Nuclear Gene Expression Changes Due to Mitochondrial Dysfunction in ARPE-19 Cells: Implications for Age-Related Macular Degeneration

Abstract

purpose. To measure changes in nuclear gene expression resulting from mitochondrial dysfunction in retinal pigment epithelial cells.

methods. ARPE-19 retinal pigment epithelial cells were depleted of their mitochondrial (mt)DNA by passaging in a low concentration of ethidium bromide. Loss of mitochondrial DNA was determined by uridine auxotrophy and quantitative real-time polymerase chain reaction of isolated DNA. Loss of mitochondrial membrane potential was estimated by uptake of JC-1. Changes in nuclear gene expression were determined by quantitative real-time reverse transcription–polymerase chain reaction of isolated total RNA from ethidium-bromide–treated and untreated cells. Morphologic and phenotypic changes were determined by phase-contrast microscopy, sensitivity to the oxidant tert-butyl hydroperoxide (tBH), and invasion assay.

results. ARPE-19 cells became auxotrophic for growth on uridine after eight passages in 50 ng/mL ethidium bromide. Quantitative PCR revealed almost complete loss of mitochondrial DNA (ρ⁰ cells). Uptake of JC-1 was reduced in the ρ⁰ cells, indicating reduction of mitochondrial membrane potential. Quantitative RT-PCR measured increased expression of genes coding for drusen components, lipid transport, extracellular matrix components, and responses to inflammation in the ρ⁰ cells. The ρ⁰ cells also exhibited an increased sensitivity to killing by tBH and increased migration and invasion through solubulized basement membrane–coated tissue culture inserts.

conclusions. ARPE-19 cells respond to loss of mitochondrial function by changes in nuclear gene expression that resemble changes observed in age-related macular degeneration. The results lead to the hypothesis that loss of mitochondrial function with age and resultant changes in nuclear gene expression may explain some of the changes in the macula that are associated with the known clinical manifestations of age-related macular degeneration.

The retinal pigment epithelium (RPE) is a single layer of cells in the vertebrate retina, and its continuous function has been shown to be necessary for vision. ^{1 2} The RPE is a terminally differentiated tissue that ages in a unique fashion and in which changes with age have been implicated in the development of age-related macular degeneration (AMD), ^{3 4} the leading cause of untreatable vision loss in persons older than 65 years in developed countries. ^{5 6} Indeed, it has been estimated that at least 8 million people ≥65 years of age in the United States are at high risk for the development of advanced AMD. ⁷  

Some of the changes that have been shown to occur in the macula and that correlate with AMD include increased deposition in the RPE of the age pigment lipofuscin, ^{8 9} the development of drusen between the RPE basal lamina and Bruch’s membrane, ^{10 11} increased sterol deposition in the elastin layer of Bruch’s membrane, ¹² thickening of Bruch’s membrane and other basement membrane changes, ^{13 14} choroidal neovascularization, ^{15 16} and RPE cell apoptosis. ¹⁷ Although the clinical manifestations of RPE aging are well described and several theories have been proposed to explain the cause of AMD, none of these theories has led to a satisfactory understanding of this unique aging phenotype. 

We have been working to understand how changes in mitochondrial function contribute to aging phenotypes in eukaryotic cells. In one model of eukaryotic mitochondrial dysfunction, yeast cells that have been depleted of their mitochondrial (mt)DNA (ρ⁰ cells) respond to loss of mitochondrial function with induction of a specific pathway that induces changes in nuclear gene expression. ¹⁸ These changes, called retrograde regulation or retrograde response, respond to this loss in a way that not only compensates for the loss of respiratory function but also results in increased replicative lifespan in cells grown on fermentable substrates. ^{19 20} Furthermore, our laboratory has demonstrated that mitochondrial dysfunction is a normal consequence of yeast aging and that the retrograde response increases with replicative age and titrates the progressive loss of mitochondrial function to enhance longevity. ²¹  

Mammalian cells also respond to loss of mitochondrial function with changes in nuclear gene expression. It has been shown that the mitochondrial-to-nuclear stress signal is probably due to a decrease in mitochondrial membrane potential, which results in increased cytosolic free [Ca]²⁺. ²² The increased in [Ca]²⁺ in turn activates Ca²⁺-sensitive protein phosphatases, kinases, and calcium-calmodulin kinase IV (CaMKIV), which activate transcription factors, resulting in changes in nuclear gene expression. ²³ Although these changes in nuclear gene expression seem to be necessary for cell survival, we have shown that the specific genes induced are common to several cell types or are cell-type specific and appear to be dependent on the differentiated functions of the cell. ²⁴  

