Selective growth of rabbit corneal epithelial
cells in culture and basement membrane
collagen synthesis
Cholappadi V. Sundar-Raj, Ian L. Freeman, and Stuart I. Brown
Selective isolation of rabbit corneal epithelial cells was achieved with a medium containing
D-valine. Long-term cultures of the epithelial cells synthesize basement membrane—type procollagen doublet chains of 160,000 and 180,000 daltons and a higher-molecular-weight species
at 280,000 daltons. Some of these components are disulfide-linked. These peptides gave rise to a
single pepsin-resistant collagen chain of approximate molecular size of 115,000 daltons.
Key words: epithelial cells, cornea, cell culture, biosynthesis,
basement membranes, collagen
ie corneal epithelium is essential to vision
and the health of the cornea in various ways.
Its expanded surface area which is the result
of its projection of microvilli helps to stabilize
the tear film, which presents a smooth
lenslike surface and prevents desiccation.
The corneal epithelium also presents a barrier to both water and bacteria and thereby
maintains corneal transparency and prevents
infection.1 The major biosynthetic product of
the epithelium is a basement membrane
which is primarily or secondarily involved in
many corneal dystrophies and slow-healing
epithelial syndromes. Many of these disorders are characterized by poor epithelial
adhesion. 2 " 4
Investigations of the properties of the epithelial cells can be approached by establishing these cells into culture. At present the
From the Department of Ophthalmology, University of
Pittsburgh School of Medicine, Pittsburgh, Pa.
Supported in part by National Institutes of Health, National Eye Institute grants EY 07032-02 and EY 01817
(C. V. S-R.).
Submitted for publication Sept. 18, 1979.
Reprint requests: Dr. C. V. Sundar-Raj, Department of
Ophthalmology, Eye and Ear Hospital, 230 Lothrop
St., Pittsburgh, Pa. 15213.
1222
only successful long-term culture of surface
epithelium has been the growth of the cells
on feeder layers of irradiated fibroblasts.5
Epithelial biosynthetic studies are impossible
with this technique because the metabolic
products of the fibroblast feeder layer will
confuse the analysis.
We present a method for long-term cultures of rabbit corneal epithelial cells and
partially isolate and characterize their collagenous biosynthetic products.
Materials and methods
Materials. Dulbecco's modified Eagle's medium (DMEM), minimal essential medium containing D-valine, fetal calf serum, penicillin, streptomycin, and pancreatin were from Grand Island
Biological Co., Grand Island, N.Y. [:tH]proline
(VV255 mCi/mmol) was from New England Nuclear, Boston, Mass. A'-Ethylmaleimide, phenylmethane sulfonyl fluoride, ethylene diamine tetracetate (tetrasodium salt), /3-aminopropionitrile
fumarate (/3-APN), and pepsin, were purchased
from Sigma Chemical Co., St. Louis, Mo. Diethylaminoethyl (DEAE)-cellulose (DE52) was from
Whatman Ltd., Maidstone, U. K. Precast Biophore polyacrylamide gels were obtained from
Bio-Rad Laboratories, Richmond, Calif. Collagenase (form III) was purchased from Advance
Biofactors Corp., Lynbrook, N.Y.
0146-0404/80/101222+09300.90/0 © 1980 Assoc. for Res. in Vis. and Ophthal., Inc.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
Volume 19
Number 10
Croivth of rabbit cornea in culture 1223
#
Fig. 1A. Phase-contrast photomicrograph of rabbit corneal epithelial cells growing directly on
tissue culture dish. Epithelial cells are migrating from the corneal tissue and spreading out.
(X250).
Corneal epithelial cell culture. Corneas from
albino rabbits were excised within the limbus,
immediately placed in sterile tissue culture dishes, and covered with DMEM supplemented with
10% fetal calf serum, penicillin (100 U/ml), and
streptomycin (100 jug/ml). An anterior corneal
layer consisting of superficial stroma and epithelium was excised with sharp dissection and transferred to separate dishes containing DMEM with
collagenase (1 mg/ml). Sterile coverslips were
placed over the tissues to keep them in position at
the bottom of the tissue culture dishes. After a 6 hr
incubation, the collagenase-containing medium
was carefully aspirated so as not to disturb the
newly loosened cells and replaced with the selective medium.* Subsequent media changes were
*The selective medium was Eagle's minimal essential
medium (Earle's salts) supplemented with nonessential
amino acids and containing D-valine substituted for
L-valine. In addition, we added 10% fetal calf serum that
had been dialyzed for 48 hr at 4° C against three changes
in 0.85% NaCl.
carried out every 48 hr. Once migration and attachment of epithelial cells onto the dishes were
evident (about 48 hr), the corneal tissue was removed and discarded. The epithelial cells were
then allowed to spread over the dish. Four to six
passages in the selective medium completely removed fibroblast contamination. Confluent monolayer cultures of these cells were serially propagated horn 60 mm tissue culture dishes into roller
bottles with pancreatin and used in the present
study.
