Biophysical aspects of epithelial
adhesion to stroma
Frank J. Holly
The affinity of various corneal boundary surfaces, such as mucus-coated epithelium, demucinized epithelium, denuded baseinent membrane, and keratectomized cornea, to water and
hydrophobic liquids was determined by contact angle goniometry on enucleated rabbit eyes.
The mucus-coated surface and the bare stromal surface without the basement membrane were
found to be more hydrophilic than the surfaces of the epithelium and the basement membrane.
If at least one of the surfaces forming a joint boundary is hydrophilic, then conditions become
favorable for the accumidation of water at the interface, forming a weak boundary layer that is
detrimental to adhesion, due to its low shear resistance. Thus, even in a normcd eye, mucus
does not adhere strongly to the epithelium; when the basement membrane is damaged or absent, the epithelium does not adhere well to the stroma. External dehydrating factors such as
hyperosmotic conditions favor adhesion, whereas edematic conditions have an abhesive (adhesion-decreasing) effect. Experimental ophthalmic solutions containing macromolecules at
hyperosmotic concentrations may demonstrate a therapeutic effect when applied to eyes with
recurrent epithelial erosion or other epithelial trauma.
Key words: epithelium, adhesion, cornea, epithelial defects
ecurrent epithelial erosion is a puzzling
R
and frustrating phenomenon for the physician and a painful, incapacitating experience
for the patient. In this condition, epithelial
defects form over the cornea, and may persist
indefinitely. Even when such defects heal,
the newly formed epithelium can be easily
detached, often by the shear effected by the
lid motion. The adhesiveness of the corneal
epithelium to the underlying stroma or
basement membrane appears to be greatly
diminished. Some in vivo work on epithelial
adhesion has been reported, the results of
From the Department of Cornea Research, Eye Research Institute of Retina Foundation, and Harvard
Medical School, Boston, Mass.
This work was supported in part by Research Grants
EY-01381 and RR-05527 from the National Institutes
of Health and by Holies Laboratories, Inc.
Submitted for publication Oct. 26, 1977.
Reprint requests: Dr. Frank J. Holly, Eye Research Institute of Retina Foundation, 20 Staniford St., Boston,
Mass. 02114.
552
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
which suggest that the strength of the adhesive joint between the epithelium and the
stroma depends on the condition of the
basement membrane.1
Aqueous boundary film as a cause of
adhesive failure.
In industrial practice, the presence of a
weak boundary layer at the interface of the
adhesive and the adherend is often the primary cause of adhesive failure.2 Such an abhesive (antonym of adhesive) layer could consist of trapped gas, adsorbed water, or lipid
contaminants. In biological systems, where
both water and lipids abound and where the
surface free energy of the boundaries can be
changed by subtle variations in molecular
configurations at the interface, a weak boundary layer could readily develop under a
given set of conditions.3
The stability of a thin aqueous film between two approaching solid surfaces immersed in water depends on the way the film
0146-O404/78/0617-0552S00.60/0 © 1978 Assoc. for Res. in Vis. and Ophthal., Inc.
Volume 17
Number 6
energy (i.e., the sum of the two interfacial
tensions of the aqueous layer) varies with the
decreasing distance between the solids. This
quantity is known as the disjoining pressure. 4
When the disjoining pressure is positive,
as is the case when two hydrophilic solids
are separated by an aqueous film, the film resists thinning, and considerable pressure is
needed to bring the two solids into close contact. On the other hand, when the solid surfaces are hydrophobic, the disjoining pressure of the aqueous film between them is
negative. In this case, the aqueous film spontaneously recedes from between the approaching hydrophobic solids. Actually, the
conditions for the formation and retention of
an aqueous weak boundary layer between
two solids are usually favorable even if only
one of the surfaces is hydrophilic. In such
cases, poor adhesion is observed.
Wettability of corneal tissue interfaces
We have studied the affinity of various
corneal boundaries to water and hydrophobic
liquids in order to estimate the tendency of
these tissue surfaces to attract and accumulate water. Fig. 1 shows a schematic view of
the surfaces studied, together with the magnitude of the advancing contact angle of
water. Instead of water, we used a 5% aqueous solution of dextran—a high-molecularweight polysaccharide with low viscosity and
negligible surface activity—which, because
its colloidal osmotic pressure equals the imbibition pressure of the stroma, avoids swelling of the substrate.
