Embed Size (px)
Citation preview
Exp. Eye Res. (1998) 67, 133–142Article Number : ey980512
The Heparan Sulfate Suleparoide Inhibits Rat Corneal
Angiogenesis and in vitro Neovascularization
UMBERTO BENELLIa, GUIDO BOCCIb, ROMANO DANESId, ANTONIO LEPRIa, NUNZIA
BERNARDINIc, FRANCESCO BIANCHIc, MARIO LUPETTIc, AMELIO DOLFIc, ANTONIO
CAMPAGNIe, CRISTIANA AGENb, MARCO NARDIa MARIO DEL TACCAb*
a Department of Neurosciences, Division of Ophthalmology; b Department of Oncology, Division of
Pharmacology and Chemotherapy and c Department of Human Morphology and Applied Biology,
University of Pisa, Pisa, Italy ; d Superior School of University Studies and Doctoral Research S. Anna,
Pisa, Italy and e C.I.S.A.M., Interforce Centre for Studies on Military Operations, S. Piero a Grado, Italy
(Received Lund 9 January 1998 and accepted in revised form 6 March 1998)
The purpose of this study was to evaluate the inhibitory activity of the heparan sulfate suleparoide onvascular cell growth in vitro and angiogenesis in vivo.
Human HUV-EC-C endothelial cell proliferation and microvessel sprouting from cultured rat aorticrings were assayed by the bioreduction of 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide.The inhibition of the neoforming capillary network in the chorioallantoic membrane of chick embryo(CAM) was evaluated by agarose disks containing suleparoide and applied on the CAM surface.AgNO
$}KNO
$injury was used to induce corneal neovascularization and to evaluate the therapeutic effect
of topical suleparoide, while the involvement of bFGF in angiogenesis was evidenced by immuno-histochemistry of corneal tissue. Quantitation of angiogenesis in the CAM and the cornea wasaccomplished by image analysis.
Suleparoide dose-dependently inhibited HUV-EC-C cell proliferation (50% inhibitory concentration[IC
&!], 197±5³15±2 µg ml−") and reduced microvessel sprouting in vitro (IC
&!, 351³22 µg ml−").
Likewise, suleparoide 150 µg in agarose disks produced an avascular area of 19±7³2±7% of the total areaof the CAM (P!0±05 as compared to controls). bFGF levels were significantly enhanced in the corneaafter AgNO
$}KNO
$injury, and the increase appeared to be time-dependent (25±6³1±8 and 43±2³7±4%,
vs. uninjured controls after 24 hr and 48 hr, respectively, P!0±05). Suleparoide 4±8 mg eye−" day−" forsix days reduced the length of blood vessels and the area of the cornea infiltrated by them (59±6³7±4%decrease vs. controls, P!0±05).
These results demonstrate that suleparoide is an active agent against angiogenesis and suggest that thetherapeutic effect of the drug could be of value to treat corneal neovascularization.
# 1998 Academic PressKey words : angiogenesis ; in vitro models ; rat cornea; suleparoide, bFGF.
1. Introduction
Angiogenesis, the process of generating new blood
vessels, plays an important role in several diseases of
the eye, including neovascular glaucoma, diabetic
retinopathy, chemical burns and viral infections of the
cornea (Epstein et al., 1987), resulting in loss of visual
acuity. Angiogenesis is also thought to predispose to
the rejection of corneal allografts by facilitating the
exposure of antigens in donor cornea to the immune
system (Hill and Maske, 1988). The mechanisms of
corneal angiogenesis have been extensively investi-
gated, and various mediators appear to be involved in
the process, particularly basic fibroblast growth factor
(bFGF) (Adamis, Meklir and Joyce, 1991), vascular
endothelial growth factor (VEGF) (Amano et al.,
1998), prostaglandins (Ziche, Jones and Gullino,
* Address correspondence to: Mario Del Tacca, Department ofOncology, Division of Pharmacology and Chemotherapy, Universityof Pisa, Via Roma, 55, I-56126 Pisa, Italy.
1982), interleukin 2 and 8 (Lipman, Epstein and
Hendricks, 1992; Koch et al., 1992), and the platelet-
derived endothelial cell growth factor (PD-ECGF)
(Risau, 1990). The angiogenic potential of the
heparin-binding peptide bFGF has been demonstrated
in several experimental assays and suggested to be a
major factor for the induction of corneal angiogenesis
(Adamis et al., 1991). The release of bFGF from cells
may represent the signal for the development of a
capillary network, as reported for inflammatory
macrophages (Sunderkotter, Roth and Sorg, 1990).
Several anti-angiogenic agents have been charac-
terized in recent years in ocular animal models.
Angiostatic steroids (BenEzra et al., 1997) in com-
bination with heparin (Crum, Szabo and Folkman,
1985), sulfated polysaccharide-peptidoglycan complex
(Tanaka et al., 1991), plasminogen fragments (Murata,
Nakagawa and Takahashi, 1997), fumagillin ana-
logues (Ingber et al., 1990), thalidomide (Kenyon,
Browne and D’Amato, 1997) and cyclosporin A
(Benelli et al., 1997) exhibit anti-angiogenic activity
in corneal assays. Electrically charged molecules show
0014–4835}98}08013310 $30.00}0 # 1998 Academic Press
134 U. BENELLI ET AL.
the characteristics to bind and inactivate bFGF; in
particular, the sulphonic derivatives of dystamicin A
and the polyanionic compound suramin prevent the
binding of bFGF to its receptors in vitro and in vivo
(Ciomei et al., 1994), and inhibit neovascularization.
Suleparoide is a semisynthetic N-desulfated and re-
N-acetylated heparin derivative with an average
molecular weight of 9 kDa (Callas et al., 1993a).
Suleparoide has been introduced as an antithrombotic
agent, with weak anticoagulant effects primarily
mediated through the activation of the heparin
cofactor-II, but not antithrombin-III (Callas et al.,
1993b). The heparan sulfate suleparoide prevents
fibrinous membrane formation in the anterior chamber
of rabbits after human plasma injection with no
evidence of ocular toxicity (Lepri et al., 1996).
