JOURNAL OF OCULAR PHARMACOLOGY AND THERAPEUTICS

Volume 23, Number 6, 2007

© Mary Ann Liebert, Inc.
DOI: 10.1089/jop.2007.0048

Larch Arabinogalactan for Dry Eye Protection and

Treatment of Corneal Lesions: Investigation in Rabbits

SUSI BURGALASSI,1 NADIA NICOSIA,1 DANIELA MONTI,1 GIULIA FALCONE,2

ENRICO BOLDRINI,2 and PATRIZIA CHETONI1

ABSTRACT

Purpose: The aim of the present study was to investigate the corneal protective and heal-
ing properties of arabinogalactan (AG), a natural polysaccharide present in conifers of the
genus Larix (Larch). AG was tested in comparison with other two polysaccharides possess-
ing well-established properties in the treatment of dry eye: tamarind seed polysaccharide and
hyaluronic acid.
Methods: The AG formulation was subjected to the following investigations: rheologic mea-
surements; evaluation of mucoadhesive properties by rheologic interaction with mucin; fer-
ning test; and in vivo evaluation on rabbits, including treatment of an experimental dry eye;
evaluation of the preocular retention; and evaluation of healing rate of experimental corneal
wound.
Results: AG dispersions showed a newtonian nonviscous behavior,   1.6 mPa  s for 10%
w/w solution; it possessed good mucoadhesive properties useful for retention on the eye sur-
face. In fact, a prolonged time of residence in rabbit eyes was ascertained using fluorescein-
labeled AG. Five percent (5.0%) w/w AG exerted a good protective effect against the appear-
ance of corneal dry spots. It also reduced significantly the healing time of an experimental
corneal lesion since 27 h after the first treatment.
Conclusions: These findings suggest that AG may be a potential therapeutic product for dry
eye protection and for the treatment of corneal wounds.

INTRODUCTION source of corneal abrasions: since approximately

C ORNEAL ABRASIONS RESULT from cutting,
scratching, or abrading the thin, protective,
clear coat of the exposed anterior portion of the
ocular epithelium. These injuries cause pain, tear-
ing, photophobia, foreign body sensation, and a
gritty feeling.1 Abrasions may be caused by many
factors, both toxic and mechanical, and by altered
physiologic functions. Contact lenses, both rigid
and soft, are a very specific and not uncommon
100 million people wear contact lenses world-
wide and this number is rapidly increasing,2,3
their risk should not be underestimated. One of
the first events in tissue repair is cell attachment
and adhesion to specific substrates of the extra-
cellular matrix, through recognition of specific
membrane receptors called integrins. We sug-
gested in a previous paper that, since integrin
recognition can be affected by natural or synthetic
polysaccharides, both polysaccharides and gly-

1Department of Bioorganic Chemistry and Biopharmaceutics, University of Pisa, Pisa, Italy.

2Opocrin SpA, Corlo di Formigine (MO), Italy.

