in lipid metabolism, which may impair lipid transport and contribute to intracellular lipid accumulation. However, no significant difference was observed in intracellular free cholesterol levels between NOR and MAL iPSC-RPE cells (Figure 4H).

Decrease and dysfunction of lysosomes in MAL iPSC-RPE. The accumulation of hexosylceramide, BMP, and α-synuclein — molecules normally degraded via lysosomal pathways — prompted us to investigate whether lysosomal biogenesis and function are impaired in MAL iPSC-RPE cells. To assess lysosomal involvement, we analyzed lysosome-related gene expression using RNA-seq. Clustering analysis revealed distinct gene expression patterns between NOR and MAL iPSC-RPE cells, classifying them into separate groups based on their respective conditions (Supplemental Figure 6). Further analysis of the highly expressed gene cluster, designated as “cluster A” in Supplemental Figure 6, revealed significantly lower expression of key lysosomal markers in MAL iPSC-RPE cells compared with NOR iPSC-RPE cells (Figure 5A). Specifically, the expression levels of lysosome-associated membrane protein-2 (LAMP2), cathepsin D (CTSD), and prosaposin (PSAP) were markedly reduced in MAL iPSC-RPE (Figure 5B). These findings suggest impaired lysosomal function in MAL iPSC-RPE, potentially contributing to disease pathology.

Figure 5

Reduced expression of lysosome-related proteins in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by an additional 4 weeks of incubation in the presence or absence of POS. RNA-seq analyses were performed on NOR1 and MAL1 in triplicate wells. (A and B) Cluster analysis comparing lysosome-associated gene expression profiles between NOR and MAL iPSC-RPE cells. The cluster showing relatively high expression levels (corresponding to the region highlighted as ‘Cluster A’ in Supplemental Figure 6) was extracted, and the expression levels of individual genes within this cluster are shown in B. Gene expression levels are presented as TPM (1 × 10³) on the y axis. *P < 0.05, **P < 0.01, ***P < 0.005, 1-way ANOVA followed by the Tukey-Kramer test; n = 3 in (B). GGA1, golgi associated, gamma adaptin ear containing, ARF binding protein 1; ARSG, arylsulfatase G; CTSL, cathepsin L; TPP1, tripeptidyl peptidase 1; ABCA2, ATP binding cassette subfamily A member 2; LIPA, lipase A, lysosomal acid type; IDS, iduronate 2-sulfatase; HEXB, hexosaminidase subunit beta; NAGA, α-N-acetylgalactosaminidase, CD68, CD68 molecule; LITAF, lipopolysaccharide-induced TNF factor; CD63, CD63 molecule; ACP2, acid phosphatase 2, lysosomal; GGA2, golgi associated, gamma adaptin ear containing, ARF binding protein 2; ACP5, acid phosphatase 5, tartrate resistant; AP1M1, adaptor related protein complex 1 subunit mu 1; GLA, galactosidase α; ASAH1, N-acylsphingosine amidohydrolase 1; NEU1, neuraminidase 1; CTSO, cathepsin O; CTSH, cathepsin H; CTSF, cathepsin F; CTSD, cathepsin D; LAMP2, lysosomal-associated membrane protein 2; PSAP, prosaposin; TPM, transcripts per million.

Western blot analysis of lysosome-related proteins (Figure 6, A and B) confirmed that CTSD expression in its double-chain forms was lower in MAL iPSC-RPE cells than in NOR iPSC-RPE cells. Similarly, LAMP2 expression was also tended to be reduced in MAL iPSC-RPE cells, although the difference did not reach statistical significance. These findings were further validated by immunostaining of iPSC-RPE cross-sections, which revealed a significant decrease in LAMP2 expression in MAL iPSC-RPE cells (Figure 6, C and D). Consistently, en face confocal imaging demonstrated a punctate intracellular distribution of LAMP2 in NOR iPSC-RPE cells, whereas this punctate lysosomal pattern was markedly reduced in MAL iPSC-RPE cells (Figure 6E). Additionally, TEM imaging shown in Figure 1D suggested a relative reduction in electron-dense vesicular structures consistent with lysosomes in MAL iPSC-RPE cells compared with NOR iPSC-RPE cells.

Figure 6

Decreased lysosomal biogenesis in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks. Subsequently, cells were incubated with or without POS for 4 weeks. (A and B) Western blot analysis of lysosome-associated markers — LAMP2, CTSD, Rab9, Rab7, and LC3 — in NOR and MAL iPSC-RPE cells. (C and D) Immunostaining of NOR and MAL iPSC-RPE cross-sections for LAMP2 (red), phalloidin (green), and DAPI (blue). (D) The quantification of LAMP2 expression is presented as fluorescence intensity per cell. (E) Immunostaining of NOR and MAL iPSC-RPE showing en face images and z stack reconstructions stained for LAMP2 (red), phalloidin (green), and DAPI (blue). (F and I–K) RNA-seq analysis of NOR and MAL iPSC-RPE cells showing expression of TFEB (F); MTOR, ATF4, and TP53 (I); CHMP4B (J), and VAMP8 (K). The y axis represents expression levels in log₂ TPM. (G and H) Immunostaining of NOR and MAL iPSC-RPE cells showing en face images and z stack reconstructions stained for TFEB (red), phalloidin (green), and DAPI (blue). (H) Quantification of nuclear-to-total TFEB fluorescence ratio in NOR and MAL iPSC-RPE cells. Each dot represents an individual cell (n = 43 cells in NOR, 38 cells in MAL). Scale bar: 20 μm (C, E, and G). *P < 0.05, **P < 0.01, ***P < 0.005; 1-way ANOVA followed by the Tukey-Kramer test (F and I–K), a linear mixed-effects model followed by Tukey’s HSD test (B and D), Welch’s t test (H), n = 5 (B), n = 10 (D), n = 3 (F and I–K). Data are shown as mean ± SD. Rab9, Ras-related protein Rab-9; Rab7, Ras-related protein Rab-7; LC3, microtubule-associated protein 1 light chain 3 α; TFEB, transcription factor EB; MTOR, mechanistic target of rapamycin kinase; ATF4, activating transcription factor 4; TP53, tumor protein p53; TPM, transcripts per million; CHMP4B, charged multivesicular body protein 4B; VAMP8, vesicle-associated membrane protein 8.

