Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

bioRxiv logoLink to bioRxiv
[Preprint]. 2024 Mar 1:2024.02.28.582405. [Version 1] doi: 10.1101/2024.02.28.582405

Metabolic phenotyping of healthy and diseased human RPE cells

Saira Rizwan 1,2,*, Beverly Toothman 1,2,*, Bo Li 1,2,3,*, Abbi J Engel 4, Rayne R Lim 4, Jinyu Lu 1,2, Jennifer R Chao 4,&, Jianhai Du 1,2,&
PMCID: PMC10925320  PMID: 38464098

Abstract

Purpose:

Metabolic defects in retinal pigment epithelium (RPE) are underlying many retinal degenerative diseases. This study aims to identify the nutrient requirements of healthy and diseased human RPE cells.

Methods:

We profiled the utilization of 183 nutrients in human RPE cells: 1) differentiated and dedifferentiated fetal RPE (fRPE), 2) induced pluripotent stem cell derived-RPE (iPSC RPE), 3) Sorsby fundus dystrophy (SFD) patient-derived iPSC RPE and its CRISPR-corrected isogenic SFD (cSFD) iPSC RPE, and 5) ARPE-19 cell lines cultured under different conditions.

Results:

Differentiated fRPE cells and healthy iPSC RPE cells can utilize 51 and 48 nutrients respectively, including sugars, intermediates from glycolysis and tricarboxylic acid (TCA) cycle, fatty acids, ketone bodies, amino acids, and dipeptides. However, when fRPE cells lose epithelial phenotype through dedifferentiated, they can only utilize 17 nutrients, primarily sugar and glutamine-related amino acids. SFD RPE cells can utilize 37 nutrients; however, Compared to cSFD RPE and healthy iPSC RPE, they are unable to utilize lactate, some TCA cycle intermediates, and short-chain fatty acids. Nonetheless, they show increased utilization of branch-chain amino acids (BCAAs) and BCAA-containing dipeptides. The dedifferentiated ARPE-19 cells in traditional culture media cannot utilize lactate and ketone bodies. In contrast, nicotinamide supplementation promotes differentiation into epithelial phenotype, restoring the ability to use these nutrients.

Conclusions:

Epithelial phenotype confers metabolic flexibility to the RPE for utilizing various nutrients. SFD RPE cells have reduced metabolic flexibility, relying on the oxidation of BCAAs. Our findings highlight the importance of nutrient availability and utilization in RPE differentiation and diseases.

Introduction

Retinal pigment epithelium (RPE) plays a vital role in supporting photoreceptor function and survival through various specific functions, including visual cycle, nutrient transport, light absorption, phagocytosis of outer segments, and formation of the blood-retina barrier1. To sustain these critical functions, the RPE relies on a robust mitochondrial metabolism by oxidizing fuels from the photoreceptors and choroidal supply2. Defects in RPE mitochondrial metabolism can cause RPE epithelial-mesenchymal transition (EMT) and dedifferentiation to lose its epithelial characteristics such as tight junctions, pigmentation and polarity, subsequently leading to photoreceptor death in retinal degenerative diseases, including Sorsby Fundus Dystrophy (SFD), and age-related macular degeneration (AMD)2-4.

Cultured human RPE cells are pivotal to studying RPE function and elucidating mechanisms underlying retinal degenerative diseases5-7. The most common types of RPE cultures include human fetal RPE (fRPE), patient-derived induced pluripotent stem cells (IPSC) RPE, and ARPE19. Cultured mature human RPE cells exhibit typical RPE morphology, function and signature gene expression closely resembling native RPE cells8. Mitochondrial metabolism regulates RPE nutrient utilization, maturation, morphology and function5, 9. For instance, mature fRPE and iPSC RPE cells preferentially oxidize proline to fuel mitochondrial metabolism, a feature positively correlated with RPE maturation. In contrast, dedifferentiated RPE lacks the ability to effectively oxidize proline10, 11. ARPE19 cells show fibroblast-like morphology under traditional culture media. However, supplementation with nicotinamide (NAM), a precursor for NAD, can rapidly induce differentiation of ARPE19 cells into RPE cells with many epithelial characteristics through revitalizing mitochondrial metabolism12, 13.

SFD, a rare macular degeneration caused by mutations in the Tissue Inhibitor of Metalloproteinase-3 (TIMP3) gene, presents clinical manifestations similar to AMD14. We demonstrate that iPSC RPE derived from SFD patients carrying the S204C mutation in TIMP3 reproduces sub-RPE deposits observed in the globes from SFD patients15. Employing CRISPR-Cas9 gene editing, correction of the S204C mutation in SFD iPSC RPE (cSFD) diminishes the presence of sub-RPE deposits. Notably, SFD RPE enhances the breakdown of the extracellular matrix (ECM) proteins but increases lipid and protein-rich deposits15.

In this study, we aimed to elucidate the metabolic phenotype of nutrient utilization in healthy and diseased human RPE. To this end, we conducted an extensive metabolic screening carbon and nitrogen sources across various human RPE cultures, including mature and dedifferentiated fRPE, healthy and SFD RPE, and ARPE19 RPE cells under different culture media. We have found healthy and mature RPE cells demonstrate robust metabolic flexibility, allowing them to utilize various nutrient sources. In contrast, the dedifferentiated and SFD RPE cells reduce their metabolic flexibility, relying on specific nutrient sources. These cell-specific differences in nutrient utilization should provide insights into the underlying mechanisms of RPE differentiation and diseases.

