T H E J O U R N A L O F C E L L B I O L O G Y
The Journal of Cell Biology, Vol. 168, No. 3, January 31, 2005 501–511
The Rockefeller University Press $8.00
function by urokinase receptor binding
1 integrin conformation and
Martin J. Humphries,
Matthias C. Kugler,
and Harold A. Chapman
Kevin K. Kim,
Department of Medicine and Pulmonary and Critical Care Division, University of California, San Francisco, San Francisco, CA 94143
Department of Cell Biology, The Scripps Research Institute, La Jolla, CA 92037
Structural Biology Program, Massachusetts General Hospital, Charlestown, MA 02129
Wellcome Trust Centre for Cell-Matrix Research, School of Biological Sciences, University of Manchester, Manchester M13 9PT, England, UK
rokinase-type plasminogen activator receptors
(uPARs), up-regulated during tumor progression,
1 integrins, localizing urokinase to
sites of cell attachment. Binding of uPAR to the
1 empowers vitronectin adhesion by this integrin.
How uPAR modifies other
1 integrins remains unknown.
Using recombinant proteins, we found uPAR directly
1 and rather than blocking, renders fibronectin
(Fn) binding by
1 Arg-Gly-Asp (RGD) resistant. This
resulted from RGD-independent binding of
to Fn type III repeats 12–15 in addition to type III repeats
9–11 bound by
by small interfering RNA in tumor cells promoted weaker,
RGD-sensitive Fn adhesion and altered overall
1 peptide (res 224NLDSPEGGF232) that
models near the known
-chain uPAR-binding region, or
1-chain Ser227Ala point mutation, abrogated effects
of uPAR on
1. Direct binding and regulation of
1 by uPAR implies a modified “bent” integrin confor-
mation can function in an alternative activation state with
this and possibly other cis-acting membrane ligands.
1. Suppression of endogenous uPAR
The urokinase-type plasminogen activator receptor (uPAR)
plays important roles in cell adhesion and migration and che-
motaxis, as well as tumor metastasis, by virtue of its interac-
tions with the urokinase-type plasminogen activator (uPA), vi-
tronectin (Vn), and integrins (Degryse et al., 2001; Ahmed et
al., 2003; Sturge et al., 2003). Because of the capacity of uPAR
to bind both integrins and uPA, this receptor contributes to
matrix remodeling and cell migration in part by focusing plasmin
activation to sites of cell attachment (Chapman and Wei,
2001). It has also become clear that uPAR–integrin complexes
can transduce intracellular signals. Several groups have reported
that the binding of uPA to uPAR stimulates intracellular signaling
(Aguirre-Ghiso et al., 2003; Tarui et al., 2003; Sitrin et al.,
2004), and much of this signaling is consistent with an integrin-
mediated pathway. In addition, uPAR expression by itself,
independently of uPA, has been reported to mediate extracellular
signal–regulated kinase (ERK) activation. Recently, Czekay et
al. (2003) showed that plasminogen activator inhibitor-1 (PAI-1)
can detach cells by binding to the uPA present in uPA–uPAR–
integrin complexes on the cell surface. Although the mechanism
remains uncertain, these investigators found that PAI-1 not
only detached cells from Vn, but also from fibronectin (Fn) and
type-1 collagen. This influence of PAI-1 on integrin function,
via uPAR, may be related to the consistent findings that high
PAI-1 is an independent risk factor for tumor metastasis
(Andreasen et al., 1997). Together, this body of evidence indi-
cates uPAR–integrin interactions likely have physiological and
Previous reports have indicated that uPAR can physically
interact with multiple integrins including
(Xue et al., 1997; Pluskota et al., 2003). We previously identi-
fied uPAR as an integrin ligand because a binding site for
uPAR on the integrin
2 mapped to the ligand-binding
region of its
-propeller (Simon et al., 2000). Among
grins, uPAR directly associates with
-propeller (W4 BC loop), but outside the laminin-5
(Ln-5) binding region. uPAR is also able to associate with
1 (Aguirre-Ghiso et al., 1999; Wei et al., 2001) and to
1 via a surface loop
Correspondence to Harold A. Chapman: email@example.com; or Ying Wei:
Abbreviations used in this paper: ERK, extracellular signal–regulated kinase;
Fn, fibronectin; LIBS, ligand-induced binding site; Ln-5, laminin-5; PAI-1, plas-
minogen activator inhibitor-1; RGD, Arg-Gly-Asp; siRNA, small interfering RNA;
suPAR, soluble uPAR; Tet, tetracycline; uPA, urokinase-type plasminogen activa-
tor; uPAR, uPA receptor; Vn, vitronectin.
