1007
Journal of Cell Science 107, 1007-1018 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
Ductus arteriosus smooth muscle cell migration on collagen: dependence on
laminin and its receptors
Ronald I. Clyman1,2,*, Jamie Tannenbaum1, Yao Qi Chen1, Douglas Cooper3,5, Peter D. Yurchenco6,
Randall H. Kramer1,3,4 and Nahid S. Waleh2
1Cardiovascular Research Institute Departments of 2Pediatrics, 3Anatomy and 4Stomatology,
5Langley Porter Institute University of California, San Francisco, CA 94143-0544, USA
6Robert Wood Johnson Medical School, Piscataway, New Jersey 08854-5635, USA
San Francisco CA 94143-0544, USA
*Author for correspondence
SUMMARY
During permanent closure of the ductus arteriosus, smooth
muscle cells migrate through the extracellular matrix
(ECM) to form intimal mounds that occlude the vessel’s
lumen. Smooth muscle cells (SMC) migrate over surfaces
coated with collagen in vitro. During the migration SMC
also synthesize fibronectin (FN) and laminin (LN). Antibodies against FN and LN inhibit migration on collagen by
30% and 67%, respectively. Because of the apparent
importance of LN in migration, we examined how SMC
interact with LN and LN fragments (P1, E8, P1′, E1′, E3,
E4, and G). Ductus SMC adhere to high concentrations of
LN and to two fragments of the molecule: P1 and E8. They
use a unique set of integrin receptors to bind to LN (α1β1,
α6β1 and αVβ3), to P1 (α1β1, αVβ3), and to E8 (α6β1, αVβ3).
The αVβ3 integrin binds to the P1 fragment of LN in an
RGD peptide-dependent manner, and to the E8 fragment
in an RGD-independent manner; the RGD site on the P1
fragment probably is not available to the cell in intact LN.
Antibodies against β1 integrins completely inhibit SMC
adhesion to LN; antibodies against the αVβ3 integrin do not
block SMC adhesion to LN, but do prevent cell spreading.
LN is also capable of interfering with SMC adhesion to
other ECM components. The antiadhesive effect of LN is
located in the E1′ domain. Both exogenous and endogenous
LN increase SMC motility on collagen I. The locomotionpromoting activity of LN resides in the E1′ antiadhesive
domain, and not in its adhesive (P1, E8) domains. LN
causes a decrease in the number of focal contacts on
collagen I. This might enable SMC to alter their mobility
as they move through the extracellular matrix to occlude
the ductus arteriosus lumen.
INTRODUCTION
precoated on the assay chamber’s surface. While rapid assays
of cell adhesion and spreading are likely to represent direct
responses of the cells to molecules precoated on the surface of
the assay chamber, assays of migration, which require several
hours to complete, might also involve responses to endogenously released ECM. Indeed, Boudreau et al. (1991) found
that migration of ductus SMC on collagen could be inhibited
by antibodies against FN. We have observed that ductus SMC
secrete and organize both FN and LN between the cell surface
and a collagen-coated substratum during the several hours
required for the migration assay (Clyman et al., 1992; and
unpublished observations). Thus, even on defined substrata,
assays of SMC migration are likely to reflect complicated cell
interactions with collagens, FN, LN and perhaps other ECM
components as well. In the present study we have concentrated
on elucidating the role of LN in SMC migration.
LN is a multifunctional basement membrane glycoprotein
formed by the association of three distinct subunits to form a
large cross-shaped trimer (Fig. 1). A family of LNs has been
described, based on the association of several alternative
Permanent closure of the ductus arteriosus at birth involves
migration of smooth muscle cells (SMC) from the muscle
media through surrounding extracellular matrix (ECM) to form
intimal mounds that occlude and eventually seal the vessel
lumen. Major components of the ECM surrounding medial
SMC include laminin (LN), fibronectin (FN), vitronectin (VN),
collagens type I (Col I) and IV (Col IV), and proteoglycans (de
Reeder et al., 1989). These ECM components have been found
to affect migration of many cell types differentially, including
SMC. Therefore, changes in ECM composition might play a
role in regulating smooth muscle migration to achieve ductus
arteriosus closure.
The relative roles of individual ECM components and their
respective cell receptors in ductus SMC migration are not
clear, in large part because ECM is released from the SMC
themselves during in vitro assays. The ECM released from the
SMC can either replace (Avnur and Geiger, 1981; Haas and
Culp, 1982) or bind to (Kleinman et al., 1981) molecules
Key words: fibronectin, cell migration, collagen, vascular smooth
muscle, integrin, adhesion, antiadhesion, heparin, laminin
1008 R. I. Clyman and others
with the enzyme modifier, α-lactalbumin (Begovac and Shur,
1990). LN itself binds heparin and heparan sulfate, primarily
through its G domain (Ott et al., 1982; Yurchenco et al., 1993),
and thus has been suggested to bind to cells through cell
surface heparan sulfate proteoglycans. A few other nonintegrin LN receptors have been proposed, but are less well
characterized (Mecham, 1991; Sato et al., 1993).
SMC will adhere, spread and migrate on high concentrations
of LN. However, compared with other ECM components, LN
is only poorly adhesive for SMC (Clyman et al., 1992). We
have previously shown that ductus arteriosus SMC express
several integrin LN receptors, α1β1, αVβ3 and a receptor that
appears to be either α6β1 or α7β1 integrin (Clyman et al.,
1992). Antibody to β1 integrins blocks SMC adhesion,
spreading and migration on LN; on the other hand, antibody to
A
B
% adhesion
subunits (Engvall et al., 1990). The relative expression of
different LN family members is tissue-specific. The prototype
and most commonly studied LN is derived from murine Engelbreth-Holm-Swarm (EHS) tumors and is composed of an A
(400 kDa), a B1 (200-220 kDa) and a B2 (200-220 kDa) chain.
LN of this composition is found surrounding SMC in blood
vessels (Engvall et al., 1990; Glukhova et al., 1993). This LN
forms a tight association with the glycoprotein, entactin (Mann
et al., 1988), and also binds less tightly to several other ECM
components, including Col IV (Aumailley et al., 1989;
Charonis et al., 1985) and heparan sulfate proteoglycan
(Kleinman et al., 1993; Yurchenco et al., 1993).
For a wide variety of cell types, LN has been shown to
promote cell attachment, migration, proliferation and differentiation (Kleinman et al., 1993). However, LN has also been
reported to inhibit substratum adhesion for some cell types
(Calof and Lander, 1991). These responses have been found to
depend on a variety of cell surface LN receptors, including both
integrin and non-integrin types. Integrins are a family of
receptors involved not only in cell interactions with LN, but also
with other ECM components and with other cells. Each integrin
receptor is a heterodimer in which one of several homologous
α subunits associates noncovalently with a β subunit. Some
integrin subunit combinations recognize multiple ligands, while
others are relatively specific. Several integrins (α1β1, α2β1,
α3β1, α6β1, α7β1 and αVβ3) bind LN. Several of the integrins
have been mapped to specific domains of the LN molecule
through the use of LN proteolytic fragments (Beck et al., 1990;
Gehlsen et al., 1992; Goodman et al., 1991; Kramer et al., 1990,
1991; Mecham, 1991; Nurcombe et al., 1989; Sonnenberg et
al., 1990, 1991; Tomaselli et al., 1990; Yurchenco et al., 1993).
The major cell attachment site in LN is apparently located in
the E8 long arm fragment (Fig. 1). Both α6β1 and α7β1 integrins
bind in this region (Kramer et al., 1991; Sonnenberg et al., 1990,
1991). The binding site for α3β1 integrin has been mapped to
the carboxy-terminal G domain (Gehlsen et al., 1992; Tomaselli
et al., 1990). A site in the LN P1 short arm fragment (Fig. 1)
contains a cryptic RGD peptide sequence that is recognized by
αVβ3 integrin (Kramer et al., 1990), but this site does not seem
to be available in the native LN molecule (Aumailley et al.,
1990). α1β1 Integrin also binds to the LN P1 fragment (Fig. 1),
but in an RGD-independent manner (Goodman et al., 1991;
Tomaselli et al., 1990). The binding site for the α2β1 integrin
has not yet been identified.
Several non-integrin receptors for LN have also been identified. Some of these recognize LN peptide sequences, and
others recognize LN oligosaccharide side-chains. A 67 kDa
receptor has been shown to mediate interactions of some cells
with LN through recognition of a YIGSR peptide sequence in
the B1 chain of the P1 fragment (Graf et al., 1987). LN is very
heavily glycosylated with oligosaccharides distributed over all
three subunits (Dean et al., 1990). Several lactose-binding
lectins have been implicated in cell interactions with LN
oligosaccharides (Mecham, 1991). The 67 kDa YIGSR
receptor also binds β-galactosides, and can be eluted from LN
with lactose (Mecham, 1991). A cell surface galactosyltransferase has been found to mediate the interaction of some cells
with LN by binding to terminal glucosamine residues on LN
oligosaccharide side-chains (Begovac and Shur, 1990).