Methods

Reagent and key resources

All the reagents and key resources were detailed in Supplemental Table S1 or methods.

Human fRPE cell culture

RPE was isolated from human fetal eye cups and cultured as previously described11, 16. The protocol was approved by the Institutional Review Boards of the University of Washington and West Virginia University, and all procedures conform to the ethical principles outlined in the Declaration of Helsinki. Briefly, human fRPE cells were cultured in MEM RPE media, consisting of MEM-α medium supplemented with N1 supplement, non-essential amino acids solution, 1% Fetal Bovine Serum (FBS), and 1% penicillin-streptomycin. (See details in Table S1). Cells at passages 4 to 6 were used for experiments, and they were seeded at a density of either 5 X 105 cells/cm2 (regular density) or 4 X 105 cells/cm2 (low density) in 96-well plates. Cell were grown for 4 weeks at regular density or 1-4 weeks at low density to induce dedifferentiation before metabolic screening.

Generation of human iPSC RPE cells

Informed consent was obtained from all subjects prior to inclusion in the study (University of Washington IRB-approved STUDY00010851). Blood drawn from subjects was separated using Ficoll-Hypaque gradient to isolate peripheral blood mononuclear cells. These cells were cultured and reprogrammed using CytoTune Sendai reprogramming kit. Following transfection, cells were cultured on irradiated mouse embryonic fibroblasts until iPSC colonies emerged. Healthy iPSC clones were manually picked and expanded in mTeSR medium on feeder-free plates for cryostorage and differentiation. Karyotyping was performed for all iPSC clones prior to their selection for inclusion in the study.

Generation of SFD patient iPSC RPE and cSFD iPSC RPE cells

SFD iPSC RPE and cSFD iPSC RPE cells were generated as reported15. Briefly, following confirmation of iPSC karyotype from SFD patients, iPSC RPE underwent electroporated with Cas9-gRNA (0.75uM, GGGGCCTATTTTCGTAGTAG, Synthego) ribonucleoprotein (RNP) complex and ssDNA donor (4uM, IDT) using Amaxa nucleofector (Human Stem Cell kit 2). Individual iPSC colonies were manually picked and seeded into 96-well plates. DNA was isolated from each colony and sequenced to identify the gene-edited iPSC colony. To assess potential off-target effects resulting from CRISPR-editing, off-target sites were identified using CRISPOR (http://crispor.tefor.net), which revealed two intronic with three mismatches. These regions were sequenced in the gene-edited iPS RPE, and no genetic alterations were detected15.

Human iPSC RPE cell culture

iPSC and gene-edited iPSCs were differentiated to RPE in a stepwise procedure as previously described15. RPE was maintained in MEM RPE media, detached with 0.25% Trypsin-EDTA and seeded at 5 X 105 cells/cm2 into 96-well plates. RPE cells at passages 4 to 6 were cultured for 4 weeks before metabolic screening.

Human ARPE19 cell culture

Human ARPE19 cells obtained from the American Type Culture Collection (ATCC) were used at passages 5-8 for the experiments. The cells were seeded at 2 X 104 cells/cm2 in 96-well plates and cultured in three different media to mimic dedifferentiated and differentiated states: 1) DMEM/F-12 media with 5% FBS, 2) MEM RPE media, and 3) MEM-NAM (MEM RPE media supplemented with 10 mM NAM as reported12, 13. All the ARPE-19 cells were grown for 4 weeks before metabolic screening.

Metabolic screening with metabolic phenotyping microarrays

We used two phenotype microarrays (PM-M1 for carbon sources and PM-M2 for nitrogen sources) (See details in Table S1) to assess nutrient utilization. Each well of the 96-well plates contained a specific nutrient except for the negative control (Table S2-3). The PM-M1 or PM-M2 plates were initially pre-incubated with 60 uL of nutrient-limiting media (IFM 1, glutamine 200 ¼M, 1% dialyzed FBS, and 1% penicillin and streptomycin) for one hour. Subsequently, the 96-well plates containing RPE cells underwent two PBS washes before being filled with the pre-incubated nutrients transferred from PM-M1 or PM-M2 microarrays. The plates were then incubated for 40 hours at 37°C in a 5% CO2 incubator, after which 10 ¼L of the redox dye mix MA was added. The plates were sealed and read at an absorbance of 595 nm every 15 minutues for 4 hours.

Statistics

All data are expressed as the mean ± standard deviation. Fold change of absorbance over negative control without nutrient source>1.5, or p<0.05 by Student’s unpaired two-tailed t-tests using GraphPad Prism 9.0, was considered significant. Heatmaps were generated in Microsoft Excel.

Results

Mature human fRPE cells demonstrate remarkable metabolic flexibility to utilize various nutrient sources.