Metabolic labeling of corneal epithelial cells. To
study the procollagen synthesis, DMEM minus
serum but supplemented with 100 /xg/ml ascorbic
acid and 50 fJLg/ml /3-APN was used. Cultures
were preincubated for 16 hr with serum-free
medium and then with medium containing 1
/xCi/ml [:)H]proline for 48 hr. The techniques utilized to purify collagen-procollagen from the radioactively labeled cells and medium have been
previously described.B However, in the current
series of experiments collagenous materials were
precipitated with 50% saturated ammonium sul-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
Invest, Ophthalmol. Vis. Sci.
October 1980
1224 Sundar-Raj, Freeman, and Brown
^4»W
f
Fig. IB. Phase-contrast photomicrograph of rabbit corneal epithelial cells growing directly on
tissue culture dish. Two week-old monolayer culture. (X250.)
fate instead of 30% saturation used previously.
With this modification, 90% to 95% of the collagenous material secreted into the culture media was
precipitated. This precipitate was redissolved and
further purified by ethanol precipitation as described previously.6
DEAE cellulose chromatography and sodium
dodecyl sulfate-polyacnjlamide gel electrophoresis (SDS-PAGE) analyses. The ethanol precipitate
was dissolved in limiting butler, dialyzed against
starting buffer, and applied to DEAE-cellulose
column as previously described.6 The column
fractions were pooled separately, dialyzed against
deionized water, and lyophilized. These samples
were further analyzed by SDS-PAGE.7 Limited
pepsin digestion of procollagen or collagen samples was carried out according to the method of
Bellamy and Bornstein.8 Duplicate sets of samples
were digested with pepsin (100 fig/mY) in 0.5N
acetic acid (40 fx\) for 6 hr at 4° C. The reactions
were terminated by addition of IN NaOH (15 //.I);
one set of samples was analyzed by SDS-PAGE
after reduction with 50 mM dithiothreitol (DTT),
and the duplicate set was electrophoresed without
reduction. Procollagen or collagen samples in duplicate were also digested with purified bacterial
collagenase according to the method of Peterkofsky and Diegelmann.9
Results
The use of D-valine as a selective agent for
the normal human and rodent epithelial cells
in culture was first demonstrated by Gilbert
and Migeon in 1975.10 Their study indicated
that the epithelial cells synthesize the enzyme D-amino acid oxidase, which converts
D-valine in the medium to the essential
amino acid L-valine. These cells are therefore
able to proliferate. Fibroblasts, however,
lack this enzyme and are thus unable to grow.
Application of this technique to corneal tissue
produced selective growth of corneal epithelial cells from tissue slices containing both
epithelial cell and stromal fibroblast populations.
Fig. 1A is a phase-contrast micrograph of
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
Volume 19
Number 10
Growth of rabbit cornea in culture 1225
Fig. 1C. Transmission electron micrograph exhibits desmosomal junctions (arrow) between
epithelial cells. (X99.750.)
rabbit corneal epithelial cells migrating out of
the corneal epithelial layer 48 hr after the
corneal tissue had been placed in the selective medium, Some of these epithelial cells
on further incubation rounded up and did not
attach to the dish. The presence of collagenase in the medium helped the epithelial
cells to disperse from the tissue and migrate
onto the culture dish. In other experiments,
we found that scrapings from the corneal epithelial layer did not result in the growth of
epithelial cells even though there were layers
of more than 20 to 30 epithelial cells attached
to each other.
Some cells with fibroblastic morphology
did grow in the initial outgrowth from the
corneal tissue in selective medium along with
epithelial cells. However, these fibroblasts
did not survive subsequent selective media
change and subcultures. It is possible that
the fibroblast survival in the early period
could be the result of these cells obtaining
their required L-valine from the turnover of
endogenous proteins.
