Comparative studies of various hyaluronic acids produced

by microbial fermentation for potential topical

ophthalmic applications

Fanny Guillaumie,1 Pascal Furrer,2 Olivia Felt-Baeyens,2 Birgit L. Fuhlendorff,1 Søren Nymand,3
Peter Westh,3 Robert Gurny,2 Khadija Schwach-Abdellaoui1
1
Novozymes Biopolymer A/S, Krogshoejvej 36, DK-2880 Bagsvaerd, Denmark
2
School of Pharmaceutical Sciences, University of Geneva, University of Lausanne, Quai Ernest-Ansermet 30,
CH-1211 Geneva 4, Switzerland
3
Department of Life Science and Chemistry, Roskilde University, PO Box 260, DK-4000 Roskilde, Denmark

Received 15 January 2007; revised 12 November 2008; accepted 2 February 2009

Published online 8 April 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32481

Abstract: This work presents a comparative study of var- the precorneal area compared to a medium MW HA at the
ious hyaluronic acids (HA) produced by fermentation of same concentration. Finally, an in vivo ocular irritation test
either Bacillus subtilis or Streptococcus towards the selection based on confocal laser scanning ophthalmoscopy (CLSO)
of an optimal molecular weight (MW) HA for the prepara- conclusively showed the excellent tolerance of both Bacil-
tion of topical ophthalmic formulations. The influence of lus-derived HA and Streptococcus-derived HA after topical
HA MW on water binding capacity, sterile filtration, rheo- instillation onto the corneal surface. Overall, this compre-
logical properties, precorneal residence time and ocular hensive work highlights the superiority of medium MW
tolerance of ophthalmic solutions was investigated. Molec- hyaluronic acid for topical ophthalmic formulations based
ular weight did not affect hydration of hyaluronic acid on their physico-chemical and biological properties, toler-
according to differential scanning calorimetry (DSC). In ance and handling. Such solutions are expected to enhance
general, medium MW HA (0.6–1 MDa) resulted in solu- tear film stability, to allow for maximum comfort, and to
tions that were superior in terms of sterile filtration and ki- exhibit high residence times, while being biocompatible
nematic viscosity requirements compared to high MW HA and easy to sterile filter. Ó 2009 Wiley Periodicals, Inc.
(>1 MDa). Moreover, all HA-based solutions exhibited J Biomed Mater Res 92A: 1421–1430, 2010
well-defined viscoelastic properties that depend on MW.
Gamma scintigraphic data indicated that HA MW at 0.1% Key words: hyaluronic acid; hyaluronan; topical ophthal-
concentration (w/v) and HA origin did not significantly mic; ocular; filterability; viscosity; viscoelasticity; precor-
affect the corneal residence time on rabbit eyes. A 0.3% so- neal residence time; tolerance; artificial tears; eye drops;
lution of high MW HA had a prolonged residence time in contact lens solutions

INTRODUCTION cous eye drops loaded with a drug, they increase the
contact time with the ocular surface, and thereby
High molecular weight (MW) polymers with water improve the bioavailability of drugs.8,9 Bioadhesive
binding properties and high viscosity have increas- properties of some polymers contribute also to the
ingly been used as hydrogels for ophthalmic applica- improvement of the bioavailability.
tions.1 Their use is threefold. Firstly, in artificial tears, The main polymeric materials that serve the afore-
they soothe, protect and lubricate the ocular surface, mentioned functions include cellulose derivatives
thus relieving the symptoms of dry eye syndrome.2–5 (e.g., hypromellose, carmellose, hydroxyethyl cellu-
Secondly, in solutions for contact lens wearers, they lose), poly(vinyl alcohol) (PVA), carbomer, poly(vinyl
also provide better and prolonged comfort to contact pyrrolidone), polyethylene glycol, and dextran. These
lens wearers through moisturizing, lubrication, and lubricating and rewetting agents are generally well
reduction of the blink frequency.6,7 Thirdly, in vis- tolerated at the concentration level found in commer-
cial eye drops. However, some of the most viscous
Correspondence to: F. Guillaumie-Longin; e-mail: fagu@
formulations are uncomfortable and lead to blurred
novozymes.com vision, stickiness, and formation of a crusty residue.
In addition, some of these polymers are Newtonian
Ó 2009 Wiley Periodicals, Inc. and do not shear thin by eye blinking, which limits
1422 GUILLAUMIE ET AL.