To study nutrient utilization in mature fRPE cells, we grew fRPE for 4 weeks into maturity with RPE characteristic cobblestone morphology and pigmentation (Supplemental Fig S1A-C). The mature RPE cultures were switched into nutrient-limiting media with or without specific nutrient from the PM-M1 plate to screen the utilization of 91 carbon sources (Fig 1A, Table S2). Remarkably, mature RPE could utilize 23 carbon sources including sugars, glycolysis intermediates (such as sugar phosphates, lactate, and pyruvate), TCA cycle intermediates, nucleosides, fatty acids and ketone bodies (Fig 1B-C). In addition to glucose, mature RPE had the capacity to utilize other sugars such as fructose, mannose, and galactose (Fig 1B-C). Nucleosides, containing ribose moiety, can serve as alternative energy sources. Mature RPE could robustly utilize inosine, adenosine and uridine.

Figure 1. Metabolic phenotyping of mature human fRPE cells.

Figure 1.

Open in a new tab

(A) A schematic for metabolic phenotyping with carbon sources. Human fRPE cells were grown for 4 weeks in a 96-well plate and then switched into nutrient-limiting media containing different carbon sources in each well from PM-M1 plate. The utilization of nutrients leads to NADH production, casuing a color change in a redox sensitive dye to purple, which is quantified by a microplate reader at 595 nm. (B) The utilization of carbon sources by mature human fRPE cells and (C) an illustration of the metabolic pathways. (D) A schematic for metabolic phenotyping of nitrogen sources using PM-M2. (E) The utilization of nitrogen sources by mature human fRPE cells and (F) an illustration of the metabolic pathways. N=3. Fold change over negative control without nutrient source (Ctr)>1.5 or *P<0.05 over Ctr. Gluc, Glucose, Gal, galactose; Mann, Mannose; F1P, fructose 1-phosphophate; R1P, ribose 1-phosphate, G6P, glucose 3-phosphate; G3P, glycerol 3-phosphate; AcAc, Acetoacetic Acid; 3HB, β-Hydroxy Butyrate; MelA, Melibionic Acid.

Similar to PM-M1 carbon source screening, we evaluated the utilization of 92 nitrogen sources, including 27 amino acids and 60 dipeptides in mature RPE using the PM-M2 plate as illustrated in Fig 1D. Mature RPE could metabolize 7 amino acids and 17 dipeptides (Fig 1E). Except for methionine and tryptophan, the remaining 5 amino acids (proline, ornithine, glutamine, glutamate, and aspartate) follow a similar metabolic pathway to generate α-ketoglutarate (αKG) or oxoacetate, which serves as fuels for the TCA cycle (Fig 1E, 1F). Notably, glutamine, glutamate, aspartate, and proline are also present in the 17 dipeptides. These findings suggest that mature fRPE has remarkable metabolic flexibility, enabling them to effectively utilize diverse nutrient sources to fuel their mitochondrial metabolism (Fig 1C, 1E).

Dedifferentiated human fRPE loses metabolic flexibility, primarily relying on sugars and glutamine as their main nutrient sources.

Seeding fRPE at low density induces their dedifferentiation into fibroblast-like phenotype5. To understand nutrient utilization in dedifferentiated RPE cells, we seeded fRPE at 10% of regular density and grew for 4 weeks. As expected, these cells transitioned from their characteristic cobblestone structure to a fibroblast-like morphology (Fig 2A, Supplemental Fig S1D-E). We performed metabolic phenotyping of carbon and nitrogen sources like mature fRPE cells (Fig 2A). Compared to mature fRPE, dedifferentiated fRPE utilized substantially fewer nutrients, with only 8 carbon and 6 nitrogen sources. Intriguingly, almost all the carbon sources were sugar and sugar phosphates, while all nitrogen sources contained glutamine (Fig 2B-C). These results suggest that dedifferentiated RPE cells lose the ability to utilize multiple nutrient sources such as lactate, fatty acid, ketone bodies and proline. Instead, they become addicted to consuming sugars and glutamine (Fig 2D). We also performed PM-M1 screening on 1-week fRPE with low-density seeding to examine potential changes in nutrient utilization over time (Fig S2A-B). Unlike fRPE cells seeded at regular density, those seeded with low density for 1 week displayed a fibroblast-like morphology resembling that of 4-week cultured fRPE cells (Fig S1D-E). Nutrient utilization in 1-week cells were largely consistent with those of 4-week fRPE, except for the capability to utilize galactose and adenosine in 1-week cells (Fig S2B), suggesting that culture time has minimal impact on metabolic phenotype of nutrient utilization in RPE cells seeded at low density. Consequently, we did not perform PM-M2 screening on 1-week fRPE.

Figure 2. Metabolic phenotyping of dedifferentiated human fRPE.

Figure 2.

Open in a new tab

(A) Human fRPE were seeded low density to induce dedifferentiation with a mesenchymal phenotype in a 96-well plate and then switched into nutrient-limiting media containing different carbon (PM-M1) or nitrogen sources (PM-M2) for metabolic phenotyping. (B) Nutrient utilization of carbon sources by dedifferentiated human fRPE. N=4. Fold change>1.5 or P<0.05 over Ctr. (C) Nutrient utilization of nitrogen sources by dedifferentiated human fRPE. N=3. Fold change>1.5 or P<0.05 over Ctr. (D) A comparison of metabolic phenotyping between mature fRPE and dedifferentiated fRPE. (E) The dedifferentiated RPE cells lose the ability to utilize multiple nutrient sources but switch to use glutamine, enhancing proline and collagen synthesis. Magenta denotes an increase and blue denotes a decrease in gene expression or metabolic pathways. PFKFB4, 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 4, LDHA, Lactate dehydrogenase A; OGDH, Oxoglutarate dehydrogenase; PRODH; Proline dehydrogenase; PYCR1, Pyrroline-5-carboxylate reductase 1; P4H4, Prolyl 4-hydroxylase subunit alpha or beta; GMPPA, GDP-mannose pyrophosphorylase A. PMM2, Phosphomannomutase 2.