541
542
cosaminoglycans may positively influence cell
adhesion and the wound-healing process.4
The continuous search for compounds favor-
ing cell adhesion and promoting ocular wound
healing prompted us to investigate a natural
polysaccharide, arabinogalactan (AG), which is
particularly abundant in plants of the Larix genus,
primarily Larix occidentalis (Western Larch). Larch
AG is approved by the Food and Drug Adminis-
tration (FDA) as a source of dietary fiber, but has
also therapeutic benefits as an immune-stimulat-
ing agent.5
The aim of the present preliminary study, car-
ried out on rabbits, was to test the eye tolerance
of AG and its protective action against experi-
mentally induced dry eye conditions and corneal
abrasions.
METHODS
Products
AG (FiberAid®AG) was purchased from
LAREX Inc., (St. Paul, MN); mannitol from Carlo
Erba (Milan, Italy); tamarind seed polysaccharide
(TSP; TSP® eyedrops) were from Farmigea S.p.A.
(Pisa, Italy); hyaluronic acid, (HA) was from
Chemofin (Milan, Italy); gastric mucin from hog
stomach (HGM), was from (TCI) Tokyo Kasei,
Ltd., (Tokyo, Japan); atropine sulfate solution
(AS; atropine 1% monodose) was from Farmigea
S.p.A., (Pisa, Italy); and fluorescein, isothio-
cyanatofluorescein, dibutyltin dilaurate, and 1-
heptanol were from Sigma-Aldrich (Milan, Italy).
All other chemicals, solvents, etc., were of an-
alytical grade.
Test formulations
AG dispersions for rheologic studies were ob-
tained by slowly adding the polymer to depu-
TABLE 1. COMPOSITION AND PROPERTIES OF THE
Polymer Mannitol
Formulation (% w/w) (% w/w)
AG-Sol 5.00000
TSP® 0.50000
HA 0.2% 0.20000
TSP 0.04% 0.04000
HA 0.00144% 0.00144
AG, arabinogalactan; TSP, tamarind seed polysaccharide; HA, hyaluronic acid.
BURGALASSI ET AL.
rated water and heating for 30 min at 80°C. TSP
and HA dispersions were prepared by gentle stir-
ring at room temperature.
Formulations for in vivo tests. A 5.0% AG dis-
persion (AG-Sol) was made isotonic by adding
4.4% w/w mannitol, whereas 0.5% TSP commer-
cial eyedrops, containing tamarind seed polysac-
charide and mannitol for isotonicity, were used
as such or diluted with mannitol isotonic solution
to 0.04% w/w. The 0.2% or 0.00144% w/w HA
formulations contained 5.0% w/w mannitol for
isotonicity. All formulations were sterilized by fil-
tration (0.22 m, Minisart; Sartorius SpA, Flo-
rence, Italy) and did not contain preservatives,
since they were used immediately after opening
the vials.
The properties of the formulations are listed in
Table 1.
Rheologic measurements
These tests were performed at 25°C on
0.2%–10% w/w AG dispersions using a Rheostress
RS 150 apparatus (Haake, Paramus, NJ) equipped
with coaxial cylinders (Z40 and Z41) at shear rates
ranging from 0 to 500 sec1.
Ferning test
This test was performed by mixing 10.0 L of
a 2.5% w/w AG dispersion with 2.0 L of artifi-
cial lachrymal fluid. The mixture was smeared on
a clean microscope slide and allowed to dry at
room temperature (25  1°C). The artificial
lachrymal fluid had the following composition
(in mg/100 mL of distilled water): MgCl2, 4.75;
CaCl2, 7.97; KHCO3, 260.00; and NaCl, 754.00.6
The specimens were examined and pho-
tographed with a light microscope (Axioskop,
Zeiss, Arese, Italy) equipped with digital photo
FORMULATIONS FOR IN VIVO STUDIES
Viscosity
pH (, mPa  s)
4.4 6.5 1.21
5.0 6.4 9.16
5.0 6.5 24.40
5.0 6.3 1.22
5.0 6.6 1.26
ARABINOGALACTAN FOR EYE PROTECTION
camera (DC100, Leica, Solms, Germany), and hu-
man lachrymal fluid was used as the control.
Rheological assessment of mucin-polymers
adhesive bond strength
The method of Hassan and Gallo,7 consisting
of the measurement of viscosity changes induced
in a mucin dispersion by the addition of the poly-
mers, was essentially followed. According to this
method, the viscosity component due to bioad-
hesion, b, is calculated from the equation b 
t  m  p, where t, m, and p are the indi-
vidual viscosity coefficients of the system, of
mucin and of the polymer, respectively. Then, the
normalized parameter “adhesion index” (AI) 
b/p was calculated. Such a normalization en-
abled polymer dispersions characterized by dif-
ferent viscosity properties to be compared.8 The
viscosity measurements were carried out using a
Rheostress RS 150 apparatus equipped with coax-
ial cylinders (Z40 and Z41) at shear rates ranging
from 0 to 500 sec1 at 32°C. The measurements
were carried out on solution containing: (i) 15%
w/w mucin alone (m); (ii) the polymer at the
usual concentration (p); and (iii) mixtures of
mucin with the polymers, at the same concentra-
tions indicated before. AG and TSP dispersions
exhibited a Newtonian behavior; the dispersions
based on HA, mucin, and mucin/polymer ex-
hibited a pseudoplastic flow, and their apparent
viscosity, , was calculated for D  200 s1 from
plots of  versus D, according to the equation  
aDb, where a and b are experimentally deter-
mined values.
Fluorescein isothiocyanate AG synthesis
Fluorescein isothiocyanate–labeled AG (FITC-
AG) was synthesized as follows: AG (1 g) was
dissolved in methyl sulfoxide (10 mL) containing
a few drops of pyridine. Isothiocyanatofluores-
cein (0.1 g) was added, followed by dibutyltin di-
laurate (20 mg), and the mixture was heated for
2 h at 95°C. After several precipitations in ethanol
to remove free dye, FITC-AG was filtered off and
dried at 80°C.9
FITC-AG solutions were prepared by heating
for 30 min at 80°C while gently stirring.
Animals
Seventy-six (76) male albino New Zealand rab-
bits, weighing 2.5–3.0 kg (Pampaloni Rabbitry,
543
Fauglia, Italy) were used and treated as pre-
scribed in the publication “Guide for the Care and
Use of Laboratory Animals” (National Institutes
of Health Publication No. 92–93, revised 1985).
All experiments conformed with the ARVO Res-
