Cultured primary human fetal retinal pigment epithelium (hfRPE) as a model for evaluating RPE metabolism

Abstract

Mitochondrial dysfunction has been shown to contribute to age-related and proliferative retinal diseases. Over the past decade, the primary human fetal RPE (hfRPE) culture model has emerged as an effective tool for studying RPE function and mechanisms of retinal diseases. This model system has been rigorously characterized and shown to closely resemble native RPE cells at the genomic and protein level, and that they are capable of accomplishing the characteristic functions of a healthy native RPE (e.g., rod phagocytosis, ion and fluid transport, and retinoid cycle). In this review, we demonstrated that the metabolic activity of the RPE is an indicator of its health and state of differentiation, and present the hfRPE culture model as a valuable in vitro system for evaluating RPE metabolism in the context of RPE differentiation and retinal disease.

Keywords: differentiation; metabolism; primary cultured human fetal RPE.

Figure 1. Morphology of hfRPE cells seeded at various densities

Fully differentiated P1 hfRPE cells on transwells were trypsinized and re-seeded into 96-well plates at 60, 30, 15, 7.5, 3.75, and 2% densities. They were cultured in complete hfRPE culture medium (Advanced MEM (Gibco; cat# 12492) supplemented with Penicillin-Streptomycin, Glutamax^®, THT, and 5% HIFBS). Media were replaced three times a week and images were taken at 10X magnification on 21 days post seeding.

Figure 2. The transition from glycolytic to oxidative metabolism orchestrates the RPE differentiation process

P1 hfRPE cells were seeded at (A) low (15%) or (B) high (30%) densities to mimic early and late stage differentiation, respectively. Cells were transfected with LDHA and LDHB (20 nM each) or SDHD (40 nM) siRNAs using Dharmafect 4 (0.2%), once upon seeding and another two days post seeding – this transfection protocol has been published in our previous work (Adijanto et al., 2012). Each condition was represented by three technical replicates. RNA was isolated seven days post-transfection, and the samples were evaluated for RPE-specific gene expression (BEST1, CLDN10, MCT3, and RPE65) using qRTPCR. RPS18 was used as endogenous control for delta-Ct-Ct analysis. P < 0.05 is considered statistically significant. Primer sequences for RPE-specific genes and RPS18 have been published (Adijanto et al., 2012).

Figure 3. Differentiated hfRPE cells possess a high oxidative capacity

P1 hfRPE cells were seeded on 96-well plate at 30% or 2% density and cultured over 21 days. (A) hfRPE cells incubated with antimycin A (20 nM; 24 hr) showed a dramatic increase in lactate release into the media whereas dedifferentiated hfRPE cells did not respond to antimycin A. (B) In differentiated hfRPE (but not in dedifferentiated hfRPE), inhibition of ETC with antimycin A drives conversion of pyruvate to lactate. MCT1 at the RPE apical membrane mediates lactate efflux. (C) Differentiated hfRPE cells expressed a higher level of UQCRC2 (complex III; Abcam Cat#: MS304), an indicator of mitochondrial density.

Figure 4. Seahorse XF24 analysis of hfRPE metabolism

P1 hfRPE cells were seeded on seahorse XF24 assay plate at 45% density and were cultured over 10 days to achieve RPE differentiation. On the day of the experiment, hfRPE cells were switched to CO₂/HCO₃-free HEPES-buffered Ringer (containing 5 mM glucose) and immediately transferred to the Seahorse analyzer and a baseline OCR measurement was obtained. Next lactate or gluconate (at a 10X concentrate) were injected into each well (n = 3 each) to achieve 5 mM final concentration and subsequent changes in OCR were recorded.

Figure 5. Water-Soluble MTT measurement of RPE metabolism (NADH/NADPH production rate)

(A) Glucose enters RPE cells via a facilitated glucose transporter (GLUT) and is metabolized to glucose-6-phosphate (G6P) by a mitochondria-associated hexokinase 2 (HK2). G6P are shunted into the PPP, a process regulated by G6PD (the first enzyme in PPP). The rest of G6P enters the glycolytic pathway, in which glyceraldehyde phosphate dehydrogenase (GAPDH) converts glyceraldehyde-3-phosphate into 1,3-biphosphoglycerate. Lactate that enters the cell via MCT1 bypasses the glycolytic pathway and is directly converted to pyruvate. Pyruvate enters the mitochondria where it is processed through the TCA cycle to generate NADH to fuel the ETC. 6-aminonicotinamide inhibits PPP by targeting G6PD. 2-deoxy D-glucose (not shown) inhibits glycolysis by acting as a competitive inhibitor for the production of G6P. 3-bromopyruvate and iodoacetate inhibits glycolysis by targeting HK2 and GAPDH, respectively. Antimycin A blocks complex III activity, thus shutting down the TCA cycle and ETC. (B) P2 Differentiated (30% density) and (C) dedifferentiated (2% density) hfRPE cells cultured over 21 days in 96-well plate were switched to glucose-free Ringer’s solution (100 µL/well) and incubated overnight at 37°C and 5% CO₂. The next day, a mixture containing 10 µL of a 13X glucose, lactate, or pyruvate stock + 20 µL of 6X Biolog reagent was injected into each well (n = 5 each). Absorbance (AU) was measured at 590 nm over 2 hr at 30 min intervals and the rate of increase in absorbance (ΔAU/hr, which directly correlate with NADH and NADPH production rate) was calculated for each well. Glucose metabolism of (D) differentiated and (E) dedifferentiated hfRPE cells was evaluated in the presence of 2-deoxy-D-glucose, Na-iodoacetate, 3-bromopyruvate, and 6-aminonicotinamide (n = 5 for each condition). In these experiments, 5 µL of a 25X inhibitor stock was manually pipetted into each well, mix by tapping the plate. 30 min later, 20 µL of glucose + BiologReagent mixture (1.25 µL of a 0.5 M glucose stock + 18.75 µL of the 6X Biolog reagent) was injected into each well using a repeater pipette. Rate of increase in absorbance at 590nm was measured as previously described. Data is presented as % metabolism relative to glucose control (without any inhibitor; 100%) and Ringer control (no substrate; baseline reads set to 0%). Statistical analysis was performed using students t-test, p < 0.05 is considered statistical significant.

Figure 6. Differentiated and dedifferentiated hfRPE cells metabolize glucose and pyruvate differently

P2 differentiated (30%) or dedifferentiated (2%) hfRPE cells on 96-well plate cultured over 21 days were switched to glucose-free Ringer (containing 1X ITS and 5% HI FBS) and incubated overnight at 37°C and 5% CO₂. On the day of the experiment, cells were pre-treated with 100 nM antimycin A (ETC complex III inhibitor) for 15 min before a 20 µL mixture of (A) glucose (5 mM) or (B) pyruvate (5 mM final conc.) + 6X Biolog reagent was added to each well (1.25 µL of 0.5 M substrate stock + 18.75 µL of 6X Biolog reagent). Absorbance at 590nm was obtained every 30 min over 2hr and the NADH and NADPH production rate was calculated.