Response of basal epithelial cell surface and Cytoskeleton to solubilized extracellular matrix molecules

Abstract

Corneal epithelium removed from underlying extracellular matrix (ECM) extends numerous cytoplasmic processes (blebs) from the formerly smooth basal surface. If blebbing epithelia are grown on collagen gels or lens capsules in vitro, the basal surface flattens and takes on the smooth contour typical of epithelium in contact with basal lamina in situ. This study examines the effect of soluble extracellular matrix components on the basal surface. Corneal epithelia from 9- to 11-d-old chick embryos were isolated with trypsin-collagenase or ethylenediamine tetraacetic acid, then placed on Millipore filters (Millipore Corp., Bedford, Mass.), and cultured at the medium-air interface. Media were prepared with no serum, with 10% of calf serum, or with serum from which plasma fibronectin was removed. Epithelia grown on filters in this medium continue to bleb for the duration of the experiments (12-14 h). If soluble collagen, laminin, or fibronectin is added to the medium, however, blebs are withdrawn and by 2-6 h the basal surface is flat. Epithelia grown on filters in the presence of albumin, IgG, or glycosaminoglycans continue to bleb. Epithelia cultured on solid substrata, such as glass, also continue to bleb if ECM is absent from the medium. The basal cell cortex in situ contains a compact cortical mat of filaments that decorate with S-1 myosin subfragments; some, if not all, of these filaments point away from the plasmalemma. The actin filaments disperse into the cytoplasmic processes during blebbing and now many appear to point toward the plasmalemma. In isolated epithelia that flatten in response to soluble collagens, laminin, and fibronectin, the actin filaments reform the basal cortical mat typical or epithelial in situ. Thus, extracellular macromolecules influence and organize not only the basal cell surface but also the actin-rich basal cell cortex of epithelial cells.

Full text

Response
of
Basal
Epithelial
Cell
Surface
and
Cytoskeleton
to
Solubilized
Extracellular
Matrix
Molecules
STEPHEN
P
.
SUGRUE
and
ELIZABETH
D
.
HAY
Department
of
Anatomy,
Harvard Medical
School,
Boston,
Massachusetts
02115
ABSTRACT
Corneal
epithelium
removed
from
underlying
extracellular
matrix
(ECM)
extends
numerous
cytoplasmic
processes
(blebs)
from
the
formerly
smooth
basal
surface
.
If
blebbing
epithelia
are
grown on
collagen
gels
or
lens
capsules
in
vitro,
the
basal
surface
flattens
and
takes
on
the
smooth
contour
typical
of
epithelium
in
contact
with
basal
lamina
in
situ
.
This
study
examines
the
effect
of
soluble
extracellular
matrix
components on
the
basal surface
.
Corneal
epithelia
from
9- to
11-d-old chick
embryos
were
isolated
with
trypsin-collagenase
or
ethylenediamine
tetraacetic
acid,
then
placed
on
Millipore
filters
(Millipore
Corp
.,
Bedford,
Mass
.),
and
cultured
at
the
medium-air
interface
.
Media
were
prepared with
no
serum,
with
10%
calf
serum,
or
with
serum
from
which
plasma
fibronectin
was
removed
.
Epithelia
grown
on
filters
in
this
medium
continue
to
bleb
for
the
duration
of
the
experiments
(12-24
h)
.
If
soluble collagen,
laminin,
or
fibronectin
is
added
to
the
medium,
however,
blebs
are
withdrawn
and
by
2-6
h
the
basal
surface
is
flat
.
Epithelia
grown on
filters
in
the
presence
of
albumin,
IgG,
or
glycosaminoglycans
continue
to
bleb
.
Epithelia
cultured
on
solid
substrata,
such
as
glass,
also
continue
to
bleb
if
ECM
is
absent
from
the
medium
.
The
basal
cell
cortex
in
situ
contains
a
compact
cortical
mat
of filaments
that
decorate with
S-1
myosin
subfragments
;
some,
if
not
all,
of these filaments
point
away
from
the
plasmalemma
.
The
actin
filaments disperse
into
the
cytoplasmic
processes
during
blebbing
and
now
many
appear
to
point
toward
the
plasma-
lemma
.
In
isolated
epithelia
that
flatten
in
response
to
soluble
collagens, laminin,
and
fibronectin,
the
actin
filaments
reform
the
basal
cortical
mat
typical
of
epithelia
in
situ
.
Thus,
extracellular
macromolecules
influence
and
organize
not
only the
basal
cell
surface
but
also
the
actin-rich
basal
cell
cortex
of epithelial
cells
.
The
basal
cell
surface
and
cortical
cytoskeleton of the
embry-
onic corneal
epithelium
are
intimately
related
to
the
underlying
extracellular
matrix
(ECM)
.
The
cortical
cytoskeleton
forms
a
dense
mat
next
to
the basal
epithelial
plasmalemma,
and
the
latter
is
relatively
smooth
or
flat
in
configuration
.
The
basal
plasmalemma
seems
to be
connected
to
the
lamina
densa
of
the
underlying
basal
lamina (basement
membrane)
by
extra-
cellular
filaments
that
traverse
the
lamina
rara
externa
(1)
.
The
laminae
rarae externa
and
interna
of the
corneal,
and
most,
if
not
all,
other
epithelial
basal
laminae
(2-5)
contain
proteogly-
can
granules
rich in
glycosaminoglycans
(GAG),
such
as
he-
paran
and/or
chondroitin
sulfates
(HS,
CS) and
hyaluronic
acid
(HA)
.
Laminin
and
fibronectin
occur
in
the
laminae
rarae
of
the
glomerular
basement
membrane
(4),
and
both
are
found
in
the
embryonic
avian
corneal
basal
lamina
(Sugrue
and
Hay,
unpublished
observations)
.
Type
IV
collagen
is
also
present
in
the
corneal
basement
membrane
(6)
.
Type
IV
and
other
col-
THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
91
OCTOBER
1981
45-54
©The
Rockefeller University
Press
"
0021-9525/81/10/0045/10
$1
.00
lagens
(e
.g
.,
type
V)
may
be
concentrated
in
the
lamina
densa
(4)
but
might
also
extend
in
the
form
of
fine
filaments
(1)
to
the
cell
surface
in
concentrations
too
dilute
to
be
detected
by
immunohistochemistry
.
The
existing
evidence
that
the organization of the basal
epithelial
cell
cytoplasm
is
dependent
on
the presence
of
the
underlying
ECM
is
mainly
circumstantial
.
In corneal
epithelia
or
epidermis
with
well-developed
hemidesmosomes,
tonofibrils
insert
in
cytoplasmic
plaques
that
seem
to
be
attached
to
specialized
regions
of
ECM
(1)
.
In
younger
corneal
epithelia
lacking
hemidesmosomes,
the
mat
of
cortical
filaments
in
the
basal
cytoplasm
seem
to
follow
the
contour
of
the
basal
lamina
(1)
.
More
direct
evidence
that
the
configuration
of the basal
cell
surface
and
cortical
cytoplasm
depends
on
ECM
derives
from
studies
using
enzymes
or
EDTA
to
remove
the
epithelial
basal
lamina
;
such
epithelia
extend
cytoplasm-filled blebs
from
their
naked
basal
surfaces
(see
references
5
and
7)
.
When
45
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