As the epithelial cells multiplied to monolayer (Fig. IB), typical polygonal-shaped cells
with contact inhibition similar to other established primary cell cultures" were observed.
These cells, when examined by transmission
electron microscopy (Fig. 1C), exhibited desmosomal connections at their points of contact, thus providing additional evidence of
pure epithelial cell growth in culture. No
evidence of the smaller, more spindle-shaped
fibroblastic cells was seen in the long-term
cultures used for the study of collagen
biosynthesis.
Our pure cultures of rabbit corneal epithelial cells grew very well in DMEM and did
not exhibit any additional requirements for
either growth factors or trace elements.
Ethanol-purified, collagenous precursor
proteins isolated from radioactively labeled
media were examined by SDS-PAGE. In the
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
1226
Invest. Ophthalmol. Vis. Sci.
October 1980
Sundar-Raj, Freeman, and Brown
CONTROL*DTT
CONTROL-DTT
10
20
30
40
50
60
GEL FRACTIONS
70
80
90
100
CONTROL*DTT
GEL FRACTIONS
COLLAGENASE'DTT
40
50
60
GEL FRACTIONS
GEL FRACTIONS
Fig. 2. SDS-PAGE of procollagen secreted by
rabbit corneal epithelial cells in culture. Ethanolprecipitated proteins (18%) were dissolved in
limiting buffer (see text). A, One aliquot of the
sample electrophoresed without reduction. B,
Another aliquot sample was reduced with DTT.
The positions of y-chain (MW 280,000 daltons),
jS-chain (MVV 190,000 daltons) and a-chain (MVV
95,000 daltons) of collagen in the same electrophoresis run are indicated. Gels were fractionated
into 1 mm sections, each fraction was extracted
with 200 (JL\ of 10% SDS overnight, and the radioactivity was determined.
r
B
c
PEPSIN'OTT
PEPSIN-DTT
h
h
I
\
A^
7
GEL FRACTIONS
absence of the reducing agent DTT, the
chain structure of the collagenous material
was shown to consist of three components. A
portion of the material (10% of the total) had a
molecular weight of approximately 280,000
daltons, although the major bands are at
180,000 and 160,000 daltons, together accounting for about 90% of the total. All the
materials were larger than interstitial collagen a-chains (approximately 95,000 daltons).
The collagen a2-chain was not present. This
Fig. 3. SDS-PAGE of 18% ethanol-precipitated
protein. A, Reduced with DTT. B, Digested with
purified bacterial collagenase. C, After limited
pepsin digestion. Gels were fractionated, and the
radioactivity was determined, y-chain, /3-chain,
and a-chain of collagen standard run in the same
electrophoresis are indicated.
indicates the absence of type I collagen which
would be synthesized by contaminating
stromal fibroblasts12' l3 in impure epithelial
cell cultures (Fig. 2, A).
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
Volume 19
Number 10
Growth of rabbit cornea in culture
10
20
30
to
50 60 70 80 90 100 110
FRACTION NUMBER
1227
120 130 KO 150
Fig. 4. DEAE-cellulose chromatograni of [;1H]proline-labeled corneal epithelial procollagencollagen. The ethanol-precipitated proteins were dissolved in limiting buffer and dialyzed
against starting buffer as described in Materials and methods. The dialysate was applied to the
column (1.5 by 20 cm) of DEAE-cellulose equilibrated in a starting buffer. The column was
developed with a linear gradient of limiting buffer. Radioactive peak 1 (fractions 89 to 98) and
peak 2 (fractions 99 to 105) were separately pooled, dialyzed extensively against deionized
water, and lyophilized.
In the presence of the reducing agent
DTT, the major components were 180,000
and 160,000 daltons (Figs. 2, B and 3, A).
Again, no material appeared at the fibrous
collagen a-chain region. Bacterial collagenase almost completely destroyed these
chains (Fig. 3, B), indicating the collagenous
nature of the material. Pepsin treatment,
which does not attack the helical portion of
the collagen molecule, resulted in destruction of the 180,000 and 160,000 dalton components and generation of single-chain material at approximately 115,000 daltons.
About 80% of the applied material appeared
in this region. The remainder appeared as
low-molecular-weight material near the lower
end of the gel (Fig. 3, C). The presence of a
collagenase-sensitive molecule which gives
rise to a single chain, larger than a\ (type I)
with limited pepsin digestion is characteristic
of basement membrane collagens.14"17 Thus
the corneal epithelial cells in culture exclusively synthesized a basement membranetype collagenous material. This property of
basement membrane collagen biosynthesis
was maintained after 10 passages in culture.