their spreadability on the ocular surface and thus viscous at the concentration range used in eye
their efficacy.10,11 To ease discomfort and overcome drops (0.1–0.3%, w/v).31 Topical ophthalmic solu-
these limitations, growing attention has thus been tions should exhibit a certain degree of viscosity to
paid to non-Newtonian and particularly to pseudo- prevent immediate drainage from the ocular surface
plastic polymers such as hyaluronic acid (HA).12 and to provide there high efficacy and long residence
HA is a high molecular weight, natural and linear time. However, the solutions should not be too vis-
polysaccharide composed of b-1,3-N-acetyl glucosa- cous to avoid blurred vision and to ease their manu-
mine and b-1,4-glucuronic acid repeating disaccha- facturing process including their sterile filtration.
ride units.13 It is found in the vitreous body and is This work has thus been aimed at investigating
also abundant in the extracellular matrix, especially the influence of HA molecular weight on its water
of soft connective tissue, and in the synovial fluid of binding capacity and on the preparation, the rheo-
articular joints.13,14 From the 1980s onward, the logical properties, the precorneal residence time and
understanding of HA properties and functions has the tolerance of ophthalmic solutions. The overall
blossomed into a truly scientific and industrial enter- objective has been to assess whether there exists a
prise so that now this biopolymer is recognized as a trend in favor of an optimal HA molecular weight.
high-value product with numerous proven and mar-
keted applications within cosmetic, dietary supple-
ments, and biomedical areas.15 Within topical oph- MATERIALS AND METHODS
thalmology, this biocompatible, nonimmunogenic
and biodegradable polymer is well-recognized and
considered as the ‘‘next-generation comfort ingredi- Materials and animal testing
ent.’’4,16–19 Indeed, it is one of the most hygroscopic
molecules in nature and hydrated hyaluronic acid Two types of hyaluronic acid (HA, sodium salt) were
can contain up to 1000-fold more water than its own used: medium MW HA from Bacillus-fermentation (molec-
ular weight 0.68, 0.77, 0.89, or 1.14 MDa) and high MW
weight.13 These exceptional water retention proper-
HA from Streptococcus-fermentation (molecular weight 1.5
ties result in enhanced hydration of the corneal sur- or 2.25 MDa). Streptococcus-HA (0.91 MDa) was produced
face.20 Moreover, applications of ophthalmic formu- from the above 1.5 MDa HA by heat treatment at 105oC
lations containing hyaluronic acid reduce tear elimi- for 4 h.
nation and enhance precorneal tear film stability, Technetium-99m (99mTc) was prepared as pertechnetate
which is a useful property against dry eye syn- ions by elution of a commercially available 99Mo-99mTc
drome.21 Its non-Newtonian and shear thinning generator (Mallinckrodt, The Netherlands) at the Univer-
properties grant HA solutions with a high viscosity sity hospital of Geneva (Department of Nuclear Medicine).
99m
at low shear rate (when the eye is open) and a low Tc-labeled diethylenetriaminepentaacetic acid (99mTc-
viscosity at high shear rate (during blinking) thus DTPA) was prepared from the eluate using a kit from
allowing an even distribution of the solution, Oryx Pharmaceutica (Switzerland).
Buffers were obtained from Sigma-Aldrich Denmark A/
improving lubrication of the ocular surface, retard-
S (Vallensbæk Strand, Denmark) or from Merck (Darm-
ing drainage, improving bioavailability, and reduc- stadt, Germany). Water used for sample preparation or
ing discomfort.11,22 Another important feature of this analysis was distilled and purified to a resistivity of
high molecular weight and anionic biopolymer is its 18.2 MX cm21 in a Milli-Q apparatus.
muco-adhesivity, which provides effective coating New Zealand white rabbits of either sex and weighing
and long-lasting protection of the cornea as well as from 4.5 to 5.5 kg were used for gamma scintigraphy anal-
extended residence times on the ocular surface.23–25 yses and tolerance studies. All animals were healthy and
Finally, when topically instilled on the eye, hyal- free of clinically observable ocular abnormalities. The
uronic acid has been shown to promote physiologi- experiments of the present investigation were run in ac-
cal wound healing by stimulating corneal epithelial cordance with the Association for Research in Vision and
migration and proliferation of keratocytes and to Ophthalmology (ARVO) resolution on the human handling
of animals in ophthalmic and vision research. They were
reduce the healing time of corneal epithelium.26,27
approved by the local ethics committees for animal experi-
At industrial scale, HA is either extracted from mentation as well as by the Novozymes’ Laboratory Ani-
rooster combs or produced by microbial fermenta- mal Review Committee (LARC).
tion using certain attenuated strains of Streptococcus
(s-HA).28,29 Recently, a new Bacillus subtilis-based fer-
mentation process has been developed for the indus- Molecular weight determination
trial production of hyaluronic acid (b-HA).30 The
two former methods typically yield high molecular The absolute molecular weight of hyaluronic acid was
weight HA (>1 MDa) while Bacillus-derived HA is a determined by Size Exclusion Chromatography combined
medium molecular weight biopolymer (0.6–1 MDa). with Multi Angle Laser Light Scattering detection and
High molecular weight solutions of HA are typically with Refractive Index detection (SEC-MALLS-RI). The