The most wettable surface was the normal
cornea (rabbit) which is ordinarily covered
with a thin mucous layer. This mucous layer,
usually together with one or two cell layers of
superficial epithelium, was removed by gentle scraping. The de-mucinized epithelial
surface was the most hydrophobic—the least
wettable by the aqueous solution. The bare
basement membrane, obtained after complete removal of the epithelium, was almost
as hydrophobic as the epithelium. Even the
bare stroma under the basement membrane
was more hydrophobic than the mucus layer
but much less so than the basement mem-
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Biophysics of epithelial adhesion to stroma
BIOSURFACE
NORMAL CORNEA
553
WATER CONTACT ANGLE
•£"
epithelial surface-*-CiVL
DEMUCINIZED CORNEA
•*?
basemen) membrane-*DE-ERTHELIAUZED
CORNEA
tore stroma ^ » ,
KERATECTOM1ZEO
CORNEA
K"
,J\
^ ^
J N
% •
Fig. 1. Schematic view of wettability of corneal
surfaces investigated.
brane. For a detailed description of contact
angle measurement on gel-like surfaces, consult Holly and Refojo.5
The adhesion tension of a liquid to a solid
is defined as the difference between the
solid-vapor and the solid-liquid interfacial
tensions and is equal to the product of the
liquid surface tension and the cosine of the
contact angle in cases of incomplete wetting.
Fig. 2 shows the adhesion tension between
the comeal boundary surfaces and water,
methylene iodide, and tricresyl phosphate.
Both methylene iodide and tricresyl phosphate are immiscible with water, except that
the latter liquid spontaneously spreads over
water whereas the former does not. Our results show that the hydrophobic liquids
exhibit much less adhesion toward these surfaces than does water. The adhesion tension
of tricresyl phosphate is about the same for
all corneal surfaces. The same is true for
methylene iodide except in the case of demucinized epithelium, where its adhesion
tension is somewhat lower.
The data obtained with the hydrophobic
liquids can be used to estimate the critical
surface tension of Zisman6 for these cornea!
surfaces, since the slope (3 in the following
equation
COS 6 = 1 - $(y]v
-
y e)
is known to be relatively invariant" and its
554
Invest. Ophthalmol. Visual Sci.
June 1978
Holly
0 water Cdextran soln)
• methylene iodide
• tricresyl phosphate;
Fig. 2. Adhesion tension of liquids on corneal surfaces.
magnitude has been determined previously.7
The critical surface tension, yc, for the normal cornea is 38 dyne/cm, which is the same
value obtained for adsorbed submaxillary
mucin. 7 Basement membrane and denuded
stroma yielded similar values, indicating that
the hydrophobic structures of these tissue
boundaries are surface chemically similar. It
is likely that at all these surfaces, glycoprotein molecules dominate and that water wettability differences are due to the varying
amount and accessibility of the carbohydrate
prosthetic groups in the glycoprotein molecules. Demucinized corneal epithelium has
a low critical surface tension, characteristic of a lipid-rich surface.8 Fig. 3 shows the
method used and the results obtained.
Contact angle hysteresis on
corneal interfaces
The relative contact angle hysteresis, defined as the difference between the advancing and the receding contact angles divided
by the advancing contact angle, was also obtained for these corneal boundary surfaces.
Fig. 4 shows the results obtained with water
and methylene iodide.
Water exhibits maximal hysteresis on all
interfaces, which reflects the fact that all receding contact angle values obtained were
equal to zero. Such a large hysteresis of contact angle cannot be due to surface roughness
alone. This assumption is supported by the
fact that the relative contact angle hysteresis
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
is much lower with methylene iodide. Furthermore, high contact angle hysteresis was
also observed for water on smooth synthetic
hydrogels,9 where it was attributed to conformational changes in the surface polymer
chains from hydrophobic to hydrophilic
configurations induced by the proximity of
the water phase. Since methylene iodide
cannot form hydrogen bonds, it is likely that
the contact angle hysteresis observed with
this substance was caused mainly by surface
roughness whereas the higher relative contact angle hysteresis obtained with water was
caused by surface roughness as well as surface stereochemical changes. Fig. 4 also contains hysteresis data obtained with crosslinked poly(acrylamide) and poly(hydroxyethyl methacrylate) hydrogels with 78% and
40% equilibrium water content, respectively.
