Download Characterization of paracellular and aqueous penetration

Characterization of Paracellular and Aqueous Penetration
Routes in Cornea, Conjunctiva, and Sclera
Kaisa Mari Hamalainen* Kirsi Kananen* Seppo Auriola* Kyosti Kontturi,^
and Arto Urtti*
Purpose. To characterize quantitatively the paracellular permeation routes in rabbit cornea,
conjunctiva, and sclera using polyethylene glycol (PEG) oligomers.
Methods. Corneal, conjunctival, and scleral tissues from New Zealand white rabbits were tested
individually in a modified two-chamber Ussing apparatus with the mixture of PEGs with mean
molecular weights 200, 400, 600, and 1000 in glutathione bicarbonated Ringer's solution
buffer on the donor side of the chamber. The samples and standards were analyzed with
high-performance liquid chromatography-thermospray mass spectrometry method. The pore
sizes and the pore densities of the corneal and conjunctival epithelia were calculated using
an effusion-like approach.
Results. The conjunctival and scleral tissues were 15 to 25 times more permeable than the
cornea and the molecular size affected the conjunctival permeability less than that of the
cornea. The palpebral and bulbar conjunctivas had equal permeabilities. The scleral permeability was approximately half of that in the conjunctiva and approximately 10 times more
than in the cornea. The conjunctival epithelia had 2 times larger pores and 16 times higher
pore density than the cornea. The total paracellular space in the conjunctiva was estimated
to be 230 times greater than that in the cornea.
Conclusions. The conjunctival epithelium, due to its higher membrane permeability and larger
absorptive and intercellular space surface areas, is the most viable route for ocular delivery
of peptides and oligonucleotides. Invest Ophthalmol Vis Sci. 1997;38:627-634.
A he cornea is considered to be the main route by which
topically applied drugs enter the eye.1'2 Another potential
route of ocular drug absorption is across the bulbar conjunctiva and sclera into the uveal tract and vitreous humor. This route has been shown to be important, for
example, in the case of inulin3 (molecular weight, 5000)
and /?-aminoclonidine4 (molecular weight, 245.1), both
of which have poor corneal permeability. Conjunctival
absorption also has been suggested as a potential route
to deliver bioactive peptides to the systemic circulation.
Both in the corneal and conjunctival epithelia, the
intercellular space is sealed by the junctional complexes
From the* Department of Pharmaceutics, University of Kuopio, Kuofrio, Finland,
and the ^Laboratory of Physical Chemistry and Electrochemistry, Department of
Chemical Engineering, Helsinki University of Technology, Espoo, Finland.
Supported by grants from the Academy of Finland and the Finnish Cultural
Foundation.
Submitted for publication June 21, 1996; revised September 5, 1996; accepted
October 24, 1996.
Proprietary interest category: N.
Reprint requests: Kaisa Mari HtimtiUiinen, Department of Pharmaceutics,
University of Kuopio, P.O. Box 1627, Ff-70211 Kuopio, Finland.
and that hinders the transport of hydrophilic compounds,
like many peptides.5"8 The corneal epithelium consists of
basal columnar cells, two to three layers of wings cells,
and one or two layers of squamous, polygonal-shaped
superficial cells. The first and second flattened apical cell
layers are the rate-determining barriers in the paracellular
permeation of molecules.6 For example, after intravenous
administration, no horseradish peroxidase activity was
seen on the apical surface of the superficial cells in corneal epithelium, but the activity was present paracellularly
in the wing cell and basal cell layers.9
The conjunctival epithelium can be divided into
three morphologically distinct epithelia: bulbar epithelium, which is continuous with that of the cornea; fomix
epithelium; and palpebral epithelium.10 The conjunctiva
contains many mucous goblet cells, and the epithelium
is two- to three-cell-layers thick, with cuboidal basal cells,
and it contains tight junctions only on the apical surface."
Consequently, the epithelial layer is the rate-limiting barrier for drug penetration in the cornea and conjunctiva.
Sclera is composed of collagen and mucopolysaccha-
Invesiigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
627
628
Investigative Ophthalmology 8c Visual Science, March 1997, Vol. 38, No. 3
ride fibers, and the primary mechanism for drug permeation likely is to be the diffusion across the intercellular
aqueous media, as in the case of the structurally resemblance corneal stroma.1'2 Therefore, the tortuosity of
these water channels affects the drug permeability in the
sclera.