Journal of Biomedical Materials Research Part A

COMPARATIVE STUDIES OF VARIOUS HYALURONIC ACIDS 1423

chromatography system consisted of a Waters Alliance concentration ranging between 0.025 and 0.5% (w/v). The
HPLC (Waters 2695, Milford, MA) equipped with Wyatt’s solutions were preserved by adding 3.25 lL of Polyhexa-
Multi Angle Laser Light Scattering (Dawn EOS, Santa Bar- methylene Biguanide (PHMB, Cosmocil1 CQ). The solu-
bara, CA) and Wyatt’s Optilab rex Refractive Index (RI) tions (50 mL each) were filtered under house vacuum
detector (Optilab rEX, Santa Barbara, CA). Three TSK col- through a Durapore1 membrane filter (0.22-lm GV)
umns (4000, 5000, and 6000 PWXL) connected in series adapted on a Büchner funnel. The filtration time was
were eluted with a buffer of 50 mM NaH2PO4 and 150 mM recorded with t 5 0 when the solution had just been
NaCl at a flow rate of 0.5 mL min21 at 308C. Dn/dc and A2 loaded on the filter.
values used were 0.153 and 2.3 3 1023, respectively. All
data were calculated in Astra software v. 5.1.3.0.
Kinematic viscosity measurements
Differential scanning calorimetry
The kinematic viscosity of each solution was measured
using a capillary viscometer type Ubbelohde (SCHOTT-
Hyaluronic acid was initially dissolved in Milli-Q water
Instruments, Mainz, Germany) at 258C. The temperature
to a concentration of 0.1% (w/v). To ensure that the sam-
was controlled with a water bath. The kinematic viscosity
ples contained only the polymer and its counter ions, the
was calculated using the following equation:
solution was extensively dialyzed (MW cut-off 12–14 kDa)
against Milli-Q water until the conductivity of the dia-
Kinematic viscosity ðcStÞ ¼ K  t
lyzate was below 0.2 lS cm21. The purified polymer was
then lyophilized. Prior to the calorimetric trials the freeze-
dried powder was dissolved in Milli-Q water. The concen- Where: K is the viscometer constant and t is the time
tration was close to 1% (w/w) and determined precisely (s) for the solution to flow through the capillary. Each
by a gravimetric method in which 100 lL aliquots was sample was measured twice and the average value was
calculated.
weighed in the calorimetric cell. The cells were dried over-
night in an oven and weighed to the nearest 0.01 mg. To
compensate for a small evaporation from the solution dur-
ing the initial weighing, the weight loss was followed for Rheological measurements
120 s and extrapolated to zero. The reproducibility of the
HA concentration was 0.05% (w/w). Rheological measurements were carried out on freshly
For the differential scanning calorimetry (DSC) trials, prepared formulations using a Rheostress 1 viscosimeter
100 lL aliquots were weighed in the calorimetric cell, (Haake, Karlsruhe, Germany) fitted with a titanium cone/
which was closed hermetically and placed in the differen- plate device (diameter 60 mm, angle 18). Temperature
tial scanning calorimeter (DSC-7707, Hart Scientific, Pleas- was controlled with a thermostatic bath Haake DC30
ant Grove, Utah). The samples were cooled from room coupled with the rheometer. Temperature was maintained
temperature to 2208C at 2008C hr21 and kept at this tem- constant at 258C (60.58C). Volume of sample used was
perature for 1 h. Subsequently, the melting was studied 1.5 mL.
during heating to 58C at 108C hr21. Throughout the calori- The viscoelastic behavior of the samples was investi-
metric trials, the DSC was frequently calibrated with 100 lL gated in triplicates by dynamic amplitude shear oscillatory
aliquots of pure water exposed to the same thermal cycle. tests, in which the material was subjected to a sinusoidal
Integration of the melting peaks for pure water showed a shear strain. In the linear viscoelastic regions, the shear
repeatability of about 0.5% (n 5 12). storage modulus (or elastic modulus, G0 ), the shear loss
The mass of nonfrozen water (B) in the HA-solutions modulus (or viscous modulus, G@) and the complex viscos-
was determined as ity (h*) values were recorded from frequency sweep
" # experiments at a constant shear stress (4 Pa) and at a fre-
DH quency between 0.1 and 50 Hz.
B ¼ msample wH2 O

DHH 2O

where msample is the total mass of the sample in the calori-

Gamma scintigraphy
metric cell, wH2O is the weight fraction of water in the HA
solution and DH and DH*H2O are respectively the meas- Sodium hyaluronate was dissolved into neutral, isotonic
ured specific melting enthalpy for the HA sample and the and sterile phosphate buffer solution at room temperature
melting enthalpy for pure water (both in J g21). to provide solutions at 0.1% (w/v) and 0.3% (w/v). The
Each HA sample was run 6–9 times and the amount of final pH of the HA-formulations was verified (pHmeter
bound water was expressed as the mass ratio (MR) of non- 691, Metrohm, Switzerland) as well as the cryoscopicity
frozen water and dry HA, MR 5 B/(msamplewH2O). (Automatic osmometer type Digital/L, Knauer, Germany).
Each measurement was repeated three times. These formu-
lations (450 lL each) were labeled by addition of 99mTc-
Filtration experiments DTPA (50 lL). In vivo precorneal drainage of each formu-
lation was assessed after instillation of 25 lL of radiola-
Solutions (0.5 L each) were prepared by dissolving HA beled solution onto the left cornea using a gamma camera
into PBS buffer at room temperature to give a final HA (Toshiba GCA 602A) adjusted to detect the radiation of