Transcription factor EB (TFEB), a master regulator of lysosomal gene expression and homeostasis (23), showed reduced transcript levels in MAL iPSC-RPE cells compared with NOR iPSC-RPE cells in our RNA-seq dataset (Figure 6F). Given that TFEB is regulated by nutrient-sensing and stress-response pathways, we assessed the transcript levels of mechanistic target of rapamycin kinase (MTOR) and activating transcription factor 4 (ATF4), the latter being a key mediator of the integrated stress response and endoplasmic reticulum stress. RNA-seq analysis revealed no apparent differences in MTOR or ATF4 transcripts, whereas tumor protein p53 (TP53) transcripts were significantly upregulated in MAL iPSC-RPE (Figure 6I). At the protein level, however, neither total TFEB abundance nor TFEB phosphorylation at Ser211 — a site associated with cytoplasmic retention — showed a significant difference between NOR and MAL iPSC-RPE cells, as assessed by Western blotting (Supplemental Figure 7, A and B). Similarly, Western blot analysis of total mTOR and phosphorylated mTOR (p-mTOR) revealed no significant differences between NOR and MAL iPSC-RPE cells (Supplemental Figure 7, A and B). Consistently, confocal z-stack imaging revealed comparable gross nuclear and cytoplasmic TFEB immunoreactivity between the 2 groups (Figure 6G). However, quantitative analysis of TFEB immunofluorescence revealed a significant increase in the nuclear-to-total TFEB fluorescence ratio in MAL iPSC-RPE cells (Figure 6H), indicating relative enrichment of TFEB in the nucleus despite unchanged total protein levels. Together, these findings indicate that the reduced lysosomal biogenesis observed in MAL iPSC-RPE cells cannot be explained solely by insufficient TFEB nuclear localization, total TFEB abundance, Ser211 phosphorylation, or mTOR signaling. Instead, they suggest that TFEB-dependent lysosomal gene programs may be functionally impaired downstream of, or uncoupled from, TFEB nuclear localization — potentially reflecting maladaptive or incomplete stress-induced TFEB activation under chronic disease conditions. Additionally, the expression of charged multivesicular body protein 4B (CHMP4B), a marker of lysosomal membrane integrity (24), showed no significant difference between NOR and MAL iPSC-RPE cells (Figure 6J). This suggests that lysosomal membrane integrity is grossly preserved in MAL iPSC-RPE cells, despite reduced lysosomal abundance and impaired lysosomal function.

The expression level of Ras-related protein Rab-9 (Rab9), a key late endosome marker (25), was significantly reduced in MAL iPSC-RPE cells compared with NOR iPSC-RPE cells (Figure 6, A and B). Similarly, Rab7, another late endosome marker (25), exhibited a downward trend in MAL iPSC-RPE cells (Figure 6, A and B). Additionally, the gene expression of vesicle-associated membrane protein 8 (VAMP8), a soluble NSF-attachment protein receptor protein that localizes to late endosomes and lysosomes and facilitates vesicle fusion with the endoplasmic reticulum and cell membranes (26), was reduced in MAL iPSC-RPE cells (Figure 6K). In contrast, the expression of microtubule-associated protein 1 light chain 3 α (LC3), an autophagy marker, remained comparable between NOR and MAL iPSC-RPE cells (Figure 6, A and B). These findings suggest that, while autophagy remains unaffected, the fusion process between late endosomes and lysosomes is impaired, likely due to the reduced expression of critical lysosome-associated proteins.

Because lysosomes play a central role in the processing of phagocytosed POS, we next directly assessed POS handling in iPSC-RPE cells. Quantitative analysis of rhodopsin levels in the culture supernatant demonstrated no significant difference in POS uptake between NOR and MAL iPSC-RPE cells, indicating that the initial phagocytic capacity is preserved in MAL iPSC-RPE cells (Supplemental Figure 8, A and B). We then evaluated intracellular degradation of phagocytosed POS by measuring cellular rhodopsin levels at defined time points after POS withdrawal. In NOR iPSC-RPE cells, intracellular rhodopsin levels rapidly decreased following POS removal, indicating efficient degradation of phagocytosed POS. In contrast, MAL iPSC-RPE cells exhibited a smaller and more variable reduction in intracellular rhodopsin levels, particularly at early time points. Although these differences did not reach statistical significance, this pattern suggests a tendency toward delayed intracellular degradation of phagocytosed POS under prolonged phagocytic stress in MAL iPSC-RPE cells (Supplemental Figure 8, C and D).

To further assess the lysosomal physiology underlying the altered degradation capacity, we evaluated lysosomal acidity using LysoTracker staining followed by flow cytometric analysis. In NOR iPSC-RPE cells, LysoTracker fluorescence intensity remained comparable between 2-week and 4-week POS exposures. In contrast, MAL iPSC-RPE cells exhibited a significant reduction in LysoTracker fluorescence after 4 weeks of POS treatment, consistent with lysosomal alkalinization (Figure 7A). These findings suggest that prolonged phagocytic stress leads to impaired lysosomal acidification in MAL iPSC-RPE cells, which may contribute to delayed intracellular degradation of phagocytosed POS.

Figure 7

Lysosomal dysfunction in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks. Subsequently, cells were incubated with or without POS for 2 or 4 weeks. (A) FACS analysis of lysosome function using LysoTracker Green. The mean fluorescence intensity at 4 weeks after POS treatment was normalized to the mean fluorescence intensity at 2 weeks of POS exposure. (B and C) Western blot and statistical analyses of phosphorylated epidermal growth factor receptor (pEGFR) in NOR and MAL iPSC-RPE cells treated with epidermal growth factor (EGF) for 6 hours, followed by measurements at 0, 15, 30, and 60 minutes after EGF removal. *P < 0.05, ***P < 0.005; P values were calculated using a linear mixed-effects model followed by Tukey’s HSD test. In C, comparisons were made versus 0 min. n = 5 (A) and n = 3 (C). Data are shown as mean ± SD.