The online version of this article includes supplemental material.
JCB • VOLUME 168 • NUMBER 3 • 2005 502
2003), and Fn matrix assembly (Monaghan et al., 2004).
1 is among many members of the integrin fam-
ily that recognize an Arg-Gly-Asp (RGD) motif within their
ligands (Takagi et al., 2003). Peptides containing this motif can
efficiently block these integrin–ligand interactions (Arnaout et al.,
1 integrin and Fn form a prototypic integrin/ligand pair
(Takagi et al., 2003), functionally important because it mediates
Fn adhesion and Fn matrix assembly, which is vital to many cell
functions in vivo (Cukierman et al., 2001). This integrin is also
shown to play a key role in promoting tumor angiogenesis and tu-
mor metastasis (Jin and Varner, 2004). In addition to the RGD se-
quence present in Fn type III module 10, a set of residues present
in the Fn type III module 9 (synergy site) contribute to high affin-
ity recognition by
1 (Redick et al., 2000). The COOH-termi-
nal heparin-binding site (HepII) of Fn also plays an important role
in regulating cell adhesion, migration, Fn fibrillogenesis, signal
transduction, and organization of focal adhesions and cytoskele-
ton (Huang et al., 2001; Kim et al., 2001). The interaction of cells
to the HepII domain is currently thought to operate through pro-
teoglycans such as syndecan 4 (Kim et al., 2001) and integrin
1 (Mould and Humphries, 1991).
To date, little is known of the molecular mechanism by
which uPAR regulates
1-mediated function. In this report,
the effect of uPAR on
1-mediated adhesion, migration, and
1 signaling (Aguirre-Ghiso et al., 2003; Tarui et al.,
1-mediated cell migration to Fn (Yebra et al.,
Fn matrix assembly was investigated. Surprisingly, using recom-
binant proteins, we found that direct binding of uPAR to
does not change overall integrin binding to Fn, but changes inte-
grin conformation, subsequently forming an additional binding
site on Fn, which is RGD independent. In the course of this work
1 peptide sequence was discovered that blocks all uPAR/
function. Mapping of this peptide near the known
1 interaction confirms this region of
important regulatory site and suggests a molecular basis for PAI-1–
mediated cell detachment. Positioning of the uPAR-binding site
near the Fn-binding site of
1 not only promotes
actions with Fn, but allows PAI-1 to reverse Fn binding, empow-
ering a mechanism of cell migration on Fn in either a protease-
rich or protease inhibitor–rich milieu.