grown
on
solid
collagenous
substrata, isolated
embryonic
avian
epithelia
withdraw
these
blebs,
become
smooth-surfaced
again,
and
step
up
synthesis
of
both
collagen
and
GAG
(7-9)
.
In
the
present
paper,
we
ask
whether
the
reaction
of
the
basal
corneal
cytoplasm
to
ECM
is
due merely
to
the
physical
support
of
the
collagenous
substratum,
or
whether
solubilized
ECM
molecules themselves
can induce
isolated
epithelia
to
withdraw
the
blebs
.
Blebbing
epithelia
impaled
on
Millipore
filters
(Millipore
Corp
.,
Bedford,
Mass
.)
are
exposed
to
solu-
tions
of
various
collagens,
glycosaminoglycans,
fibronectin,
and
laminin
to
judge
the
effect
of
these
molecules
on
the
structure
of the
basal
epithelial
surface
.
The
appearance
of
the
basal
surface
on
solid
but
inert
substrata
is
also
examined
.
Finally,
we
address
the
question of
the
organization
and
nature
of the
cortical
cytoskeleton
in
blebbing
and
ECM-treated
em-
bryonic corneal
epithelia
by
examining
the
actin
components
labeled with
S-1
fragments
of
heavy
meromyosin
.
MATERIALS
AND
METHODS
Corneas
harvested
from
chick
embryos
after
9-11
d
of
incubation
were
treated
at
25°C
with
0.1%
trypsin
(Sigma
Chemical
Co
.,
St
.
Louis,
Mo
.)
and
0
.1%
collagen-
ase
(Sigma Chemical
Co
.)
in
Hanks'
solution,
pH
7
.4,
for
9-12
min
or
were
soaked
in
0
.04%
EDTA
(Sigma Chemical
Co
.)
in
calcium-magnesium-free
Hanks'
solution,
pH
7
.4,
for
30
min
.
The
epithelia
were
removed
with
forceps
from
the
stroma
as
a
sheet,
transferred
to
a
Millipore
filter
(HATF
0
.45-pm
pore
size)
.
Such
epithelia
are
completely
clean
of
contaminating
ECM
(8,
9)
.
Epithelia
on
filters
were then
cultured
at
the
air-medium
interface
of
Falcon
culture
dishes
(Falcon
Plastics,
Oxnard,
Calif) in
Ham's
F-12
medium
(Gibco
Laboratories,
Grand
Island Biological
Co
.,
Grand
Island,
N
.
Y
.)
without
serum
or with
10%
fetal
calf
serum(Flow
Laboratories,
McLean,
Va
.)
from whichplasma
fibronectin
(CIG)
was
removed
by
affinity
chromatography
with
gelatin
.
Medium
was
supplemented
with
50
itg/ml
ascorbic acid
and
1%
antibiotic-antimycotic
(peni-
cillin,
fungizone,
and
streptomycin,
Gibco
Laboratories)
.
Each
isolated
corneal
epithelium
was
usually
placed
basal-side
down
on
a
disk
(3-mm
Diam)
made
of
Millipore
filter
and
transferred
within
2 h
after
isolation
to
standard
medium
(above)
or
to
medium
containing
soluble
ECM
molecules,
albumin,
or
IgG
.
Some
epithelia
were
cultured
at
the
air-medium
interface
on
glass
cover
slips
coated
with
0
.1%
poly-L-lysine
(Sigma Chemical
Co
.)
or
tissue
culture
plastic
(Falcon
Plastics)
.
Other
epithelia
were
attached
by
their
apical
surfaces
to
Millipore
filter
disks
coated
with
poly-L-lysine
.
The
disks
were
then
inverted across
holes
2
.3-mm
Diam
in
another
filter
;
as
a
result,
a
large
stretch
of
lim
.
46

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
91,
1981
the
basal
epithelial
surface
was
exposed
to
the
underlying
medium
without
physical
support
on
the
basal
side
.
Still
other
epithelia
were
grown
on
Millipore
filters
that
had
been soaked
for
18
h
in
collagen
(type
1,
100
!rg/ml,
see
below)
and
washed
briefly
in
three
changes
of
Hanks'
solution
.
Purified
collagens
were
dissolved
in
0
.5
M
acetic
acid
at
a
concentration
of
1
mg/ml,
then
dialyzed
against
0
.1
M
phosphate-buffered
saline
(PBS),
pH
7
.4
.
Type
I
collagen
was
denatured
by
heating
at
50°C
for
20
min
.
Type
I
collagen
was
purified
from
rat
tail
tendons
by our
laboratory
(8, 9),
and
type
II
collagen
from
sternal
cartilage
by
Dr
.
Thomas
Linsenmayer (Department
of
Anatomy,
Harvard Medical
School,
Boston,
Mass
.)
(10)
.
Type
IV
collagen
and
laminin
from
a
murine
tumor
(1l,
12)
were
provided
by Dr
.
George
Martin
(National
Institute
of
Dental
Research,
Bethesda,
Md
.)
.
Rat
tendon
a2(1)
chains
were
also
obtained
from
Dr
.
Martin
.
Cellular
fibronectin
was
extracted
from
NIL8
cells,
and
plasma
fibronectin
was
purified
from
human
plasma by
Dr
.
Richard
Hynes
(Massachusetts
Institute
of
Technology,
Cambridge,
Mass
.)
(13)
.
HA
and
CS
mixed
chains
(Sigma
Chemical
Co
.),
HS
chains
(Dr
.
J
.
A
.
Cifonelli,
University
of
Chicago,
Chicago,
III
.),
and
heparin
(Upjohn
Co
.,
Kalamazoo,
Mich
.)
were
dissolved
double-strength
in
H
2
O
and
diluted
in
double-strength
media
.
Bovine
serum
albumin
(Sigma Chemical
Co
.)
and
nonspecific
rabbit
1gG
(Miles
Labo-
ratories
Inc
.,
Elkhart,
Ind
.)
were
dissolved
directly
into
medium
in
concentrations
as
stated
in
Results
and
in
Table
1
.
After
incubation,
cultures
were
routinely
fixed
for
30
min
in
2%
paraformal-
dehyde,
2.5%
glutaraldehyde
in
cacodylate
buffer
(0
.1
M),
and
postfixed
in
I%
osmium
tetroxide
in
0
.1
M
cacodylate
buffer,
pH
7
.4,
at
4°C
for
30-60
min
.
They
were
stained
en
block
in
1%
uranyl
acetate,
dehydrated,
and
embedded
in
Spurr
(D
.
E
.
R
.
736
embedding
kit,
Tousimis
Research
Corp
.,
Rockville,
Md
.)
.
Myosin
S-1
was
prepared
by
the
method
of
Margossian
and
Lowey
(see
reference
14)
and
stored
at
a
concentration
of
30
mg/ml
in
50%
glycerol
.
Epithelia
were
extracted
with
Triton
X100
in
0
.05-0
.1%
in
0
.1
M
PIPES
buffer
(Sigma
Chemical
Co
.),
2
mM
MgC12,
2
mM
EGTA,
l
mM
phenylmethyl
sulfonyl
fluoride
(PMSF),
1
MM
P-tosyl-L-argone
methyl
ester-HCI
.
S-1
decoration
was
performed
in
0
.3
M
PIPES
buffer
containing
S-1
fragments,
0
.5-1
mg/ml,
for
10
min
at
room
temperature
.
Fixation
was
carried
out
subsequently
in
1%
glutaral-
dehyde
with
0.2%
tannic acid (Mallinckrodt, St
.
Louis,
Mo
.,
batch
AR
1764)
in
0
.1
M
PIPES
buffer
(14)
for
30
min
.
After
thorough
washing,
tissue
was
postfixed
in
1%
OsO,
in
0
.1
M
phosphate
buffer,
pH
6
.0,
for
30
min
at
4°C,
and
prepared
as
described
above
for
electron
microscopy
.
RESULTS
The
avian corneal
epithelium
at
9-11
d
of
embryonic
devel-
opment
is
two
to
three
cells
thick
.
The
cells
of
the
basal
layer
are
cuboidal
to
columnar
in
shape
and
possess extensive
rough
endoplasmic
reticulum
and
prominent Golgi complexes
(1)
.
The
cells
of
the
middle
layer
are
round
in
shape
and
the
FIGURE
1

The
tissues
shown
in
Figs
.
1-7
were
routinely
fixed
and
processed
for
electron
microscopy
.
Fig
.
1 is
a
micrograph
of
the
basal
cytoplasm
of
a
corneal
epithelial
cell
fixed
in situ
.
The
basal
cortex
contains
a
mat
(parentheses)
of
microfilaments
(mf)
running
close
to
the
plasmalemma
and
parallel
to
the
basal
lamina
(bl)
.
The
cytoplasm
is
rich in
granular
endoplasmic
reticulum
at this
stage
(10
d
of
incubation)
.
Mitochondrion
(m),
nucleus
(n),
collagen
fibrils
in
underlying
corneal
stroma
(cf)
.
Bar,
0
.5
/Am
.
FIGURE
2

