Content uploaded by Jean-Yves Brossas
Author content
All content in this area was uploaded by Jean-Yves Brossas on Jan 04, 2022
Content may be subject to copyright.
Original Contribution
Prematurely senescent ARPE-19 cells display features of
age-related macular degeneration
Anne-Lise Glotin
a,b,c
, Florence Debacq-Chainiaux
d
, Jean-Yves Brossas
a,b,c
, Anne-Marie Faussat
a
,
Jacques Tréton
a,b,c
, Anna Zubielewicz
e
, Olivier Toussaint
d
, Frédéric Mascarelli
a,b,c,
⁎
a
Centre de Recherche des Cordeliers, Université Pierre et Marie Curie–Paris 6, UMR S 872, Paris F-75006, France
b
Université Paris Descartes, UMR S 872, Paris F-75006, France
c
INSERM, U872, Paris F-75006, France
d
Laboratory of Biochemistry and Cellular Biology, Department of Biology, University of Namur (FUNDP), rue de Bruxelles, 61, 5000 Namur, Belgium
e
Ophthalmology Department, Medical School, Lublin, Poland
Received 3 October 2007; revised 7 December 2007; accepted 12 December 2007
Available online 28 January 2008
Abstract
The etiology of age-related macular degeneration (AMD), the leading cause of blindness in the developed world, remains poorly understood,
but may be related to cumulative oxidative stress. The prime target of the disease is the retinal pigmented epithelium (RPE). To study the
molecular mechanisms underlying RPE degeneration, we investigated whether repetitive oxidative stress induced premature senescence in RPE
cells from the human ARPE-19 cell line. After exposure to 8 mM tert-butylhydroperoxide (tert-BHP) for 1 h daily for 5 days, the cells showed
four well-known senescence biomarkers: hypertrophy, senescence-associated β-galactosidase activity, growth arrest, and cell cycle arrest in G1. A
specific low-density array followed by qRT-PCR validation allowed us to identify 36 senescence-associated genes differentially expressed in the
prematurely senescent cells. Functional analysis demonstrated that premature senescence induced amyloid βsecretion, resistance to acute stress by
tert-BHP and amyloid β, and defects in adhesion and transepithelial permeability. Coculture assays with choroidal endothelial cells showed the
proangiogenic properties of the senescent RPE cells. These results demonstrate that chronic oxidative stress induces premature senescence in RPE
cells that modifies the transcriptome and substantially alters cell processes involved in the pathophysiology of AMD. Oxidative stress-induced
premature senescence may represent an in vitro model for screening therapeutics against AMD and other retinal degeneration disorders.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Aging; Retina; Transcriptome; Tight junction; Amyloid β; Angiogenesis; Free radicals
Age-related macular degeneration (AMD) is a retinal de-
generative disease and the leading cause of vision loss in Western
countries, affecting 5 to 10% of the population older than 60 years
[1]. The prime target of its early development is the retinal
pigmented epithelium (RPE), a monolayer of quiescent cells that
perform various processes essential for maintaining photoreceptor
(PR) functions and survival. The RPE is essential for the retinoid
cycle and for maintaining the blood–retina barrier (BRB). The RPE
degenerates during AMD and leads to BRB disturbance and PR
apoptosis. The neovascular form of AMD is characterized by newly
formed vessels that originate in the choroidal vascular endothelial
cells and break through Bruch’s membrane into the subretinal
pigment epithelial space. Many genetic and environmental factors
A
vailable online at www.sciencedirect.com
Free Radical Biology & Medicine 44 (2008) 1348 –1361
www.elsevier.com/locate/freeradbiomed
Abbreviations: 4-HNE, 4-hydroxynonenal; 8-OHG, 8-hydroxyguanosine;
AEBSF, 4-(2-aminoethyl)benzenesulfonyl fluoride; AMD, age-related macular
degeneration; Aβ,amyloidβ;BRB,blood–retina barrier; CEC, choroidal endo-
thelial cells; ELISA, enzyme-linked immunosorbent assay; FGF, fibroblast
growth factor; FITC, fluorescent isothiocyanate; GSH, glutathione; H
2
DCFDA,
2′,7′-dichlorodihydrofluorescein diacetate; LDH, lactate dehydrogenase; MTT,
3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide; PBS, phosphate-
buffered saline; PI, propidium iodide; qRT-PCR, quantitative real-time–polymerase
chain reaction; ROS, reactive oxygen species; RPE, retinal pigmented epithelium;
SA β-gal, senescence-associated β-galactosidase; SEM, standard error of the mean;
tert-BHP, tert-butylhydroperoxide; tPA, tissue plasminogen activator.
⁎Corresponding author. Centre de Recherchedes Cordeliers, Université Pierre et
Marie Curie–Paris 6, UMR S 872, Paris F-75006, France. Fax: +33 1 40 46 78 65.
E-mail address: frederic.mascarelli@idf.inserm.fr (F. Mascarelli).
0891-5849/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.freeradbiomed.2007.12.023
play a role in AMD but may also confound the role of other factors.
RPE cell dysfunction, inflammatory processes, and oxidative stress
have been proposed as pathogenic pathways. The recent finding
that oxidative stress modulates the expression of the complement
factor H (CFH) gene, which is associated with AMD, suggests
interactions between AMD risk factors, the strongest of which is
advanced age [2].
For several reasons, the retina is an ideal environment for
the generation of reactive oxygen species (ROS) and oxidative
damage [3]: (1) it consumes much more oxygen than any other
tissue; (2) it is subject to high levels of cumulative irradiation;
(3) membranes of the PR outer segments are rich in polyun-
saturated fatty acids; (4) the PR and RPE contain an abundance
of photosensitizers; and (5) phagocytosis of rod outer segments
by the RPE is itself an oxidative stress and generates ROS.
Chronic and cumulative oxidative stress may also be involved in
the pathogenesis of AMD [3–5]. The free radical theory of
aging, which proposes that aging and age-related disorders result
from cumulative damage from reactions involving ROS and
oxidative damage, may thus help explain AMD.
Premature senescence can be induced by the hyperactivation
of oncogenes or by chronic exposure to cellular stress, such as
DNA-damaging agents or oxidative stress [6–8]. Age-related
changes in the transcriptional profile of RPE have previously
been analyzed with microarray and PCR in mice and rats [9–11].
In this study, we investigated whether ARPE-19 cells undergo
premature senescence after exposure to a direct oxidative agent at
sublethal concentrations. We also used a low-density DNA array,
quantitative PCR analyses, and various biological assays to assess
whether this premature senescence affects RPE-specific func-
tional properties through a transcriptional process. We show that
chronic oxidative stress induces premature senescence, affects the
expression of RPE-specific genes, and alters several RPE func-
tional properties essential for retinal homeostasis.
Materials and methods
Cultures and cell treatments
ARPE-19 (generously provided by Dr. Hjelmeland, University
of California, Davis, CA, USA), a nontransformed human RPE
cell line that displays many differentiated properties typical of RPE
in vivo, was cultured in DMEM:F12 (Invitrogen), supplemented
with 10% serum, 2 mM glutamine, and 15 mM Hepes (complete
culture medium). Premature senescence was induced by oxidative
stress on ARPE-19 cells, as previously described [12].Briefly,
confluent human ARPE-19 cells were incubated daily in the pre-
sence of 8 mM tert-BHP for 1 h for 5 days (Fig. 1). The medium
contained serum to avoid serum depletion-associated oxidative
stress [13]. After each stress, the cells were thoroughly washed
with HBSS and provided with complete culture medium for 24 h.
Complete culture medium was replaced daily in parallel cultured
control cells during the 5-day period (Fig. 1). After the fifth stress,
the cells were allowed to recover for 3 days before we conducted
further experiments. The cells were then replated at various
densities, ranging from 10,000 to 40,000 cells/cm
2
,dependingon
the particular assay.
Choroidal endothelial cells (CEC) were isolated from bovine
eyes and purified as previously described [14]. They were cultured
in EGM-2MV medium (Cambrex) on 50 μg/ml fibronectin-coated
dishes (Invitrogen) and cocultured with ARPE-19 cells as follows:
ARPE-19 cells were seeded on permeable-insert-system mem-
branes (Transwell, Costar) at 30,000 cells/cm
2
(day 0) and cultured
for 24 h. CEC (passage 4 to 7) were then seeded on fibronectin-
coated wells at 5000 cells/cm
2
(day 1) and placed 24 h later under
the ARPE-19 insert in minimum culture medium (without serum,
FGF, or VEGF) (day 2).
