Download Structure and composition of rat precorneal tear film. A study

Structure and Composition of Rat Precorneal Tear Film
A Study by an In Vivo Cryofixation
Hai-Bo Chen, Shigeki Yamabayashi, Bo Ou, Yuko Tanaka, Shinichi Ohno,*
and Shigeo Tsukahara
Purpose. To visualize the in vivo structure and to investigate the composition of rat precorneal
tear film.
Methods. An in vivo cryofixation with freeze substitution method of electron microscopy was
used for the study. For light and transmission electron microscopy, a small amount of aluminum powder was used as a tracer spread on the corneal surface. The eyeballs were immediately
and quickly frozen by pouring an isopentane-propane mixture cooled by liquid nitrogen
directly over the eyes. For scanning electron microscopy, the corneal surface was freezefractured after the cryofixation. The specimens were then freeze-substituted and prepared
conventionally for microscopic observation.
Results. The tear film appeared as a layer of homogeneous and fine network-like structures
varying from 2 to 6 //m in thickness on the corneal surface, with a membrane-like layer
covering its surface. The aluminum powder was located on the surface of the tear film. The
tear film could be removed completely by applying 10% or 20% acetylcysteine, but not by
phosphate buffer.
Conclusions. The in vivo structure of the rat tear film is composed primarily of mucus, with a
lipid layer covering its surface but without a free aqueous layer. The "three layers theory" of
tear film structure requires revisions. Invest Ophthalmol Vis Sci. 1997; 38:381-387.
1. he optical integrity and normal function of the eye
can not be maintained without the presence of the
precorneal tear film, a very thin fluid film on the exposed part of the ocular surface. The precorneal tear
film serves an optical function by maintaining an optically uniform corneal surface, a mechanical function
by flushing cellular debris and foreign matter from
the cornea and conjunctival sac and by lubricating
the surface, a corneal nutritional function, and an
antibacterial function. 12 In spite of these important
functions, however, little information is available
about the morphology of the tear film. It is generally
thought that it consists of three layers: the superficial
lipid layer (less than 0.1 fxm in thickness) from Meibomian secretion; the middle aqueous layer (approxiFrom the Department of Ophthalmology and the * Department of Anatomy,
Yamanashi Medical University, Tamaho, Yamanashi, 409-38 Japan.
Supported by a Scientific Research Grant ofJapan Monbusyou (No. 07407050).
Submitted for publication June 17, 1996; revised September 3, 1996; accepted
September 16, 1996.
Proprietary interest categmy.N
Reprint requests: Shigeo Tsukahara, Department of Ophthalmology, Yamanashi
Medical University, Tamaho, Yamanashi, 409-38 Japan.
mate 7 /xm), which is the main component, secreted
by the lacrimal gland and the accessory glands of
Krause and Wolfring; and the deepest mucus layer
(0.02 to 0.05 fim) elaborated by goblet cells of the
conjunctiva.1'2 This "three layers theory" of the structure and composition of tear film was proposed initially by Wolff,3 based on histologic studies of the eye
surface and on observation of the corneal surface with
a slitlamp. Many investigators have studied and measured the tear film with various methods.4"8 No study
has so far provided enough evidence to confirm the
three layers theory or demonstrated the entire structure of the precorneal tear film morphologically, because the tear film could not be preserved by conventional chemical fixation.910 Accordingly, the three layers theory of tear film structure remains unconfirmed.
Recent studies of the tear film have reported results that refute the three layers theory and suggest
that this theory has to be revised.11"17 We have reported that a layer of homogeneous material considered to be tear film could be preserved on the rat
corneal surface by an in vivo cryofixation with a freeze
Investigative Ophthalmology & Visual Science, February 1997, Vol. 38, No. 2
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
381
382
Investigative Ophthalmology & Visual Science, February 1997, Vol. 38, No. 2
substitution (VC-FS) method of electron microscopy,9'10 but whether that was its entire structure remains unresolved. In the current study, we again used
the VC-FS method to confirm the entire structure of
the tear film with a tracer, to investigate the composition of tear film by a rnucolytic agent, and to estimate
the thickness of the tear film.