Next, we evaluated lysosomal function in iPSC-RPE cells. Upon stimulation with epidermal growth factor (EGF), EGF receptors on the cell membrane undergo phosphorylation and are subsequently internalized and degraded by lysosomes (27). Using this mechanism, we assessed lysosomal function in MAL iPSC-RPE cells by measuring the degradation rate of phosphorylated EGFR (pEGFR) following EGF stimulation. The degradation of pEGFR was significantly slower in MAL iPSC-RPE cells than in NOR iPSC-RPE cells (Figure 7B). Notably, in MAL iPSC-RPE cells cultured with POS for 2 weeks, pEGFR degradation was observed within 60 minutes after EGF stimulation. However, in cells subjected to prolonged POS exposure (a total of 4 weeks), pEGFR degradation was markedly impaired, with no significant reduction even after 60 minutes (Figure 7, B and C). Thus, these findings indicate that lysosomal function is inherently compromised in MAL iPSC-RPE cells and that extended POS phagocytosis under near-physiological conditions exacerbates this dysfunction.

Effect of trehalose treatment to relieve lysosomal decrease and dysfunction. Because lysosomal reduction and dysfunction were observed in MAL iPSC-RPE cells, we next tested whether trehalose, a lysosome/autophagy-modulating compound reported to enhance lysosomal gene programs including TFEB-dependent pathways (23, 28), could restore lysosomal biogenesis and function. To assess whether trehalose affects TFEB activity at the level of subcellular localization, we performed immunofluorescence analysis and quantified the nuclear-to-total TFEB fluorescence intensity ratio. Trehalose treatment significantly increased the nuclear TFEB fraction in both NOR and MAL iPSC-RPE cells (Figure 8, A and B, and Supplemental Figure 9A), indicating enhanced TFEB nuclear localization under trehalose-treated conditions. Notably, total TFEB and pTFEB protein levels were not increased at the experimental endpoint, as assessed by Western blotting (Supplemental Figure 9, B and C). Consistent with this change, Western blot analysis demonstrated that trehalose treatment increased the levels of lysosome-associated proteins, LAMP2 and mature CTSD, both of which were previously reduced in MAL iPSC-RPE cells (Figure 8, C and D). Immunostaining of iPSC-RPE cells further confirmed a significant increase in punctate LAMP2 immunoreactivity following trehalose treatment (Figure 8, E–G), along with increased punctate immunoreactivity for the mannose-6-phosphate receptor (M6PR), a marker of late endosome–lysosome trafficking (29) (Figure 8H). Additionally, trehalose treatment led to an increasing trend in the expression of late endosome-related proteins Rab7 and Rab9, as well as the autophagy-related protein LC3 (Figure 8, C and D).

Figure 8

Improvement of lysosomal function by trehalose in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by the addition of POS and treatment with or without 100 mM trehalose (tre) for an additional 4 weeks. (A and B) Immunostaining of MAL iPSC-RPE cells treated with trehalose showing z stack reconstructions stained for transcription factor EB (TFEB, red), phalloidin (green), and DAPI (blue). (B) Quantification of the nuclear-to-total TFEB fluorescence intensity ratio in NOR, NOR + trehalose, MAL, and MAL + trehalose iPSC-RPE cells. Sample sizes: NOR, n = 43 cells; NOR + trehalose, n = 41; MAL, n = 38; MAL + trehalose, n = 20. (C and D) Western blot analysis and statistical analyses of LAMP2, CTSD, Rab7, Rab9, and LC3 expression in NOR and MAL iPSC-RPE cells. (E and H) Immunostaining of NOR and MAL iPSC-RPE cross-sections for LAMP2 (red), phalloidin (green), and DAPI (blue, E), as well as separate cross-sectional images stained for mannose-6-phosphate receptor (M6PR) (red), phalloidin (green), and DAPI (blue, H). (F) The quantification of LAMP2 expression is presented as fluorescence intensity per cell. (G) En face image and z stack reconstructions of MAL iPSC-RPE cells treated with trehalose stained for LAMP2 (red), phalloidin (green), and DAPI (blue). (I and J) Western blot analysis of pEGFR in MAL iPSC-RPE cells treated with EGF and assessed at 0, 15, 30, and 60 minutes after EGF removal. *P < 0.05, ***P < 0.005, 1-way ANOVA followed by the Tukey-Kramer test (B), a linear mixed-effects model followed by Tukey’s HSD test (D, F, and J). In J, comparisons were made versus 0 min, n = 5 (D), n = 10 (F), n = 3 (J). Data are shown as mean ± SD. (-), not treated with trehalose; tre, treated with trehalose. Scale bar: 20 μm (A, E, G, and H).

To evaluate whether trehalose improved lysosomal function, we assessed pEGFR degradation following EGF stimulation. In trehalose-treated MAL iPSC-RPE cells, pEGFR was significantly degraded within 60 minutes after stimulation (Figure 8, I and J), suggesting a functional restoration of lysosomal activity. These findings indicate that trehalose treatment enhances both lysosomal quantity and function, offering potential therapeutic benefits for mitigating lysosomal dysfunction in MAL iPSC-RPE cells.

Suppression of drusen-like deposit formation by trehalose treatment. To determine whether restoration of lysosomal function translates into improvement of MAL-associated pathological phenotypes, we examined the effects of trehalose on drusen-like deposit formation, extracellular matrix integrity, and cell survival. In long-term cultures, MAL iPSC-RPE cells exhibited morphological degeneration, including cell flattening and excessive pigment deposition. However, trehalose treatment effectively prevented these changes, maintaining normal cellular morphology (Figure 9, A and B, and Supplemental Figure 10, A–C). Immunostaining for pERM showed no appreciable change in its apical localization pattern following trehalose treatment (Figure 9C). Additionally, trehalose significantly reduced the number of TUNEL⁺ apoptotic cells, restoring them to levels comparable to those in NOR iPSC-RPE cells (Figure 9, D and E).