olution on the Use of Animals in Research: They
were carried out under veterinary supervision,
after approval of the protocols by the Ethical-Sci-
entific Committee of the University of Pisa (Pisa,
Italy).
The animals were housed singly in standard
cages in a room with controlled lighting at 19 
1°C and 50%  5% R.H. (relative humidity), with
no restriction of food or water. During the ex-
periments, the rabbits were placed in restraining
boxes, to which they had been habituated, in a
room with dim lighting; they were allowed to
move their heads freely, and their eye movements
were not restricted.
Induction and treatment of dry eye conditions
The animals were preliminarily submitted to
the Schirmer I test and to a slit-lamp examination
of the corneal surface to verify the integrity of the
corneal epithelium and the function of lachrymal
system. Then, 12 animals were treated as reported
by Burgalassi and colleagues.6 Briefly, all animals
received in the lower conjunctival sac of both eyes
50 L of 1% AS at 9.00 AM, 1.00 PM, and 5.00 PM.
Five (5) min after each administration of AS, they
received an eye 50 L of the formulation under
study, AG-Sol, while the polymer-free vehicle
was administered in the control eye. All treat-
ments were discontinued after 5 days.
The Schirmer I test was performed 2, 3, 4, and
5 days after the first administration of AS, at
10.00 AM. The test was performed on both eyes
(nonanesthetized) of all animals by maintaining
for 3 min a standardized test strip (Alfa Intes,
Casoria, Italy) into the external third of the
lower conjunctival fornix. The wetted length, in
millimeters, of the strip was taken as the test
score.
After staining with fluorescein (Fluorets; Smith
& Nephew Pharmaceuticals Ltd., Romford, UK),
the corneal surface was observed with a slit-lamp
biomicroscope fitted with a blue filter. The test
was performed at 2.00 PM at 3, 4, and 5 days af-
ter the first administration of AS. The occurrence
of dotted staining, revealing the presence of dry
spots on the ocular surface, was considered as a
symptom of corneal desiccation.
544
Evaluation of the preocular retention of the
FITC-AG solution
The retention time of the polymer in rabbit eyes
was assessed by applying 50 L of 5.0% w/w
FITC-AG solution into the lower conjunctival sac
of both eyes of 8 animals; the eyelids of the ani-
mals were then gently kept closed for 30 sec. As
a reference, 1.0% w/w fluorescein aqueous solu-
tion (Fluo), presenting the same fluorescence in-
tensity at 5.0% FITC-AG, was administered in the
same way to 8 animals. At 1, 3, 5, 10, 20, 30, 45,
and 60 min after administration, tear fluid sam-
ples were collected from the lower marginal tear
strip using 1.0 L disposable glass capillaries
(Drummond “Microcaps”; Fisher Scientific, St.
Louis, MO). The tear fluid was transferred into
microtubes, and the capillaries were flushed with
distilled water. After dilution with distilled wa-
ter (50 L), the fluorescence of FITC-AG in the
samples was measured at 514 nm (excitation 490
nm; RF-551 Fluorometer; Shimadzu, Kyoto,
Japan).
Induction and treatment of an experimental
corneal lesion
The rabbits were divided into six groups, each
consisting of 8 animals. The corneal surface was
preliminarily examined by slit lamp to verify the
integrity of the corneal epithelium. All animals
were anesthetized by an intramuscular (i.m.) in-
jection of 0.15 mL/kg of Zoletil® 100 (Laboratories
Virdac, Carros, France). The right eye was then
kept open with a blepharostat and anesthetized
with 10 L of oxybuprocaine hydrochloride
(Novesina®, MiPharm, Milano, Italy). The corneal
epithelium was removed by applying, for 1 min,
a paper disc (diameter 6 mm; Whatman No. 50 fil-
ter paper, Maidstone, UK) shaped like a small con-
tact lens and soaked with 10 L of n-heptanol.10
The eyes were then carefully rinsed with normal
saline, and only the damaged eye of each animal
of groups 1–5 was treated three times daily with
50 L of the formulations under test, while the
sixth group was used as the control and treated
with 50 L of normal saline solution.
The treated groups 1–5 received the following
formulations: AG-Sol; 0.5% TSP; 0.04% TSP; 0.2%
HA; and 0.00144% HA.
Immediately after producing the epithelial
damage and before each daily treatment, the right
BURGALASSI ET AL.
eye was stained with 10 L of fluorescein (1.0%
w/w in water) to visualize the damaged area. The
diameter of corneal wounds was measured using
a slit-lamp biomicroscope fitted with a blue filter
and a micrometer.4
Statistical data analysis
The variability of individual observations pre-
sented throughout this paper were reported as
the standard deviation of the mean.
The statistical significance of differences be-
tween means for Schirmer I test data and for
corneal epithelial defect area, were assessed by
one way analysis of variance (ANOVA; StatView
software, Abacus Concepts, Inc., Berkeley, CA).
When statistical differences were detected, mul-
tiple comparisons were made among the data cor-
responding to various measuring times using
Fisher’s Protected Least Significant Difference
(PLSD) test. In the relevant figures, the symbols
(* and §) indicate a significant difference at the
P  0.05 level.
The statistical significance for the corneal ep-
ithelium integrity observations was tested by a con-
tingency table analysis (Total chi-square, StatView
software, Abacus Software, Berkeley, CA).
RESULTS
Characteristics of the AG dispersion
The viscosity (mPa.s) versus concentration (%
w/w) profile of AG dispersions is reported in Fig-
ure 1. All tested AG dispersions exhibited a New-
tonian behavior.
FIG. 1. Rheological properties of a 5.0% arabinogalac-
tan dispersion.
ARABINOGALACTAN FOR EYE PROTECTION 545