•
/
METHODS
Animal experiments conformed with the recommendations of the ARVO Statement for the Use of Animals
in Ophthalmic and Vision Research. Nineteen Wistar
rats (19 eyes) at the age of approximately 15 weeks
were used for the study. Sodium pentobarbital was
injected intraperitoneally (4 mg/100 g) to anesthetize
the rats for the experiments.
Specimens Prepared by the VC-FS Method
The eyelids were artificially opened not beyond a physiologic width after anesthesia. A small amount of aluminum powder, used as a tracer, was spread over the
surface of five eyes (five rats) to locate the surface of
the tear film. Then 10 eyeballs of 10 rats (including
the above five eyes) were quickly frozen and freezesubstituted by the VC-FS method described previously.9'10 Then, for light microscopy (LM) and transmission electron microscopy (TEM), the five eyes that
had the aluminum powder spread on them were embedded in epoxy resin and semithin or ultrathin sections were cut on a microtome (MT2-B; Sorvall). After
staining the semithin sections with toluidine blue and
the ultrathin sections with uranyl acetate and lead citrate, the specimens were observed under LM and
TEM (H-500; Hitachi). For scanning electron microscopy (SEM), the corneal surface of another five eyes
was freeze-fractured with a scalpel in liquid nitrogen
after in vivo freezing. After freeze-substitution, the
specimens were coated with platinum in a vacuum
evaporator (E-1030; Hitachi) after being dried in a
critical point-drying apparatus (CP-2; Hitachi) and
then observed under a SEM (S-4500; Hitachi).
Specimens Prepared by the VC-FS Method
After Mucolytic Agent Treatment
To investigate the composition of tear film, 10% and
20% solutions of acetylcysteine, a type of mucolytic
agent,18 in 0.1 M phosphate buffer (PB) were applied
according to the method described by Prydal et al.15'16
The solutions were applied for 5 minutes to each of
three eyes, after which the eyes were irrigated for 30
seconds with PB. Three control eyes were treated only
with PB in the same way. After these procedures, the
eyes were in vivo cryofixed, and the specimens for LM,
TEM, and SEM were prepared as described above.
FIGURE l. Photomicrograph of transmission electron microscopic image. Tear film (asteiish) appears to be a layer of a
homogeneous material on the corneal surface. Glycocalices
(arrows) are denser and mainly distributed on the tip of the
microvilli. (inset) Photomicrograph of LM image. The tracer
of aluminum powder, appearing as black masses (arrows),
are confirmed to be on the surface of the tearfilm(ast&isks).
E = corneal epithelial cell..
RESULTS
Specimens Prepared Directly by the VC-FS
Method
Under LM, the tear film appeared as a layer of homogeneous material on the corneal surface. The aluminum powder, which appeared as black masses, was
present on the surface of the tear film (Fig. 1, inset).
Under TEM, xhe tear film appeared to be a layer of a
homogeneous and fine network-like structure of 2 to
6 jum thickness on the corneal surface (Fig. 1). A
dense line measuring about 12 nm in thickness covered the surface of the network-like structure (Figs.
1,2). Under high magnification, the dense line consisted of two parts differing in electron density. The
inner part, in contact with the network-like structure
and almost as thick as half of the phospholipid bilayer
of the epithelial cell membrane, was denser than the
outer one that faced the atmosphere (Fig, 2). The
glycocalices of the surface epithelial cells, which were
denser and mainly distributed on the tip of the microvilli, overlapped the network-like structure of the tear
film; no interface could be identified between them
(Figs. 1,2). Under SEM, contrary to the findings with
the conventional fixation,1019'20 the outlines and types
of surface epithelial cells (light, dark, and medium
cells) could not be seen because of the presence of
the tear film (Fig. 3). When the edge of the freezefractured region was observed, however, the tear film
showed network-like structures as under TEM three-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
383
In Vivo Structure of Rat Precorneal Tear Film
ture of the tear film was almost completely removed
after application of 10% acetylcysteine solution, but
the glycocalices still remained between the microvilli
and no definite damage was found (Fig. 7).
DISCUSSION
Previous studies19"21 as well as our reports910 have
shown that preparation of corneal specimens by conventional chemical fixation removes the tear film on
the corneal surface. In the current study, we investigated the structure and composition of the tear film
using the VC-FS method. We have found that the tear
film appears as a layer of homogeneous and fine network-like structure of 2 to 6 fim in thickness with a
membrane-like layer covering its surface. Aluminum
powder, which was used as a tracer, was confirmed to
be located on the surface of the tear film. The structure of the tear film could be removed completely by
applying 10% or 20% acetylcysteine, but not by PB.