Further purification of the precursor pro-
teins from ethanol precipitate was achieved
on DEAE-cellulose columns (Fig. 4). Three
major peaks were observed on chromatography. The material eluted from the column in
the starting buffer wash (fractions 20 to 40)
when analyzed by SDS-PAGE consisted
mainly of a single band at about 115,000 daltons (not shown). The other two peaks
(marked 1 and 2) eluted in salt gradients of 5
to 7 mmho. Each of the peaks, 1 (fractions 89
to 98) and 2 (fractions 99 to 105), was pooled
separately, dialyzed against deionized water,
lyophilized, and analyzed individually on
SDS-PAGE in the presence and absence of
the reducing agent DTT.
Column-purified peaks 1 and 2 were similar when analyzed by SDS-PAGE analyses.
Both the peak materials were completely
destroyed by collagenase. Peak 1 material
(Fig. 5, A) in the presence of DTT resolved
into a single polypeptide of 140,000 daltons
and in the absence of DTT resolved in three
minor components of molecular weights
280,000, 180,000, and 160,000 and a major peak around 140,000. Pepsin digestion
brought them into a single polypeptide of approximately 115,000 daltons (Fig. 5, B).
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
1228
Invest. Ophthalmol. Vis. Sci.
October 1980
Sundar-Raj, Freeman, and Brown
A
CONTROL-DTT
CONTROL+DTT
32
40
GEL
48
56
64
72
80
88
A
CONTROL-DTT
CONTROL*DTT
32
96
40
48
56
64
GEL FRACTIONS
FRACTIONS
B
PEPSIN -DTT
PEPSIN + DTT
PEPSIN-DTT
PEPSIN+DTT
r
b
"S 12
5
Q_
O 10
5
Q.
O 10
0
<1
-
UJ
3
o
cc
8
a- 6
i
4
I'A
'
2
8
16
24
32
40
GEL
48
56
64
72
80
88
24
96
32
Fig. 5. Electrophoretic analyses of DEAE-cellulose column-purified collagenous.protein fractions
(peak 1). Lyophilized samples were dissolved in
limiting buffer, and aliquots were electrophoresed.
A, Without DTT (—) and after reduction with
DTT (
). B, Samples were subjected to limited
pepsin digestion and then electrophoresed without
DTT (-—) and after reduction with DTT (
).
Gels were fractionated, and the radioactivity was
determined. The positions ofy-, /3-, and a-chains of
collagen in the same electrophoresis run are indicated.
DEAE-cellulose column-eluted peak 2 material (Fig. 6, A), when analyzed, revealed
two individual peptides 180,000 and 160,000
daltons in the presence of reducing agents.
Also, under unreduced conditions, two other
polypeptides, one greater than 300,000 and
another around 280,000 daltons, were resolved. On pepsin digestion, similar to the
material from peak 1, peak 2 proteins resolved into a single protein band corresponding to 115,000 daltons (Fig. 6, B). The presence of lower-molecular-weight material in
40
GEL
FRACTIONS
48
56
(
72
80
88
96
FRACTIONS
Fig. 6. Electrophoretic analyses of DEAE-cellulose column-purified collagenous protein fractions
(peak 2). For explanation see legend to Fig. 5.
peak 1 indicates that this peak contained
some partly processed chains.
Discussion
We have utilized a selective growth medium containing D-valine which for the first
time has established corneal epithelial cells
in long-term cultures. We speculate that an
undisturbed basal cell layer is important for
long-term culture, since our earlier trials of
culturing epithelium scraped from the cornea
have uniformly failed. In order to obtain an
undisturbed basal layer, we included superficial corneal stroma with the epithelium for
our initial culture. Fibroblast proliferation
was prevented by D-valine because these
cells cannot convert D- to L-valine. This
seems to allow the slow-growing epithelial
cells (which can utilize D-valine for their
growth) to predominate.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
Volume 19
Number 10
The establishment of pure corneal epithelial cells in culture provides a system for
studying the biosynthetic mechanisms involved in the formation of normal basement
membrane.
Several tissues like isolated rat renal glomeruli,18 chick lens cells,19 rat parietal yolk
sac endoderm,20 and developing mouse embryos21 have been investigated for their ability to synthesize basement membrane components. Conflicting reports have arisen from
these reports regarding the molecular size of
the basement membrane collagen molecules.
