Induction of an Aging mRNA Retinal Pigment Epithelial Cell Phenotype by Matrix-Containing Advanced Glycation End Products In Vitro

Abstract

purpose. To determine an extensive mRNA phenotype of the established RPE cell line ARPE-19 when grown on a matrix modified by advanced glycation end products (AGEs).

methods. Growth Factor Reduced Matrigel (Collaborative Biomedical Products, Bedford, MA) was nonenzymatically glycated with glycolaldehyde. ARPE-19 cells were seeded on both AGE-Matrigel and Matrigel and grown to confluence, and serum was withdrawn for 3 days. RNA was extracted, and microarray analysis was performed to characterize the genes, which are altered by a matrix modified by AGEs. Gene expression changes were confirmed by RT-PCR/Southern and Northern blot analysis. Apoptosis was measured by annexin V/propidium iodide labeling.

results. Clusters of genes with altered expression were found related to cell differentiation, growth factors that regulate the RPE cell and basement membrane, and apoptosis. RT-PCR/Southern and Northern blot analysis confirmed the expression patterns of selected genes, and flow cytometry showed increased annexin V/propidium iodide-labeled cells when grown on AGE-Matrigel.

conclusions. Microarray analysis identified clusters of genes that could promote an aging RPE phenotype in vitro induced by a matrix modified with AGEs.

Advanced glycation end products (AGEs) are a heterogeneous group of structures formed over time by nonenzymatic Maillard reactions between a protein’s primary amino group and carbohydrate-derived aldehyde groups. AGEs have been implicated in a variety of age-related diseases such as cataract, Alzheimer’s disease, and atherosclerosis. ^{1 2 3} Our laboratory recently quantified an age-dependent increase of AGEs in human Bruch’s membranes and identified AGEs in basal deposits and drusen. ^{4 5} We also determined that physiological doses of the AGE pentosidine upregulated PDGF-B chain in RPE cells, which suggests that AGEs can influence the RPE phenotype. ⁶ Although providing the rationale for further investigation, this study was limited by characterizing the expression of only one gene. We hypothesize that AGEs induce an aging mRNA phenotype to the RPE that would promote degeneration to the RPE-Bruch’s membrane complex. Our long-range goal is to identify an extensive mRNA phenotype of the RPE in health and aging in vivo and to determine the subset of genes that are regulated by AGEs. The emergence of cDNA microarrays allows for the simultaneous examination of the expression of large sets of genes that could characterize the molecular events during aging and define a role for AGEs during RPE aging. As initial work to address our hypothesis, we used microarray analysis to characterize an extensive mRNA phenotype of the established, nonimmortalized ARPE-19 cell line when grown on matrix containing AGEs. 

Materials and Methods

Nonenzymatic Glycation of Matrigel

A 1:16 dilution of Growth Factor Reduced Matrigel (Collaborative Biomedical Products, Bedford, MA), which in preliminary experiments did not impair the proliferation of ARPE-19 cells, was incubated for 4 hours at 37°C with glycoaldehyde (50 mM; Sigma, St. Louis, MO) in 0.2 M phosphate buffer using a modification of a previously published protocol. ⁷ Control Growth Factor Reduced Matrigel was treated for 4 hours with 0.2 M phosphate buffer without glycolaldehyde. The mixtures were then rinsed extensively with Dulbecco’s phosphate-buffered saline (PBS). A sample was analyzed for fluorescence with a Hitachi F-2000 fluorescence spectrometer using 370 nm excitation/440 nm emission to estimate the quantity of AGEs, a typical indicator of AGE formation. ⁸ In this study, we designate “AGE-MG” to indicate glycolaldehyde-treated Growth Factor Reduced Matrigel and “MG” for Growth Factor Reduced Matrigel. 