The
basal surface
of
the
corneal
epithelium
begins
to bleb as
soon
as
the
basal
lamina
is
removed
.
The
blebs
vary
in
size
and
contain
filamentous
ground
substance, ribosomes,
and
even
mitochondria
and
endoplasmic
reticulum
(as
in
the
bleb
labeled
by
the
asterisk)
.
Neither
trypsin-collagenase
nor
EDTA
in
the
concentrations
used here
affects
the
free
surface
(
fs),
which
possesses
microvilli
and a
surface
coat
.
Lateral
cell
surfaces
and
junctions
also
appear
unaffected
by
the
treatment
.
Nucleus
(n)
.
Bar, 5
ym
.
FIGURE
3

The
blebs
disrupt
the
basal
cytoskeleton
.
Microfilaments
(mf) attached
to lateral
cell
junctions
(as at
the
arrow)
may
persist,
but the
cytoplasm
extending
into
the
blebs
appears
unorganized
.
The
enzyme-isolated
epithelium
shown
here
was
incubated
for
6 h
on
a
Millipore
filter
(Mp)
before
fixation
in
aldehydes
and
osmium
tetroxide
.
Nucleus
(n)
.
Bar, 0
.5
ttm
.
FIGURE
4

The
blebs
persist
even
when
isolated
epithelia are
grown
on
inert
solid
substrata
such
as
plastic
(shown
here)
or
glass
.
The
blebs
are
flattened
slightly
where
attached
to
the
plastic
.
The
plastic
itself
was
removed
after
embedding
the
tissue
.
Bar,
0
.5
JIM
.
FIGURE
5

Isolated
epithelia
grown
on
Millipore
filters
in
media
containing
solubilized
collagens
lose
their
blebs
as
early
as
2
h
after
collagen
is
added
to
the
medium
.
The
specimen
shown
here
was
treated
for
4
h
with
100
lag/ml
of
type
IV
collagen
.
The
epithelium
is
flattened
and
stretches across
a
pore
in
the
filter
.
Millipore
filter
(MP),
reorganized
basal
microfilaments
(mf
) .
Bar,
0
.5
FIGURE
6

The
isolated
epithelium
shown
here
was
treated
with
a2(I)
chains
(100
yg/ml)
for
6 h
.
The
basal
epithelial
surface
has
flattened
and
reorganized
the
cortical
microfilamentous
mat
(mf)
.
The
unsupported
basal
surface
spans
distances
of
1-2
tim
without
filter
contact
.
Millipore
filter
(Mp)
.
Bar,
0
.5
ttm
.
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

SUCRUE
AND HAY

ECM
Organizes
Basal
Epithelial
Surface

47
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

outermost
cells
(periderm)
are
flat
.
The
cells
are
connected
laterally
by
desmosomes,
which
are
increasing
in
number
at
this
time,
and
they
are
joined
at
the apical
surface
by
typical
junctional
complexes
(1)
.
The
free
surface
of
the
periderm
is
covered
with
microvilli
and
a glycoprotein
surface
coat
.
The
basal
surface
of
the
epithelium
is
smooth
and
follows
the
contour
of
the
basal
lamina
(Fig
.
1)
.
Just
above
the
basal
plasma
membrane,
the
cortical
cytoskeleton
forms
a
dense
mat,
150
nm
wide,
consisting
of
numerous
microfilaments
(mf,
Fig
.
1)
running
parallel
to
the
plasma
membrane
.
Beneath
the basal
lamina
is
a
cell-free
zone
of collagen
fibrils
(cf,
Fig
.
1),
rich
in
proteoglycan
and
fibronectin
.
Below
this
are
the
fibroblasts
and
orthogonally
arranged
collagen
fibrils
of
the
stoma
proper,
and,
below
them,
the
corneal
endothelium
.
Epithelia
isolated
with
trypsin-collagenase
(Fig
.
2)
or
EDTA
exhibit
dramatic blebbing
of
their
basal
surfaces
upon
removal
of
the
basal
lamina
.
The
blebs
persist
indefinitely
when
epithe-
lia
are
grown
on
Millipore
filters
(Fig
.
3)
.
The
blebs take
the
form of
cytoplasm-filled
protrusions
from
the basal
epithelial
surface
.
Their
plasmalemma
is
completely
free
of
visible
extra-
cellular
material
(Figs
.
2
and
3)
.
They
contain
ribosomes
(Fig
.
3),
filaments,
and
even
mitochondria
and
endoplasmic
reticu-
lum
(Fig
.
2)
.
These
protrusions
on
living
cells
are
not
to
be
confused
with
empty
membrane-bounded
blisters
(see
refer-
ence
15),
which
are
most
likely
an
artifact
of
aldehyde
fixation
(16)
.
The
cytoskeletal
elements
associated
with
the
lateral
and
apical
surfaces
of
isolated
epithelia
appear
unaffected
by
the
treatment,
but
the
microfilamentous
cortical
mat
is
partly
or
completely
dispersed
.
The
general configuration of
the apical
and
lateral
surfaces
of
blebbing
epithelia
(Fig
.
2)
appears
essentially
normal
.
The
cells
remain connected
by
desmosomes,
and
cytoplasmic
organelles
also
seem
little
affected
.
On
Millipore
filters
(Fig
.
3),
the
blebs
move
into
the
pores
(for
a
distance
of
up
to
20
,um)
.
Such
epithelia
continue
to
bleb
indefinitely
in
ECM-free
media
.
Isolated
epithelia
also
con-
tinue
to
bleb
on
solid
glass
or
plastic
substrata
(Fig
.
4),
even
though
attached
to
the
substratum
.
Coating
the
dish
with
polylysine
does
not
cause
the
blebs
to
disappear
.
On
solid
substrata
composed
of
collagen,
however,
epithelia
quickly
withdraw
the
blebs
(8)
.
To
test
the
hypothesis
that
collagen
molecules themselves
interact
with the
basal
epithelial
surface,
we
added
solubilized
collagen
to
the
medium
of
isolated
epithelia
grown
on
Millipore
filters
.
Within 2-4
h,
such
epithelia
withdraw
their
blebs
from
the pores of
the
filter,
reorganize
their
basal
cytoskeleton,
and
become
flattened
.
Types
1,
II,
and
IV
(Fig
.
5)
all
have
similar
activities
(Table
1)
.
Although
some
effect
is
observed
at
a
concentration of 50
Fig/ml,
collagen
proved
to
be
most
effective
at
concentrations
of
100
lLtg/ml
.
Because
heat-denatured
type
I
collagen
was
as
effective
as
intact
molecules,
we
added
purified
a2(I)
chains
to
the
culture
medium
.
a2(I)
chains
were
also
quite
effective,
at
a concentration
of 100
jig/ml,
in
inducing
blebbing
epithelia
to
reorganize
the
cortical
cytoskeleton
(Fig
.
6
and
Table
1)
.
The
reorganized basal
microfilamentous
mat
(mf,
Figs
.
5
and
6)
is
similar
to
that
observed
in
situ
(Fig
.
1)
.
Because
the
flattened
surface
of
the
basal
epithelial
cells
bridges
the
pores
of
the
filter,
and
because
no
basal
lamina
reforms
(Figs
.
5
and
6),
it
seems
likely
that
it is
the
cortical
cytoskeleton
that
supports
the
cells
.
The
idea
that
physical
support
from
the
substratum
is
not
required
for
maintenance
of
the
cortical
cytoskeleton
derives
from
the
following
considerations
.
Although
the
effective
pore
4
8