The aim of this study was to assess whether
suleparoide affects the growth of new blood vessels in
the rat cornea after chemical cauterization. Since the
most widely used drug combination to inhibit ex-
perimental corneal angiogenesis is heparin and
steroids, the effect of suleparoide was tested against
this treatment. To study the involvement of bFGF in
neovascularization and establish a possible correlation
with the mechanism of action of suleparoide, the
immunohistochemistry of bFGF in the rat cornea after
cauterization was investigated. Finally, the antiangio-
genic effects of suleparoide were tested on the human
endothelial cells HUV-EC-C, in the rat aortic ring assay
and the chorioallantoic membrane of chick embryo.
2. Materials and Methods
Chemicals, Drugs and Animals
The following reagents were used from the desig-
nated sources : heparin sodium salt grade II from
porcine intestinal mucosa (140 U mg−"), phosphate
buffered saline (PBS), hydrocortisone 21-phosphate,
and bovine serum albumin (BSA) (Sigma, St. Louis,
MO, U.S.A.) ; fetal bovine serum (FBS) and MCDB 131
medium (Gibco, Paisley, U.K.) ; bovine brain bFGF and
anti-bovine brain bFGF antibody (R&D System,
Minneapolis, MN, U.S.A.) ; unconjugated secondary
antibody and rabbit peroxidase-antiperoxidase (PAP)
complex (Dakopatts, Ballestrup, Denmark) ; 3,3«-di-
aminobenzidine tetrahydrochloride (DAB) (Fluka,
Buchs, Switzerland) ; gelatine (Kind & Knox INC.,
Sioux City, IA, U.S.A.). The MTT (3-[4,5-dimethyl-
thiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay
kit (Cell titer 964) was purchased from Promega
(Madison, WI, U.S.A.). The heparan sulfate suleparoide
(poly-hexuronyl-1,4--N-acetyl-(N-sulfonyl-) glucosa-
mine 6-O-sulfate sodium salt) was a generous gift of L.
Manetti-H. Roberts (Florence, Italy) and was dissolved
in NaCl 0±9% as vehicle. Female Wistar rats (body
weight, 200 g) were from Nossan (Comerio, Italy) and
were allowed unrestricted access to food and tap
water ; their care and handling were in accordance
with the recommendation of the European Economic
Community on animal experimentation.
Evaluation of HUV-EC-C Cell Proliferation
The human umbilical vein endothelial cell line
HUV-EC-C was purchased from the American Type
Culture Collection (Rockville, MD, U.S.A.). HUV-EC-C
cells were cultured in Ham’s F12 supplemented with
20% FBS, 10 ng ml−" bFGF, 4 m -glutamine,
100 IU ml−" penicillin, and 100 µg ml−" streptomycin
in 75 cm# flasks and incubated at 37°C under 5% CO#.
Briefly, cells (1¬10% well−") were plated in 96-well
plates (Costar, Cambridge, MA, U.S.A.) and incubated
for 2 days at 37°C in a final volume of 100 µl per well
of medium used for cell propagation. Cultures were
then incubated in medium containing suleparoide
(0±25–250 µg ml−") or vehicle alone for 24 hr. Cell
proliferation was estimated by the MTT assay by
reading the absorbance at 570 nm of the formazan
produced using a model 450 Microplate Reader (Bio-
Rad, Melville, NY, U.S.A.). The percent reduction of
cell proliferation, as compared to controls, was
expressed as the mean³.. of triplicate experiments.
Assay of Microvessel Growth from Rat Aortic Rings
The procedure described by Diglio et al. (1989) was
followed with minor modifications. Briefly, rats were
anesthetized by urethane (1 g kg−" i.p.) and a tho-
racotomy was performed. The thoracic aorta was
excised and, after removal of adventitia, 1 mm-long
rings were cut. The rings of aorta were positioned in
the center of each flat bottom well of a sterile 96-well
cluster plate (BioBraun, Milan, Italy), containing 60 µl
of serum-free MCDB 131 medium for 24 hr and then
supplemented with 2% FCS until the end of the
experiments. The explants were then treated at day 4
for 6 days with 100–800 µg ml−" of suleparoide or
drug vehicle and incubated at 37°C and 5% CO#. The
aortic rings were then removed and cell proliferation
was assessed by the MTT assay by reading the
absorbance at 570 nm. The percent reduction of cell
proliferation, as compared to controls, was expressed
as the mean³.. of three independent experiments.
Assay of Angiogenesis in the Chick Chorioallantoic
Membrane (CAM )
The effect of suleparoide on CAM angiogenesis, was
evaluated by a previously described method (Danesi et
al., 1997). Six-day old embryos were incubated at
37±5°C, 90% humidity and treated with suleparoide
(50–150 µg) included in 2±5% agarose disks (volume,
50 µl ; diameter, 5 mm) for sustained release ; control
disks contained drug vehicle. This dosage was chosen
in agreement with previously published data about
heparin and heparain sulfate molecules used in the
ANGIOGENESIS INHIBITION BY SULEPAROIDE 135
CAM (Hahnenberger and Jakobson, 1991). Each CAM
received one disk and embryos were returned to the
incubator for 24 hr. The procedure for assessing
vascularization has been described in detail elsewhere
(Danesi et al., 1993). Briefly, after removal of the
agarose disk, a 20% fat emulsion for intravenous
injection (Intralipid2, Pierrel, Milan, Italy) was injected
into the chorioallantois of cultured embryos to
enhance the red color of blood vessels against the
surrounding tissues. The CAMs were photographed
with a Polaroid model MP-4 camera system equipped
with 105 mm lenses. Each print of the same magni-
fication was digitized and subjected to image analysis.