In this study, we attempted to establish a link between the loss of mitochondrial function in RPE cells and changes that contribute to the development of AMD, by investigating changes in nuclear gene expression in ARPE-19 cells brought on by loss of mtDNA (ρ⁰ARPE-19). These cells contain mitochondria but cannot respire, because of the absence of the subset of 13 respiratory transport chain proteins that are coded for by the mitochondria. We have previously reported that ρ⁰ARPE-19 cells induce the transcription of glycolytic pathway genes to maintain energy production, genes to maintain cellular redox potential and mitochondrial membrane potential, and genes that may allow for increased stress resistance and protect the cell against cell death. ²⁴ Furthermore, there is evidence in the literature that a reduction in mitochondrial function may contribute to the development of AMD. ^{25 26} Therefore, we hypothesized that changes in RPE nuclear gene expression due to loss of mitochondrial function may help explain some of the phenotypic changes in aged RPE cells that lead to the clinical manifestations in AMD. In this study, we more thoroughly characterized ρ⁰ARPE-19 cells as a model system in which to study loss of mitochondrial function and demonstrate some specific changes in nuclear gene expression and in cell phenotype due to mitochondrial dysfunction. 

Materials and Methods

Cell Culture

The retinal pigment epithelial cell line ARPE-19 was obtained from ATCC (Manassas, VA). ARPE-19 is an established cell line that has been shown to express many of the characteristics of RPE cells in culture. ²⁷ The cells were grown in DMEM/Ham’s F12 1:1 (Invitrogen-Gibco, Grand Island, NY) containing 24 mM sodium bicarbonate, 10% FBS, 10 μg/mL uridine, and 1.0 mM sodium pyruvate. All cells were maintained in a 5% CO₂ humidified atmosphere at 37°C and grown without added antibiotics or fungicides. 

The ARPE-19 cells were depleted of their mtDNA with ethidium bromide (EtBr). ²⁸ ARPE-19 cells were grown in T75 flasks and then passaged 1:3 into medium containing 50 ng/mL EtBr. Cells were grown until confluent and then serially passaged in medium containing EtBr. The cells were screened for loss of mtDNA by passaging into medium without added uridine containing 10% dialyzed FBS. Cells that have lost their mtDNA become auxotrophic for growth on uridine due to the loss of functional dihydroorotate dehydrogenase, which is located on the outer surface of the mitochondrial inner membrane. After five to seven passages, the ARPE-19 cells were unable to grow without added uridine. Loss of the mitochondrial genome was confirmed by quantitative real-time polymerase chain reaction (PCR), with a primer set that spanned the origin of replication of the mitochondrial heavy strand, as described later. Control cells were treated in an identical fashion, except that EtBr was omitted from the medium. 

Nucleic Acid Isolation and Quantitative Real-Time RT-PCR

DNA and RNA were isolated from trypsinized cells with isolation kits (DNeasy or RNeasy; Qiagen, Valencia, CA). RNA was isolated with on-column DNase I digestion from recently confluent cells that had been fed 6 hours earlier. Final yields and concentrations were calculated by measuring the absorbance at 260 nm in 10 mM Tris, 1 mM EDTA (pH 7.5; TE buffer). 

The genes of interest were chosen because their protein products have been identified as components of drusen, because they have been shown to be upregulated in RPE or retina of individuals with AMD, because of their association with extracellular components of Bruch’s membrane, or because of their association with outer segment phagocytosis or lipid transport into the RPE. Table 1is a list of the genes analyzed and their descriptions. 