To investigate the mechanism underlying these distinct metabolic phenotypes between mature and dedifferentiated RPE cells, we analyzed the expression of metabolic genes related to nutrient utilization in two RNA-Seq data sets: 1) mature fRPE vs dedifferentiated RPE induced by multiple passages17 and 2) mature primary adult RPE vs dedifferentiated RPE induced by transforming growth factor-beta (TGFβ) and tumor necrosis factor α (TNFα)18. Both dedifferentiated RPE models had significant downregulation of genes invloved in fatty acid oxidation, TCA cycle, transporters for lactate and ketone bodies, nucleotide metabolism, and amino acid metabolism especially proline catabolism and glutamine synthetase (Table S4, Fig 2E). These data corroborate our findings that dedifferentiated RPE cannot efficiently utilize fatty acid, lactate, ketone bodies, proline and other nucleosides for mitochondrial metabolism. Importantly, genes involved in proline synthesis, glycolysis, collagen and heparin sulfate were substantially upregulated (Table S4), suggesting dedifferentiated RPE cells utilize sugar and glutamine to activate proline synthesis and the hexosamine biosynthetic pathway to produce collagen and glycan for ECM remodeling (Fig 2E).

Human iPSC RPE cells have similar metabolic flexibility to fRPE cells but utilize more sugars and TCA cycle intermediates.

To study nutrient utilization in human iPSC RPE cells, we cultured iPSC RPE for 4 weeks to reach maturity (Fig S3) and then replaced media with carbon or nitrogen sources from PM-M1 or PM-M2 for metabolic phenotyping (Fig 3A). iPSC RPE could utilize 31 carbon sources and 17 nitrogen sources (Fig 3B-D). Similar to fRPE, iPSC RPE demonstrated robust flexibility to utilize various sugars, fatty acids, lactate, ketone bodies, nucleosides, amino acids, and dipeptides (Fig 3B-D). However, unlike fRPE, iPSC RPE showed a preference for more sugars and TCA cycle intermediates, but less amino acids and dipeptides (Fig 3E-F). These results suggest both healthy fRPE and iPSC RPE possess metabolic flexiblility to adapt to different substrates, yet also have cell type-specific preferences for certain substrates.

Figure 3. Metabolic phenotyping of healthy human iPSC RPE cells.

Figure 3.

Open in a new tab

(A) Human iPSC RPE cells were differentiated in RPE media for 4 weeks and switched into nutrient-limit media from PM1 or PM2 for metabolic phenotyping. (B-D) Nutrient utilization of carbon sources and nitrogen sources by human iPSC RPE cells. N=3. Fold change>1.5 or P<0.05 over Ctr. (E) A comparison of number of nutrients that were utilized between fRPE and iPSC RPE cells. (F) An illustration of nutrient utilization in iPSC RPE cells. GlcNAc, N-Acetyl-D-Glucosamine; Tcba, Tricarballylic Acid; AlaN, L-Alaninamide.

SFD iPSC RPE cells utilize less intermediates from glycolytic and TCA cycle intermediates but more branch-chain amino acids (BCAAs).

We reported SFD iPSC RPE cells harboring the TIMP3 S204 mutation have increased extracellular deposits and elevated intracellular 4-hydroxyproline15. These iPSC RPE and their CRISPR-corrected cSFD were grown into maturity for 4 weeks before undergoing metabolic phenotyping with PM-M1 and PM-M2 (Fig 4A). Both SFD and cSFD had normal cobblestone structure and pigmentation (Fig S4A-B). While SFD RPE cells utilized 19 carbon sources and 19 nitrogen sources, cSFD RPE cells utilized 28 carbon sources and 7 nitrogen sources (Fig 4B-E). Similar to mature fRPE and normal iPSC RPE, SFD RPE utilized different types of sugar, nucleosides, fatty acids, ketone bodies, amino acids and dipeptides (Fig 4B, 4D). However, SFD RPE could not utilize lactate, sugar phosphate, short-chain fatty acids (acetate and butyrate), succinate, αKG and gamma-aminobutyric acid (GABA). Notably, the CRISPR-corrected RPE cells restored the utilization of these nutrients (Fig 4C-E). SFD RPE could use proline and dipeptides containing glutamine, alanine and arginine. Intriguingly, many of the dipeptides utilized by SFD RPE cells contained BCAAs inlcuding leucine, isoleucine and valine, along with the ability to catabolize free leucine (Fig 4D). These metabolic phenotypes are different from fRPE and healthy iPSC RPE cells. While the cSFD RPE cells did not use the BCAAs but they also utilized significantly less nitrogen sources (Fig 4E, Fig S4C). These results suggest that SFD RPE cells have metabolic defects in recycling nutrients such as lactate from the neural retina but rely more on BCAAs, which may contribute to the formation of sub-RPE deposits of lipids and ECM proteins (Fig 4F).

