Acta Ophthalmologica 2012

The wound healing effects of

vitamin A eye drops after a
corneal alkali burn in rats
Eun Chul Kim,1 Tai Kyong Kim,1 Sang Hee Park2 and Man Soo Kim1
1

Department of Ophthalmology & Visual Science, College of Medicine, Catholic

University of Korea, Seoul, Korea
2
Institute of Clinical Medicine Research, College of Medicine, Catholic University
of Korea, Seoul, Korea

ABSTRACT.
Purpose: We investigated the wound healing effect of retinyl palmitate eyedrops following a corneal alkali burn in rats.
Methods: A total of 160 Sprague-Dawley male rats were divided into two
groups and central corneas were injured by contacting eyes with lter paper
saturated with 0.01 m NaOH for 45 seconds. Vitamin A group was treated
with retinyl palmitate and antibiotic (Cravit: 0.5% levooxacin) eye drops
four times daily for 3 days and the control group with vehicle and antibiotic
eye drops. Corneal wound healing by uorescein staining and impression cytologic analysis were conducted at 0, 24, 48 and 72 hr after injury. Vascular
endothelial growth factor A (VEGF-A), thrombospondin 2, matrix metalloproteinase 9 (MMP 9) and transforming growth factor-b (TGF-b) were measured
in corneas by ELISA, immunouorescent staining and real-time PCR.
Results: Corneal wound healing was better in the vitamin A group than in the
control group. Early sprouting of new vessel was observed in the control group
at 72 hr, but not in the vitamin A group. Corneal thrombospondin 2 proteins
in ELISA were higher in the vitamin A group, but VEGF-A, MMP 9 and
TGF-b proteins were higher in the control group (p < 0.05). Similarly, thrombospondin 2 immunouorescent staining was stronger, whereas VEGF-A,
MMP 9 and TGF-b staining were weaker in the vitamin A group (p < 0.05).
In addition, thrombospondin 2 mRNA levels were higher, whereas VEGF-A,
MMP 9 and TGF-b mRNA levels were lower in the vitamin A group
(p < 0.05).
Conclusions: Retinyl palmitate eye drops can inhibit VEGF-A and activate
thrombospondin 2 and improve conjunctival impression cytologic ndings. Furthermore, retinyl palmitate eye drops were found to promote corneal healing
after an alkali burn in rats.
Key words: corneal alkali burn corneal wound healing impression cytology vitamin A
eyedrops

Acta Ophthalmol. 2012: 90: e540e546

2012 The Authors
Acta Ophthalmologica 2012 Acta Ophthalmologica Scandinavica Foundation

doi: 10.1111/j.1755-3768.2012.02496.x

e540

Introduction
Corneal alkali burns cause corneal
infection, ulceration, perforation, neovascularization (NV) and opacication
(Salman & Gundogdu 2010). Because
the cornea is a physiologically transparent tissue, the balance between
angiogenic and antiangiogenic factors
determines the regulation of corneal
NV after a corneal alkali burn (Chang
et al. 2001).
Vascular endothelial growth factor
A (VEGF-A) promotes the division
and proliferation of endothelial cells
and enhances blood capillary permeability (Byrne et al. 2005). On the
other hand, thrombospondin 2 (TSP2)
is expressed strongly during several
states of cell migration and proliferation and inhibits angiogenesis (Armstrong & Bornstein 2003). During the
course of wound healing, TSP2 participates in the regulation of cell
de-adherence and represses the
growths of new vessels into the cornea
(Armstrong & Bornstein 2003). In
addition to VEGF-A, other angiogenic
stimulators, including basic broblast
growth factor (bFGF) and transforming growth factor (TGF)-a and b, can
cause capillary endothelial cells to
proliferate and cause corneal neovascularization (NV) (Chang et al. 2001).
TGF-b also has the ability to
induce the expressions of other cytokines, such as matrix metalloproteinase 9 (MMP 9), VEGF-A and
monocyte macrophage
chemotactic