Assay of Angiogenesis in Chemically Injured Rat
Corneas
Rats were deeply anesthetized with ether and both
eyes, as previously described by Proia et al. (1988) and
Culton et al. (1990), were cauterized by an AgNO$}
KNO$(75:25, w}v) applicator (Graham-field Surgical,
New Hyde Park, NY, U.S.A.) applied to the surface of
the cornea approximately 2 mm from the corneo-
scleral limbus (Scroggs et al., 1991). The peripheral
lesion was chosen to preserve vision and to minimize
discomfort in animals. Differences in the severity of
corneal injury were minimized by allowing one
investigator to perform the procedure. The applicator
was held in place for 7 s and excess AgNO$}KNO
$was
removed by gentle blotting with tissue paper. After
corneal cauterization, animalswere randomly assigned
to the control and treatment group, respectively, and
then returned to cages. Thirty minutes after chemical
lesion, five groups of ten animals each were given one
of the following treatments four times daily for six
days: suleparoide 4±8 mg eye−" day−", sodium heparin
3±4 mg eye−" day−", hydrocortisone 21-phosphate
0±21 mg eye−" day−", sodium heparin 3±4 mg eye−"
day−"hydrocortisone 21-phosphate 0±21 mg eye−"
day−", or drug vehicle (NaCl 0±9% and carboxy-
methylcellulose 2±5%). To define these treatments, in
preliminary experiments, dose adjustments were made
to find the doses of the drugs with optimum activity.
Two drops (20 µl each) were given a few seconds apart
by using a micropipette and animals were allowed to
blink between them. The concentration for each
eyedrop of suleparoide was 1±5%, while that for
sodium heparin was 1±1% and that for hydrocortisone
21-phosphate was 0±7%. Solutions were stored at 4°Cand were vortexed vigorously with sterile glass beads
just before applications. The effect of drugs on corneal
vascularization was also evaluated when the treat-
ment was started 6 days after chemical cauterization
and continued for 6 days thereafter. At the end of
treatment, rats were killed with i.p. 8% chloral hydrate
(280 mg kg−") ; the upper body was perfused with
50 ml of Ringer’s solution to completely remove blood
from vessels and then with a mixture of 6% gelatine
and 10% India ink, filtered with Supervelox2 filters
(Da Franceschi, Pisa, Italy) in Ringer’s solution. The
gelatine mixture within the corneal vessels was
solidified by freezing the eyes by dichlorodifluoro-
methane. Samples of cornea, including 1 mm rim of
adjacent scleral tissue, were removed from the rest of
the globe and fixed in 4% phosphate-buffered neutral
formaldehyde for 24 hr. Three radial cuts at full
thickness were made to allow flattening of the sample
and corneas were placed on a glass slide in mounting
media, magnified, and photographed.
bFGF Immunohistochemistry in the Cornea
The eyes of bilaterally cauterized animals were
enucleated 12 and 48 hr after the lesion was induced.
At each time point, the eyes of 3 injured animals and
3 uninjured controls were fixed by immersion in
freshly prepared 4% paraformaldehyde in 0±1
phosphate buffer (pH 7±4) for 10 hr (Cauchi et al.,
1996), dehydrated through graded concentrations of
ethanol and embedded in paraffin. The corneal tissue
was then cut in 7 µm sections and stained with
hematoxylin and eosin (H&E) or processed for
immunohistochemistry. The endogenous peroxidase
activity and the non-specific binding were quenched
with a 30-min treatment of 0±3% H#O#
in absolute
methanol and 3% FCS. The sections were incubated
overnight at 4°C with the anti-bFGF antibody
(1:1000), treated with the swine anti-rabbit antibody
(1:50) for 30 min, and then incubated for 30 min
with the rabbit PAP (1:100). Finally, sections were
exposed for 10 min to 0±5 mg ml−" of DAB containing
0±1% H#O#. All steps were separated by washes in
0±01 PBS, (pH 7±2)-0±3% Triton X-100. Anti-bFGF
antibodies were diluted in 0±01 PBS (pH 7±2)
containing 1% BSA. Controls included: [i] omission of
the first antibody, [ii] its replacement with preimmune
serum, or [iii] its preadsorption with excess of bFGF
before the incubation. Basal keratinocytes of rat
epidermis were chosen as positive controls and the
sections stained for bFGF were then subjected to image
analysis.
Image Analysis and Statistics
Photographs obtained from the CAM and corneal
angiogenesis assay were digitized in a 512¬512-pixel
matrix, using a video camera TK-1280E (JVC, Tokyo,
Japan) and pictures were visualized on a high
resolution color display (SAMPO, Tao-Yuan Hsien,
Taiwan). One cm# of the image array contained 1509
pixels and 270 different gray levels could be dis-
tinguished for each pixel. The image analysis software
package KS 300 v.1.2 (Kontron Elektronik GmbH,
Eching, Germany) was run for interactive manipu-
lation, quantification of the images and data collection.
CAM and corneal samples were analysed in a masked
fashion to minimize observer bias. Using the digitizer
136 U. BENELLI ET AL.
tablet, a line was drawn to delineate the total perimeter
of the CAMs and the treated areas showing reduction
in vascular network; their extension was then
calculated and averaged. In the corneal samples the
perimeter of cornea and of the vascular area of both
control and treated groups was delineated in the
digitized images by a freehand command; these areas
were then measured and the data obtained expressed
in mm#. Each eye was considered as an independent
variable. Where appropriate, the changes in HUV-EC-
C cell proliferation and in neovascularization of aortic
rings and of chemically-injured rat corneas induced by
treatments were expressed as percent values as
obtained by the following formula:
[(control value—treatment value)}control value]
¬100¯% decrease by treatment
Results are given as mean³.. of n experiments.
Statistical significance of differences was calculated by
the Student’s t-test and a P value less than 0±05 was
considered to be significant.