Cell Culture

All experiments used the ARPE-19 established but nonimmortalized human RPE cell line. This cell line exhibits significant differentiated characteristics such as cobblestone morphology, apical microvilli, basolateral infoldings, and expression of CRALBP and RPE-65. ⁹ The routine maintenance of ARPE-19 has been previously described. ⁶ Preliminary experiments, such as the work of Haitoglou et al., ¹⁰ revealed that the attachment of cells onto AGE-matrix is reduced by 25%. To obtain equal confluency for experiments, ARPE-19 cells were seeded in T-75 cm² flasks at a density of 80,000/cm² when plated on MG and 100,000/cm² when plated on AGE-MG and grown for 1 day in Dulbecco’s modified Eagle’s medium/Nutrient mixture F12 with 15 mM Hepes buffer (DMEM/F12; Gibco BRL, Gaithersburg, MD) + 10% fetal bovine serum (FBS; UBI Upstate, Lake Placid, NY), 0.348% additional sodium bicarbonate, 2 mM l-glutamine solution (Gibco BRL), at 37°C in 10% CO₂. Cells were then rendered quiescent in DMEM/F12 + 1% bovine albumin (BSA fraction V; Sigma). 

RNA Extraction for Microarray Analysis

Total RNA was extracted using the RNeasy Midi kit (Qiagen Inc., Valencia, CA) twice according to the manufacturer’s recommendations and treated with DNase (Amplification grade; Gibco BRL). PolyA⁺ RNA was isolated with the Oligotex mRNA kit (Qiagen Inc.) after the quality of total RNA was assessed by subjecting a sample of RNA to 1% agarose gel electrophoresis and to verify that it was free of DNA contamination by PCR amplification using protein kinase 1 and 2 primers. 

Microarray Analysis

The Atlas Human cDNA expression (588 genes) and Human Cell Interaction (265 genes) Microarrays (Clontech Laboratories, Inc., Palo Alto, CA) were used according to the manufacturer’s protocol. Briefly, ³²P-labeled first-strand cDNA probes were synthesized from 1 μg polyA⁺ RNA using Superscript II RNase H- (Gibco BRL, Grand Island, NY) in the master mix, which included [α-³²P]dATP. Probes were purified from unincorporated ³²P-labeled nucleotides and small cDNA fragments by column chromatography, and the incorporated ³²P into the probe was verified by scintillation counting. After prehybridizing with ExpressHyb (Clontech) with sheared salmon testes DNA for 30 minutes at 65°C, radiolabeled cDNA probes at a final probe concentration of 0.5 to 2 × 10⁶ cpm/ml were hybridized to the array at 68°C overnight. The arrays were washed once with 2× SSC, 1% SDS solution for 30 minutes at 68°C, and twice with 0.1× SSC, 0.5% SDS at 68°C for 30 minutes. The relative abundance of the signals for each gene was quantified by phosphorimager analysis using ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA). Experiments were repeated once. 

Microarray Quantification and Statistical Analysis

The intensity of each gene on the array was quantified using a previously published method. ¹¹ Briefly, the signal intensity of each cDNA spot was corrected by subtracting the immediate surrounding background. The corrected intensities were normalized for each cDNA as follows: (corrected intensity/75th percentile value of the intensity of the mean value of the entire array) ×1000. To determine nonspecific hybridization, 75th percentile values were calculated from the individual averages of an “irrelevant” set of cDNAs on the array. The values were then analyzed using the Cluster and TreeView software (http://rana.stanford/software) using a previously published method. ¹²  

Northern Blot Analysis

Northern blot analysis for all samples was performed with random-primed ³²P-labeled cDNA probes using our previously published protocol. ⁶ Total RNA (10 μg) was subjected to electrophoresis through a 1.2% agarose gel in 6.6% formaldehyde, transferred to a nylon membrane, prehybridized for 2 hours at 42°C, and hybridized by addition of the denatured probe in 50% formamide, 5× SSC, 0.1% sodium dodecyl sulfate (SDS), 5× Denhardt’s solution, and 100 μg denatured salmon sperm DNA at 42°C for 15 hours. Blots were then washed twice in 0.1× SSC/0.1% SDS at room temperature for 5 minutes and in 0.1× SSC/0.1% SDS at 50°C for 2 hours, and the signal was quantified by phosphorimager analysis. Values were normalized using a 28S rRNA probe. Three independent experiments were conducted. Statistical significance was determined by the Mann–Whitney test, with P < 0.05 as significant. 