Acta Ophthalmologica 2012

protein-1, which are believed to be

involved in matrix degradation, local
neovascularization and inammation,
respectively (Saika 2004; Saika et al.
2005). TGF-b also initiates keratocyte
to myobroblast conversion after a
corneal burn (Saika et al. 2005), and
this has been suggested to be the main
mechanism of corneal haze formation
after injury, because myobroblasts
exhibit a different level of crystalline
production and are less transparent
than keratocytes (Jester et al. 1999;
Zieske et al. 2001).
Vitamin A is necessary for the normal growth and differentiation of epithelium (Kobayashi et al. 1997), and
vitamin A deciency causes the loss of
goblet cells, leading to increased epidermal keratinization and squamous
metaplasia of mucous membranes,
including the cornea and conjunctiva
(Hatchell & Sommer 1984). Furthermore, vitamin A may modulate the
expression of TSP1 to accelerate the
re-epithelialization of wounded corneas (Uno et al. 2005), and TSP1 can
also inhibit VEGF-A signalling by
directly binding to the protein and by
competing with VEGF-A for binding
to cell surface heparan sulphate proteoglycans (Gupta et al. 1999). The
objective of this study was to investigate the effects of Vitamin A eye
drops on wound healing following
corneal alkali burn in rats.

Materials and Methods

Animals

For the wound healing and eye drop

application studies, 160 Sprague-Dawley male rats (weight range, 250
300 g) were used. Animals were
divided into two groups of 80 rats.
The animal care committee of
Bucheon St. Marys Hospital of the
Catholic University Medical Center,
in compliance with the ARVO Statement for the Use of Animals in
Ophthalmic and Vision Research,
approved all animal protocols.
In vivo animal model of corneal epithelial
wound healing

Rats were deeply anesthetized by an

intraperitoneal (i.p.) injection of
50 mg kg tiletamine plus zolazepam
(Zoletil; Virbac, Carros, France) and
15 mg kg
xylazine
hydrochloride

(Rompun; Bayer, Leuverkeusen, Germany). In each case, the central cornea

of the right eye was injured by placing
a lter paper (circular 3.0 mm diameter) saturated with 0.01 m NaOH on it
for 45 seconds. The wound surface
was then washed with 0.9% physiological saline. Vitamin A ophthalmic
solutions were prepared by dissolving
retinyl palmitate (Sigma-Aldrich, St
Louis, MO, USA) in saline with the
aid of a surfactant (0.5% polyoxyethylene hydrogenated castor oil 60). The
vitamin A group was treated with
aqueous ophthalmic retinyl palmitate
solution (1500 IU ml, 100 ll time)
and antibiotic (Cravit: 0.5% levooxacin) eye drops four times daily for
3 days, and the control group was
treated with vehicle and antibiotic eye
drops in the same manner.
Wound sizes were determined by
staining with 1% uorescein and photographing immediately after injury
and at 24, 48 and 72 hr after injury.
Areas of wounds were quantied from
photographs using a computerassisted image analyzer (Cell F; Olympus GmbH, Mainz, Germany). The
percentage of wound healing was calculated using the following formula:
initialwoundarea  woundarea=
initialwoundarea  100.
Impression cytology of bulbar conjunctiva

Impression cytological analysis was

performed on lower bulbar conjunctivae by applying strips of cellulose acetate lter paper for 3 seconds (MFS
membrane lters; Advantec MFS,
Dublin, CA, USA), as previously
described (Anshu et al. 2001) at 0, 24,
48 and 72 hr postinjury. Single animals
were tested at 0, 1, 2 and 3 days to
determine the grade of impression
cytology. Specimens were xed in formalin, stained with periodic acid-Schiff
and photographed at a magnication
of 200. Degrees of squamous metaplasia were graded from zero to six as
described by (Tseng 1985a), and goblet
cell density was dened as the number
of cells per square millimetre. All 160
right eyes were included in the analysis.
Protein extraction and ELISA for
VEGF-A, thrombospondin 2, MMP 9
and TGF-b

Corneas were trephined using a

4.0-mm-diameter trephine at 24, 48

and 72 hr postinjury (n = 20 per

group). Excised corneas were frozen
at )80C and homogenized by thawing and freezing three times in a lysis
buffer containing 50 mm Pipes KOH
(pH 6.5), 2 mm EDTA, 0.1% Chaps,
20 mg ml Leupeptin, 10 mg ml Pepstatin, 10 mg ml Aprotinin, 5 mm DTT
and protease inhibitors. Homogenates
were centrifuged at 14 000 g for
30 min, and the supernatants so
obtained were subjected to VEGF-A,
thrombospondin 2, MMP 9 and
TGF-b assays. VEGF-A, thrombospondin 2, MMP 9 and TGF-b levels
were measured before injury and at
24, 48 and 72 hr postinjury. VEGF-A
(R&D Systems, Minneapolis, MN,
USA), thrombospondin 2 (Uscn Life
Science Inc.,Wuhan, China), MMP-9
(Cusabio Biotech, Wuhan, China) and
TGF-b
(AbFRONTIER,
Seoul,
Korea) levels were measured using an
ELISA systems developed for rats.
Immunouorescence staining