The sections stained for bFGF and including corneal
epithelium and stroma were analysed in a masked
fashion with the Quantimet 500 image analysis
system (Leica, Heerbrugg, Switzerland) based on true
color detection. The input device was a CCD JVC
videocamera mounted on a light microscope and
observation and video grabbing were carried out with
a 45¬ lens. After selecting the color range cor-
responding to the bFGF-positive areas, 30 fields for
each of the five sections were randomly chosen,
grabbed and stored on the computer hard disk. Each
field was then analysed for the presence of bFGF
immunoreactivity by automatically counting the
pixels of the digitized images corresponding to the FGF
color threshold and the results were expressed as
percentages of bFGF positive areas (µm#) with respect
to the total area of each field. The 50% inhibitory
concentration (IC&!
) of suleparoide on HUV-EC-C cells
and aortic sprouting was calculated by non-linear
least square fitting of the data using a computer
program (GraphPad PRISM4, San Diego, CA, U.S.A.).
3. Results
Inhibition of HUV-EC-C Cell Proliferation by
Suleparoide
A dose-dependent inhibition of cultured endothelial
cell proliferation was demonstrated after exposure to
the heparan sulfate suleparoide for 24 hr [Fig. 1(A)]. A
9±4³2±1% reduction of cell growth was obtained with
suleparoide 0±25 µg ml−" and further decreased to
55±7³1±5% at a concentration of 250 µg ml−" [P!0±05 vs. controls ; Fig. 1(A)]. The mean IC
&!of cell
proliferation, as calculated with a non-linear least
square fitting, was 197±5³15±2 µg ml−".
A
B
F. 1. Inhibition by suleparoide of human HUV-EC-C cellproliferation (A) and vascular cell growth in the rat aorticring explant assay (B). Results are the means of threeindependent experiments³.. (vertical bars). *P!0±05 vs.controls.
Suleparoide Inhibits Sprouting from Rat Aortic Rings
Aortic rings formed microvascular-like sprouts after
four days of culture ; the sprouts were composed of
vascular cells that extended radially from the explants
and grew on the surface of plastic wells. The vascular
cell population was composed of endothelial and
smooth muscle cells that proliferated rapidly during
four days after the explant and formed a monostrate
on the surface of the culture microwells, reaching the
plateau at the tenth day. Treatment with heparan
sulfate resulted in a concentration-dependent inhib-
ition of the proliferation rate of vascular cells ; their
growth in the presence of suleparoide 100 and
800 µg ml−" was reduced to 30±5³2±9% and
53±5³4±8% of controls, respectively [n¯3; P!0±05; Fig. 1(B)], with a calculated IC
&!of
351³22 µg ml−".
Suleparoide Inhibits Angiogenesis in the CAM
The area of the CAM below disks without suleparoide
did not show changes in vascular density, and a
normal branching pattern of blood vessels was present,
indicating that the disk weight did not affect their
growth [Fig. 2(A)]. After 24 hr of treatment, the
ANGIOGENESIS INHIBITION BY SULEPAROIDE 137
F. 2. Effect of suleparoide on angiogenesis in the CAM.An avascular area produced by treatment with suleparoide150 µg disk−" (black circle) is shown by arrowheads.
T I
Effect of agarose disks containing suleparoide on
neovascularization in the CAM of 6-day old embryos. Each
value represents the mean³S.E. of 20 CAMs.
Suleparoide(µg disk−")
Avascular area(cm# ; mean³..)
Percent of total CAM(mean³..)
0 0 050 0±52³0±06 2±3³0±26
100 3±62³0±5* 16³2±2*150 4±45³0±6* 19±7³2±7*
*P!0±05 vs. 50 µg disk−"
branching pattern of blood vessels below disks con-
taining suleparoide 50 µg had a minor decrease and
an avascular area of 0±52³0±06 cm# was produced,
corresponding to 2±3³0±26% of total area of CAM
(Table I). On the contrary, in the CAM around the
disks containing suleparoide 100 µg the tiny vessel
loops were absent and the avascular area was
3±62³0±5 cm#, corresponding to 16³2±2% of the
total area of the CAM (P!0±05 vs. controls ; Table I).
At the highest dose of 150 µg disk−", the area showing
reduced vascularity was 4±45³0±6 cm# (19±7³2±7%
of total CAM) [P!0±05 vs. controls ; Table I ; Fig.
2(B)]. The faint boundaries and capillaries of the CAM
disappeared after treatment, while the larger vessels
were reduced in caliber. Moreover, there were not
signs of thrombosis of the blood vessels, maybe due to
the well-known antithrombotic property of the hepa-
rin sulfate suleparoide and also the vasoconstriction
could be excluded because of the biological properties
of the drug.
Suleparoide Inhibits Angiogenesis in Chemically Injured
Rat Corneas
In the eyes of untreated rats numerous vessels were
usually seen invading the cornea by day 2 or 3, and
reaching the lesioned area within 5–6 days. In rats
treated with vehicle alone, a dense vascular network
extending from the corneoscleral limbus to the
cauterized site was observed after 6 days [Fig. 3(A),
Table II]. Treatment with suleparoide 4±8 mg eye−"
day−" induced a marked reduction in the length of
blood vessels in the neovascularized cornea [Fig. 3(B),
Table II] ; angiogenesis inhibition was also obtained
with the combination of heparin 3±4 mg eye−" day−"
hydrocortisone0±21 mg eye−" day−" [Fig. 3(D), Table
II], while heparin 3±4 mg eye−" day−" produced a
modest reduction of angiogenesis [Fig. 3(C), Table II].
Table II shows the area of the cornea occupied by
newly formed blood vessels, as calculated by image
analysis, and the degree of inhibition produced by
different treatments as compared to controls. Heparin
alone produced a modest inhibition of angiogenesis,
while its combination with hydrocortisone resulted to
be the most effective treatment (78±7³9±3% decrease,
P!0±05 vs. controls ; Table II). A significant inhib-
ition of neovascularization was obtained with the
heparan sulfate suleparoide (59±6³7±4% decrease,
P!0±05 vs. controls ; Table II).