Reverse Transcriptase–Polymerase Chain Reaction/Southern Blot Analysis

First strand cDNAs were produced from total RNA using random primers and Sensiscript reverse transcriptase (RT; Qiagen) in a 20 μl RT-mix at 37°C for 75 minutes, as recommended by the manufacturer. The resulting cDNAs were amplified in a 50 μl volume using the primers and conditions listed in Table 1 . Polymerase chain reaction (PCR) products were visualized by agarose gel electrophoresis and ethidium bromide staining. After Southern blot transfer, confirmation of the PCR reaction fidelity was assessed using ³²P end-labeled internal oligonucleotide primers as described previously. ^{13 14} Signals were quantified by phosphorimager analysis, and values were normalized to GAPDH. Statistical significance was assessed by the Mann–Whitney test, with P < 0.05 as significant. Three independent experiments were performed. 

Annexin V/Propidium Iodide Assay

ARPE-19 cells were plated at 80,000/cm² on MG and 100,000/cm² on AGE-MG in DMEM/F12 + 10% FBS for 24 hours and then grown in DMEM/F12 + 1% BSA for an additional 3 to 7 days, as described above. Cells were then trypsinized and rinsed in 1× binding buffer, and 10⁵ cells were resuspended in 100 μl 1× binding buffer. Annexin V-FITC (5 μl) and propidium iodide (10 μl) were incubated in the dark for 15 minutes, as recommended by the manufacturer (BD PharMingen, Inc., San Diego, CA). Phosphatidylserine externalization and nuclear propidium iodide labeling were measured using FACS analysis with excitation wavelength at 488 nm. Quadruplicate samples of each condition were used for each experiment, and statistical significance was assessed by the Mann–Whitney test, with P < 0.05 as significant. Three independent experiments were conducted. 

Results

Morphologic Alterations of ARPE-19 Cells Grown on AGE-MG

Confluent ARPE-19 cells grown under serum withdrawn conditions on Growth Factor Reduced Matrigel (1:16 dilution) developed a regular cobblestone morphology that simulates the morphologic appearance of the RPE in vivo (Fig. 1A) . When incubated with 50 mM glycolaldehyde for 4 hours, a twofold increase in the fluorescence of AGE-MG was detected by spectrophotometric analysis. In contrast, confluent serum–withdrawn ARPE-19 cells grown on AGE-MG developed a spindly, irregular“ fibroblastic” like morphology after 10 days (Fig. 1B) . 

Altered mRNA Phenotype of ARPE-19 Cells Grown on AGE-MG

Anderson et al. ¹⁵ have shown that cells grown on AGE-matrix first show altered gene expression after 3 days in culture. To determine whether an extensive mRNA phenotype change also developed along with the above-described morphologic changes by matrix containing AGEs, the RNA from confluent ARPE-19 cells grown under the identical conditions described above was assessed for gene expression changes by microarray analysis 3 days after serum withdrawal. Using an expression ratio of ≥ 2 or ≤ 0.5 as an initial filter, a total of 52 genes were differentially regulated by AGE-matrix changes. All the relative changes in gene expression were between 0.3- to 0.5-fold and 2.0- to 3.1-fold. This study will focus on the expression profiles of genes that could be involved in cell differentiation, Bruch’s membrane regulation, and early apoptosis. Clusters of genes with fold changes outside of the initial filter are included in this analysis because of the inter-related function of specific genes. 

Altered Expression of Cell Differentiation Genes by AGEs

Our analysis revealed a cluster of genes whose expression pattern could alter cell differentiation, which included ligands, receptors, intracellular signal molecules, and transcription factors from the inter-related Wingless, Notch, Hedgehog, and EGF receptor signaling pathways (Table 2) . For example, the Wingless ligand WNT-10B and early intracellular mediators disheveled 1 and disheveled 3 were downregulated by ARPE-19 cells grown on AGE-MG. The Notch modulator Manic fringe was upregulated, whereas the Hedgehog pathway receptor Smoothened was downregulated in ARPE-19 cells grown on AGE-MG. EGF ligand expression (kidney EGF precursor, heparin-binding EGF-like growth factor, and EGF-like cripto protein CR1) was mildly upregulated by ARPE-19 cells grown on AGE-MG. Although the EGF receptor ERBB3 was downregulated, the EGF receptor HER4 and eps15, a substrate for EGF receptors, ¹⁶ were upregulated. Several transcription factors that regulate genes that promote cell differentiation, including GATA-2, GATA-3, and ID-3, were downregulated in ARPE-19 cells grown on AGE-MG, as listed in Table 2 . 