Analysis of cell viability and senescence
ARPE-19 cell viability was assessed by (1) counting trypan
blue-excluding cells after adding 0.5% trypan blue, (2) using
MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium
bromide) staining (Sigma), and (3) monitoring lactate dehydro-
genase (LDH) release into the culture medium, with a cytotoxicity
detection kit (Roche). Senescence was investigated with the
senescence-associated β-galactosidase (SA β-gal) staining kit
(Cell Signaling), according to the manufacturer’s instructions.
Measurement of intracellular oxidation
We measured intracellular ROS levels with a 2′,7′-dichlorodi-
hydrofluorescein diacetate (H
2
DCFDA) probe (Interchim). Cells
were incubated with 1 μMH
2
DCFDA for 15 min at 37°C,
collected in 500 μl 1% PAF, and analyzed by flow cytometry
(Epics ALTRA; Beckman Coulter). Intracellular glutathione
(GSH) concentrations were determined by colorimetric assays
according to the manufacturer’s instructions (GT10; Oxford Bio-
medical Research).
Cell cycle analysis
DNA content was determined by staining cells with pro-
pidium iodide (PI). Cells were washed in PBS and fixed in 70%
Fig. 1. Experimental design and reference time frame.
1349A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
ethanol for 30 min at 4°C. Then the cells were rehydrated in
PBS, treated with 1 mg/ml RNase A, stained with 50 mg/ml
PI for 15min at 4°C, and analyzed by flow cytometry (Epics
ALTRA, Beckman Coulter).
Low-density DNA array
Design of the array
We used the DualChip human aging microarray (Eppendorf
Array Technology), a low-density DNA array containing 240
genes involved in senescence or stress response (see Table S1 in
the supplementary material) [6,7]. The method is based on a
system with two assays (a control and a test) per glass slide
with three subarrays per assay. Several positive and negative
hybridization controls plus detection controls were spotted on
the array to verify the reliability of the experimental data. Five
hundred nanograms of mRNA were retrotranscribed with Su-
perScript II reverse transcriptase (Invitrogen). Three synthetic
poly(A)
+
-tailed RNA standards were spiked in three different
amounts (10, 1, and 0.1 ng per reaction) into the purified mRNA.
Two independent experiments were performed in duplicate, with
direct and cross comparisons. There were accordingly hybridi-
zations on six arrays. Hybridization on the DualChip human
aging microarray was carried out as described by the manu-
facturer. Detection was performed using a Cy3-conjugated IgG
anti-biotin (Jackson Immuno Research Laboratories). Fluores-
cence of the hybridized arrays was scanned with the Packard
ScanArray (Perkin–Elmer) at a resolution of 10 μm. To maxi-
mize the dynamic range of detection, the same arrays were
scanned at three photomultiplier gains for quantifying high- and
low-copy-number expressed genes. The scanned 16-bit images
were imported into ImaGene 4.1 software (BioDiscovery). The
fluorescence intensity of each DNA spot (average intensity of
each pixel present within the spot) was calculated and local mean
background subtracted. A signal was accepted when the average
intensity after background subtraction was at least 2.5 times
higher than its local background. The three intensity values of
the triplicate DNA spots were averaged and used to calculate the
intensity ratio between the reference and the test samples. To
overcome the detection limits inherent to DNA array-based
methods of transcriptome analysis, we performed a qRT-PCR
analysis for several genes of interest.
Data normalization
The data were normalized in two steps, as previously described
[6,7]. First, the values were corrected by using a factor calculated
from the intensity ratios of the internal standards in the references
and test samples. The presence of the three internal standard
probes at two different locations of the array allowed measurement
of local background and evaluation of the array homogeneity,
which is considered in normalization. Because the internal stan-
dard control does not take the purity and quality of the mRNA
into account, however, a second normalization step calculated the
average intensity for a set of eight housekeeping genes. The var-
iance of the normalized set of housekeeping genes was used to
generate an estimate of expected variance and thus a predicted
confidence interval for testing the significance of the ratios ob-
tained. Ratios outside the 95% confidence interval were deter-
mined to be significantly different [6,7].
Quantitative real-time–polymerase chain reaction
The Ribopure extraction kit (Applied Biosystems) was used to
extract total RNA from three independent cultures and Super-
Script II reverse transcriptase (Invitrogen) was used to retrotrans-
cribe 2 μg of mRNA. Amplification reaction assays contained
either 1× SYBR Green PCR master mix or 1× TaqMan PCR
master mix (Applied Biosystems), depending on the type of
primers in the mix. A hot start at 95°C for 5 min was followed by
40 cycles at 95 °C for 15 s and 65°C for 1 min with the 7300 SDS
thermal cycler (Applied Biosystems). Controls with no reverse
transcriptase were run for each assay to confirm the lack of
genomic DNA contamination. Each sample was tested in trip-
licate. In-house designed primers are listed (see Table S2 in the
supplementary material) and were used with the SYBR Green
chemistry. PLAU and NEP primers were purchased from Super
Array and used with the SYBR Green chemistry. RLBP1,RBP1,
LRAT,andRDH5 primers were purchased from Applied Bio-
systems and used with the TaqMan chemistry. GAPDH was used
as a suitable reference gene. The standard curve method (Prism
7700 sequence detection system; ABI User Bulletin No. 2) was
used for relative quantification of gene expression.
Fig. 2. Effects of cumulative tert-BHP treatment on ARPE-19 cell viability.
Confluent cultures of quiescent ARPE-19 cells were either treated daily with tert-
BHP at the indicated concentrations or stimulated by serum (Control) for 1 h daily
for 5 days. Cells were then replated after 3 daysof recovery.Cellular and molecular
analyses were undertaken 3 days after replating. Cell survival was analyzed with
(A) MTT and (B) trypan blue exclusion assays, and cellular toxicity was assessed
by (C) LDH release, as described under Materials and methods. Similar results
were obtained in three independent experiments. Values are means± SEM. Dif-
ferences between groups wereanalyzed by the Kruskal–Wallis test (A: p= 0.0047).
Differences between pairs were analyzed by the Mann–Whitney test (ns, non-
significant; ⁎pb0.05; ⁎⁎pb0.005).
1350 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
Western blot analysis
Cells were lysed in ice-cold buffer and mixed with Laemmli
buffer, as previously described [15]. Proteins (30 μg) were then
resolved by SDS–PAGE (12–15% polyacrylamide gel) and
electroblotted onto PVDF filters before being probed with anti-
bodies directed against apolipoprotein J, β-actin, BAX (Santa
Cruz), SOD2 (Stressgen), and APP (Abcam). We used β-actin as
an internal standard to check for protein loading. Primary anti-
bodies were detected with horseradish peroxidase-conjugated
secondary antibodies (Cell Signaling). The enhanced chemilu-
minescence substrate was used to detect the secondary antibody
according to the manufacturer’s instructions.
Immunocytochemistry
Cells were fixed with 4% PAF and permeabilized with 0.1%
Triton X-100. We used PBS/milk (1× PBS, 15% milk) for satu-
ration and incubated cells with antibodies directed against acro-
lein (Novus), 4-HNE, 8-OHG (Alpha Diagnostics), or occludin
(Santa Cruz). Primary antibodies were detected with fluorescein-
conjugated secondary antibodies. Equal exposure parameters
were used to compare the fluorescence intensity of staining of
control versus tert-BHP-treated ARPE-19 cells, measured with a
fluorescence microscope (Olympus).
Quantification of tissue plasminogen activator (tPA) and
amyloid β(Aβ) secretion
tPA and Aβ(1–42) secretion in ARPE-19 cell supernatants was
detected by using Assaypro and Signet ELISA kits, respectively.
To enable Aβ(1–42) detection, 20 μMphospharamidon(Sigma),
a metalloprotease inhibitor, and a cocktail of anti-protease inhi-
bitors (1 μg/ml leupeptin, 5 μg/ml pepstatin, 1 mM AEBSF, and
4μg/ml aprotinin) were added to the supernatants, which were
then concentrated with 100- and 3-kDa cut-off centricons (Mil-
lipore). We then followed the manufacturer’sinstructionsfor
detecting tPA and Aβ.