The contact angle data obtained with
methylene iodide were used to calculate the
apparent surface roughness, r, of the corneal
and hydrogel surfaces, employing a modified
form of the Wenzel equations:10
r = cos0R/cos0A
assuming that dA ~ dtme
where r is the ratio of the actual surface area
and the geometric apparent surface area and
6A and OR are the advancing and receding
contact angles, respectively. The calculated
values are shown in Fig. 4. The roughness
factor appears to be about the same for all
corneal surfaces except for the demucinized
Volume 17
Number 6
Biophysics of epithelial adhesion to stroma
30
555
40
LIQUID SURFACE TENSION (dynes/cm)
Fig. 3. Approximate critical surface tension of corneal layers.
'apparent roughness factor: r = cose R / cos8 A
I water (dextran sol'n)
I methylene iodide
1.0
5.68
or
1.77
1.83
1.84
1.69
Fig. 4. Relative contact angle hysteresis on corneal surfaces and hydrogels.
epithelium, where it is much higher, due
probably to its cellular structure and the demucinizing method employed.
Adhesion and abhesion of corneal
tissue layers
The presence of mechanical interlinking in
the form of hemidesmosomes at the adhesive
joint of two adjacent biological tissues is well
established,11 and the strength of the adhesive biological joints is often attributed to
these anatomical structures. Actually, the
small cohesive strength of these structures
and their relatively small area of contact render these cell-coupling structures secondary
in importance in determining the stress resistance of tissue joints.
We propose that the adhesive joint of two
adjacent tissue layers such as corneal epithelium and stroma, or retina and choroid, can
be represented by a simple model consisting
of two adherends that are joined by a thin
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
layer of viscous adhesive. In such a system,
the resistance to separation depends only on
the rheological properties of the adhesive
layer.
Specifically, the integrity of the adhesive
bond in such systems is determined by its
tackiness, which is the product of the amount
of stress and the length of time during which
it must be applied in order to achieve separation. When the stress is created by a pulling
force, the tackiness varies inversely with the
second power of the thickness of the adhesive
layer and is linearly proportional to the viscosity of the adhesive. When one adherend is
peeled off, a different relation is obtained.12
When a tangential shear is applied, such as
lid motion, the integrity of the adhesive joint
is determined by the shear resistance of the
intertissue adhesive. Even a thin fluid layer
has a negligibly small shear resistance provided that its viscosity is low.
Hence, if water were to accumulate be-
556
Invest, Ophthalmol. Visual Sci.
June 1978
Holly
NORMAL
POOR
EPITHELIAL ADHESION
MUCOUS LAYER (Y)
EPITHELIUM (XI
BASEMENT MEMBRANE (X>
S T R O M A
THIN LAYER OF DEHYDRATED
INTERTISSUE ADHESIVE
DILUTED AND THICKENED
INTERTISSUE ADHESIVE
hiqh viscosity
low viscosity
(X) relatively hydrophobic
(Yl relatively hydrophilic
Fig. 5. Weak-boundary-layer model of epithelial adhesion.
tween the epithelium and the stroma, diluting the intertissue adhesive, thus lowering its
viscosity and increasing its thickness, the
shear resistance of the adhesive joint would
be greatly diminished. Fig. 5 illustrates both
good and poor tissue adhesion for the corneal
epithelium.
In light of present knowledge concerning
transport, rheological, and surface chemical
properties of the tissues involved, it is
possible to enumerate the factors that may
improve the adhesion of epithelium and
those that tend to diminish it.
Positively contributing to the adhesive effect are (1) the hydrophobic nature of the epithelium and the basement membrane, (2) the
transport of water toward the tear film by the
epithelium, (3) the imbibition pressure of the
stroma, and (4) the hyperosmolality of macromolecules in the tear film and at the tearepithelium interface. All these factors tend to
dehydrate the epithelium-stroma adhesive
joint.
