The matricellular protein CCN5 inhibits fibrotic deformation of retinal pigment epithelium
The matricellular protein CCN5 inhibits fibrotic deformation of retinal pigment epithelium
Aeri Yoon 0 1 2
Sora Im 0 1 2
Juyeon Lee 0 1 2
Daeho Park 0 1 2
Dong Hyun Jo 0 2
Jin Hyoung Kim 0 2
Jeong Hun Kim 0 2
Woo Jin ParkID 0 1 2
0 Medical Technology Development Program of the National Research Foundation , and MSIP (NRF- 2015M3A9E6028951, 2015M3A9E6028949, and 2017M3A9B4062654) , the Development of Platform Technology for Innovative Medical Measurements Program from the Korea Research Institute of Standards and Science (KRISS-2017- GP2017-0020), the Basic Research Laboratory
1 College of Life Sciences, Gwangju Institute of Science and Technology , Cheomdangwagi-ro, Buk-gu, Gwangju , Republic of Korea, 2 Department of Biomedical Sciences, Seoul National University College of Medicine , Daehak-ro, Jongno-gu, Seoul , Republic of Korea, 3 Fight Against Angiogenesis-Related Blindness Laboratory, Biomedical Research Institute, Seoul National University Hospital , Daehak-ro, Jongno-gu, Seoul , Republic of Korea
2 Editor: Paloma B. Liton, Duke University , UNITED STATES
Retinal pigment epithelium (RPE) plays an essential role in maintaining retinal function, and its defect is thought to be critically implicated in various ocular disorders. This study demonstrated that the matricellular protein CCN5 was down-regulated in ARPE-19 cells treated with the pro-fibrotic agent transforming growth factor (TGF)-β. A recombinant adenovirus expressing CCN5 (AdCCN5) was used to restore the level of CCN5 in these cells. AdCCN5 prevented TGF-β-induced fibrotic changes, including disruption of tight junctions, up-regulation of mesenchymal marker proteins, and down-regulation of epithelial marker proteins. In addition, AdCCN5 prevented TGF-β-induced functional defects, including increased migratory activity and reduced phagocytic activity. Notably, AdCCN5 reversed morphological and functional defects pre-established by TGF-β prior to viral infection. The CCN5 level was down-regulated in RPE of 18-month-old Ccl2-/- mice, which exhibited retinal defects. Restoration of the CCN5 level via intravitreal injection of a recombinant adeno-associated virus expressing CCN5 (AAV9-CCN5) normalized the altered expression of mesenchymal, epithelial, and functional marker proteins, as assessed by western blotting and immunohistochemistry. Taken together, these data suggest that down-regulation of CCN5 is associated with fibrotic deformation of RPE under pathological conditions and that restoration of the CCN5 level effectively promotes recovery of deformed RPE.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Retinal pigment epithelium (RPE) is a highly polarized monolayer of cuboidal cells located
between the neural retina and choroid that plays a pivotal role in maintaining retinal function
]. Normal RPE has a mature epithelial phenotype with morphologically and functionally
asymmetric structures [
]. Deformation of RPE cells is regarded as a primary event leading to
progressive fibrotic diseases of the eye, including proliferative vitreoretinopathy (PVR),
diabetic retinopathy (DR), and age-related macular degeneration (AMD) [
]. In pathological
(2016R1A4A1009895), the Basic Science
Research Program through the National Research
Foundation of Korea funded by the Ministry of
Education (2017R1A2B4007340 and
2017R1A6A3A04004741), and a grant from the
Systems Biology Infrastructure Establishment
provided by the Gwangju Institute of Science and
conditions, RPE cells undergo fibrotic deformation characterized by loss of cellular integrity
and perturbation of functions, such as phagocytosis and directional transcytosis [
pathological process is thought to be triggered by a variety of intravitreal cytokines, among
which transforming growth factor (TGF)-β is the best characterized [
]. In vitro experiments
demonstrated that TGF-β contributes to an aberrant wound response in RPE [
] and that its
level significantly correlates with excessive deposition of extracellular matrix components in
]. In addition, the TGF-β level is significantly elevated in patients with PVR, DR, and
]. These observations suggest that TGF-β-mediated fibrotic deformation of RPE
plays a role in various retinal pathologies.
CCN5 (also known as Wnt-1-induced signaling protein-2; WISP2) is a member of the CCN
family of matricellular proteins (CCN1–6) [
] and has been implicated in a variety of human
]. The role of CCN5 in human pathology is mainly attributed to its function in
proliferation and mobilization of breast cancer and pancreatic adenocarcinoma cells [
proliferation of vascular smooth muscle cells . We previously showed that CCN5 prevents
pressure overload-induced cardiac fibrosis partly by inhibiting the TGF-β-SMAD signaling
pathway in mice [
]. At the cellular level, CCN5 inhibits endothelial-mesenchymal transition
and transdifferentiation of fibroblasts into myofibroblasts in the heart, the two most critical
processes underlying cardiac fibrosis. We further showed that CCN5 can reverse
pre-established cardiac fibrosis by inducing apoptosis specifically in myofibroblasts, but not in
fibroblasts or myocytes [
]. CCN5 also induces phenotypic reversion in some populations of
myofibroblasts (unpublished observations).
The present study demonstrated that CCN5 was markedly down-regulated in RPE cells
treated with TGF-β. Restoration of the CCN5 level via infection of a recombinant adenovirus
harboring CCN5 (AdCCN5) prevented and reversed TGF-β-mediated fibrotic deformation of RPE
cells. CCN5 elicited similar beneficial effects in aged Ccl2-/- mice, which exhibited RPE
deformation. Collectively, this study suggests that CCN5 is a useful target for the treatment of various
ocular diseases that are primarily caused or accompanied by fibrotic deformation of RPE.