The ferning test showed that AG crystallized

forming type I fern-like structures similar to those
produced by human lachrymal fluid (Fig. 2A–2B).

Rheological assessment of mucin-polymers

adhesive bond strength
The results of the rheologic measurements are
summarized in Table 2. The polymeric solutions
showed a wide viscosity range, from 1.38 mPa.s
of AG to 24.40 mPa.s of HA, but an inspection of
the AI values in Table 2 shows that AG interacts
with mucin “in vitro” more than the other tested
polymers.

A Corneal epithelium dryness

B with statistically significant differences on the
FIG. 2. Ferning test: (A) A 2.5% arabinogalactan dis-
persion and (B) human lachrymal fluid (scale bar  50
m).
The Schirmer test scores (reported as millime-
ters of wet strip 3 min after insertion) obtained
before (basal values) and after (dry eye) treatment
with AS, and relevant to the treatment with the
formulation under test, are reported in Fig. 3A.
A significantly decreased tear production (P 
0.05, Fisher’s PLSD test) was observed since the
third day after the beginning of AS treatment: at
this time, the average Schirmer test score in dry
eyes was reduced from 14.60  2.99 mm (basal
value) to 11.33  4.85 mm. AG-Sol-treated eyes
showed greater scores with respect to dry eyes at
all experimental times, with values ranging be-
tween 15.08  5.09 and 16.20  3.79 mm, and
3rd, 4th, and 5th days of treatment (P-levels 
0.0031, 0.0027, and 0.014, respectively, on the
Fisher’s PLSD test).
The results of the slit-lamp examination of the
fluorescein-stained corneas, expressed as the per-

TABLE 2. RHEOLOGIC ASSESSMENT OF MUCIN-POLYMERS ADHESIVE BOND STRENGTH


HGM AG TSP HA (m, p, t) b AI
Formulation (% w/w) (% w/w) (% w/w) (% w/w) (mPa  s) (mPa  s) (b/p)