Real-time quantitative RT-PCR was performed on a commercial system (Prism 7000 Sequence Detection real-time PCR system; Applied Biosystems, Inc. [ABI], Foster City, CA) in a two-step reaction. Total RNA was reverse transcribed in a 50-μL volume with the RT reagents (ABI). Each reaction contained 1 μg of RNA, 5 μL of 10× RT buffer, 5.5 mM MgCL₂, 2 mM dNTPs, 2.5 μM random hexamers, 20 U RNase inhibitor, and 62.5 U reverse transcriptase (MultiScribe; ABI). For real-time PCR, primers for the genes of interest were designed with commercial software (Primer Express; ABI) and purchased from Integrated DNA Technologies, Inc. (Coralville, IA). In general the primer pair with the lowest penalty score identified by the software was chosen. All primer pairs were check by a BLAST search against the human mRNA database, to assure homology only with the gene of interest. In cases in which there was more than one transcript variant, primer pairs were chosen that mapped to all known variants. Each 25-μL reaction contained 0.5 μL of the RT reaction as template, 0.2 μM each of the forward and reverse primers, and 12.5 μL 2× PCR mix (SYBR Green PCR Master Mix; ABI). All reactions were performed in triplicate and consisted of 1 cycle of 50°C for 2 minutes, 1 cycle of 95°C for 10 minutes, and 40 cycles of 95°C for 15 seconds and 60°C for 1 minute (two-step PCR). The appropriate housekeeping gene to use as a control for DNA loading was determined by screening a set of commonly used housekeeping genes (β-actin [ACTB] peptidylprolyl isomerase A [PPIA], glyceraldehyde-3-phosphate dehydrogenase [GAPDH], and ribosomal protein S9 [RPS9]) and choosing the gene that showed no significant difference in expression between control and experimental cells. With this method, we chose PPIA as the control gene. Standard curves were generated from dilutions of the cDNA template obtained from ARPE-19 cells, with primers for PPIA, and relative concentrations of the genes of interest were calculated from the standard curves after normalization with the housekeeping gene. In some cases, dilutions of the cDNA templates were analyzed with primer pairs of the genes of interest, to be certain that the genes of interest were amplified similarly to the housekeeping genes. 

Estimation of Mitochondrial Membrane Potential

ARPE-19 cells were plated on two-well chamber slides (InterMed, Naperville, IL) and grown until confluent. The cells were incubated with 10 nM of the lipophilic dye JC-1 in complete medium for 10 minutes at 37°C, ²⁹ and either rinsed with PBS, coverslipped, and observed under a fluorescence microscope or trypsinized and analyzed in a flow cytometer. For microscopy the cells were observed by deconvolution microscope equipped with a mercury lamp and filters for epifluorescence applications (model DMRXA; Leica, Deerfield, IL). Digital images were then obtained (Sensicam QE 12-bit, cooled CCD system; PCO AG, Kelheim, Germany; and SlideBook image-processing software; Intelligent Imaging Innovations, Santa Monica, CA). Alternatively, the cells were analyzed with a flow cytometer (EPICS; Beckman Coulter, Inc., Fullerton, CA) and the data collected in both the FITC and rhodamine channels. 

Cell Viability MTT Assay

Cell viability after oxidative stress was assessed with 3-(4,5-dimethylthiazol-2,5-diphenyl tetrazolium bromide (MTT). Cells plated in 24-well plates were grown to confluence and treated with 0 to 2000 μM tert-butyl hydroperoxide (tBH), for 6 hours in the absence of serum. After treatment, the cells were fed with growth medium and maintained overnight. The next day, the cells were incubated with DMEM without phenol red, containing 5% FBS and 1.0 mM MTT for 2 hours at 37°. Afterward, the cells and the formazan MTT reduction product were solubilized with 0.1 g/mL SDS in 0.01 M HCl and read at 570 nm. 

Cell Invasion Assay

Cell invasion assays were performed in growth-factor–reduced, solubulized basement membrane (Matrigel)–coated, 24-well invasion chambers (BD Biosciences, Franklin Lakes, NJ). The precoated wells were rehydrated by incubation with 500 μL of culture medium in the lower chamber, and then 5 × 10⁴ cells in 300 μL of medium were added to the upper chamber. The wells were incubated at 37°C for 48 hours under normal culture conditions. The medium was removed, and the cells adhering to the upper side of the membrane were removed by scrubbing three times with a cotton swab moistened with culture medium. The cells remaining on the underside of the membrane were then fixed with 0.5 mL 2% formaldehyde and 0.2% glutaraldehyde in PBS for 10 minutes. After they were rinsed, the cells were stained with 0.5 mL hematoxylin, counterstained with eosin, and rinsed with distilled water. The number of cells that migrated to the underside of the membrane was determined with an inverted microscope. In all, the assay was performed in duplicate two times, and three fields per insert were counted and averaged. 