Figure 4. Metabolic phenotyping of IPSC SFD and cSFD RPE cells.

Figure 4.

Open in a new tab

(A) Human iPSC SFD RPE carring TIMP3 204C mutation and mutation corrected cSFD RPE cells were differentiated in RPE media for 4 weeks and then switched to nutrient-limit media with carbon or nitrogen sources from PM-M1 or PM-M2 for metabolic phenotyping. (B-E) Nutrient utilization of (B-C) carbon sources and (D-E) nitrogen sources in iPSC SFD RPE cells. N=3. Fold change>1.5 or P<0.05 over Ctr. (F) An illustration of altered nutrient utilization in SFD iPSC RPE cells. L-Mal, L-Malic Acid; mmSuc, mono methyl succinate; 3HB, β-Hydroxy Butyrate; BCAAs, Branch-Chained Amino Acids.

Culture media composition controls the differentiation and nutrient utilization in APRE19 cells.

ARPE19 cells are typically cultured in DMEM/F-12 media by manufacturer’s instruction. However, when ARPE19 cells are cultured in MEM-α-based RPE culture media supplemented with NAM, these cells can undergo rapid differentiatiation into a mature RPE-like phenotype12, 13. As a proof of principal that medium composition can influence nutrient utilization in ARPE19 cells, we cultured them into three media (DMEM/F12, MEM RPE media and MEM RPE media plus NAM) for 4 weeks and focused only on the screening of carbon sources using PM-M1 plate (Fig 5A). ARPE19 cells in DMEM/F12 had fibroblast-like morphology (Fig S5A). MEM RPE media partially differentiated the cell line into RPE-like cobblestone structure with clumps of cells on the top. However, NAM almost completely differentiated APRE-19 cells into mature RPE-like morphology as reported (Fig S5B-C). The DMEM/F12 cultured cells could utilize 16 carbon sources including sugar, fatty acids, nucleosides, pyruvate, mono-methyl-succinate and two intermediates from amino acid metabolism (Fig 5B). Similar to dedifferentiated fRPE, APRE19 cells cultured in DMEM/F12 could not utilize lactate, ketone bodies and xylitol (Fig 5B, Fig 6A), suggesting a common metabolic defect in dedifferentiated RPE cells. ARPE19 cells cultured in MEM RPE media could utilize 24 carbon sources, which covered 94% of nutrients that were used by DMEM/F12 cultured cells. Cells cultured in MEM RPE media could utilize ketone bodies, glycogen and other sugar, but still could not metabolize lactate and xylitol (Fig 5C). The supplementation with NAM enabled ARPE-19 cells to utilize lactate, ketone bodies, sugar phosphate, xylose and tagatose (Fig 5D-E), demonstrating patterns similar to those observed in mature fRPE and iPSC RPE, except for the reduced utilization of short-chain fatty acids and maltose (Fig 6A). These results further underscore the critical roles of nutrient availability and utilization in RPE differentiation.

Figure 5. Metabolic phenotyping of ARPE19 cells cultured in different media.

Figure 5.

Open in a new tab

(A) ARPE19 cells were cultured in three different media: DMEM/F-12, MEM RPE media, and MEM RPE media with 10 mM NAM for 4 weeks. All the cells were then changed into nutrient-limit media with carbon sources from PM-M1 for metabolic phenotyping. (B-D) Nutrient utilization of carbon sources of ARPE19 cells cultured in (B) DMEM/F12 (C) MEM RPE media, (D) MEM RPE media with NAM. Fold change>1.5 or P<0.05 over Ctr. (E) A comparison of number of nutrients that were utilized in ARPE19 cells grown in different culture media. (F-G) An illustration of altered nutrient utilization in APRE19 cells cultured between DMEM/F12 and MEM RPE media with NAM. Meso-Tar, Meso Tartatic acid; mmSuc, mono methyl succinate; AcAc, Acetoacetic Acid; 3HB, β-Hydroxy Butyrate.

Figure 6. Comparison of metabolic phenotyping profiles of human fRPE cells, dedifferentiated fRPE, iPSC RPE cells, SFD RPE, cSFD RPE and ARPE-19 cells grown under different culture media.

Figure 6.

Open in a new tab

Heat map visualization of nutrients that were differentially utilized by different RPE cells from (A) PM-M1 nutrients and (B) PM-M2 nutrient microarray plates. De-fRPE, dedifferentied fRPW with low density seedings; ARPE, ARPE19 cells; MelA, Melibionic Acid; GlcNAc, N-Acetyl-D-Glucosamine; Tcba, Tricarballylic acid; mmSuc, mono methyl succinate; AlaN, Alaninamide.

Discussion

In this study, we have identified distinctive metabolic phenotypes of nutrient utilization in healthy and diseased human RPE cells (Fig 6). Mature fRPE and iPSC RPE are metabolically flexible, capable of adapting to various nutrient sources. Conversely, dedifferentiated RPE cells and SFD patient-derived iPSC RPE cells reduce their flexibility to dependend on specific substrates. Our findings suggest the critical roles of nutrient utilization in RPE differentiation and dedifferentiation.