This confusion mainly appears to be due to
differences in the methods of extraction and
analysis in different laboratories. Pepsin digestion has been the conventional method
applied to solubilize and characterize collagens.14"16 With this method, nonhelical
crosslinked species are removed by the action of the enzyme, giving rise to pepsinresistant helical basement membrane chains.
The pepsin-resistant basement membrane
collagen molecules produced by this method
have been reported to range in molecular
size between 110,000 to 195,000 daltons,21
although it is now generally accepted that the
major component after pepsin treatment is
approximately 110,000 to 140,000.17
Our present studies indicate that corneal
epithelial cells in long-term culture produce
a high-molecular-weight species of basement
membrane precursor material composed of
major polypeptide chains of 160,000 and
180,000 daltons and higher-molecular-weight
species of 280,000 daltons, some of which are
disulfide-linked. When subjected to limited
digestion conditions with pepsin, these peptides give rise to a single pepsin-resistant
polypeptide with a molecular weight of approximately 115,000 daltons. This chain is
collagenase-sensitive and is similar in molecular size to other reported basement
membrane collagens.
In 1977, Sundar-Raj et al.22 reported that
long-term cultures of rabbit corneal endothelial cells in culture synthesized two precursor chains of basement membrane collagen which migrated in close proximity
on SDS-PAGE with approximate molecular
Growth of rabbit cornea in culture 1229
weight of 160,000 and 200,000 daltons. These
two chains yielded a single pepsin-resistant
collagen peptide of 115,000 daltons on limited
pepsin digestion. The presence of a doublet
on SDS-PAGE gels might very well be a
characteristic feature of cells in culture which
synthesize basement membrane-type collagenous precursors. The study of Crouch
and Bornstein23 on the synthesis of type IVlike procollagen by human amniotic fluid
(AF) cells in culture also shows this doublet
on SDS-PAGE gels. The doublet migrated as
two closely spaced chains between /3-cOmponents and type I procollagen chains (approximately 150,000 daltons). However, unlike the corneal epithelial cells reported here
and endothelial precursor collagen reported
previously,6 AF procollagen eluted prior to
the start of the gradient when examined
by DEAE-cellulose chromatography. Pepsin
treatment of type I precursor molecules synthesized by AF cells (AF1) produced a collagen chain which migrated slightly more
slowly than collagen al(I) chains.
The basement membrane precursor collagen protein from rabbit corneal endothelial
cells in culture eluted as a single peak between 2 and 3 mmho on DEAE-cellulose
chromatography. In the present study under
similar conditions, the collagenous proteins differed in molecular size (160,000 and
180,000 daltons) and eluted between 5 and 7
mmho.
It is possible that basement membrane collagen precursors synthesized by rabbit corneal epithelial and endothelial cells are related but are separate entities. Perhaps they
represent a family of similar molecules as has
been reported for collagen type I24 and type
II.25
It has been suggested that collagen type A
chain, abundant in skin and liver, might originate in epithelial basement membranes,26 but
in the studies reported here this molecular
species was not seen. In addition, the absence
of type I al-chains strongly indicates that the
corneal epithelial cells in culture synthesize
only type IV procollagen. This epithelial cell
product has distinguishing properties on
DEAE-cellulose columns which suggest that
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017
1230
Invest. Ophthalmol. Vis. Sci.
October 1980
Sunclar-Raj, Freeman, and Brown
it is different from AF cell type IV,23 EHS
sarcoma17 and rabbit corneal endothelial cell
collagen,6 although all species mentioned
share the common property of appearing as a
doublet on SDS-PAGE under reducing conditions and, when pepsin-digested, give rise
to a single pepsin-resistant collagen chain with
molecular weight between 110,000 and
140,000 daltons.
We are grateful to Mr. Dan Waruszewski for his expert technical assistance. Our appreciation and thanks
are extended to Dr. Takashi Inoue and Mrs. Gloria
Limetti for transmission electron microscopy.
REFERENCES
1. Newell FW and Ernest JT: Ophthalmology: Principles and Concepts. St. Louis, 1974, The C. V.
Mosby Co., p. 67.
2. Nakanishi I and Brown SI: infrastructure of the epithelial dystrophy of Meesmann. Arch Ophthalmol
93:259, 1975.