Statistical Analysis

Testing of statistical significance was performed on computer (StatMost; DataMost Corp., Sandy, UT). For the real-time RT-PCR data, Student’s paired t-test was performed on the raw values corrected for DNA loading as described earlier. P ≤ 0.05 was taken as the criterion of significance. 

Results

The EtBr treatment of ARPE-19 cells resulted in a morphologic change in the cells and an almost complete loss of the mtDNA. Figure 1Ashows photomicrographs of ARPE-19 cells (ρ⁺) and ARPE-19 cells that had been treated with EtBr for eight passages (ρ⁰). In general ρ⁰ARPE-19 cells grew more slowly, were unable to grow without added uridine, and appeared to spread out more. Partial characterization of the ρ⁰ARPE-19 cells has been reported. ²⁴  

Figure 1Bshows an amplification plot of a real-time quantitative PCR analysis of DNA isolated from ρ⁺ARPE-19 and ρ⁰ARPE-19 cells. The primers were chosen to amplify a region of the mtDNA that spans the origin of replication of the mtDNA heavy strand. Quantitative PCR showed mtDNA at <2.5 × 10⁻⁴ that of control cells. 

We estimated mitochondrial membrane potential in the ρ⁰ARPE-19 cells by using the lipophilic cationic fluorescent probe JC-1. Transport of JC-1 into mitochondria has been shown to be dependent on the mitochondrial membrane potential, and the fluorescence of JC-1 has been shown to change from green to red due to the formation of J aggregates at higher membrane potentials. Figure 2shows fluorescence microscope images of ρ⁺ARPE-19 and ρ⁰ARPE-19 cells that were treated with JC-1. The micrographs demonstrate that the ρ⁰ARPE-19 cells had reduced uptake of the dye, which indicates a lower mitochondrial membrane potential than in the untreated ρ⁺ARPE-19 cells. This result was confirmed by flow cytometry. 

When we measured changes in nuclear gene expression resulting from loss of mitochondrial function in a select set of genes, the results were dramatic. We arranged this select set of genes in four groups, and the results are shown in Figure 3(see Table 1for accepted gene names). 

Group 3A contained genes that code for proteins that have been shown to be involved in outer segment phagocytosis or to be components of drusen. There was an increase in expression of ITGAV, ITGB5, CLU, and FBLN and a decrease in expression of VTN in ρ⁰ARPE-19 cells compared with ARPE-19 cells. 

Group 3B contained genes that code for components of the extracellular matrix. We measured a large increase in expression of COL1A1, COL5A1, MMP2, and PLAU; modest increased expression of PLAUR and LOX; and decreased expression of TIMP3. 

Group 3C contained genes associated with lipid transport in the RPE. We noted a large increase in APOE expression, and increased expression of VLDLR, SCARB1, SCARB2, and CD68. 

Group 3D contained genes associated with a response to inflammation or oxidative stress. We noted a large increase in expression of VEGF and increases in TGFB1, FGF5, CD97, MGST1, EPHX1 and SOD2. 

Because oxidative stress is considered one of the contributing factors in AMD, we also were interested in whether ρ⁰ARPE-19 cells were more or less sensitive to oxidative stress than their ρ⁺ counterparts. Figure 4shows the results of a cell viability assay on cells treated with the oxidant tBH. These results show that the ρ⁰ARPE-19 cells are much more sensitive to the oxidant than are the control cells, indicating that the ρ⁰ARPE-19 cells are more sensitive to oxidative stress than their ρ⁺ counterparts. 

Finally, because pigmentary changes due to clumping and migration of the RPE are characteristic of AMD and because of the observed changes in extracellular matrix and matrix metalloproteinase gene expression observed in this study, we were interested in whether we could demonstrate a change in invasive phenotype in our ρ⁰ cells. Figure 5shows the result of a migration assay using solubulized basement membrane–coated tissue culture inserts. The results show an approximate sixfold increase in the invasive properties of the ρ⁰ cells compared with the control cells. 