Every day, RPE must manage a plentiful nutrient influx origninating from phagocytosed lipid- and protein-rich photoreceptor outer segments, from “waste products” such as lactate produced by the neural retina, and rich from nutrient-laden blood through choroid circulation1, 2. In addition to direct transport of nutrients to the neural retina, RPE can recycle and generate nutrients via lysosomes and mitochondria to support retinal metabolism, thereby establishing a metabolic ecosystem between the outer retina and RPE1, 2, 19. The ability to utilize various nutrients is crucial for maintaining this metabolic ecosystem and disruption of this balance due to mitochondrial and/or lysosomal dysfunction in RPE can lead to retinal degeneration3, 4, 20. Healthy human RPE cells are known to be versatile in utilizing different nutrients, including glucose, galactose, lactate, fatty acids, succinate, proline, glutamine and other amino acids11, 19, 21-24. Consistent with these findings, our results demonstrate both healthy fRPE and iPSC RPE cells are capable of utilizing these nutrients.

Furthermore, we have found human RPE cells utilize a wide range of other nutrients including various sugars, ketone bodies, nucleosides and dipeptides. Mannose and galactose, for example, are components of glycoproteins found abundantly in the outer segment proteins such as rhodopsin and peripherin 225-27. The ability to utilize these sugars from daily degradation of outer segment may preserve more glucose for the photoreceptors. Additionally, like the liver, human RPE cells can synthesize ketone bodies, which can be utilized by the neural retina22, 28. Interestingly, we have found mature human RPE can also readily utilize ketone bodies but not the dedifferentiated RPE cells, suggesting that the ability to utilizing ketone bodies might be a characteristic of RPE metabolism. Consistent with this metabolic capability, mature RPE expresses key genes involved in ketone bodies degradation22, a feature distinct from the liver, which can not utilize ketone bodies. Our findings undscore the metabolic flexibility of healthy mature RPE cells in nutrient utilization, an essential RPE function to support retinal metabolism and health.

RPE EMT and dedifferentiation are important events in wound healing and pathogenesis of proliferative vitreoretinopathy and AMD29. We have found dedifferentiated RPE cells are unable to utilize alternative fuels like lactate, fatty acids, ketone bodies and proline, but instead, they prefer to utilize sugars and glutamine. RNA-seq data from EMT models show reduced genes in TCA cycle, lactate transport, ketone body utilization and proline catabolism and glutamine synthesis, whereas genes associated in glycolysis, proline synthesis, collagen synthesis and glycoproteins are upregulated (Fig 2E, Table S4). Consistently, dedifferentiated RPE cells have been reported to switch from oxidative phosphorylation to glycolysis and lose the ability to utilize proline as a fuel5, 16. Inhibition of mitochondrial metabolism in human RPE cells or mouse RPE is sufficient to cause the activation of glycolysis, RPE dedifferentiation and retinal degeneration4, 30. Furthermore, mitochondrial respiration diminishes in primarily cultured RPE cells from AMD donors31. Inhibiting mitochondrial respiration impedes proline consumption while increasing glucose consumption and lactate production9. The inability to utilize exogenous lactate and other alternative nutrient sources in RPE may be an important mechanism in RPE differentiation and AMD pathogenesis.

A key feature in dedifferentiated cells is their massive production of ECM proteins, primarily collagens and glycoproteins, through activating cytokine signaling such as TGFβ and TNFα32, 33. In fibroblasts, TGFβ stimulates the oxidation of glucose and glutamine to synthesize proline and other amino acids crucial for ECM protein production34. Interestingly, proline synthesis acts as a vent for growth when cells are under mitochondrial redox stress or hypoxia by recycling NADH into NAD+, thus enhancing the oxidation of glutamine and glucose34, 35. Proline synthesis may be particularly crucial for dedifferentiated RPE cells, given their inhibited mitochondrial metabolism and heavy reliance on lactate production for NAD+ regeneration. Strategies aimed at relieving mitochondrial redox stress and reverting proline synthesis to proline catabolism could hold promise for treating RPE dedifferentiation.

TIMP3, a risk gene for AMD, is secreted by the RPE and plays a critical role in ECM remodeling by inhibiting matrix metalloproteinases36. Mutations in TIMP3 cause SFD, characterized by drusen-like sub-RPE deposits containing lipids and ECM proteins15, 37. Unlike healthy or mutation-corrected RPE cells, SFD RPE cells utilize more free BCAAs or dipeptides containing BCAAs (Fig 4F, Fig 6). BCAAs are ketogenic amino acids and their oxidation promotes the synthesis and transport of fatty acids and cholesterol38, 39. BCAA oxidation is closely associated with dysregulated glucose and lipid metabolism in diabetes and non-alcoholic fatty liver diseases38, 40, 41. Interestingly, intermediates of the BCAA degradation pathway are significantly elevated in the plasma of AMD patients42, and BCAT1, a key gene in BCAA degradation, upregulates in the RPE/choroid from AMD donors43. Moreover, in addition to proline, ECM proteins are highly enriched with BCAAs44 . Our findings suggest that SFD RPE might upregulate BCAA degradation pathway by utilizing BCAAs from free amino acids or protein degradation, contributing to the formation of their sub-RPE deposits.