Fig. 3. Analysis of the prematurely senescent phenotype of tert-BHP-treated ARPE-19 cells. Confluent cultures of quiescent ARPE-19 cells were either treated daily
with tert-BHP at 8 mM or stimulated by serum (Control) for 1 h daily for 5 days. Cells were then replated after 3 days of recovery. (A) Cell surface area was measured
3 days after cell replating by phase-contrast microscopy at 40× magnification (see pictures below the graph, scale bar, 100 μm), with Visilog image processing software
(Noesis). (B) Cells were observed 3 days after replating by phase-contrast microscopy at magnification 20× (see picture below the graph, scale bar, 200 μm) and the
population of SA β-gal-positive cells was determined by counting 200 cells per dish. (C) Cell proliferation was determined by counting cells during a 7-day culture
period after cells were replated. Culture medium was replaced every 2 days. (D) Cell cycle analyses were performed 1 day after cell replating by determining the DNA
content with propidium iodide. Similar results were obtained in three independent experiments. Values are means ± SEM. Differences between groups were tested by
the Kruskal–Wallis test (C: p=0.0001) followed by pairwise comparisons with the Mann–Whitney test (⁎pb0.05, ⁎⁎pb0.005).
1351A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
Measurement of cell permeability
ARPE-19 cells were seeded on permeable insert system mem-
branes (Transwell, Costar) as previously described [16].Con-
fluent ARPE-19 cells were then incubated for 6 h with 1 mg/ml
40-kDa fluorescent isothiocyanate (FITC)–dextran (Sigma) in
culture medium containing neither serum nor growth factors.
Permeability was assessed in both directions: upper to lower well
(“inward”) and lower to upper (“outward”), as described pre-
viously [17]. Empty filters without cells reflected FITC–dextran
passive flux. Fluorescence was measured with a plate fluorometer
(Safire microplate reader; Tecan) at an excitation/emission of 492/
535 nm. Barrier properties of ARPE-19 cells were investigated in
confluent cells 1 week after replating. The difference between the
FITC–dextran flux in ARPE-19 cells compared with a filter
without cells showed barrier functions in the ARPE-19 cells.
Statistics
Statistical analyses were performed with GraphPAD software.
Differences between groups were compared with the nonpara-
metric Kruskal–Wallis test, and paired comparisons were per-
formed with the Mann–Whitney test. Data are expressed as
means ±SEM.
Results
Chronic sublethal tert-BHP-mediated oxidative stress induces
premature senescence in ARPE-19 cells
We first investigated what culture conditions might induce
ARPE-19 senescence without cell death. ARPE-19 cells were
treated with tert-BHP (1.5 to 12 mM) daily for 5 consecutive
Fig. 4. Effects of cumulative tert-BHP treatments on intracellular oxidation state in ARPE-19 cells. Premature senescence was induced in ARPE-19 cells, as described
in the legend for Fig. 3. Cellular and molecular analyses were undertaken 3 days after replating. (A) Intracellular ROS levels were measured with an H
2
DCFDA probe
and FACS analysis with excitation and emission settings of 488 and 530 nm, respectively. Oxidation of (B and C) lipids, and (D) nucleic acids, was assessed by
immunocytochemistry by measuring 4-HNE and acrolein, and 8-OHG, respectively, as described under Materials and methods. Similar results were obtained in three
independent experiments. Equal exposure parameters were used to compare the fluorescence intensity of staining of control versus tert-BHP-treated ARPE-19 cells.
Scale bar, 50 μm.
1352 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
days. Cell treatment with 8 mM tert-BHP was not cytotoxic,
whereas 12 or 15 mM tert-BHP concentrations were, according
to the LDH activity assay, trypan blue-excluding cell counting,
and the MTT assay (Fig. 2). We then determined whether the
cumulative treatment of the cells with 8 mM tert-BHP induced
their premature senescence, by investigating four major markers
of cell senescence: (1) the development of an enlarged, flat-
tened, and irregular morphology and decreased cell density at
confluence [18]; (2) SA β-gal staining [19]; (3) cell failure to
respond to mitogens [20]; and (4) cell growth arrest in the G1 or
G2 phase of the cell cycle [20]. The cell surface area of the tert-
BHP-treated ARPE-19 cells was double that of the control
ARPE-19 cells (Fig. 3A). SA β-gal staining was 12 times
greater in the tert-BHP-treated than in the control ARPE-19
cells (82% versus 7%) (Fig. 3B). Tert-BHP-treated ARPE-19
cells were no longer able to proliferate after serum stimulation
(Fig. 3C), and flow cytometry analysis showed that in the S
phase of the cell cycle the proportion of tert-BHP-treated cells
(19%) was dramatically lower than the proportion of control
cells (71%). Most of the tert-BHP-treated cells were in the
G1 phase (57%) (Fig. 3D). We can thus conclude that chronic
sublethal stress produced a prematurely senescent phenotype in
ARPE-19 cells and we will henceforth refer to the tert-BHP-
treated ARPE-19 cells as prematurely senescent ARPE-19 cells.
Note that prematurely senescent ARPE-19 cells presented a
stable phenotype over a period of at least 6 weeks of culture
(data not shown).
Analysis of the redox status of the prematurely senescent
ARPE-19 cells indicated that intracellular ROS production had
multiplied by 13 (Fig. 4A), whereas immunocytochemical ana-
lysis showed that the levels of lipid peroxidation and oxidized
nucleic acids were higher in the prematurely senescent than in
the control cells (Fig. 4B). This suggests that tert-BHP-induced
premature senescence involves an intracellular oxidative stress
that affects lipids and nucleic acids in ARPE-19 cells.
Senescence-associated transcriptomic changes induced by
repeated exposure to tert-BHP
To investigate the transcriptional changes in prematurely
senescent ARPE-19 cells, we used a recently developed low-
Table 1
Genes from the DualChip Human Aging microarray that are differentially
expressed in prematurely senescent ARPE-19 cells
Gene
symbol
Gene name Accession
No.
Change
(fold)
Cell cycle control
DHFR Dihydrofolate reductase NM_000791 −2.7
CENPA Centromere protein-A U14518 −2.3
MYBL2 b-Myb X13293 −2.0
H4FM Histone 4 member M consensus NM_003495 −2.0
PLK Polo-like kinase U01038 −2.0
KNSL6 Mitotic centromere-associated kinesin NM_006845 −2.0
KNSL5 Mitotic kinesin-like protein 1 NM_004856 −1.9
UBE2C Mitotic kinesin-like protein 1 NM_004856 −1.9
CDK4 Cyclin-dependent kinase 4 U79269 −1.8
E2F1 E2F transcription factor 1 NM_005225 −1.7
ADPRT ADP-ribosyltransferase J03473 −1.6
CDK2 Cyclin-dependent kinase 2 NM_001798 −1.5
POLA2 Polymerase αNM_002689 −1.5
CCND1 Cyclin D1 NM_053056 1.9
TGFA TGF-αNM_003236 2.1
H2B/S Histone2b member B/S consensus NM_080593 3.3
DNA synthesis
TK1 Thymidine kinase NM_003258 −2.6
MCM2 Mitotin D21063 −2.1
TYMS Thymidylate synthetase NM_001071 −2.1
TOP2A Topoisomerase 2αNM_001067 −1.9
PCNA Proliferating cell nuclear antigen NM002592 −1.7
RRM1 Ribonucleotide-reductase M1 NM_001033 −1.6
Cell remodeling/extracellular matrix
SM22 Transgelin M95787 −5.8
FN1 Fibronectin X02761 −2.6
THBS1 Thrombospondin X14787 −1.8
IVL Involucrin M13903 2.9
TPA Tissue plasminogen activator NM_000930 4.9
MMP3 Matrix metalloproteinase 3 NM_002422 31.6
Protein degradation
CTSZ Cathepsin Z AF136273 −2.0
CTSH Cathepsin H NM_004390 −1.8
PSMA3 Proteasome subunit, αtype, 3 NM_002788 −1.2
CTSD Cathepsin D NM_001904 1.7
Growth factors
BMP2 Bone morphogenetic protein 2 NM_001200 2.1
Apoptosis/defense system
BAX BCL2-associated X protein NM_004324 −1.7
CASP8 Caspase 8 X98172 −1.3
APOJ Apolipoprotein J J02908 3.0
ARPE-19 cell premature senescence was induced as described in the legend for
Fig. 3. Total mRNA was extracted 3 days after cell replating. The effect of
premature senescence on gene expression was determined with the DualChip
Human Aging microarray.
Table 2
Comparison of gene expression detected by the DualChip Human Aging
microarray and qRT-PCR in prematurely senescent ARPE-19 cells
Gene
symbol
Gene name Accession
No.