Discussion

The results show ρ⁰ARPE-19 cells had a lower mitochondrial membrane potential than the ρ⁺ARPE-19 cells. It has been suggested that it is the reduced mitochondrial membrane potential in ρ⁰ cells and the resultant increase in intracellular free [Ca]²⁺ that activates Ca²⁺-sensitive kinases, resulting in changes in nuclear gene expression. ²²  

Furthermore, ρ⁰ARPE-19 cells upregulated several nuclear genes known to be associated with an aging RPE phenotype and/or with AMD. Some of the changes in nuclear gene expression in response to the loss of mitochondrial function are most likely an attempt to compensate for the energy deficit. This may explain the upregulation of genes for proteins involved in lipid transport such as APOE and VLDLR, and increased expression of the genes for the phagocytosis receptors ITGAV and ITGB5. This may be because the RPE, in contrast to the neural retina, apparently derives a large portion of its energy needs from mitochondrial fatty acid β-oxidation. ^{54 55} Because mitochondrial fatty acid β-oxidation is normally linked to electron transport, nonmetabolized lipids would have to be excreted or they would accumulate intracellularly. This could account for the increase of esterified lipid shown to accumulate in basal linear deposits in AMD. The cell may also increase expression of VEGF and other angiogenic factors to attempt to compensate for low energy levels, in a fashion similar to tumor cells, which upregulate angiogenic cytokines as a response to hypoxia. This upregulation of VEGF in RPE cells would contribute to the choroidal neovascularization in advanced AMD. Other upregulated genes may be due to a state of chronic inflammatory response brought on by increased cellular oxidative stress. Increased oxidative stress was evidenced by the increased sensitivity of the ρ⁰ARPE-19 cells to an oxidant challenge (Fig. 4)and by the increased expression of SOD2, EPHX1, and MGST1 (Fig. 3D) . Polymorphisms in SOD2 and EPHX1 have also been shown to be associated with AMD. ⁵²  

Although the effect of increased gene expression on protein translation was not measured, it is probably not coincidental that genes coding for proteins that have been identified as components of drusen were shown to be upregulated. This includes APOE, CLU, and FBLN. Vitronectin gene expression, the product of which is found in high concentrations in drusen, was shown to be markedly decreased in ρ⁰ARPE-19 cells. However, it has been shown that vitronectin is abundantly expressed in the neural retina as well as the RPE, ⁵⁶ and the vitronectin receptor genes ITGAV and ITGB5 were upregulated in ρ⁰ARPE-19 cells. Therefore, vitronectin may be deposited in drusen as nonmetabolized receptor-vitronectin complexes. A more complete picture of the effect of mitochondrial dysfunction on cellular metabolism awaits further studies that should include measurements of changes in protein expression and enzyme activity. 

In our study, ρ⁰ARPE-19 cells had a more invasive phenotype that was the result of increased expression of inflammatory cytokines such as TGFβ and of changes in ECM and MMP gene expression. These changes could explain observations of pigmentary changes in the maculas of patients with AMD and the observation that RPE cells have been identified as components of choroidal neovascular (CNV) membranes. ⁵⁷ RPE cells in CNV membranes have also been reported to be immunoreactive for VEGF which correlates with the findings in the current study. Because other genes that are associated with an inflammatory response are also upregulated in ρ⁰ARPE-19 cells, this VEGF immunoreactivity may indicate that these cells contribute to a chronic low level of inflammation in the retina and Bruch’s membrane. There is ample evidence that drusen deposition in the RPE-choroid is associated with an inflammatory response. ⁵⁸  

The question is, what is the evidence of mitochondrial dysfunction in the aging human RPE cell? mtDNA deletions are clearly associated with human disease. This diverse group of diseases collectively termed mitochondrial myopathies include Kearns-Sayre syndrome (KSS); chronic progressive external ophthalmoplegia (CPEO); mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episode (MELAS); myoclonic epilepsy and ragged-red fibers (MERRF); and Leber hereditary optic neuropathy (LHON). These diseases are generally fatal, and it is the brain and muscles that are the tissues principally affected. However, it has also been reported that the retina, particularly the macular RPE, is also likely to be involved in these diseases. ^{59 60}  