Nichols et al" have demonstrated a similar structure on the corneal surface of guinea pig eyes by cryofixation with a metal contact method. Their conclusion that this structure was only the mucus layer of
the tear film was probably based on the unconfirmed
three layers theory. No other evidence supporting the
conclusion was provided. Moreover, some loss of, or
FIGURE 2. High magnification of transmission electron miinterference with, the tear film probably occurred
croscopic image, showing that the tear film appears as a
when
the eyes were enucleated and had the cooled
homogeneous and fine network-like structure (asterisk) with
metal
applied. These problems, however, were
a dense line covering its surface (arroiuhead). The dense line
avoided
in the current study by using the VC-FS
consists of two parts differing in density. The inner part,
method,
a
better way to preserve the ultrastructure of
which is in contact with the network-like structure and almost as thick as half of the lipid bilayer of the epithelial
cell membrane, is denser than the outer one that faces the
atmosphere. Glycocalices [arrows) overlap the network-like
structure of the tear film. E = corneal epithelial cell.
dimensionally (Fig, 4). The dense line on the surface
of the network-like structure was confirmed to be a
membrane-like layer (Figs. 2,4).
Specimens After Mucolytic Agent Treatment
In contrast to control eyes washed with PB, in which
the structure of the tear film was still observed on the
corneal surface and the outlines or microvilli of the
surface cells could not be seen (Fig. 5), neither tear
film nor glycocalices of the surface cells could be
found on the corneal surface after treatment with 20%
acetylcysteine solution. The superficial epithelial cells
seemed to be damaged by this concentration, so that
cytolysis occurred and the cytoplasm looked transparent under TEM (Fig. 6a). The cell membrane showed
spotty breakage under SEM (Fig. 6b). Outlines and
microvilli of the surface cells could be seen as in the
conventional method (Fig. 6b).10ll9>2° Also, the struc-
FIGURE 3. Photomicrograph of scanning electron microscopic image, showing the freeze-fractured region (F) and
unfractured region (UF) of the corneal surface. Arrows indicate the freeze-fractured edge. Outlines or the microvilli of
the surface epithelial cells cannot be seen when the unfractured region is studied.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
384
Investigative Ophthalmology & Visual Science, February 1997, Vol. 38, No. 2
of these molecules through disulfide bridges produces
polymers of high molecular weight that bind similar
polymers to form a gel by weak interactions of their
carbohydrate chains. The mucus has been demonstrated to be a fine network-like structure. 2e ~ 28 The
homogeneous and fine network-like structure, the
main part of the tear film found in the current study,
was similar to that of mucus. It overlapped with the
glycocalices of the epithelial cells, which also consist
primarily of glycoproteins,29'30 without any morphologic difference between them except for electron
density. Furthermore, this structure could be completely removed by a mucolytic agent but not by PB.
Accordingly, we conclude that it is a layer of mucus.
Mucus in dilute solution is known to be highly hydrated. 24 Tear fluid from the lacrimal gland provides
such a diluent in the precorneal tear film. Therefore,
the mucus of the tear film might be dilute. This proposal also derives support from unpublished data of a
histochemical study for polysaccharides: the networklike structure was weakly stained by the periodic
acid-thiocarbohydrazide-silver protein (PA-TCHSP) method in TEM. A tear film in this state might
be more suitable for its moistening and protective
functions, 12 according to the properties and functions
of the mucus.24'25
FIGURE 4. Part of Figure 3 under high magnification. At the
freeze-fractured edge, the tear film is also composed of the
network-like structure (asterisks), as seem under transmission
electron microscopic, and the dense line that is three-dimensionally confirmed to be a membrane-like layer (M). The
network-like structure shown here on scanning electron microscopic is looser than on transmission electron microscopic, probably because of loss of soluble proteins in the
tear film during freeze-substitution. Arrows indicate the microvilli of the surface epithelial cell, arrowheads indicate the
underlying epithelial cells. E = corneal epithelial cells.
tissue in its natural state. 9 The fact that the aluminum
powder was observed on the surface of the tear film
confirms that all the constituents on the corneal surface were preserved. In other words, the structure of
the tear film shown in the current study was complete.