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
91,
1981
size
of
the
Millipore
filter
is
0
.45
pin,
the
actual
space
between
cellulose
acetate
strands
facing
the
basal
epithelial
surface
ranges
from
1-21Lm
.
The
flattened
epithelial
surface
spanning
these
spaces thus
must
be
supported
by
the
cytoskeleton
in
response
to
collagen
molecules
on
the
cell
surface
.
To
examine
whether
or
not
collagen
affected
the epithelia
by
binding
to
the
filter
rather
than
the
cells,
epithelia
were
placed
on
filters
that
had
been
previously
immersed
in
collagen
solution
and
then
rinsed
in
Hanks'
solution
.
Such
epithelia
continued
to
bleb
.
To
demonstrate
further
that
the
reorganization
of
the
basal
epithelial
surface
is
independent
of
attachment
to
a
physical
substratum,
epithelia
were
placed
on
polylysine-
coated Millipore
filters
with
apical
surface
facing
the
filter
and
then
were
grown
basal-side
down
across
holes 2
.3
mm
Diam
TABLE
I
Response of
Basal
Surface
to
ECM
*The
number
of
epithelia
that
flattened
after
the
treatment
indicated
is
given
as
the
numerator,
and
total
number
of
experiments
as
the
denominator
.
An
epithelium
is
only
scored
as
positive after
examination
of
the
entire
basal
surface
.
Substance
1
Hours
2
after
4
treatment
6
12-48
ftg/ml
Collagen
I
10-20
0/2*
0/2 0/2
0/5
1/2
50
-
1/2
-
2/4
-
100 0/3 3/4 2/2 6/7 4/4
200
0/3
- - -
2/2
Collagen
11
10
0/2
-
0/2 1/5
50
0/2 1/3 2/3 2/3
100
1/4
2/3 4/5 4/4
200
0/4
-
3/3
-
Collagen
IV
10-20
0/2 0/2 0/4 1/2
50
012 1/2 3/4
-
100 2/4 3/3 212
-
200 2/2
-
2/2
-
a2(I)
chains
50
0/2
-
114
100 1/3
-
3/4
Cellular
fibronectin
10
-
0/3
-
0/6
-
25
-
0/3
-
3/6 2/2
50
0/2 0/7 1/4
7/9
8/8
100
-
0/3
-
212
-
Plasma
fibronectin
-
0/4
-
4/5
-
(CIG)
50
Laminin
0
.5
-
0/2 012 112
1/2
1
-
214
2/3
3/3
-
5
-
1/3
3/3 2/2 2/2
10
0/2
1/4
3/3 3/3
-
20
-
0/4
-
212
-
F-12,
no
serum
-
0/6
-
0/8
0/6
F-12,
10%
serum
-
0/3
-
0/4
0/4
F-12
10%
serum
without
0/4 0/7
0/3
1/18
0/11
CIG
Hyaluro
nic
acid
250
-
0/8
-
Chondroitin
sulfate
250
-
0/4 0/4
Heparan
sulfate
250 0/4 0/5
-
Heparin
250
-
0/3
-
Serum
albumin
250
-
0/3
-
IgG
250
-
0/4
-
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

FIGURE
7

The
epithelium
shown
on
this
light
micrograph
was
cultured
for
6
h
without
any
filter
support
for
the
stretch
of
basal
surface
depicted
.
The
medium
contained
1001g/ml
type
I
collagen
.
The
basal
surface
has
flattened
without
any
physical
substrate
.
Free
surface
(Fs)
square,
area
shown
in Fig
.
8
.
Bar,
10
um
.
FIGURE
8

This electron
micrograph
shows
the
area
depicted
by
the square
in Fig
.
7
.
The
basal
surface
is
flat
and no
polymerized
material
can be
resolved
on
the
surface
.
Microfilamentous
mat
(mf)
.
Bar,
0
.5
tm
.
FIGURE
9

This
light
micrograph
shows
an
isolated
epithelia
suspended
as
in
Fig
.
7
for
6
h
over
media
without
basal
filter
support
.
In this
case,
no
extracellular
molecules
were
added
.
The
basal
surface
is
decorated
with
numerous
blebs
(arrows)
.
Bar,
10,
um
.
FIGURE
10

Isolated
epithelium
grown
on
Millipore
filters
in
the
presence
of
laminin
in
the
medium
reorganizes
its
microfilamen-
tous
basal
cortical
mat
(mf)
and
appears
flat
in
4-6 h
.
The
specimen
shown
here
was
treated
for
6
h
with
1
pg/ml
of
laminin
.
The
blebs
have
retracted
.
Nucleus
(n),
endoplasmic
reticulum
(er)
.
Bar,
0
.5
Am
.
FIGURE
11
Isolated
epithelium
cultured
in
the
presence
of 50
lAg/ml
fibronectin flattened
after 6
h
of
incubation
and
the
microfilamentous
mat (mf)
reorganized
.
Nucleus
(n)
;
Millipore
filter
(Mp)
.
Bar,
0
.5
Im
.
SUGRUE
AND HAY

ECM
Organizes
Basal
Epithelial
Surface

49
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

(see
Materials
and
Methods)
.
The
basal
side
thus
had
no
physical
support
over
most
of
its
surface
and,
again,
no
detect-
able
ECM
polymerized
on
the basal
surface
(Fig
.
8)
.
Yet,
these
epithelia
flattened
in
response
to
soluble
collagen
molecules
in
medium
(Figs
.
7
and
9)
.
We
next
asked
whether
or
not
other
molecular
components
of
the
basal lamina,
if
added
in
soluble
form
to
isolated
epithelia
on
Millipore
filters,
would
also
cause the
cortical
cytoskeleton
to
organize
.
When
the
blebbing
epithelia
are
confronted
with
laminin
or
fibronectin,
they
respond
by
with-
drawing
the
blebs
and
flattening
the
basal
surface
(Figs
.
10
and
11)
.
It is
possible
to
observe
the
effect
even
if
epithelia are
suspended
across
2
.3-mm
holes
as
described
above
(Fig
.
7)
.
Laminin
at
concentrations
of
1-10
ILg/ml
causes
epithelia
to
withdraw
blebs
from
the
filters
by
2-4
h
.
At
higher concentra-
tions
(>20
[Lg/ml),
laminin
is
not
so
effective
(Table
I)
and
at
100 itg/ml
laminin
precipitates
from
solution
.
Fibronectin,
in
concentrations
as
low
as
50
ftg/ml,
is
effective
after
6
h
of
incubation,
but
fibronectin
does not
appear
to
affect
the
epithelium
at
earlier
intervals,
regardless
of the
concentration (Table
I)
.
Interestingly,
although
we
used
CIG-
free
serum
or
omitted
serum
for
the
experiments,
we
found
that
whole
fetal
calf
serum
has
no
effect
on
blebbing
(Table
I)
.
The
CIG
content of
commerical
serum
presumably
is
too
low,
because
plasma
fibronectin
is
as
effective
as cellular
fibronectin
in
causing the
epithelial
surface
to
flatten
(Table
I)
.
Earlier
studies
demonstrated
that
the
isolated
corneal
epi-
thelium
responds
to
GAG
by
increasing
its
synthesis
of
GAG
(17)
.
To
examine
the
effect
of
such
macromolecules
on
the
blebbing
morphology
of
epithelia,
we
confronted
isolated
epi-
thelia
on
Millipore
filters
with
HA,
CS, HS,
and
heparin
.
None
of the
molecules
at
concentrations of
250
Fig/ml
had
any
effect
on
the
blebbing
(Table
I)
.
To
examine
whether
or not the
flattening
of
the
basal
surface
is
a
nonspecific
effect
of
protein
binding
randomly
to
the
epithelium,
we
cultured
isolated
epithelia
on
Millipore
filters
in
the
presence of
bovine
serum
albumin
or
IgG
(nonspecific
rabbit)
.
The
epithelia
exhibited
no
response
to
the
addition
of
these
molecules
at
concentrations
of
2501íg/
ml
(Table
I)
.
To
investigate
further
the
dramatic
effects
of
ECM
molecules
on
the basal
cytoarchitecture
of
epithelial
cells,
we
decorated
the
actin
filaments
with
S-1
subfragments
of
heavy mero-
myosin
.
The
in
situ
basal
epithelial
cell
(Fig
.
12)
on
basal
lamina
has
a
microfilamentous
cortical
mat
composed
of
a
meshwork
of
actin
filaments
.
Actin
filaments
can
be
seen
branching
out
from
this
meshwork
both
downward,
toward
the
basal
plasma
membrane, and
upward,
intermingling with or
attaching
to
the
secretory
organelles
and
ribosomes
(Fig
.
12)
.
There
does
not
seem
to
be
a
unidirectional
polarity
to
the
actin
filaments
in
the
cortical
mat
;
they resemble
stress
fibers
(14)
in
this
regard
.
Actin
filaments
approaching
the
basal
plasma
membrane
mainly
point
away
from
the
membrane,
as
observed
in a
number
of other
cells
(14)
.
In
striking
contrast
to
their
organization
in
situ,
the
basal
microfilaments
of
isolated
epithelia are
dispersed
into
a
mesh-
work
in the blebs
(Figs
.
13-15)
.
The
decorated
actin
filaments
frequently
approach
the
plasma
membrane
and
occasionally
there
appears
an
electron-dense
plaque
at
the
putative
insertion
site
(dp,
Fig
.
14)
.
Interestingly,
the
polarity
of the
decorated
filaments
"inserting"
into
the
plasmalemma
is
often
with
ar-
rowheads
pointing
toward
the
membrane
(Fig
.
13
inset
and
Fig
.
14)
.
Filaments
may
also
run
parallel to
the
plasmalemma
in
the
blebs (Fig
.
15)
.
If
such
a
blebbing
epithelium
is
confronted
with
collagens,
laminin, or
fibronectin,
incubated
for
6
h,
then
extracted
with
detergent
and
decorated
with
S-1,
a
morphology
results
that
is
remarkably
similar
to
that
seen
in
the
in situ
epithelium
.
The
actin-rich
microfilamentous
mat
reorganizes
(Fig
.
16)
.
Fila-
ments
descending
from
the
actin
complex
to
approach
the
basal
plasma
membrane
again
more
often
point
away
from
the
membrane
.
It
is
difficult
to
discern
the exact
relation
of
actin
filaments
to
the
plasmalemma
because
of
extraction
and
swell-
ing
caused
by
the
detergent
treatment
.
DISCUSSION
The
present
study
demonstrates
that
embryonic
corneal
epithe-
lial
cells
are
capable
of
interacting
with,
and
responding
to,
three
classes
of
extracellular
matrix
glycoproteins
in
soluble
form,
namely,
collagens,
laminin,
and
fibronectin
(Fig
.
17)
.
In
FIGURE
12