1 15.0 — — — 47.97 — —
2 — 5.0 — — 1.38 — —
3 — — 0.5 — 9.16 — —
4 — — — 0.2 24.40 — —
5 15.0 5.0 — — 73.10 23.74 17.21
6 15.0 — 0.5 — 190.41 133.27 14.55
7 15.0 — — 0.2 145.97 73.59 3.02

HGM, hog gastric mucin; AG, arbinogalactan; TSP, tamarind seed polysaccharide; HA, hyaluronic acid; m, p, t,
b, viscosity coefficients of mucin, polymer, the mucin-polymer system, and bioadhesive component; AI, adhesion
index.
546
A peared from 30 min for Fluo.
B TSP and 0.2% HA in enhancing corneal epithe-
FIG. 3. Corneal epithelium dryness. (A) Schirmer test
scores obtained with arabinogalactan (AG)-Sol formula-
tion in the rabbit dry eye model (n  12). *Significantly
different from control (P  0.05, Fisher’s Protected Least
Significant Difference test). (B) Percentage of fluorescein-
positive eyes in control (“dry”) eyes and in eyes treated
with the AG-Sol formulation (n  12). P-levels (total chi-
square) by contingency table analysis.
centage of eyes showing dotted staining, are il-
lustrated in Figure 3B, reporting the data ob-
tained on the 3rd, 4th, and 5th days of treatment.
The values corresponding to “dry” eyes (i.e.,
treated with AS alone) were 33.33%, 50.00%, and
58.33%, respectively. In the case of AG-Sol, at all
observation times, a lower percentage (ranging
from 8.33% to 16.67%) of stained eyes with re-
spect to dry eyes was observed, even if it was sig-
nificantly different (P  0.05, total chi-square)
only on the fifth day.
Evaluation of the time of residence of the
formulation in rabbit eyes
The behavior in the rabbit eyes of the aqueous
solutions containing 5.0% w/w FITC-AG and
1.0% w/w Fluo is illustrated in Figure 4 as the
percentage of the fluorescence at 1 min in tear
fluid (mean  standard deviation [SD]) vs. time.
Observation of the eye surface by slit-lamp bio-
BURGALASSI ET AL.
microscope fitted with a blue filter evidenced a
fluorescent corneal film lasting 20 min for both
solutions. The quantitative analysis showed flu-
orescence in tear fluid decreased very rapidly af-
ter administration, even if was still quite high af-
ter 10 min: 14.49% and 9.74% for FITC-AG and
Fluo, respectively. The fluorescence produced by
FITC-AG was still detectable 60 min after ad-
ministration (0.74%  0.13%), while it disap-
Experimental corneal lesion in rabbits
The produced corneal lesions were very regu-
lar in shape and had a constant area of 28.27 mm2,
possibly because of the small amount of n-hep-
tanol (10 L) used to soak the paper disc: A larger
amount would spread onto the cornea, produc-
ing irregularly shaped lesions. The mean epithe-
lial defect areas for each treatment group are re-
ported in Figures 5A and 5B as the percentage of
the initially damaged area (mean  SD) versus
time. AG-Sol showed the same efficiency as 0.5%
lium wound healing. When TSP and HA were
tested at the same viscosity of AG-Sol, some dif-
ferences appeared. The AG-Sol and 0.04% TSP
eyedrop groups produced a small decrease in the
damaged area 24 h after the first treatment with
respect to the control group. However, as shown
by statistical analysis, only AG-Sol produced sig-
nificantly different results (P  0.05; Fisher’s
PLSD test) from the controls at 27, 29, 31, 34, and
41 h after the first treatment, and 0.04% TSP was
FIG. 4. Evaluation of the time of residence of floures-
cein isothiocyanate–labeled arabinogalactan and floures-
cein aqueous solutions in rabbit eyes (n  8).
ARABINOGALACTAN FOR EYE PROTECTION
A a Newtonian rheologic behavior and viscosity
B viscous formulations renders administration
FIG. 5. (A–B). Reduction of the corneal defect area in
rabbits after different treatments (n  8). *arabinogalac-
tan (AG) and § tamarind seed polysaccharide (TSP) sig-
nificantly different from control (P  0.05, Fisher’s Pro-
tected Least Significant Difference test). HA, hyaluronic
acid.
significantly different only at 29 h. No time points