The pore radius of the intercellular spaces is an important factor regulating paracellular permeation of peptides and oligonucleotides in biomembranes. Promotion
of peptide absorption through paracellular pathway has
gained interest because of the perception that the paracellular route is deficient in proteolytic activity, thereby minimizing the enzymatic penetration barrier of peptide.12
Polyethylene glycols (PEGs) display some basic features of peptides and oligonucleotides (hydrophilicity, hydrogen bonding capacity, size) and, therefore, they are
suitable model compounds to study the physical barrier
of die epithelium.13 PEGs often have been used to investigate the influence of molecular size on the paracellular
permeability of polar compounds in biologic membranes. H - |b We studied the corneal, conjunctival, and
scleral permeabilities of PEG oligomers. To quantitate
the paracellular permeation route, we calculated the pore
sizes, porosities, and the number of pores in the corneal
and conjunctival epithelia of rabbits using die in vitro
permeabilities of 17 PEG oligomers.
MATERIALS AND METHODS
Animals
Male and female albino New Zealand rabbits weighing
between 3.0 and 4.5 kg were used. The animals were
housed in standard laboratory rabbit cages and fed
regular diets with no restrictions on the amount of
food or water consumed. All experiments conformed
to the ARVO Resolution on the Use of Animals in
Research.
Materials
PEGs with mean molecular weights of 200 (Mw/Mn =
1.11), 400 (Mw/Mn = 1.07), 600 (Mw/Mn = 1.10), and
1000 (Mw/Mn = 1.05) were obtained from Chemical
Pressure (Pittsburgh, PA). Polypropylene glycol 425
(Aldrich, Steinheim, Germany) was used as an internal
standard. All other chemicals were of analytic grade.
Polyethylene Glycol Solutions
PEG 200 (final concentration, 2.0 mg/ml), PEG 400
(4.0 mg/ml), PEG 600 (6.0 mg/ml), and PEG 1000
(10.0 mg/ml) were weighed and dissolved in glutathione bicarbonated Ringer's solution (GBR).17 On the
basis of mean molecular weight, the concentrations
are 0.01 M.
The osmolarity of solutions was between 300 and
309 mOsm as determined on an Osmostat osmometer
(Kyoto Kaqaku, Kyoto, Japan). The pH of the solution
was adjusted to 7.65 at 37°C with oxygen-carbon dioxide (95:5) bubbling.
In Vitro Permeability Experiment
The rabbits were killed by a marginal vein injection of
a lethal dose of T-61-vet (Hoechst, Munich, Germany).
Conjunctivas were dissected from the palpebral (lower
eyelid) or bulbar side (lower cul-de-sac). The bulbar
conjunctivas were dissected without Tenon's capsule.
Then, an incision was made along scleral tissue and
corneas were dissected, leaving a scleral ring (4 mm
laterally from the limbus). Lens and iris were removed,
and the remaining sclera with episclera was detached
from the orbit. The sclera samples for permeability
experiments were obtained from the superior temporal quadrant. The samples were without any muscle
attachments or blood vessels. Six to seven tissues were
used for each permeability determination.
The tissues were positioned between two plastic
rings and placed within 25 minutes from the death
in the center of the modified perfusion chamber as
described previously.18 The exposed surface areas of
the conjunctiva, sclera, and cornea were 0.28, 0.28,
and 1.17 cm2, respectively. GBR solution (6.5 ml) was
added to the receptor side. Immediately thereafter,
an equal volume of PEG in GBR was added to the
donor side. Constant mixing of the reservoir solution
was achieved by bubbling an oxygen-carbon dioxide
(95:5) mixture, which maintained the pH at 7.65. The
experiments were conducted at 37°C for a period of
4 hours. Samples of 1 ml were collected from the
receptor chamber at 30-minute intervals and replaced
with an equal volume of blank GBR buffer.
After each permeability experiment, the cornea
was removed from the mounting rings, and the remaining scleral tissue was cut away. The cornea was
weighed and dried at 50°C overnight. After reweighing, the water content of the cornea was calculated. In all cases, the corneal hydration levels were
in the normal range (76% to 80%).
Analytic Procedure
Pure PEG oligomers to be used as analytic standards
were obtained by preparative high-performance liquid
chromatography (HPLC)19 using Shimadzu LC-6A
Liquid Chromatograph (Kyoto,Japan), equipped with
a Kromasil C-8 column, 1 cm X 25 cm (Eka Nobel,
Bolus, Sweden) at a flow rate of 3 ml/minute. The
eluent consisted of acetonitrile/water, 15/85 for PEG
300, 20/80 for PEG 600, and 22.5/77.5 for PEG 1000
(Fluka, Buchs, Switzerland). PEG oligomers were detected at 195 nm, the fractions were collected, and
the eluent was evaporated in vacuum. Each fraction
of oligomer was weighed, and the molecular weights
and purities (>95%) of each fraction were deter-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
Anterior Paracellular Routes in the Eye
mined by thermospray liquid chromatography-mass
spectrometry (LC-MS). These purified oligomers
with known concentration were used as standards for
determination of oligomer concentrations in the polydisperse PEG mixtures used in permeation studies.