The
tissues
depicted
in
the
micrographs
in
Figs
.
8-12
were
detergent
extracted
and
treated
with
S-1
fragments
of
heavy
meromyosin
before
fixation
in
aldehydes,
tannic
acid,
and
osmium
tetroxide
.
Fig
.
12
shows
a
section
of
basal
epithelial
cytoplasm
in situ
.
The
microfilamentous
basal
cortical
mat
(mf)
appears
swollen
due
to
the
detergent
treatment
.
It
is
well labeled
with
S-1
fragments,
and
the
arrowheads
may
point
in
opposite
directions
(small
arrows)
.
Above
the
mat,
ribosomes
seem
to
be
suspended
on
actin
filaments
.
Coated
vesicle
(cv),
microtubule
(mt),
basal
lamina
(bl)
.
Bar,
0
.2
Am
.
FIGURE
13
In
a
typical
bleb,
the
direction
of the
S-1
decoration
on
actin
filaments
may
appear
random
(small
arrows,
main
figure)
.
The
arrowheads,
however,
often
point
toward
the
plasmalemma
(small
arrow,
inset)
.
The
detergent-treated
preparations
contain
vesicles
(v)
inside
or
outside
of
the
cells
that
derive
from
the
solubilized
membranes
.
All
the
blebs
shown
in Figs
.
9-12
were
on
the
undersurface
of
enzyme-isolated
epithelia
grown
for
6
h
on
Millipore
filters
.
Main
figure
:
bar,
0
.2
;
Inset
:
bar,
0
.1
pm
.
FIGURE
14

In
the
bleb
depicted
here,
several
actin
filaments
are
decorated
with
arrowheads
pointing
toward
the
plasmalemma
(small
arrows)
.
Some
appear
to
insert
in
a
dense
plaque
(dp)
in
the
membrane
.
The
cytoplasm
above
the
bleb
shown
here
is
rich
in
intermediate-sized filaments
(if),
which
are
associated
with
nearby
desmosomes
(not
shown)
.
Bar,
0
.2
Am
.
FIGURE
15

In
the
bleb depicted
here,
some
actin
filaments
(large
arrowheads)
are
running
parallel
to
the
plasmalemma
.
Others
are
running
in
a
perpendicular
direction
(small
arrows)
.
This
is
a
small
bleb,
which
may
be
partly
retracting
into
the
cell
.
Bar,
0
.2
Am
.
FIGURE
16

After
addition
of
solubilized
collagens,
laminin, or
fibronectin,
the
organized
actin
complex
(mf)
that
comprises
the
basal
cytoskeleton
re-forms
.
The
epithelium
depicted
here
was
treated
for
6 h with
type
IV
collagen
in
solution (100
wg/ml)
.
The
section
is
slightly
tangential
to
the
poorly
preserved
plasmalemma
(pm)
.
This
plane
of section
and
the
detergent
treatment
tend
to
exaggerate
the
distance
between
the
cortical
mat
(mf)
and
the
plasmalemma
(pm)
.
Actin
filaments
(small
arrows)
may
be
oriented
in
opposite
directions
.
S-1
debris
is
present
in
the
space
(lower
left)
between
the
epithelium
and
filter
.
Bar,
0
.2
Am
.
5
0

THE
JOURNAL
of
CELL
BIOLOGY
"
VOLUME
91,
1981
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

SUGRUE
AND HAY

ECM
Organizes
Basal
Epithelial
Surface

51
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

FIGURE
17
Diagram
summarizing
the
effects
of
several
different
molecules
on
the
organization
of
the
basal
corneal
cell
surface
.
The
basal
surface
blebs
when
the
basal
lamina
is
removed
by
EDTA
or
enzyme
treatment,
and
this
blebbing
persists
(left)
on
Millipore
filters
in
the presence
of
nonmatrix
proteins
or
GAG
.
Soluble
col-
lagens, fibronectin,
and
laminin
added
to
the
medium
(right)
cause
the
blebs
to
retract
and
the
cytoskeleton
to
reform
the
basal
mat
of
actin
filaments
.
response
to
removal
of
the
glycoprotein-rich basal
lamina
from
the
basal
surface,
the
epithelial
cells
send
out
numerous
round
or
irregularly
shaped
cytoplasm-filled
protrusions
(blebs)
.
Staining
of the basal
cytoskeleton
of
the
blebbing
epithelial
cells
with
S-1
fragments
of
heavy
meromyosin
shows
that
the
actin-rich
cortical
microfilamentous
mat,
which
characterizes
the
basal
cells
in
vivo,
is
disrupted
and
that
actin
filaments
flow
out
into
the
cell
protrusions
during
blebbing
.
The
epithelia
continue
to
bleb
when
grown
on
filters
in
the
presence
of
IgG,
albumin,
and
GAG
(Fig
.
17)
.
We
also
demonstrate
that
inert
solid
substrata
do
not cause
retraction
of the blebs
.
We
found
in
this
study
that
solubilized
collagens
are
just
as
effective
as
solid
collagenous
substrata
in
causing
the
epithelial
surface
to
flatten
and
the
basal
cytoplasm
to
reorganize
.
Soluble
laminin
and
fibronectin
also
cause the basal
epithelial
surface
to
resume
its
in
situ
appearance
(Fig
.
17)
.
The
results
suggest
that
these
ECM
molecules themselves
interact
with the basal
cell
surface,
because
it is
clear
that
they
need
not
be
polymerized
into
solid
substrata
to
exert their
effect
on
epithelial
cytoarchitecture
.
In
the
discussion,
we
will
consider
the
effects
of
each
of the
glycoproteins
separately,
and
then
speculate
on
the
possible
interactions
that
might
occur
among
these
molecules,
the
cell
surface,
and
the
cytoskeleton
.
Any
type of
solid
collagenous substratum,
including
lens
capsule
(type
IV
collagen)
and
collagen
gels
(type
I
or type
II
collagen),
induces
flattening
of
the
blebbing
basal
surface
of
enzyme-
or
EDTA-isolated
corneal
epithelium
(8, 9,
18)
.
In the
present
study,
a
similar
nonspecificity
was
observed
.
Types
I,
II,
and
IV
collagens
in
solution
are
capable
of
inducing
bleb-
bing
epithelia
to
retract
basal
cytoplasmic
protrusions
within
2
h
(at
concentrations
of
100
ttg/ml)
.
Heat-denatured
type
I
collagen
and
purified
a2(I)
collagen
chains
seem
to
be
as
effective
as
undenatured
collagen in
eliciting
the
recovery
of
basal cytoarchitecture
.
These
results
indicate
that
the
epithelial-
ECM
interaction
is
not
dependent
on
the
native
helical
config-
uration
and
is
mediated
by components
of
collagen
structure
that
the
various
molecules
share
.
Rubin
et
al
.
(19)
have
recently
reported
that
rat
hepatocytes
attach
equally
well
to
all
collagens
tested
(types
I-V),
as well
5
2