Journal of Biomedical Materials Research Part A

1424 GUILLAUMIE ET AL.

99m
Tc (140 KeV) and fitted with a 4-mm pinhole. The ac-
tivity instilled ranged between 1 and 2 MBq per 25 lL
dose. A small plastic vial containing 25 lL aliquot of solu-
tion to be tested was placed near the eye of the rabbit,
and used as a position tracer. The rabbit was kept on a ta-
ble without a restraining box, its head being supported by
the experimenter hand with its left eye in front of the col-
limator aperture at a distance of 6 cm. Frames are
recorded during 10 min using a 128 3 128 pixel matrix.
Each formulation was tested on eight rabbits. No anesthe-
sia was performed.
The data were analyzed as follows. Individual 63
frames recorded during the first 10 min were aligned Figure 1. Influence of the HA molecular weight on the
and summed to obtain an overall picture of the distribu- amount of bound water.
tion of the label. The final image was divided into five
regions of interests (ROIs) which were: (i) the position
center, (ii) the precorneal reference, (iii) the inner can- RESULTS AND DISCUSSION
thus, (iv) the lachrymal duct, and (v) the background,
respectively. The raw data were transferred to a compati-
ble computer and curve treatment was performed using Influence of HA molecular weight on water
binding capacity
Excell Macros. The area under the curve of the percent-
age activity remaining in the precorneal ROI versus time
[AUC(0,10min)], which represents the corneal residence The capacity of hyaluronic acids of various chain
time of the formulation, was calculated using the mathe- lengths to bind water was assessed from the amount
matical trapezoidal rule. of nonfrozen water in DSC. The mass ratio, MR, of
non-frozen water to HA (see ‘‘Materials and Meth-
ods’’ section) was determined for b-HA, 770 kDa; b-
HA, 1140 kDa; s-HA, 1500 kDa and plotted as a
function of the molecular weight in Figure 1. Over-
Tolerance studies
all, the results showed that all HA samples bound
very high amounts of water, that is, 4–5 g g21, com-
Tolerance evaluation has been performed using a confo-
pared to both other biopolymers (e.g., proteins) and
cal laser scanning ophthalmoscope (CLSO Zeiss, Oberko-
chen, Germany) modified by addition of a set of lenses to hydrophilic synthetic polymers which typically bind
view the cornea instead of the retina. An argon ion laser 0.5–1.5 g g21.32,33 Moreover, the amount of bound
operating at 488 nm wavelength was used as the excitation water was independent on the molecular weight and
light source. The fluorescence signal was detected by a the origin of hyaluronic acid. This is in accordance
photomultiplier. Images were obtained using Epiplan-Neo- with earlier observations based on FTIR analyses of
fluar 2.53/0.075 NA objective lens (Zeiss, Oberkochen, freeze-dried medium molecular weight HA (750 kDa)
Germany). Optical sectioning was performed parallel to and HA oligosaccharides (817 Da) at controlled
the corneal surface, at 16 equidistant different focal planes, humidity.34 In that study, HA MW did not affect
the focus shifting (from 0 to 470 lm) covering the whole hydration of sodium hyaluronate. The amount of
corneal thickness. The images were displayed on a digital
nonfrozen water in dilute HA solutions measured in
video monitor and recorded on a S-VHS videotape.
the current work is much higher than the amounts
An image processing system (Analysis SIS, Münster,
Germany) carried out the following operation steps: addi- previously found in moist powders and concentrated
tion of the 16 digitized images in a stack to produce a solution of the polymer.35,36 In the latter work, only
three-dimensional reconstruction, projection of this stack 0.6–0.8 g g21 (i.e., 15 water molecules per disaccha-
and calculation of the total surface of the fluorescent areas ride unit) did not crystallize upon cooling. This pro-
on the projection. The experimental protocol was the fol- nounced dependence of MR on the water content
lowing: 25 lL of formulation to be studied were instilled emphasizes that full hydration of HA involves a sub-
onto the cornea four times per day during 3 days. After stantial number of water molecules. Thus, MR values
the last instillation of the fourth day, rabbits were sedated illustrated in Figure 1 corresponds to 100 water
with intramuscular injection of a 1:1 mixture of ketamine molecules per disaccharide unit and additional DSC
hydrochloride (15 mg kg21 of body weight) and xylazine
work (not shown) on still more dilute HA (0.5%)
hydrochloride (3 mg kg21 of body weight). Twenty five
suggested an even higher MR.
microliters of sodium fluorescein (0.5% w/v) were instilled
in the right eye to selectively marked injured areas. After a Many important roles of hyaluronic acid in biolog-
2 min dying time, the eye was rinsed during 1 min with a ical systems are related to its ability to interact with
NaCl 0.9% solution at 378C. The fluorescent image of the water through hydrogen bonding, polar interactions
central corneal apex was finally taken. Each formulation with hydroxyl groups, as well as electrostatic inter-
has been tested on 3 rabbits. actions involving charged carboxyl groups. When

Journal of Biomedical Materials Research Part A

COMPARATIVE STUDIES OF VARIOUS HYALURONIC ACIDS 1425

thus manufacturing costs during the production of

ophthalmic solutions. Finally, it also allows increas-
ing HA concentrations in eye drops and thus provid-
ing a better efficacy.

Influence of HA molecular weight on the

kinematic viscosity of ophthalmic formulations

The viscosity of a topical ophthalmic formulation

is a crucial parameter in that it influences its efficacy.
Figure 2. Filtration time of HA solutions at various con-
Indeed, it is well-established that addition of visco-
centrations; & b-HA, 0.89 MDa; n s-HA 2.25 MDa. enhancing agents, such as PVA, to ophthalmic
vehicles decreases the drainage rate and that this is
all the more pronounced than the viscosity is high.38
topically instilled on the eye, HA-based solutions Moreover, in drug delivery, a highly viscous solu-
decreased evaporative rates of water, which indi- tion prolongs the contact time of the drug with the
cates that HA enhances water retention on the cor- cornea, resulting in improved bioavailability.39 How-
neal surface and increases corneal wettability, owing ever, some disadvantages related to the use of thick-
to its water-retentive properties.20 In several clinical ened eye drops include ocular discomfort, increase
studies involving patients with severe dry eye, pre- of eye blink frequency and possible interference with
corneal tear film stability was significantly increased vision.40
in eyes treated with 0.1 or 0.3% HA-solutions and The rheological properties of eye drops and artifi-
dry eye symptoms were alleviated.21,37 cial tears depend to a large extent on the concentra-
tion, the molecular weight and the nature of viscos-
ity-enhancing agents.38,39 Even a slight degradation
Influence of HA molecular weight on the of the polymers will result in a significant change of
filterability of ophthalmic formulations
the rheological properties of the formulation. Here,
the effect of the HA molecular weight and the con-
Topical ophthalmic formulations containing ther- centration on the viscosity of the resulting solution
molabile drugs are typically sterilized by filtration at was studied in an effort to determine the optimum
production scale. However, those containing hyal- eye drop. More precisely, the kinematic viscosity of
uronic acid may be difficult to filter due to the rela- formulations containing HA at either 0.1 or 0.2%
tively high viscosity of this ingredient even at low (w/v) concentration and with a MW ranging from
concentrations. The ability to filter solutions contain- 774 kDa to 1.5 MDa was measured (Fig. 3). The data
ing medium MW HA (0.89 MDa) or high MW HA indicate that the kinematic viscosity is strongly de-
(2.25 MDa) at 0.025, 0.1, 0.2, and 0.3%, respectively, pendent on the MW and the concentration of HA. In
was studied. The time to filter 50 mL of each solu- case of 0.1% HA formulations, the kinematic viscos-
tion through a 0.22-lm filter was recorded (Fig. 2). ity varied from 2.35 to 3.83 cSt in a linear fashion
The molecular weight does not affect the filtration when HA MW increased from 774 kDa (medium
time up to 0.1%. However, at 0.2% concentration, it MW) to 1.5 MDa (high MW). At higher concentra-
takes up to three times longer to filter a high MW
HA solution than a medium MW HA solution. The
effect is even more pronounced when the concen-
tration is increased to 0.3%: a medium MW HA
solution is filtered within less than 10 min while fil-
tration of a high MW HA solution requires nearly
1 hour.
Today, commercial eye drops recommended for
the treatment of dry eye contain from 0.1 to 0.3%
HA. Sterile filtration of these ophthalmic solutions is
a crucial step in the manufacturing process in that
its optimization can not only reduce the number of
adverse events linked to the handling of highly vis-
cous solutions, but also lower the production costs.
In view of this, the use of medium MW HA appears Figure 3. Influence of HA molecular weight on the kine-
highly advantageous compared to high MW HA. It matic viscosity of HA solutions; & 0.1% (w/v); n 0.2%
is expected to reduce filtration time at large scale (w/v).