Binding of the galactosyltransferase to LN can be inhibited
Fig. 1. Ductus arteriosus SMC adhesion to LN and its fragments.
(A) Schematic model of laminin. The amino termini of the A, B1 and
B2 chains are located at the ends of the short arms. The positions of
different elastase-derived fragments (E1′, E3, E4, E8), pepsinderived fragments (P1, P1′), and the large globular domain (G) at the
end of the long arm are indicated. (Model based on Beck et al., 1990;
Ott et al., 1982; Sasaki et al., 1988). (B) Cells were allowed to attach
to wells coated with different concentrations of LN and LN
fragments. Values represent the mean percentage (± s.d.) of
hexosaminidase activity that remained in the wells in: 11 separate
experiments for LN, 6 for P1, 3 for E8, 3 for P1′, 3 for E1′, 2 for E3,
2 for E4, and 3 for G. Data from each experiment were expressed as
the mean of three wells. The percentage of cells adhering to BSAcoated wells was 2±1%.
Laminin, integrins and muscle migration 1009
αVβ3 integrin only blocks spreading and migration on LN, not
adhesion. Possible roles for non-integrin LN receptors in this
SMC behavior have not been previously investigated.
In the present study, we report that lamb ductus SMC use
both LN and FN to migrate on other ECM components, but the
influence of LN seems to predominate. On the basis of PCR
and northern analysis, we demonstrate that the α6/α7-like
integrin subunit of ductus SMC most likely represents α6. We
present evidence that ductus SMC adhere and spread on P1 and
E8, but not on other LN fragments. SMC α1β1 integrin binds
to the LN P1 fragment; α6β1 integrin binds to the LN E8
fragment; and the αVβ3 integrin binds to both fragments.
Binding of αVβ3 integrin to the P1 fragment is RGDdependent, but binding to the E8 fragment is RGD-independent. The major non-integrin LN receptors discussed above do
not seem to be important for ductus SMC interaction with LN.
We conclude that LN stimulates SMC migration on other
matrix components through an antiadhesive effect, which is
mediated by a different LN domain, E1′, and by a different
mechanism from those mediating cell adhesion to LN.
MATERIALS AND METHODS
Cell culture
Vascular SMC were isolated from medial explants of ductus arteriosus obtained from fetal lambs (106 days gestation; term, 145 days)
and characterized by their ‘hill and valley’ morphology at confluence
and their ability to be recognized by a monoclonal antibody against
smooth muscle actin (Clyman et al., 1992). Smooth muscle cells were
also isolated from medial explants of rat aorta and human fetal (19
week) pulmonary artery and similarly characterized (Clyman et al.,
1990). The smooth muscle cells were grown in monolayer culture in
Iscove’s modified Dulbecco’s medium (IDME) supplemented with
4.5 g/l glucose, 0.29 g/l glutamine, 10% fetal calf serum, 100 units/ml
penicillin, 100 µg/ml streptomycin, and 1.25 µg/ml fungizone. Cells
were passaged with trypsin-EDTA and were used between passages
3 and 8. During the experiments, the cells were incubated in IDME
with 1 mg/ml bovine serum albumin (BSA) in place of fetal calf
serum.
Antibodies
Rat
The following rat monoclonal antibodies were used: anti-human α6
integrin subunit (GoH3) (provided by Dr A. Sonnenberg, Netherlands
Cancer Institute), anti-integrin β1 subunit (AIIB2), and anti-human α5
subunit (BIE5) (provided as ascites fluid by Dr C. Damsky, University of California, San Francisco).
Mouse
The following mouse monoclonal antibodies were used: anti-integrin
β3 subunit (provided by Dr Larry Fitzgerald, University of California, San Francisco), anti-human integrin αV subunit (LM 142)
(provided by Dr D. Cheresh, Research Institute of Scripps Clinic), and
anti-vinculin (clone VIN-II-5, ICN Immunobiologicals, Lisle, IL).
Rabbit
The following polyclonal antibodies were used: anti-integrin β3
subunit (poly-β3) from platelet IIbIIIa complex (provided by Dr K.
Knudsen, Lankenau Hospital Research Center, Philadelphia, PA);
anti-integrin αVβ3 complex (Telios Pharmaceuticals, San Diego, CA);
anti-integrin αV (anti-αV) derived from the cytoplasmic domain of the
integrin αV subunit (characterized by competition ELISA and by
Western Immunoblot, and provided by Dr L. Reichardt, University of
California, San Francisco); anti-human fibronectin (anti-FN) (Cappel,
Westchester, PA); two different antisera against laminin: anti-LNGibco (Gibco BRL, Gaithersburg, MD); and anti-LN-CR (antimurine
laminin 3.7 mg/ml; Collaborative Research, Bedford, MA); antilaminin E8 fragment and anti-laminin P1′ fragment (characterized by
Western immunoblot and competition ELISA (data not shown, P.
Yurchenko), and provided by P. Yurchenko); and anti-entactin
(provided by A. Chung, University of Pennsylvania, Philadelphia,
PA).
Neither the mouse monoclonal antibody against human αV (LM
142) nor the rat monoclonal antibody against human α5 (BIE5) bound
to the corresponding ovine α subunits; they were used as control antibodies when other mouse and rat antibodies were tested. Heat-inactivated normal rabbit serum was used as a control for rabbit antisera.
Antibodies were used in concentrations that maximized their
inhibitory effects on cell adhesion or migration. The anti-human
fibronectin (anti-FN) antibody (diluted 1:40) blocked ductus arteriosus SMC adhesion (see below) to fibronectin (10 µg/ml, coating concentration) by 78±23% (m ± s.d., n=4) without affecting cell adhesion
to laminin (50 µg/ml) or collagen I (2.5 µg/ml). The anti-laminin
antibody (anti-LN-Gibco) (diluted 1:100) blocked SMC adhesion to
laminin by 67±6% (n=6) without affecting cell adhesion to fibronectin
or collagen I. The anti-laminin antibody (anti-LN-CR) (diluted 1:40)
blocked SMC adhesion to laminin by 94±7% (n=6) without affecting
cell adhesion to fibronectin or collagen I.
Extracellular matrix
Type I collagen from bovine skin was purchased from Celtrix, Palo
Alto, CA (Vitrogen 100). Fibronectin was purified from outdated
human plasma by using gelatin-Sepharose affinity chromatography
(Ruoslahti et al., 1982). Vitronectin was purified from plasma by glass
bead column chromatography and heparin-Sepharose chromatography (Ruoslahti et al., 1987). Both laminin and type IV collagen were
isolated from Englebreth-Holm-Swarm tumors grown in C57 BL/6
mice (Kleinman et al., 1982). The laminin used in these experiments
was found to be free of type IV collagen and entactin (Clyman et al.,
1990). Laminin fragments E8 and P1 (Sonnenberg et al., 1990)
(provided by R. Timpl, Max Planck Institut für Biochemie, Martinsried, Germany) and fragments E1′, E3, E4, P1′, G (Schittny and
Yurchenco, 1990; Yurchenco et al., 1993) were prepared by previously published techniques.
Other reagents
The RGD (Gly-Arg-Gly-Asp-Ser-Pro; GRGDSP), RGE (Gly-ArgGly-Glu-Ser-Pro; GRGESP) and YIGSR (Tyr-Ile-Gly-Ser-Arg)
peptides were obtained from Telios Pharmaceuticals (San Diego, CA).
Poly-L-lysine (500,000 Mr), heparin, bovine serum albumen (BSA)
and lectins from Bandeiraea simplicifolia (BS-1) and Triticum
vulgaris (wheat germ agglutinin; WGA) were obtained from Sigma,
St Louis, MO.
Cell surface radioiodination, affinity chromatography and
immunoprecipitation
The cell monolayer from three confluent 10 cm dishes was surfacelabeled with Na125I in the culture disk as previously described
(Clyman et al., 1990). Membrane proteins were solubilized by extracting the radiolabeled cells for 1 hour at 4˚C with a solution of 200 mM
octyl-β-glucopyranoside, 50 mM Tris-HCl, pH 7.4, and 1 mM
MnSO4. Protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 2
µg/ml aprotinin, and 10 mM N-ethylmaleimide) were added during
solubilization and throughout the subsequent procedures.