THE
JOURNAL
OF
CELL
BIOLOGY
"
VOLUME
91,
1981
as
to
denatured
collagen,
a
1(I)
chains,
cyanogen
bromide
peptides,
and
collagenlike
synthetic
peptides
(albeit
less
effi-
ciently)
.
Epidermal
cells
show
a
preference
for
adhering
to
type
IV
collagen,
although
there
is
significant
attachment
to
type
I
collagen
(20)
.
A
preference
for
type
IV
collagen
was
also
demonstrated
in
mammary
gland
epithelium
attachment
in
vitro
(21)
.
Thus,
hepatocytes
resemble
corneal
epithelial
cells
in
the
nonspecificity
of
their
requirement
for
collagen,
whereas
certain
epithelia
seem
to
be
more demanding
.
It
is
important
to point
out,
however,
that
our study
is
not
an
attachment
assay
.
Indeed,
we
showed
that
attachment
to
a
physical
substratum
is
not necessary
for
the
epithelial
basal
surface
to interact
with
and
respond
to
ECM
molecules
.
Bleb-
bing
epithelia
suspended
over
holes
2
.3-mm
Diam
flatten
in
response
to
soluble
ECM
molecules
.
Additional evidence
that
the corneal
epithelial
surface
can
interact
with
ECM
molecules
in
soluble
form
is
as
follows
.
Filters
presoaked
in
collagen
and
washed
briefly
do
not
have an
effect
on
the
basal
epithelial
surface
.
Indeed,
if
the
epithelium
had
attached
to
the
filter
via
filter-bound
collagen, the
epithelium
would
have
followed
the
contour
of the
filter
pores,
rather
than spanning
the pores
.
Moreover,
neither
basal
lamina,
collagen
fibrils,
nor
other
visible
ECM
polymerizes
on
the basal
epithelial
surface
during
the
2-
to
6-h
period
in
which
the
basal
surface
flattens
.
Laminin
may
be involved
in
the
attachment
to
collagen
of
epidermis,
mammary
and
yolk
sac
epithelia,
and
several
epi-
thelial
cell
lines
that
prefer
type
IV
collagen
(20-22),
whereas
endothelial
cells
(23),
lens
epithelium
(24),
and
some
hepato-
cytes
(25)
seem
to
use
fibronectin
to
attach
to
collagen
.
There-
fore,
we
also
studied
the
effects
of
these
two
molecules
on
the
basal
epithelial
surface
.
We
found
that
both
laminin
and
fibronectin
in
soluble
form have
a
dramatic
effect
on
the
corneal
epithelial
basal
surface
.
Laminin
in
very
low
concen-
trations
eliminates
blebbing
in
2-4
h,
but
both
cellular
and
plasma
fibronectin
require
4-6
h
to
begin
to
flatten
the basal
cell
surface
.
The
effect
of laminin,
then,
is
as
fast
as
the
effect
of
collagen,
but the
effect
of
fibronectin
is
somewhat
delayed
.
It
will
be
of
considerable
interest
to
determine
the
interre-
lationships,
if
any,
among
these
ECM
molecules
during
their
interaction
with
the
basal
epithelial
surface
.
The
corneal
epi-
thelium
synthesizes
types
I
and
II
collagens (26)
and
probably
also
type
IV
(18)
.
In
all
likelihood,
this
epithelium
also
syn-
thesizes
laminin
and
fibronectin
;
both
glycoproteins
are
present
in the corneal
basement
membrane
.
Therefore,
the
possibility
exists
that
the
effect
of
exogenous
collagen
and
laminin
on
epithelial
blebbing,
even
though
very
rapid,
uses
an
endoge-
nous
source
of
one
or
the other
ECM
molecules
.
Terranova
et
al
.
(27)
demonstrated
that
certain
epithelial
cells
in
the
presence
of a
protein
synthesis
inhibitor,
cycloheximide,
fail
to
adhere
to
collagen
unless
exogenous
laminin
is
added
.
In
preliminary
studies,
however,
we
find
that
cycloheximide
does
not
inhibit
the
effect
of
either
collagen
or
laminin
on
corneal
epithelial
blebbing
but
does
abolish
the
effect
of
fibronectin,
as
do
inhibitors
of collagen
secretion
(28)
.
The
simplest
explanation
for
our
data
is
that
collagen
and
lamimn
are
capable
of
independent,
direct
interaction
with
the
epithelial
cell
surface
to affect
the
basal
cytoskeleton,
whereas
fibronectin
interacts
indirectly
(e
.g
.,
via
collagen)
.
It
is
tempting
to
believe
that
these
effects
on
cell
surface
and
cytoskeletal
morphology
are
mediated
via
receptors
for
one
or
more
of
these
ECM
molecules
.
Kleinman
et
al
.
(29)
suggested
a
ganglioside
type
receptor
for fibronectin,
which
then
indi-
rectly
binds
to
collagen
;
however,
fibronectin
binding
to
cells
is
increased
by adding
soluble
collagen
(30)
.
Goldberg
(31)
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