It seems unlikely that any additional layers, such as an
aqueous layer, could have been present and not found
by our techniques.
Although there may be other sources, such as that
from conjuncdval nongoblet epithelial cells or corneal
epithelial cells, 21314 the mucus of the tear film has
been concluded to be mainly elaborated by conjunctival goblet cells. 1 " 31122 Mucus from different sources
has a similar structure. 1822 " 25 Mucus is composed of
glycoproteins containing many carbohydrate groups.
These glycoproteins are large, elongated molecules
with a protein backbone to which oligosaccharides are
attached in a botde-brush configuration. Crosslinking
With slitlamp microscopy, the interference patterns of a lipid layer can be observed on the corneal
surface1'3'5'12 and this lipid layer has been physically
proven to be present on the anterior surface of the
tear film.31 This lipid layer plays an important role in
reducing the evaporation rate of the tear film, thereby
stabilizing it.1"3'6 In the current study, the dense line
or the membrane-like layer was located on the outermost surface of the tear film. Its composition has been
shown to be different from that of the network-like
structure by means of the PA-TCH-SP method, with
which it failed to stain, whereas the network-like structure did stain (unpublished data). We conclude,
therefore, that it is most likely to be the lipid layer of
the tear film. The tipid layer of the tear film is secreted
by the Meibomian glands1'2'3 and is composed both of
nonpolar lipids, such as wax esters, and polar lipids,
such as phospholipids. 2 ' 32 ' 33 The phospholipids are capable of forming a monolayer molecular membrane
on a water/air interface and of forming the bilayer
structure of the living biological membrane. 3435 The
cell membrane, a phospholipid bilayer membrane, is
visible as two parallel railroad tracks on electron micrographs when their polar hydrophilic heads are reduced by osmium tetroxide. 29 ' 35 In the current study,
the inner part of the dense line covering the surface
of the network-like structure (Fig. 2) was almost as
thick as half of the phospholipid bilayer of corneal
epithelial cell membranes (like a single railroad track)
and, therefore, is probably a monolayer membrane
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
385
In Vivo Structure of Rat Precorneal Tear Film
FIGURES. Specimens after PB washing, (a) Photomicrograph of transmission electron microscopic image. The tear film is still present on the corneal surface and shows almost the same
findings as in Figure 1 or Figure 2. The loose network-like structure (asterisks) results from
poor cryofixation, which might be due to an increase in the aqueous component when
washing. Arrow indicates the dense line, (b) Photomicrograph of scanning electron microscopic image. The corneal surface also shows the same findings as in the unfractured region
in Figure 3; surface epithelial cells cannot be seen. E = corneal epithelial cells.
formed by the polar lipids, whereas the outer one
might be the nonpolar lipids, such as wax esters. Recently Greiner et al33 have suggested that the phospholipids secreted by the Meibomian glands form a planar
noncellular membrane separating hydrophobic lipids
from an aqueous environment. Our findings seem to
provide morphologic evidence supporting this suggestion.
The lipid layer has been measured to be from 0.04
to 0.5 /xm in the human 1 ' 2 5 8 ; this has been determined
6. Specimens after 20% acetylcysteine solution treatment, (a) Photomicrograph of
transmission electron microscopic image. Neither tear film nor glycocalices of the surface
cells can be found on the corneal surface (arrow). The superficial epithelial cells (E) seem
to be damaged by this concentration, so that cytolysis has occurred and the cytoplasm looks
transparent, (b) Photomicrograph of scanning electron microscopic image. The tear film
has been completely removed and the structures of the corneal surface were the same as
that found by the conventional method, in which the oudines or the three types of surface
cells can be seen clearly. Breakage of the cell membrane (arrows) resulting from damage is
mainly found in light (L) or medium (M) cells. D = dark cell.
FIGURE
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
386
Investigative Ophthalmology & Visual Science, February 1997, Vol. 38, No. 2
FIGURE 7. Specimens after 10% acetylcysteine solution treatment, (a) Photomicrograph of
transmission electron microscopic image, (b) Photomicrograph of scanning electron microscopic image, (c) Part of b was examined under high magnification. Tear film is completely
removed, but glycocalices still remain between the microvilli (mrows). No definite damage
is found. E = corneal epithelial cell, L = light cell; M = medium cell; D = dark cell.
mainly by means of measuring the reflection of the
tear film and depends on various theoretical foundations.5'8 It is possible that the different thickness of
the lipid layer seen in the rat tear film is a species
difference.