The HPLC system with column of 4.5 mm X 15 cm
was used. The acetonitrile-water gradient was 10% to
15% in 7.5 minutes and 25% in 30 minutes. Calibration standards were made at concentrations 0.5 mg/
ml, 1 mg/ml, and 1.5 mg/ml for each oligomer. Six
replicate samples of the PEG solution used in the donor side of the permeation chamber were analyzed.
The samples and the quantitation standards were
analyzed with the HPLC-thermospray mass spectrometry (TSP-LC-MS) method. The method has been
described in detail previously.20 Model 2900-0374 solvent delivery system (Applied Biosystems), Rheodyne
7125 injector (50 //I), and PRP-1 column (150 X 4.1
mm inner diameter, 10 //m particle size; Hamilton,
Reno, NV) were used. The isocratic eluent consisted of
ammonium acetate (0.1 M, pH 6)-acetonitrile (79:21,
vol/vol). The flow rate was 1 ml/minute.
The LC-MS system was a vacuum generator (VG)
thermospray probe coupled to a VG Trio-2 quadropole mass spectrometer (VG Masslab, Manchester,
UK). The ion source was modified by changing the
original blunt tip repeller electrode to a needle tip
electrode located 3 mm from the ion exit aperture.
The ion source was at 210°C, the vaporized temperature 175° to 200°C, and the ion repeller potential was
300 V. Other source parameters were tuned daily with
[PEG + NH,,] + ions at m/z 388, 608, and 916 by
injecting PEG through a loop.
The oligomers in the PEG sample mixtures were
quantitated using selected ion recording. Selected ion
recording was based on the ammoniated molecules
(MNH<,+ ions): m/z 384 for internal standard, polypropylene glycol, and m/z 256, 300, 344, 388, 432,
476, 520, 564, 608, 652, 696, 740, 784, 828, 872, 916,
and 960 for PEGs. The method allows direct measurement of the permeation percentage for all 17 PEG
oligomers in the mixture.
The HPLC-TSP-mass spectrometer was calibrated
with PEG-spectrum and reference table of PEGs. Retention times of each oligomer were controlled with
standards. The internal standard (polypropylene glycol, MNH,,+, at m/z 384) had retention time of 8
minutes. The samples were analyzed without any extraction procedure.
The quantitation standards were prepared by diluting the donor-side PEG solution with GBR buffer.
These dilutions correspond to 0.002%, 0.05%, 0.1%,
0.3%, and 1% transmembrane permeation of PEG,
respectively. Four to five calibration point graphs (duplicate injections) were collected for the range corresponding to 0.01% to 0.3% permeation by plotting
629
the ratios of analyte and internal standard peak areas
versus the amounts of analyte (r = 0.998 to 0.999).
The precision of the TSP-LC-MS system was tested
using within-day and day-to-day precision. Three same
samples were analyzed repeatedly on six different
days. The precision of the analyzing method was best
by PEG in the molecular weights ranging from 238 to
810 g/mol (RSD, 4.7% to 19.8%). The bigger the
molecular weight was, the worse was the precision.
The reason for weak precision (at molecular weight
>810) could be the contamination of the ion source
with PEG.
Estimation of the Pore Sizes and Pore Density
of the Epithelia
The penetrated amounts (milligrams) of PEGs were
plotted versus time (minutes). The permeability coefficient Papp (centimeters/second) was calculated as
oligomer flux (Jh) divided by its initial concentration
in the donor chamber (milligram/milliliter). The lag
time (T|ag) of permeation is the time required to reach
the steady-state concentration gradient in the membrane. This is prerequisite for steady-state flux. It was
denned as the extrapolated intercept of the linear part
of the permeation curve on the time axis.
Effusion approach was used to calculate the limiting pore size and pore density. It is based on the low
probability of molecules to find a pore as a limitingfactor and therefore, permeation is not hindered by
the transport in the pore. Another criteria for effusion-like process is the small rate-limiting barrier thickness. The theory and the derivation formula are
given.21
In effusion-based approach, pore size and porosity
are obtained from the following relation:
RTe
C
127T77 rsNA\
where J h /C, permeability of PEGs; rs, radius of PEG
oligomer molecules; e, porosity of the membrane; X,
jump length (i.e, 3.1 A); 77, viscosity of water; R, gas
law constant; T, temperature; and NA, Avogadro's
number.