Journal of Biomedical Materials Research Part A

1426 GUILLAUMIE ET AL.

tion (0.2%), the influence of HA MW on the kinematic the crossover frequency, decreased with increasing
viscosity of the resulting formulation was greater due HA molecular weight. These findings are in good
to higher entanglement of the HA molecules: a solu- agreement with those reported in the literature.41–44
tion containing medium MW HA (774 kDa) has a For instance, Falcone and coworkers observed on
viscosity of 4.89 cSt, whereas a 1.5 MDa HA gives a three solutions prepared from 0.35, 0.68, and 1.8 MDa
solution with a viscosity of 11.24 cSt. HA, respectively, that the crossover frequency
Manufacturers of eye drops for dry eye treatment decreased as the molecular weights increased.41
usually set defined specifications on the viscosity of At concentrations lower than 1% (w/v), the HA-
their formulations and the kinematic viscosity of based solutions may form weak gels with a predom-
commercial eye drops typically varies from 2 to inant viscous behavior. A wide range of other poly-
7 cSt (data not shown). This kinematic viscosity can saccharides including carrageenan, scleroglucan, and
be met using both medium and high molecular xanthan gum have been selected for potential ocular
weight HA given that the HA concentration is kept use based on the fact that they formed weak visco-
at 0.1%. However, the use of high MW HA at higher elastic gels.45 It is noteworthy that the HA visco-
concentrations may result in too viscous solutions elastic parameters are influenced by the solubiliza-
according to the market’s specifications. In that case, tion medium, particularly by saccharides (e.g., glu-
medium MW HA appears to be the product of cose) and by salts (e.g., NaCl) commonly used to
choice to prepare topical formulations with a higher prepare isotonic solutions.44,46 More precisely, sac-
content of HA (0.2%) without compromising their charides promote network formation by creating
comfort on the eye. hydrogen bonds with HA, whereas salts may reduce
G0 and G@ due to cation shielding around the HA
carboxyl groups and subsequent electrostatic repul-
Influence of HA molecular weight on the sion of HA molecules.
rheological properties of ophthalmic formulations

Polymeric substances used to increase viscosity Influence of HA molecular weight and HA

display various rheological behaviors, which have concentration on the in vivo residence
been shown to have significant effect on the proper- time of ophthalmic formulations
ties of eye drops.31 Moreover, the rheological charac-
teristics of polymer solutions are a combined effect The type and characteristics of the various HA-
of molecular size and concentration. A series of based formulations used for in vivo residence time
dynamic experiments was carried out on hyaluronic studies are summarized in Table I. In all cases, both
acids originating from either Streptococcus-fermenta- pH and osmolality fit in the physiological range.
tion or Bacillus-fermentation to evaluate their visco- Clearance curves from the ocular surface were
elastic properties. plotted for each formulation containing b-HA
Figure 4 shows the rheological behavior of HA (0.68 MDa), s-HA (0.91 MDa) and s-HA (2.25 MDa)
solutions at 1% (w/v) in water. The concentration of at 0.1% concentration (Fig. 5) as well as for two HA
hyaluronic derivatives in ophthalmology varies molecular weights (0.68 and 2.25 MDa) at 0.1% [Fig.
according to the intended use: for the topical appli- 6(a)] and 0.3% [Fig. 6(b)] using mean data from
cation of artificial tears, a 0.15% concentration of HA 8 rabbits. The area under the curve values, which
is common, whereas commercial viscosurgical devi- provide an index of residence time, are reported for
ces contain generally between 1.0 and 3.0% HA. The each HA-based solution (Table I). In case of 0.1%
investigated concentration chosen was 1.0%. In Fig- HA solutions, the AUC values ranged between 5563
ure 4, the complex viscosity (h*) decreased with 6 2810 and 6356 6 2375. This indicates that the cor-
increasing frequency due to the entangled network neal residence time of HA at 0.1% was not signifi-
conformational and dynamic properties of HA in cantly influenced by the biopolymer origin (Bacillus-
solutions. In general, the HA-based solutions studied derived HA versus Streptococcus-derived HA) nor by
herein behaved as viscous liquids (G@ > G0 ) at very its molecular weight (medium MW versus high
low frequencies: molecular chains can release stress MW). Interestingly, such a corneal residence time is
by disentanglement and molecular rearrangement similar to that obtained with a saline solution, which
during the oscillation period (Fig. 4). At higher fre- typically has an AUC of 6000 (%.min). In other
quencies, HA solutions were elastic (G0 > G@), which words, hyaluronic acid does not seem to prolong the
means that the oscillation period is too short for the ocular residence time of ophthalmic formulations at
molecules to disentangle, to rearrange and that they 0.1%. This had previously been reported by Gurny
behave as temporary cross-linked networks. How- and coworkers who showed that there was no differ-
ever, the frequency at which the viscous (G@) and ence in precorneal residence time between an iso-
the elastic (G0 ) curves cross each other, referred to as tonic phosphate buffer and a 0.125% solution of a

Journal of Biomedical Materials Research Part A

COMPARATIVE STUDIES OF VARIOUS HYALURONIC ACIDS 1427

TABLE I
Physico-Chemical Characteristics of Hyaluronic Acid Formulations Used for In Vivo Studies and
Their Residence Time [Area Under the Curve (AUC) Value] Onto The Corneal Surface
Concentration AUC(0,10 min) 6 Standard
Hyaluronic Acid (source, MW) (%, w/v) pH Osmolality (mOsm/kg) Deviation (%.min)