-chain site of
1 integrins as an
PAI-1 mediates cell detachment by
reversing ligand binding to
PAI-1 can detach cells from Vn and Fn by binding to uPA
present in uPA–uPAR–integrin complexes on the cell surface
(Czekay et al., 2003). However, the mechanism of detachment
is uncertain. To explore a possible mechanism, we took advan-
tage of a recently described point mutant of the Ln-5 receptor,
1, incapable of uPAR interaction but fully capable of Ln-5
adhesion and signaling (Zhang et al., 2003). This mutation,
for uPA–PAI-1 mediated matrix detachment. (A) Cell
detachment from Ln-5 and Fn. 96-well plates were coated
with Fn or Ln-5 and blocked with BSA. Wt ?3, wt ?3/U,
mut ?3, or mut ?3/U cells were allowed to attach for 1.5 h,
were acid washed, and then were incubated with uPA,
PAI-1, or PAI-1 followed by uPA. The wells were washed
as described in Materials and methods, and the number
of remaining attached cells was determined. Each column
is expressed as a percentage of the control (acid washed
only; 100%). n ? 3. (B) Biotin-III 9-11 Fn fragment binding
to HT1080 cells. HT1080 cells were acid washed and
incubated with or without uPA, or with uPA followed by
PAI-1. The cells were then incubated with biotinylated III
9-11. After washing, the cells were lysed and the bound
biotin III 9-11 was detected by avidin-HRP. The densitometry
of the bands was quantified and graphed. All the steps
were done at 4?C. Each column is expressed as a per-
centage of the control (100%). Data shown are represen-
tative from three independent experiments with similar
results. (C) Effect of uPA–PAI-1 on purified ?5?1 binding
to Fn. Recombinant ?5?1-Fc integrin was analyzed by
Western blotting with peroxidase-conjugated anti–human
Fc antibody. The ?5-Fc and ?1-Fc proteins are shown as
an inset. Nunc high binding plates precoated with Fn
were incubated with ?5?1-Fc with or without suPAR in
the presence of 1 mM Mn2?. Increasing amounts of uPA–
PAI-1 mixture were then added. Bound ?5?1-Fc was
detected by protein A–HRP and quantified by OD at 490
nm. Data are expressed as specific binding (total binding
subtract the binding to BSA-coated wells). n ? 3. Each
experiment was done with triplicate determinations.
uPAR and ligand-engaged integrin are required
UROKINASE RECEPTOR REGULATES
1 INTEGRIN • WEI ET AL.503
His245Ala, allowed us to test the hypothesis that PAI-1 acts
directly on matrix-engaged
complexes bound to
1 and not through indirect effects
mediated by another integrin. Mouse kidney epithelial cells
?1 (wt ?3) attach to Ln-5– or Fn-coated sur-
faces, and this attachment is not reversed by the presence of
uPA or uPA–PAI-1 complexes (Fig. 1 A). Co-expression of
human uPAR in these cells (wt ?3/U) resulted in marked sensi-
tivity of matrix attachment to the presence of uPA–PAI-1, as
predicted by prior studies of Czekay et al. (2003). However,
epithelial cells bearing the H245A ?3 mutation (mut ?3), al-
though attaching normally to Ln-5, failed to detach from Ln-5
in the presence of uPA–PAI-1 complexes whether or not there
was coexpression of uPAR. In contrast, mut ?3 cells express-
ing uPAR (mut ?3/U) adhered to Fn and detached ?80% in the
presence of uPA–PAI-1. Similar results were obtained in at
least three separate experiments. These data indicate that ma-
trix detachment initiated by uPA–PAI-1 requires uPAR inter-
action with the specific integrin that is matrix engaged. If so,
this finding raises the possibility that uPA–PAI-1 directly af-
fects matrix ligand binding to uPAR–integrin complexes, re-
sulting in cellular detachment. This possibility was tested in
HT1080 fibrosarcoma cells, cells known to express uPAR and
bind Fn through ?5?1 (Xue et al., 1997). The binding of biotin-
ylated, soluble III 9-11 (Fn type III repeats 9–11, containing
1 by binding to uPA–uPAR
the RGD integrin-binding site) to HT1080 cells was almost
completely blocked when cells were pretreated with active
uPA and PAI-1 during the 4?C binding assay (Fig. 1 B). These
results indicate that uPA–PAI-1 complexes interfere with ?1
integrin ligand binding in a uPAR-dependent manner.
To explore this finding further, we performed Fn-binding
assays using recombinant purified soluble ?5?1 and uPAR. Re-
combinant ?5?1-Fc fusion protein was expressed in HEK 293
cells and purified using protein A–agarose beads (Coe et al.,
2001). The ?5?1-Fc fusion was found to bind to Fn-coated sur-
faces (Fig. 1 C) and the binding is Mn2?-dependent (unpublished
data). Integrin ?5?1 binding to Fn was not affected by the pres-
ence of either uPA–PAI-1 complexes or soluble uPAR (suPAR).