Figure 9

Rescue of cellular morphology and suppression of apoptosis by trehalose in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by the addition of POS and treatment with or without 100 mM trehalose (tre) for an additional 4 weeks. (A) Close-up view by bright-field micrographs of NOR and MAL iPSC-RPE cells. (B) Transmission electron microscope images of vertical sections of iPSC-RPE cells. (C) Immunostaining of NOR and MAL iPSC-RPE cross-sections for pERM (red), phalloidin (green), and DAPI (blue). (D and E) TUNEL staining (green) of cross-sections of NOR and MAL iPSC-RPE cells. Scale bar: 200 μm (A), 2 μm (B), 20 μm (C and D). *P < 0.05, a linear mixed-effects model followed by Tukey’s HSD test, n = 10 (E). Data are shown as mean ± SD. (-), not treated with trehalose; tre, treated with trehalose.

While overall fibulin-3 expression levels remained unchanged (Figure 10, A and B), trehalose treatment altered the distribution pattern of fibulin-3 immunoreactivity, with a reduction in dense fibulin-3⁺ structures (Figure 10, C and D). Likewise, α-synuclein aggregation was diminished following treatment (Supplemental Figure 10, D–F). Trehalose treatment also altered the intracellular distribution of ApoE, reducing the intensity and aggregation of enlarged intracellular ApoE⁺ puncta (Figure 10, E–G), while total ApoE protein levels were maintained or increased (Figure 10, A and B). In parallel, ApoE immunoreactivity became more prominent as fine signals in the membrane-adjacent region, consistent with redistribution rather than enhanced intracellular aggregation. Similarly, C3 aggregates, which are typically abundant in MAL iPSC-RPE cells, were significantly reduced (Figure 10, H and I). Moreover, trehalose treatment disrupted the fibrotic accumulation of collagen IV (Figure 10, J and K) and led to a reduction in MMP2 expression (Figure 10, L and M), suggesting improved extracellular matrix integrity. Collectively, these results suggest that trehalose restores lysosomal function in MAL iPSC-RPE cells, which may help inhibit drusen-like deposit formation, reinforce the extracellular matrix and prevent cell death, highlighting its potential therapeutic benefits.

Figure 10

Improvement of drusen-like deposits and extracellular matrix dysregulation by trehalose in MAL iPSC-RPE cells. iPSC-RPE cells were cultured for 4 weeks, followed by the addition of POS and treatment with or without 100 mM trehalose (tre) for an additional 4 weeks. (A and B) Western blot analysis of fibulin-3, ApoE, and collagen type IV in NOR and MAL iPSC-RPE cells. (C–F and H–K) Immunostaining of cross-sections of MAL iPSC-RPE cells with or without trehalose treatment for fibulin-3 (red, C), ApoE (red, E), C3 (red, H), or collagen IV (red, J), together with phalloidin (green) and DAPI (blue). (G) Immunostaining of MAL iPSC-RPE cells treated with trehalose showing en face images and z stack reconstructions stained for LAMP2 (red), phalloidin (green), and DAPI (blue). (D, F, I, and K) Quantification of fibulin-3 (D) and collagen type IV (K) fluorescence intensity; intracellular ApoE metrics, including the proportion of cells with intracellular ApoE⁺ puncta-containing cells, intracellular ApoE [ApoE (intra)] fluorescence intensity, and basal ApoE fluorescence intensity in NOR and MAL iPSC-RPE cells treated with or without trehalose (F); or C3⁺ puncta (I), comparing NOR, MAL, and trehalose-treated MAL iPSC-RPE cells. Student’s t test; n = 10. (L and M) Western blot analysis of MMP2 in NOR and MAL iPSC-RPE cells. Scale bar: 20 μm (C, E, G, H, and J). *P < 0.05, **P < 0.01, ***P < 0.005, a linear mixed-effects model followed by Tukey’s HSD test, n = 5 (B and M), n = 10 (D, F, I, and K). Data are shown as mean ± SD. (-), not treated with trehalose; tre, treated with trehalose.

Given previous reports implicating intracellular lipid dysregulation in MAL (14) and our prior demonstration that lipid extraction with 2-hydroxypropyl-β-cyclodextrin (HPBCD) improves lysosomal function in iPSC-RPE cells derived from patients with Bietti crystalline dystrophy (30), we also evaluated HPBCD treatment as a comparative approach in MAL iPSC-RPE cells. However, treatment with HPBCD failed to restore lysosomal enzyme expression. Additionally, HPBCD treatment did not prevent drusen-like deposit formation, extracellular matrix weakening, or apoptosis in MAL iPSC-RPE cells (Supplemental Figure 11).

In this study, we demonstrated that iPSC-derived RPE cells from patients with MAL, a hereditary macular degeneration, exhibit significant qualitative and quantitative lysosomal impairment. These defects led to reduced intracellular degradation, resulting in the accumulation of undigested materials, drusen-like deposits, and increased cytotoxicity, including apoptosis. Additionally, extracellular matrix degradation was exacerbated by MMP2 activation. Importantly, treatment with a lysosome-activating agent alleviated these pathological changes by reducing drusen-like deposits, mitigating cytotoxicity, and preserving extracellular matrix integrity.

We confirmed the presence of drusen-like deposits beneath MAL iPSC-RPE cells, consistent with prior reports (31). In vivo, RPE cells are postmitotic and continuously phagocytose large amounts of POS, imposing a substantial lysosomal and lipid metabolic burden. Although POS exposure did not globally alter gene expression, it acted as a physiologically relevant stressor that unmasked disease-specific defects in lipid handling and lysosomal function in MAL iPSC-RPE cells. In other words, POS primarily functions as a physiological stressor that challenges lysosomal and metabolic capacity rather than triggering broad transcriptional remodeling. These findings indicate that MAL iPSC-RPE cells are particularly vulnerable to POS-induced metabolic burden, which tends to enhance drusen-like deposition and reveal underlying lysosomal dysfunction. Given the critical role of RPE cells in lipid processing, their dysfunction can lead to the accumulation of waste products such as drusen, a hallmark of early-stage AMD.