Bacillus-derived, 0.68 MDa 0.1 6.99 269 5563 6 2810

0.3 7.18 260 5487 6 1434
Streptococcus-derived, 0.91 MDa 0.1 6.88 266 6356 6 2375
0.3 6.53 276 7558 6 1553
Streptococcus-derived, 2.25 MDa 0.1 6.95 270 5810 6 1722
0.3 7.07 282 10,504 6 2784

high MW HA (3 MDa).47 Moreover, the performance neal residence time, except for high MW s-HA (Fig.
of medium MW HA in terms of ocular surface resi- 6). In fact, the latter polymer exhibits a twofold
dence time is equal to that of high MW. The lack of higher AUC value at 0.3% than at 0.1% [Table I, Fig.
significant influence of molecular weight of some 6(b)]. Its complex viscosity at 0.3% (68 mPa s) is
biopolymers on the residence time has been previ- largely higher than that of all other solutions and
ously observed for chitosan at concentrations could in part explain this difference. In another
between 0.5 and 1.5%.48 study, a 0.25% HA solution (with a high MW HA)
Increasing the HA concentration to 0.3% did not had a prolonged residence time in the precorneal
necessarily induce a corresponding increase in cor- area compared to a 0.125% solution prepared with
the same HA.47 This positive attribute has for
instance been exploited for the ocular delivery of
gentamicin which bioavailability was increased for
at least 10 min.8
Moreover, a formulation containing 0.3% of high
MW HA with extended residence time would be
expected to enhance even more the tear film stabil-
ity, to reduce the drainage rate of the instilled solu-
tion and to allow longer intervals between instilla-
tions. However, from a practical point of view, such
a highly viscous formulation will be difficult to filter
and thus to manufacture and its use in topical oph-
thalmology may lead to discomfort and blurred
vision.

Figure 4. Frequency dependence of G0 (~, elastic modu-

lus), G@ (, loss modulus) and complex viscosity (continu-
ous line) for two hyaluronic acid solutions (1%, w/v); n 5 Figure 5. Influence of the polymer on corneal residence
3, standard deviation not shown. (A) Streptococcus-derived time of HA formulations at 0.1% (w/v); n b-HA 0.68
HA, 2.25 MDa; (B) Bacillus-derived HA, 0.68 MDa. [Color MDa; ^ s-HA 0.91 MDa; ~ s-HA 2.25 MDa. [Color figure
figure can be viewed in the online issue, which is available can be viewed in the online issue, which is available at
at www.interscience.wiley.com.] www.interscience.wiley.com.]

Journal of Biomedical Materials Research Part A

1428 GUILLAUMIE ET AL.

formulations were very well tolerated, according to

the scale of tolerance previously established by
Kälin.50 Moreover, neither the origin nor the molecu-
lar weight nor the concentration modify significantly
the tolerance of the HA-based formulations tested,
all of them being acceptable for topical ophthalmic
use. Finally, this in vivo study highlights the biocom-
patibility of Bacillus-derived HA on the corneal
surface.

CONCLUSION

Topical ophthalmic solutions containing hyal-

uronic acid are more and more popular owing to the
general acceptance of this natural polymer. More-
over, its numerous properties make it a rewetting
agent of choice compared to traditional polymers.
Recent development in the biotechnology field has
Figure 6. Influence of the concentration of polymer led to the availability of a range of molecular
on corneal residence time of HA formulations. (a) b-HA weights for eye care applications, particularly high
0.68 MDa, (b) s-HA 2.25 MDa; ^ 0.1%; n 0.3%. [Color (>1 MDa) and medium (0.6–1 MDa) molecular
figure can be viewed in the online issue, which is available
at www.interscience.wiley.com.] weights. The characteristics of formulations contain-
ing either medium MW HA or high MW HA have
been compared. Our results indicate that medium
The remaining activity on the corneal surface after MW HA is superior to high MW HA because it
3 min is at least 5% in all cases and remains at this leads to ophthalmic solutions that: (i) may maintain
level until the end of the study (Fig. 6). Additional or enhance the tear film stability through high water
investigation would be required to determine the binding capacity; (ii) are easy to sterilize by filtra-
effects of the remaining hyaluronic acid in the ther- tion; (iii) have viscoelastic properties allowing for
apy of dry eye. maximum comfort; (iv) exhibit high residence times
on the corneal surface for enhanced efficacy; (v) are
biocompatible. The use of high MW HA may not be
Tolerance and biocompatibility evaluation
optimal in terms of viscosity and comfort, filtration
and manufacturing. Whereas most of the reported
In this study, s-HA and b-HA of either high (2.25 studies on the use of polymeric substances to
MDa) or medium molecular (0.68 or 0.91 MDa) improve the performance of topical ophthalmic
weight were used (Table I). Two concentrations were applications focused on contact time and bioavaila-
tested, namely 0.1 and 0.3%. The osmolality and pH bility only, we have taken into account very impor-
of the tested formulations, as mentioned earlier,
were within a physiologically compatible range
(Table I).
Figure 7 illustrates the percentage of corneal injury
in term of epithelial cell loss, measured by confocal
laser scanning ophthalmoscopy, following instillation
of HA-containing formulations in eyes of three rab-
bits. The high reproducibility and high sensitivity of
this method for tolerance assessment are well-recog-
nized advantages.49 In this study, the normal corneal
cell turn over was about 2.5% as demonstrated by
instillation of a control solution of 0.9% NaCl.
Administration of HA-containing formulations
induced a slightly higher level of irritation than the Figure 7. Percentage of corneal lesion after instillation of
HA formulations with various concentrations and with
control sample, with percentage of corneal lesions various HA molecular weights; & b-HA 0.68 MDa; s-HA
ranging between 7.34 6 2.75 and 10.71 6 1.24. These 0.91 MDa; n s-HA 2.25 MDa (*Scale established by
levels were lower than 25% and indicated that all Kälin50).