were significantly different for the 0.00144% HA
formulation.
DISCUSSION
AGs are a class of long, densely branched poly-
saccharides with molecular weights ranging from
10,000 to 120,000 Da. As mentioned in the Intro-
duction, AGs are FDA-approved for use in food
applications and are orally safe, even in large
doses. In nature, AGs are found in several mi-
crobial systems, where they are complexed be-
tween peptidoglycans and mycolic acids as com-
ponents of the cell wall. Although many edible
and inedible plants are rich sources of AGs,
mostly in glycoprotein form and bound to a pro-
tein spine of either threonine, proline, or serine,
they are most abundant in the inner bark of
conifers of the genus Larix (larch). All AGs iso-
lated from larch trees are water-soluble, nitrogen-
free, highly branched polysaccharides of the 3,6-
547
beta-D-galactan type. Side chains consist of com-
binations of single galactose sugars, as well as
longer side chains comprised of both beta-galac-
tose and beta-arabinose residues in a molar ratio
of approximately 6:1.5 A molecular modeling
study has revealed that the resulting triple helix,
applicable to Western larch AG, can assume quite
different morphologies, since the side group has
access to several allowed conformational states.11
All tested AG dispersions (0.2 to 10%) showed
values ranging from 1.00 to 1.58 mPa.s. The vis-
cosity of tear fluid is reported to be 1.02–1.93
mPa  s,12 and viscosizers are commonly added
to ophthalmic formulations to lengthen their pre-
ocular residence time. The maximum allowable
viscosity of ophthalmic solutions has been indi-
cated to be in the range 15–30 mPa  s,13 but in
the case of mucoadhesive polymers, lower values
may produce the same results, as evidenced by
Saettone and colleagues14 and by Oechsner and
Keipert.15 Moreover, the application of highly
more difficult and may cause ocular discomfort,
usually resulting in a reduced patient compli-
ance.16
An interesting physical characteristic of tears,
when allowed to dry on a microscope slide at
room temperature, is their ability to crystallize
forming fern-like structures.17 Rolando18 classi-
fied tear-film crystallization (i.e., ferning or ar-
borization) in four types according to their ap-
pearance. Basically, type I consists of large,
homogeneous ferns; in type II, fern patterns are
smaller and sparsely distributed; in type III, ar-
borisation is scarcely present; while it is totally
absent in type IV. An attempt was subsequently
made by Filippello and colleagues19 to apply the
ferning test to the evaluation of polymeric artifi-
cial tear substitutes: The assumption was made
that polymers showing the best muco-mimetic
properties would crystallize in fern-like patterns,
analogous to those produced by lachrymal mu-
cus itself. Such properties are apparent in AG so-
lutions.
Several methods have been employed to in-
vestigate mucin-polymer interactions and to
classify polymers as mucoadhesives. Rheologic
interaction to evidence the mucoadhesive prop-
erties of polymeric substances is a time-honored
technique in the ophthalmic field,14,20,21 even if
spectroscopic methods have been recently em-
ployed to characterize the mucoadhesive-mucus
548
interface.22–24 Our rheologic results indicated
strong interactions between AG and mucin, of the
same order as those shown by TSP and about
fivefold greater than HA, in spite of the nonvis-
cous properties of AG.
Inclusion of a fluorescent marker in polymeric
ophthalmic vehicles is a well-known technique
for investigations on the ocular retention time,
and on the influence of polymers on the thickness
of precorneal tear film.25 These factors are con-
sidered of great importance in ophthalmic vehi-
cles, since they can ensure a long contact time
with the ocular surface and a high and constant
drug concentration in the tear film, and hence, a
high ocular bioavailability.26 Gamma scinti-