To the best of our knowledge, our study is the first to demonstrate that lysosomal reduction and dysfunction are critical contributors to MAL pathophysiology. We found that lysosomal dysfunction disrupts intracellular digestion of lipids and other cellular components, leading to the extrusion of partially digested waste products into the extracellular space and excessive drusen deposition in the macular region. This dysfunction was associated with altered regulation of lysosomal gene programs, including reduced transcript levels of TFEB, a key regulator of lysosomal biogenesis. While TFEB is influenced by multiple pathways — including the mTOR pathway (32), the endoplasmic reticulum stress pathway (33), and the p53 pathway (34) — RNA-seq analysis revealed that TP53 transcripts were significantly elevated in MAL iPSC-RPE cells, whereas the involvement of other pathways appeared to be minimal. Given that p53 signaling has been reported to suppress TFEB transcription and lysosomal gene programs under conditions of chronic cellular stress (34), elevated p53 expression may contribute to dysregulation of lysosomal gene programs in MAL. Importantly, although total TFEB protein abundance did not differ between NOR and MAL iPSC-RPE cells under basal conditions and was not increased at the experimental endpoint, trehalose treatment was associated with enhanced nuclear enrichment of TFEB. These findings suggest that trehalose may modulate TFEB activity through dynamic regulation of subcellular localization rather than through sustained increases in total TFEB protein levels. Notably, although previous studies have associated MAL with intracellular free cholesterol accumulation due to carboxylesterase 1 downregulation (14), we did not detect significant cholesterol buildup in MAL iPSC-RPE cells. Additionally, while cyclodextrin has been shown to reduce intracellular cholesterol and improve lysosomal function in Bietti crystalline dystrophy (30), it failed to restore lysosomal enzyme expression, suppress drusen-like deposit formation, or prevent apoptosis in MAL iPSC-RPE cells. These findings strongly suggest that lysosomal dysfunction in MAL is not primarily driven by lipid accumulation.

Based on these findings, we propose a working model in which intracellular accumulation of mutant fibulin-3 induces chronic cellular stress in MAL iPSC-RPE cells. This sustained stress state is associated with dysregulation of lysosomal gene programs, including those normally governed by TFEB, resulting in reduced lysosomal content and impaired degradation of lipids and proteins. Importantly, our data indicate that this lysosomal impairment cannot be explained solely by static differences in TFEB protein abundance or phosphorylation at the examined time point, suggesting the involvement of additional regulatory layers affecting TFEB transcriptional output and/or lysosomal function. As a consequence, partially processed cellular components, including lipid species and ApoE-associated material, accumulate intracellularly and are aberrantly deposited in the subcellular space beneath the RPE cells, promoting drusen-like deposit formation. In parallel, impaired lysosomal and autophagic clearance exacerbates cellular toxicity and apoptosis, while extracellular matrix remodeling through MMP2 activation further contributes to disease progression. Lysosomal activation by trehalose appears to interrupt this pathogenic cascade by restoring lysosomal homeostasis and improving intracellular proteostasis and lipid handling, potentially through dynamic regulation of TFEB nuclear localization and/or additional TFEB-independent mechanisms, even in the absence of sustained increases in total TFEB protein levels at the experimental endpoint.

In addition to lysosomal impairment, altered processing of ApoE appears to be a key downstream consequence of lysosomal dysfunction in MAL iPSC-RPE cells. Under long-term POS exposure, MAL iPSC-RPE cells exhibited abnormal intracellular ApoE⁺ puncta and ApoE accumulation in the subcellular space between the cells and the culture membrane, whereas NOR iPSC-RPE cells showed predominantly fine punctate ApoE signals within the membrane, consistent with physiological secretion. These findings suggest that lysosomal dysfunction in MAL disrupts normal ApoE processing and trafficking, leading to pathological intracellular retention and mislocalization rather than impaired ApoE production per se. Notably, trehalose treatment did not alter total ApoE protein levels but reduced intracellular ApoE signal intensity. Although the proportion of cells harboring intracellular ApoE⁺ puncta remained unchanged, these findings indicate improved intracellular handling and secretion of ApoE. The fine ApoE signals observed within the membrane likely reflect enhanced basal secretion from RPE cells, whereas the reduction of intracellular and subcellular ApoE deposits suggests alleviation of pathological accumulation. Together, these results support the notion that lysosomal activation restores ApoE proteostasis and prevents aberrant ApoE deposition, thereby contributing to the suppression of drusen-like deposit formation in MAL. Consistent with this interpretation, lysosomal dysfunction has also been implicated in the pathophysiology of AMD (35, 36). This study demonstrated that lysosomal reduction and dysfunction contribute to drusen-like deposit formation, whereas lysosomal activation suppresses its accumulation. These findings provide valuable insights into potential therapeutic strategies for AMD and other drusen-associated diseases, including MAL.

A further downstream consequence of lysosomal dysfunction was the elevation of BMP levels in MAL iPSC-RPE cells. Recent studies indicate that BMP synthesis and stabilization are regulated by PLD3 and phospholipase D family member 4 (PLD4) (20). While PLD4 expression was barely detectable in iPSC-RPE (data not shown), its expression was increased in MAL iPSC-RPE cells. Moreover, zinc finger protein 212, a known regulator of PLD3, was also upregulated (data not shown). Because endoplasmic reticulum stress induces PLD3 expression (37), this mechanism may underlie BMP stabilization in MAL. In contrast, ABHD6, a BMP degradation enzyme (21), was downregulated, suggesting reduced BMP degradation. These findings indicate that although lysosomes and late endosomes were reduced in MAL iPSC-RPE cells, BMP levels remained higher than those in NOR iPSC-RPE cells. Overall, the relationship between increased BMP levels and lysosomal dysfunction requires further investigation.