Twenty rats in each group were used

for immunouorescent staining. Rat
corneas at 24, 48 and 72 hr postinjury
were embedded in an optical cutting
temperature compound (Sakura Finetek, Torrance, CA, USA), frozen and
cut into cryosections (4 lm). Frozen
sections were thawed, dehydrated and
xed in cold acetone for 10 min.
Sections were incubated in blocking
solution (normal goat serum; Zymed,
South San Francisco, CA, USA) for
30 min, and immunouorescent staining was performed using afnity-puried mouse monoclonal antibodies
against VEGF-A (C-1; Santa Cruz
Biotechnology, Santa Cruz, CA,
USA), thrombospondin 2 (N-15;
Santa Cruz Biotechnology), MMP 9
(EP1254; Abcam, Cambridge, MA,
USA) and TGF-b (MAB240; R&D
systems) as previously described
(Manzano et al. 2007; Bock et al.
2009). Primary antibodies against
VEGF-A, thrombospondin 2, MMP 9
and TGF-b were used at a dilution of
1:50 and were incubated for 1 hr at
room temperature. Fluorescein-5-isothiocyanate (FITC) labelled secondary
antibodies (Santa Cruz Biotechnology)
were then applied to sections, which
were incubated in a dark chamber for
45 min and counterstaining with
DAPI (4_6-diamidino-2-phenylindole,
dihydrochloride; 1:1000) for 5 min.

e541

Acta Ophthalmologica 2012

After washing sections with PBS,

antifade mounting medium (Gel
Mount; Biomedia Corp., Foster City,
CA, USA) and coverslips were
applied. Staining was evaluated under
a uorescence microscope (BX50;
Olympus, Tokyo, Japan) and photographed with a digital camera (DP70Set2; Olympus).

Table 1. Gene expression assays used for real-time polymerase chain reaction.

Gene

Abbreviation

Reference
sequence

Assay number

Vascular endothelial growth

factor A
Thrombospondin 2
Matrix metallopeptidase 9
Transforming growth factor,
beta 1
Glyceraldehyde 3-phosphate

VEGF164

NM_031836.2

Rn00582935_m1

Thbs2
MMP-9
TGF-b1

NM_001169138.1
NM_031055.1
NM_021578.2

Rn02111874_s1
Rn00579162_m1
Rn99999016_m1

GAPDH

NM_017008.3

Rn99999916_s1

Real-time polymerase chain reaction

RNA samples were collected using the

RNeasy Mini kit (QIAGEN, Valencia,
CA, USA) and treated with deoxyribonuclease 1 (QIAGEN) to eliminate
genomic DNA. Aliquots (1 lg) of
puried RNA were reverse transcribed
into rst-strand cDNA using an
iScript reverse transcription kit (BioRad Laboratories, Hercules, CA,
USA), which included a genomic
DNA elimination step (QIAGEN).
Real-time polymerase chain reaction (PCR) amplication and the relative
quantications
of
VEGF,
thrombospondin 2, MMP 9, and
TGF-b and Acta2 were performed
using TaqMan gene expression assays
(Table 1) (Applied Biosystems, Foster
City, CA, USA) using a lightcycler
480 PCR system (Roche, Mannheim,
Germany). Assays had similar amplication efciencies, and relative quantication was performed using a DCT
experimental design. Reactions were
performed in duplicates in total volumes of 20-ll using TaqMan probe
Master Mix (Roche); 10 ng of cDNA
was used in each reaction. Glyceraldehyde
3-phosphate
dehydrogenase
(GAPDH) served as an endogenous
control. No-template control was
included during quantitative real-time
PCR to conrm the absence of DNA
contamination in the reagents used
for amplication. Results were analysed using Lightcycler 480 instrument
software 1.2 (Roche) and were
expressed relative to average healthy
cornea results. The statistical analysis
of quantitative real-time PCR data
was performed using two-way anova
and DataAssist Software v3.0
(Applied Biosystems, Foster City, CA,
USA). Statistical signicance was
accepted for p-values of < 0.05. All
samples were analysed separately, and
the average gene expression value
(geometric mean) of each tissue type
is presented. Data are expressed as
relative quantities (RQ), and differ-

e542

ences are shown in the gures as

expression ratios versus the normalized target gene.
Statistical analysis

Results obtained in the two groups

were expressed as means standard
deviations (SD). Two-way anova was
used to determine whether there were
signicant changes in corneal wound
healing rates, corneal VEGF-A,
thrombospondin 2, MMP 9 and
TGF-b levels, and VEGF-A, thrombospondin 2, MMP 9 and TGF-b
mRNA levels after epithelial debridement. The chi-square was used to
analyse semi-quantitative immunouorescence staining results. Statistical
signicance was accepted for p-values
of < 0.05 throughout.