This equation predicts that the measured permeability, Jh/C, inversely is proportional to the radius of
the drug molecule, and from the slope of this relation,
J h /C = f(l/r s ), the porosity e can be evaluated. Furthermore, die extrapolation into die zero permeability gives
information for die critical value of the molecular radius still able to permeate into the hydrophilic pore of
the membrane. Number of the pores is obtained from
the relation e = Ah/A, where Ah is the effective surface
area of die hydrophilic pathways (A,, = mai,; m is the
number of paracellular routes in the area A, and ah is
the surface area of an individual orifice).
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
630
Investigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3
The molecular size (rs) and diffusion coefficient
(Doo) of PEGs in water are needed to calculate the
pore size and pore number. The radius of PEG molecules (rs) was calculated based on the radius of gyration (rg)22'23 for a spherical solute. Diffusion coefficients (Deo) of PEGs were based on the treatment of
Chin et al.24 The viscosity (77) of water was used, and
the jump length was determined by the solvent (water,
3.1 A). R, T, and NA had their usual meanings.
The corneal and conjunctival clearances (microliter/minute) for PEG 238 and 942 from donor compartment were calculated using equation CL = P X
S, where P is the corneal or conjunctival permeability
(centimeter/sec) and S is the corneal (1.59 cm2) or
conjunctival (15.13 cm2) surface area obtained.25 Consequently, the first-order rate constant for elimination
from tear compartment (K^ [%/minute]) was calculated from the relation K^,, = CL/Vd, where Vd is the
volume of PEG distribution in tear compartment (i.e.,
combined volumes of eyedrop and lacrimal fluid, 50
/zl). The expected noncorneal permeabilities of PEG
238 and 942 (i.e., through bulbar conjunctiva +
sclera) were calculated using equation 1/Pnc = 1/PCS
+ 1/PSC,26 where Pnc is the permeability in the noncorneal route, Pcs is the permeability of the bulbar conjunctiva, and Psc is the permeability of sclera.
The statistical significance between the groups was
tested using one-way analysis of variance. Values of P
< 0.05 were considered to represent statistically significant differences.
RESULTS
1.2l-
1
0.80.6-h
* 0.4"
60
r-f—r^TI
240
120
180
Time (min)
1 .<L
1"
£ 0.8S 0.6-
1 0.4-
0.2-
>—<
>—<>-*
t
C\ -t
0^
B
60
120
180
Time (min)
240
60
120
180
Time (min)
240
1.2-r
PEG oligomers permeated through the cornea, sclera
and bulbar, and palpebral conjunctiva at a constant
rate (Fig. 1). Because of the small fraction of PEG
permeation (<2%) during 4 hours, PEG concentration in the donor chamber did not change significantly during the experiment.
Cornea
Corneal permeability of PEGs decreased with increasing molecular weight (Fig. 2A). Permeability coefficient of PEG was 1.03 X 10~6 cm/second at molecular
weight 238, and almost five times smaller 0.22 X 10~6
cm/second at the molecular weight of 942. The permeability of PEGs decreased most rapidly at molecular
weights below 414. At higher molecular weights, the
permeabilities decreased slower, in particular above
500. The permeated fractions (percentage) of PEG
varied from 0.25% ± 0.05% (PEG 238) to 0.06% ±
0.03% (PEG 942), respectively (Fig. 2A), whereas Tlag
ranged from 47 ± 5 minutes to 96 ± 36 minutes.
Conjunctivas
In general, both palpebral and bulbar conjunctiva
were approximately 15 to 25 times more permeable
0
FIGURE 1. The penetration profiles (%) of polyethylene glycols with molecular weights of 282 (•) and 854 g/mol (O)
versus time (minutes). (A) Cornea, (B) sclera, and (C) palpebral conjunctiva. Means ± standard error of the mean.
than was the cornea (Figs. 2A, 2B, and 2C). Similarly,
permeabilities of hydrophilic /5-blockers,27 mannitol, 28
and fluorescein isothiocyanate-dextran (molecular
weight, 4400) 29 were shown to be higher in bulbar
conjunctiva than in cornea. In our study, permeabilit-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
Anterior Paracellular Routes in the Eye
i—1—Hh-H -
1.2 -
631
H
h-H -
1 J
1 0.8 I
0.6- -
- <^J
* < <
-_
1 0.4 -
-
--
^^=:^:%4;^
:
A
o -H h-H
238 326 414 502 590 678 766 854 942
Molecular weight (g/mol)
238 326 414 502 590 678 766 854 942
Molecular weight (g/mol)
B
238 326 414 502 590 678 766 854 942
Molecular weight (g/mol)
238 326 414 502 590 678 766 854 942
Molecular weight (g/mol)
FIGURE 2.