We performed RT-PCR/Southern blot analysis to confirm the expression pattern of WNT-10B, manic fringe, and smoothened using identical culture conditions as described above (Fig. 2) . WNT-10B and manic fringe showed a 70% downregulation (n = 3; P = 0.05) and a 20% upregulation (n = 3; P = 0.05), respectively, by ARPE-19 cells grown on AGE-MG. We found a 13% decrease in smoothened expression (n = 3; P = 0.2) by ARPE-19 cells grown on AGE-MG, although this difference was not significant. 

Altered Expression of Genes Involved in Bruch’s Membrane Regulation by AGEs

Growth factors that regulate the extracellular matrix were upregulated by RPE cells grown on AGE-MG, including connective tissue growth factor (CTGF; 3.1-fold induction), TGF-β2 (2-fold induction), and to a lesser extent, PDGF-B (1.3-fold induction). We confirmed the upregulation of CTGF (4.1 ± 1.4-fold, P = 0.037), TGF-β2 (2.3 ± 0.1-fold, P = 0.037), and PDGF-B (2.0 ± 0.4-fold, P = 0.012) three times by Northern blot analysis using identical growth conditions (Fig. 3) . The selectivity of an AGE-induced upregulation of TGF-β2 is suggested by lack of TGF-β1 and -β3 expression by both microarray and Northern analyses (data not shown). In addition, the microarray experiments identified upregulation of Stem cell factor (2-fold), PAI-1 (2.1-fold), and the elastin protease myeloblastin (2.0-fold) by ARPE-19 cells grown on AGE-MG. 

Induction of Early Pathway Apoptosis Genes by AGEs

A set of genes involved in apoptosis showed altered expression by ARPE-19 cells grown on AGE-MG, as shown in Table 3 . Our initial filtering found only Bcl-2 as significantly downregulated, but further analysis also identified 11 other apoptosis-related genes, mainly from the TNF-α superfamily, with mildly upregulated expression, and because of their functional relationship, they were included in our analysis. Several DNA repair enzymes are listed that are significantly downregulated, which could impair the cell’s ability to repair DNA damage during apoptosis. 

To determine whether functional evidence of early apoptosis was induced by cells grown on AGE-MG, ARPE-19 cells grown under the previously described identical conditions were assessed for externalization of phosphatidylserine in the plasma membrane as an early indicator of apoptosis with annexin V staining and loss of viability with nuclear propidium iodide staining. Flow cytometric analysis revealed no evidence of apoptosis by ARPE-19 cells 3 days after serum withdrawal, but by 7 days, increased annexin V and propidium iodide staining was present in cells grown on AGE-MG. As shown in Figure 4a representative experiment shows that the percentage of cells with evidence of early apoptosis but still viable cells (annexin V staining only) increased from 1.1% ± 0.34% when grown on MG to 4.5% ± 1.0% when grown on AGE-MG (n = 4, P = 0.016). In addition, the percentage of nonviable apoptotic cells increased from 3.4% ± 0.34% to 8.9% ± 1.7% when grown on AGE-MG (n = 4, P = 0.02). 

Discussion

We report an altered mRNA phenotype induced by ARPE-19 cells in vitro grown on Growth Factor Reduced Matrigel modified by AGEs. The microarray approach allowed us to examine a relatively large gene set, which uncovered an expression profile that could influence aging of RPE cells. Although the fold changes induced by AGEs were moderate, we confirmed the expression patterns of a selected subset of genes by RT-PCR/Southern and Northern analyses. The gene clusters with altered expression include those involved in cell differentiation, maintenance of the basement membrane, and early apoptosis. This analysis identified genes that were previously not known to be regulated by AGEs, expressed by RPE cells, or implicated in aging changes. 