Results
Effect of vitamin A eye drops on corneal
epithelial wound healing

Initial areas of injury ranged from

15.3 to 17.6 mm2. Quantitative analysis revealed that the percentage wound
healing in the vitamin A group
(35.8 7.1%) was signicantly greater
than in the control group (18.8

4.5%)
at
24 hr
postinjury
(p < 0.05). At 48 and 72 hr, percentage wound healing in the vitamin
A group (69.1 16.5% and 82.1
17.9%, respectively) was signicantly
larger than in the control group
(38.2 9.5% and 44.8 10.2%)
(p < 0.05 for both) (p < 0.05) (Fig. 1).
Corneal photographs in the control
group show stromal haze at 24 hr,
corneal opacity at 48 hr and early
sprouting of new vessels and slight
limbal damage at 72 hr. However, in
the vitamin A group, epithelial defects
healed and corneal opacity was not
observed (Fig. 2). There was no crosscorrelation between the groups and
time in two-way anova. Wound healing increased with time in both
groups.
Impression cytological ndings

Immediately after injury, mean

impression cytology grades in the two
groups ranged between 3.85 and 4.25
(on a scale of 05), and no signicant
difference was observed between the
groups. The decreases of impression
cytology grades from immediately
after injury in the vitamin A group
()0.46 0.09, )1.21 0.12 and
)1.53 0.16)
were
signicantly

Fig. 1. Slit lamp photographs of uorescein staining of corneal epithelial defects in the vitamin
A group and the control group immediately after and at 24, 48 and 72 hr after a corneal alkali
burn. Photographs of uorescein staining showing better wound healing in the vitamin A group
than in the control group at 24, 48 and 72 hr after a corneal alkali burn.

Acta Ophthalmologica 2012

in the control group (p < 0.05)

(Fig. 6B). No cross-correlation was
found between groups and times in
two-way anova.
Immunouorescence staining for VEGF,
thrombospondin 2, MMP 9 and TGF-b

Fig. 2. Slit lamp photographs of corneas in the vitamin A group and the control group immediately after and at 24, 48 and 72 hr after a corneal alkali burn. Photographs in the control group
show stromal haze at 24 hr postinjury, corneal opacity obscured pupils at 48 hr, and early new
vessel sprouting (red arrow) and slight limbal damage (blue arrow) were observed at 72 hr.
However, in the vitamin A group, epithelial defects healed with little haze.

VEGF-A, thrombospondin 2, MMP 9

and TGF-b protein levels in corneas

Fig. 3. Changes in impression cytology grade

in the vitamin A and control groups at 24,
48 and 72 hr postinjury versus immediately
after injury. Impression cytology grades in
the vitamin A group improved more than in
the control group at 24, 48 and 72 hr
postinjury (*p < 0.05; n = 20 per group;
two-way anova). Data represent means
standard deviations.

greater than in the control group

()0.10 0.03, )0.17 0.05 and )0.35
0.08) at 24, 48 and 72 hr postinjury
(p < 0.05) (Fig. 3). Impression cytology pictures taken at the different
times are shown in Fig. 4.

VEGF-A levels (9.7 0.82, 10.6

0.97 and 12.1 1.07 pg ml) in the
vitamin A group were signicantly
lower than in the control group
(27.4 2.12, 56.9 3.43 and 45.6
3.26 pg ml) at 24, 48 and 72 hr postinjury (p < 0.05) (Fig. 5A). However, at the same times, thrombospondin
2 levels (6.9 0.78, 15.9 1.56 and
11.4 1.25 pg ml) in the vitamin A
group were signicantly higher than in
the control group (1.4 0.12, 5.6
0.59 and 4.5 0.43 pg ml) (p <
0.05) (Fig. 5B). On the other hand,
MMP 9 levels (31.5 2.15, 48.7
2.81 and 23.3 1.75 pg ml) in the
vitamin A group were signicantly
lower than in the control group
(81.9 3.45, 117.8 3.87 and 91.6
3.28 pg ml) (p < 0.05) (Fig. 6A),
as were TGF-b levels (21.7 1.47,
42.5 2.14 and 37.8 1.98 pg ml in
the vitamin A group and 68.7 3.06,
102.8 4.25 and 72.3 3.18 pg ml