However, in the presence of suPAR, increasing amounts of uPA–
PAI-1 complexes progressively blocked ?5?1 binding to Fn
(Fig. 1 C). Addition of uPA or PAI-1 separately had no effect on
Fn binding. In additional experiments, biotinylated suPAR could
also be shown to bind to immobilized ?5?1-Fc, as expected
(unpublished data). Together, these data indicate that uPAR
binds ?5?1, and such binding modifies ?5?1 integrin binding to
Fn and enables PAI-1 to release Fn from ?5?1 through uPAR-
bound uPA. All of these events can happen on the extracellular
domains of the integrin independent of cell signaling.
These observations imply that uPAR can interact with at
least two and possibly multiple ?1 integrins, suggesting the
sequence alignment. (A) An energy-minimized model
of integrin ?5?1 structure was developed based on
the atomic coordinates of ?v?3 crystal structure
(Xiong et al., 2001). The putative W4 BC loop (NLTY)
of the predicted ?-propeller structure of the ?5-chain
(blue) is marked in red. The two ?1 peptides ?1P1
and ?1P2 from ?1-chain (gray) are indicated in yellow
and purple, respectively. Both side and top view of
the model are shown. (B) The sequence alignment of
the two ?1 peptide regions in ?1-chain shows homol-
ogies among different ?-chains. ?1P1 peptide is un-
derlined yellow and ?1P2 purple. In the ?3?1 model,
these are the two ?1 sequences closest to the reported
?3 interaction site with uPAR (W4 BC loop of the
?-propeller structure). Asterisk highlights Ser227 within
the ?1P1 sequence because only the ?1-chain contains
this Ser and a Ser→Ala point mutation in the ?1P1
peptide was made and used in some experiments.
Model of ?5?1 structure and homologous
JCB • VOLUME 168 • NUMBER 3 • 2005504
involvement of the common ?1-chain itself in uPAR binding.
Moreover, the direct effect of uPA–PAI-1 on Fn binding to
uPAR–?5?1 suggests the uPAR-binding site may be positioned
close to the Fn-binding site. As demonstrated in Fig. 2 A, an en-
ergy-minimized model of integrin ?5?1 structure was generated
based on the atomic coordinates of the ?v?3 crystal structure
(Xiong et al., 2001). Recently, we have found that the blade 4
BC loop of the proposed ?-propeller structure of integrin ?3 is
important for uPAR association (Wei et al., 2001). The corre-
sponding BC loop in the ?5?1 model is highlighted in red. In-
spection of the model reveals two loops on the ?1-chain
(224NLDSPEGGF232 in yellow, 262FHFAGDGKL270 in pur-
ple) that are very close to the blade 4 BC loop of the ?-propeller
(in red). We hypothesized that the two ?1-chain loops may also
be involved in integrin–uPAR association. The alignment of
these ?1-chain sequences with that of other integrin ?-chains is
shown in Fig. 2 B. Interestingly, the NH2-terminal Asn 224 in
?1P1 has been implicated in bonding to the Asp (D) of RGD in
the ?v?3 crystal, placing these loops very close to the putative
RGD-binding pocket of ?5?1. This could potentially explain the
direct effect of uPA–PAI-1 on Fn binding to uPAR–?5?1.
uPAR associates with ?5?1 and changes
Fn adhesion properties
To explore whether the ?1 loops discussed above are in-
volved in uPAR and ?5?1 interactions, a series of Fn adhe-
sion experiments were performed. Kidney epithelial cells
used in Fig. 1 and their uPAR cotransfectants were again
used. The adhesion of these cells to Fn is mediated to a large
extent by the ?5?1 integrin because the adhesion can be
completely blocked by an ?5?1-blocking antibody 5H10-27
(unpublished data). The observation that uPAR associates
with ?3?1 integrin (Wei et al., 2001) and the complex medi-
ates RGD-independent cell adhesion to Vn (unpublished
data) prompted us to test whether adhesion of uPAR-expressing
cells to Fn can be inhibited by RGD peptides. Surprisingly,
overexpression of uPAR strikingly increased cell adhesion to
Fn in the presence of RGD-containing peptides (Fig. 3 A).