Affinity columns were prepared by coupling the ligand (BSA,
laminin or laminin fragments P1 and E8) to CNBr-activated
Sepharose (Pharmacia, Pleasant Hill, CA). Affinity columns, 0.5 cm
× 1.5 cm, were equilibrated in wash buffer (50 mM Tris-HCl, pH 7.4,
50 mM octyl-β-glucopyranoside, 1 mM MnSO4, 1 µg/ml BSA, and
1010 R. I. Clyman and others
0.005% NaN3). Solubilized whole-cell extract was applied, and the
columns were washed sequentially with 10 column volumes of wash
buffer, with 8 column volumes of 0.2 M NaCl in wash buffer, and
with 5 column volumes of 10 mM EDTA in wash buffer without
divalent cation (Clyman et al., 1990). In some experiments the
columns were washed with 5 column volumes of either GRGDSP (2
mM) or GRGESP (2 mM) before the EDTA wash. Fractions (1.5 ml)
were collected and either analyzed directly by SDS-PAGE or first
subjected to immunoprecipitation and then analyzed by SDS-PAGE
as previously described (Clyman et al., 1990). In addition, parallel
immunoprecipitations with control-irrelevant antibodies were
performed. Negligible radioactivity was recovered in control precipitates (not shown).
Cell adhesion/antiadhesion assay
Adhesion of ductus arteriosus SMC to protein-coated microtiter plates
was assayed as previously described (Clyman et al., 1990). Confluent
cells were removed from tissue culture plates by brief incubation (2
minutes) at room temperature with trypsin-EDTA. Following inactivation of the trypsin with 10% fetal bovine serum, the cell pellet was
washed twice with cold IDME and resuspended in cold IDME with
BSA (1 mg/ml).
The wells of polystyrene 96-well microtiter plates (Serocluster
Costar Corp., Cambridge, MA) were precoated with different concentrations of extracellular matrix proteins dissolved in sterile
phosphate-buffered saline (PBS) for 1 hour at 37˚C. The amount of
ligand that adsorbed to the polystyrene plate has been shown to be
directly related to the concentration of ligand used to coat the plate
(Calof and Lander, 1991; Hall et al., 1987; Sonnenberg et al., 1990).
The wells were then washed with PBS, and nonspecific adherence to
the coated wells was blocked with 1 mg/ml BSA in IDME for 1 hour
at 37˚C. Appropriate dilutions of antibodies or peptides were added
to the wells; following this, cells were added (2×104 cells/well) and
allowed to attach to the wells at 37˚C for 30 minutes when attaching
to FN, VN, Col I or Col IV or for 60 minutes when attaching to LN
or its fragments. The attachment interval was determined by the time
required for 95% of maximal cell attachment to a particular ligand
(Clyman et al., 1992). After washing unattached cells from the wells,
adherent cells were quantified by a colorimetric assay for hexosaminidase, a lysosomal enzyme, and the data from each experiment
were expressed as the mean of triplicate wells (Landegran, 1984).
To test the antiadhesive behavior of LN and its fragments, 96-well
polystyrene microtiter plates were precoated with I, FN, IV or VN as
described above. After the wells were rinsed with PBS, they were
coated a second time (1 hour at 37˚C) with the antiadhesive substratum to be tested. They were finally blocked with 1 mg/ml BSA as
described above. Previous experiments have shown that coating with
a second substratum does not alter the amount of the first substratum
that initially bound to the plastic (Calof and Lander, 1991; Smalheiser,
1991). In addition, the antiadhesive effects of LN and its fragments
were independent of the order in which the substrata were coated onto
the plate (e.g. Col I followed by LN; LN followed by Col I; or LN
and Col I coated simultaneously) (data not shown). The cells were
allowed to attach to the wells for 30 minutes, and the number of
adherent cells in the wells were expressed as a percentage of the
number of cells that adhered to wells coated with the original substratum alone.
Migration
To examine SMC migration, we used the circular outgrowth assay
described previously (Clyman et al., 1992). The 96-well upper
chamber of a Minifold filtration apparatus (Schleicher and Schuell,
Keene, NH) was used to coat different ligands onto an uncharged
polystyrene sheet by filling the individual wells with either collagen
I or collagen I plus LN or LN fragments as described above under
‘Cell Adhesion Assay’. The nonspecific binding sites on the polystyrene sheet were subsequently blocked by incubation with 1 mg/ml
BSA in PBS. A stainless-steel screen with 940 µm diameter circular
perforations was then inserted as a barrier between the substratumcoated polystyrene sheet and the upper 96-well template. Ductus SMC
(2×104 cells/well) were placed in the upper wells and allowed to
attach to the substratum-coated sheet for 1k hour at 37˚C. The metal
screen prevented the cells from attaching to the coated sheet except
at the perforated areas. Once the cells had attached, the screen was
removed, leaving discrete circular areas of cells attached to the coated
sheet. The upper chamber was filled with 1 mg/ml BSA in IDME
(with appropriate dilutions of antibodies when used). The cells then
were allowed to migrate out, on top of the substratum-coated sheet,
from their original 940 µm diameter circular area, to form an
enlarging circle. After 7 hours the cells were fixed with 70% ethanol
and stained with hematoxylin. The migration rate was calculated as
the increase in diameter of the circular area covered by the cells
between 0 and 7 hours. The data from each experiment were
expressed as the mean of eight wells.
Immunofluorescent staining
Glass coverslips were coated with appropriate extracellular matrix
substrata as described above (see Clyman et al., 1992). Ductus SMC
were suspended in IDME with 1 mg/ml BSA and overlaid (2×104
cells) on coverslips. After 1G hours the cells were fixed and permeabilized as previously described (Kramer et al., 1989). The samples
were incubated with anti-vinculin (1:100) followed by goat antimouse IgG antisera coupled with rhodamine (1:350) (Boehringer
Mannheim, Indianapolis, IN). The coverslips were mounted in Fluoromount (Fisher Scientific Co., Santa Clara, CA) and viewed on a
Nikon microscope equipped with epiluminescence optics.
Molecular biology
A 15 µg sample of total RNA, isolated from cultured cells by the
guanidinium-cesium chloride method (Ausubel et al., 1987), was electrophoresed in 1% agarose gel containing 6% formaldehyde (Maniatis
et al., 1982). The RNAs were transferred to nylon membranes
(Amersham Hybond N) and hybridized with the appropriate cDNA
probe (Maniatis et al., 1982). The filters were washed twice with 1×
SSC, 0.1% SDS at room temperature for 15 minutes and once at 55˚C
with 0.1× SSC, 0.1% SDS for 1 hour. They were then exposed to Xray film at −70˚C.
Total RNA (20 µg) was used as a template for cDNA synthesis by
the polymerase chain reaction (PCR). PCR was carried out using a
commercially available thermal cycler (Cetus, Emeryville, CA). The
degenerate PCR primers were designed on the basis of the known
sequences of the extracellular regions of the α6 and α7 integrin
subunits (Song et al., 1992; Tamura et al., 1990).
The PCR primers for α7 and the cDNA probe for α6 (ovine) were
provided by Dr Robert Pytela (University of California, San
Francisco, CA). The cDNA probes homologous to the extracellular
domain of the respective molecules for α6 (murine), α7 (murine) and
α7 (human) were generated in our laboratory, and their identities were
confirmed by sequencing (Sanger et al., 1977).
Glycosidase digestion
To examine the influence of laminin glycosylation on cell adhesion
or antiadhesion, 96-well microtiter plates were precoated with either
LN (50 µg/ml) (for LN adhesion assay) or collagen I (2.5 µg/ml) with
or without LN (10 µg/ml) (for LN antiadhesion assay). The wells were
washed with PBS, and blocked with 1 mg/ml BSA as described above.
The coated wells were incubated overnight at 37˚C with combinations
of α-galactosidase (0.3 units/ml), neuraminidase (0.3 units/ml), endoβ-galactosidase (0.1 units/ml), hexosaminidase (0.1 units/ml), and βgalactosidase (0.2 units/ml) in 0.1 M sodium citrate (pH 5) with 1 mM
EDTA and 1 mg/ml BSA. The wells were washed three times with 5
mg/ml BSA in PBS and used in the cell adhesion or antiadhesion
assay. Using various labeled plant lectins to quantify exposed saccharides, we have found that such digestion removes 60% of terminal
Laminin, integrins and muscle migration 1011
α- and β-galactose, 100% of terminal sialic acid, 90% of terminal glucosamine, and 95% of linear polylactosamine (data not shown). All
enzymes were obtained from Sigma Company (St Louis, MO), except
for endo-β-galactosidase, which was generously provided by Dr
Michiko Fukuda (La Jolla Cancer Research Foundation, La Jolla,
CA).