graphic studies by Wilson and his associates on
the residence time of polymers in the eye27,28
demonstrated that the mean half-time for the
clearance of saline solution was 22 sec, and was
significantly increased by the addition of poly-
mers: half-times were 1089 and 89 sec for 0.6%
w/w gellan gum and 0.5% w/w hydroxyethyl-
cellulose, respectively, and retention times
reached 35 and 116 min after the application of
Gel Tears (a commercially available gel based on
carbomer 940), TSP, and HA solutions at differ-
ent concentrations. Our study showed that AG-
Sol was present in rabbits’ lachrymal fluid for at
least 60 min after administration.
Many studies on corneal wound healing, using
different methods to produce lesions, have been
published.10,29–31 The heptanol model used in this
study10 only produces removal of the surface ep-
ithelium, without damage either to the basement
membrane or to the extracellular stromal com-
ponents. We found that AG significantly in-
creases corneal wound healing rate 27 h after the
first treatment.
CONCLUSIONS
The present preliminary results indicate that
AG, a natural polysaccharide producing New-
tonian, nonviscous solutions, shows mucoadhe-
sive properties useful for retention on the eye sur-
face. Five percent (5%) w/w AG formulations,
tested on a dry eye model, exerted a protective
effect against the appearance of dry spots on the
corneal epithelium, and significantly increased
the healing rate of corneal wounds, with respect
to other polymers commonly used as adjuvants
in ophthalmic vehicles.
BURGALASSI ET AL.
Though, for the time being, they are demon-
strated only on rabbits, these findings suggest the
potentiality of AG in dry eye conditions, and for
the prevention and treatment of corneal wounds.
AG solutions might particularly benefit contact
lens wearers because of their tolerability, pro-
longed permanence, and noninterference with vi-
sion owing to low viscosity. Further research
meant to verify the compatibility of AG with con-
tact lens and microscopy studies, aimed at a more
thorough survey of its involvement at tissue level,
are in progress.
Finally, clinical trials will definitely need to
verify the transferability of these results to hu-
mans.
ACKNOWLEDGMENTS
The authors wish to thank Professor M. F. Saet-
tone for useful discussions and Dr. E. Napolitano
for assistance in the synthesis of FITC-AG.
REFERENCES
1. Wilson, S.A., and Last, A. Management of corneal
abrasions. Am. Fam. Physician. 70:123–128, 2004.
2. Jalbert, I., and Stapleton, F. The corneal stroma dur-
ing contact lens wear. Contact Lens Ant. Eye 28:3–12,
2005.
3. Casser, L., and Lingel, N.J. Diseases of the cornea. In:
Bartlett, J.D., and Jaanus, S.D., eds. Clinical Ocular
Pharmacology. Boston: Butterworth-Heinemann; 1995:
679–745.
4. Burgalassi, S., Raimondi, L., Pirisino, R., et al. Effect
of xyloglucan (tamarind seed polysaccharide) on con-
junctival cell adhesion to laminin and on corneal ep-
ithelium wound healing. Eur. J. Ophthalmol. 10:71–76,
2000.
5. Gregory, S., and Kelly, N.D. Larch arabinogalactan:
Clinical relevance of a novel immune-enhancing poly-
saccharide. Altern. Med. Rev. 4:96–103, 1999.
6. Burgalassi, S., Panichi, L., Chetoni, P., et al. Develop-
ment of a simple dry eye model in the albino rabbit
and evaluation of some tear substitutes. Ophthalmic
Res. 31:229–235, 1999.
7. Hassan, E.E., and Gallo, J.M. A simple rheological
method for the in vitro assessment of mucin-polymer
bioadhesive bond strength. Pharm. Res. 7:491–495,
1990.
8. Sandri, G., Rossi, S., Ferrari, F., et al. Assessment of
chitosan derivatives as buccal and vaginal penetra-
tion enhancers. Eur. J. Pharm. Sci. 21:351–359, 2004.
9. De Belder, A.N., and Granath, K. Preparation and
properties of fluorescein-labeled dextrans. Carb. Res.
30:375–378, 1973.
ARABINOGALACTAN FOR EYE PROTECTION
10. Cintron, C., Hassinger, L., Kublin, C.L., et al. A sim-