In conclusion, metabolic phenotyping offers a sensitive and powerful platform for measuring nutrient utilization in healthy and diseased human RPE cells. Healthy and mature human fRPE and iPSC RPE have similar features of high metabolic flexibility in utilizing various nutrient sources. However, dedifferentiated RPE or SFD RPE cells demonstrate selective preferences, favoring the use of sugars, glutamine or BCAAs, which may contribute to their morphological and pathological alterations in AMD and SFD.

Supplementary Material

Supplement 1
media-1.pdf (819.6KB, pdf)
Supplement 2
media-2.xlsx (47.3KB, xlsx)

Acknowledgment

This work was supported by National Institutes of Health Grant EY034364 (J. R. C. and J. D), EY03459 (J. R. C. and J. D), EY031324 (JD), EY032462(JD), the Retina Research Foundation (JD), Human Development grant R24HD000836 (The Birth Defects Research Laboratory from the University of Washington), and funds for Core facilities P20 GM103434 and P20 GM144230 (WV INBRE grant).

Funding:

The funders had no role in the study design, data collection and analysis, decision on publishing, or manuscript preparation. This work was supported by National Institutes of Health Grant EY034364 (J. R. C. and J. D), EY03459 (J. R. C. and J. D), EY031324 (JD), EY032462(JD), the Retina Research Foundation (JD) and funds for Core facilities P20 GM103434 and P20 GM144230 (WV INBRE grant).

Footnotes

Commercial relationships disclosures: None.

Conflicts of interest

The authors declare no conflicts of interest.