The permeability coefficients (cm/sec) (•) and penetrated amounts (%) of PEGs
(•) in the molecular weight range of 238 to 942 g/mol. Mean ± standard error of the
mean. (A) Cornea. (B) Palpebral conjunctiva. (C) Bulbar conjunctiva, and (D) sclera. The
percentage values are not comparable between membranes, because corneas had different
exposed area than did conjunctivas and scleras.
ies of PEGs in palpebral and bulbar conjunctiva were
equal.
Although increasing molecular weight decreased
the penetration of PEG in both palpebral and bulbar
conjunctiva, the effect of molecular weight was minor
(approximately twofold to threefold), and statistically
significant differences were seen only between molecular weight extremes 238, 898, and 942 (analysis of variance, P < 0.05) in bulbar conjunctiva (Figs. 2B, 2C).
Sclera
In sclera, the permeability of PEGs between molecular
weights 238 and 942 decreased nearly three times.
PEG 238 had the permeability coefficient of 8.80 X
10"6 cm/second and PEG 942 3.08 X 10"6 cm/second
(Fig. ID). The permeability of PEGs decreased linearly
with increasing molecular weight, and the permeated
amounts of PEG 238 and PEG 942 were 0.53% ±
0.15% and 0.19% ± 0.08%, respectively.
The estimated permeability of the noncorneal
route (bulbar conjunctiva + sclera) was calculated to
be 6.25 X 10"6 cm/second for PEG 238 and 2.03 X
10 6 cm/second for PEG 942. Compared to that of
the cornea, the scleral permeabilities of PEG 238 and
PEG 942 were six times and nine times higher, respectively.
Epithelial Pore Size and Pore Density
Conjunctival epithelium has larger (P< 0.05) paracellular pores than does the corneal epithelium. The
pore diameter in the apical rate-limiting cell layer of
corneal epithelium was 2.0 nm ± 0.2 (n=7). In palpebral conjunctiva and bulbar conjunctiva, the pore diameters were 4.9 nm ± 2.5 and 3.0 nm ± 1.6,
respectively. For peptides with molecular weights <550
(tetrapeptides and smaller), the molecular diameter
is approximately 1.0 to 2.0 nm,30 whereas lysozyme
(molecular weight, 14,000) has a diameter of 4.1 nm."
Therefore, the paracellular space of the conjunctiva
appears to be large enough for permeation of small
peptides and oligonucleotides (molecular weight,
5000 to 10,000). In contrast, the pores in the cornea
allow paracellular permeation of only small molecules
(molecular weight, <500).
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
632
Investigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3
Porosities of the corneal and conjunctival ratelimiting layers were 0.13 X 10~6 and 1.71 X 10~6, respectively, and the number of pores in the corneal
and conjunctival surfaces was 4.3 X 106/cm2 and 20.2
X 106/cm2, respectively. The intercellular spaces occupy only a fraction of 30 X 10~7 and 2 X 10"7 of
the total surface area of the conjunctival and corneal
epithelia, respectively. Consequently, in the cornea
(area, 1.59 cm2), the total intercellular space is 0.03
X 10~3 mm2 and in the conjunctivas (area, 15.1 cm2),
it is 7 X 10~3 mm2, respectively. For comparison, it
was estimated that in the highly permeable intestinal
epithelium, the intercellular spaces occupy approximately 0.01% of the total surface area of the epithelium (i.e., fraction of 1000 X 10~7).32'33 The paracellular corneal clearance values for PEG 238 and PEG 942
from tear compartment were estimated to be 0.1 [A/
minute (Ke, = 0.20%/minute) and 0.02 /zl/minute
(Kel = 0.04 %/minute), respectively (assuming Vd =
50 /il). Conjunctival (palpebral) clearances of the PEG
238 and 942 from the lacrimal fluid were calculated
to be in vivo 15.0 /il/minute (Kei = 30%/minute) and
6.3 //1/minute (K^ = 12.7%/minute), respectively, assuming 50-/il volume in tear fluid and contact with the
entire membrane. Because of the great paracellular
porosity and permeability, the paracellular elimination from the tear fluid through conjunctiva is expected to be much greater than the elimination
through cornea.