Journal of Biomedical Materials Research Part A

COMPARATIVE STUDIES OF VARIOUS HYALURONIC ACIDS 1429

tant aspects such as pharmaceutical handling, for 15. Balazs EA, Laurent TC. Round table discussion: New applica-
example, sterile filtration. tions for hyaluronan. In: Laurent TC, editor. The Chemistry,
Biology and Medical Applications of Hyaluronan and Its
Additional work is now required to validate these Derivatives. London: Portland Press Ltd; 1998. p 325–336.
initial findings. For instance, clinical studies should 16. Szczotka-Flynn LB. Chemical properties of contact lens
be performed to evaluate the comparative efficacy rewetters. A review on hyaluronic acid as a contemporary in-
and acceptability of medium MW and high MW gredient in contact lens rewetters. Contact Lens Spectr Apr
HA. The superiority of one hyaluronic acid over the 2006;40–45.
17. Aragona P, Di Stefano G, Ferreri F, Spinella R, Stilo A. So-
other should also be confirmed by corneal wound dium hyaluronate eye drops of different osmolarity for the
healing studies. treatment of dry eye in Sjogren’s syndrome patients. Br
J Ophthalmol 2002;86:879–884.
The authors thank Catherine Siegfried for her technical 18. Papa V, Aragona P, Russo S, Di Bella A, Russo P, Milazzo G.
Comparison of hypotonic and isotonic solutions containing
assistance during the rheological measurements.
sodium hyaluronate on the symptomatic treatment of dry eye
patients. Ophthalmologica 2001;215:124–127.
19. Condon PI, McEwen CG, Wright M, Mackintosh G, Prescott
References RJ, McDonald C. Double blind, randomised, placebo con-
trolled, crossover, multicentre study to determine the efficacy
1. Felt O, Einmahl S, Furrer P, Baeyens V, Gurny R. Polymeric of a 0.1% (w/v) sodium hyaluronate solution (Fermavisc) in
system for ophthalmic drug delivery. In: Dumitriu S, editor. the treatment of dry eye syndrome. Br J Ophthalmol 1999;
Polymeric Biomaterials. New York: Marcel Dekker; 2002. 83:1121–1124.
p 377–421. 20. Nakamura M, Hikida M; Nakano T, Ito S, Hamano T,
2. Bron AJ, Mangat H, Quinlan M, Foley-Nolan A, Eustace P, Kinoshita S. Characterization of water retentive properties of
Fsadni M, Sunder Raj P. Polyacrylic acid gel in patients with hyaluronan. Cornea 1993;12:433–436.
dry eyes: a randomised comparison with polyvinyl alcohol. 21. Hamano T, Horimoto K, Lee M, Komemushi S. Sodium hya-
Eur J Ophthalmol 1998;8:81–89. luronate eye drops enhance tear film stability. Jpn J Ophthal-
3. Pouliquen P. Carbomer gels in dry eye treatment: Literature mol 1996;40:62–65.
review. J Fr Ophthalmol 1999;22:903–913. 22. Berriaud N, Milas M, Rinaudo M. Characterization and prop-
4. McDonald CC, Kaye SB, Figueiredo FC, Macintosh G, Lockett erties of hyaluronic acid (hyaluronan). In: Dumitriu S, editor.
C. A randomised, crossover, multicentre study to compare Polysaccharides. Structural Diversity and Functional Versatil-
the performance of 0.1% (w/v) sodium hyaluronate with ity. New York: Marcel Dekker Inc; 1998. p 313–334.
1.4% (w/v) polyvinyl alcohol in the alleviation of symptoms 23. Saettone MF, Monti D, Torracca MT, Chetoni P. Mucoadhe-
associated with dry eye syndrome. Eye 2002;16:601–607. sive ophthalmic vehicles: evaluation of polymeric low-viscos-
5. Christensen MT, Cohen S, Rinehart J, Akers F, Pemberton B, ity formulations. J Ocul Pharmacol 1994;10:83–92.
Bloomenstein M, Lesher M, Kaplan D, Meadows D, Meuse P, 24. Greaves JL, Wilson CG. Treatment of diseases of the eye with
Hearn C, Stein JM. Clinical evaluation of an HP-guar gellable mucoadhesive delivery systems. Adv Drug Delivery Rev
lubricant eye drop for the relief of dryness of the eye. Curr 1993;11:349–383.
Eye Res 2004;28:55–62. 25. Saettone MF, Chetoni P, Torraca MT, Burgalassi S, Giannac-
6. Acosta MC, Gallar J, Belmonte C. The influence of eye solu- cini B. Evaluation of muco-adhesive properties and in vivo
tions on blinking and ocular comfort at rest and during work activity of ophthalmic vehicles based on hyaluronic acid. Int J
at video display terminals. Exp Eye Res 1999;68:663–669. Pharm 1989;51:203–212.
7. Sclafani LA. Lubricant eye drops for contact lens wearers. 26. Tani E, Katakami C, Negi A. Effects of various eye drops on
Rev Cornea Contact Lens Mar 2005;1–4. corneal wound healing after superficial keratectomy in rab-
8. Bernatchez SF, Tabatabay C, Gurny R. Sodium hyaluronate bits. Jpn J Ophthalmol 2002;46:488–495.
0.25% used as a vehicle increases the bioavailability of topi- 27. Nishida T, Nakamura M, Mishima H, Otori T. Hyaluronan
cally administered gentamicin. Graefe’s Arch Clin Exp Oph- stimulates corneal epithelial migration. Exp Eye Res 1991;53:
thalmol 1993;231:157–161. 753–758.
9. Saettone MF, Giannaccini B, Ravecca S, La Marca P, Tota G. 28. Chong BF, Blank LM, Mclaughlin R, Nielsen LK. Microbial
Polymer effects on ocular bioavailability-the influence of dif- hyaluronic acid production. Appl Microbiol Biotechnol 2005;
ferent liquid vehicles on the mydriatic response of tropica- 66:341–351.
mide in humans and in rabbits. Int J Pharm 1984;20:187–202. 29. Balazs EA. Ultra pure, high molecular wt. hyaluronic acid—
10. Snibson GR, Greaves JL, Soper NDW, Tiffany JM, Wilson Is non-antigenic and used as synthetic aq. humour and syno-
CG, Bron AJ. Ocular surface residence times of artificial tear vial fluid, as nerve tissue and wound protector and drug car-
solutions. Cornea 1992;11:288–293. rier. US patent 4,141,973; 1979.
11. Ludwig A, Van Ooteghem M. Evaluation of sodium hyaluro- 30. Widner W, Sloma A, Thomas M, Tang M. Producing a hyal-
nate as viscous vehicle for eye drops. J Pharm Belg uronic acid, by cultivating Bacillus cell comprising a nucleic
1989;44:391–397. acid construct comprising a variant amyL promoter, a con-
12. Camber O, Edman P, Gurny R. Influence of sodium hyaluro- sensus promoter, and a cryIIIA promoter, and isolating hyal-
nate on the miotic effect of pilocarpine in rabbits. Curr Eye uronic acid from the cultivation medium. WO patent applica-
Res 1987;6:779–784. tion 2005098016-A2; 2005.
13. Laurent TC, Fraser JRE. Hyaluronan. FASEB J. 1992;6:2397– 31. Bothner H, Waaler T, Wik O. Rheological characterization of
2404. tear substitutes. Drug Dev Ind Pharm 1990;16:755–768.
14. Balazs EA, Armand G. Glycosaminoglycans and proteogly- 32. Ciesla K, Rahier H, Zakrzewska-Trznadel G. Interaction of
cans of ocular tissues. In: Varma RS, Varma R, editors. Gly- water with the regenerated cellulose membrane studied by
cosaminoglycans and Proteoglycans in Physiological and DSC. J Therm Anal Calorim 2004;77:279–293.
Pathological Processes of Body Systems. Basel: Karger; 1992. 33. Ali S, Bettelheim FA. Non-freezing water in protein solutions.
p 480–499. Colloid Polym Sci 1985;263:396–398.