Change (fold) qRT-PCR/
array (%)
Array qRT-PCR
SM22 Transgelin M95787 −5.8 −4.2 ⁎⁎ 73
FN1 Fibronectin X02761 −2.6 −1.7 ⁎⁎ 66
E2F1 E2F transcription
factor 1
NM_005225 −1.7 −1.7 ⁎⁎ 101
CDK2 Cyclin-dependent
kinase 2
NM_001798 −1.5 −1.6 ⁎103
CTSD Cathepsin D NM_001904 1.7 1.8 ⁎⁎ 106
CCND1 Cyclin D1 NM_053056 1.9 1.7 ⁎⁎ 90
APOJ Apolipoprotein J J02908 3.0 2.8 ⁎⁎⁎ 92
TPA Tissue plasminogen
activator
NM_000930 4.9 4.5 ⁎⁎⁎ 93
qRT-PCR analysis of the expression of eight genes found by the DualChip assay
to be differentially expressed in prematurely senescent and control ARPE-19
cells. Data normalization of the DualChip results was performed as described
under Material and methods. The differences between pairs in the qRT-PCR
analysis were determined by the Mann–Whitney test.
⁎pb0.05.
⁎⁎ pb0.005.
⁎⁎⁎ pb0.0005.
1353A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
density DNA array to analyze 240 genes involved in senescence
[7]. Prematurely senescent ARPE-19 cells expressed 36 genes
differentially, with more abundant mRNA seen for 8 genes and
decreased mRNA levels for 28 (Tab l e 1). The microarray-based
quantification was very consistent with qRT-PCR analysis for 8
representative genes (Tab le 2). The process most affected by
senescence was cell cycle regulation. Expression of the principal
genes involved in positive regulation of the cell cycle, including
E2F1,CDK2,andCDK4, was decreased in the prematurely
senescent ARPE-19 cells (Ta b le 1). Cyclin D1 mRNA became
more abundant, as it is reported to do in replicative senescence of
human fibroblasts [21]. qRT-PCR analysis showed that premature
senescence induced an increase in the expression of p21
waf-1
,
which is a major gene responsible for cell cycle arrest during
senescence (Tab le 3). Changes in p21
waf-1
, E2F1, and cyclin D1
mRNA levels showed similar changes in the amounts of the
respective proteins (data not shown). These results were consistent
with cell cycle arrest detected by flow cytometry and a prematurely
senescent phenotype (Fig. 3D). The transcriptome analysis also
demonstrated that DNA synthesis was affected by premature
senescence in ARPE-19 cells. A decrease in mRNA levels for six
major genes involved in the positive control of DNA synthesis,
including MCM2 and PCNA (Tab le 1), was consistent with our
observation that prematurely senescent ARPE-19 cells failed to
enter the S phase of the cell cycle (Fig. 3D).
Five other major cell processes that are altered in sene-
scence were affected in these prematurely senescent ARPE-19
cells. First, genes involved in cell remodeling were affected: we
observed less abundant FN1 and THBS1 mRNA and more
abundant MMP3 mRNA (Table 1), consistent with the altered
morphology of the senescent cells. Second, premature senes-
cence greatly affected gene expression for both the lysosomal
(CTSD,CTSH, and CTSZ) and the nonlysosomal (PSMA3)
degradation pathways (Table 1), suggesting an alteration in the
process of degradation of abnormal or damaged proteins in
ARPE-19 cells. Third, gene expression changed for the growth
factor BMP2, which was found overexpressed in the prematurely
senescent cells (Table 1). Fourth, we observed the underexpres-
sion of two proapoptotic genes, CASP8 and BAX. Finally, APOJ
was highly overexpressed (by a factor of 5.5) (Table 1), as shown
previously in all models of premature senescence of fibroblasts,
as well as in replicative senescence [6,22,23].
Premature senescence and resistance to lethal oxidative stress
It has previously been shown that senescent cells are resistant
to cell death [24,25]. Because we observed decreased amounts
of caspase 8 and BAX mRNA in the prematurely senescent
ARPE-19 cells, we extended the transcriptomic analyses to genes
involved in stress response and defense systems. qRT-PCR
showed a significant increase in the expression of GPX1A and
SOD2, both involved in antioxidant defense, and of CRYBA1,
which encodes the heat-shock protein αβ-crystallin (Tabl e 3 ).
Western blot analysis showed that the quantity of BAX protein
decreased, whereas SOD2 protein production increased (Fig. 5A).
Intracellular production of GSH, an essential cell detoxifier, was
2.5 times higher than in the control cell (Fig. 5B), thereby sug-
gesting that the defense system against oxidative stress is bet-
ter developed in prematurely senescent than in control ARPE-19
cells. Acute treatment with tert-BHP induces ARPE-19 cell
apoptosis through oxidative stress [15]. Prematurely senescent
ARPE-19 cells showed greater resistance than control cells to
treatment with a lethal concentration of tert-BHP, with a survival
rate 8.8 times higher than that of the control ARPE-19 cells
Table 3
Gene expression analysis by qPCR in prematurely senescent ARPE-19 cells
Gene symbol Gene name Change (fold)
Cell cycle control
P21WAF-1 p21
waf-1
1.7 ±0.3 ⁎⁎
P53 p53 −1.4 ±0.1 ⁎⁎
Defense system
CAT Catalase 1 ± 0.2 ns
CRYBA1 αβ-Crystallin 1 2.6± 0.9 ns
GPX1A Glutathione peroxidase 1 1.4± 0.1 ns
HSP27 Heat shock 27-kDa protein 1 −1.1 ± 0.2 ns
HSP70 Heat shock 70-kDa protein 1 −1.4 ± 0.1 ⁎
HSP90 Heat shock protein 90-kDa α−1.3 ±0.2 ⁎⁎
SOD1 Cu/Zn superoxide dismutase −1.7± 0.1 ns
SOD2 Mn superoxide dismutase 1.9± 0.2 ⁎⁎
Amyloidogenesis
APOJ Apolipoprotein J 5.1 ±1.4 ⁎⁎
APP Amyloid β(A4) precursor protein 1.5± 0.4⁎
BACE1 β-site APP-cleaving enzyme 1 1.2 ±0.1 ⁎⁎
NEP Neprilysin −1.7 ±0.2 ⁎⁎⁎
Retinoid cycle
LRAT Lecithin retinol acyltransferase −1.7 ±0.0 ⁎⁎
RBP1 Cellular retinol-binding protein 1 4.9 ±1.8 ⁎⁎
RDH5 Retinol dehydrogenase 5 (11-cis and 9-cis) 1.4 ± 0.1 ⁎
RLBP1 Retinaldehyde-binding protein 1 1.4 ±0.4 ⁎⁎
Adhesion
ACTA2 α-smooth muscle actin 1.1 ± 0.2 ⁎⁎
COL1 Collagen 1 −3.3 ± 0.1 ns
FN1 Fibronectin 1 −2.5 ±0.0 ⁎⁎
VIM Vimentin −1.1± 0.2 ns
Tight junction
CLDN1 Claudin 1 2 ± 0.6 ns
CX43 Connexin 43 −2 ± 0.1 ⁎⁎
OCLN Occludin −1.3± 0.2 ns
PAR3 Par-3 partitioning-defective 3 homolog −1.3 ±0.1 ⁎⁎
Angiogenesis
PEDF Pigment epithelium-derived factor 1 ± 0.4 ns
TGFB2 Transforming growth factor-β2−1.1 ± 0.2 ⁎⁎
TGFB1 Transforming growth factor-β1 1.6 ± 0.1⁎⁎
PLAU Plasminogen activator urokinase 2.2 ± 0.2 ns
VEGFA Vascular endothelial growth factor A −2.5 ± 0.2 ⁎⁎
TPA Tissue plasminogen activator 4.6 ± 0.4 ⁎⁎
ARPE-19 cell premature senescence was induced as described in the legend
for Fig. 3. Total mRNA was extracted 3 days after cell replating. The effect of
premature senescence on gene expression was determined by qPCR. The differences
between pairs were determined by the Mann–Whitney test.
ns, nonsignificant.
⁎pb0.05.
⁎⁎ pb0.005.
⁎⁎⁎ pb0.0005.
1354 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
(Fig. 5C). The Aβpeptides (1–40) and (1–42) have also been
shown to induce oxidative cell toxicity in RPE cells [27]. Treat-
ment of prematurely senescent ARPE-19 cells with Aβ(1–42) or
Aβ(1–40) induced cell death rates of 20 and 5%, respectively,
compared with rates of 30 and 15% for the control cells (Fig. 5C).