The following would have an abhesive effect: (1) absence of or damage to the basement membrane, which would result in mak-
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
ing one side of the joint more hydrophilic, (2)
diminished epithelial pumping, (3) edematous cornea, and (4) the hypo-osmolality of
the tear film.
Possible clinical applications
If the presence of an aqueous weak boundary layer adversely affects epithelial adhesion, and perhaps healing, then an effective
colloidal dehydrating agent which would not
penetrate traumatized epithelium and could
not even enter denuded stroma would have a
therapeutic value. Such colloidal preparations have been tested abroad in patients
suffering from corneal edema complicated by
epithelial discontinuity. Such preparations
have been generally effective when applied
in patients with recurrent epithelial erosion,
superficial corneal trauma, and dendritic
keratitis, all accompanied by stromal edema.
Conclusions
The basement membrane of the stroma,
like the demucinized epithelium, is relatively hydrophobic, which accountsforthe
maintenance of a fairly dehydrated adhesive
Volume 17
Number 6
joint at the boundary of these two tissues. If
fluid removal is inadequate or if hydrophobic
property or low water permeability is lost,
fluid accumulation at this boundary may take
place, forming a weak boundary layer of low
shear resistance. Such a state could lead to
recurrent epithelial loss or to an epithelial
layer of poor integrity. In this case, stromal
and epithelial edema could not be treated
effectively with a topical hypertonic salt solution, since the electrolyte itself would readily
penetrate the epithelium and accumulate in
the stroma. Isotonic aqueous solutions containing certain macromolecular dehydrating
agents may have significant therapeutic value
when applied topically for a number of conditions involving corneal edema in the presence of epithelial discontinuities.
Biophysics of epithelial adhesion to stroma 557
4.
5.
6.
7.
8.
9.
Helpful comments by David W. Lamberts and editorial assistance by S. Flavia Blackwell are hereby acknowledged.
REFERENCES
1. Khodadoust, A. A., Silverstein, A. M., Kenyon, K.
R., and Dowling, J. E.: Adhesion of regenerating
corneal epithelium. The role of the basement membrane, Am. J. Ophthalmol. 65:339, 1968.
2. Bikerman, J. J.: The tackiness of liquid adhesives,
Trans. Soc. Rheol. 1:3, 1957.
3. Baier, R. E., Shafrin, E. G., and Zisman, W. A.:
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
10.
11.
12.
Mechanisms assisting and impeding adhesion in
biological systems, NRL Report 6691, Naval Research Laboratory, Washington, D. C , 1968.
Derjaguin, B. V.: On the repulsive forces between
charged colloid particles and on the theory of slow
coagulation and stability of lyophobe sols, Trans.
Faraday Soc. 36:203, 1940.
Holly, F. J., and Refojo, M. F.: Wettability of hydrogels. I. Poly(2-hydroxyethyl methacrylate), J.
Biomed. Mater. Res. 9:315, 1975.
Zisman, W. E.: Relation of equilibrium contact
angle to liquid and solid constitution. In Fowkes,
F. M., editor: Contact Angle, Wettability, and Adhesion, Adv. Chem. Ser. No. 43, Washington,
D.C., 1964, American Chemical Society, p. 3.
Holly, F. J., and Lemp, M. A.: Wettability and wetting of corneal epithelium, Exp. Eye Res. 11:239,
1971.
Lemp, M. A., Holly, F. J., Iwata, S., and Dohlman,
C. H.: The precorneal tear film. I. Factors in
spreading, Arch. Ophthalmol. 83:89, 1970.
Holly, F. J., and Refojo, M. F.: Water wettability of
hydrogels. In Andrade, J. D., editor: Hydrogels for
Medical and Related Applications, ACS Symposium
Series No. 31, Washington, D.C., 1976, American
Chemical Society, p. 252.
Adamson, A. W.: Physical Chemistry of Surfaces,
ed. 2, New York, 1968, Interscience Publishers, p.
358.
Goldman, J. N., and Kuwabara, T.: Histopathology
of corneal edema, Int. Ophthalmol. Clin. 8:561,
1968.
/
Zauberman, H., deGuillebon, H., and Holly, F. J.:
Retinal traction in vivo: biophysical aspects, INVEST.
OPHTHALMOL. 11:46, 1972.