Since RPE cells are postmitotic, autophagy plays a crucial role in intracellular clearance. When lysosomal dysfunction impairs autophagic degradation, as observed in MAL, it not only promotes drusen formation but also directly damages RPE cells. Furthermore, MAL is associated with macular neovascularization, and the observed MMP2 activation and extracellular matrix weakening resemble features of Bruch membrane degradation in patients, contributing to a permissive environment for neovascularization. Together, these converging defects in lysosomal function, lipid homeostasis, autophagic clearance, and extracellular matrix integrity provide a mechanistic framework linking intracellular stress to the complex pathological features of MAL.

This study has a few limitations. First, while lysosomal reduction and dysfunction were identified as key contributors to MAL pathophysiology, the precise temporal sequence and molecular interactions linking EFEMP1 mutations, chronic cellular stress, lysosomal gene dysregulation, and extracellular matrix alterations could not be fully resolved. Static endpoint analyses may not capture the dynamic regulation of TFEB activity or TFEB-independent mechanisms affecting lysosomal function. Second, although using isogenic controls with correction of the EFEMP1 mutation would provide the most rigorous assessment of mutation-specific effects, generating and fully characterizing such lines was beyond the scope of this study. To mitigate this limitation, we analyzed 2 independently derived iPSC-RPE lines from the same patient with MAL and confirmed consistent disease-associated phenotypes across both lines. In addition, whole-exome sequencing of the patient did not identify other known pathogenic variants associated with inherited retinal degenerations, reducing the likelihood that the observed phenotypes are driven by additional genetic factors. Third, given the rarity of MAL, establishing and analyzing iPSC-derived RPE cells from multiple patients was not feasible. As a result, analyses were conducted using iPSCs from a single patient. Nevertheless, the consistency of observed phenotypes across independent patient-derived lines supports the robustness and reliability of these findings. Finally, while this study utilized RPE cells differentiated from patient-derived iPSCs and successfully recapitulated several key pathological features of MAL, this in vitro model does not fully capture the complex interplay of systemic factors and tissue-specific microenvironments present in vivo. Future studies incorporating EFEMP1-corrected isogenic lines, additional patient-derived iPSC-RPE cells, and in vivo models will be crucial for further validation.

In summary, this study successfully replicated drusen formation, a hallmark of MAL, and elucidated its underlying pathology and mechanisms using patient-derived iPSC-RPE cells. Findings revealed cell damage, likely responsible for retinal and choroidal degeneration, as well as MMP2 activation, which may contribute to neovascularization. These pathological changes appear to be driven by lysosomal reduction and dysfunction. Importantly, this study underscores the pivotal role of lysosomal impairment in MAL pathophysiology, providing a foundation for future therapeutic interventions. Notably, lysosomal activation alleviated key pathological features, including drusen-like deposit accumulation, extracellular matrix degeneration, and cell death, highlighting the potential of targeting lysosomal pathways as a therapeutic strategy for MAL and related macular degenerative diseases, including AMD. Further investigation into the mechanisms underlying lysosomal impairment will facilitate the development of novel, mechanism-based treatments for vision-threatening conditions.

Sex as a biological variable. The iPSC line used in this study was established from a male patient. Because the experiments were conducted using iPSC-derived RPE cells generated from a single patient-derived line, sex was not analyzed as a biological variable. MAL is not known to exhibit strong sex-specific differences in its underlying molecular pathology. Therefore, the mechanisms identified in this study are expected to be broadly relevant to both sexes.

Establishment and culture of iPSCs and differentiation of iPSC-RPE cells. Skin cells were collected from 2 healthy individuals and one patient with MAL carrying an EFEMP1 mutation (R345W). These cells were reprogrammed into iPSCs using episomal vectors containing reprogramming factors (pCE-hOCT3/4, pCE-mp53DD, pCE-hSK, pCE-hUL, and pCXB-EBNA1) (38). We tested 2 different strains from a patient with MAL (MAL1 [16001MAL 13] and MAL2 [16001MAL 14]) and 2 strains from 2 healthy individuals (NOR1 [14100NOR(α) 4f1] and NOR2 [14101NOR(α) 9]). The maintenance of human iPSCs and their differentiation into RPE cells have been described previously (39). Briefly, iPSCs were differentiated into RPE cells through a stepwise process. This process utilized a KO serum replacement differentiation medium with gradual reductions in serum concentration, followed by floating culture and purification to obtain a uniform RPE population. iPSC-RPE cells were cultured in clinical trial specification CELLstart-coated (Gibco, A10142-01) dishes using a preconfluent medium composed of a 1:1 mixture of RPE maintenance medium and RPE differentiation medium (30, 40). The RPE maintenance medium was prepared by mixing Dulbecco’s Modified Eagle Medium (DMEM; MilliporeSigma, D6046) and F12HAM (MilliporeSigma, N6658) in a 7:3 ratio, supplemented with B-27 (Gibco, 17504-044) and 2 mM L-glutamine (Nacalai, 16948-04). The RPE differentiation medium consisted of DMEM/F12 (Gibco, 10565-018) supplemented with 10% fetal bovine serum. The culture medium was refreshed every 2–3 days. After completion of differentiation, iPSC-RPE cells were replated onto experimental substrates and cultured for an additional 4 weeks to allow maturation before subsequent experiments. For the analyses shown in Figure 1, iPSC-RPE cells were examined at this maturation stage to confirm RPE differentiation and identity.