In control animals, the neovascular areas obtained
6 and 12 days after the chemical lesion were not
significantly different. However, the pattern of vas-
cular reaction was different, since after 12 days the
blood vessels were fewer but larger and more radially
oriented. When the treatment was started 6 days after
chemical cauterization, no significant difference was
noted in the degree of corneal vascularization between
controls and animals treated with suleparoide (Table
III) ; however, hydrocortisone and heparinhydro-
cortisone produced a modest, yet significant reduction
of angiogenesis as compared to controls (Table III).
bFGF Immunohistochemistry in the Corneal Tissue
Uninjured corneal epithelium and stromal cells from
control rats showed a weak bFGF immunostaining
[Fig. 4(a), and the percent value of immunoreactive
areas was 1±2³0±3 (n¯12 corneas) as compared to
the total area of samples. The production of bFGF in
the corneal epithelium was significantly enhanced
after chemical cauterization [Fig. 4(b), (c)] and the
increase appeared to be time-dependent. Corneas
obtained 12 hr after the cautery displayed an intense
bFGF immunostaining (percent value of immuno-
138 U. BENELLI ET AL.
F. 3. Angiogenesis in the cornea as shown by the ink-filled corneal and pericorneal vasculature. (A) Cornea of a rat treatedwith drug vehicle showing numerous blood vessels. (B) Cornea of a rat treated with suleparoide for six days : the treatmentmarkedly reduced corneal neovascularization. (C) Cornea of a rat treated with heparin showing a modest reduction ofneovascularization. (D) Cornea of a rat treated with heparinhydrocortisone: a marked reduction of neovascularization isshown. * cautery site ; Co, cornea; S, sclera.
T II
Effects of suleparoide, heparin and hydrocortisone on corneal neovascularization in animals treated for 6 days, starting
30 min after chemical lesion. * P!0±05 vs. controls. §P!0±05 vs. suleparoide treatment. Each value represents the
mean³S.E. of 20 corneas from 10 rats
TreatmentVascular area
(mm# ; mean³..)Percent reduction
(mean³..)
Control (drug vehicle) 7±15³0±52 —Suleparoide 2±89³0±36* 59±6³7±4*Heparin 6±65³0±46 6±88³0±5Hydrocortisone 4±17³0±51* 41±57³5±1*Heparinhydrocortisone 1±52³0±18*§ 78±72³9±3*§
T III
Effects of suleparoide, heparin and hydrocortisone on corneal neovascularization in animals treated for 6 days, starting
6 days after chemical lesion. *P!0±05 vs. controls. Each value represents the mean³S.E. of 20 corneas from 10 rats
TreatmentVascular area
(mm# ; mean³..)Percent reduction
(mean³..)
Control (drug vehicle) 6±98³0±55 —Suleparoide 6±52³0±44 6±6³0±45Heparin 6±80³0±37 2±57³0±14Hydrocortisone 6±33³0±28* 9±28³0±41*Heparinhydrocortisone 5±98³0±35* 14±32³0±84*
ANGIOGENESIS INHIBITION BY SULEPAROIDE 139
F. 4. Rat corneas immunostained with anti-bFGF(¬530; A, B, C) or stained with H&E (¬530; D). (a)Uninjured control cornea is weakly immunostained; (b)12 hr after chemical lesion the corneal epithelium isimmunostained near the cautery site ; some cells in thecorneal stroma are also reacting; (c) 48 hr later a strongreaction is present in the corneal epithelium; exfoliation ofthe distal layers of the corneal epithelium are shown. (d)Neocapillaries (arrowheads) are demonstrated in the cornealstroma 6 days after cauterization.
reactive areas : 25±6³1±8, n¯12), which resulted in
being significantly higher than that observed in the
undamaged corneas (P!0±05 vs. controls). bFGF was
also found within some fibroblast-like cells in the
connective tissue of the cauterized cornea [Fig. 4(b)].
Forty-eight hours following corneal injury, the cautery
site was surrounded by damaged exfoliating cells and
the corneal epithelium showed a marked bFGF-positive
immunostaining (percent value of immunoreactive
areas 43±2³7±4, n¯12, P!0±05 vs. controls) [Fig.
4(c)]. When the anti-bFGF antibodies were replaced by
1% BSA-PBS, preimmune serum, or the antibodies
were preadsorbed with bFGF, no immunostaining was
observed; furthermore anti-bFGF antibodies, tested on
sections of rat epidermis, distinctly reacted with the
basal keratinocytes (data not shown). Furthermore, in
the corneal stroma of H&E-stained sections, small
capillaries were observed 6 days after cauterization
[Fig. 4(d)].
4. Discussion
Angiogenesis plays an important role in a variety of
normal and pathological conditions involving the eye
and has been the subject of an extensive investigation.
The growth of blood vessels is a complex process that
involves the interplay among a large number of
growth factors, including bFGF and VEGF, and the
endothelial cells, with stimulation of their prolifer-
ation, migration, and tubule formation (Cockerill,
Gamble and Vadas, 1995).
In the present study, topically applied suleparoide
reduced the neovascularization after corneal cauteri-
zation with AgNO$}KNO
$; furthermore, suleparoide
inhibited the HUV-EC-C endothelial cell proliferation
in vitro, the sprouting of rat aortic rings and the
angiogenesis in the CAM. The data presented in this
report provide evidence that suleparoide is an angio-
genesis inhibitor for potential topical use in the cornea.
Although heparan sulfate was a less potent inhibitor
of corneal angiogenesis as compared to the com-
bination of heparinhydrocortisone, suleparoide has
the advantage of lacking the side effects of cortico-
steroids on the cornea, and among them partial
adrenal suppression in adults, reduced resistance to
fungal, bacterial and viral infections, and thinning of
the cornea. Furthermore, repeated local administration
of corticosteroids may lead to a clinically important
increase of intraocular pressure, with possible damage
of optic nerve, and to a posterior subcapsular cataract
during long-term therapy (American Medical As-
sociation, 1995). Indeed, suleparoide is endowed of a
favorable toxicity profile in animal models (Lepri et al.,
1996), confirming the potential utilization in ocular
treatments for humans.