This demonstrates that prematurely senescent ARPE-19 cells are
resistant to toxic Aβpeptides.
Premature senescence in ARPE-19 cells induces increased
amyloidogenesis
Drusen, which are Aβ-containing extracellular deposits, are
common in the retinas of elderly people and especially AMD
patients [26]. Aging also modulates APP processing [27–29].qRT-
PCR analysis showed that the expression of APP, which encodes
the amyloid precursor protein, was 1.5 times higher in prematurely
senescent than in control ARPE-19 cells (Table 3). BACE1,en-
coding the β-secretase 1 enzyme, which cleaves APP to produce
Aβ, was also overexpressed in the prematurely senescent ARPE-19
cells, as was APOJ, which encodes a protein that influences Aβ
formation [30] (Table 3). NEP, encoding a rate-limiting peptidase
involved in Aβcatabolism, was underexpressed in prematurely
senescent ARPE-19 cells (Table 3). We analyzed APP and apo-
lipoprotein J protein expression by Western blotting and found
higher levels of APP and apolipoprotein J production in the
prematurely senescent cells (Fig. 6A). Quantification of Aβ(1–42)
by ELISA showed that Aβsecretion was significantly higher in the
prematurely senescent than in the control cells (Fig. 6B). This
strongly suggests that the increased Aβsecretion induced by
premature senescence involves upregulation of both its precursor
protein and the major enzymes involved in its formation.
Effect of premature senescence on the retinoid cycle
We investigated the effect of premature senescence on the
retinoid cycle, a process specific to RPE that consists of rege-
nerating the 11-cis-retinaldehyde chromophore, which is essen-
tial for PR function and vision. Because RPE cells downregulate
Rpe65 at the protein level but not the mRNA level when grown
in culture [31], we analyzed the expression levels of four other
key genes of the retinoid cycle. We found that premature
senescence altered this pathway in ARPE-19 cells. RBP1 gene
expression was substantially upregulated (4.9 times higher than
in control cells), whereas RLBP1 and RDH5 gene expression
increased modestly (Table 3). On the other hand, LRAT gene
expression decreased (Table 3).
Fig. 5. Effects of premature senescence on the expression ofkey genesinvolved in the defense system and the resistance of ARPE-19 cells to various types of cell-death inducers.
Premature senescence was induced in ARPE-19 cells, as described for Fig. 3. Analysis of protein levels and functional assays was undertaken 3 days after cell replating. (A) Protein
(30 μg) was reduced and subjected to SDS–PAGE and Western blot analysis with specific anti-BAX, anti-SOD2, anti-HSP90, and anti-b-actin, as described under Materials and
methods. (B) Intracellular production of GSH was detected and quantified using a colorimetric method, as described under Materials and methods. (C) The effects of acute treatment
with 500 μMtert-BHP, 2 μMAβ(1–42), or 2 μMAβ(1–40) were analyzed after 24 h culture by the MTT assay. The percentage of cell survival was calculated and compared for
untreated control and prematurely senescent ARPE-19 cells. Values are means ± SEM. Differences between groups were tested by the Kruskal–Wallis test (D: p b0.0001). Pairwise
comparisons were performed with the Mann–Whitney test (⁎pb0.05, ⁎⁎pb0.005). Similar results were obtained in three independent experiments.
1355A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
Premature senescence induces loss of adhesion strength and
alters permeability in ARPE-19 cells
RPE cell detachment and pathological changes in Bruch’s
membrane are associated with RPE aging and AMD [32]. Our
results from the array, which show the effects of premature
senescence on the expression of genes involved in the cell
remodeling process (Table 1), suggest that these prematurely
senescent ARPE-19 cells may have adhesion defects. We
looked first at several genes involved in cell–extracellular ma-
trix interactions (Fig. 8A). Expression of COL1 and FN1 de-
creased by 75 and 62%, respectively (Table 3), suggesting that
premature senescence reduced the attachment ability of ARPE-
19 cells. Accordingly we compared the attachment strength of
prematurely senescent and control cells. Attachment of the latter
to the surface of the dish increased time-dependently and in
great numbers for 120 min (Fig. 7A). The number of attached
cells then reached a plateau (Fig. 7A). In contrast, prematurely
senescent ARPE-19 cells attached much more poorly, with cell
attachment 23 and 33% of that of the control cells at 120 and
360 min, respectively. A fibronectin coating did not improve
adhesion by the control cells, but it completely reversed the
prematurely senescent phenotype of the ARPE-19 cells, increa-
sing their ability to attach (Fig. 7B). This demonstrates a key
role for fibronectin in ARPE-19 cell premature senescence and
adhesion strength.
Cell adhesion is crucial to the formation and maintenance of
coherent multicellular structures, including the tight junctions
that maintain the selective permeability of the BRB [33].We
analyzed the expression patterns of several key genes involved
in tight junction complexes and found that premature sene-
scence induced changes in the mRNA levels of several, inclu-
ding CX43 and CLDN1 (Table 3). Changes in the expression of
these genes accompanied alteration of the cytoskeleton in pre-
maturely senescent ARPE-19 cells. The morphology of the
control remained normal, with cortical actin staining surround-
ing the cytosol and very few stress fibers (Fig. 8A). In contrast,
cortical actin staining of the prematurely senescent ARPE-19
cells showed numerous actin stress fibers and defects (Fig. 8A).
Occludin staining encircled the periphery of the control cells,
whereas there were clear breaks in this in the prematurely
senescent cells (Fig. 8A). OCLN expression nonetheless re-
mained unchanged (Table 3). Premature senescence signifi-
cantly affected transepithelial permeability, both inward and
outward for each FITC–dextran probe (Fig. 8B). Specifically,
after 6 h of transepithelial flux, outward permeability was 3
times greater and inward permeability 2.7 times greater for the
40-kDa FITC–dextran probe in the prematurely senescent cells.
These results strongly indicate that premature senescence cau-
ses permeability dysfunction at the molecular and cellular level
in ARPE-19 cells.
Prematurely senescent ARPE-19 cells have proangiogenic
properties
Array analysis showed an increase in BMP2 expression.
BMP2 is a member of the TGF-βfamily, which is involved in
various processes including premature and replicative senes-
cence, permeability, and angiogenesis [34–36]. We used qRT-
PCR for additional characterization of the expression of several
growth factors regulating angiogenesis, a process involved in
neovascular AMD. Expression of antiangiogenic TGFβ2and
PEDF did not change in the prematurely senescent cells, but
antiangiogenic TGFβ1mRNA was more abundant and proan-
giogenic VEGF mRNA less abundant than in control ARPE-19
cells (Table 3). TPA and PLAU, encoding the proangiogenic
tPA and urokinase enzymes, respectively, participate in angio-
genesis by degrading the extracellular matrix [37]. The pre-
maturely senescent cells had more abundant mRNA for both
enzymes (Table 3) as well as greater tPA production and secre-
tion than the control cells (Fig. 9A). Premature senescence thus
disturbed the balance in the expression of pro- and antiangio-
genic factors. We therefore performed a coculture assay to
determine whether premature senescence of ARPE-19 cells
could modulate CEC proliferation. This proliferation increased
by a factor of 1.4 at day 3 and of 1.8 at day 5 of coculture in
prematurely senescent compared with control cells (Fig. 9B).
Fig. 6. Effects of premature senescence on amyloidogenesis. Prematuresenescence
was induced in ARPE-19 cells, as described in the legend for Fig. 3.Analysisof
protein levels and functional assays were undertaken 3 days after cell replating.
(A) Protein (30 μg) was reduced and subjected to SDS–PAGE and Western blot
analysiswith specific anti-APP, anti-apolipoprotein J, and anti-b-actin, as described
under Materials and methods. (B)Aβ(1–42) secretionwas quantified by ELISA, as
described under Materials and methods. Values are means ± SEM. Pairwise
comparisons were analyzed with the Mann–Whitney test (*pb0.05). Similar
results were obtained in three independent experiments.
1356 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
FGF2 and VEGF have been suggested as the main proangio-
genic factors in the choroid. The failure of treatment with either
the FGF-specific pharmacological inhibitor SU5402 or an anti-
FGF2 blocking antibody (data not shown) to reverse the in-
creased CEC proliferation showed that this process is not
mediated through FGF. At the same time, the decrease in VEGF
expression in prematurely senescent cells ruled out its direct
role in this proliferation.