RESULTS
SMC migration is modulated by endogenous
fibronectin and laminin
Previous metabolic labeling studies have shown that ductus
arteriosus SMC in culture make FN and LN (Boudreau and
Rabinovitch, 1991). By immunostaining, we found that
ductus SMC also deposited FN and LN under the cell surface
when plated on collagen I-coated glass coverslips for 7 hours
(the duration of our migration assay) (data not shown). When
added exogenously, both FN and LN increased the migration
of SMC on collagen I (Fig. 2). Even in the absence of added
FN and LN, antisera against these proteins inhibited
migration of SMC on Col I (Fig. 2). Therefore, ductus SMC
appeared to use the FN and LN that they secreted during the
migration assay to facilitate their migration on collagen I. LN
appeared to play a more significant role, because the RGD
peptide, which inhibited ductus SMC migration on FN but
Fig. 2. Migration of ductus arteriosus smooth muscle cells on
collagen Col I (5 µg/ml). Cells were allowed to out-migrate on
polystyrene sheets coated with either Col I (5 µg/ml), Col I (5 µg/ml)
plus FN (10 µg/ml) (+FN), or Col I (5 µg/ml) plus LN (10 µg/ml)
(+LN). Antisera against LN (anti-LN (Gibco), diluted 1:100) or
against FN (anti-FN, diluted 1:40) were added to the media of some
wells. The initial diameter at 0 hours was 940±5 µm. Values
represent the means ± s.d. of the increase in diameter over 7 hours of
the circular cell area in six experiments. Data from each experiment
were expressed as the mean of eight wells. By using a paired
Student’s t-test with a Bonferoni correction for multiple
comparisons, we found that in comparison with the migration of
SMCs on collagen I alone (control), the addition of FN (P<0.05) or
LN (P<0.05) increased the migration of SMCs, whereas the addition
of antisera against LN (P<0.05) or against FN (P<0.05) inhibited the
migration of SMCs. SMCs migrating in the presence of anti-FN
antisera, on sheets coated with FN alone, were completely inhibited
by the antisera, and their diameter was 0±4% (n=3) of the diameter
of cells migrating on FN alone. Similarly, SMCs migrating in the
presence of anti-LN (Gibco), on LN, increased their diameter to only
11±9% (n=4) of cells migrating on LN alone.
not on LN, did not affect migration on collagen I (Clyman et
al., 1992).
SMC use integrins to adhere to and spread on
laminin and its fragments
Ductus arteriosus SMC adhered to and spread on high coating
concentrations of LN (Table 1, Fig. 1). Fragmentation of LN,
by limited proteolysis with elastase or pepsin, revealed that
ductus SMC adhered primarily to two regions of the LN
molecule: the P1 and E8 regions (Fig. 1). Antibodies to β1
integrins blocked SMC adhesion and spreading on P1 or E8,
while anti-αVβ3 antiserum merely inhibited cell spreading on
these fragments (Table 1). These effects were similar to those
seen with intact LN.
Several experiments suggested that SMC interact with the P1
fragment in a different manner than with intact LN: (1) heating
LN to 80˚C destroyed its ability to support cell adhesion, but
did not significantly alter cell adhesion to P1; (2) the RGD
peptide blocked the cells’ ability to adhere to and spread on P1,
but had no effect on their adhesion and spreading on intact LN;
and (3) anti-αVβ3 antiserum blocked cell spreading on LN, but
had no effect on spreading on P1 (Table 1).
SMC interactions with the E8 fragment were more akin to
those with intact LN. Antibodies to β1 integrins blocked
spreading and adhesion, whereas antiserum to αVβ3 only
blocked spreading. However, while antiserum against E8
blocked SMC adhesion and spreading on E8, it had no effect
on adhesion and spreading on intact LN. Even when combined,
the antisera to E8 and P1 had no effect on SMC adhesion to
intact LN. Therefore, there might be another binding site for
β1 integrins on intact LN that was not present in the tested
fragments.
Although antibodies against β1 integrins almost completely
blocked SMC adhesion and spreading on LN, it remained
possible that SMC also used nonintegrin receptors to interact
with LN. However, we found no evidence for involvement of
either galactosyltransferase or lactose-binding lectins. Neither
α-lactalbumin (5 µg/ml) nor digestion of LN carbohydrate
side-chains with various glycosidases, alone or in combination
(see Materials and Methods), had any noticeable effect on the
adhesion and spreading of ductus SMC (data not shown).
Identification of SMC integrins that bind LN and LN
fragments
We used a series of antibodies and antisera to individual
integrin α and β subunits to identify cell surface proteins that
were eluted from a LN affinity column with EDTA. From our
previous results (Clyman et al., 1992), ductus arteriosus SMC
used α1β1 (185/116 kDa: Fig. 3, lane 3) and αVβ3 (145/90 kDa,
Fig. 3, lanes 4, 5) to bind to LN. Anti-β1 antibody precipitated
a second α subunit (130 kDa), in addition to the α1 subunit,
which migrated just above the β1 subunit under nonreducing
conditions (Fig. 3, lane 3), and ahead (110 kDa) of the β1
subunit under reducing conditions (Fig. 3, lane 7). These electrophorectic characteristics resembled those of either α6 or α7.
The rat monoclonal antibody to α6, GoH3, did not cross-react
with the sheep α6 subunit, and therefore failed to identify this
as α6 (Fig. 3, lane 6). As a result, we examined the SMC for
the presence of either α6 or α7 mRNA. Using a sheep-specific
cDNA probe for α6 and 3′ and 5′ oligonucleotide PCR primers
for α6 and α7, we were able to detect α6 mRNA by both
1012 R. I. Clyman and others
Table 1. Factors affecting ductus SMC adhesion to laminin and its P1 and E8 fragments
A. Adhesion (%)*
LN (50 µg/ml)
P1 (25 µg/ml)
E8 (25 µg/ml)
Control
Anti-β1
Anti-αVβ3
Anti-PI′
Anti-E8
36(±6)
51 (±5)
15 (±2)
2 (±0)
24 (±3)
1 (±0)
35 (±5)
55 (±2)
13 (±2)
38 (±5)
4 (±2)
14 (±2)
34 (±6)
50 (±4)
3 (±2)
Control
Anti-β1
Anti-αVβ3
Anti-PI′
Anti-E8
++
+++
++
–
±
–
±
+++
–
++
–
++
++
+++
–
Anti-PI′ +
anti-E8
39 (±6)
RGD
Heat
34 (±5)
3 (±1)
13 (±2)
2 (±1)
40 (±3)
0 (±0)
B. Spreading†
LN (50 µg/ml)
P1 (25 µg/ml)
E8 (25 µg/ml)
Anti-PI′ +
anti-E8
++
RGD
++
–
++
*Values represent the mean percentage (±s.d.) of hexosaminidase activity that remained in the wells in four separate experiments. Data from each experiment
were expressed as the mean of three wells.
†Cell spreading in four experiments was determined qualitatively: +++ → –, well spread → round.
Cells were allowed to attach and spread on LN, P1 and E8 for 60 minutes. The assay was performed in the presence of control antiserum or antibody (see
Materials and Methods: Antibodies), anti-β1 (diluted 1:1000), anti-αVβ3 (diluted 1:50), anti-laminin P1′ fragment (anti-P1′, 20 µg/ml), anti-laminin E8 fragment
(anti-E8, 40 µg/ml), the combination of anti-P1′ and anti-E8, or GRGDSP (RGD) peptide, 0.025 mM. In some experiments, LN, E8 and P1 were heated to 80˚C
for 10 minutes, then cooled on ice before coating onto the plastic wells. The percentage of cells adhering to BSA-coated wells was 2 ± 1%.
northern analysis and PCR. No α7 mRNA could be detected
by PCR (data not shown). These findings were similar to those
with vascular SMC from other sources: (1) in rat aorta SMC
α6 but not α7 mRNA could be detected by northern analysis
and PCR; and (2) in human pulmonary arterial SMC, the α6
antibody, GoH3, both inhibited SMC adhesion to LN and
immunoprecipitated the α6 subunit; in contrast, in human
pulmonary arterial SMC, α7 mRNA could not be detected by
either northern analysis or PCR (data not shown). On the basis
of these observations, we concluded that the 130/116 kDa αβ1
integrin that bound LN represented α6β1.
EDTA eluted three integrins from the P1-LN fragment
affinity column: α1β1 and minor amounts of αVβ1 and αVβ3.
Integrin α6β1 was not detected in this eluate (Fig. 3, lanes 8-
12). EDTA eluted two integrins from the E8-LN fragment
affinity column: α6β1 and smaller amounts of αVβ3 (Fig. 3,
lanes 13-18).
Whereas the RGD peptide eluted αVβ3 from the P1Sepharose column, it had no effect on integrin binding
(including αVβ3) to either the LN- or E8-Sepharose columns
(data not shown). This finding supported the cell adhesion data,
which suggested that SMC recognize an RGD sequence in the
P1 fragment, but not in intact LN.
Cell antiadhesion by laminin and its fragments
Ductus SMC adhered to collagen I-coated wells. However,
when the wells were coated with both Col I and LN, cell
adhesion was poorer than with Col I alone (Fig. 4). This anti-
Fig. 3. Affinity chromatography of detergent-solubilized ductus SMC on laminin (1-7), and laminin fragments P1 (8-12) and E8 (13-18).