reported
that
a
receptor
for
collagen
itself
exists
on
the
cell
surface
of
fibroblasts
.
When
he
added
labeled
type
I
collagen
to
the
top
of
epithelial
sheets,
it
failed
to
bind, suggesting
that
there
is
not such
a
receptor
site
on
the
epithelial
cell
surface
(31)
.
Because
epithelial
cells
on
culture
dishes
are
polarized
with
the
apical
surface
up,
however,
basal
surfaces
would
not
have been exposed
to
collagen
in
Goldberg's
studies
.
The
binding
site
on
the
collagen
molecule
for
the
fibroblast
cell
surface
is
in
the
helical
portion
of
al
and
a2
chains
and
seems
to
be
determined
by
primary
structure
rather
than
helical
conformation
(31)
.
This
type of
binding
would
be
consistent
with
the
results
of
the
present
study,
because
both
collagen
chains
and
heat-denatured
collagens
interact
with
corneal
ep-
ithelial
cells
.
More
recently,
Rubin
et al
.
(19)
have
hypothesized
that
the
attachment
of
rat
hepatocytes
to
collagen,
a
process
that
does
not
require
fibronectin
(32,
33),
is
mediated
by
receptors
that
recognize
multiple
repeating
sites
(e
.g
.,
Gly-Pro-Hyp)
along
the
collagen
molecule
.
They
suggest
that
these
receptors
have
a low
affinity
for
collagen
and
are
mobile
in
the
plane
of
the
cell
membrane
.
The
tertiary
structure
of
the
collagen
molecules
may
also
be
important,
because
native
collagen
substrates
promote
hepatocyte
attachment
better
than
denatured
collagen
or
synthetic collagenlike
substrates
(19)
.
Regardless
of
the
possible
mode
of
interaction
of
collagen,
laminin,
and
fibronectin
with
the
cell
surface,
either
via
inde-
pendent
receptors
or
by
means
of
molecular
complexes,
the
end
result
observed
in
the
present
study
is
a
dramatic
effect
on
the
basal
epithelial
cytoskeleton
.
In
the
presence
of
solubilized
ECM,
the
basal
cytoskeleton
is
composed
of
organized
actin
filaments
in
parallel
array,
some
of
which
may
attach
directly
or
indirectly
(34) to
the
cell
membrane
.
The
arrangement
is
unlike
the
relationship
of
polymerized
fibronectin
to
actin
cables
in
cultured
fibroblasts
(35),
in
that
whole
actin
bundles
do
not
appear
to
insert
into
the
plasmalemma
.
The
blebs,
which
form
instantly
on
the
basal
epithelial
surface
when
basal
lamina
is
removed,
contain
a
disorganized
meshwork
of
actin
filaments
and
are
undoubtedly
mobile
because
they
move
into
the
pores
of
filters
.
Interestingly,
individual actin
filaments
in
the
blebs
may
decorate
with
S-1
fragments
in
a
pattern
pointing
toward
the
cell
membrane,
whereas
in
microvilli
the
pattern usually
points
away
from
the
plasmalemma
(14)
.
For
an
actin-myosin
sliding
mechanism
to
work,
the
actin
pattern
should
point
to
the
membrane
if
processes
are
moving
away
from
the
cell
(15,
36),
as
seems
to
be
the
case
for the
blebs
.
Further
work
is
necessary,
however,
to
establish
that
an
actin-myosin
interaction
occurs
in
the
blebs
.
The
functional
implications
of
the
structural
organization
of
the
basal
epithelial
surface
have
received
little
attention
in
the
past,
even though
blebs
have been
reported
to
form on a
number
of
different
types
of
epithelia
when
isolated
by
EDTA
or
trypsin
(see
references
5,
37,
and
38)
.
Our
study
suggests
that
one
effect
of
ECM
molecules
may
be
to
immobilize
basal
plasmalemmal
receptors
and
cytoskeleton
.
Along
these
lines,
it
is
interesting
to
note
that
certain
malignant
epithelial
cells,
after
digesting
the
basal
lamina,
become
mobile
and
invade
the
underlying
spaces
(39)
.
Another
effect,
directly
or
indirectly
related
to
the
organization of
the
cytoskeleton,
is
the
effect
of
ECM
molecules,
such
as
collagen,
on
epithelial
polarity
(40),
cell
shape,
and
metabolism
(9)
.
The
present
study
calls
atten-
tion to
the
role
of
ECM
in
controlling
the
organization of
the
basal
epithelial
cytoplasm
and
invites
further
study
of
its
structurally
important
and
potentially
motile
cytoskeletal
ele-
ments
.
We
thank
Dr
.
George
Martin
for
samples
of
purified
type
IV
collagen,
a2(I)
chains,
and
laminin
;
Dr
.
Richard
Hynes
for
fibronectin
;
and
Dr,
Thomas
Linsenmayer
for
type
II
collagen
.
We
also
thank
Drs
.
David
Begg
and
David
Albertini
for
supplying
S-1
fragments
and
for
help
in
using
them,
Ms
.
Deborah
Almeida
and
Ms
.
Nancy
Pollack
for
their
technical
assistance,
and
Ms
.
Susan
Hunt
for
her
secretarial
talents
.
This
research
was
supported
by
a
National
Institutes
of
Health
grant,
HD-00143,
to
Dr
.
Hay
.
The
laboratories
supplying
purified
ECM
components were
supported
by
CA-17007
(Dr
.
Hynes),
EY-
02261
(Dr
.
Linsenmayer),
and
the
National
Institute
of
Dental
Re-
search
(Dr
.
Martin)
.
Received
for
publication
23
April
1981,
and
in
revised
form
16
June
1981
.
REFERENCES
I
.
Hay,
E
.
D
.,
and
J
.
P
.
Revel
.
1969
.
Fin
e
Structure
of
the
Developing
Avian
Cornea
.
S
.
Karger,
Basel
.
144
pp
.
2
.
Trelstad,
R
.
L
.,
K
.
Hayashi,
and B
.
P
.
Toole
.
1974
.
Epithelia
l
collagens
and
glycosami-
noglycans
in
the
embryonic
cornea
.
Macromolecular
order
and
morphogenesis
in
the
basement
membrane
.
J
.
Cell Biol
.
62
:815-830
.
3
.
Hay, E
.
D
.,
and
S
.
Meier
.
1974
.
Glycosaminoglycans
synthesis
by
embryonic
inductors
:
neural
tube,
notochord
and
lens
.
J
Cell Biol
.
62
:889-898
.
4
.
Farquhar,
M
.
G
.
1981
.
The
glomerular
basement
membrane
:
a
selective
macromoleculer
filter
.
In Cell
Biology
of
Extracellular
Matrix
.
E
.
D
.
Hay,
editor
.
Plenum
Press,
New
York
.
In
press
.
5
.
Banerjee,
S
.
D
.,
R
.
H
.
Cohn,
and
M
.
R
.
Bemfield
.
1977
.
Basa
l
lamina
of embryonic
salivary
epithelia
:
production
by
the
epithelium
and
role
in
maintaining
lobular
mor-
phology
.
J
.
Cell
Biol
.
73
:445-463
.
6
.
Hay,
E
.
D
.,
T
.
L
.
Linsenmayer,
R
.
L
.
Trelstad,
and
K
.
von
der
Mark
.
1979
.
Origin
and
distribution of
collagens
in
the
developing avian
comes
.
Curr
.
Top
.
Eye
Res
.
1
:1-35
.
7
.
Hay,
E
.
D
.
1973
.
Origin
and
role
of
collagen
in
the
embryo
.
Am
.
Zool
.
13
:1085-1107
.
8
.
Dodson,
J
.
W
.,
and
E
.
D
.
Hay
.
1974
.
Secretio
nof
collagen
by
corneal
epithelium
.
If
.
Effect
of
the
underlying substratum
on
secretion
and
polymerization
of
epithelial
products
.
J
.
Exp
.
Zool
.
189
:51-72
.
9
.
Meier,
S
.,
and
E
.
D
.
Hay
.
1974
.
Contro
l
of
corneal
differentiation
by
extracellular
materials
.
Collagen
as
a
promoter
and
stabilizer
of
epithelial
stroma
production
.
Dev
.
Biol
.
38
:249-270
.
10
.
Linsenmayer,
T
.
F
.,
and
M
.
J
.
C
.
Hendrix
.
1980
.
Monoclona
l
antibodies
to
connective
tissue
macromolecules
:
type
II
collagen
.
Biochem
.
Biophys
.
Res
.
Commun
.
92
:440-446
.
11
.
Timpl,
R
.,
G
.
R
.
Martin,
P
.
Bruckner,
G
.
Wick,
and
H
.
Wiedemann,
1978
.
Nature
of
the
collagenous
protein
in
a
tumor
basement
membrane
.
Eur
.
J
.
Biochem
.
84
:43-52
.
12
.
Timpl,
R
.,
H
.
Rohde, P
.
G
.
Robey, S
.
I
.
Rennard,
J
.
M
.
Foidart,
and
G
.
R
.
Martin
.
1979
.
Laminin-
A
glycoprotem
from basement
membranes
.
J
.
Biol
.
Chem, 254
:9933-9937
.
13
.
Hynes,
R
.
O
.,
A
.
T
.
Destree,
M
.
E
.
Perkins,
and
D
.
D
.
Wagner
.
1979
.
Cel
l
surface
fibronectin
and
oncogenic
transformation
.
J
.
Supramol
.
Siruct
.
11
:95-104
.
l4
.
Begg,
D
.
A
.,
R
.
Rodewald,
and
L
.
1
.
Rebhun
.
1978
.
Th
e
visualization
of
actin
filament
polarity
in thin
sections
.
Evidence
for
uniform
polarity
of
membrane-associated
filaments
.
J
.
Cell Biol
.
79
:846-852
.
15
.
Trinkaus,
J
.
P
.
1980
.
Formation of
protrusions
of
the
cell
surface
during
tissue
cell
movement
.
In
Tumor
Cell
Surfaces
and Malignancy
.
R
.
O
.
Hynes,
and
C
.
F
.
Fox,
editors
.
Alan
R
.
Liss,
Inc
.,
New
York
.
887-906
.
16
.
Hasty,
D
.
L
.,
and
E
.
D
.
Hay
.
1978
.
Freeze-fractur
e
studies
of
the
developing
cell
surface
.
11
.
Formation of
particle-free
membrane
blebs
during
glutaraldehyde
fixation
.
J
.
Cell
Biol
.
78
:756-768
.
17
.
Meier,
S
.,
and
E
.
D
.
Hay
.
1974
.
Stimulatio
n
of
extracellular
matrix
synthesis
in
the
developing
comes
by
glycosaminoglycans
.
Proc
.
Nall
.
Acad
.
Sci
.
U
.
S
.
A
.
71
:2310-2313
.
18
.
Hay,
E
.
D
.,
and
J
.
W
.
Dodson
.
1973
.
Secretio
n of
collagen
by
corneal epithelium
. I .
Morphology
of
the
collagenous
products
produced
by
isolated
epithelia
grow on
frozen-
killed
lens
.
J
.
Cell Biol
.
57
:190-213
.
19
.
Rubin,
K
.,
M
.
Hook,
B
.
ÖBrink,
and
R
.
Timpl
.
1981
.
Substrate
adhesion
of
rat
hepatocytes
:
mechanism
of
attachment
to collagen substrates
.
Cell
.
24
:463-470
.
20
.
Murray,
J
.
C