AGE modification to the matrix altered the expression of genes involved in cell–cell signaling that regulate cell differentiation, which could have influenced in part the morphologic changes observed in our experiments. Ligand to transcription factor genes from the Wingless, Notch, Hedgehog, and EGF receptor pathways showed altered expression in a manner that would promote loss of differentiation by AGEs. These highly conserved, complex inter-related pathways control developmental decisions during morphogenesis in the embryo and cell polarity in adult tissue, through cell–cell interaction. ^{17 18 19 20 21} The Notch pathway has gained interest in the aging community because the presenilins have been associated with abnormal β-amyloid processing in Alzheimer’s disease. ^{18 22} Tight temporal and spatial regulation of Hedgehog receptor Smoothened and the Notch-modulating Fringe proteins are necessary for mouse skin epidermal differentiation. ^{20 23} The EGF receptor pathway can influence cell differentiation through the Ras/Raf/MAPK pathway ²⁴ and/or indirectly through Hedgehog and Wingless signaling via disheveled. ^{25 26} An intriguing finding was the upregulation of eps-15, which promotes the endocytotic downregulation of EGF receptors and directly influences actin cytoskeletal-mediated morphologic changes. ¹⁶ The downregulation of GATA-2 and GATA-3 is of potential significance because the GATA family transcription factors have been linked to the Wingless signaling pathway. ²⁷ Although a complete characterization of these pathways was not possible because of the limited number of genes on the microarray system used in this study, these results raise the possibility that AGE-induced alterations on RPE differentiation are influenced by these inter-related pathways. 

The upregulation of TGF-β2, CTGF, and PDGF-B by AGEs could promote a phenotype that expands Bruch’s membrane, a finding found during aging and age-related macular degeneration (AMD). Our confirmatory Northern blot analysis showed that microarray analysis underestimated the relative changes in expression of these genes, an effect that was seen by Luo et al., ¹¹ possibly from suboptimal hybridization conditions. Furthermore, Northern analyses of the TGF-β isoforms confirmed our microarray results that TGF-β2, the predominant isoform expressed by the RPE, was selectively upregulated. ²⁸ TGF-β2 and CTGF are coordinately expressed in every fibrotic disorder examined to date, through a TGF-β responsive site in the CTGF promoter. ^{29 30 31 32 33 34} The upregulation of TGF-β2 and CTGF could promote matrix protein accumulation either by increased matrix protein synthesis or decreased degradation. AGEs promote matrix fibrosis by altering both matrix protein and matrix protease expression. ¹⁵ Of the matrix metalloproteinases (MMPs), their inhibitors (TIMPs), and the plasminogen system, only PAI-1 expression was altered by ARPE-19 cells grown on AGE-MG, which could expand the matrix by decreasing plasmin activation and directly reduce proteolytic degradation or MMP activation. ^{35 36} Because an interruption in the elastic layer of Bruch’s membrane is an identifiable histopathologic change in AMD, it is of potential interest that the elastase myeloblastin was upregulated by AGEs. ³⁷  

AGEs induce apoptosis in part by generating reactive oxygen species. ^{38 39 40 41 42} Alternatively, TGF-β or TNF-α, both of which were upregulated by AGEs in our studies, promote apoptosis in a number of cell types including RPE cells. ^{43 44 45 46 47} Our analysis found a cluster of genes predominantly from the early apoptosis pathways, such as death signalers and central modulators, with an expression pattern that would promote apoptosis. ^{48 49} Interestingly, the death signalers upregulated in our study are from the TNF superfamily. Although FAS is almost exclusively linked to apoptosis, a role for TNF R1 and R2 activation is less clear, because they mediate a variety of biological effects. ^{49 50} However, the associated upregulation of TRADD, a TNF R1 adaptor protein, supports apoptosis signaling by TNF receptor activation. ⁵⁰ TNF receptor–induced apoptosis signaling can also be mediated through ceramide. ⁴⁹ It is intriguing that our analysis also found upregulation of phospholipase C, which promotes ceramide generation, by ARPE-19 cells grown on AGE-MG. ⁴⁹ Our annexin V studies provide functional evidence of apoptosis potentially induced by this cluster of genes. Because RPE apoptosis has been commonly hypothesized to play a role in the development of AMD, ⁵¹ these results warrant further investigation of this gene set. 