Polypeptides eluted with 10 mM EDTA from their respective affinity columns (lanes 1, 8, 13) were immunoprecipitated with monoclonal
antibodies against the β1 (AIIB2) (lanes 3, 7, 10, 15, 18), the β3 (lanes 4, 11, 16), and the human α6 (GoH3) (lane 6) integrin subunits as well as
with antiserum against the αV (anti-αV) (lanes 5, 12, 17) subunit. Control rabbit (lane 2), mouse (lane 9), and rat (lane 14) antibodies (see
Materials and Methods) were also used. The gels were run under nonreducing conditions, except lanes 7 and 18 (reduced). Molecular mass
markers, in kDa, are shown on the right.
Laminin, integrins and muscle migration 1013
Fig. 4. The effect of laminin and its fragments on ductus arteriosus
SMC adhesion to collagen Col I. Wells were first coated with Col I
(2.5 µg/ml). Next they were coated with different concentrations of
LN or its fragments. The percentage of cells adhering to Col I alone
was 61±11%. Values represent the mean number of cells adhering to
the combination of Col I+LN (or one of its fragments) as a
percentage of the number of cells adhering to Col I alone. Values
represent the mean (± s.d.) of 5 separate experiments for LN (25), 12
for LN (10), 10 for LN (5), 4 for LN (2.5), 3 for P1 (25), 2 for P1′
(25), 3 for E8 (25), 4 for E1′ (25), 4 for E1′ (10), 2 for E3 (25), 2 for
E4 (25), and 2 for G (25). Data from each experiment were expressed
as the mean of three wells. The percentage of cells adhering to BSAcoated wells was 5±2%. In each and every one of the 5 experiments
with LN (25), the 12 experiments with LN (10), the 10 experiments
with LN (5), the 4 experiments with E1′ (25), and the 4 experiments
with E1′ (10), the number of cells that adhered to collagen I alone
was greater than the number that adhered to the combination of
collagen I+LN or collagen I+E1′.
adhesive effect of LN could be mimicked by the E1′ fragment
of LN, but not by the P1, E8, E3, E4, G or P1′ fragments (Fig.
4). The antiadhesive effect of LN was seen at LN coating concentrations (5-10 µg/ml) below those needed for cell attachment to LN (Figs 1, 4).
The ability of LN to inhibit cell adhesion was not limited to
a Col I substratum. It also was seen with other substrata that
depended on integrin-mediated adhesion (FN, IV and VN; not
shown) as well as with one that did not (poly-L-lysine) (Fig.
5). The degree of LN inhibition depended on the coating concentration of the adhesive substratum. When substrata were
coated with concentrations that exceeded the amount needed
to obtain maximal cell adhesion (see Clyman et al., 1992),
inhibition by LN was markedly reduced (Fig. 5).
To investigate the molecular basis for the antiadhesive
activity of LN, we used heat denaturation, polyclonal antisera,
peptides, glycosidase digestion and plant lectins: heating LN
(or E1′ (10 µg/ml); data not shown) to 80˚C before use in the
assay virtually eliminated the antiadhesive activity (Fig. 6). Of
two polyclonal antisera that specifically blocked SMC
adhesion to LN (anti-LN (Gibco) and anti-LN (CR); see
Materials and Methods), only anti-LN (Gibco) also blocked
LN’s antiadhesive effects. Neither antiserum affected the
ability of the cells to adhere to Col I (Fig. 6). Antisera against
LN fragments P1′ and E8 also had no effect on LN’s antiadhesive activity. These results, together with those obtained with
LN fragments (above), indicated that the antiadhesive activity
Fig. 5. The effect of laminin on ductus arteriosus SMC adhesion to
fibronectin (FN), collagen IV (IV), and poly-L-lysine (PLL). Wells
were first coated with different concentrations of FN (10, 25 µg/ml),
IV (0.5, 5.0 µg/ml), and PLL (2.5 and 5.0 µg/ml). Next they were
coated with LN (10 µg/ml). Values represent the mean percentage (±
s.d.) of hexosaminidase activity that remained in the wells in 4
separate experiments for FN, 3 for IV, and 2 for PLL. Data from
each experiment were expressed as the mean of three wells. The
percentage of cells adhering to BSA-coated wells was 5±3%. In each
and every experiment, the number of cells that adhered to either FN
(10, 25 µg/ml), IV (0.5 µg/ml) or PLL (2.5 µg/ml) exceeded the
number of cells that adhered to LN + the respective ligand.
was mediated by LN domain(s) distinct from those mediating
SMC adhesion.
A variety of evidence argued against involvement of wellknown integrin or nonintegrin receptors in the antiadhesive
effect of LN on SMC. Antibodies against the β1 integrins and
the αVβ3 integrin did not alter LN’s antiadhesive action (Fig.
6). Neither the LN peptides RGD (0.5-0.1 mM) nor YIGSR
(1.0-0.1 mM) had any significant effect on LN’s antiadhesive
behavior (data not shown).
The lectins WGA and BS-1 inhibited the antiadhesive properties of LN without altering the ability of the cells to bind to
Col I (Fig. 6). These lectins also restored the ability of the cells
to spread on Col I in the presence of LN. The blocking action
of BS-1 could be reversed by including the competitive sugar,
galactose (10 mM) in the incubation medium (Fig. 6).
Although these lectins inhibited the antiadhesive activity, it is
unlikely that the carbohydrate chains that decorate LN were
responsible for this behavior, since: (1) heat destroyed LN’s
antiadhesive activity; (2) high concentrations of lactose (25
mM), mannose (25 mM), maltose (25 mM), galactose (25
mM), α-lactalbumin (5 µg/ml) and heparin (1.0 and 0.1 mg/ml)
had no effect on LN’s antiadhesive behavior (data not shown);
and (3) extensive deglycosylation of LN with various glycosidases (alone or in combination, see Materials and Methods) did
not alter its ability to block adhesion to Col I (data not shown).
These results suggested that cross-linking of LN saccharides
by the polyvalent lectins could overcome LN’s antiadhesive
effect.
Effects of laminin on cell migration
To test whether LN affected SMC migration on Col I through
its adhesive or its antiadhesive domains, we used specific LN
1014 R. I. Clyman and others
Fig. 6. The effect of laminin on ductus
arteriosus SMC adhesion to collagen Col I
and vitronectin. Wells were first coated with
Col I (2.5 µg/ml) or VN (10 µg/ml). Some
wells were then coated with LN (5 µg/ml). In
some experiments, wells were also coated
with WGA (100 µg/ml) or BS-1 (100 µg/ml).
In some experiments, either galactose (10
mM) or antibodies were added to the cell
suspension in the well: anti-LN (Gibco)
(1:100 dilution), anti-LN (CR) (1:40), anti P1′
(20 µg/ml), anti-E8 (40 µg/ml), anti-entactin
(1:10), anti-αVβ3 (1:50) and anti-β1 (AIIB2)
(1:1000). Values represent the mean (± s.d.)
number of cells adhering to the wells of a
particular condition as a percentage of the
number of cells adhering to either Col I alone
or VN alone. n represents the number of
separate experiments for each condition. Data
from each experiment were expressed as the
mean of six wells. Because ductus SMC
primarily use β1 integrins to bind to collagen
Col I and β3 integrins to bind to VN (Clyman
et al., 1992), we used anti-β1 antibody when
cells adhered to VN and anti-αVβ3 antiserum
when cells adhered to collagen Col I to
examine the role of integrins in LN’s
antiadhesive activity. The ability of LN to
inhibit cell adhesion to a collagen I
substratum was not blocked by the following
antisera – anti-LN (CR), anti-P1′, anti-E8,
anti-entactin and anti-αVβ3 – because in each
and every experiment, the number of cells that
adhered to collagen I alone was greater than
the number of cells that adhered to collagen
I+LN in the presence of the respective
antiserum. In each of the 4 experiments using
the lectin BS-1 plus collagen I+LN, the
presence of galactose reversed the blocking
action of BS-1 on LN’s antiadhesive effects.
fragments, lectins and antibodies that had different actions on
cell adhesion and antiadhesion (Fig. 7). When surfaces were
coated with both Col I and LN, cell migration was increased
compared with migration on Col I alone. This could be
mimicked when the antiadhesive LN fragment E1′ was used
instead of LN, but not when the adhesive fragments P1 and E8
were used. The anti-LN (CR) antibody, which blocked LN’s
adhesive properties but not its antiadhesive properties, had no
effect on SMC migration on substrata other than LN (e.g. Col
I or FN) (Fig. 7 and legend, Fig. 7). In contrast, the anti-LN
(Gibco) antibody, which blocked both LN’s adhesive and antiadhesive properties, markedly inhibited SMC migration on
collagen Col I and FN as well as migration on LN (Fig. 7 and
legend). The lectins WGA and BS-1, which inhibited the antiadhesive properties of LN, blocked LN’s stimulatory effect on
migration without affecting the ability of cells to migrate on
Col I alone. These results suggest that LN affected SMC
migration on Col I through its antiadhesive activity.