Discussion
Cumulative oxidative stress induces premature senescence of
RPE cells
The aim of this study was to investigate the role of cumu-
lative oxidative stress on human RPE cell senescence and to
analyze the key changes in gene expression in prematurely
senescent RPE cells. Cumulative oxidative stress has been
implicated in aging. We demonstrated that repeated treatments
of ARPE-19 cells with sublethal tert-BHP induced premature
senescence, characterized by the appearance of four major bio-
markers of senescence, including SA β-gal activity, irreversible
growth arrest, cell enlargement, and lack of response to mito-
genic stimuli.
SA β-gal activity in our experimental ARPE-19 cells is con-
sistent with that observed in replicative senescence of the RPE
340 cell line and in the RPE of aged primates [38,39].Cellgrowth
arrest of prematurely senescent ARPE-19 cells is also consistent
with the correlation between chronological age and the number
of large nondividing human RPE cells [40]. Moreover, this en-
largement of prematurely senescent ARPE-19 cells may be
related to RPE cell hyperplasia in the elderly [41]. In the repli-
catively senescent RPE 340 cell line, the greatest changes detect-
ed were the overexpression of IGFBP2 and the underexpression
of FGF2 and FGF5, whereas expression of PEDF,TGFβ2,
VEGF,andPDGF remained constant [42].Wealsosawno
changes in expression for PEDF or TGFβ2, but found under-
expression of VEGF in prematurely senescent RPE cells.
Senescent cells often underexpress CDK2 and CDK4 and
overexpress MMP3 and CCDN1 [21,43], as we observed in
prematurely senescent ARPE-19 cells. Prematurely senescent
ARPE-19 cells also overexpressed APOJ, as previously observed
in replicatively senescent primary cultures of RPE cells [44].
Thus, although no marker of senescence has so far been associated
with either normal RPE aging or the pathophysiology of AMD,
cell senescence markers detected in prematurely senescent ARPE-
19 cells may be potential markers for RPE aging. We cannot,
however, rule out the possibility that some transcriptomic changes,
such as SOD2 overexpression, may be due to general stress rather
than to premature senescence. However, several results are
consistent with previous transcriptional analysis of senescent RPE
cells. For example, GPX3 has been found to be overexpressed in
old mouse RPE [10], which may be related to the overexpression
of GPX1A observed in prematurely senescent ARPE-19 cells.
Fig. 7. Effects of premature senescence on cell adhesion. Premature senescence was induced in ARPE-19 cells, as described in the legend for Fig. 3. (A) Cell adhesion
was followed by counting surface-attached cells 10, 30, 120, and 360 min after cell replating, as described under Materials and Methods. (B) The effect of fibronectin
on adhesion was analyzed as described under Materials and methods ((+) with, (−) without). Cell adhesion was quantified as in (A) 360 min after cell replating. Values
are means ± SEM. Differences between groups were tested by the Kruskal–Wallis test (B: p=0.0197). Pairwise comparisons were analyzed with the Mann–Whitney
test (ns, nonsignificant; ⁎pb0.05; ⁎⁎pb0.005). Similar results were obtained in three independent experiments.
1357A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
Are premature senescence-associated altered ARPE-19 cell
genes related to AMD?
Although our interpretation is limited by an experimental
system based on an RPE cell line, we observed several coherent
transcriptomic and functional changes characteristic of senescent
RPE cells, aged animal models, human aging, and even AMD
features. They are (1) oxidative stress; (2) drusen; (3) alteration of
RPE cytoskeleton, adhesion, migration, and BRB breakdown;
and (4) neovascularization [26,45,46].
(1) RPE cells are continuously assaulted by high levels of
ROS, and growing evidence 7suggests that cumulative oxidant
injury to the RPE is involved in AMD [3,5]. Oxidative damage to
lipids and nucleic acids was observed in prematurely senescent
ARPE-19 cells. Increased oxidative damage in the retina of
patients with AMD has recently been demonstrated directly, with
the same markers we used [47]. It therefore seemed relevant to
study the effect of cumulative sublethal oxidative stress on RPE
cell functions to understand the mechanisms underlying RPE
aging and age-related RPE pathologies such as AMD.
Fig. 8. Effects of premature senescence on ARPE-19 cell tight junctions and monolayer permeability. Premature senescence was induced in ARPE-19 cells, as
described in the legend for Fig. 3. Immunocytochemistry and functional assays were undertaken 1 week after cell replating. (A) Actin and occludin localization was
determined by immunocytochemistry, as describedunder Materials and methods. DAPI was used to counterstain cell nuclei. (⁎, stress fibers and increased cytoplasmic
localization; arrowheads, clear breaks in the cell periphery staining; scale bar, 100 μm.) (B) The effect of senescence on the transepithelial permeability was analyzed
with FITC–dextran over a 6-h culture period, inward (upper to lower chamber) and outward (lower to upper chamber). Transwell inserts without cells reflected FITC–
dextran passive flux. Control and prematurely senescent ARPE-19 cells were observed by phase-contrast microscopy at 5× magnification to ensure the confluence of
the ARPE-19 monolayer. Values are means± SEM. Differences between groups were tested by the Kruskal–Wallis test (C: pb0.0001). Pairwise comparisons were
analyzed with the Mann–Whitney test (⁎pb0.05, ⁎⁎pb0.005). Similar results were obtained in three independent experiments.
1358 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
(2) Lysosomes and the ubiquitin proteasome pathway are the
two major proteolytic systems in which functional decline and
loss have been reported in aging and postmitotic senescence
[48,49]. This decline was confirmed by a decrease in the
expression of CTSZ and CTSH, which encode cathepsin Z and
cathepsin H, respectively. Interestingly, levels of cathepsin D
expression increased in prematurely senescent ARPE-19 cells.
This increase may be linked to the increase in cathepsin D
mRNA detected in old mouse RPE [10] and to cathepsin D
immunoreactivity observed around hyalinized drusen of AMD
patients [50]. Premature senescence of ARPE-19 cells modu-
lated the expression of other proteolytic enzymes involved in
the production of drusen components. Drusen, which are cold
spots for proteolysis and Aβ, are associated with aging, RPE
atrophy, and retinal degeneration [51]. Their accumulation is a
major risk factor for AMD. Some of the metabolic products that
accumulate in drusen, such as Aβ, are not inert and may be
involved in retinal degeneration [32,51,52]. We observed a
synergic transcriptome modulation of enzymes involved in Aβ
formation (BACE1,NEP,APOJ ), which may be consistent
with the Aβdeposition observed within drusen during aging
and in AMD patients [51]. Overexpression of TPA and over-
secretion of tPA in prematurely senescent RPE cells can also be
linked to Aβtoxicity: high levels of tPA localize to Aβ-rich
areas and mediate Aβ-induced apoptosis of hippocampal
neurons [53]. The functional consequences of these altera-
tions are not known in vivo. Therefore, it would be of interest
to test our hypothesis by analyzing the effect of Aβtoxicity
in vivo.
(3) The transcription of genes involved in adhesion,
cytoskeleton stability, and the assembly of tight junctions—
primordial for maintaining selective permeability of the BRB—
also changed significantly in prematurely senescent ARPE-19
cells. In our experimental model, underexpression of COL1 and
FN1 was associated with a defect in cell adhesion. This may be
related to RPE detachment observed during early AMD [45].
Breaks in the pattern of occludin staining were also observed in
the periphery of the cells. This finding indicates that senescence
causes permeability dysfunctions by affecting the cytoskeleton,
cell adhesion, and tight junctions at the molecular and cellular
levels in RPE cells. This is consistent with the report by Bailey
et al. that acute oxidative stress can affect RPE cell tight
junctions [16] and may participate to the BRB breakdown
during AMD [54].
(4) The prematurely senescent ARPE-19 cells showed pro-
angiogenic properties due neither to the overexpression of
proangiogenic VEGF or FGF2 nor to the underexpression of
antiangiogenic TGFβ1or TGFβ2. tPA, which mediates vascular
endothelial cell proliferation [55], may be involved in the
stimulation of choroidal vascular endothelial cell proliferation
by the prematurely senescent cells of our model, which
overexpress and oversecrete this plasminogen activator. It
would be of interest to investigate the expression levels of tPA
in AMD patients.
In conclusion, we showed that chronic exposure of ARPE-
19 cells to oxidative stress induced their premature senescence
and that this phenotype showed several features of RPE aging
and several processes observed during AMD. We suggest
that oxidative stress-induced prematurely senescent ARPE-19
cells may represent a potential in vitro model for RPE aging
analysis.
Acknowledgments
The authors thank Jo Ann Cahn for her careful proofrea-
ding of the manuscript. This work was supported by a grant
from European Grant EVI-GenoRet (LSHG-CT-2005-512036)
(A.-L.G., J.-Y.B., J.T., F.M.), the Ministère de la Recherche
(A.-L.G.), Retina France (A.-L.G.), and the French Federation
for the Blind and Visually Impaired.
Fig. 9. Effect of premature senescence on angiogenic properties of ARPE-19 cells. Premature senescence was induced in ARPE-19 cells, as described in the legend for
Fig. 3. (A) tPA secretion was measured by ELISA 3 days after cell replating as described under Materials and methods. (B) The angiogenic properties of ARPE-19 cells
were analyzed by a coculture system in Transwells with ARPE-19 cells in the upper chamber and CEC in the lower chamber, as described under Materials and
methods. CEC proliferation was assessed 3 and 5 days after the start of the ARPE-19/CEC coculture with an MTT assay. ARPE-19 and CEC media were replaced on
day 3 of the coculture. Values are means± SEM. Differences between groups were tested by the Kruskal–Wallis test (B: p= 0.040). Pairwise comparisons were
analyzed with Mann–Whitney test (⁎pb0.05, ⁎⁎⁎pb0.0005). Similar results were obtained in three independent experiments.
1359A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.freeradbiomed.2007.12.023.
References
[1] Tomany, S. C.; Wang, J. J.; Van Leeuwen, R.; Klein, R.; Mitchell, P.;
Vingerling, J. R.; Klein, B. E.; Smith, W.; De Jong, P. T. Risk factors for
incidentage-related macular degeneration:pooled findings from 3 continents.
Ophthalmology 111: 1280–1287; 2004.
[2] Wu, Z.; Lauer, T. W.; Sick, A.; Hackett, S. F.; Campochiaro, P. A.
Oxidative stress modulates complement factor H expression in retinal
pigmented epithelial cells by acetylation of FOXO3. J. Biol. Chem.
282:22414–22425; 2007.
[3] Beatty, S.; Koh, H.; Phil, M.; Henson, D.; Boulton, M. The role of
oxidative stress in the pathogenesis of age-related macular degeneration.
Surv. Ophthalmol. 45:115–134; 2000.
[4] Liang, F. Q.; Godley, B. F. Oxidative stress-induced mitochondrial
DNA damage in human retinal pigment epithelial cells: a possible mecha-
nism for RPE aging and age-related macular degeneration. Exp. Eye Res.
76:397–403; 2003.
[5] Decanini, A.; Nordgaard, C. L.; Feng, X.; Ferrington, D. A.; Olsen, T. W.
Changes in select redox proteins of the retinal pigment epithelium in age-
related macular degeneration. Am. J. Ophthalmol. 143:607–615; 2007.
[6] Debacq-Chainiaux, F.; Borlon, C.; Pascal,T.; Royer, V.; Eliaers, F.; Ninane, N.;
Carrard, G.; Friguet, B.; de Longueville, F.; Boffe, S.; Remacle, J.; Toussaint,
O. Repeated exposure of human skin fibroblasts to UVB at subcytotoxic level
triggers premature senescence through the TGF-beta1 signaling pathway.
J. Cell Sci. 118:743–758; 2005.
[7] de Magalhaes, J. P.; Chainiaux, F.; de Longueville, F.; Mainfroid, V.;
Migeot, V.; Marcq, L.; Remacle, J.; Salmon, M.; Toussaint, O. Gene
expression and regulation in H
2
O
2
-induced premature senescence of
human foreskin fibroblasts expressing or not telomerase. Exp. Gerontol.
39:1379–1389; 2004.
[8] Serrano, M.; Lin, A. W.; McCurrach, M. E.; Beach, D.; Lowe, S. W. Oncogenic
ras provokes premature cell senescence associated with accumulation of p53
and p16INK4a. Cell 88:593–602; 1997.
[9] Choo, K. W.; Kong, W. Identification of differentially expressed genes in
multiple microarray experiments using discrete Fourier transform. Front.
Biosci. 12:1845–1851; 2007.
[10] Ida, H.; Boylan, S. A.; Weigel, A. L.; Hjelmeland, L. M. Age-related changes
in the transcriptional profile of mouse RPE/choroid. Physiol. Genomics
15:258–262; 2003.
[11] Smith, C. P.; Sharma, S.; Steinle, J. J. Age-related changes in sympathetic
neurotransmission in rat retina and choroid. Exp. Eye Res. 84:75–81;
2007.
[12] de Magalhaes, J. P.; Chainiaux, F.; Remacle, J.; Toussaint, O. Stress-induced
premature senescence in BJ and hTERT-BJ1 human foreskin fibroblasts.
FEBS Lett. 523:157–162; 2002.
[13] Hecquet, C.; Lefevre, G.;Valtink, M.;E ngelmann, K.;Mascarelli, F. Activation
and role of MAP kinase-dependent pathways in retinal pigment epithelium
cells: JNK1, P38 kinase, and cell death. Invest. Ophthalmol. Visual Sci.
44:1320–1329; 2003.
[14] Hoffmann, S.; Spee, C.;Murata, T.; Cui, J. Z.; Ryan, S. J.; Hinton, D. R. Rapid
isolation of choriocapillary endothelial cells by Lycopersicon esculentum-
coated Dynabeads. Graefe Arch. Clin. Exp. Ophthalmol. 236:779–784;
1998.
[15] Glotin, A. L.; Calipel, A.; Brossas,J. Y.;Faussat, A. M.;Treton, J.; Mascarelli,
F. Sustained versus transient ERK1/2 signaling underlies the anti- and
proapoptotic effects of oxidative stress in human RPE cells. Invest. Oph-
thalmol. Visual Sci. 47:4614–4623; 2006.
[16] Bailey, T. A.; Kanuga, N.; Romero, I. A.; Greenwood, J.; Luthert, P. J.;
Cheetham, M. E. Oxidative stress affects the junctional integrity of retinal
pigment epithelial cells. Invest. Ophthalmol. Visual Sci. 45:675–684;
2004.
[17] Pitkanen, L.; Ranta, V. P.; Moilanen, H.; Urtti, A. Permeability of retinal
pigment epithelium: effects of permeant molecular weight and lipophili-
city. Invest. Ophthalmol. Visual Sci. 46:641–646; 2005.
[18] Stein, G. H.; Dulic, V. Origins of G1 arrest in senescent human fibroblasts.
Bioessays 17:537–543; 1995.
[19] Dimri, G. P.; Lee, X.; Basile, G.; Acosta, M.; Scott, G.; Roskelley, C.;
Medrano, E. E.; Linskens, M.; Rubelj, I.; Pereira-Smith, O.; et al. A
biomarkerthat identifies senescent human cells in culture and in aging skin in
vivo. Proc. Natl. Acad. Sci. USA 92:9363–9367; 1995.
[20] Vojta, P. J.; Barrett, J. C. Genetic analysis of cellular senescence. Biochim.
Biophys. Acta 1242:29–41; 1995.
[21] Fukami, J.; Anno, K.; Ueda, K.; Takahashi, T.; Ide, T. Enhanced ex-
pression of cyclin D1 in senescent human fibroblasts. Mech. Ageing Dev.
81:139–157; 1995.
[22] Dumont, P.; Burton, M.; Chen, Q. M.; Gonos,E . S.; Frippiat, C.; Mazarati, J. B.;
Eliaers, F.; Remacle, J.; Toussaint, O. Induction of replicative senescence
biomarkers by sublethal oxidative stresses in normal human fibroblast. Free
Radic. Biol. Med. 28:361–373; 2000.
[23] Trougakos, I. P.; Pawelec, G.; Tzavelas, C.; Ntouroupi, T.; Gonos, E. S.
Clusterin/Apolipoprotein J up-regulation after zinc exposure, replicative
senescence or differentiation of human haematopoietic cells. Biogeron-
tology 7:375–382; 2006.
[24] Marcotte, R.; Lacelle, C.; Wang, E. Senescent fibroblasts resist apoptosis
by downregulating caspase-3. Mech. Ageing Dev. 125:777–783; 2004.
[25] Wang, E. Senescent human fibroblasts resist programmed cell death, and
failure to suppress bcl2 is involved. Cancer Res. 55:2284–2292; 1995.
[26] Johnson, L. V.; Leitner, W. P.; Rivest, A. J.; Staples, M. K.; Radeke, M. J.;
Anderson, D. H. The Alzheimer’s A beta-peptide is deposited at sites of
complement activation in pathologic deposits associated with aging
and age-related macular degeneration. Proc. Natl. Acad. Sci. USA
99:11830–11835; 2002.
[27] Dewachter, I.; Van Dorpe, J.; Smeijers, L.; Gilis, M.; Kuiperi, C.; Laenen, I.;
Caluwaerts, N.; Moechars, D.; Checler, F.; Vanderstichele, H.; Van Leuven,
F. Aging increased amyloid peptide and caused amyloid plaques in brain of
old APP/V717I transgenic mice by a different mechanism than mutant
presenilin 1. J. Neurosci. 20:6452–6458; 2000.
[28] Kern, A.; Roempp, B.; Prager, K.; Walter, J.; Behl, C. Down-regulation of
endogenous amyloid precursor protein processing due to cellular aging.
J. Biol. Chem. 281:2405–2413; 2006.
[29] Kimura, N.; Imamura, O.; Ono, F.; Terao, K. Aging attenuates dynactin–
dynein interaction: down-regulation of dynein causes accumulation of
endogenous tau and amyloid precursor protein in human neuroblastoma
cells. J. Neurosci. Res. 85:2909–2916; 2007.
[30] Yerbury, J. J.; Poon, S.; Meehan, S.; Thompson, B.; Kumita, J. R.; Dobson,
C. M.; Wilson, M. R. The extracellular chaperone clusterin influences
amyloid formation and toxicity by interacting with prefibrillar structures.
FASEB J. 21:2312–2322; 2007.
[31] Hamel, C. P.; Tsilou, E.; Pfeffer, B. A.; Hooks, J. J.; Detrick, B.; Redmond,
T. M. Molecular cloning and expression of RPE65, a novel retinal pigment
epithelium-specific microsomal protein that is post-transcriptionally
regulated in vitro. J. Biol. Chem. 268:15751–15757; 1993.
[32] Hageman, G. S.; Luthert, P. J.; Victor Chong, N. H.; Johnson, L. V.;
Anderson, D. H.; Mullins, R. F. An integrated hypothesis that considers
drusen as biomarkers of immune-mediated processes at the RPE–Bruch’s
membrane interface in aging and age-related macular degeneration. Prog.
Retin. Eye Res. 20:705–732; 2001.
[33] Rizzolo, L. J. Polarity and the development of the outer blood–retinal
barrier. Histol. Histopathol. 12:1057–1067; 1997.
[34] Bani-Yaghoub, M.; Felker, J. M.; Sans, C.; Naus, C. C. The effects of bone
morphogenetic protein 2 and 4 (BMP2 and BMP4) on gap junctions during
neurodevelopment. Exp. Neurol. 162:13–26; 2000.
[35] Frippiat, C.; Chen,Q. M.; Zdanov, S.; Magalhaes, J. P.; Remacle, J.; Toussaint,
O. Subcytotoxic H
2
O
2
stress triggers a release of transforming growth factor-
beta 1, which induces biomarkers of cellular senescence of human diploid
fibroblasts. J. Biol. Chem. 276:2531–2537; 2001.
[36] Rothhammer,T.; Poser, I.;Soncin, F.; Bataille, F.; Moser, M.; Bosserhoff, A. K.
Bone morphogenetic proteins are overexpressed in malignant melanoma and
promote cell invasion and migration. Cancer Res. 65:448–456; 2005.
1360 A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
[37] Mignatti, P.; Rifkin, D. B. Plasminogen activators and matrix metallopro-
teinases in angiogenesis. Enzyme Protein 49:117–137; 1996.
[38] Mishima, K.; Handa, J. T.; Aotaki-Keen, A.; Lutty, G. A.; Morse, L. S.;
Hjelmeland, L. M. Senescence-associated beta-galactosidase histochemistry
for the primate eye. Invest. Ophthalmol. Visual Sci. 40:1590–1593; 1999.
[39] Matsunaga, H.; Handa, J. T.; Aotaki-Keen,A.; Sherwood, S. W.; West,M. D.;
Hjelmeland, L. M. Beta-galactosidase histochemistry and telomere loss in
senescent retinal pigment epithelial cells. Invest. Ophthalmol. Visual Sci.
40:197–202; 1999.
[40] Flood, M. T.; Gouras, P.; Kjeldbye, H. Growth characteristics and ultra-
structure of human retinal pigment epithelium in vitro. Invest. Ophthalmol.
Visual Sci. 19:1309–1320; 1980.
[41] Boulton, M.; Roanowska, M.; Wess, T. Ageing of the retinal pigment
epithelium: implications for transplantation. Graefe Arch. Clin. Exp. Ophthal-
mol. 242:76–84; 2004.
[42] Matsunaga, H.; Handa, J. T.; Gelfman, C. M.; Hjelmeland, L. M. The mRNA
phenotype of a human RPE cell line at replicative senescence. Mol. Vis. 5:39;
1999.
[43] Lucibello, F. C.; Sewing, A.; Brusselbach, S.; Burger, C.; Muller, R.
Deregulation of cyclins D1 and E and suppression of cdk2 and cdk4 in
senescent human fibroblasts. J. Cell Sci. 105 (Pt 1):123–133; 1993.
[44] Wang, X. F.; Cui, J. Z.; Nie, W.; Prasad, S. S.; Matsubara, J. A. Differential
gene expression of early and late passage retinal pigment epithelial cells. Exp.
Eye Res. 79:209–221; 2004.
[45] Ambati, J.; Ambati, B. K.; Yoo, S. H.; Ianchulev, S.; Adamis, A. P. Age-related
macular degeneration: etiology, pathogenesis, and therapeutic strategies. Surv.
Ophthalmol. 48:257–293; 2003.
[46] Chong, N. H.; Keonin, J.; Luthert, P. J.; Frennesson, C. I.; Weingeist, D. M.;
Wolf, R. L.; Mullins, R. F.; Hageman, G. S. Decreased thickness and integrity
of the macular elastic layer of Bruch’s membrane correspond to the dis-
tribution of lesions associated with age-related macular degeneration. Am.
J. Pathol. 166:241–251; 2005.
[47] Shen, J. K.;Dong, A.; Hackett,S. F.; Bell,W. R.; Green,W. R.; Campochiaro,
P. A. Oxidative damage in age-related macular degeneration. Histol. His-
topathol. 22:1301–1308; 2007.
[48] Chondrogianni, N.; Gonos, E. S. Proteasome dysfunction in mammalian
aging: steps and factors involved. Exp. Gerontol. 40:931–938; 2005.
[49] Grune, T.; Merker, K.; Jung, T.; Sitte, N.; Davies, K. J. Protein oxidation
and degradation during postmitotic senescence. Free Radic. Biol. Med.
39:1208–1215; 2005.
[50] Rakoczy, P. E.; Sarks, S. H.; Daw, N.; Constable, I. J. Distribution of
cathepsin D in human eyes with or without age-related maculopathy. Exp.
Eye Res. 69:367–374; 1999.
[51] Luibl, V.; Isas, J. M.; Kayed, R.; Glabe, C. G.; Langen, R.; Chen, J. Drusen
deposits associated with aging and age-related macular degeneration
contain nonfibrillar amyloid oligomers. J. Clin. Invest. 116:378–385; 2006.
[52] Yoshida, T.; Ohno-Matsui, K.; Ichinose, S.; Sato, T.; Iwata, N.; Saido, T. C.;
Hisatomi, T.; Mochizuki, M.; Morita, I. The potential role of amyloid beta
in the pathogenesis of age-related macular degeneration. J. Clin. Invest.
115:2793–2800; 2005.
[53] Medina, M. G.; Ledesma, M. D.; Dominguez, J. E.; Medina, M.; Zafra, D.;
Alameda, F.; Dotti, C. G.; Navarro, P. Tissue plasminogen activator medi-
ates amyloid-induced neurotoxicity via Erk1/2 activation. EMBO J.
24:1706–1716; 2005.
[54] Campochiaro, P. A. Retinal and choroidal neovascularization. J. Cell.
Physiol. 184:301–310; 2000.
[55] Welling, T. H.; Huber, T. S.; Messina, L. M.; Stanley, J. C. Tissue plas-
minogen activator increases canine endothelial cell proliferation rate
through a plasmin-independent, receptor-mediated mechanism. J. Surg.
Res. 66:36–42; 1996.
1361A.-L. Glotin et al. / Free Radical Biology & Medicine 44 (2008) 1348–1361