Materials and methods
All animal experimental methods and protocols in this study were approved by the
Institutional Animal Care and Use Committee of the School of Life Sciences, Gwangju Institute of
Science and Technology, and carried out in accordance with their approved guidelines
(IACUC GIST-2015-24). Ccl2-/- mice (Jackson Laboratories, USA) generated as described
] were backcrossed ten times with C57BL/6 wild type (WT) mice (Damul
Science, Korea). AAV9-VLP or AAV9-CCN5 was delivered into 18-month-old Ccl2-/- mice for 12
weeks. Age-matched WT littermates administered AAV9-VLP were used as a control. Surgery
was performed under deeply anesthetized via intraperitoneally injecting a mixture of Zoletil 50
(Virbac, France) and Rompun (Bayer Korea, Korea) at a ratio of 3:1 (1 ml/kg), and all efforts
were made to minimize suffering. After surgical procedure, mice were monitored in every
other day to check whether any adverse events were occurred. As far as we have observed, we
have not found any other adverse clinical signs or gross abnormalities in other organ systems
to date. At the end of experiments, we administered CO2 inhalation for euthanasia and the
eyes were enucleated immediately.
AAV9-CCN5 was intravitreally injected using a Nanofil syringe with a 33G blunt needle
(World Precision Instruments Inc., USA) underneath an operating microscope (Leica
2 / 15
Microsystems Ltd., Germany). The same concentration of AAV9-VLP was injected into
Cells and cell culture
ARPE-19 cells were obtained from the American Type Culture Collection (ATCC, USA) and
cultured in a 1:1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F12 Nutrient
Mixture (DMEM/F12; Welgene, Korea) containing 10% fetal bovine serum (FBS; HyClone,
USA), 1 mM glutamine, and 100 U/ml penicillin/streptomycin at 37˚C in a humidified
atmosphere containing 5% CO2. The medium was changed every 2 days. Confluent cells were
dissociated with 0.25% trypsin/0.02% ethylenediaminetetraacetic acid solution (Gibco, USA). Fully
confluent ARPE-19 cells were serum-starved for 24 hours, and then the medium was replaced
by that containing AdCCN5 at a MOI of 50 or 10 ng/ml TGF-β2 (PeproTech Korea, Korea)
prior to TGF-β treatment or AdCCN5 infection, respectively.
AAV and adenovirus vectors
An AAV serotype 9 construct harboring the human CCN5 gene under the control of the CMV
promoter was produced and purified by Virovek (USA). Full-length mouse CCN5 cDNA
tagged with hemagglutinin at the amino-terminal was generated and purified as previously
Western blot analysis
ARPE-19 cells were washed with phosphate-buffered saline (PBS) and harvested by scraping.
Cell pellets were suspended in cold RIPA lysis buffer (1% NP-40, 50 mM Tris-HCl [pH 7.4],
150 mM NaCl, and 10 mM NaF) supplemented with a mammalian cell protease inhibitor
cocktail (Roche, Germany) and sonicated at an amplitude of 50% for 5 minutes with a cycle of
2 seconds on/1 second off. To extract proteins from the RPE/choroid/sclera complex of
Ccl2-/mice, enucleated eyes without the neural retina were suspended in cold RIPA lysis buffer
containing protease inhibitor and sonicated at an amplitude of 50% for 10 seconds with a cycle of
2 seconds on/2 seconds off using a probe sonicator with a 1/8 inch microtip. Lysates were
centrifuged at 13,000 rpm for 20 minutes, and the protein concentration of the supernatant was
determined using a bicinchoninic acid protein assay kit (Thermo Fisher Scientific, USA).
Equal amounts of protein samples were mixed with 5× SDS sample buffer, subjected to
SDS-PAGE, and transferred to polyvinylidene difluoride membranes (Millipore, USA).
Membranes were blocked with Tris-buffered saline-Tween 20 containing 5% non-fat dry milk
powder for 1 hour, incubated overnight at 4˚C with specific primary antibodies (S1 Table), and
then labeled with the appropriate horseradish peroxidase-conjugated secondary antibodies
(Invitrogen, USA). Immune complexes were visualized using an enhanced chemiluminescence
reagent (Amersham, UK) and an ImageQuant LAS 4000 Mini imager (GE Healthcare, UK).
The intensity of each protein band was quantified using NIH ImageJ software (NIH, USA).
Cells cultured on coverslips in a 12-well culture plate were fixed with 4% paraformaldehyde
for 10 minutes and permeabilized with 0.2% Triton X-100 prepared in PBS for 10 minutes.
After blocking for 1 hour with 5% bovine serum albumin prepared in PBS at room
temperature, cells were incubated with primary antibodies overnight at 4˚C and then with FITC- or
Alexa Fluor 594-conjugated secondary antibodies (Invitrogen, USA) for 1 hour at room
temperature. Cells were stained with Texas Red-conjugated phalloidin to visualize F-actin. Nuclei
3 / 15
were stained with Hoechst 33342 (Invitrogen, USA). Samples were analyzed underneath a
microscope equipped with a 20× objective lens and epifluorescence filters (Zeiss, Germany).
For flat mount of RPE/choroid/sclera complex, enucleated eyes were dissected to remove
neural retina and gently flat mounted and fixed in methanol for 15 minutes at -20˚C. After
washing with PBS, RPE complex were incubated in 0.2% Triton X-100 for 2 hours at 37˚C. After
blocking for an hour at 37˚C, the RPE complex were incubated at 4˚C overnight with primary
antibody. After washing with PBS, the RPE complex were incubated for 2 hours at 37˚C with
secondary antibodies. After washing with PBS, flat mounts were counterstained with 10 mg/
mL of 4, 6-diamidino-2-phenolindole (DAPI; Sigma-Aldrich, USA). After washing with PBS,
the RPE complex were mounted with aqueous mounting medium and observed under
fluorescence microscope. Information about the primary antibodies is provided in S1 Table.
RNA extraction and quantitative Real-time PCR analysis
Total RNA was isolated from treated cells using Trizol reagent (Invitrogen, USA) according to
the manufacturer’s instruction. cDNA was synthesized with a reverse transcription kit
(Promega, USA) and amplified via Real-time PCR using the SYBR green (Takara, Japan). The
specificity of the amplification reactions was confirmed by melting curve analysis. Samples were
assayed in triplicate or quadruplicate. RNA was quantified using the ΔΔCt method of relative
quantification (RQ) and the transcript levels were normalized to the endogenous control 18s
rRNA. Information of the primer sequence is shown in S2 Table.
Collagen gel contraction assay
Collagen gel contraction assays were performed as previously described [
]. In brief, 1.2 mg/
ml collagen solution (Invitrogen, USA), 10× DMEM/F12, 0.5N NaOH and ARPE-19 cells
(1 × 105 cells per well) were mixed and added to a 24-well culture plate. After incubation for 1
hour at 37˚C to allow collagen polymerization, the gelatinous suspension was detached from
the well edges using a sterile pipette tip. The collagen gel was allowed to float in serum free
DMEM/F12 media containing TGF-β and AdLacZ or AdCCN5. ImageJ software was used to
calculate the extent of collagen gel contraction.
In vitro cell scratch assay
ARPE-19 cells were plated on coverslips in a 12-well culture plate in DMEM/F12 medium.
When the monolayer was fully confluent, the medium was replaced by serum-free medium
and cells were cultured for 12 hours. Following starvation, cells were infected with AdCCN5
for 24 hours prior to treatment with 10 ng/ml TGF-β for 48 hours in the prevention model.
Cells were treated with AdCCN5 and TGF-β in the opposite order in the reversion model. The
monolayer was scratched by gently and slowly dragging a pipette tip across the center of the
well. After 24 hours, cells were stained with DAPI and closure of the gap via cell migration was
monitored by microscopy. The gap width was measured using NIH ImageJ software.
ARPE-19 cells were seeded into a 24-well culture plate at a density of 5 × 104 cells per well and
treated with the indicated virus (AdLacZ or AdCCN5) and TGF-β. The effect of photodynamic
stress on phagocytic activity of the cells was determined by challenging control and
photodynamically treated cells with TAMRA-stained apoptotic thymocytes or 1 mg/ml pHrodo Red
BioParticles conjugates (Invitrogen, USA) in 5% CO2 at 37˚C for 6 and 4 hours, respectively.
To generate TAMRA-labeled apoptotic thymocytes, a thymus was acquired from a
5–6-week4 / 15
old C57BL/6 mouse and gently dissociated using a 5 ml syringe piston and a cell strainer to
separate single thymocytes. Thymocytes were stained with 50 μM TAMRA-SE (Invitrogen,
USA) prepared in HBSS in a CO2 incubator at 37˚C for 30 minutes, de-stained in RPMI
containing 10% FBS and 1% penicillin/ streptomycin/ glutamine in 5% CO2 at 37˚C for 20
minutes, and washed once with complete RPMI. Apoptosis of thymocytes was induced by
treatment with 50 μM dexamethasone (Calbiochem, Germany) in a CO2 incubator at 37˚C
for 4 hours. Thereafter, cells were washed thrice with complete RPMI, and 20 × 105 apoptotic
thymocytes were resuspended in 300 μl of phagocyte culture medium. Apoptotic thymocytes
were incubated with treated ARPE-19 cells in 5% CO2 at 37˚C. Thereafter, phagocytes were
washed five times with ice-cold PBS, trypsinized, suspended in complete culture medium,
and analyzed using a FACSCantoII flow cytometer (BD, USA). ARPE-19 cells were gated
according to the forward scatter/side scatter plot to distinguish non-ingested apoptotic
thymocytes or BioParticles, cell fragments, and other debris from single cells. An appropriate
negative control (cells without apoptotic thymocytes or BioParticles) was used in each
experiment. After using the marker M1 to gate positive cells, the percentage of fluorescence-positive
events from 20,000 live cells per sample was calculated. Data were analyzed using FlowJo
Quantitative functional and gene expression assays were performed at least three times. Mean
averages ± S.D. were calculated. Two groups were compared using the two-tailed Student’s
ttest. Statistical significance was indicated by a single asterisk ( , p < 0.05) or a double asterisk
( , p < 0.01).
CCN5 prevents TGF-β-induced fibrotic deformation of ARPE-19 cells
ARPE-19 is a cell line spontaneously derived from human RPE. Post-confluent ARPE-19 cells
form a cobblestone-like monolayer and express typical epithelial marker proteins. ARPE-19
cells underwent phenotypic transformation and acquired an elongated and fibroblastic
morphology upon treatment with 5 or 10 ng/ml TGF-β for 48 hours (S1A Fig). Notably, TGF-β
treatment significantly decreased the CCN5 protein level. In parallel, levels of mesenchymal
marker proteins, including α smooth muscle actin (α-SMA), vimentin, fibronectin, and type I
collagen, were elevated, whereas levels of epithelial marker proteins, including zonula
occludens (ZO)-1 and occludin, were significantly reduced (S1B Fig). Furthermore,
immunocytochemistry revealed that the levels of α-SMA and filamentous (f-) actin were elevated upon
treatment with TGF-β (S2 Fig). This confirmed that TGF-β treatment induced fibrosis-like
phenotypic transformation of ARPE-19 cells. We thus treated ARPE-19 cells with 10 ng/ml
TGF-β for 48 hours to induce fibrotic deformation in all subsequent experiments. Cells were
infected with the recombinant adenovirus AdCCN5, which expressed CCN5 under the control
of the CMV promoter, or with the control virus AdLacZ. At 2 days post-infection, cells were
treated with TGF-β for an additional 2 days (Fig 1A). AdCCN5 markedly prevented the
morphological changes induced by TGF-β (Fig 1B). Western blotting showed that AdCCN5
prevented the TGF-β-induced increases and decreases in levels of mesenchymal and epithelial
marker proteins, respectively (Fig 1C). These changes were further confirmed by qRT-PCR
analyses (S3 Fig). Immunocytochemical staining of ZO-1 revealed that TGF-β markedly
disrupted tight junctions and that this was prevented by AdCCN5. AdCCN5 also prevented
TGFβ-mediated induction of α-SMA and appearance of f-actin, as detected by immunostaining
and phalloidin staining (Fig 1D). Enhanced contractility due to increased α-SMA expression is
5 / 15
Fig 1. CCN5 prevents TGF-β-induced fibroblast-like phenotypic transformation of ARPE-19 cells. (A) ARPE-19 cells were
grown until 100% confluent and cultured in the presence of TGF-β and AdLacZ or AdCCN5. (B) Bright field microscopy images
showing the change in the morphology of ARPE-19 cells from a monolayer of cuboidal cells to mesenchymal-like cells upon
treatment with 10 ng/ml TGF-β. This change was not observed in cells pretreated with AdCCN5. Original magnification: 100×. Scale
bar: 200 μm. (C) Total cell lysates obtained under the same conditions as in A were immunoblotted with antibodies against CCN5,
αSMA, vimentin, fibronectin, ZO-1, and occludin. GAPDH was used as a loading control. Quantified protein levels were normalized
against the loading control (bottom). Error bars = S.D. p < 0.05, p < 0.01. (D) Immunofluorescence images of ARPE-19 cells
stained for ZO-1 and α-SMA. Nuclei were counterstained with Hoechst. Images are representative of at least three independent
experiments. Original magnification: 200×. Scale bar: 50 μm. (E) Collagen gel lattices were generated and cultured under the same
conditions as shown in A. Areas of collagen gel contraction were measured after 48 hours. Representative images of collagen gel
lattices are shown (left), and areas of gel contraction were quantified (right) (n = 5). Error bars = S.D. p < 0.05, p < 0.01.
a hallmark of fibrotic transformation of epithelial cells. The collagen gel contraction assay
demonstrated that TGF-β increased contractility and that this was significantly inhibited by
AdCCN5 (Fig 1E). These results indicate that CCN5 prevents TGF-β-induced fibrotic
deformation of ARPE-19 cells.
6 / 15
CCN5 prevents TGF-β-induced functional deterioration of ARPE-19 cells
TGF-β triggers functional deterioration, as well as morphological transformation, of ARPE-19
cells. ARPE-19 cells were infected with AdCCN5 or AdLacZ for 2 days and then treated with
TGF-β for an additional 2 days. Thereafter, a scratch was made across the monolayer and cells
were examined underneath a microscope after 24 hours. TGF-β enhanced the migratory
activity of ARPE-19 cells, and this was significantly inhibited by AdCCN5 (Fig 2A). The RPE65
enzyme converts all-trans retinal into 11-cis retinal during the vision cycle and is essential for
Fig 2. CCN5 prevents TGF-β-induced functional deterioration of ARPE-19 cells. (A) ARPE-19 cells were cultured in the
presence of TGF-β and AdLacZ or AdCCN5 for 48 hours, scratched, fixed, and stained with Hoechst. Representative images are
shown (left), and migration distances were quantified (right) (n = 3). Error bars = S.D. p < 0.05, p < 0.01. (B) Expression of
the RPE function-related proteins RPE65 and MerTK was analyzed by western blotting under the same conditions as in A. Error
bars = S.D. p < 0.05. (C) Phagocytic function of ARPE-19 cells was analyzed using pHrodo BioParticles conjugates under the
same conditions as in A. TGF-β reduced phagocytic activity, and this was inhibited by pretreatment with AdCCN5. Phagocytosis
was quantified by flow cytometry. The relative fluorescence intensity of internalized pHrodo BioParticles was quantified (right)
(n = 9). Error bars = S.D. p < 0.01. (D) TAMRA-labeled apoptotic thymocytes were added to ARPE-19 cells under the same
conditions as shown in A and examined by flow cytometry. The relative fluorescence intensity of internalized apoptotic
thymocytes was quantified (right) (n = 6). Error bars = S.D. p < 0.01.
7 / 15
normal vision, while MerTK plays a critical role in phagocytosis by RPE cells. Western blotting
showed that the protein levels of RPE65 and MerTK were significantly reduced by TGF-β and
that this was inhibited by AdCCN5 (Fig 2B). We used two pH-sensitive fluorophore-labeled
materials, pHrodo Red BioParticles and tetramethyl-6-carboxyrhodamine (TAMRA)-labeled
apoptotic thymocytes, to analyze the phagocytic function of RPE cells. These fluorophores are
activated when the materials are engulfed by acidic phagosomes. RPE cells were incubated
with these materials and analyzed by flow cytometry. TGF-β markedly reduced the phagocytic
activity of RPE cells, and this was significantly attenuated by AdCCN5 (Fig 2C and 2D). These
data indicate that CCN5 prevents TGF-β-induced functional deterioration of ARPE-19 cells.
CCN5 reverses TGF-β-induced fibrotic deformation of ARPE-19 cells
We next examined whether CCN5 could reverse pre-established TGF-β-induced fibrotic
deformation of ARPE-19 cells. To this end, confluent ARPE-19 cells were treated with TGF-β for 2
days and then infected with AdCCN5 or AdLacZ (Fig 3A). Microscopic analysis showed that
AdCCN5 markedly reversed TGF-β-induced morphological transformation of ARPE-19 cells
(Fig 3B). Western blotting demonstrated that AdCCN5 significantly normalized the
TGF-βinduced increases in levels of mesenchymal marker proteins (α-SMA, vimentin, and
fibronectin) and decreases in levels of epithelial marker proteins (ZO-1 and occludin) (Fig 3C).
Immunostaining with an anti-ZO-1 antibody revealed that tight junctions disrupted by TGF-β were
markedly restored by AdCCN5. In addition, AdCCN5 normalized the α-SMA level, which was
increased by TGF-β. Phalloidin staining revealed that AdCCN5 completely eliminated the
TGFβ-induced formation of f-actin (Fig 3D). The collagen gel contraction assay also demonstrated
that TGF-β increased contractility and that this was restored by AdCCN5 (Fig 3E). These results
show that CCN5 can reverse TGF-β-induced fibrotic deformation of ARPE-19 cells.
CCN5 reverses TGF-β-induced functional deterioration of ARPE-19 cells
ARPE-19 cells treated as described in the previous section were subjected to functional
analyses. TGF-β-induced migration of ARPE-19 cells was reduced by AdCCN5 to a level
comparable to that in the control group (Fig 4A). AdCCN5 significantly normalized the
TGF-βinduced decreases in the protein levels of RPE65 and MerTK (Fig 4B). Flow cytometric
analyses revealed that TGF-β-induced suppression of phagocytic activity was significantly recovered
by AdCCN5 (Fig 4C and 4D). These results indicate that CCN5 can reverse TGF-β-induced
functional defects of ARPE-19 cells.