Indeed up to 1 mM RGD had no discernible affect on ad-
hesion of uPAR-expressing cells to Fn. By contrast, the ad-
hesion of non-uPAR–expressing cells to Fn was totally
abolished by 500 ?M RGD peptides. The RGD-resistant ad-
hesion of uPAR-expressing cells was observed at all Fn con-
centrations supporting adhesion (0.5–5 ?g/ml; Fig. S3, available
at http://www.jcb.org/cgi/content/full/jcb.200404112/DC1). The
?1 peptides had no effect on Fn adhesion of non-uPAR–
expressing cells at any concentration tested, but 400 ?M of
either of the two ?1 peptides completely inhibited cell adhe-
sion to Fn in uPAR-expressing cells. The scrambled pep-
tides had no effect (Fig. 3 A). The dose inhibition effect of
one of the two ?1 peptides (?1P1) on adhesion to Fn is
shown in Fig. 3 B.
binding mechanism. (A) Adhesion to Fn. mut ?3
(?uPAR), or mut ?3/U (?uPAR) cells pretreated
with different peptides were plated on Fn and cell
adhesion was measured as described in Materials
and methods. RGD and RAD: 500 ?M; ?1 pep-
tides: 400 ?M. A representative of three indepen-
dent experiments with triplicate wells is shown.
(B) Dose effect of ?1P1 peptide on Fn adhesion.
Wt ?3/U cells were seeded on Fn-coated wells
with ?1P1 peptide (50–400 ?M) and the adhesion
was assessed as above. Data are expressed as
percentage of control (no peptide added). n ? 3.
(C) Effect of ?1P1 peptide on uPAR–?5?1 com-
plex formation. HT1080 cells were lysed in 1%
Triton X-100 lysis buffer and the lysates were incu-
bated with peptide ?1P1, its scrambled control
(sc?1p1), or left untreated. Lysates were immuno-
precipitated (P1D6) and the lysates and immuno-
precipitates separated by SDS-PAGE and blotted
for uPAR (R2) and integrin ?5. Data shown are
representative of three independent experiments.
White lines indicate that intervening lanes have
been spliced out. (D) Biotin-suPAR binding to
?5?1–Fn complex. 20 nM biotinylated suPAR
was added to Fn-coated wells and incubated with
or without ?5?1-Fc or ?5?1SA-Fc (20 ?g/ml).
The bound biotin-suPAR was detected by avidin-
HRP and Fn-bound integrin was detected by Protein
A–HRP. “?” represents the background binding
on Fn or BSA. Both are quantified by measuring
OD at 490 nm. n ? 3.