Caution must be used in fully comparing the AGE-induced gene expression profile found in this study to aging of the RPE or AMD. Although the ARPE-19 cell line has been characterized in depth and displays many of the characteristics of the RPE in vivo, full applicability to other RPE cell lines or to the RPE in vivo is open to speculation. It is unclear if these gene expression changes translate into similar protein profiles or if there is a correlative functional change. The morphologic and annexin V labeling changes by ARPE-19 cells grown on AGE-MG are at least very preliminary evidence of a functional consequence of these gene expression changes. We acknowledge that Matrigel, although the best-known basement membrane approximation in vitro, could induce artifactual gene expression changes or that nonphysiologic AGEs could be produced by glycolaldehyde and influence our results. Our objective was to determine whether AGEs could induce an altered mRNA phenotype. To this end, our work addressed this purpose. A glimpse at the diversity of gene expression alterations induced by AGEs was made possible with a high through-put approach. Further work characterizing AGE-induced changes, determining what changes are “cause or effect,” and identifying an AGE mediated aging effect to the RPE in vivo is warranted and underway. 

 Supported by National Institutes of Health, National Eye Institute Grants 00344 (JTH) and 06473 (LMH), Research to Prevent Blindness Manpower Award (JTH), and an unrestricted Research to Prevent Blindness grant to the Department of Ophthalmology.

 Submitted for publication January 26, 2001; revised May 9, 2001; accepted May 23, 2001.

 Commercial relationships policy: N.

 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.

 

PCR Conditions, Primer Sequences, and Product Size

Gene	Oligonucleotide Sequence	Product Size (bp)
Manic fringe (30 cycles, T anneal = 55°C)
Sense	5′-CTGGCTAATGTCTCTCAGTC-3′	247
Antisense	5′-CCGTTATGCTCCATCATCTG-3′
Internal probe	5′-ACCTGCTGTTGTTGCCAACCA-3′
Smoothened* (35 cycles, T anneal = 68°C)
Sense	5′-CTGGTACGAGGACGTGGAGG-3′	140
Antisense	5′-AGGGTGAAGAGCGTGCAGAG-3′
Internal probe	5′-ACCAGGACATGCACAGCTACA-3′
WNT-10B, † (35 cycles, T anneal = 45°C)
Sense	5′-TAAGATGAAATGCACTGT-3′	411
Antisense	5′-GAGGCTCCAAGAGTCATGGG-3′
Internal probe	5′-TGAATCCTCAGAGAGTTG-3′
GAPDH (30 cycles, T anneal = 60°C)
Sense	5′-TCTGGTAAAGTGGATATTGTTG-3′	157
Antisense	5′-GATGGTGATGGGATTTCC-3′
Internal probe	5′-ATTCCACCCATGGCAAATTCCATGGC-3′

*  From Ref. ⁵² .

†  From Ref. ⁵³ .

Figure 1.

 

Morphologic changes induced by AGE-MG. (A) Confluent ARPE-19 cells grown in DMEM/F12 + 1% BSA for 10 days on Growth Factor Reduced Matrigel (1:16 dilution) developed regular morphology. (B) When grown in identical serum withdrawn conditions for 10 days on Growth Factor Reduced Matrigel (1:16 dilution) treated with glycolaldehyde, cells appeared “spindle-shaped.” Scale bar, 25μ m.

Table 2.