Previous studies reported conflicting data on the
effect of glycosaminoglycans on angiogenesis (Chiarugi
et al., 1986). Although heparin itself does not initiate
angiogenesis and inhibits the growth of capillary
endothelial cells (Crum et al., 1985), it appears to gain
stimulatory activity when bound to copper ions (Brem
et al., 1990). However, heparin becomes inhibitory in
the presence of glucocorticoids and a combined
heparin-steroid treatment inhibited neovascularization
140 U. BENELLI ET AL.
and induced the regression of the M5076 reticulum
cell sarcoma in mice (Folkman et al., 1983). In other
studies heparin alone inhibited endothelial cell pro-
liferation as well as collagen and collagenase synthesis,
while dexamethasone had no effects ; however, the
corticoid inhibited myofibroblastic cell proliferation
and induced dissolution of capillary basement mem-
brane (Ingber, Madri and Folkman, 1986). These data
reveal that the biologic effects of heparin and steroids
on neovascularization are distinct, but when given
together they show a potent antiangiogenic effect.
There is increasing evidence that bFGF may play a
central role in the early steps of angiogenesis (Rifkin
and Moscatelli, 1989). In the rat cornea bFGF was
detected in endothelial cells before the onset of
neovascularization: these cells may represent a source
of bFGF for the developing capillary network, as
reported for inflammatory macrophages (Baird, Mor-
mede and Bohlen, 1985). Even if a high degree of
immunostaining for bFGF was detected in the cornea,
the role of other factors, including TNF-α (Leibovich et
al., 1987), fibronectin and laminin, which are com-
ponents of the extracellular matrix with growth factor-
like repeats, is likely to be important in the early
neoangiogenic events (Ausprunk et al., 1991). Re-
cently, the heparin-binding growth factor VEGF was
identified as a functional endogenous corneal angio-
genic peptide. VEGF mRNA and protein were induced
to high levels after corneal injury and were temporally
and spatially correlated with inflammation and
neovascularization. The specific inhibition of VEGF
bioactivity with neutralizing antibodies suppressed
corneal neovascularization (Amano et al., 1998).
Nonetheless, the strong immunostaining for bFGF in
the endothelial cells around the cautery site of the rat
eyes suggests that also bFGF play an important role in
the regeneration of corneal epithelium, as also
reported in the mouse cornea (Sunderkotter et al.,
1990) and in studies in vitro with primary bovine
cultures and keratinocytes (Wooste et al., 1985).
Endogenous bFGF is implicated in the vascularization
of the chick embryo chorioallantoic membrane;
experimental evidences indicate that this growth
factor has an important role in the vascularization of
the CAM during chick embryogenesis (Ribatti et al.,
1995).
Suleparoide is a polyanionic polysulfated molecule
characterized by a strong negative charge; some
growth factors, including bFGF and transforming
growth factor β (TGF-β), are bound by polysulfated
molecules of the extracellular matrix (Ruoslahti and
Yamaguchi, 1991), and the heparan sulfate proteo-
glycans are involved in bFGF receptor binding (Saksela
et al., 1988). Yayon et al. (1991) showed that the
bFGF-receptor interaction requires prior binding to the
heparan sulfate side chains of the receptor itself. In this
light it could be possible that the excess of suleparoide
may alter the stoichiometry of the process and inhibit
the binding of bFGF to its receptor and its biologic
activity. Previous works demonstrated that the injury
to the corneal epithelium results in the release and
binding of bFGF to the heparan sulfate proteoglycan in
Bowman’s layer of the cornea (Adamis et al., 1991;
Soubrane et al., 1990). The results of the present study
indicate that damaged corneal epithelium actively
produces bFGF, whereas uninjured corneas contain
very low levels of the growth factor.
The inhibition of angiogenesis in the CAM, obtained
in the present experiments by suleparoide, is in
agreement with the previous report by Hahnenberger
and Jakobson (1991) who demonstrated the inhibitory
effect of sulfated and nonsulfated glycosaminoglycans
and polysaccharides on the normal outgrowth of
capillaries in the CAM. These findings support the
evidence that polysulfated molecules, including sule-
paroide, could be active against CAM neovasculari-
zation, by inhibiting the action of endogenous bFGF.
Furthermore, the IC&!
in HUV-EC-C cells treated with
suleparoide and cultured in a medium containing
bFGF, was markedly lower than that observed in rat
aortic rings cultured in serum-free medium and then
switched to 2%-FBS, where the amount of bFGF is
likely to be very low. These two models of bFGF-
dependent (HUV-EC-C cells) and -independent growth
(aortic explants) indirectly suggested that the inhib-
ition by suleparoide of vascular cell proliferation is
largely dependent on its activity on bFGF growth
stimulation of these cells, possibly by reducing the
interaction of the bFGF to its receptor. The lower
activity of suleparoide in vitro as opposed to the
marked effect in inhibiting blood vessel growth in vivo
suggests that the release by angiogenic tissue of a
number of heparin-binding growth factors, including
bFGF and VEGF, plays a determinant role in angio-
genesis.
The new capillary formation is initiated by local
degradation of vascular basement membrane in
response to angiogenic factors. Indeed, endothelial
cells secrete enzymes such as collagenase type IV and
heparanase (Kalebic et al., 1983). SCM-chitin III, a
homogeneous polysaccharide composed of N-acetyl-
glucosamine residues, reduces angiogenesis by in-
hibiting the enzyme activities released by endothelial
cells (Murata et al., 1991); a similar biologic effect
could be common to other polysulfated molecules,
including suleparoide. Moreover, it has been demon-
strated that suleparoide inhibits HUV-EC-C endothelial
cell growth in vitro and this effect may play a role in
the overall antiangiogenic activity of this drug. The
effect of suleparoide seems to be specific to endothelial
cells, since preliminary data from our laboratory
showed that suleparoide did not inhibit the pro-
liferation of normal epithelial cells (data not shown).