Immunouorescent
staining
for
thrombospondin 2 was stronger in the
vitamin A group than in the control
group, but immunouorescent staining for VEGF-A, MMP 9 and TGF-b
were weaker at 24, 48 and 72 hr
postinjury. Immunouorescent staining for VEGF-A, thrombospondin 2,
MMP 9 and TGF-b were strongest at
48 postinjury (Fig. 7). Twenty-four
samples were positive for thrombospondin 2 immunouorescent staining,
and 3 was negative in vitamin A
group (total = 27), but only 6 samples were positive for thrombospondin
2, and 21 was negative in control
group (total = 27). The rate of
Thrombospondin 2 staining was statistically signicantly greater in vitamin A group compared to control
group (chi-square test; p < 0.05).
Real-time polymerase chain reaction

Real-time PCR showed that levels of

thrombospondin 2 mRNA transcripts
were 1.53.2-fold higher in the vitamin
A group than that in the control
group at 24, 48 and 72 hr postinjury
(p < 0.05).
However,
VEGF-A,
MMP 9 and TGF-b mRNA expressions in the vitamin A group were
0.350.56-fold lower than in the control group at 24, 48 and 72 hr
(p < 0.05) (Fig. 8).

Discussion

Fig. 4. Photomicrographs showing changes in impression cytologic analysis grades in the vitamin A and control groups at 24, 48 and 72 hr postinjury versus immediately after injury.
Improvement in impression cytological analysis results was observed in the vitamin A group:
(stages 4, 3, 2, 1 after initial wounding and at 24, 48 and 72 hr postinjury, respectively). However, impression cytological analysis in the control group showed less improvement (stages 4, 4,
3, 3 after initial wounding and at 24, 48 and 72 hr postinjury).

The cornea and the conjunctiva have

an absolute requirement for vitamin
A (retinol), which acts through its
metabolites, such as retinaldehyde,
which forms the visual chromophore
and retinoic acids, which regulate
gene expression (Nezzar et al. 2007).
A lack of vitamin A causes abnormal
differentiation of the ocular surface
and results in keratinization, ulceration, epithelial squamous metaplasia
and a deciency of conjunctival
goblet cells (Hatchell & Sommer
1984; Macsai et al. 1998). Owing to
these effects, topical retinoids have
been tested as pharmacological agents

e543

Acta Ophthalmologica 2012

(A)

(B)

Fig. 5. Corneal vascular endothelial growth factor A (VEGF-A) (A) and thrombospondin 2 (B)
concentrations after a corneal alkali burn in the vitamin A and control groups. Corneal VEGFA concentrations in the vitamin A group were decreased, but thrombospondin 2 concentrations
were increased versus the control group at 24, 48 and 72 hr postinjury (*p < 0.05; n = 20 per
group; two-way anova). Data represent means standard deviations.

(A)

(B)

Fig. 6. Corneal matrix metalloproteinase 9 (MMP 9) (A) and transforming growth factor-b
(TGF-b) (B) concentrations in the vitamin A and control groups. Corneal MMP 9 and TGF-b
concentrations in the vitamin A group were lower than in the control group at 24, 48 and 72 hr
postinjury (*p < 0.05; n = 20 per group; two-way anova). Data represent means standard
deviations.

Fig. 7. Immunouorescent staining for vascular endothelial growth factor A (VEGF-A), thrombospondin 2, matrix metalloproteinase 9 (MMP 9) and transforming growth factor-b (TGF-b)
in the vitamin A and control groups at 48 hr after corneal alkali burn injury. DAPI was used
for nuclear staining (blue). Original magnication, 200X. At 48 hr postinjury, immunouorescent staining for thrombospondin 2 was stronger in the vitamin A group, but immunouorescent stainings for VEGF-A, MMP 9 and TGF-b were weaker than in the control group.

for the treatment of ocular surface

diseases associated with wounds,
squamous metaplasia or dry eye
(Kim et al. 2009). Tseng demonstrated that topical all-trans retinoic
acid ointment was effective at treating four patients severely affected by,
keratoconjunctivitis sicca, StevensJohnson
syndrome,
drug-induced

e544

pseudopemphigoid or surgery-induced
dry eye (Tseng 1985b). The vitamin
A eye drops used in the present contained retinyl palmitate (solubilized
by surfactant) in saline.
Acyclic retinoid was found to suppress both angiogenesis and hepatocellular carcinoma-induced angiogenesis
in a xenografted chicken chorioallan-