DISCUSSION
The permeability result of cornea is comparable with
the reported progressively decreasing urinary recovery
of PEG between molecular weights 200 and 400 after
dosing of PEG as a 5% (wt/vol) solution into the stomach of rat.15'34 The result, however, differs from the
corneal permeation data of Liaw and Robinson,35 who
reported essentially constant apparent permeability
coefficients for PEGs 206 to 415 and substantial decrease in the permeability at higher molecular
weights. The differences may be because of different
experimental and analytical methods. Liaw and Robinson35 used approximately 10 times greater initial
PEG concentrations in the donor phase, the sample
bathing medium was 0.16 M sodium chloride, and
the analytic method was HPLC with refractometric
detection. In concentrated solutions, the PEGs are in
states of association, which differ from the individual
molecular species, whereas in dilute solutions, PEG
chains are expanded.36 It also is known that the salts
in the aqueous solution may affect the conformation
of PEG chains.3788
The results of permeability studies with rabbit conjunctivas show that neutral, hydrophilic molecules
with molecular weights below 1000 are able to perme-
ate equally through bulbar and palpebral conjunctiva,
and, further, the solute size has not pronounced effect
on permeability. Therefore, in terms of epithelial permeability, absorption of hydrophilic compounds to
the palpebral side is not favored over bulbar side. In
vivo, however, there are other factors (e.g., surface
area, metabolic activity, vascularization) that may affect drug absorption. It is known that more than 20%
to 50% of the instilled small molecular weight drug
absorbs systemically through palpebral conjunctiva.39'40 Systemic absorption of [D-Ala]2-metenkephalinamide (molecular weight, 647)41 and insulin (molecular weight, 5778)42 through conjunctiva was 36%
and 1%, respectively.
The permeability of sclera was approximately half
of that in the conjunctiva and approximately order of
magnitude greater than in the cornea. The sclera has
been shown to be more permeable than is the cornea
for inulin43 and hydrophilic beta-blockers.27 Previously, Huang et al9 reported that solute size has more
pronounced effect on scleral permeability than on
corneal permeability, so that sucrose (molecular
weight, 342) permeated eightfold faster than did inulin (molecular weight, 5000) in the cornea, whereas
in sclera, the difference was 16-fold. In our study, the
molecular weight had a most pronounced effect on
the corneal permeability, and in sclera, the permeability was less affected by the molecular weight. Our observations are in line with the data of Ahmed and
Patton,43 which show importance of noncorneal route
in the ocular absorption of hydrophilic inulin.
Intercellular space was characterized in the cornea and conjunctiva. In these membranes, the ratelimiting barrier for hydrophilic compounds is the
thin, most-apical cell layer in the epithelium, and anatomically, the intercellular space is much smaller than
the area of the cells.611 Sclera does not fulfill these
criteria.
The values for the critical pore size and porosity of
the ocular surface based on controlled series of permeating analogues are the first quantitative determinations.
These values are in line with published permeation of
molecules. Because the surface of the corneal epithelium is impermeable to horseradish peroxidase (molecular weight, 40,000) ^ and fluorescein isothiocyanatedextran (molecular weight, 20,000),29 the intercellular
spaces of the apical layer of corneal epithelium should
be smaller than 3 nm. Furthermore, particles of 5 nm
do not penetrate to the cornea.45 Other previous estimates for the pore size in the corneal epithelium are
the molecular diameters of glycerol (1.2 nm)6 and inulin
(3.0 nm).3 No attempts to estimate the pore size or
porosity of conjunctiva have been reported.
In conclusion, using effusion-like calculations and
polyethylene-glycol permeation, we determined the
drug absorption-limiting pore sizes and porosities of
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
633
Anterior Paracellular Routes in the Eye
the ocular surface tissues. Both intercellular pore size
and pore density in the cornea are much smaller than
in the conjunctiva. Therefore, bulbar conjunctiva and
sclera may constitute a more viable route for ocular
peptide and oligonucleotide delivery than would the
cornea.
14.
15.
Key Words
intercellular space, membrane porosity, mol<
molecular weight,
pafinn nnivpfhvipnp
crlvrol
paracellular permeation,
polyethylene glycol
16.
Acknowledgments
The authors thank Mr. Jukka Knuutinen, Mr. Pekka Suhonen, and Mr. Petri Nykvist for their skillful assistance.
17.
References
1. Doane MG, Jensen AD, Dohlman CH. Penetration
routes of topically applied eye medications. Am] Opthalmol 1978; 85:383-386.