Effects of laminin on focal adhesion plaques
When surfaces were co-coated with LN in addition to either
Col I or FN, there was a change in the distribution and a loss
in the number of focal adhesion plaques that formed under the
cells after adhering to the surface (Fig. 8).
DISCUSSION
Ductus arteriosus SMC make both LN and FN during the time
they are being assayed for in vitro migration. FN facilitates the
migration of ductus arteriosus SMC through collagen gels
(Boudreau et al., 1991). LN, also, facilitates ductus arteriosus
SMC migration on collagen, since antibodies to LN can inhibit
SMC migration. In the present study we examined how ductus
arteriosus SMC interact with LN.
Laminin, integrins and muscle migration 1015
Fig. 7. SMC migration on collagen Col I. Polystyrene sheets were
first coated with Col I (2.5 µg/ml). Some sheets were then coated
with LN (5 µg/ml) or its fragments E1′ (25 µg/ml), P1′ (25 µg/ml),
or E8 (25 µg/ml). In some experiments, sheets were also coated with
lectins WGA (100 µg/ml) or BS-1 (100 µg/ml). Antisera against LN
(anti-LN (Gibco) diluted 1:100 or anti-LN (CR) diluted 1:40) were
added to the media during the migration phase of some experiments.
Values represent the mean (± s.d.) increase in diameter over 7 hours
during a particular migration condition as a percentage of the
increase in diameter of cells migrating on Col I alone. n represents
the number of separate experiments for each condition. Data from
each experiment were expressed as the mean of six wells. In each
and every one of the 6 experiments with LN and the 3 experiments
with E1′, cells migrating on collagen I alone increased their diameter
less than those migrating on either collagen I+LN or collagen I+E1′.
In each and every one of the 6 experiments with anti-LN (Gibco),
cells migrating on collagen I alone increased their diameter more
than those migrating in the presence of the antiserum. SMCs
migrating in the presence of anti-LN-Gibco antisera, on sheets
coated with LN (50 µg/ml), increased their diameter only to 11±9%
(n=4) of cells migrating on LN alone; similarly, those migrating in
the presence of anti-LN-Gibco antisera, on FN (10 µg/ml), increased
their diameter only to 34±6% (n=3) of cells migrating on FN alone.
In contrast, cells migrating in the presence of anti-LN (CR) antisera,
on LN, increased their diameter only to 40±15% (n=4) of cells
migrating on LN alone, whereas cells migrating in the presence of
anti-LN (CR) antisera, on FN, were not inhibited in their migration
and increased their diameter to 108±16% (n=3) of cells migrating on
FN alone.
SMC adhesion
Ductus arteriosus SMC express several integrin receptors of
the β1 and β3 families that have been found to bind LN in other
cell types: αVβ3, α1β1, α2β1, α3β1 and an integrin with electrophoretic properties like α6β1 or α7β1 (Clyman et al., 1992).
In the present study, we found that the α6/α7β1-like integrin
represents the α6β1 integrin, based on PCR and northern
analysis of mRNA extracted from ductus SMC. Consistent
results were also found for two other vascular SMC types.
Ductus SMC use a unique set of integrins to bind to LN: α1β1,
α6β1 and αVβ3 (Clyman et al., 1992). In these cells, α2β1 and
α3β1 do not appear to bind to LN. The β1 integrins on ductus
SMC bind to two different domains of the LN molecule: P1
(by α1β1) and E8 (by α6β1) (Table 1, Figs 1, 3). The αVβ3
integrin binds to both the P1 and E8 domains of LN (Fig. 3):
binding of the αVβ3 to the P1 domain occurs in an RGDdependent manner. This RGD site is probably not available to
the cell in intact LN because RGD does not elute the αVβ3
integrin from an intact LN affinity column, nor does it alter the
cells’ ability to bind to intact LN; conversely, heating LN
destroys its ability to support cell adhesion, but only slightly
reduces cell adhesion to the P1 fragment (Table 1). Binding of
the αVβ3 integrin to the E8 domain of LN occurs in an RGDindependent manner, and appears to play a role in cell
spreading (Table 1) and migration (Clyman et al., 1992) on
intact LN.
Although antisera against the P1′ or E8 fragments blocked
cell adhesion to the P1 and E8 fragments, they did not effectively block adhesion to intact LN. This suggests that there may
be another binding site for β1 integrins on intact LN that is
either destroyed or too weak to support adhesion by itself
following enzymatic digestion.
Laminin as an anti-adhesive substratum
Another prominent function of LN is its ability to interfere with
SMC adhesion to other ECM components. This effect cannot
simply be attributed to laminin’s displacement of other ECM
components from the culture surface (Calof and Lander, 1991;
Dean et al., 1990; Smalheiser, 1991). The antiadhesive effect
also does not depend on the order in which the surface is coated
with LN or other ECM components. Our results indicate that
the antiadhesive effect of LN is located in a specific domain,
the E1′ domain, at the opposite end of the molecule from the
major cell binding region, the E8 domain.
The mechanism by which LN inhibits SMC adhesion to
collagen and other ECM substrata is still unknown. We have
considered two general possibilities: (1) based on its large size,
its high degree of glycosylation, and its poor intrinsic ability
to promote SMC adhesion, LN might simply sterically
interfere with cell recognition of other adhesive substratum
molecules; or (2) a specific receptor might recognize LN and
trigger an antiadhesive response. There is a maximum spacing
for any adhesive molecule beyond which cells no longer will
be able to form local contacts or even to spread (Massia and
Hubbell, 1991). We think it unlikely that LN decreases cell
adhesion by hiding other adhesive molecules on the plate;
according to this hypothesis, when the lectins WGA and BS-1
and the anti-LN (Gibco) antibody bind to the LN molecule,
they should increase the bulk of the molecule, thereby increasing its antiadhesive effect; however, they have the opposite
effect and block LN’s antiadhesive action. In addition, the antiadhesive effect is eliminated by heat denaturation.
If the antiadhesive effect of LN is mediated by a specific
receptor, it does not appear to be the 67 kDa YIGSR binding
protein or a β1 or β3 integrin as neither the YIGSR peptide nor
specific anti-integrin antibodies inhibit this activity (Fig. 6).
Two plant lectins, WGA and BS-1, inhibit the antiadhesive and
antispreading properties of LN (Fig. 6), which suggests that
carbohydrate chains might be important for antiadhesion. Carbohydrate chains on LN previously have been implicated in
cell spreading and neurite outgrowth on LN (Dean et al., 1990).
1016 R. I. Clyman and others
Fig. 8. Effect of laminin on focal adhesion plaque formation. Ductus arteriosus SMC adhered to glass coverslips coated with either Col I (2.5
µg/ml) (with or without LN; 10 µg/ml) or FN (10 µg/ml) (with or without LN; 10 µg/ml). The cells were stained with antibodies to vinculin to
visualize the adhesion plaques. In the presence of LN, there was both a decrease in the number of adhesion plaques and a loss of adhesion
plaques under the center of the cell.
A cell surface galactosyltransferase has been shown to bind to
exposed glucosamine residues on LN and thereby promote cell
spreading and migration of some cell types (Begovac and Shur,
1990). Several members of the lactose-binding lectin family
bind to polylactosamine chains on LN and have been postulated to play a role in cell spreading and migration (Mecham,
1991; Sato et al., 1993). We therefore considered the possibility that the antiadhesive effect of LN might be mediated by LN
carbohydrate chains and that inhibition by WGA and BS-1
might result from their ability to shield specific carbohydrate
moieties. However, neither α-lactalbumin, a specific inhibitor
of galactosyltransferase binding to LN, nor extensive deglycosylation of LN reduced its antiadhesive activity. It therefore
seems most likely that when the plant lectins bind to LN carbohydrate chains, they sterically block the recognition of
adjacent polypeptide sequences by a specific antiadhesion
receptor.
The antiadhesive action of LN appears to be located only in
the E1′ fragment. The antiadhesive effects are seen with
coating concentrations of LN that are below those necessary to
promote cell adhesion to LN. As LN concentration is
increased, ultimately the cell is able to bind to the cell binding
domain at the opposite end of the LN molecule. Therefore, at
coating concentrations above 25 µg/ml, LN has less of an antiadhesive effect than it does at concentrations below 10 µg/ml.
The antiadhesive effect of LN, at high concentrations, can be
restored by the presence of heparin in the assay, which specifically blocks the cell’s ability to attach to LN (data not shown).
Different plastic surfaces have previously been shown to
evoke different cell attachment activities from the same LN
fragments (Sonnenberg et al., 1990). This does not appear to
be due to differential protein adsorption (Goodman et al.,
1991); rather, the different plastic surfaces are believed to alter
the conformation of the adsorbed LN so that different domains
are more or less available to the cells (Lindon et al., 1986). In
our experiments, we used tissue culture-treated plastic as well
as non-tissue-culture-treated plastic wells. There was no
noticeable difference between the two surfaces in the ability of
cells to attach to different fragments of LN. In contrast, when
the experiments were performed with tissue culture plastic,
both LN and E1′ had a much greater antiadhesive effect than
when the experiments were performed with non-tissue-culturetreated polystyrene plates (experiments not shown).