 

Expression of Cell Differentiation Genes Altered by AGE-MG

Name	GenBank No.	Fold Change	Function
Wingless, Notch, Hedgehog, EGF pathway
WNT-10B	X97057	0.4	Wingless ligand
Disheveled 1	U46461	0.5	Wingless component
Disheveled 3	U49262	0.5	Wingless component
Manic Fringe	U94352	2.0	Notch ligand modulator
Smoothened	U84401	0.5	Hedgehog receptor
Kidney EGF precursor	X04571	1.4	Regulates differentiation
EGF-like GF	M60278	1.4	Wound healing, activates HER4
EGF-like crypto protein-1	M96956	1.8	Expressed in cancer cells
Eps 15	U07707	2.0	EGF rec substrated involved in endocytosis and cytoskeletal rearrangement
EGFR HER4	L07868	2.4	Tyrosine kinase EGF Rec
EGFR ERBB3	M29366	0.3	EGF Receptor
Transcription factors (TF) involved in cell differentiation
GATA-2	M68891	0.3	Possible Wingless TF
GATA-3	X55122	0.4	Possible Wingless TF
ID3	X69111	0.5	Inhibitor of basic HLH TFs
Sp2 protein	M97190	0.5	Ubiquitous expression, GT box
Hap2	M59079	0.4	Ubiquitous expression, CCAAT motif

Figure 2.

 

Representative RT-PCR analysis with Southern blot analysis of ARPE-19 cells grown on AGE-MG and MG. ARPE-19 cells were grown under identical conditions as for the microarray experiments. Total RNA was extracted and subjected to RT-PCR (column 1), Southern transferred, and hybridized with a ³²P-labeled internal probe (column 2). The Southern blot analysis phosphorimager signals were normalized to that of GAPDH. WNT, Wingless 10B; MFNG, Manic Fringe; smoh, Smoothened; M, ARPE-19 cells grown on Matrigel; A, ARPE-19 cells grown on AGE-Matrigel.

Figure 3.

 

Representative Northern blot analysis of ARPE-19 cells grown on AGE-MG and MG. ARPE-19 cells were grown under identical conditions as for the microarray experiments. Total RNA was extracted and subjected to Northern blot analysis for CTGF, TGF-β2, and PDGF-B. Signals were normalized to 28S rRNA.

Table 3.

 

Expression of Apoptosis-Related Genes

Name	GenBank	Fold Change	Function
Death Signalers
TGF-β2	M19154	2.0	Initiates apoptosis signal
TNF-α	X01394	1.7	Ligand for TNF Rs
TNF-α receptor-1	33294	1.5	Initiates apoptosis signal
TNF-α receptor-2	M32315	1.7	Initiates apoptosis signal
Fas antigen (CD95)	M67454	1.5	Initiates apoptosis signal
TNF-α receptor CD27	L08096	1.5	Initiates apoptosis signal
Central Modulators
TRADD	L41690	1.5	TNF R1 adaptor protein, recruits “death domain” proteins
p53	M14694	1.5	Multiple roles in cell cycle and apoptosis regulation
Bik	X89986	1.5	Antagonizes Bcl-2, promotes apoptosis
Bcl-2	M14745	0.5	Protective against apoptosis
HSP70	M11717	0.6	Protective against TNF-generated apoptosis
Phospholipase C	D42108	2.0	Induces ceramide production for apoptosis signaling
Execution Stage
Caspase 8	U60520	1.5	Most upstream execution stage enzyme
DNA repair enzymes
Excision repair ERCC2	X52221	0.5	Nucleotide excision repair
Excision repair ERCC5	L20046	0.5	Nucleotide excision repair
DNase X	X90392	0.5	DNase I-like endonuclease

Figure 4.

 

Effect of AGE-MG on inducing apoptosis. ARPE-19 cells were grown in DMEM/F12 + 10% FBS for 1 day and then in DMEM/F12 + 1% BSA for 7 days on MG (A) and AGE-MG (B). Apoptosis was assessed with two-parameter analysis by staining with annexin V (early apoptotic externalization of phosphatidylserine) and propidium iodide (cell viability) and measured by flow cytometry. Normal cells (bottom left quadrant) had low annexin V and low PI staining. The early apoptotic cells (bottom right quadrant) had high annexin V and low PI staining. The dead cells (top right quadrant) had high annexin V and high PI staining. Percentages of cells in each quadrant are indicated. Results are one replicate sample of one experiment that is representative of three independent experiments that were conducted.