ple method for the removal of rabbit corneal epithe-
lium utilizing n-heptanol. Ophthalmic Res. 11:90–96,
1979.
11. Chandrasekaran, R., and Janaswamy, S. Morphology
of Western larch arabinogalactan. Carb. Res. 337:
2211–2222, 2002.
12. Lemp, M.A., Holly, F.J., Iwata, S., et al. The precorneal
tear film. I. Factors in spreading and maintaining a
continuous tear film over the corneal surface. Arch.
Ophthalmol. 83:89–94, 1970.
13. Chrai, S.S., and Robinson, J.R. Ocular evaluation of
methylcellulose vehicle in albino rabbits. J. Pharm. Sci.
63:1218–1223, 1974.
14. Saettone, M.F., Monti, D., Torracca, M.T., et al. Mu-
coadhesive ophthalmic vehicles: Evaluation of poly-
meric low-viscosity formulations. J. Ocul. Pharmacol.
Ther. 10:83–92, 1994.
15. Oechsner, M., and Keipert, S. Polyacrylic acid/
polyvinylpyrrolidone bipolymeric systems. I. Rhel-
ogical and mucoadhesive properties of formulations
potentially useful for the treatment of dry-eye syn-
drome. Eur. J. Pharm. Biopharm. 47:113–118, 1999.
16. Winfield, A.J., Jessiman, D., Williams, A., et al. A
study of the causes of non-compliance by patients pre-
scribed eyedrops. Br. J. Ophthalmol. 74:477–480, 1990.
17. Kogbe, O., Liotet, S., and Tiffany, J.M. Factors re-
sponsible for tear ferning. Cornea 10:433–444, 1991.
18. Rolando, M. Tear mucus ferning test in normal and
keratoconjunctivitis sicca eyes. Chibret Int. J. Ophthal-
mol. 2:32–41, 1984.
19. Filippello, M., Scimone, G., Mazzone, M.G., et al. Fer-
ning test and artificial tears: Comparative evaluation.
In: Hurwitz, J., Miglior, M., Spinelli, D., et al., eds.
Lacrimal System, Proceed Symp Lacrimal System,
Toronto, June 25, 1994. Kugler & Ghedini,
1995:97–106. New York.
20. Hartmann, V., and Keipert, S. Physico-chemical, in
vitro and in vivo characterisation of polymers for oc-
ular use. Pharmazie 55:440–443, 2000.
21. Weyenberg, W., and Ludwig, A. In vitro assessment
of the interaction mechanisms between mucin and a
cationic cellulosic derivative by rheological methods
using an experimental design. Pharmazie 57:628–631,
2002.
22. Patel, M.M., Smart, J.D., Nevell, T.G., et al. Mucin/
poly(acrylic acid) interactions: A spectroscopic inves-
549
tigation of mucoadhesion. Biomacromology 4:1184–
1190, 2003.
23. Haogerstrom, H., Strømme, M., and Edsman, K. Drug
molecules as probe for studying the compatibility be-
tween gels and mucous tissue with dielectric spec-
troscopy. J. Pharm. Sci. 94:1090–1100, 2005.
24. Jabbari, E., Wisniewski, N., and Peppas, N.A. Evi-
dence of mucoadhesion by chain interpenetration at
a poly(acrylic acid)/mucin interface using ATR-FTIR
spectroscopy. J. Contr. Rel. 26:99–108, 1993.
25. Benedetto, D.A., Shah, D.O., and Kaufman, H.E. The
instilled fluid dynamics and surface chemistry of
polymers in the preocular tear film. Invest. Ophthal-
mol. 14:887–902, 1975.
26. Maurice, D.M., and Mishima, S. Ocular pharmacoki-
netics. In: Sears, M.L., ed. Pharmacology of the Eye.
Berlin: Springer, 1984:28–31.
27. Greaves, J.L., Wilson, C.G., Rozier, A., et al. Scinti-
graphic assessment of an ophthalmic gelling vehicle
in man and rabbit. Curr. Eye Res. 9:415–420, 1990.
28. Wilson, C.G. Topical drug delivery in the eye. Exp.
Eye Res. 78:737–743, 2004.
29. Obenberger, J. Paper strips and rings as simple tools
for standardization of experimental eye injuries. Oph-
thalmic Res. 7:363–366, 1975.
30. Moses, R., Parkinson, G., and Schuchardt, R. A stan-
dard large wound of the corneal epithelium in the rab-
bit. Invest. Ophthalmol. Vis. Sci. 18:103–106, 1979.
31. Tonelian, D.L., and Bisla, K. A new in vitro corneal
preparation to study epithelial wound healing. Invest.
Ophthalmol. Vis. Sci. 33:3024–3028, 1992.
Received: May 10, 2007
Accepted: August 30, 2007
Reprint Requests: Susi Burgalassi
Department of Bioorganic Chemistry
and Biopharmaceutics
University of Pisa
Via Bonanno 33
I-56126 Pisa
Italy
E-mail: burgal@farm.unipi.it