G
.
Stingl,
H
.
K
.
Klienman,
G
.
R
.
Martin,
and S
. 1 .
Katz
.
1979
.
Epiderma
l
cells
adhere
preferentially
to
type
IV
(basement
membrane)
collagen
.
J
.
CellBiol
.
80
:197-
202
.
21
.
Wicha,
M
.,
L
.
A
.
Liotta,
S
.
Garbisa,
and
W
.
R
.
Kidwell
.
1979
.
Basemen
t
membrane
collagen
requirements
for
attachment
and
growth
of
mammary
epithelium
.
Exp
.
Cell
Res
.
124
:181-190
.
22
.
Kleinman,
H
.
K
.,
R
.
J
.
Klebe,
and
G
.
R
.
Martin
.
1981
.
Role
of
collagenous
matrices
in
the
adhesion
and growth of
cells
.
J
.
Cell Biol
.
88
:473-485
.
23
.
Gold,
L
.
L,
and
E
.
Pearlstein
.
l9ß0
.
Fibronectin-collagen
binding
and
requirement
during
cellular
adhesion
.
Biochem
.
J
.
186
:551-559
.
24
.
Hughes,
R
.
C
.,
G
.
Mills,
and
Y
.
Courtois
.
1979
.
Roleof
fibronectin
in
the
adhesiveness
of
bovine
lens
epithelial
cells
.
Biol
.
Cell
.
36
:321-330
.
25
.
Berman,
M
.
D
.,
1
.
G
.
Eaggoner
.
J
.
N
.
Foidart,
and
H
.
K
.
Kleinman,
1980
.
Attachment
to
collagen
by
isolated
hepatocytes
from
rats
with
induced
hepatic
fibrosis
.
J
.
Lab
.
Clin
.
Med
.
95
:660-665
.
26
.
Linsenmayer,
T
.
F
.,
G
.
N
.
Smith,
and
E
.
D
.
Hay
.
1977
.
Synthesi
s
oftwo
collagen
types
of
embryonic
chick
corneal epithelium
in
vitro
.
Proc
.
Nail
.
Acad
.
Sci
.
U
.
S
.
A
.
74:39-43
.
27
.
Terranova,
V
.
P
.,
D
.
H
.
Bohrbach,
and
G
.
R
.
Martin
.
1980
.
Rol
e
of
laminin
in
the
attachment
of
PAM
212
(epithelial)
cells
to
basement
membrane
collagen
.
Cell
.
22
:719-
726
.
28
.
Sugrue,
S
.
P
.,
and
E
.
D
.
Hay
.
1981
.
Effec
t
of
protein
synthesis
inhibitors
on
interaction
of
epithelium
with
extracellular
matrix
.
Anat
.
Ree
.
199
:249A
.
(Abstr
.)
.
29
.
Kleinman,
H
.
K
.,
G
.
R
.
Martin,
and P
.
H
.
Fishman
.
1979
.
Ganglinsid
e
inhibition
of
fibronectin-mediated
cell
adhesion
to
collagen
.
Proc
Nall
.
Acad
.
Set
.
U
.
S
.
A
.
76
:3367-
3371
.
30
.
Rennard,
S
. I .,
M
.
L
.
Wind,
A
.
T
.
Hewitt,
and
H
.
K
.
Kleinman
.
1981
.
Effec
t
of
collagen
and
cell
shape
on
binding
of
fibronectin
to
cells
.
Arch
.
Biochem
.
Biophys
.
206
:205-212
.
31
.
Goldberg,
B
.
1979
.
Binding
of
soluble
type
I
collagen
molecules
to
the
fibroblast
plasma
membrane
.
Cell
.
16
:265-275
.
32
.
Rubin,
K
.,
A
.
Oldberg,
M
.
Hook, and B
.
Öbrink
.
1978
.
Adhesion of
rat
hepatocytes
to
SUGRUE
AND
HAY

ECM
Organizes
Basal
Epithelial
Surface

53
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981

collagen
.
Exp
.
Cell
Res
.
117
:165-177
.
33
.
Rubin,
K
.,
S
.
Johansson,
1
.
Pettersson,
C
.
Ocklind,
B
.
Obrink,
and
M
.
H66k
.
1979
.
Attachment of
rat
hepatocytes
to
collagens
and
fibronectin
:
a
study
using antibodies
directed
against
cell
surface
components
.
Biochem
.
Biophys
.
Res
.
Commun
.
91
:86-94
.
34
.
Geiger,
B
.,
K
.
T
.
Tokuyasu,
A
.
H
.
Dutton,
and S
.
J
.
Singer
.
1980
.
Vinculin,
and
intracellular
protein
localized
at
specialized
sites
where
microfilament
bundles
terminate
at
cell
membranes
.
Proc
.
Nail
.
Acad
Sci
.
U
.
S
.
A
.
77
:4127-4131
.
35
.
Singer,
1
.
1 .
1979
.
The
fibronexus
:
a
transmembrane
association
of
fibronectin-containing
fibers
and
bundles
of
5
nm
microfdamnents
in
hamster and
human
fibroblasts
.
Cell
.
16
:
675-685
.
36
.
Pollard,
T
.
D
.
1977
.
Cytoplasmic
contractile
proteins
.
In
International
Cell
Biology,
1976-
1977
.
B
.
R
.
Brinkley
and
K
.
R
.
Porter, editors
.
The
Rockefeller
University
Press,
New
54

THE
JOURNAL
OF
(_ELL
BIOLOGY
"
VOLUME
91,
1981
York
.
378-387
.
37
.
Dodson,
J
.
W
.
1967
.
The
differentiation
of
epidermia
.
1
.
The
interrelationship
of
epidermis
and
dermis
in
embryonic
chicken
skin
.
J
.
Embryol
Exp
.
Morphol
17
:83-105
.
38
.
Jensen,
H
.
M
.,
and
N
.
K
.
Mottet
.
1970
.
Ultrastructura
l
changes
in
keratinizing
epithelium
following
trypsinization,
epidermal
detachment
and
apposition
to
mesenchymes
.
J
.
Cell
Sci
.
6
:511-535
.
39
.
Liotta,
L
.
A
.,
S
.
Abe,
P
.
Gehron-Robey,
and
G
.
R
.
Martin
.
1979
.
Preferentia
l
digestion
of
basement
membrane
collagen
by an
enzyme
derived
from
a
metastatic
murine
tumor
.
Proc
.
Nail
.
A
cad
.
Sci
.
U
.
S
.
A
.
76
:2268-2272
.
40
.
Mauchamp,
J
.,
A
.
Margotat,
M
.
Chambard,
B
.
Charrier,
L
.
Remy,
and
M
.
Michel-Bechet
.
1979
.
Polarity
of
three-dimensional
structures
derived
from
isolated
hog
thyroid
cells
in
primary
culture
.
Cell Tissue
Res
.
204
:417-430
.
on July 12, 2011jcb.rupress.orgDownloaded from
Published October 1, 1981