In conclusion, overall results indicate that sule-
paroide is an effective inhibitor of pathological angio-
genesis induced after chemical injury in the rat cornea,
and this effect is similar to that obtained by the
combination of heparin and hydrocortisone. These
ANGIOGENESIS INHIBITION BY SULEPAROIDE 141
data provide evidence and the rationale in favor of the
possible therapeutic use of suleparoide in ophthal-
mology, for the treatment of selected cases of corneal
neovascularization.
Acknowledgements
The experiments were performed with the technicalassistance and collaboration of Mr Bruno Stacchini. Thisstudy was supported in part by a grant from the ItalianAssociation for Cancer Research (AIRC, Milano, Italy).
References
Adamis, A. P., Meklir, B. and Joyce, N. C. (1991). In situinjury-induced release of basic-fibroblast growth factorfrom corneal endothelial cells. Am. J. Pathol. 139,961–6.
Amano, S., Rohan, R., Kuroki, M., Tolentino, M. andAdamis, A. P. (1998). Requirement for vascular en-dothelial growth factor in wound- and inflammation-related corneal neovascularization. Invest. Ophthalmol.Vis. Sci. 39, 18–22.
American Medical Association (1995). Ophthalmic anti-inflammatory agents, decongestants, and antiallergyagents. In Drug evaluations. Annual 1995 (Eds Bennet,D. R., Proudfit, C. M., Dickinson, B. D., Rodgers, B. J.,McVeigh, S., Meyer, M. and Schwartz, J. M.) Pp.2263–9. A.M.A.
Ausprunk, D. H., Dethlefsen, S. M. and Higgins, E. R. (1991).Distribution of fibronectin, laminin and types IVcollagen during development of blood vessels in thechick chorioallantoic membrane. In The development ofthe vascular system (Eds Feinberg, R. N., Sherer, G. K.and Auerbach, R.) Pp. 95–108. Basel : Issues Biomed,Karger.
Baird, A., Mormede, P. and Bohlen, P. (1985). Immuno-reactive fibroblast growth factor in cells of peritonealexudate suggest its identity with macrophage-derivedgrowth factor. Biochem. Biophys. Res. Commun. 126,358–64.
Benelli, U., Ross, J. R., Nardi, M. and Klintworth, G. K.(1997). Corneal neovascularization induced by xeno-grafts or chemical cautery. Inhibition by cyclosporin A.Invest. Ophthalmol. Vis. Sci. 38, 274–82.
BenEzra, D., Griffin, B. W., Maftzir, G., Sharif, N. A. andClark, A. F. (1997). Topical formulations of novelangiostatic steroids inhibit rabbit corneal neovasculari-zation. Invest. Ophthalmol. Vis. Sci. 38, 1954–62.
Brem, S. S., Zagzag, D., Tsanaclis, A. M. C., Gately, S.,Elkouby, M. P. and Brien, S. E. (1990). Inhibition ofangiogenesis and tumour growth in the brain. Sup-pression of endothelial cell turnover by penicillamineand the depletion of copper, an angiogenic cofactor.Am. J. Pathol. 137, 1121–8.
Callas, D. D., Ahsan, A., Iqbal, O. and Fareed, J. (1993a).Molecular weight and biochemical profile of achemically modified heparin derivative, suleparoide.Thromb. Res. 69, 369–76.
Callas, D. D., Hoppensteadt, D. A., Jeske, W., Iqbal, O.,Bacher, P., Ahsan, A. and Fareed, J. (1993b). Com-parative pharmacologic profile of a glycosaminoglycanmixture, Sulodexide, and a chemically modified heparinderivative, Suleparoide. Semin. Thromb. Hemost. 19Suppl. 1, 49–57.
Cauchi, J., Alcorn, D., Cancilla, B., Key, B., Berka, J. L.,Nurcombe, V., Ryan, G. B. and Bertram, J. F. (1996).
Light-microscopic immunolocalization of fibroblastgrowth factor-1 and -2 in adult rat kidney. Cell TissueRes. 285, 179–87.
Chiarugi, V., Ruggiero, M., Porciatti, F., Vannucchi, S. andZiche, M. (1986). Cooperation of heparin with otherangiogenetic effectors. Int. J. Tissue React. 8, 129–33.
Ciomei, M., Pastor, I. W., Mariani, M., Sola, F., Grandi, M.and Mongelli, N. (1994). New sulfonated distamycin Aderivatives with bFGF complexing activity. Biochem.Pharmacol. 47, 295–302.
Cockerill, G. W., Gamble, J. R. and Vadas, M. A. (1995).Angiogenesis : models and modulators. Int. Rev. Cytol.159, 113–60.
Crum, R., Szabo, S. and Folkman, J. (1985). A new class ofsteroids inhibits angiogenesis in the presence of heparinor a heparin fragment. Science 230, 1375–8.
Culton, M., Chandler, D. B., Proia, A. D., Hickingbotham, D.and Klintworth, G. K. (1990). The effect of oxygen oncorneal neovascularization. Invest. Ophthalmol. Vis. Sci.31, 1277–81.
Danesi, R., Del Bianchi, S., Soldani, P., Campagni, A., LaRocca, R. V., Myers, C. E., Paparelli, A. and Del Tacca,M. (1993). Suramin inhibits bFGF-induced endothelialcell proliferation and angiogenesis in the chick chorio-allantoic membrane. Br. J. Cancer 68, 932–8.
Danesi, R., Agen, C., Benelli, U., Di Paolo, A., Nardini, D.,Bocci, G., Basolo, F., Campagni, A. and Del Tacca, M.(1997). Inhibition of experimental angiogenesis by thesomatostatin analogue octreotide acetate (SMS 201–995). Clin. Cancer Res. 3, 265–72.
Diglio, C. A., Grammas, P., Giacomelli, F. and Wiener, J.(1989). Angiogenesis in rat aorta ring explant cultures.Lab. Invest. 60, 523–31.
Epstein, R. J., Stulting, R. D., Hendricks, R. L. and Harris,D. M. (1987). Corneal neovascularization: pathogenesisand inhibition. Cornea 6, 250–8.