Collectively, our data indicate that, at least in rat,
the structure of die tear film is mainly composed of
dilute mucus with a lipid layer covering its surface,
but without a free aqueous layer. Holly 6 once suggested that the aqueous layer of the tear film consisted
of two parts of dilute semigel near the epithelium and
a highly dilute mucin solution toward the surface of
the tear film. Measuring the thickness of the tear film
after the application of a mucolytic agent, Prydal et
al15'16 have proposed that the tear film was largely composed of mucus without a free aqueous layer. Observing the lipid layer of the tear film widi noncontact
specular microscopy, Danjo et al12 have shown that
the lipid layer must be in contact widi mucin for interaction. Recendy Tiffany et al17 also suggested that the
theories of tear film structure need revision based on
their study of the soluble mucins in tears. Our findings
support these authors' suggestions.
The thickness of the tear film has been measured
by various methods. 4 ' 7 '" Prydal et al 1516 have criticized
these methods and estimated the tiiickness of tear film
in various animal species; the tear film thickness was
reported to be 34 to 45 fira in human and 12 fim in
rat. These results are larger dian earlier reports of 7
^ m i,2,4.7 Moreover, their results showed that there was
no variation in thickness at different sites on die cornea surface. Our results, in which the tear film varied
from 2 to 6 //m in thickness, do not agree with those
of Prydal et al.15 The variation in tear film thickness
might be partially due to the process of exfoliation of
aged cells, resulting in their elevation so that the tear
film on them becomes thinner. 9 It is still uncertain
why the thickness of the tear film in the current study
differs so much from that of Prydal et al.15 In our
results, the thickness of the tear film seemed to be
affected by the width of the palpebral fissure. The tear
film fissure was reported to measure 0.8 to 1 ^m when
the eyes were cryofixed in the protruded posidon for
easy enucleation after being frozen.9 It might be that
protruding die eyes resulted in extremely wide opening of the palpebral fissure so that die tear film became thinner, whereas in the current study the eyes
were cryofixed when the eyelids were not held open
beyond a physiologic widdi. The dependence of the
thickness of the tear film on the width of the palpebral
fissure suggests that die tear film is so dilute diat it
flows easily. Some difference may be explained by the
measuring methods used by Prydal et all;j'16; for example, difficulty in focusing the laser beam on the illdefined (even with an interferometer and slitlamp microscope) surface of the tear film and the effect of oil
on the tear film when the confocal microscope was
applied.
Key Words
in vivo cryofixation, microscopy, mucus, precorneal tear
film, three-layered structure
References
1. Milder B. The lacrimal apparatus. In: Moses RA, Hart
WMJ, eds. ADLER'S Physiofogyof the Eye, Clinical Applica-
tion. Washington: The C. V. Mosby Company; 1987:
15-35.
2. Holly FJ and Lemp MA. Tear physiology and dry eyes.
Surv Ophthalmol 1977; 22:69-87.
3. Wolff E. The mucocutaneous junction of the lid margin and the distribution of the tear fluid. Trans Ophthalmol Soc UK. 1946;66:291-308.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017
387
In Vivo Structure of Rat Precorneal Tear Film
4. Mishima S. Some physiological aspects of the precorneal tear film. Arch Ophthalmol. 1965; 73:233-241.
5. McDonald JE. Surface phenomena of the tear film.
Am J Ophthalmol 1969; 67:56-64.
6. Holly FJ. Formation and rupture of the tear film. Exp
Eye Res. 1973; 15:515-525.
7. Benedetto DA, Shah DO, Kaufman HE. The instilled
fluid dynamics and surface chemistry of polymers in
the preocular tear film. Invest Ophthalmol. 1975;
14:887-902.
8. Olsen T. Reflectometry of the precorneal film. Ada
Ophthalmol. 1985; 63:432-438.
9. Chen HB, Yamabayashi S, Ou B, Ohno S, Tsukahara
S. Ultrastructural studies on the corneal superficial
epithelium of rats by in vivo cryofixation with freeze
substitution. Ophthalmic Res. 1995;27:286-295.