uPAR binding to ?5?1 changes the Fn-
UROKINASE RECEPTOR REGULATES ?5?1 INTEGRIN • WEI ET AL.505
Because the ?1 peptide sequence contains a natural
Ser227Ala polymorphism, inviting a mutational analysis, we
explored the functional properties of this peptide in more de-
tail. We tested the effect of the ?1P1 peptide on the biochemi-
cal association of ?5?1 and uPAR. HT1080 cell lysates were
immunoprecipitated with ?5 antibodies in the presence of
?1P1 or its scrambled control, and the presence of uPAR was
determined by immunoblotting. As indicated in Fig. 3 C,
uPAR could be reliably coprecipitated with ?5?1 and this as-
sociation was completely blocked by the functionally active
peptide, suggesting residues 224–232 in the ?1-chain are in-
volved in uPAR–?5?1 physical association. To test this point
further, a ?5?1-Fc fusion protein containing the entire extra-
cellular domains of ?5?1 was expressed in which Ser227 was
mutated to Ala (?5?1SA-Fc). As shown in Fig. 3 D, the
?5?1SA-Fc fusion was found to bind Fn identically to wt. We
verified that recombinant biotinylated suPAR bound to ?5?1-
Fc/Fn, but not appreciably to Fn alone. However, the Ser227
to Ala mutant integrin (?5?1SA-Fc) totally lost interaction
with uPAR, indicating that Ser227 on the integrin ?1-chain,
and by inference the 224–232 ?1 peptide loop, is critical for
uPAR binding to ?5?1 induces an
additional binding site for Fn
Next, we examined whether the apparent switch in the mech-
anism of ?5?1 adhesion to Fn initiated by the presence of
uPAR (Fig. 3 A) was evident with purified proteins. Fn was
immobilized and ?5?1-Fc binding to Fn was measured by
protein A–HRP as before. Fn–?5?1 binding in the presence
or absence of suPAR is similar (Fig. 4 A), and both can be
blocked by an ?5-blocking antibody (unpublished data). The
binding of ?5?1 to Fn in the absence of suPAR was abol-
ished by RGD peptides, as expected, whereas binding in the
presence of suPAR was vice versa. Interestingly, the binding
of mutant integrin (?5?1SA) to Fn remained RGD sensitive
in the presence of suPAR (Fig. 4 A). These findings com-
pletely recapitulate the pattern seen with live cells and indi-
cate that the presence of uPAR markedly changes the matrix-
binding properties of ?5?1. This raises two possibilities:
uPAR binding switches the integrin-binding site from the
central RGD binding domain (III 10) to a different site on
Fn. Or, uPAR binding to ?5?1 creates an additional binding
site for Fn. To determine which possibility is more likely, we
made biotinylated RGD peptides and performed binding as-
says with purified proteins. Binding of biotin-RGD (closed
bars) and ?5?1-Fc (open bars) to Fn were measured sepa-
rately and graphed together (Fig. 4 B). In the absence of
uPAR, biotin-RGD (0.5 mM) competed with the RGD-bind-
ing site on immobilized Fn, blocking ?5?1 binding to the
plate. However, in the presence of suPAR, biotin-RGD ro-
bustly bound to uPAR–?5?1–Fn unless 10-fold excess unla-
beled RGD was added (Fig. 4 B), strongly suggesting the ex-
istence of an additional RGD-independent binding site for
Fn. Nevertheless, the RGD-binding site on ?5?1 must still
be intact in the presence of suPAR, otherwise biotin-RGD
would not be able to bind ?5?1.
To determine where the additional binding site interacts
on Fn, we examined the effect of uPAR on cell adhesion to Fn
fragments in the presence or absence of RGD peptides. Our ini-
tial experiments show that adhesion to NH2-terminal 70-kD Fn
was RGD-resistant, but unaffected by the presence of uPAR
and not blocked by ?5 integrin blocking antibodies (unpub-
lished data). Thus, we focused on the Fn type III repeats. Cells
with or without uPAR were allowed to attach to immobilized
of RGD peptides on wt or mutant ?5?1 binding to Fn. Purified ?5?1-Fc or
mutant ?5?1SA-Fc (20 ?g/ml) was added to Fn-coated wells in the pres-
ence or absence of suPAR and incubated with RGD peptides. The bound
?5?1 was detected by protein A–HRP. n ? 3. (B) Binding of biotin-RGD to
uPAR–?5?1–Fn complexes. 20 ?g/ml purified ?5?1 was allowed to bind
immobilized Fn in the presence or absence of 20 nM suPAR. 0.5 mM biotin-
ylated RGD peptides or buffer were then added and incubated without or
with excess unlabeled peptides (5 mM). The bound biotin was detected by
avidin-HRP and bound ?5?1 in parallel plates was detected by protein
A–HRP. The data are expressed as absorbance at 490 nm. n ? 3. (C) Cell
adhesion to Fn fragments. Mut ?3 (?uPAR) and mut ?3/U (?uPAR) cells
were plated in Fn or Fn fragment (III 9-11 or III 12-15)–coated wells and in-
cubated without or with different peptides: RGD or control RAD peptides (500
?M); ?1P1 or control sc?1P1 peptide (400 ?M). The attached cells were
quantified and the data from a representative experiment are shown. n ? 3.