CCN5 reverses phenotypic deformation of RPE in vivo
Aged Ccl2-/- mice exhibit a subset of phenotypes of dry-type AMD, including fibrotic
deformation of RPE [
]. A recombinant adeno-associated virus (AAV) expressing CCN5
] or a control virus (AAV9-VLP) was intravitreally injected into 18-month-old
Ccl2-/- mice. RPE cells were obtained and analyzed at 12 weeks post-injection (Fig 5A). The
CCN5 level was markedly reduced in Ccl2-/- mice but was restored by AAV9-CCN5. Levels of
the mesenchymal marker proteins (α-SMA, vimentin, and fibronectin) were significantly
increased in RPE of Ccl2-/- mice, whereas levels of the tight junction marker proteins (ZO-1
and occludin) and of the functional RPE marker proteins (RPE65 and MerTK) were
significantly decreased, and AAV9-CCN5 normalized these alterations (Fig 5B). To investigate the
integrity of tight junction, RPE/choroid/sclera complex were mounted. Compared to the
organized hexagonal shape of tight junction in the AAV9-VLP delivered 18-month-old wild-type
(WT) littermate mice, age-matched AAV9-VLP delivered Ccl2-/- mice showed irregular and
vague shape of tight junction. In contrast, tight junction integrity was maintained in
Ccl2-/8 / 15
Fig 3. CCN5 reverses TGF-β-induced fibrotic deformation of ARPE-19 cells. (A) Fully confluent ARPE-19 cells were
serumstarved for 24 hours, treated with TGF-β, and then infected with AdLacZ or AdCCN5. (B) Bright field microscopy images showing the
change in the morphology of ARPE-19 cells from an organized epithelial cobblestone-like monolayer to fibroblast-like cells upon
treatment with TGF-β in the absence of AdCCN5. Original magnification: 100×. Scale bar: 200 μm. (C) Total cell lysates prepared
under the same conditions as shown in A were immunoblotted with antibodies against CCN5, α-SMA, vimentin, fibronectin, ZO-1,
and occludin. GAPDH was used as a loading control. Quantified protein levels were normalized against the loading control (left
bottom). Error bars = S.D. p < 0.05, p < 0.01. (D) Immunofluorescence images of ARPE-19 cells stained for ZO-1 and α-SMA.
Nuclei were counterstained with Hoechst. Images are representative of at least three independent experiments. Original
magnification: 200×. Scale bar: 50 μm. (E) Collagen gel lattices were generated and cultured under the same conditions as shown in A.
Representative images of collagen gel lattices are shown (left), and areas of gel contraction were quantified (right) (n = 3). Error
bars = S.D. p < 0.05, p < 0.01.
mice under the intravitreal injection of AAV9-CCN5 as assessed by immunocytochemical
staining of ZO-1 (Fig 5C). The RPE65 level was markedly reduced and the α-SMA level was
prominently elevated in Ccl2-/- mice, which were significantly normalized by the injection of
AAV9-CCN5 (Fig 5C). These data suggest that RPE undergoes fibrotic deformation in aged
Ccl2-/- mice and that this can be reversed by restoration of the CCN5 level.
9 / 15
Fig 4. CCN5 reverses TGF-β-induced functional deterioration of ARPE-19 cells. (A) ARPE-19 cells were treated with TGF-β
and then infected with AdLacZ or AdCCN5. After 48 hours, cells were scratched, fixed, and stained with Hoechst. Representative
images are shown (left), and migration distances were quantified (right) (n = 3). Error bars = S.D. p < 0.05. (B) Expression of the
RPE function-related proteins RPE65 and MerTK was determined by western blotting under the same conditions as in A. Error
bars = S.D. p < 0.05. (C) Phagocytic function of ARPE-19 cells was analyzed using pHrodo BioParticles conjugates under the
same conditions as shown in A. TGF-β-induced phagocytic dysfunction was recovered in the presence of AdCCN5. Phagocytosis
of ARPE-19 cells was quantified by flow cytometry. The relative fluorescence intensity of internalized pHrodo BioParticles was
quantified (right) (n = 9). Error bars = S.D. p < 0.01. (D) TAMRA-labeled apoptotic thymocytes were added to ARPE-19 cells
under the same conditions as shown in A and analyzed by flow cytometry. Relative fluorescence intensity of internalized
apoptotic thymocytes was quantified (right) (n = 6). Error bars = S.D. p < 0.01.
Phenotypic transformation of RPE has been implicated in various ocular diseases. This change
is often described as epithelial-mesenchymal transition (EMT), which leads to loss of epithelial
characteristics, including down-regulation of tight junction proteins, and gain of mesenchymal
phenotypes, including up-regulation of α-SMA, vimentin, and fibronectin. However, more
comprehensive cellular and molecular analyses are required to clarify the TGF-β-mediated
phenotypic transformation of RPE. We thus called this phenomenon “fibrotic deformation”
throughout this study. TGF-β can induce fibrotic deformation of RPE in vitro [
10 / 15
WT or Ccl2-/- mice CCN5 ZO-1
Ccl2-/Fig 5. CCN5 is protective in a mouse model of AMD. (A) The effects of intravitreally injected AAV9-CCN5 in
18-month-old Ccl2-/- mice were evaluated at 12 weeks post-injection. Viral particles corresponding to 2 × 109 viral
genomes in 2 μl were used per eye. (B) Western blot analysis of RPE/choroid/sclera complex lysates isolated from
18-month-old Ccl2-/- mice and age-matched WT littermates using antibodies against CCN5, α-SMA, vimentin,
fibronectin, ZO-1, occludin, RPE65, and MerTK. GAPDH was used as a loading control. Quantified protein levels
were normalized against the loading control (left bottom) (n = 4, WT AAV9-VLP; n = 6, Ccl2-/- AAV9-VLP; and n = 6,
Ccl2-/- AAV9-CCN5). Error bars = S.D. p < 0.05, p < 0.01. (C) RPE flat mounts were co-immunostained against
ZO-1 (green) and CCN5 (red) or RPE65 (red) or α-SMA (red). Using DAPI for nucleus staining (blue). The
representative images are at least three independent experiments. Scale bar: 150 μm. Original magnifications: 200×.
effect is of physiological relevance. For example, TGF-β reactivity has been detected in vitreous
samples, subretinal fluids, and epiretinal membranes surgically removed from patients with
retinal detachment [
In this study, we report that CCN5 is expressed in RPE, as detected by western blotting (Fig
1 and S1 Fig) and qRT-PCR (S3 Fig), although the role of CCN5 in normal RPE is unknown.
Intriguingly, the CCN5 level was markedly reduced in TGF-β-treated RPE cells, and
restoration of the CCN5 level prevented TGF-β-induced deformation of RPE cells (Figs 1 and 2, S1–
S3 Figs). As a control, an antagonist of TGF-β receptor, A83-01, was also treated along with
TGF-β to RPE cells. Western blotting showed that the effects of CCN5 and A83-01 in
preventing the TGF-β-induced malformation of RPE cells were comparable (S4 Fig). The
TGF-β11 / 15
SMAD signaling pathway was significantly inhibited by CCN5 as assessed by western blotting
(S5 Fig). These observations are consistent with our previous findings that CCN5 inhibits the
TGF-β-SMAD signaling pathway in the heart [
] by up-regulating inhibitory SMAD7
(unpublished observations). It was also shown that CCN5 is strongly expressed in non-invasive
breast cancer cells in which it down-regulates the expression of TGF-β receptor II (TGF-βRII)
at the transcriptional level. Loss of CCN5 promoted EMT of the breast cancer cells through
up-regulation of the TGF-βRII level . Therefore, the inhibitory role of CCN5 in the
TGFβ-SMAD signaling appears to be conserved in diverse cells and tissues.
CCN5 also antagonizes the pro-fibrotic activity of CCN2 (also known as connective tissue
growth factor), although the underlying mechanism is unknown [
]. CCN2 was significantly
up-regulated in RPE cells treated with TGF-β, and this change was prevented by CCN5 (S4
and S5 Figs). Treatment with a combination of EGTA, epidermal growth factor, and fibroblast
growth factor was suggested to be a more physiologically relevant means of inducing fibrotic
deformation of RPE [
]. CCN5 prevented RPE deformation induced by this combination of
reagents (data not shown). Collectively, these data suggest that CCN5 is required to maintain
the epithelial phenotypes of RPE and that restoration of the CCN5 level prevents fibrotic
deformation of RPE induced by diverse pathological stimuli.
More clinically important is the finding that CCN5 reversed pre-established fibrotic
deformation of RPE, as shown by the reappearance of tight junction proteins and the disappearance
of mesenchymal marker proteins (Fig 3). This phenomenon is reminiscent of
mesenchymalepithelial transition (MET). Notably, CCN5 enhances MET of pancreatic cancer cells [
loss of CCN5 promotes EMT and acquisition of stem cell-like phenotypes in
estrogen-dependent MCF7 breast cancer cells [
]. Therefore, the role of CCN5 in MET-like phenotypic
transformation is not confined to RPE. In this regard, it is of note that the ARPE-19 cell line
spontaneously acquired immortality [
] and thus is not truly representative of native RPE
cells. Further analyses of primary cultures of RPE cells are needed to confirm these findings
and to elucidate the underlying signaling pathways.
We extended our findings from in vitro experiments to an in vivo model of fibrotic ocular
disease. Ccl2-/- mice have morphological, ultrastructural, and functional features characteristic
of AMD. Substantial Bruch’s membrane thickening and drusen deposition are observed in aged
Ccl2-/- mice. Lipofuscin granules, which are thought to promote RPE dysfunction in dry-type
AMD, accumulate in RPE of 15-month-old Ccl2-/- mice. In addition, attenuation of RPE and
chorio-capillaries are observed in 16-month-old Ccl2-/- mice [
]. We focused on the
pathogenic alteration of RPE cells in Ccl2-/- mice. The CCN5 level was significantly reduced
concomitant with apparent EMT-like fibrotic deformation of RPE in 18-month-old Ccl2-/- mice, and
this was rescued by restoration of the CCN5 level, as assessed by western blotting and
immunohistochemistry (Fig 5). Although extensive functional and anatomical analyses were not
performed, our data show that CCN5 induced phenotypic reversion of deformed RPE cells in vivo.
Taken together, our data suggest that CCN5 is a novel modality to prevent and reverse
fibrotic deformation of RPE cells.
S1 Fig. Fibrotic conformation of ARPE-19 cells was induced by TGF-β. ARPE-19 cells were
cultured for 48 hours with 5 ng/ml or 10 ng/ml of TGF-β. (A) Compared with control, TGF-β
treatment cells were transformed to mesenchymal morphology by bright field microscopy.
The representative images are at least three independent experiments. Original magnifications:
100×. Scale bar: 200 μm. (B) CCN5 and tight junction markers are down-regulated, whereas
mesenchymal markers are up-regulated during TGF-β-induced fibroblast-like phenotype in
12 / 15
ARPE-19 cells. Error bars = S.D. p<0.05, p<0.01.
S2 Fig. TGF-β induces fibrotic changes in ARPE-19 cells. Cells were visualized by staining
with anti-α-SMA and phalloidin. The representative images are at least three independent
experiments. Original magnification: 200×. Scale bar: 50 μm.
S3 Fig. CCN5 maintains mRNA levels of tight junction markers and attenuates
mesenchymal markers from TGF-β-induced fibrotic changes of ARPE-19 cells. Quantitative RT-PCR
analyses of CCN5, ZO-1, occludin, RPE65, α-SMA, vimentin, and fibronectin (A, B, and C)
were performed (n = 4). Error bars = S.D. p<0.05, p<0.01.
S4 Fig. An antagonist of TGF-β receptor, A83-01, prevents TGF-β-induced fibroblast-like
RPE disruption. ARPE-19 cells were pretreated with 1 μM of A83-01 for 1 hour and cultured
with TGF-β for 48 hours. Cell lysates were subjected to Western blotting. Error bars = S.D.
S5 Fig. CCN5 inhibits TGF-β-SMAD signaling pathway. ARPE-19 cells were cultured in the
presence of AdLacZ or AdCCN5 prior to TGF-β treatment. Cell lysates were subjected to
Western blotting. Error bars = S.D. p<0.05, p<0.01.