2. Lee VHL. Review. New directions in the optimization
of ocular drug delivery. J Ocul Pharmacol Ther.
1990;6:157-162.
3. Lee VHL, Carson W, Takemoto KA. Macromolecular
drug absorption in the albino rabbit eye. IntJPharmacol. 1986; 29:43-51.
4. Chien D-S, Sasaki H, Bundgaard H, Buur A, Lee VHL.
Role of enzymatic lability in the corneal and conjunctival penetration of timolol ester prodrugs in the pigmented rabbit. Pharmacol Res. 1991;8:728-733.
5. Schoenwald RD. Ocular drug delivery. Pharmacokinetic considerations. ClinPharmacokinet. 1990; 18:255269.
6. Grass GM, Robinson JR. Mechanisms of corneal penetration I: In vivo and in vitro kinetics. J Pharmacol Sci.
1988; 77:3-14.
7. Stratford RE, Carson LW, Dodda-Kashi S, Lee VHL.
Systemic absorption of ocularly administered enkephalinamide and insulin in the albino rabbit: Extent,
pathways, and vehicle effects. J Pharmacol Sci.
1988; 77:343-357.
8. Rojanasakul Y, Paddock SW, Robinson JR. Confocal
laser scanning microscopic examination of transport
pathways and barriers of some peptides across the cornea. Int J Pharmacol. 199O;61:163-172.
9. Huang AJW, Tseng SCG, Kenyon KR. Paracellular permeability of corneal and conjunctival epithelia. Invest
Ophthalmol Vis Sci. 1989; 30:684-689.
10. Wei Z-G, Wu R-L, Lavker RM, Sun T-T. In vitro growth
and differentiation of rabbit bulbar, fornix, and palpebral conjunctival epithelia. Invest Ophthalmol Vis Sci.
1993;34:1814-1828.
11. Pfister RR. The normal surface of conjunctiva epithelium: A scanning electron microscopic study. Invest
Ophthalmol Vis Sci. 1975; 14:267-279.
12. HochmanJ, Artursson P. Mechanisms of absorption
enhancement and tight junction regulation. J Control
Rel. 1994; 29:253-267.
13. MaxfieldJ, Shepherd IW. Conformation of polyethyl-
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
ene oxide) in die solid state, melt and solution measured by Raman scattering. Polymer. 1975; 16:505-509.
Tagesson C, Andersson P-A, Andersson T, Bolin T,
Kallber M, Sjodahl R. Passage of molecules through
die wall of the gastrointestinal tract. Scand J Gastroenterol. 1983; 18:481-486.
Donovan MD, Flynn GL, Amidon GL. Absorption of
polyethylene glycols 600 through 2000: The molecular
weight dependence of gastrointestinal and nasal absorption. Pharmacol Res. 1990; 7:863-868.
Ruddy SB, Hadzija BW. Iontophoretic permeability of
polyethylene glycols through hairless rat skin: Application of hydrodynamic theory for hindered transport
through liquid-filled pores. Drug- Des Discov. 1992;
8:207-224.
O'Brien W, Edelhauser HF. The corneal penetration
of trifluorothymidine, adenine arabinoside, and idoxuridine: A comparative study. Invest Ophthalmol Vis Sci.
1977;16:1093-1103.
Schoenwald RD, Huang H-S. Corneal penetration behavior of beta-blocking agents. I: Physicochemical factors. J Pharmacol Sci. 1983; 72:1266-1272.
Escott REA, Mortimer N. Analysis of polyethylene glycol and derivatives by high performance liquid chromatography using elevated temperatures and low
wave-length ultraviolet detection, and supercritical
fluid chromatography. J Chromatogr. 1991; 553:423432.
Auriola S, Ronkko KM, Urtti A. Determination of polyediylene glycols by high-performance liquid chromatography thermospray mass spectrometry. JPharm BiomedAnal. 1993; 11:1027-1032.
Hamalainen KM, Kontturi K, Auriola S, Murtomaki
L, Urtti A. Estimation of pore size and pore density
of biomembranes from permeability measurements of
polyethylene glycols using an effusion-like approach.
J Control Rel. 1997; in press.
Zeman L, Wales M. Steric rejection of polymeric solutes by membranes with uniform pore size distribution. Sep Sci Technol. 1981; 16:275-290.
Deen WM. Hindered transport of large molecules in
liquid-filled pores. AlCheJ. 1987; 33:1409-1425.
Chin KP, Li SFY, Yao YJ, Yue LS. Infinite dilution diffusion coefficients of poly(ethylene glycol) and poly(propylene glycol) in water in the temperature range
303-318 K/Chem EngData. 1991;36:329-331.
Watsky MA, Jablonski MM, Edelhauser HF. Comparison of conjunctival and corneal surface areas in rabbit
and human. Curr Eye Res. 1988; 7:483-486.
Flynn GL, Yalkowsky SH, Roseman TJ. Review: Mass
transport phenomena and models: Theoretical concepts. J Pharmacol Sci. 1974; 63:479-510.
Sasaki H, Igarashi Y, Nagano T, Nishida K, Nakamura
J. Different effects of absorption promoters on corneal
and conjunctival penetration of ophthalmic betablockers. Pharmacol Res. 1995a;12:1146-1150.
Ahmed I, Gokhale RD, Shah MV, Patton TF. Physicochemical determinants of drug diffusion across the
conjunctiva, sclera and cornea. / Pharmacol Sci. 1987;
76:583-586.
Sasaki H, Yamamura K, Nishida K, NakamuraJ. Ocular
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017
634
30.
31.
32.
33.
34.
35.
36.
37.
Investigative Ophthalmology & Visual Science, March 1997, Vol. 38, No. 3
permeability of FITC-dextran with absorption promoter for ocular delivery of peptide drug. JDrug Target. 1995; 3:129-135.
Conradi RA, Hilgers AR, Ho NFH, Burton PS. The
influence of peptide structure on transport across
CaCo 2-cells. Pharmacol Res. 1991;8:1453-1460.
Hennink WE, Talsma H, Borchert JCH, De Smedt
SC, DemeesterJ. Controlled release of proteins from
dextran hydrogels. / Control Rel. 1996;39:47-55.
Madara JL, Trier JS. Functional morphology of the
mucosa of the small intestine. In: Johnson LR, ed.
Physiology of the Gastrointestinal Tract. 2nd ed. New York:
Raven Press; 1977.
Rojanasakul Y, Wang L-Y, Bhat M, Glover DD, Malanga
CJ, Ma JKH. The transport barrier of epithelia: A comparative study on membrane permeability and charge
selectivity in the rabbit. Pharmacol Res. 1992;9:10291034.
Hollander D, Koyama S, Dadulfalza V, et al. Polyethylene glycol 900 permeability of rat intestinal and coIonic segments in vivo and brush border membrane
vesicles in vitro. JLab Clin Med. 1989; 113:505-515.
LiawJ, Robinson JR. The effect of polyethylene glycol
molecular weight on corneal transport and the related
influence of penetration enhancers. Int J Pharmacol.
1992;88:125-140.
Liu K-J. Nuclear magnetic resonance studies of polymer solutions. IV: Polyethylene glycols. Macromolecules
1969;2:529-533.
Bailey FE, Koleske JV. Configuration and Hydrodynamic
38.
39.
40.
41.
42.
43.
44.
45.
Properties of the Polyoxyethylene Chain in Solution. New
York: Marcel Dekker, Inc; 1967:794-822.
Graessley WW. Polymer chain dimensions and the dependence of viscoelastic properties on concentration,
molecular weight and solvent power. Polymer. 1980;
21:258-262.
Urtti A, Salminen L, Miinalainen O. Systemic absorption of ocular pilocarpine is modified by polymer matrices. Inl J Pharmacol. 1985;23:147-161.
Chang SC, Lee VHL. Nasal and conjunctival contributions to the systemic absorption of topical timolol in
the pigmented rabbit: Implications in the design of
strategies to maximize the ratio of ocular to systemic
absorption. / Ocul Pharmacol. 1987;3:158-167.
Stratford RE, Carson LW, Dodda-Kashi S, Lee VHL.
Systemic absorption of ocularly administered enkephalinamide and inulin in the albino rabbit: Extent,
pathways, and vehicle effects. J Pharmacol Sci. 1988;
77:838-842.
Yamamoto A, Luo AM, Dodda-Kashi S, Lee VHL. The
ocular route for systemic insulin delivery in the albino
rabbit, f Pharmacol Exp Ther. 1989;249:249-255.
Ahmed I, Patton TF. Importance of the non-corneal
route in topical ophthalmic drug delivery. Invest Ophthalmol Vis Sci. 1985;26:584-587.
Tonjum AM. Permeability of horseradish peroxidase
in the rabbit corneal epithelium. Ada Ophthalmol.
1974;52:650-658.
Grass GM, Robinson JR. Mechanisms of corneal penetration II: Ultrastructural analysis of potential pathways for drug movement. /Pharmacol Sci. 1988; 77:1523.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933197/ on 05/10/2017