RPE cells are highly specialized, postmitotic cells that phagocytose large amounts of POS daily in vivo and play a central role in lipid metabolism and waste clearance. While POS exposure is not required for RPE differentiation or maintenance of basic RPE identity, POS phagocytosis imposes a substantial lysosomal and metabolic burden on RPE cells and is therefore critical for modeling disease-relevant stress conditions. Accordingly, in this study, POS were omitted in experiments aimed at confirming RPE differentiation, identity, and baseline cellular properties, whereas POS were introduced in assays designed to assess cytotoxicity, drusen-like deposit formation, lysosomal function, and lipid metabolism to better recapitulate physiological conditions in vivo. For most subsequent experiments, matured iPSC-RPE cells were further cultured for an additional 4 weeks with or without bovine POS (purchased from InVision BioResources; 2 μg/well in a 24-well plate), as indicated. The duration of POS treatment (2 or 4 weeks) and the presence or absence of POS were varied depending on the experimental purpose, particularly for assays assessing lysosomal function, and are specified in the corresponding figure legends.

For the treatment of MAL iPSC-RPE cells, trehalose (MilliporeSigma, T9449) was administered at a concentration of 100 mM for 4 weeks.

Immunocytochemistry. iPSC-RPE cells were seeded onto Lab-Tek II Chamber Slides with CC2-coated glass coverslips (Thermo Fisher Scientific, 154941) and cultured for 4 weeks. For immunostaining of cross-sections, cells were seeded onto Millicell cell culture inserts (MilliporeSigma, PIHA01250) and cultured for 4 weeks. When indicated, bovine POS was subsequently added, and cultures were maintained for an additional 4 weeks. The presence or absence of POS and the duration of POS exposure varied depending on the experimental purpose and are specified in the corresponding figure legends. Following fixation with 4% paraformaldehyde for 15 minutes, the cells with membranes were embedded in Tissue-Tek OCT compound 4583 (Sakura, 4583), frozen, and sectioned at a thickness of 14 μm.

Immunostaining was performed as previously described (30, 40). The primary antibodies used were as follows: rabbit anti–ZO-1 (1:200; Invitrogen, 61-7300), rabbit anti–claudin-16 (1:250; Proteintech, 82954-1-RR), rabbit anti-pERM (1:200; Cell Signaling, 3726), mouse anti–fibulin-3 (1:250; Santa Cruz Biotechnology, sc-33722), rabbit anti-HSPA5 (1 μg/mL; Abcam, ab21685), rabbit anti-ApoE (1:500; Abcam, ab183597), rabbit anti–complement C3 (1:500; Novus, NBPI-32080), rabbit anti–C5b-9 (1:200; Abcam, ab55811), rabbit anti–collagen IV (1:500; Abcam, ab6586), mouse anti-LBPA (1:500; Echelon Bioscience, Z-PLBPA-2), mouse anti-LAMP2 (1:250; Abcam, ab25631), rabbit anti-TFEB (1:1000; Proteintech, 13372-1-AP), and rabbit anti-M6PR (1 μg/mL; Abcam, ab124767). The secondary antibodies used included Alexa Fluor 594 anti–rabbit IgG (1:1000; Invitrogen, A21207), Alexa Fluor 594 anti–mouse IgG (1:1000; Invitrogen, A21203), and Alexa Fluor 488 anti–mouse IgG (1:1000; Invitrogen, A21202 antibodies. Additionally, Alexa Fluor 488 phalloidin (1:250; Invitrogen, A12379) or Alexa Fluor 594 phalloidin (1:400; Invitrogen, A12381) was used to stain actin filaments, while cell nuclei were counterstained with DAPI (1 μg/mL). Cells were then mounted using FluorSave Reagent (MilliporeSigma, 345789) and imaged using a laser-scanning confocal microscope (Leica TCS SP8) and a fluorescence microscope (KEYENCE BZ-9000).

Detection of apoptosis. Apoptosis was detected using the MEBSTAIN Apoptosis TUNEL Kit Direct (MBL, 8445), and cell nuclei were counterstained with DAPI (1 μg/mL) (Molecular Probes). The cells were then mounted with FluorSave Reagent (MilliporeSigma, 345789) and imaged using a laser-scanning confocal microscope (Leica TCS SP8). Images were captured at 10 different locations, and the percentage of apoptotic cells was calculated by dividing the number of TUNEL⁺ cells within a 290 μm–wide region by the total number of cells.

TEM and FIB-SEM tomography. iPSC-RPE cells were seeded onto Lab-Tek Chamber slides without a Permanox Slide cover (Thermo Fisher Scientific, 177445). The cultured iPSC-RPE cells were washed with phosphate-buffered saline and fixed in a solution containing 4% formaldehyde and 2% glutaraldehyde.

For the TEM assay, sections were exposed to 1% osmium tetroxide in 100 mM PB (pH 7.4) for 1 hour at 4°C, followed by dehydration in a graded ethanol series. The cells were then embedded in Epon 812 resin (Nacalai Tesque, 20829-05) and polymerized at 60°C for 3 days. Ultrathin sections were obtained using an ultramicrotome and stained with uranyl acetate and lead citrate before examination under a TEM (model H-7650; Hitachi Co.).

For focused ion beam scanning electron microscopy tomography, sections were incubated with 1.5% potassium ferrocyanide, followed by 2% osmium tetroxide at 4°C for 2 hours. Next, they were washed with distilled water and fixed with 2% osmium tetroxide at room temperature for 1 hour. Fixed specimens were dehydrated using a graded ethanol and propylene oxide series and were embedded in Epon 812. Imaging was performed using a Crossbeam540 (Zeiss) with Smart FIB software (Zeiss). Image stacks were processed and edited using Dragonfly 2022 software (ORS Inc.).