We have demonstrated that mitochondrial dysfunction is a normal consequence of aging in yeast. ²¹ The loss of mitochondrial function with age has been postulated to contribute to aging in higher eukaryotes as well, although definitive demonstration of widespread mitochondrial dysfunction with age has not been shown, perhaps because of the large number of mtDNA mutations known and the difficulty of quantifying all potential mtDNA mutations within a single sample. It has been demonstrated that mutations in mtDNA occur with aging in postmitotic tissues and can reach high levels in old individuals, ⁶¹ especially in the control region responsible for mtDNA replication and transcription. ⁶²  

The RPE mtDNA is particularly susceptible to oxidative damage. It is terminally differentiated, and damaged cells cannot be replaced by mitosis, whereas damage to the mtDNA can be propagated as the mitochondria in the cells divide. The RPE has a high metabolic demand and may produce high levels of oxidants in the mitochondria due to β oxidation of fatty acids. We showed an increase in expression of SOD2 in our ρ⁰ARPE-19 cells, suggesting increased superoxide production in the mitochondria. It has been reported that SOD2 activity increases with donor age in freshly isolated RPE cells, ⁶³ and there is one report of an association of SOD2 polymorphism with exudative AMD. ⁵² Although there has been speculation as to how oxidative damage to mitochondria may lead to an RPE energy deficit or to RPE apoptotic or necrotic cell death, there has been no previous consideration of the response of nuclear gene expression to the mitochondrial dysfunction caused by oxidative damage to protein or DNA and how changes in nuclear gene expression may relate to clinical manifestations of AMD. 

We postulate that accumulation of oxidative damage in mtDNA is the primary event leading to mitochondrial dysfunction and an inflammatory phenotype in the macula, but it is also possible that other cellular events could lead to drusen deposition that could precipitate an inflammatory response in the RPE or Bruch’s membrane. The inflammatory response in turn could lead to the accumulation of toxic proteins such as β-amyloid which may be targeted to the mitochondria and also precipitate changes in nuclear gene expression. It has been recently reported that amyloid-P may be targeted to the mitochondria, resulting in mitochondrial dysfunction and loss of calcium homeostasis. ⁶⁴ Because there is ample evidence of calcium deregulation in Alzheimer’s disease, a similar scenario could be envisioned for the RPE-choroid, although the contribution of inflammatory protein deposition to mitochondrial dysfunction in the RPE must be demonstrated. 

In Figure 6we present a schematic representation of our hypothesis that the manifestations of AMD can be explained by damage to the mtDNA of the RPE, resulting in mitochondria containing dysfunctional electron transport chains. This dysfunction results in mitochondria with a lower mitochondrial membrane potential and increased superoxide production and in release of Ca²⁺ into the cytoplasm. This in turn upregulates calcium-dependent kinases and nuclear transcription factors, which upregulate genes that aid in the survival of the RPE. However, although these changes in gene expression may contribute to RPE survival, they can also contribute to a pathologic state in the retina. 

 Supported by grants from the National Institute on Aging of the National Institutes of Health.

 Submitted for publication November 12, 2004; revised January 8, 2005; accepted January 24, 2005.

 Disclosure: M.V. Miceli, None; S.M. Jazwinski, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Accepted Names of Genes Analyzed, Descriptions, and References

Gene	Description	References
Genes coding for proteins involved in ROS uptake or components of drusen
ITGAV	Integrin, alpha V (vitronectin receptor)	30 , 31
ITGB5	Integrin, beta 5	30 , 31
CLU	Clusterin (apolipoprotein J)	32
FBLN	Fibulin 3 (EFEMP1)	33
VTN	Vitronectin	34
Genes coding for components of the extracellular matrix
COL1A1	Collagen, type I, alpha 1	35 , 36
COL5A1	Collagen, type V, alpha 1	35 , 36
MMP2	Matrix metalloproteinase 2 (gelatinase A)	37
PLAU	Plasminogen activator, urokinase	38 , 39
PLAUR	Plasminogen activator, urokinase receptor	38 , 39
LOX	Lysyl oxidase	40
TIMP3	Tissue inhibitor of metalloproteinase 3	41
Genes coding for proteins associated with lipid transport
APOE	Apolipoprotein E	42 , 43
VLDLR	Very-low-density lipoprotein receptor	44
CD68	Scavenger receptor class D, member 1 (SCARD1)	45
SCARB1	Scavenger receptor class B, member 1	46
SCARB2	Scavenger receptor class B, member 2	46
LDLR	Low-density lipoprotein receptor	47
Genes associated with responses to inflammatory or oxidative stress
TGFB1	Transforming growth factor, beta 1	48
FGF5	Fibroblast growth factor 5	49
CD97	CD97 antigen	50
VEGF	Vascular endothelial growth factor	51
EPHX1	Epoxide hydrolase 1, microsomal	52
MGST1	Microsomal glutathione S-transferase 1	53
SOD2	Superoxide dismutase 2, mitochondrial	52

Figure 1.