The antiadhesive effects observed in our study are not
unique to LN or to the ductus SMC. LN also inhibits adhesion
of olfactory epithelial neuronal cells (Calof and Lander, 1991).
Similarly, tenascin (Spring et al., 1989) and thrombospondin
(Murphy-Ullrich and Hook, 1989) can inhibit adhesion of
certain cell types to other matrix components.
Antiadhesion and SMC migration
The antiadhesive effect of LN seems to facilitate cell
Laminin, integrins and muscle migration 1017
migration. Both exogenously added and endogenously synthesized LN increase ductus arteriosus SMC motility on Col I
(Fig. 7). The locomotion-promoting activity of LN resides in
the E1′ antiadhesive domain of LN and not in its adhesive (P1,
E8) domains. SMC migration on Col I is inhibited by lectins
(WGA and BS-1) and anti-LN antisera (anti-LN; Gibco) that
inhibit the antiadhesive properties of LN. In contrast, an
antiserum (anti-LN; CR) that only inhibits the adhesive properties of LN has no effect on SMC migration on collagen. Calof
and Lander (1991) similarly noted a correlation between LN’s
antiadhesive activity and its stimulation of cell migration.
The mechanism by which LN promotes migration through
its antiadhesive domain is unclear. Strong substratum attachment might be antithetical to motility (Neugebauer and
Reichardt, 1991; Trinkaus, 1985). Agents that decrease passive
substratum attachment sites have been shown to accelerate the
migration of neuronal cells (Smalheiser, 1991). However,
simply limiting the availability of attachment sites on the substratum does not necessarily increase motility (Calof and
Lander, 1991; Chan et al., 1992). In fact, antibodies that block
substratum cell binding domains (anti-FN antibody or anti-LN
(CR) antibody) can inhibit SMC migration on the respective
substratum (FN or LN) (legend, Fig. 2). Some balanced level
of substratum attachment might be optimal for migration
(Akiyama et al., 1989; Chan et al., 1992; DiMilla et al., 1993;
Giancotti and Ruoslahti, 1990). This might involve particular
types and organizations of substratum receptors. For example,
the integrin receptors on locomoting cells have high lateral
mobility, whereas those on stationary cells are relatively
immobile (Duband et al., 1988). Stationary cells form focal
contacts that seem to act as localized sites of very strong
adhesion, ill suited for migration (Tucker et al., 1985). In
contrast, migrating cells have fewer vinculin-positive focal
contacts and diminished stress fiber organization (Abercrombie et al., 1971; Couchman and Rees, 1979; Duband et al.,
1988). We found that LN inhibits SMC spreading and
formation of focal contacts (Fig. 8). If LN’s antiadhesive
signals cause cytoskeletal reorganization or destabilization,
this may enable it to facilitate migration. Our data support the
idea that during closure of the ductus arteriosus, SMC facilitate their own migration by secreting LN, which interferes with
adhesion to collagen I and other matrix components.
We thank Ms Françoise Mauray for her help in developing the
migration assay and Mr Paul Sagan for his expert editorial assistance.
This work was supported by a grant from the US Public Health
Service, National Heart, Lung and Blood Institute HL 46691, and by
a gift from Perinatal Associates Research Foundation.
REFERENCES
Abercrombie, M., Heaysman, J. and Pegrum, S. M. (1971). The locomotion
of fibroblasts in culture. Exp. Cell Res. 65, 359-367.
Akiyama, S. K., Yamada, S. S., Chen, W.-T. and Yamada, K. M. (1989).
Analysis of fibronectin receptor function with monoclonal antibodies: roles
in cell adhesion, migration, matrix assembly, and cytoskeletal organization.
J. Cell Biol. 109, 863-875.
Aumailley, M., Wiedemann, H., Mann, K. and Timpl, R. (1989). Binding of
nidogen and the laminin-nidogen complex to basement membrane collagen
type IV. Eur. J. Biochem. 184, 241-248.
Aumailley, M., Gerl, M., Sonnenberg, A., Beutzmann, R. and Timpl, R.
(1990). Identification of the Arg-Gly-Asp sequence in laminin A chain as a
latent cell binding site being exposed in fragment P1. FEBS Lett. 262, 82-86.
Ausubel, F., Brent, R., Kingston, R. E., Seidman, J. G., Smith, J. A. and
Struhl, K., eds (1987). Current Protocols in Molecular Biology, vol. I.
Green Publishing Assoc. and Wiley-Interscience, New York.
Avnur, Z. and Geiger, B. (1981). The removal of extracellular fibronectin
from areas of cell-substratum contact. Cell 25, 121-132.
Beck, K., Huynter, I. and Engel, J. (1990). Structure and function of laminin:
anatomy of a multidomain glycoprotein. FASEB J. 4, 148-160.
Begovac, P. C. and Shur, B. D. (1990). Cell surface galactosyltransferase
mediates the initiation of neurite outgrowth from PC12 cells on laminin. J.
Cell Biol. 110, 461-470.
Boudreau, N. and Rabinovitch, M. (1991). Developmentally regulated
changes in extracellular matrix in endothelial and smooth muscle cells in the
ductus arteriosus may be related to intimal proliferation. Lab. Invest. 64, 187199.
Boudreau, N., Turley, E. and Rabinovitch, M. (1991). Fibronectin, hyaluron
and a hyaluron binding protein contribute to increased ductus arteriosus
smooth muscle cell migration. Dev. Biol. 143, 235-247.
Calof, A. L. and Lander, A. D. (1991). Relationship between neuronal
migration and cell-substratum adhesion: laminin and merosin promote
olfactory neuronal migration but are anti-adhesive. J. Cell Biol. 115, 779794.
Chan, B. M. C., Kassner, P. D., Schiro, J. A., Byers, H. R., Kupper, T. S.
and Hemler, M. E. (1992). Distinct cellular functions mediated by different
VLA integrin α subunit cytoplasmic domains. Cell 68, 1051-1060.
Charonis, A. S., Tsilibary, E. C., Yurchenco, P. D. and Furthmayr, H.
(1985). Binding of laminin to type IV collagen: a morphological study. J.
Cell Biol. 100, 1848-1853.
Clyman, R. I., Turner, D. C. and Kramer, R. H. (1990). An α1/β1-like
integrin receptor on rat aortic smooth muscle cells mediates adhesion to
laminin and collagen types I and IV. Arteriosclerosis 10, 402-409.
Clyman, R. I., Mauray, F. and Kramer, R. H. (1992). β1 and β3 integrins
have different roles in the adhesion and migration of vascular smooth muscle
cells on extracellular matrix. Exp. Cell Res. 200, 272-284.
Couchman, J. R. and Rees, D. A. (1979). The behaviour of fibroblasts
migrating from chick heart explants: changes in adhesion, locomotion and
growth, and in the distribution of actomyosin and fibronectin. J. Cell Sci. 39,
149-165.
Dean, J. W., Chandrasekaran, S. and Tanzer, M. L. (1990). Biological role
of the carbohydrate moieties of laminin. J. Biol. Chem. 265, 12553-12564.
de Reeder, E. G., Poelmann, R. E., van Munsteren, J. C., Patterson, D. F.
and Gittenberger-de-Groot, A. C. (1989). Ultrastructural and
immunohistochemical changes of the extracellular matrix during intimal
cushion formation in the ductus arteriosus of the dog. Atherosclerosis 79, 2940.
DiMilla, P. A., Stone, J. A., Quinn, J. A., Albelda, S. M. and Lauffenburger,
D. A. (1993). Maximal migration of human smooth muscle cells on
fibronectin and type IV collagen occurs at an intermediate attachment
strength. J. Cell Biol. 122, 729-737.
Duband, J.-L., Nuckolls, G. H., Ishihara, A., Hasegawa, T., Yamada, K.
M., Thiery, J. P. and Jacobson, K. (1988). Fibronectin receptor exhibits
high lateral mobility in embryonic locomoting cells but is immobile in focal
contacts and fibrillar streaks in stationary cells. J. Cell Biol. 107, 1385-1396.
Engvall, E., Earwicker, D., Haaparanta, T., Ruoslahti, E. and Sanes, J. R.
(1990). Distribution and isolation of four laminin variants: tissue restricted
distribution of heterotrimers assembled from five different subunits. Cell
Reg. 1, 731-740.
Gehlsen, K. R., Sriramarao, P., Furcht, L. T. and Skubitz, A. P. (1992). A
synthetic peptide derived from the carboxy terminus of the laminin A chain
represents a binding site for the α3β1 integrin. J. Cell Biol. 117, 449-459.
Giancotti, F. G. and Ruoslahti, E. (1990). Elevated levels of the α5β1
fibronectin receptor suppress the transformed phenotype of Chinese hamster
ovary cells. Cell 60, 849-859.