Western blot analysis. iPSC-RPE cells were lysed in radioimmunoprecipitation buffer, and protein samples (5 μg) were subjected to Western blot analysis as previously described (30, 40). The primary antibodies used were as follows: mouse anti-RPE65 (1:10000; Santa Cruz Biotechnology, sc-390787), mouse anti-BEST1 (1:1000; Abcam, ab2182), mouse anti-CRALBP (1:1000; Abcam, ab15051), mouse anti–fibulin-3 (1:1000; Santa Cruz Biotechnology, sc-33722), rabbit anti–α-synuclein (1:1000; Abcam, ab138501), rabbit anti–cathepsin D (1:5000; Abcam, ab75852), mouse anti-LAMP2 (1:1000; Abcam, ab25631), rabbit anti–caspase-3 (1:1000; Abcam, ab2302), rabbit anti-MMP2 (1:1000; Abcam, ab92536), rabbit anti-TIMP1 (1:1000; Abcam, ab211926), rabbit anti-TIMP3 (1 μg/mL; Abcam, ab85926), rabbit anti–collagen IV (1:2500; Abcam, ab6586), rabbit anti-Rab7 (1:5000; Abcam, ab137029), rabbit anti–phospho-EGF receptor (1:1000; Cell Signaling, 3777), rabbit anti-Rab9 (1:5000; Abcam, ab179815), rabbit anti-LC3 (0.5 μg/mL; Abcam, ab48394), rabbit anti-ApoE (1:2000; Abcam, ab183597), and mouse anti–β-actin (1:5000; Sigma, A5441). Anti-mouse (NA931) and anti-rabbit (NA934) secondary antibodies were obtained from Cytiva and used at a 1:5,000 dilution.

RNA-seq and pathway analysis. Analyses were conducted using NOR1 and MAL1. Total RNA was extracted using the RNeasy Mini Plus Kit (Qiagen, 74104) and treated with deoxyribonuclease I (DNase 1, QIAGEN, 79254) to eliminate genomic DNA contamination. Sequencing was performed on an Illumina NovaSeq 6000 platform in 150-base paired-end mode. Raw read data were mapped to the human reference genome using the DRAGEN Bio-IT Platform (version 3.5.7, Illumina), and the transcript abundance was quantified in transcripts per million. Heatmap visualization was generated using MultiExperiment Viewer (version 4.9.0). Pathway analysis was conducted using the Database for Annotation, Visualization, and Integrated Discovery web tool (National Cancer Institute, United states of America) and the Kyoto Encyclopedia of Genes and Genomes (Kyoto University Institute for Chemical Research, Bioinformatics Center, Japan).

Gelatin zymography. iPSC-RPE cells were seeded onto Millicell culture inserts (MilliporeSigma, PIHA01205) and cultured for 4 weeks, followed by POS supplementation for an additional 4 weeks. The medium beneath each insert was collected, and 5 μL of the collected medium was loaded onto a fluorescently labeled precast gel (Cosmo Bio, AK83). Zymography assays were performed according to the manufacturer’s instructions. The gel was visualized using a Molecular Imager ChemiDoc XRS+ system (Bio-Rad), and gelatinase activity was quantified by measuring band intensity using ImageJ software (NIH).

Untargeted lipidomics. Liquid chromatography–tandem mass spectrometry (LC-MS/MS) was used for comprehensive untargeted lipidomics analysis, as previously described (41). Briefly, the LC system consisted of a Waters Acquity UPLC system equipped with an ACQUITY UPLC MEH C18 column (50 x 2.1 mm i.d., 1.7 μm particle size; Waters). Lipid detection was performed using a quadrupole/time-of-flight mass spectrometer (TripleTOF 6600; SCIEX, Framingham, MA, USA). Qualitative and quantitative lipid analyses were conducted using MS and data-dependent MS/MS acquisition in both negative and positive ion modes.

Cellular measurements of free cholesterol. Free cholesterol was quantified using the Total Cholesterol/Cholesteryl Ester Quantitation Kit II (BioVision, K623-100) following the manufacturer’s instructions. Briefly, 1 × 10⁶ cells were homogenized in 200 μL of a chloroform:isopropanol:NP-40 solution (7:11:0.1). After centrifugation at 15,000g for 5 minutes, the liquid phase was transferred to a new tube and air-dried at 50°C. The dried lipids were then dissolved in 200 μL of assay buffer, and the optical density was measured using a fluorescence plate reader (ARVO Multilabel Counter, Wallac). Values were normalized to cell number.

Functional analysis of lysosomes. To evaluate lysosomal acidity, iPSC-RPE cells were stained with LysoTracker Green (Cell Signaling Technology) according to the manufacturer’s instructions after incubation with the dye for 30 minutes at 37°C. The proportion of LysoTracker⁺ cells was measured by flow cytometry (BD FACS Calibur). To assess lysosomal degradative capacity, iPSC-RPE cells were incubated overnight in OPTI-MEM (Gibco, 31985-062) and subsequently stimulated with 100 ng/mL EGF (R&D SYSTEMS, 236-EG) for 6 hours. Cell lysates were collected at 0, 5, 15, 30, and 60 minutes after EGF removal and analyzed by Western blotting using an anti-phospho-EGFR antibody.

Statistics. Statistical significance was assessed using 1-way ANOVA followed by Tukey’s multiple-comparison test. Pairwise comparisons were performed using 2-tailed Student’s t test or 2-tailed Welch’s t test, as appropriate. Data derived from multiple iPSC lines were analyzed using a linear mixed-effects models, with experimental group/treatment or time as fixed effects and iPSC line identity as a random effect. Multiple comparisons for mixed-effects models were performed using Tukey’s honestly significant difference (HSD) test. All statistical analyses were performed using JMP Pro 17 or Student Edition (JMP Statistical Discovery LLC). P < 0.05 was considered statistically significant.

Study approval. This study adhered to the tenets of the Declaration of Helsinki and was approved by the IRB of Kyoto University Graduate School of Medicine (R0091). Informed consent was obtained from all participants.

Data availability. Untargeted lipidomic data have been deposited in the MB-POST repository (https://repository.massbank.jp/) under accession no. MPST000187 and are publicly available. The RNA-seq data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession no. GSE329104 and are publicly available. The distribution of MAL iPSCs requires the establishment of a Material Transfer Agreement (MTA). Requests for data access and other materials should be directed to Hanako Ikeda. Supporting data values for all figures are provided in the Supporting Data Values file, with separate tabs corresponding to each figure panel.