Fig. 8. Real-time PCR of the relative expressions of vascular endothelial growth factor A
(VEGF-A), thrombospondin 2, matrix metalloproteinases 9 (MMP 9) and transforming
growth factor-b (TGF-b) mRNA in the vitamin A and control groups. VEGF-A, thrombospondin 2, MMP 9 and TGF-b expressions
in the vitamin A group are presented relative
to their expressions in the control group
(RQ = 1). VEGF-A, MMP 9 and TGF-b
mRNA expressions in the vitamin A group
were signicantly depressed, but thrombospondin 2 mRNA expressions were signicantly enhanced at 24, 48 and 72 hr
postinjury (*p < 0.05; n = 20 per group;
two-way anova). Data are presented as
means standard deviations.

toic membrane model (Komi et al.

2010). Furthermore, thrombospondin-1,
which can inhibit VEGF-A and thus
inhibit angiogenesis and accelerate
corneal wound healing like thrombospondin 2 (Yan et al. 2007),
expression was found to be impaired
during epithelial wound healing in the
corneas of vitamin A-decient mice
(Uno et al. 2005). In addition, TSP2
expression was found to inhibit
VEGF during the early stage after an
alkali burn (Yan et al. 2007). However, no reports have been issued on
the association between vitamin A
and thrombospondin 2 after an alkali
burn. On the basis of our results, we
hypothesize that vitamin A may promote thrombospondin 2 expression,
and thus inhibit VEGF.
After corneal alkali injury, the corneal epithelium secretes growth factors
and cytokines, including TFG-b,
which bind to cell surface receptors
and initiate multiple signal transduction cascades (Myrna et al. 2009).
Transforming growth factor initiates
keratocyte to myobroblast conversion (Saika et al. 2010) and this process plays a vital role in the contraction
and closure of incisional corneal
wounds and surface re-epithelialization
(Saika et al. 2010). However, it has
also been implicated in a number of
pathologic processes including the formation of corneal haze (Ebihara et al.

Acta Ophthalmologica 2012

2007). Myobroblasts also participate

in wound remodelling through phagocytosis and by expressing transglutaminase (which cross-links the extracellular
matrix) and matrix metalloproteinases
(MMPs) (Priglinger et al. 2005; Ollivier et al. 2007). Furthermore, the
rapid degradation of corneal stroma
associated with corneal ulcers appears
to be caused by these MMPs (Qiu
et al. 2010). However, all-trans retinoic acid (ATRA) was found to
downregulate MMP-9 by upregulating
TIMP-1 and E-cadherin and downregulating EGFR, NFkB, FAK, and
a5b1 and aVb3 integrin receptors
in vitro (Dutta et al. 2010).
MMP-9 is a well-known regulator
and effector of many cellular processes, including wound healing
(Gordon et al. 2009). Furthermore,
MMP-9 is important for the migration
of basal epithelial cells. For example,
following epithelial closure, MMP-9
expression spreads distally throughout
wound sites and the pattern of its
expression matches remodelling in the
basement membrane zone beneath the
epithelium (Mohan et al. 1998). However, the increased expressions and
elevated activities of a wide range of
MMPs are responsible for corneal
melting in primary Sjorgens syndrome (Brejchova et al. 2009).
In the present study, MMP-9 was
highly expressed in the control group,
and thus, corneal wound healing was
delayed. On the other hand, in the
vitamin A group, MMP-9 expression
was depressed, and thus, basal epithelial cells migration might have been
greater than in the control group.
In this study, wound healing was
improved in the vitamin A group than
in the control group (Fig. 1). According to our ELISA studies, corneal
VEGF-A, MMP 9 and TGF-b proteins in the vitamin A group were
lower and corneal thrombospondin 2
protein was higher than in the control
group (p < 0.05) (Figs 5 and 6). Furthermore, immunouorescent staining
for thrombospondin 2 was stronger in
vitamin A corneas, but those of
VEGF-A, MMP 9 and TGF-b were
weaker (Fig. 7). In addition, thrombospondin 2 expression in the vitamin A
group was upregulated and VEGF-A,
MMP 9 and TGF-b expressions were
downregulated versus the control
group (p < 0.05) (Fig. 8).