Folkman, J., Langer, R., Linhardt, R. J., Haudenschild, C. andTaylor, S. (1983). Angiogenesis inhibition and tumorregression caused by heparin or a heparin fragment inthe presence of cortisone. Science 221, 719–25.
Hahnenberger, R. and Jakobson, A. M. (1991). Antiangio-genic effect of sulphated and nonsulphated glycos-aminoglycans and polysaccharides in the chick embryochorioallantoic membrane. Glycoconj. J. 8, 350–3.
Hill, J. C. and Maske, R. (1988). An animal model for cornealgraft rejection in high-risk keratoplasty. Transplantation46, 26–30.
Ingber, D., Fujita, T., Kishimoto, S., Sudo, K., Kanamaru, T.,Brem, H. and Folkman, J. (1990). Synthetic analoguesof fumagillin that inhibit angiogenesis and suppresstumour growth. Nature 348, 555–7.
Ingber, D., Madri, J. A. and Folkman, J. (1986). A possiblemechanism for inhibition of angiogenesis by angiostaticsteroids : induction of capillary basement membranedissolution. Endocrinology 119, 1768–75.
Kalebic, T., Garbisa, S., Glaser, B. and Liotta, L. A. (1983).Basement membrane collagen: degradation by mi-grating endothelial cells. Science 221, 281–4.
Kenyon, B. M., Browne, F. and D’Amato, R. J. (1997). Effectsof thalidomide and related metabolites in a mousecorneal model of neovascularization. Exp. Eye Res. 64,971–8.
Koch, A. E., Polverini, P. J., Kunkel, S. L., Harlow, L. A., DiPietro, L. A., Elner, V. M., Elner, S. G. and Strieter, R. M.(1992). Interleukin-8 as a macrophage-derived me-diator of angiogenesis. Science 258, 1798–801.
Leibovich, S. J., Polverini, P. J., Shepard, H. M., Wisemann,D. M., Shively, V. and Nuseir, N. (1987). Macrophage-induced angiogenesis is mediated by tumor necrosisfactor-α. Nature 329, 630–2.
142 U. BENELLI ET AL.
Lepri, A., Benelli, U., Chetoni, P., Perini, G., Saettone, M. F.andNardi,M. (1996). Effect of suleparoide on fibrinolysisin the anterior chamber of rabbits. Ophthalmic Res. 28,176–83.
Lipman, R. M., Epstein, R. J. and Hendricks, R. L. (1992).Suppression of corneal neovascularization with cyclo-sporine. Arch. Ophthalmol. 110, 405–7.
Murata, J., Saiki, I., Makabe, T., Tsuta, Y., Tokura, S. andAzuma, I. (1991). Inhibition of tumor-induced angio-genesis by sulfated chitin derivatives. Cancer Res. 51,22–6.
Murata, M., Nakagawa, M. and Takahashi, S. (1997).Inhibitory effects of plasminogen fragment on ex-perimentally induced neovascularization of rat corneas.Graefes Arch. Clin. Exp. Ophthalmol. 235, 584–9.
Proia, A. D., Chandler, D. B., Haynes, W. L., Smith, C. F.,Suvarnamani, C., Erkel, F. H. and Klintworth, G.(1988). Quantitation of corneal neovascularizationusing computerized image analysis. Lab. Invest. 58,473–9.
Ribatti, D., Urbinati, C., Nico, B., Rusnati, M., Roncali, L. andPresta, M. (1995). Endogenous basic fibroblast growthfactor is implicated in the vascularization of the chickembryo chorioallantoic membrane. Dev. Biol. 170,39–49.
Rifkin, D. B. and Moscatelli, D. (1989). Recent developmentsin the cell biology of basic fibroblast growth factor. J.Cell Biol. 109, 1–6.
Risau, W. (1990). Angiogenic growth factors. Prog. GrowthFactor Res. 2, 71–9.
Ruoslahti, E. and Yamaguchi, Y. (1991). Proteoglycans asmodulators of growth factor activities. Cell 64, 867–9.
Saksela, O., Moscatelli, D., Sommer, A. and Rifkin, D. B.
(1988). Endothelial cell-derived heparan sulfate bindsbasic fibroblast growth factor and protects it fromproteolytic degradation. J. Cell Biol. 107, 743–51.
Scroggs, M. W., Proia, A. D., Smith, C. F., Halperin, E. C. andKlintworth, G. K. (1991). The effect of total-bodyirradiation on corneal neovascularization in the Fischer344 rat after chemical cauterization. Invest. Ophthalmol.Vis. Sci. 32, 2105–11.
Soubrane, G., Jerdan, J., Karpouzas, I., Fayein, N. A., Glaser,B., Coscas, G., Courtois, Y. and Jeanny, J. C. (1990).Binding of basic fibroblast growth factor to normal andneovascularized rabbit cornea. Invest. Ophthalmol. Vis.Sci. 31, 323–33.
Sunderkotter, C., Roth, J. and Sorg, C. (1990). Immuno-histochemical detection of bFGF and TNF-α in thecourse of inflammatory angiogenesis in the mousecornea. Am. J. Pathol. 137, 511–15.
Tanaka, N. G., Sakamoto, N., Korenaga, H., Inoue, K.,Ogawa, H. and Osada, Y. (1991). The combination of abacterial polysaccharide and tamoxifen inhibits angio-genesis and tumour growth. Int. J. Radiat. Biol. 60,79–83.
Wooste, P., Brightwell, J., Eiferman, R. and Schultz, G.(1985). Effects of growth factors with dexamethasoneon healing of corneal stromal incision. Exp. Eye Res. 40,47–60.
Yayon, A., Klagsbrun, M., Esko, J. D., Leder, P. and Ornitz,D. M. (1991). Cell surface heparin like molecules arerequired for binding of basic fibroblast growth factor toits high affinity receptor. Cell 64, 841–8.
Ziche, M., Jones, J. and Gullino, P. M. (1982). Role ofprostaglandin E
"and copper in angiogenesis. J. Natl.
Cancer Inst. 69, 475–80.