Journal of Biomedical Materials Research Part A

1430 GUILLAUMIE ET AL.

34. Gilli R, Kacurakova M, Mathlouthi M, Navarini L, Paoletti S. of hyaluronan and its derivatives. London: Portland Press
FTIR studies of sodium hyaluronate and its oligomers in the Ltd; 1998. p 25–32.
amorphous solid phase and in aqueous solution. Carbohydr 43. Ambrosio L, Borzacchiello A, Netti PA, Nicolais L. Rheologi-
Res 1994;263:315–326. cal study on hyaluronic acid and its derivative solutions.
35. Jouon N, Rinaudo M, Milas M, Desbrières J. Hyadration of J Macromol Sci Part A: Pure Appl Chem 1999;36:991–1000.
hyaluronic acid as a function of the counterion type and rela- 44. Gibbs DA, Merrill EW, Smith KA, Balazs EA. Rheology of
tive humidity. Carbohydr Polym 1995;26:69–73. hyaluronic acid. Biopolymers 1968;6:777–791.
36. Joshi HN, Topp EM. Hydration in hyaluronic acid and its 45. Albasani M, Ludwig A. Evaluation of polysaccharides
esters using differential scanning calorimetry. Int J Pharm intended for ophthalmic use in ocular dosage forms. Il Farm-
1992;80:213–225. aco 1995;50:633–642.
37. Mengher LS, Pandher KS, Bron AJ, Davey CC. Effect of so- 46. Kobayashi Y, Okamoto A, Nishinari K. Viscoelasticity of hy-
dium hyaluronate (0.1%) on break-up time (NIBUT) in aluronic acid with different molecular weights. Biorheology
patients with dry eyes. Br J Ophthalmol 1986;70:442–447. 1994;31:235–244.
38. Patton TF, Robinson JR. Ocular evaluation of polyvinyl alco- 47. Gurny R, Ryser JE, Tabatabay C, Martenet M, Edman P,
hol vehicle in rabbits. J Pharm Sci 1975;64:1312–1316. Camber O. Precorneal residence time in humans of sodium
39. Saettone MF, Giannaccini B, Chetoni P, Torracca MT, Monti hyaluronate as measured by gamma scintigraphy. Graefe’s
D. Evaluation of high- and low-molecular-weight fractions of Arch Clin Exp Ophthalmol 1990;228:510–512.
sodium hyaluronate and an ionic complex as adjuvants for 48. Felt O, Furrer P, Mayer JM, Plazonnet B, Buri P, Gurny R.
topical ophthalmic vehicles containing pilocarpine. Int J Topical use of chitosan in ophthalmology: Tolerance assess-
Pharm 1991;72:131–139. ment and evaluation of precorneal retention. Int J Pharm
40. Dudinski O, Finnin BC, Reed BL. Acceptability of thickened 1999;180:185–193.
eye drops to human subjects. Curr Ther Res 1983;33:322–337. 49. Furrer P, Plazonnet B, Mayer JM, Gurny R. Application of
41. Falcone SJ, Palmeri DM, Berg RA. Rheological and cohesive in vivo confocal microscopy to the objective evaluation of oc-
properties of hyaluronic acid. J Biomed Mater Res Part A ular irritation induced by surfactants. Int J Pharm 2000;207:
2006;76A:721–728. 89–98.
42. Bothner Wik H, Wik O. Rheology of hyaluronan. In: Laurent 50. Kälin P. Contribution à la validation d’un test de tolérance
TC, editor. The chemistry, biology and medical applications oculaire sur la souris. Ph.D. thesis, Lausanne, 1994.

Journal of Biomedical Materials Research Part A