Vitamin A palmitate may promote

the repair of mechanical corneal epithelial defects and the development
of intracellular conjunction (Qiu
et al. 2010). Impression cytology, dry
eye symptoms, visual acuity, keratopathy and the Schirmer test were
reported to be improved after the
application of topical all-trans retinoic acid ointment (Tseng 1985b).
Vitamin A eye drops have also been
reported to improve symptoms of
blurred vision, tear lm tBUT, Schirmer I score and impression cytological ndings in patients with severe
dry eye syndrome (Kim et al. 2009).
In the present study, impression
cytology grade and goblet cell densities were improved in the vitamin A
group versus the control group
(Figs 3 and 4). Impression cytology
can inuence wound healing, but in
the present study, strips of cellulose
acetate lter paper were applied to
lower bulbar conjunctivae for 3 seconds and not to the cornea. Thus,
we believe that impression cytology
had little inuence on the corneal
wound healing observed.
The present study, also demonstrates that vitamin A eye drops
containing antibiotics activate thrombospondin 2 and inhibit VEGF-A
more so than control vehicle containing antibiotics after a corneal
alkali burn in rats. In addition, it
shows that vitamin A eye drops
may inhibit proteinases such as,
MMP 9 and inhibit myobroblast
transformation by downregulating
TGF-b.
To the best of our knowledge, this
study is the rst to show that vitamin A eye drops can promote
wound healing by activating thrombospondin 2 and inhibiting VEGFA, MMP 9 and TGF-b following a
corneal alkali burn in rats. The
study also demonstrates that vitamin
A eye drops improve impression
cytology ndings and inhibit corneal
degradation and corneal haze formation following a corneal alkali burn
in rats.

Acknowledgements
This work was supported by the Institute of Clinical Medicine Research of
Bucheon
St.
Marys
Hospital,
Research Fund, 2010.

References
Anshu, Munshi MM, Sathe V & Ganar A
(2001): Conjunctival impression cytology in
contact lens wearers. Cytopathology 12:
314320.
Armstrong LC & Bornstein P (2003): Thrombospondins 1 and 2 function as inhibitors
of angiogenesis. Matrix Bio 22: 6371.
Bock F, Onderka J, Rummelt C, Dietrich T,
Bachmann B, Kruse FE, Schlotzer-Schrehardt U & Cursiefen C (2009): Safety prole of topical VEGF neutralization at the
cornea. Invest Ophthalmol Vis Sci 50:
20952102.
Brejchova K, Liskova P, Hrdlickova E, Filipec M & Jirsova K (2009): Matrix metalloproteinases in recurrent corneal melting
associated with primary Sjorgens syndrome. Mol Vis 15: 23642372.
Byrne AM, Bouchier-Hayes DJ & Harmey
JH (2005): Angiogenic and cell survival
functions of vascular endothelial growth
factor (VEGF). J Cell Mol Med 9: 777
794.
Chang JH, Gabison EE, Kato T & Azar DT
(2001): Corneal neovascularization. Curr
Opin Ophthalmol 12: 242249.
Dutta A, Sen T & Chatterjee A (2010): Alltrans retinoic acid (ATRA) downregulates
MMP-9 by modulating its regulatory molecules. Cell Adh Migr 4: 409418.
Ebihara N, Yamagami S, Chen L, Tokura T,
Iwatsu M, Ushio H & Murakami A (2007):
Expression and function of toll-like receptor-3 and -9 in human corneal myobroblasts. Invest Ophthalmol Vis Sci 48: 3069
3076.
Gordon GM, Ledee DR, Feuer WJ & Fini
ME (2009): Cytokines and signaling pathways regulating matrix metalloproteinase-9
(MMP-9) expression in corneal epithelial
cells. J Cell Physiol 221: 402411.
Gupta K, Gupta P, Wild R, Ramakrishnan S
& Hebbel RP (1999): Binding and displacement of vascular endothelial growth factor
(VEGF) by thrombospondin: effect on
human microvascular endothelial cell proliferation and angiogenesis. Angiogenesis 3:
147158.
Hatchell DL & Sommer A (1984): Detection
of ocular surface abnormalities in experimental vitamin A deciency. Arch Ophthalmol 102: 13891393.
Jester JV, Moller-Pedersen T, Huang J, Sax
CM, Kays WT, Cavangh HD, Petroll WM
& Piatigorsky J (1999): The cellular basis
of corneal transparency: evidence for corneal crystallins. J Cell Sci 112: 613622.
Kim EC, Choi JS & Joo CK (2009): A comparison of vitamin a and cyclosporine a
0.05% eye drops for treatment of dry eye
syndrome. Am J Ophthalmol 147: 206
213.
Kobayashi TK, Tsubota K, Takamura E,
Sawa M, Ohashi Y & Usui M (1997):
Effect of retinol palmitate as a treatment
for dry eye: a cytological evaluation. Ophthalmologica 211: 358361.