S1 Table. List of primary antibodies.
S2 Table. Sequences of primers used for qRT-PCR.
We thank Dr. Sung Wook Park for help with the animal experiments.
Conceptualization: Dong Hyun Jo, Jeong Hun Kim, Woo Jin Park.
Data curation: Aeri Yoon.
Funding acquisition: Jeong Hun Kim, Woo Jin Park.
Methodology: Sora Im, Juyeon Lee, Daeho Park.
Supervision: Jeong Hun Kim.
Validation: Jin Hyoung Kim.
Writing – original draft: Aeri Yoon, Woo Jin Park.
Writing – review & editing: Dong Hyun Jo.
13 / 15
14 / 15
1. Burke JM . Epithelial phenotype and the RPE: is the answer blowing in the Wnt? Prog Retin Eye Res . 2008 ; 27 ( 6 ): 579 - 95 . Epub 2008/09/09. https://doi.org/10.1016/j.preteyeres. 2008 . 08 .002 PMID: 18775790; PubMed Central PMCID : PMCPMC2584165 .
2. Bonilha VL . Retinal pigment epithelium (RPE) cytoskeleton in vivo and in vitro . Experimental eye research . 2014 ; 126 : 38 - 45 . https://doi.org/10.1016/j.exer. 2013 . 09 .015 PMID: 24090540
3. Saika S. TGFβ pathobiology in the eye . Laboratory investigation . 2006 ; 86 ( 2 ): 106 . https://doi.org/10. 1038/labinvest.3700375 PMID: 16341020
4. Kalluri R , Weinberg RA . The basics of epithelial-mesenchymal transition . The Journal of clinical investigation . 2009 ; 119 ( 6 ): 1420 - 8 . https://doi.org/10.1172/JCI39104 PMID: 19487818
5. Grisanti S , Guidry C . Transdifferentiation of retinal pigment epithelial cells from epithelial to mesenchymal phenotype . Investigative ophthalmology & visual science . 1995 ; 36 ( 2 ): 391 - 405 .
6. Thiery JP . Epithelial-mesenchymal transitions in development and pathologies . Current opinion in cell biology . 2003 ; 15 ( 6 ): 740 - 6 . PMID: 14644200
7. Baudouin C , Hofman P , Brignole F , Bayle J , Loubiere R , Gastaud P. Immunocytology of cellular components in vitreous and subretinal fluid from patients with proliferative vitreoretinopathy . Ophthalmologica . 1991 ; 203 ( 1 ): 38 - 46 . Epub 1991/01/01. https://doi.org/10.1159/000310223 PMID: 1766640 .
8. Radeke MJ , Radeke CM , Shih Y-H , Hu J , Bok D , Johnson LV , et al. Restoration of mesenchymal retinal pigmented epithelial cells by TGFβ pathway inhibitors: implications for age-related macular degeneration . Genome Med . 2015 ; 7 ( 1 ): 58 . https://doi.org/10.1186/s13073-015-0183 -x PMID : 26150894
9. Khankan R , Oliver N , He S , Ryan SJ , Hinton DR . Regulation of Fibronectin-EDA through CTGF Domain-Specific Interactions with TGFβ2 and Its Receptor TGFβRII . Investigative ophthalmology & visual science . 2011 ; 52 ( 8 ): 5068 - 78 .
10. Connor T , Roberts AB , Sporn M , Danielpour D , Dart LL , Michels RG , et al. Correlation of fibrosis and transforming growth factor-beta type 2 levels in the eye . The Journal of clinical investigation . 1989 ; 83 ( 5 ): 1661 - 6 . https://doi.org/10.1172/JCI114065 PMID: 2708527
11. Kita T , Hata Y , Kano K , Miura M , Nakao S , Noda Y , et al. Transforming growth factor-β2 and connective tissue growth factor in proliferative vitreoretinal diseases: possible involvement of hyalocytes and therapeutic potential of Rho kinase inhibitor . Diabetes . 2007 ; 56 ( 1 ): 231 - 8 . https://doi.org/10.2337/db06- 0581 PMID: 17192487
12. Amin R , Puklin JE , Frank RN . Growth factor localization in choroidal neovascular membranes of agerelated macular degeneration . Investigative ophthalmology & visual science . 1994 ; 35 ( 8 ): 3178 - 88 .
13. Hirase K , Ikeda T , Sotozono C , Nishida K , Sawa H , Kinoshita S . Transforming growth factor β2 in the vitreous in proliferative diabetic retinopathy . Archives of ophthalmology . 1998 ; 116 ( 6 ): 738 - 41 . PMID: 9639441
14. Holbourn KP , Acharya KR , Perbal B. The CCN family of proteins: structure-function relationships . Trends Biochem Sci . 2008 ; 33 ( 10 ): 461 - 73 . Epub 2008/09/16. https://doi.org/10.1016/j.tibs. 2008 . 07 . 006 PMID: 18789696; PubMed Central PMCID : PMCPMC2683937 .
15. Russo JW , Castellot JJ . CCN5: biology and pathophysiology . J Cell Commun Signal . 2010 ; 4 ( 3 ): 119 - 30 . Epub 2010/11/11. https://doi.org/10.1007/s12079-010-0098-7 PMID: 21063502; PubMed Central PMCID : PMCPMC2948116 .
16. Inadera H , Shimomura A , Tachibana S . Effect of Wnt-1 inducible signaling pathway protein-2 (WISP-2/ CCN5), a downstream protein of Wnt signaling, on adipocyte differentiation . Biochemical and biophysical research communications . 2009 ; 379 ( 4 ): 969 - 74 . https://doi.org/10.1016/j.bbrc. 2008 . 12 .185 PMID: 19135425
17. Fritah A , Saucier C , De Wever O , Bracke M , Bièche I , Lidereau R , et al. Role of WISP-2/CCN5 in the maintenance of a differentiated and noninvasive phenotype in human breast cancer cells . Molecular and cellular biology . 2008 ; 28 ( 3 ): 1114 - 23 . https://doi.org/10.1128/ MCB . 01335 -07 PMID: 18070926
18. Lake AC , Castellot JJ . CCN5 modulates the antiproliferative effect of heparin and regulates cell motility in vascular smooth muscle cells . Cell Communication and Signaling . 2003 ; 1(1):5 . https://doi.org/10. 1186/ 1478 -811X-1-5 PMID: 14636425
19. Yoon PO , Lee MA , Cha H , Jeong MH , Kim J , Jang SP , et al. The opposing effects of CCN2 and CCN5 on the development of cardiac hypertrophy and fibrosis . J Mol Cell Cardiol . 2010 ; 49 ( 2 ): 294 - 303 . Epub 2010/05/01. https://doi.org/10.1016/j.yjmcc. 2010 . 04 .010 PMID: 20430035 .