3. Rice NSC, Ashton N, Jay B, and Blach RK: ReisBuchlers' dystrophy: a clinico-pathological study. Br
J Ophthalmol 62:577, 1968.
4. Cogan DG, Kuwarbara T, Donaldson DD, and Collins E: Microcystic dystrophy of the cornea: a partial
explanation of its pathogenesis. Arch Ophthalmol
92:470, 1974.
5. Sun TT and Green H: Cultured epithelial cells of
cornea, conjunctiva and skin: absence of marked intrinsic divergence of their differentiated states. Nature 269:489, 1977.
6. Sundar-Raj CV, Freeman IL, Church RL, and
Brown SI: Biochemical characterization of procollagen-collagen synthesized by rabbit corneal endo-
12. Stoesser TR, Church RL, and Brown SI: Partial
characterization of human collagen and procollagen
secreted by human corneal stromal fibroblasts in cell
culture. INVEST OPHTHALMOL VIS SCI 17:264, 1978.
13. Yue BYJT, Baum JL, and Smith BD: Collagen synthesis by cultures of stromal cells from normal
human and keratoconus corneas. Biochem Biophys
Res Commun 86:465, 1979.
14. Kefalides NA: Basement membranes: structural and
biosynthetic considerations. J Invest Dermatol 65:
85, 1975.
15. Daniels JR and Chu CH: Basement membrane collagen of renal glomerulus. J Biol Chem 250:3531,
1975.
16. Tryggvason K and Kivirikko K: Heterogeneity of
pepsin-solubilized human glomerular basement
membrane collagen. Nephron 21:230, 1978.
17. Timpl R, Bruckner P, and Fietzek P: Characterization of pepsin fragments of basement membrane collagen obtained from a mouse tumor. Eur J Biochem
95:255, 1979.
18. Cohen MP and Klein CV: Evidence for heterogeneous origin of glomerular basement membrane.
Biochem Biophys Res Commun 77:1326, 1977.
19. Grant ME, Kefalides NA, and Prockop DJ: The
biosynthesis of basement membrane collagen in
embryonic chick lens. J Biol Chem 247:3539, 1972.
20. Clark CC, Tomichek EA, Koszalka TR, and Minor
RR: The embryonic parietal yolk sac: the role of the
parietal endoderm in the biosynthesis of basement
membrane collagen and glycoprotein in vitro. J Biol
Chem 250:5259, 1975.
21. Adam son E D and Ayers SE: The location and synthesis of some collagen types in developing mouse
embryos. Cell 16:953, 1979.
22. Sundar-Raj CV, Freeman IL, and Church RL: Collagen biosynthesis by rabbit corneal endothelial cells in
culture.
thelial cells in culture. INVEST OPHTHALMOL VIS SCI
18:75, 1979.
7. Fairbanks G, Steck TL, and Wallach DFH: Electrophoretic analysis of the major polypeptides of the
human erythrocyte membrane. Biochemistry 10:
2606, 1971.
8. Bellamy G and Bornstein P: Evidence for procollagen, a biosynthetic precursor of collagen. Proc Natl
Acad Sci USA 68:1138, 1971.
9. Peterkofsky B and Diegelmann R: Use of a mixture
of protease-free collagenases for the specific assay of
radioactive collagen in the presence of other proteins. Biochemistry 10:988, 1971.
10. Gilbert SF and Migeon BR: D-valine as a selective
agent for normal human and rodent epithelial cells
in culture. Cell 5:11, 1975.
11. Owens RB: Selective cultivation of mammalian epithelial cells. Methods Cell Biol 14:341, 1976.
23.
24.
25.
26.
INVEST OPHTHALMOL VIS SCI 16(ARVO
Suppl):102, 1977.
Crouch E and Bornstein P: Characterization of a
type IV procollagen synthesized by human amniotic
fluid cells in culture. J Biol Chem 254:4197, 1979.
Linsenmayer TF, Hendrix MJC, and Little CD:
Production and characterization of a monoclonal antibody to chick type I collagen. Proc Natl Acad Sci
USA 76:3703, 1979.
Butler WT, Finch JE Jr, and Miller EJ: The covalent
structure of cartilage collagen. Evidence for sequence heterogeneity of bovine alpha I(II) chains. J
Biol Chem 252:639, 1977.
Stenn KS, Madri JA, and Roll FJ: Migrating
epidermis produces AB2 collagen and requires continual collagen synthesis for movement. Nature
277:229, 1979.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933320/ on 06/15/2017