e545

Acta Ophthalmologica 2012

Komi Y, Sogabe Y, Ishibashi N, Sato Y,

Moriwaki H, Shimokado K & Kojima S
(2010): Acyclic retinoid inhibits angiogenesis by suppressing the MAPK pathway.
Lab Invest 90: 5260.
Macsai MS, Agarwal S & Gamponia E
(1998): Bilateral corneal ulcers in primary
vitamin A deciency. Cornea 17: 227229.
Manzano RP, Peyman GA, Khan P, Dietrich
T, Bachmann B, Kruse FE, SchlotzerSchrehardt U & Cursiefen C (2007): Inhibition
of experimental corneal neovascularization
by bevacizumab (Avastin). Br J Ophthalmol
91: 804807.
Mohan R, Rinehart WB, Bargagna-Mohan P
& Fini ME (1998): Gelatinase B lacZ
transgenic mice, a model for mapping gelatinase B expression during developmental
and injury-related tissue remodeling. J Biol
Chem 273: 2590325914.
Myrna KE, Pot SA & Murphy CJ (2009):
Meet the corneal myobroblast: the role of
myobroblast transformation in corneal
wound healing and pathology. Vet Ophthalmol 12: 2527.
Nezzar H, Chiambaretta F, Marceau G,
Blanchon L, Faye B, Dechelotte P, Rigal
D & Sapin V (2007): Molecular and metabolic retinoid pathways in the human ocular surface. Mol Vis 13: 16411650.
Ollivier FJ, Gilger BC, Barrie KP, Kallberg
ME, Plummer CE, OReilly S, Gelatt KN
& Brooks DE (2007): Proteinases of the
cornea and preocular tear lm. Vet Ophthalmol 10: 199206.
Priglinger SG, May CA, Alge CS, Wolf A,
Neubauer AS, Kampik A & Welge-Luessen

e546

U (2005): Absence of scar formation in

human donor cornea with prior laser in situ
keratomileusis. J Cataract Refract Surg 31:
14031408.
Qiu XD, Gong L & Chen MJ (2010):
Research on effects of vitamin A palmitate
on repair of mechanical corneal epithelial
defects and conjunctival goblet cells in rabbits. Zhonghua Yan Ke Za Zhi 46: 151
160.
Saika S (2004): TGF-beta signal transduction
in corneal wound healing as a therapeutic
target. Cornea 23: S25S30.
Saika S, Ikeda K, Yamanaka O et al. (2005):
Expression of Smad7 in mouse eyes accelerates healing of corneal tissue after
exposure to alkali. Am J Pathol 166: 1405
1418.
Saika S, Yamanaka O, Sumioka T, Okada Y,
Miyamoto T, Shirai K, Kitano A &
Tanaka S (2010): Transforming growth
factor beta signal transduction: a potential
target for maintenance restoration of
transparency of the cornea. Eye Contact
Lens 36: 286289.
Salman IA & Gundogdu C (2010): Epithelial
healing in experimental corneal alkali
wounds with nondiluted autologous serum
eye drops. Cutan Ocul Toxicol 29: 116
121.
Tseng SC (1985a): Staging of conjunctival
squamous metaplasia by impression cytology. Ophthalmology 92: 728733.
Tseng SC (1985b): Topical retinoid treatment
for dry eye disorders. Trans Ophthalmol
Soc UK 104: 489495.

Uno K, Kuroki M, Hayashi H, Uchida H,

Kuroki M & Oshima K (2005): Impairment
of thrombospondin-1 expression during
epithelial wound healing in corneas of vitamin A-decient mice. Histol Histopathol
20: 493499.
Yan J, Zeng Y, Jiang J, Zhou J, Yin Z,
Wang Z & Zhu P (2007): The expression
patterns of vascular endothelial growth factor and thrombospondin 2 after corneal
alkali burn. Colloids Surf B Biointerfaces
60: 105109.
Zieske JD, Hutcheon AE, Guo X, Chung
EH & Joyce NC (2001): TGF-beta receptor types I and II are differentially
expressed during corneal epithelial wound
repair. Invest Ophthalmol Vis Sci 42:
14651471.

Received on December 8th, 2011.

Accepted on June 4th, 2012.
Correspondence:
Man Soo Kim, MD, PhD
Department of Ophthalmology
Seoul St. Marys Hospital
No. 505 Ban-po Dong
Seocho-Ku
Seoul 137-040
Korea
Tel: + 82 2 2258 6197
Fax: + 82 2 599 7405
Email: mskim@catholic.ac.kr