20. Jeong D , Lee MA , Li Y , Yang DK , Kho C , Oh JG , et al. Matricellular Protein CCN5 Reverses Established Cardiac Fibrosis . J Am Coll Cardiol . 2016 ; 67 ( 13 ): 1556 - 68 . Epub 2016/05/07. https://doi.org/10.1016/j. jacc. 2016 . 01 .030 PMID: 27150688 .
21. Kuziel WA , Morgan SJ , Dawson TC , Griffin S , Smithies O , Ley K , et al. Severe reduction in leukocyte adhesion and monocyte extravasation in mice deficient in CC chemokine receptor 2 . Proc Natl Acad Sci U S A . 1997 ; 94 ( 22 ): 12053 - 8 . Epub 1997/10/29. PMID: 9342361; PubMed Central PMCID : PMCPMC23699 .
22. Lu B , Rutledge BJ , Gu L , Fiorillo J , Lukacs NW , Kunkel SL , et al. Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1-deficient mice . J Exp Med . 1998 ; 187 ( 4 ): 601 - 8 . Epub 1998/03/28. PMID: 9463410; PubMed Central PMCID : PMCPMC2212142 .
23. Liu Y , Ko J-A , Yanai R , Kimura K , Chikama T-i , Sagara T, et al. Induction by latanoprost of collagen gel contraction mediated by human tenon fibroblasts: role of intracellular signaling molecules . Investigative ophthalmology & visual science . 2008 ; 49 ( 4 ): 1429 - 36 .
24. Ambati J , Anand A , Fernandez S , Sakurai E , Lynn BC , Kuziel WA , et al. An animal model of age-related macular degeneration in senescent Ccl-2- or Ccr-2-deficient mice . Nat Med . 2003 ; 9 ( 11 ): 1390 - 7 . Epub 2003/10/21. https://doi.org/10.1038/nm950 PMID: 14566334 .
25. Lee H , O'Meara SJ , O'Brien C , Kane R . The role of gremlin, a BMP antagonist, and epithelial-to-mesenchymal transition in proliferative vitreoretinopathy . Invest Ophthalmol Vis Sci . 2007 ; 48 ( 9 ): 4291 - 9 . Epub 2007/08/29. https://doi.org/10.1167/iovs.07-0086 PMID: 17724219 .
26. Lee J , Ko M , Joo CK . Rho plays a key role in TGF-beta1-induced cytoskeletal rearrangement in human retinal pigment epithelium . J Cell Physiol . 2008 ; 216 ( 2 ): 520 - 6 . Epub 2008/03/05. https://doi.org/10. 1002/jcp.21424 PMID: 18314880 .
27. Li H , Wang H , Wang F , Gu Q , Xu X . Snail involves in the transforming growth factor beta1-mediated epithelial-mesenchymal transition of retinal pigment epithelial cells . PLoS One . 2011 ; 6 ( 8 ): e23322 . Epub 2011 /08/20. https://doi.org/10.1371/journal.pone.0023322 PMID: 21853110; PubMed Central PMCID : PMCPMC3154444 .
28. Baudouin C , Fredj-Reygrobellet D , Brignole F , Negre F , Lapalus P , Gastaud P . Growth factors in vitreous and subretinal fluid cells from patients with proliferative vitreoretinopathy . Ophthalmic Res . 1993 ; 25 ( 1 ): 52 - 9 . Epub 1993/01/01. https://doi.org/10.1159/000267221 PMID: 8446368 .
29. Sabbah M , Prunier C , Ferrand N , Megalophonos V , Lambein K , De Wever O , et al. CCN5, a novel transcriptional repressor of the transforming growth factor beta signaling pathway . Mol Cell Biol . 2011 ; 31 ( 7 ): 1459 - 69 . Epub 2011/01/26. https://doi.org/10.1128/ MCB . 01316 -10 PMID: 21262769; PubMed Central PMCID : PMCPMC3135290 .
30. Chen H-C , Zhu Y-T , Chen S-Y , Tseng SC . Selective activation of p120ctn-Kaiso signaling to unlock contact inhibition of ARPE-19 cells without epithelial-mesenchymal transition . PLoS One . 2012 ; 7 ( 5 ): e36864. https://doi.org/10.1371/journal.pone. 0036864 PMID: 22590627
31. Dhar G , Mehta S , Banerjee S , Gardner A , McCarty BM , Mathur SC , et al. Loss of WISP-2/CCN5 signaling in human pancreatic cancer: a potential mechanism for epithelial-mesenchymal-transition . Cancer letters . 2007 ; 254 ( 1 ): 63 - 70 . https://doi.org/10.1016/j.canlet. 2007 . 02 .012 PMID: 17383817
32. Ferrand N , Gnanapragasam A , Dorothee G , Redeuilh G , Larsen AK , Sabbah M. Loss of WISP2/CCN5 in estrogen-dependent MCF7 human breast cancer cells promotes a stem-like cell phenotype . PloS one . 2014 ; 9 ( 2 ):e87878. https://doi.org/10.1371/journal.pone. 0087878 PMID: 24498388
33. Dunn KC , Aotaki-Keen AE , Putkey FR , Hjelmeland LM . ARPE-19, a human retinal pigment epithelial cell line with differentiated properties . Exp Eye Res . 1996 ; 62 ( 2 ): 155 - 69 . Epub 1996/02/01. https://doi. org/10.1006/exer. 1996 .0020 PMID: 8698076 .