10. Chen HB, Ou B, Yamabayashi S, Ohno S, Tsukahara
S. Ultrastructural study on rat precorneal tear film by
the quick freezing freeze-substitution method. J Jpn
Ophthalmol Soc. 1996; 100:453-457.
11. Nichols BA, Chiappino ML, Dawson CR. Demonstration of the mucus layer of the tear film by electron
microscopy. Invest Ophthalmol Vis Sci. 1985; 26:464473.
12. Danjo Y, Hamano T. Observation of precorneal tear
film in patients with Sjogren's syndrome. Ada Ophthalmol Scand. 1995; 73:501-505.
13. Gipson IK, Yankauckas M, Spurr-Michaud SJ, Tisdale
AS, Rinehart W. Characteristics of a glycoprotein in
the ocular surface glycocalyx. Invest Ophthalmol Vis Sci.
1992;33:218-227.
14. Watanabe H, Fabricant M, Tisdale AS, Spurr-Michaud
SJ, Lindberg K, Gipson IK. Human corneal and conjunctival epithelia produce a mucin-like glycoprotein
for the apical surface. Invest Ophthalmol Vis Sci.
1995; 36:337-344.
15. PrydalJI, Campbell FW. Study of precorneal tear film
thickness and structure by interferometry and confocal microscopy. Invest Ophthalmol Vis Sci. 1992; 33:
1996-2005.
16. Prydal JI, Artal P, Woon H, Campbell FW. Study of
human precorneal tear film thickness and structure
using laser interferometry. Invest Ophthalmol Vis Sci.
1992;33:2006-2011.
17. Tiffany JM, Pandit JC and Bron AJ. Soluble mucins
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
and the physical properties of tears. ARVQ Abstracts.
Invest Ophthalmol Vis Sci. 1996;37:S845.
Creeth JM. Constituents of mucus and their separation. BrMedBull. 1978;34:17-24.
Hoffmann F. The surface of epithelial cells of the
cornea under the scanning electron microscope. Ophthalmic Res. 1972;3:207-214.
Hazlett LD. Corneal and ocular surface histochemistry
(Review). ProgHistochem Cytochem. 1993;25:1-60.
Hogan MJ, Alvarado JA, Weddell JE. Histology of the
Human Eye. Philadelphia: WB Saunders Company;
1971:55-109.
Moore JC, Tiffany JM. Human ocular mucus. Origins
and preliminary characterization. Exp Eye Res. 1979;
29:291-301.
Allen A, Pain RH, Robson TR. Model for the structure
of the gastric mucous gel. Nature. 1976;264:88-89.
Allen A. Mucus-a protective secretion of complexity.
Trends Biochem Sci. 1983;8:169-173.
Meyer FA. Comparison of structural glycoproteins
from mucus of different sources. Biochim Biophys Ada.
1977;493:272-282.
Specian R, Neutra MR. Mechanism of rapid mucus
secretion in goblet cells stimulated by acetylcholine.
JCellBiol. 1980; 85:626-640.
Sandborn EB. Cells and Tissues by Light and Electron
Microscopy. Volume 2. Tokyo: Academic Press, Inc;
1970:25-89.
Schumacher GFB. Biochemistry of cervical mucus. FertilSteril. 1970;21:697-705.
Darnell J, Lodish H, Baltimore D. Molecular Cell Biology. 2nd ed. New York: WH Freeman; 1990:491-530.
Lehninger AL. Biochemistry. 2nd ed. New York: Worth
Publishers, Inc; 1979:249-277.
Brauninger GE, Shah DO and Kaufman HE. Direct
physical demonstration of oily layer on tear film surface. Am J Ophthalmol. 1972; 73:132-134.
Tiffany JM. The meibomian lipids of the rabbit. I.
Overall composition. Exp Eye Res. 1979;29:195-202.
Greiner JV, Glonek T, Korb DR, Booth R and Leahy
CD. Phospholipids in meibomian gland secretion.
Ophthalmic Res. 1996;28:44-49,
Rawn JD. Biochemistry. Burlington: Neil Patterson Publishers; 1989:209-232.
Lehninger AL. Biochemistry. 2nd ed. New York: Worth
Publishers; 1979:279-306.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933422/ on 06/18/2017