References

  • 1.Lakkaraju A, Umapathy A, Tan LX, et al. The cell biology of the retinal pigment epithelium. Prog Retin Eye Res 2020;100846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hurley JB. Retina Metabolism and Metabolism in the Pigmented Epithelium: A Busy Intersection. Annu Rev Vis Sci 2021;7:665–692. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Ferrington DA, Fisher CR, Kowluru RA. Mitochondrial Defects Drive Degenerative Retinal Diseases. Trends Mol Med 2020;26:105–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Zhao C, Yasumura D, Li X, et al. mTOR-mediated dedifferentiation of the retinal pigment epithelium initiates photoreceptor degeneration in mice. J Clin Invest 2011;121:369–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Adijanto J, Philp NJ. Cultured primary human fetal retinal pigment epithelium (hfRPE) as a model for evaluating RPE metabolism. Exp Eye Res 2014;126:77–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Bharti K, den Hollander AI, Lakkaraju A, et al. Cell culture models to study retinal pigment epithelium-related pathogenesis in age-related macular degeneration. Exp Eye Res 2022;222:109170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nazari H, Zhang L, Zhu D, et al. Stem cell based therapies for age-related macular degeneration: The promises and the challenges. Prog Retin Eye Res 2015;48:1–39. [DOI] [PubMed] [Google Scholar]
  • 8.Blenkinsop TA, Saini JS, Maminishkis A, et al. Human Adult Retinal Pigment Epithelial Stem Cell-Derived RPE Monolayers Exhibit Key Physiological Characteristics of Native Tissue. Invest Ophthalmol Vis Sci 2015;56:7085–7099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zhang R, Engel AL, Wang Y, et al. Inhibition of Mitochondrial Respiration Impairs Nutrient Consumption and Metabolite Transport in Human Retinal Pigment Epithelium. J Proteome Res 2021;20:909–922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Du J, Zhu S, Lim RR, Chao JR. Proline metabolism and transport in retinal health and disease. Amino Acids 2021;53:1789–1806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Chao JR, Knight K, Engel AL, et al. Human retinal pigment epithelial cells prefer proline as a nutrient and transport metabolic intermediates to the retinal side. J Biol Chem 2017;292:12895–12905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hazim RA, Volland S, Yen A, Burgess BL, Williams DS. Rapid differentiation of the human RPE cell line, ARPE-19, induced by nicotinamide. Exp Eye Res 2019;179:18–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hazim RA, Paniagua AE, Tang L, et al. Vitamin B3, nicotinamide, enhances mitochondrial metabolism to promote differentiation of the retinal pigment epithelium. J Biol Chem 2022;298:102286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Anand-Apte B, Chao JR, Singh R, Stohr H. Sorsby fundus dystrophy: Insights from the past and looking to the future. J Neurosci Res 2019;97:88–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Engel AL, Wang Y, Khuu TH, et al. Extracellular matrix dysfunction in Sorsby patient-derived retinal pigment epithelium. Exp Eye Res 2022;215:108899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Yam M, Engel AL, Wang Y, et al. Proline mediates metabolic communication between retinal pigment epithelial cells and the retina. J Biol Chem 2019;294:10278–10289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Radeke MJ, Radeke CM, Shih YH, et al. Restoration of mesenchymal retinal pigmented epithelial cells by TGFbeta pathway inhibitors: implications for age-related macular degeneration. Genome Med 2015;7:58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Boles NC, Fernandes M, Swigut T, et al. Epigenomic and Transcriptomic Changes During Human RPE EMT in a Stem Cell Model of Epiretinal Membrane Pathogenesis and Prevention by Nicotinamide. Stem Cell Reports 2020;14:631–647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kanow MA, Giarmarco MM, Jankowski CS, et al. Biochemical adaptations of the retina and retinal pigment epithelium support a metabolic ecosystem in the vertebrate eye. Elife 2017;6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Boya P, Kaarniranta K, Handa JT, Sinha D. Lysosomes in retinal health and disease. Trends Neurosci 2023;46:1067–1082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Liu B, Calton MA, Abell NS, et al. Genetic analyses of human fetal retinal pigment epithelium gene expression suggest ocular disease mechanisms. Commun Biol 2019;2:186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Adijanto J, Du J, Moffat C, Seifert EL, Hurle JB, Philp NJ. The retinal pigment epithelium utilizes fatty acids for ketogenesis. J Biol Chem 2014;289:20570–20582. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bisbach CM, Hass DT, Robbings BM, et al. Succinate Can Shuttle Reducing Power from the Hypoxic Retina to the O(2)-Rich Pigment Epithelium. Cell Rep 2020;31:107606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Li B, Zhang T, Liu W, et al. Metabolic Features of Mouse and Human Retinas: Rods versus Cones, Macula versus Periphery, Retina versus RPE. iScience 2020;23:101672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Skiba NP, Lewis TR, Spencer WJ, Castillo CM, Shevchenko A, Arshavsky VY. Absolute Quantification of Photoreceptor Outer Segment Proteins. J Proteome Res 2023;22:2703–2713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Stuck MW, Conley SM, Naash MI. PRPH2/RDS and ROM-1: Historical context, current views and future considerations. Prog Retin Eye Res 2016;52:47–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Daniele LL, Han JYS, Samuels IS, et al. Glucose uptake by GLUT1 in photoreceptors is essential for outer segment renewal and rod photoreceptor survival. FASEB J 2022;36:e22428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Reyes-Reveles J, Dhingra A, Alexander D, Bragin A, Philp NJ, Boesze-Battaglia K. Phagocytosis-dependent ketogenesis in retinal pigment epithelium. J Biol Chem 2017;292:8038–8047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Zhou M, Geathers JS, Grillo SL, et al. Role of Epithelial-Mesenchymal Transition in Retinal Pigment Epithelium Dysfunction. Front Cell Dev Biol 2020;8:501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Hu X, Calton MA, Tang S, Vollrath D. Depletion of Mitochondrial DNA in Differentiated Retinal Pigment Epithelial Cells. Sci Rep 2019;9:15355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fisher CR, Ferrington DA. Perspective on AMD Pathobiology: A Bioenergetic Crisis in the RPE. Invest Ophthalmol Vis Sci 2018;59:AMD41–AMD47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Scott LE, Weinberg SH, Lemmon CA. Mechanochemical Signaling of the Extracellular Matrix in Epithelial-Mesenchymal Transition. Front Cell Dev Biol 2019;7:135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Shu DY, Butcher E, Saint-Geniez M. EMT and EndMT: Emerging Roles in Age-Related Macular Degeneration. Int J Mol Sci 2020;21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Schworer S, Berisa M, Violante S, et al. Proline biosynthesis is a vent for TGFbeta-induced mitochondrial redox stress. EMBO J 2020;39:e103334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Westbrook RL, Bridges E, Roberts J, et al. Proline synthesis through PYCR1 is required to support cancer cell proliferation and survival in oxygen-limiting conditions. Cell Rep 2022;38:110320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Alexander JP, Bradley JM, Gabourel JD, Acott TS. Expression of matrix metalloproteinases and inhibitor by human retinal pigment epithelium. Invest Ophthalmol Vis Sci 1990;31:2520–2528. [PubMed] [Google Scholar]
  • 37.Weber BH, Vogt G, Pruett RC, Stohr H, Felbor U. Mutations in the tissue inhibitor of metalloproteinases-3 (TIMP3) in patients with Sorsby's fundus dystrophy. Nat Genet 1994;8:352–356. [DOI] [PubMed] [Google Scholar]
  • 38.Neinast MD, Jang C, Hui S, et al. Quantitative Analysis of the Whole-Body Metabolic Fate of Branched-Chain Amino Acids. Cell Metab 2019;29:417–429 e414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Jang C, Oh SF, Wada S, et al. A branched-chain amino acid metabolite drives vascular fatty acid transport and causes insulin resistance. Nat Med 2016;22:421–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.White PJ, McGarrah RW, Grimsrud PA, et al. The BCKDH Kinase and Phosphatase Integrate BCAA and Lipid Metabolism via Regulation of ATP-Citrate Lyase. Cell Metab 2018;27:1281–1293 e1287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.White PJ, Lapworth AL, An J, et al. Branched-chain amino acid restriction in Zucker-fatty rats improves muscle insulin sensitivity by enhancing efficiency of fatty acid oxidation and acyl-glycine export. Mol Metab 2016;5:538–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Haijes HA, Jans JJM, van der Ham M, van Hasselt PM, Verhoeven-Duif NM. Understanding acute metabolic decompensation in propionic and methylmalonic acidemias: a deep metabolic phenotyping approach. Orphanet J Rare Dis 2020;15:68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Newman AM, Gallo NB, Hancox LS, et al. Systems-level analysis of age-related macular degeneration reveals global biomarkers and phenotype-specific functional networks. Genome Med 2012;4:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Li P, Wu G. Roles of dietary glycine, proline, and hydroxyproline in collagen synthesis and animal growth. Amino Acids 2018;50:29–38. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
media-1.pdf (819.6KB, pdf)
Supplement 2
media-2.xlsx (47.3KB, xlsx)

Articles from bioRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES