FGF-2– and TGF-?1–Induced Downregulation of
Lumican and Keratocan in Activated Corneal
Keratocytes by JNK Signaling Pathway
Jian Chen,1Julie Wong-Chong,1and Nirmala SundarRaj2
PURPOSE. Downregulation of lumican and keratocan expres-
sion is an undesirable phenotypic change that occurs during
corneal wound healing. The present study was intended to
determine whether the activation of Jun N-terminal kinase
(JNK)-signaling pathway is involved in their downregulation
in TGF-?1– and FGF-2–activated keratocytes.
METHODS. Keratocytes, isolated from rabbit corneal stroma, and
cultured in a serum-free medium, pretreated or not treated
with JNK inhibitor (SP600125), were activated with FGF-2/
heparin sulfate (HS) or TGF-?1 in the presence or absence of
SP600125. In another set of experiments, keratocytes were
transfected with JNK1/2 Dicer-substrate RNA (DsiRNA) and
then activated with TGF-?1 or FGF-2/HS. Specific phenotypic
changes were analyzed immunocytochemically and correlated
with Western blot analyses. The relative levels of specific
mRNAs were estimated by quantitative RT-PCR using specific
RESULTS. The FGF-2/HS– or TGF-?–induced activation of cor-
neal stromal keratocytes to fibroblast- or myofibroblast-pheno-
type, respectively, resulted in marked decreases in cell surface-
associated and secreted keratan sulfate proteoglycans (KSPGs).
Both keratocan and lumican proteins and their mRNAs were
downregulated in the activated keratocytes. However, JNK
inhibition during the activation of keratocytes, pretreated with
the JNK inhibitor, suppressed the reduction in the cell–surface
associated and secreted KSPGs (lumican and keratocan), and
their mRNA transcripts. Downregulation of total KSPGs and
their mRNAs was also inhibited by decreasing JNK1 and JNK2
levels via JNK1/2 DsiRNA transfection of keratocytes before
CONCLUSIONS. Extrapolating from the present study, FGF-2– and
TGF-?1–activation of JNK signaling pathway may be partly
responsible for the downregulation of keratocan and lumican
expression in activated corneal keratocytes observed during
corneal stromal wound healing. (Invest Ophthalmol Vis Sci.
endothelium. A well-organized extracellular matrix (ECM) con-
taining densely packed and regularly spaced thin collagen
fibrils of uniform diameter, is largely responsible for transpar-
ency in the corneal stroma.1–3Keratan sulfate proteoglycans
(KSPGs) in the stromal ECM play a critical role in the develop-
ment and maintenance of corneal transparency. Keratocan and
lumican, the major KSPGs in the corneal stroma, regulate both
fibril diameter and interfibrillar spacing as evident from the
phenotype of lumican and keratocan knockout mice.4–9Ker-
atocan knockout mice have a thinner corneal stroma with
irregular collagen fibril organization compared with the normal
mice,8and lumican knockout mice have increased collagen
fibril diameter and develop opaque corneas.4,5
Corneal stromal cells (keratocytes), which synthesize ker-
atocan and lumican during development, become quiescent in
a fully-developed cornea. However, after an injury to the cor-
nea, growth factors and cytokines originating from corneal
epithelial cells, inflammatory cells, and tear fluid activate the
keratocytes to fibroblast or myofibroblast phenotypes (re-
views, see Refs. 10–14). Keratocan and lumican synthesis is
downregulated in the activated keratocytes during wound heal-
ing.15–18KSPG expression is also downregulated in vitro when
cultured keratocytes are activated with growth factors includ-
ing FGF-2 and TGF-?1.19–23Therefore, an in vitro model of
keratocyte activation is useful to study the signaling mecha-
nisms that downregulate the expression of KSPGs. We had
previously demonstrated that activation of the small GTPase
Rho and its downstream target Rho kinase (ROCK) regulate
several undesirable phenotypic changes including the down-
regulation of KSPGs in the activated keratocytes.23Jun N-ter-
minal kinase (JNK), a member of the mitogen activated protein
kinase (MAPK) family, has been shown to mediate some of the
Rho/ROCK regulated events.24–26The present study was de-
signed to investigate whether JNK activation is responsible for
the observed TGF-?1– and FGF-2–induced downregulation of
KSPGs in activated keratocytes.
he cornea is a transparent avascular refractive structure
consisting of three tissue layers, epithelium, stroma, and
All procedures involving rabbits were performed in compliance with
the ARVO Statement for the Use of Animals in Ophthalmology and
Vision Research. Keratocytes were isolated from corneas excised from
whole rabbit eyes obtained from Pel-Freez Biological (Rogers, AR) and
cultured using a modified23procedure of Jester et al.27Isolated ker-
atocytes were suspended in DMEM/F-12 containing 0.021% L-glutamine
medium (glutaMAX; Invitrogen/Gibco, Carlsbad, CA) 0.011% pyruvate,
100 units /mL penicillin, and 100 ?g/mL streptomycin (Invitrogen/
Gibco) and the suspension was filtered through a cell strainer (70 ?m
BD, Falcon, Bedford, MA). Keratocytes were then centrifuged, resus-
pended, and plated at a density of 1.5 ? 104cells/cm2into 60 mm
Ophthalmology and Visual Science Research Center, Eye and Ear Insti-
tute, and the2Department of Cell Biology and Physiology, University of
Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
Supported by NIH Grant EY03263 (NS) and Core Grant EY008098;
Research to Prevent Blindness; and The Eye and Ear Foundation (Pitts-
Submitted for publication June 17, 2011; revised August 22 and
September 20, 2011; accepted October 14, 2011.
Disclosure: J. Chen, None; J. Wong-Chong, None; N. Sundar-
Corresponding author: Nirmala SundarRaj, Eye and Ear Institute,
Ophthalmology Department, Room 1029, 203 Lothrop Street, Pitts-
burgh, PA 15213; firstname.lastname@example.org.
1Department of Ophthalmology, UPMC Eye Center,
Investigative Ophthalmology & Visual Science, November 2011, Vol. 52, No. 12
Copyright 2011 The Association for Research in Vision and Ophthalmology, Inc.
dishes (Falcon Primaria; Becton Dickinson, Lincoln Park, NY), in the
serum-free DMEM/F12 with 0.1 mM L-ascorbic acid 2-phosphate (SFM).
After incubating the dishes at 37°C in a humidified 5% CO2/95% air
incubator for 24 hours, the media were replaced with fresh SFM and
the cells were incubated for an additional 48 hours. The cells were
then incubated with SFM or without (controls) JNK inhibitor (20 ?M,
SP600125) for 3 hours. The cells were then activated by adding FGF-2
(40 ng/mL) and heparin sulfate (HS) (5 ?g/mL) or TGF-?1 (2 ng/mL) to
the media in the culture dishes. After 60 hours, the culture superna-
tants were removed and stored at 4°C after the addition of protease
inhibitors (PMSF, 1 mM; NEM, 2mM, and pepstatin, 1 ?g/mL) for the
analyses of secreted KSPGs. Total RNAs and proteins were extracted
from the cells in the respective buffers.
For immunocytochemical analyses, the cells were fixed with 4% para-
formaldehyde and permeabilized with 0.25% Triton X-100 in PBS.28
After blocking with 10% heat-inactivated goat serum in PBS for 1 hour,
the cells were treated with primary and secondary antibodies as de-
scribed previously.28Primary antibodies were ascites monoclonal
mouse anti-KS (J19) at 1:100 dilution or rabbit anti–p-c-Jun (ser 63)
(Santa Cruz Biotechnology, Santa Cruz, CA) at 1:50 dilution. The
secondary antibodies were goat anti-mouse IgG or anti-rabbit IgG
conjugated to either Alexa Fluor 488 or Alexa Fluor 647 (Molecular
Probes Inc./Invitrogen, San Diego, CA) at 1:2500 and 1:1500 dilutions,
respectively. To stain actin filaments, Alexa 546 phalloidin (Molecular
Probes) was included at 1:50 dilutions with the secondary antibody.
Coverslips were mounted on the top of the cells (Immu-mount; Shan-
don, Pittsburgh, PA). Fluorescent images were captured with a scan-
ning laser system attached to an inverted microscope (Olympus IX70;
Olympus America Inc., Center Valley, PA) using the same settings for
comparisons of the intensities of staining.
Western Blot Analyses
Cells were lysed in RIPA buffer (9.1 mM dibasic sodium phosphate, 1.7
mM monobasic sodium phosphate, pH 7.4; 150 mM NaCl, 1% Nonidet
P-40; 0.5% sodium deoxycholate; 0.1% SDS; 0.03 TIU/mL aprotinin
[Sigma-Aldrich, St. Louis, MO]; 1 mM PMSF and 1 mM sodium or-
thovanadate) for protein extraction. Because the final cell density
depended on the treatment received, for the comparative Western blot
analyses of secreted KSPGs from equal numbers of cells, the volume of
culture supernatant in each sample was adjusted with PBS to 5 mL/100
mg of proteins in the corresponding cell extracts. Equal volumes of
normalized samples were then treated with1⁄4 the volume of five times
sample buffer for SDS-PAGE. For Western blotting of keratocan and
lumican core proteins, the culture supernatants were concentrated
10-fold and then subjected to keratanase treatment.29Protein bands on
SDS-PAGE were electrophoretically transferred (Immobilon-P mem-
brane; Millipore Corp, Bedford, MA), and the blots were reacted with
anti-KS antibody or rabbit polyclonal anti-lumican or anti-keratocan
antibodies (kind gift from John Hassell, University of South Florida,
Tampa, FL) followed by HRP-conjugated secondary antibody as de-
scribed previously.28The immunoreactive bands were detected (Im-
mobilon Western Chemiluminescent HRP Substrate; Millipore Corp.)
following the manufacturer’s protocols. The bands on x-ray film were
scanned and analyzed using publicly available software (ImageJ, devel-
oped by Wayne Rasband, National Institutes of Health, Bethesda, MD;
available at http://rsb.info.nih.gov/ij/index.html). In some experi-
ments, the protein bands in the gel were blotted on transfer mem-
branes (Millipore Immobilon-FL PVDF). The membranes were then
blocked with blocking buffer (Odyssey; LI-COR, Lincoln, NE); and
secondary antibodies (IRDye 680LT and 800CW, LI-COR) at 1:10,000
dilution were used to detect the primary antibody using an infrared
imager (Odyssey; LI-COR). Densitometric analyses were performed
using the infrared imager software (Odyssey; LI-COR).
Total RNA from the cells was isolated using a kit (RNeasy Mini; Qiagen,
Valencia, CA) and RNA was isolated using quantitative RT-PCR. The
quantification of specific mRNAs was carried out using RT-PCR re-
agents (SYBR Green; Fermentas; Thermo Scientific, Waltham, MA)
according to the manufacturer’s instructions. The reactions were car-
ried out using a sequence detection system (ABI 7700; Applied Biosys-
tems, Foster City, CA) as described previously.23To design the primers
used for quantitative RT-PCR (Table 1), partial sequences of the genes
encoding rabbit keratocan, lumican, JNK1, and JNK2 were first deter-
mined from PCR of thermo-amplified cDNAs encoding these proteins.
The primers used for PCR were designed from the conserved se-
quences of these genes in other animal species.
After plating, the keratocytes were incubated in SFM for 48 hours, and
then the medium was replaced with fresh SFM. The cells were then
transfected with 10 nM Dicer-substrate RNA (DsiRNA) for JNK1 and
JNK2 or with nonspecific scrambled DsiRNA (Integrated DNA Tech-
nologies Inc., Coralville, IA) using lipid reagent (SilentFect; Bio-Rad,
Hercules, CA) according to the manufacturer’s instructions. The se-
quences of the DsiRNAs are shown in Table 2.
Six hours after the transfection, the media were replaced with SFM
or SFM containing TGF-?1 or FGF-2/HS. After 60 more hours of incu-
bation, the cells were analyzed immunocytochemically or by Western
blotting analysis for proteins, and by qRT-PCR for mRNAs.
All data are presented as the mean ? SD. Statistical analyses of the data
from three or more separate experiments were performed with repeat-
ed-measures ANOVA. The differences were considered significant at
P ? 0.05.
As expected30, keratocytes isolated from rabbit corneas, when
cultured in SFM, exhibited dendritic morphology similar to that
exhibited in vivo and also expressed KSPGs which were se-
creted in the culture media as well as associated with the cell
surface (Figs. 1, 2). Activation of keratocytes with FGF-2 (40
ng/mL) and HS (5 ?g/mL) or TGF-?1 (2 ng/mL) to fibroblast or
myofibroblast phenotype, respectively (in the present study
these cells will be referred to as FGF-2/HS– or TGF-?1–acti-
vated keratocytes), resulted in expected changes in the cell
morphology, and assembly of actin stress fibers as seen by
phalloidin staining (Fig. 1). The stress fiber-network was more
robust in TGF-?1–activated keratocytes. By double staining
with anti-KS antibody the cell surface–associated KSPG(s) was
evident in nonactivated keratocytes, but was reduced on acti-
vation with FGF-2/HS or TGF-?1 (Figs. 2A, 2B). However, JNK
TABLE 1. Primers Used in Quantitative RT-PCR
Genes Primer Sequence (5? to 3?)
18S (GenBank X 00640) Forward: CTCAACACGGGAAACCTCAC
Forward: TGCAGCTTACCCACA ACA AG
8958 Chen et al.
IOVS, November 2011, Vol. 52, No. 12
inhibition with SP600125 during FGF-2– and TGF-?1–induced
activation of keratocytes, pretreated with SP600125, resulted
in the inhibition of the stress fiber assembly and changes in the
cell morphology (Fig. 1). JNK inhibition also prevented the loss
in cell-associated KSPG staining in FGF-2/HS– and TGF-?1–
activated keratocytes (Figs. 2A and 2B, respectively). The inhi-
bition of JNK activity by the JNK inhibitor was confirmed by a
resulting decrease in its downstream target, nuclear p-c-Jun, as
seen by immunostaining (Figs. 2A, 2B).
The relative levels of KSPGs in the culture supernatants,
secreted by comparable numbers of cells, were analyzed by
Western blotting. Based on the reduction in the densities of
KSPGs bands (Fig. 3A), TGF-?1 and FGF-2/HS were found to
inhibit KSPG synthesis. However, JNK inhibition suppressed
the TGF-?1– and FGF-2/HS–induced decrease in the secreted
KSPGs, and it also increased the levels of KSPG secreted by
nonactivated keratocytes (Fig. 3A). To determine which of the
KSPGs were regulated by JNK, keratanase- digested culture
supernatants were analyzed for keratocan and lumican levels
by Western blot analysis. Both keratocan and lumican were
decreased in the culture supernatants of TGF-?1– as well as
FGF-2/HS–activated keratocytes, which was evident from the
decreases in the band densities in the Western blots (Fig. 3B).
However, JNK inhibition occluded the loss in secreted lumican
and keratocan. The increases in secreted lumican and kerato-
can on JNK inhibiton in FGF-2– and TGF-?1–activated cells
were greater than those in the nonactivated cells.
The levels of mRNA encoding lumican and keratocan were
also decreased in keratocytes when activated with FGF-2/HS
and TGF-?1 (Fig. 4). Activation of keratocytes with FGF-2/HS
resulted in the reduction in lumican mRNA levels by 49.5% ?
5.2% and keratocan mRNA levels by 79.4% ? 6.5%. However,
the inhibition of JNK activity during the FGF-2/HS–induced
activation of keratocytes, pretreated with JNK inhibitor, re-
sulted in a170% ? 70% increase in the levels of lumican mRNA
and a 470% ? 50% increase in the levels of keratocan mRNA in
the activated keratocytes (P ? 0.05). The increased levels of
expression were slightly higher than those in the nonactivated
keratocytes (Fig. 4). Similarly, in TGF-?1–activated cells, lumi-
can and keratocan mRNA levels were reduced by 77% ? 9%
and 85% ? 14%, respectively. JNK inhibition during TGF-?1–
induced activation of JNK-inhibited keratocytes resulted in a
240% ? 60% and a 511% ? 180% increase in the levels of
lumican and keratocan mRNAs, respectively (P ? 0.05). The
increased levels of lumican and keratocan, resulting from the
inhibition of JNK during the activation, were close to those in
the nonactivated keratocytes. JNK inhibition in the nonacti-
vated keratocytes also resulted in a 100% ? 30% and a 75% ?
30% increase in lumican and keratocan mRNA levels, respec-
tively, indicating the presence of some JNK activity in the
nonactivated keratocytes. SP600125-induced increases in the
levels of keratocan and lumican mRNA levels were not signif-
icantly different in the growth factor-activated cells than those
in the nonactivated keratocytes.
The results of the JNK inhibitor studies were verified
using DsiRNA to knockdown JNK1 and JNK2 in the cells.
FGF-2/HS–induced activation of keratocytes, transfected
with scrambled DsiRNA (controls), resulted in a 140% ?
40% and a 130% ? 20% increase in JNK1 and JNK2 mRNA
levels, respectively (Fig. 5). The levels of lumican and ker-
atocan in these cells were reduced by 65% ? 3% and 45% ?
15%, respectively. When nonactivated keratocytes were
transfected with JNK1/2 DsiRNA, the levels of JNK1 and
JNK2 mRNAs were reduced by 53% ? 6% and 66% ? 7%,
respectively. The levels of lumican and keratocan in these
cells were 27% ? 5% and 44% ? 6% higher, respectively,
than those in corresponding scrambled DsiRNA-transfected
controls. Similarly, when the keratocytes, transfected with
JNK1/2 DsiRNA, were activated with FGF-2/HS, the levels of
JNK1 and JNK2 mRNAs were reduced by 75% ? 13% and
57% ? 8%, respectively. The levels of lumican and keratocan
in these cells were 70% ? 20% and 100% ? 10% higher than
in the corresponding scrambled DsiRNA controls (P ? 0.05).
Similarly, TGF-?1–induced activation of these keratocytes,
TABLE 2. DsiRNA Sequences
Gene DsiRNA Sequence (5? to 3?)
FGF-2/HS– and TGF-?1–induced as-
sembly and organization of actin fila-
ments in activated stromal kerato-
cytes. Corneal stromal keratocytes,
isolated from rabbit corneas, were
cultured in DMEM/F12 without FBS
(SFM). After pretreating with or with-
out SP600125 for 3 hours in the SFM,
the keratocytes were activated by
supplementing the medium in the
culture dishes with 40 ng/mL FGF-2
and 5?g/mL HS or 2 ng/mL TGF-?1.
After 60 hours of incubation, the
cells were reacted with Alexa Fluor
546-phalloidin. Bar, 50 ?m.
JNK inhibition affects
IOVS, November 2011, Vol. 52, No. 12
Regulation of Keratan Sulfate Proteoglycan Expression 8959
transfected with scrambled DsiRNA, resulted in a 70% ?
20% and a 50% ? 16% increase in JNK-1 and JNK-2 mRNA
levels, respectively, compared with those in nonactivated
keratocytes. The levels of lumican and keratocan in TGF-?1–
activated keratocytes were reduced by 68% ? 4% and 85% ?
5%, respectively. However, activation of JNK1/2 DsiRNA-
transfected keratocytes with TGF-?1 resulted in a 55% ? 7%
and a 69% ? 3% reduction in JNK1 and JNK2 mRNA levels,
creases in cell surface–associated KS
in the activated corneal stromal ker-
atocytes. Corneal stromal kerato-
cytes, isolated from rabbit corneas,
were cultured in DMEM/F12 without
FBS (SFM). After pretreating with or
without SP600125 for 3 hours in the
SFM, the keratocytes were activated
by supplementing the medium in the
culture dishes with: (A) 40 ng/mL
of TGF-?1. After 60 hours of incuba-
tion, cells were analyzed by double
fluorescence staining using mouse
anti-KS antibody or anti-p-c-Jun anti-
body followed by Alexa Fluor 488 anti-
mouse IgG antibodies and Alexa Fluor
546 phalloidin. Bar, 50 ?m.
JNK inhibition occludes
8960Chen et al.
IOVS, November 2011, Vol. 52, No. 12
respectively, compared with those in the controls trans-
fected with scrambled DsiRNA. The corresponding levels of
lumican and keratocan in these cells were 100% ? 15% and
110% ? 30% higher, respectively, than those in the scram-
bled DsiRNA transfected controls. The above changes in the
levels of JNK1, JNK2, lumican, and keratocan mRNAs, re-
sulting from JNK1/2 DsiRNA transfection were statistically
significant (P ? 0.05). The increases in the levels of kerato-
can and lumican, resulting from JNK1/2 DsiRNA transfec-
tion, were not significantly different in nonactivated and
Western blot analyses of the cell extracts demonstrated that
JNK1 and JNK2 were indeed present in nonactivated kerato-
cytes, cultured in SFM (Fig. 6A; a representative Western blot).
However, on activation with FGF-2/HS or TGF-?1 the densities
of JNK1 and JNK2 bands increased by 380% and 350%, and
190% and 160%, respectively. As evident from the densities of
the bands, JNK1 and JNK2 levels in JNK1/2 DsiRNA transfected
nonactivated, FGF-2 activated, and TGF-?1 activated kerato-
cytes, were less than those in the scrambled DsiRNA controls.
In JNK1/2 DsiRNA transfected nonactivated keratocytes the
band densities of JNK1 and JNK2 were 46.5% and 43.4% less,
in FGF-2/HS-activated keratocytes they were 49.3% and 45.9%
less, and in TGF-?1–activated keratocytes they 52% and 78%
less, respectively, than those in the corresponding scrambled
DsiRNA-transfected controls (Fig. 6A). The levels of secreted
keratocytes and occludes TGF-?1– and FGF-2–induced decreases in
KSPG secretion (A). The KSPGs, downregulated in the JNK-inhibited
activated keratocytes, are characterized to be lumican and kertocan
(B). Rabbit corneal keratocytes, cultured in DMEM/F12 without FBS
(SFM), were incubated with or without SP600125 for 3 hours. The cells
were then activated by supplementing the medium in the culture
dishes with 40 ng/mL FGF-2 and 5?g/mL HS or 2 ng/mL TGF-?1 or not
activated. After 60 hours, culture supernatants were analyzed by West-
ern blotting using mouse anti-KS antibody (A). Aliquots of the super-
natants were concentrated, digested with keratanase, and then ana-
lyzed by Western blotting using rabbit polyclonal anti-lumican or anti-
keratocan antibodies (B).
JNK inhibition increases KSPG secretion in nonactivated
and keratocan mRNAs in TGF-?1– and FGF-2–activated keratocytes.
For mRNA analysis rabbit corneal keratocytes cultured in DMEM/F12
without FBS (SFM) were either not activated (controls) or activated by
supplementing the medium in the culture dishes with 40 ng/mL FGF-2
and 5 ?g/mL HS or 2 ng/mL TGF-?1, without or with 20 ?M JNK
inhibitor (SP600125). After 60 hours in culture, the levels of specific
mRNA in the cells were analyzed by real time RT-PCR using levels of
18S rRNA to normalize values in each sample. The values presented in
the bar graph are the mean ? SD from three different experiments
(*P ? 0.05).
JNK inhibition occludes decreases in the levels of lumican
levels of lumican and keratocan mRNAs. Keratocytes cultured in
DMEM/F12 without FBS (SFM) were first transfected with JNK1/2
DsiRNAs or scrambled DsiRNA, and then either activated with 40
ng/mL FGF-2 and 5 ?g/mL HS or TGF-?1 (2 ng/mL) or not activated (SF
transfected controls) and incubated for 60 hours. JNK1, JNK2, lumi-
can, and keratocan mRNA levels were then analyzed by real time
RT-PCR using levels of 18S rRNA to normalize values in each experi-
ment. The values presented in the bar graph are the mean ? SD from
three different experiments (*P ? 0.05).
Downegulation of JNK1/2 with JNK1/2 DsiRNA increases
IOVS, November 2011, Vol. 52, No. 12
Regulation of Keratan Sulfate Proteoglycan Expression8961
KSPGs in the JNK1/2 DsiRNA transfected cells were higher
than those in the corresponding controls (Fig. 6B).
In the above experiments, where JNK1 and JNK2 were
documented to be dowregulated by JNK1/2 DsiRNA trans-
fection, a duplicate set of cells was analyzed immunocyto-
chemically. The intensities of the fluorescent signal for cell-
surface–associated KSPG(s) were diminished in the TGF-?1–
or FGF-2/HS– activated keratocytes, which were transfected
with scrambled DsiRNA (controls). However, the fluores-
cent staining of cell-surface–associated KS was evident in
activated cells previously transfected with JNK1/2 DsiRNA
(Fig. 7). These results indicated that JNK signaling pathway,
at least in part, was responsible for the decreased KS stain-
ing in the FGF-2/HS– and TGF-?1–activated keratocytes.
An injury to the corneal stroma activates keratocytes to
differentiate into fibroblasts and myofibroblasts which re-
pair the wound. The phenotypic characteristics of the acti-
vated cells are influenced by extracellular components, in-
cluding growth factors and cytokines. The activated
keratocytes in the corneal wounds are different from those
of keratocytes engaged in the embryonic and early postnatal
development of corneal stroma. Notably, the ECM assem-
bled during wound healing is not well organized and often
results in the formation of nontransparent scar tissue. The
growth factors, TGF-?1 and FGF-2, are responsible for many
of the phenotypic changes in the activated keratocytes,
including the downregulation or the loss of the expression
of KSPGs which are detrimental to collagen fiber organiza-
tion. These growth factors have been shown to induce
downregulation of KSPGs in activated bovine21and rabbit
corneal keratocytes23in culture. We have previously shown
that Rho/ROCK inhibition suppressed FGF-2– and TGF-?1–
mediated loss in the expression of keratocan and lumican.23
Therefore, we concluded that Rho-activation leads to the
loss in KSPGs in the activated keratocytes. Several recent
reports indicate that some of the downstream effects of Rho
are regulated via the JNK signaling pathway.24,31Therefore,
in the present study, we examined whether JNK is involved
in the downregulation of KSPGs in TGF-?1– and FGF-2–
Taken together, our results demonstrate that JNK activa-
tion downregulates expression of keratocan and lumican in
cultured nonactivated as well as in TGF-?1– and FGF-2–
activated keratocytes. Here we show that JNK inhibition
increased KSPG expression in keratocytes cultured in SFM
and also partially occluded the growth factor–induced de-
crease in the secreted and cell-associated KSPG proteins in
face–associated KS is occluded by JNK1/2 transfection of keratocytes.
A portion of DsiRNA transfected keratocytes, from the experiments
described in Figure 5, activated with 40 ng/mL FGF-2 and 5 ?g/mL HS
(A), or with 2 ng/mL of TGF-?1 (B), were plated for immunocytochem-
ical analyses. Double fluorescence staining was performed using
mouse anti-KS antibody followed by Alexa Fluor 488-anti-mouse IgG
antibodies and Alexa Fluor 546-phalloidin. Bar, 50 ?m.
TGF-?1– and FGF-2/HS–induced decrease in the cell sur-
occluded by JNK1/2 DsiRNA transfection of keratocytes. Keratocytes,
cultured in DMEM/F12 without FBS (SFM), were first transfected with
JNK1/2 DsiRNAs or scrambled DsiRNA, then activated with 40 ng/mL
FGF-2 and 5 ?g/mL HS or 2 ng/mL of TGF-?1, or not activated (SF
controls) for 60 hours. (A) Western blot analyses of JNK1 and JNK2 in
the extracts of nonactivated (SF), and FGF-2/HS- and of TGF-?1–acti-
vated cells using secondary antibodies (IRDye 680) and an infrared
imager (Odyssey; LI-COR). (A) Lanes A, control; B, transfected with
scrambled DsiRNA, and C, transfected with JNK1/2 DsiRNA. Bottom
panel shows tubulin bands (loading controls) in the corresponding
lanes. (B) Western blot analyses of KSPGs in culture supernatants as
determined using anti-KS antibody. Lanes A, D, and H, SFM; E and I,
nontransfected controls; B, F, and J, transfected with scrambled
DsiRNA, and C, G, and K, transfected with JNK1/2 DsiRNA.
TGF-?1– and FGF-2/HS–induced decrease in KSPGs is
8962Chen et al.
IOVS, November 2011, Vol. 52, No. 12
the activated keratocytes. While total increases in the
amounts of secreted keratocan and lumican, resulting from
JNK inhibition with SP600125, were slightly higher in the
activated than in the nonactivated keratocytes, those result-
ing from DSiRNA transfection were not significantly differ-
ent. The increases in lumican and keratocan mRNA levels in
nonactivated or activated cells, resulting from SP600125
treatments or JNK1/2 transfection, were not significantly
different. Low levels of JNK1 and JNK2 were present in the
nonactivated keratocytes, and they significantly increased
on TGF-?1– or FGF-2–induced activation. Because the in-
creases in the levels of expression of KSPGs, resulting from
JNK inhibition, were not significantly different between
activated and nonactivated keratocytes, and JNK inhibition
did not completely prevent TGF-?1– or FGF-2–induced de-
crease in either keratocan or lumican, we speculate that
activation or inhibition of other signaling pathway(s) con-
tributes to further loss in expression of these KSPGs, inde-
pendently of the JNK signaling pathway.
The occlusion of TGF-?1– or FGF-2–induced decreases in
the KSPGs was relatively less when JNK1/2 expression was
inhibited with JNK1/2 DsiRNA, than when JNK1/2 activity
was inhibited using the chemical inhibitor SP600125. These
differences were possibly due to incomplete or nonsus-
tained downregulation of JNK1/2 activity during the activa-
tion, using the DsiRNA method. The possibility also exists
that SP600125 may have some nonspecific effects which
also block the loss in KSPGs. Nonetheless, the changes in
KSPG protein and mRNA levels correlated with correspond-
ing inverse changes in the JNK1/2 proteins and mRNAs.
Therefore, we conclude that JNK acts, at least in part, to
downregulate the transcription of lumican and keratocan
during the growth factor–induced activation of keratocytes.
Zhang et al.32reported that MAPK kinase kinase
(MEKK)1-deficient mouse fetuses had normal corneal mor-
phology and thickness, but reduced transcription of kerato-
can, lumican, and collagen I. Contrary to our findings, their
observations suggest that the JNK pathway induces kerato-
can and lumican expression because MEKK1 preferentially
regulates the JNK pathway. This discrepancy indicates that
the downstream effects of JNK in the corneal stromal ker-
atocytes of the developing cornea may be different from
those in the reactivated ketatocytes during wound healing.
JNK has been reported to regulate TGF-?1–mediated expres-
sion of connective tissue growth factor (CTGF) and fi-
bronectin; thus, it regulates other phenotypic changes
which contribute to scar tissue formation.33,34JNK activa-
tion also regulates thromspondin- induced corneal neovas-
cularization,35and Toll-like receptor 2–induced corneal in-
flammation.36Based on our present findings that JNK
inhibition increases KSPG expression in activated kerato-
cytes and the reported observations described above, JNK
inhibition may potentially be a valuable approach to inhibit
scar tissue formation following corneal stromal injury or
other diseased states.
1. Benedek GB. Theory of transparency of the eye. Appl Opt. 1971;
2. Meek KM, Boote C. The organization of collagen in the corneal
stroma. Exp Eye Res. 2004;78:503–512.
3. Hassell JR, Birk DE. The molecular basis of corneal transparency.
Exp Eye Res. 2010;91:326–335.
4. Chakravarti S, Magnuson T, Lass JH, Jepsen KJ, LaMantia C, Carroll
H. Lumican regulates collagen fibril assembly: skin fragility and
corneal opacity in the absence of lumican. J Cell Biol. 1998;141:
5. Chakravarti S, Petroll WM, Hassell JR, et al. Corneal opacity in
lumican-null mice: defects in collagen fibril structure and packing
in the posterior stroma. Invest Ophthalmol Vis Sci. 2000;41:3365–
6. Chakravarti S. Functions of lumican and fibromodulin: lessons
from knockout mice 1. Glycoconj J. 2002;19:287–293.
7. Kao WW, Liu CY. Roles of lumican and keratocan on corneal
transparency 1. Glycoconj J. 2002;19:275–285.
8. Liu CY, Birk DE, Hassell JR, Kane B, Kao WW. Keratocan-deficient
mice display alterations in corneal structure. J Biol Chem. 2003;
9. Chakravarti S, Zhang G, Chervoneva I, Roberts L, Birk DE. Collagen
fibril assembly during postnatal development and dysfunctional
regulation in the lumican-deficient murine cornea 1. Dev Dyn.
10. Ahmadi AJ, Jakobiec FA. Corneal wound healing: cytokines and ex-
tracellular matrix proteins. Int Ophthalmol Clin. 2002;42:13–22.
11. Baldwin HC, Marshall J. Growth factors in corneal wound healing
following refractive surgery: a review. Acta Ophthalmol Scand.
12. Imanishi J, Kamiyama K, Iguchi I, Kita M, Sotozono C, Kinoshita S.
Growth factors: importance in wound healing and maintenance of
transparency of the cornea. Prog Retin Eye Res. 2000;19:113–129.
13. Nishida T, Tanaka T. Extracellular matrix and growth factors in
corneal wound healing. Curr Opin Ophthalmol. 1996;7:2–11.
14. Fini ME. Keratocyte and fibroblast phenotypes in the repairing
cornea. Prog Retin Eye Res. 1999;18:529–551.
15. Funderburgh JL, Cintron C, Covington HI, Conrad GW. Immuno-
analysis of keratan sulfate proteoglycan from corneal scars. Invest
Ophthalmol Vis Sci. 1988;29:1116–1124.
16. SundarRaj N, Fite D, Belak R, et al. Proteoglycan distribution
during healing of corneal stromal wounds in chick 266. Exp Eye
17. Brown CT, Applebaum E, Banwatt R, Trinkaus-Randall V. Synthesis
of stromal glycosaminoglycans in response to injury. J Cell
18. Carlson EC, Wang IJ, Liu CY, Brannan P, Kao CW, Kao WW.
Altered KSPG expression by keratocytes following corneal injury.
Mol Vis. 2003;9:615–623.
19. Funderburgh JL, Funderburgh ML, Mann MM, Prakash S, Conrad
GW. Synthesis of corneal keratan sulfate proteoglycans by bovine
keratocytes in vitro. J Biol Chem. 1996;271:31431–31436.
20. Beales MP, Funderburgh JL, Jester JV, Hassell JR. Proteoglycan
synthesis by bovine keratocytes and corneal fibroblasts: mainte-
nance of the keratocyte phenotype in culture. Invest Ophthalmol
Vis Sci. 1999;40:1658–1663.
21. Funderburgh JL, Funderburgh ML, Mann MM, Corpuz L, Roth MR.
Proteoglycan expression during transforming growth factor beta-
induced keratocyte-myofibroblast transdifferentiation 9. J Biol
22. Funderburgh JL, Mann MM, Funderburgh ML. Keratocyte pheno-
type mediates proteoglycan structure: a role for fibroblasts in
corneal fibrosis 6. J Biol Chem. 2003;278:45629–45637.
23. Chen J, Guerriero E, Sado Y, SundarRaj N. Rho-mediated regulation
of TGF-beta1- and FGF-2-induced activation of corneal stromal
keratocytes. Invest Ophthalmol Vis Sci. 2009;50:3662–3670.
24. Hall A. Rho GTPases and the control of cell behaviour. Biochem
Soc Trans. 2005;33:891–895.
25. Atfi A, Djelloul S, Chastre E, Davis R, Gespach C. Evidence for a
role of Rho-like GTPases and stress-activated protein kinase/c-Jun
N-terminal kinase (SAPK/JNK) in transforming growth factor beta-
mediated signaling. J Biol Chem. 1997;272:1429–1432.
26. Shimada H, Rajagopalan LE. Rho-kinase mediates lysophospha-
tidic acid-induced IL-8 and MCP-1 production via p38 and JNK
pathways in human endothelial cells. FEBS Lett. 2010;584:
27. Jester JV, Barry-Lane PA, Cavanagh HD, Petroll WM. Induction
of alpha-smooth muscle actin expression and myofibroblast
transformation in cultured corneal keratocytes. Cornea. 1996;
28. Guerriero E, Chen J, Sado Y, et al. Loss of alpha3(IV) collagen
expression associated with corneal keratocyte activation 3. Invest
Ophthalmol Vis Sci. 2007;48:627–635.
IOVS, November 2011, Vol. 52, No. 12
Regulation of Keratan Sulfate Proteoglycan Expression8963
29. Oike Y, Kimata K, Shinomura T, Nakazawa K, Suzuki S. Structural
analysis of chick-embryo cartilage proteoglycan by selective deg-
radation with chondroitin lyases (chondroitinases) and endo-beta-
D-galactosidase (keratanase). Biochem J. 1980;191:193–207.
30. Chen J, Guerriero E, Lathrop K, SundarRaj N. Rho/ROCK signaling
in regulation of corneal epithelial cell cycle progression 1. Invest
Ophthalmol Vis Sci. 2008;49:175–183.
31. Shirai H, Autieri M, Eguchi S. Small GTP-binding proteins and
mitogen-activated protein kinases as promising therapeutic targets
of vascular remodeling. Curr Opin Nephrol Hypertens. 2007;16:
32. Zhang L, Deng M, Kao CW, Kao WW, Xia Y. MEK kinase 1
regulates c-Jun phosphorylation in the control of corneal morpho-
genesis. Mol Vis. 2003;9:584–593.
33. Chang Y, Wu XY. The role of c-Jun N-terminal kinases 1/2 in
transforming growth factor beta(1)-induced expression of connec-
tive tissue growth factor and scar formation in the cornea. J Int
Med Res. 2009;37:727–736.
34. Chang Y, Wu XY. JNK1/2 siRNA inhibits transforming-growth
factor-beta1-induced connective tissue growth factor expres-
sion and fibrotic function in THSFs. Mol Cell Biochem. 2010;
35. Jimenez B, Volpert OV, Reiher F, et al. c-Jun N-terminal kinase
activation is required for the inhibition of neovascularization by
thrombospondin-1. Oncogene. 2001;20:3443–3448.
36. Adhikary G, Sun Y, Pearlman E. C-Jun NH2 terminal kinase (JNK)
is an essential mediator of Toll-like receptor 2-induced corneal
inflammation. J Leukoc Biol. 2008;83:991–997.
8964 Chen et al.
IOVS, November 2011, Vol. 52, No. 12