 

Characterization of ρ⁰ARPE-19 cells. (A) Phase-contrast photomicrographs showing control ρ⁺ARPE-19 cells and ρ⁰ARPE-19 cells at confluent density. ρ⁰ARPE-19 cells were more spread out and grew to a lower confluent density than did control cells. (B) Quantification of mtDNA in ρ⁰ARPE-19 cells. Amplification plots from a quantitative real-time PCR of DNA extracted from ARPE-19 cells and ARPE-19 cells that had been treated with 50 ng/mL EtBr (ρ⁰ARPE-19). The plot is a graph of the baseline-corrected reaction fluorescence (Delta Rn) versus the cycle number. The primer pair was chosen to amplify a product spanning the origin of replication of the mitochondrial genome. Each sample was run in triplicate and contained 10 ng of total DNA. ρ⁰ARPE-19 cells contained 2.5 × 10⁻⁴ the amount of mitochondrial DNA compared with the untreated control cells. Bars, 175 μm.

Figure 2.

 

Fluorescence images of cells on chamber slides that were treated with the lipophilic dye JC-1. (A) ρ⁺ARPE-19 cells observed with the FITC filter set. (B) The same field observed with the Texas red filter set, showing the formation of J aggregates in the mitochondria. (C) ρ⁰ARPE-19 cells with the FITC filter set. (D) The same field observed with the Texas red filter set, showing very little J aggregate formation. The ρ⁰ARPE-19 cells took up less dye than the untreated control cells, demonstrating reduced mitochondrial membrane potential. Bars, 20 μm.

Figure 3.

 

Changes in gene expression determined by quantitative real-time RT-PCR. Shown is the percentage change in gene expression of the experimental (ρ⁰) cell population compared with the normal control (ρ⁺). Data are the average of three to four determinations from at least three RNA isolations. Gene groups: (A) those coding for proteins involved in ROS uptake of components of drusen; (B) those coding for components of the extracellular matrix; (C) those coding for proteins associated with lipid transport; and (D) those associated with responses to inflammatory or oxidative stress. Gene descriptions are given in Table 1 . Error bars, ±SEM. Statistical significance was calculated from raw data normalized to cyclophilin A as the housekeeping gene. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Figure 4.

 

Cell viability assay for cells treated with tBH for 6 hours and assayed after 24 hours with MTT. Results are the average of six determinations ± SEM. The ρ⁰ARPE-19 cells were much more susceptible to the oxidant than were the ρ⁺ARPE-19 cells.

Figure 5.

 

RPE cell invasion assay. (A, B) Images of fixed and stained cells that had migrated to the underside of solubulized basement membrane–coated tissue culture inserts. (C) Summary graph of the photomicrograph data for the two cell types. Data are the average number of cells per field from two separate experiments performed in duplicate. Three fields per insert were counted. Data are the mean ± SEM. The ρ⁰ARPE-19 cells showed a more invasive phenotype than did the ρ⁺ARPE-19 cells. Bars, 55 μm.

Figure 6.

 

Schematic representation of a hypothesis of how mitochondrial dysfunction could lead to a RPE phenotype, resulting in AMD. In this view, accumulated damage to mtDNA results in a defective electron transport chain due to an imbalance of transport chain components. This defect causes lower mitochondrial membrane potential (Δψ_m), increased superoxide production, and increased cytosolic [Ca²⁺]. The resultant increase in Ca²⁺ concentration activates calcium-dependent transcription factors that signal a nuclear response in an attempt to compensate for loss of energy production and to increase RPE cell viability. These gene expression changes contribute to some of the manifestations in AMD, such as RPE cell migration angiogenesis and lipid deposition. The excess superoxide production also contributes to an increased level of inflammation in the retina and Bruch’s membrane.