Glukhova, M., Koteliansky, V., Fondacci, C., Marotte, F. and Rappaport,
L. (1993). Laminin variants and integrin laminin receptors in developing and
adult human smooth muscle. Dev. Biol. 157, 437-447.
Goodman, S., Auxmailley, M. and von der Mark, M. (1991). Multiple cell
surface receptors for the short arms of laminin: α1β1 integrin and RGD
dependent proteins mediate cell attachment only to domains III in murine
tumor laminin. J. Cell Biol. 113, 931-941.
Graf, J., Iwamoto, Y., Sasaki, M., Martin, G. R., Kleinman, H. K., Robey,
F. A. and Yamada, Y. (1987). Identification of an amino acid sequence in
1018 R. I. Clyman and others
laminin mediating cell attachment, chemotaxis, and receptor binding. Cell
48, 989-996.
Haas, R. and Culp, L. A. (1982). Properties and fate of plasma fibronectin
bound to the tissue culture substratum. J. Cell. Physiol. 113, 289-297.
Hall, D. E., Neugebauer, K. M. and Reichardt, L. F. (1987). Embryonic
neural retinal cell response to extracellular matrix proteins: developmental
changes and effects of the CSAT antibody. J. Cell Biol. 104, 623-634.
Kleinman, H. K., Klebe, R. J. and Martin, G. R. (1981). Role of collagen
matrices in the adhesion and growth of cells. J. Cell Biol. 88, 473-485.
Kleinman, H. K., McGarvey, M. L., Liotta, L. A., Robey, P. G.,
Tryggvason, K. and Martin, G. R. (1982). Isolation and characterization of
type IV procollagen, laminin, and heparan sulfate proteoglycan from the
EHS sarcoma. Biochemistry 21, 6188-6193.
Kleinman, H. K., Weeks, B. S., Schnaper, H. W., Kibbey, M. C.,
Yamamura, K. and Grant, D. S. (1993). The laminins: a family of
basement membrane glycoproteins important in cell differentiation and
tumor metastases. Vitamins and Hormones 47, 161-186.
Kramer, R. H., McDonald, K. A., Crowley, E., Ramos, D. and Damsky, C.
H. (1989). Integrin related complexes mediate the adhesion of melanoma
cells to basement membrane. Cancer Res. 49, 393-402.
Kramer, R. H., Cheng, Y. F. and Clyman, R. (1990). Human microvascular
endothelial cells use β1 and β3 integrin receptor complexes to attach to
laminin. J. Cell Biol. 111, 1233-1243.
Kramer, R. H., Vu, M. P., Cheng, Y.-F., Ramos, D. M., Timpl, R. and
Waleh, N. (1991). Laminin-binding integrin integrin α7β1: functional
characterization and expression in normal and malignant melanocytes. Cell
Reg. 2, 805-817.
Landegran, U. (1984). Measurement of cell numbers by means of the
endogenous enzyme hexosaminidase. Applications to detection of
lymphokines and cell surface antigens. J. Immunol. Meth. 67, 379-388.
Lindon, J. N., McManama, G., Kushner, L., Merrill, E. W. and Salzman, E.
W. (1986). Does the conformation of adsorbed fibrinogen dictate platelet
interactions with artificial surfaces? Blood 68, 355-362.
Maniatis, T., Fritsch, E. F. and Sambrook, J. (1982). Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Mann, K., Deutzmann, R. and Timpl, R. (1988). Characterization of
proteolytic fragments of the laminin-nidogen complex and their activity in
ligand-binding assays. Eur. J. Biochem. 178, 71-80.
Massia, S. P. and Hubbell, J. A. (1991). An RGD spacing of 440 nm is
sufficient for integrin αvβ3-mediated fibroblast spreading and 140 nm for
focal contact and stress fiber formation. J. Cell Biol. 114, 1089-1100.
Mecham, R. P. (1991). Laminin receptors. Annu. Rev. Cell Biol. 7, 71-91.
Murphy-Ullrich, J. and Hook, M. (1989). Thrombospondin modulates focal
adhesions in endothelial cells. J. Cell Biol. 109, 1309-1319.
Neugebauer, K. M. and Reichardt, L. F. (1991). Cell-surface regulation of
B1-integrin activity on developing retinal neurons. Nature 350, 68-71.
Nurcombe, V., Aumailley, M., Timpl, R. and Edgar, D. (1989). The highaffinity binding of laminin to cells. Assignation of a major cell-binding site to
the long arm of laminin and of a latent cell-binding site to its short arms. Eur.
J. Biochem. 180, 9-14.
Ott, U., Odermatt, E., Engel, J., Furthmayr, H. and Timpl, R. (1982).
Protease resistance and conformation of laminin. Eur. J. Biochem. 123, 6372.
Ruoslahti, E., Hayman, E. G., Pierschbacher, M. D. and Engvall, E. (1982).
Fibronectin: purification, immunochemical properties and biological
activities. Meth. Enzymol. 82, 803-831.
Ruoslahti, E., Suzuki, S., Hayman, E. G., Ill, C. and Pierschbacher, M. D.
(1987). Purification and characterization of vitronectin. Meth. Enzymol. 144,
430-437.
Sanger, F., Nicklen, S. and Coulson, A. R. (1977). DNA sequencing with
chain-terminating inhibitors. Proc. Nat. Acad. Sci. USA 74, 5463-5467.
Sasaki, M., Kleinman, H. K., Huber, H., Deutzmann, R. and Yamada, Y.
(1988). Laminin, a multidomain protein: the A chain has a unique globular
domain and homology with the basement membrane proteoglycan and the
laminin B chains. J. Biol. Chem. 263, 16536-16544.
Sato, S., Burdett, I. and Hughes, R. C. (1993). Secretion of the baby hamster
kidney 30 kDa galactose-binding lectin from polarized and nonpolarized
cells: a pathway independent of the endoplasmic reticulum-Golgi complex.
Exp. Cell Res. 207, 8-18.
Schittny, J. C. and Yurchenco, P. D. (1990). Terminal short arm domains of
basement membrane laminin are critical for its self-assembly. J. Cell Biol.
110, 825-832.
Smalheiser, N. R. (1991). Role of laminin in stimulating rapid-onset neurites in
NG108-15 cells: relative contribution of attachment and motility responses.
Brain Res. Dev. Brain Res. 62, 81-89.
Song, W. K., Wang, W., Foster, R. F., Bielser, D. A. and Kaufman, S. J.
(1992). H36-alpha 7 is a novel integrin alpha chain that is developmentally
regulated during skeletal myogenesis. J. Cell Biol. 117, 643-657.
Sonnenberg, A., Linders, C. J. T., Modderman, P. W., Damsky, C. H.,
Aumailley, M. and Timpl, R. (1990). Integrin recognition of different cellbinding fragments of laminin (P1, E3, E8) and evidence that α6β1 but not
α6β4 functions as a major receptor for fragment E8. J. Cell Biol. 110, 21452155.
Sonnenberg, A., Gehlsen, K. R., Aumailley, M. and Timpl, R. (1991).
Isolation of α6β1 integrins from platelets and adherent cells by affinity
chromatography on mouse laminin fragment E8 and human laminin pepsin
fragment. Exp. Cell Res. 197, 234-244.
Spring, J., Beck, K. and Chiquet-Ehrismann, R. (1989). Two contrary
functions of tenascin: dissection of the active sites by recombinant tenascin
fragments. Cell 59, 325-334.
Tamura, R. N., Rozzo, C., Starr, L., Chambers, J., Reichardt, L. F.,
Cooper, H. M. and Quaranta, V. (1990). Epithelial integrin alpha 6 beta 4:
complete primary structure of alpha 6 and variant forms of beta 4. J. Cell
Biol. 111, 1593-1604.
Tomaselli, K. J., Hall, D. E., Flier, L. A., Gehlsen, K. R., Turner, D. C.,
Carbonetto, S. and Reichardt, L. F. (1990). A neuronal cell line (PC12)
expresses two β1-class integrins – α1β1 and α3β1 – that recognize different
neurite outgrowth-promoting domains in laminin. Neuron 5, 651-662.
Laminin, integrins and muscle migration 1019
Trinkaus, J. P. (1985). Further thoughts on directional cell movement during
morphogenesis. J. Neurosci. Res. 13, 1-19.
Tucker, R. P., Edwards, B. F. and Erickson, C. A. (1985). Tension in the
culture dish: microfilament organization and migratory behavior of quail
neural crest cells. Cell Motil. 5, 225-237.
Yurchenco, P. D., Sung, U., Ward, M. D., Yamada, Y. and O’Rear, J. J.
(1993). Recombinant laminin G domain mediates myoblast adhesion and
heparin binding. J. Biol. Chem. 268, 8356-8365.
(Received 1 October 1993 - Accepted 21 December 1993)