uPAR–?5?1 binds to heparin-binding domain II of Fn. (A) Effect
JCB • VOLUME 168 • NUMBER 3 • 2005 506
Fn fragments containing type III repeats 9–11 (III 9-11) or type
III repeats 12–15 (III 12-15). Cells without or with uPAR ex-
pression adhered strongly to the RGD-containing III 9-11 and
the adhesion of both cells was blocked by RGD peptides, con-
firming that the uPAR-expressing cells maintained an RGD-
binding site on Fn for ?5?1. As expected, ?1P1 had no effect
on this adhesion. However, only uPAR-expressing cells at-
tached to III 12-15, and this adhesion was resistant to RGD
peptides and now sensitive to the ?1 peptide (?1P1), indicating
uPAR induces at least one ?5?1-binding site within the
COOH-terminal heparin-binding domain of Fn. This additional
site conveys RGD resistance to binding of cells to the whole Fn
molecule (Fig. 4 C). As expected, the Ser227Ala point mutant
peptide (?1P1SA), like scrambled ?1P1, had no effect.
uPAR alters ?5?1 conformation and
changes ?5?1 integrin–dependent
adhesion and detachment in tumor cells
The above data indicate that binding of uPAR to ?5?1 alters
integrin conformation and changes its matrix ligand binding
properties. To probe this idea further and determine whether
uPAR-mediated changes in ?5?1 function are observable in
nontransfected cells, several tumor cell lines expressing vari-
ous amounts of uPAR were evaluated. HT1080 (fibrosarcoma),
MDA-MB-231 (breast carcinoma), and Skov-3 (ovarian carci-
noma) cells were transfected with a small interfering RNA
(siRNA) previously shown to suppress uPAR mRNA (Vial et
al., 2003) or control and suppression of surface uPAR expression
verified 48 h later by FACS analysis (Fig. 5 A). Suppression of
surface uPAR had no effect on total ?1 integrin expression
(JB1A). However, suppression of uPAR had clear effects on
integrin conformation as judged by altered binding of the con-
formation-sensitive mAbs, HUTS-21, and 9EG7, in all of the
cell lines examined. Suppression of surface uPAR was accom-
panied by increased binding of both HUTS-21 and 9EG7 anti-
bodies (Fig. 5 A), confirming that endogenous uPAR expres-
sion modifies integrin ?1-chain conformation.
Are the changes in ?5?1 conformation in tumor cells in-
duced by knockdown of uPAR accompanied by changes in mech-
anism of adhesion? Previous studies show that these cell lines
duces LIBS epitope and changes ?5?1-mediated
Fn binding in tumor cells. (A) FACS analysis of
HT1080, MDA-MB-231, and Skov-3 cells with
uPAR (siRNA uPAR) or control siRNA (control)
transfection. Cells were harvested 48 h after trans-
fection and incubated with antibodies against
uPAR (uPAR), total ?1 (JB1A), or conformation-
sensitive ?1 integrin antibodies (HUTS-21, 9EG7),
followed by FITC-conjugated secondary antibod-
ies. (B) HT1080 adhesion to Fn. siRNA uPAR or
control cells were seeded to Fn-coated wells and
incubated with different peptides. All the above
experiments were performed at least three times
with similar results.
Suppression of uPAR expression in-
UROKINASE RECEPTOR REGULATES ?5?1 INTEGRIN • WEI ET AL.507
express substantial levels of uPAR and ?5?1 integrin (Xue et al.,
1997; van der Pluijm et al., 2001). Indeed, we confirmed that Fn
adhesion was mainly mediated by ?5?1 as the ?5-blocking anti-
body (P1D6) totally blocked the Fn adhesion (unpublished data).
To investigate whether increased RGD resistance of Fn adhesion
by uPAR expression also exists in nontransfected cells, we tested
the relation between uPAR level and Fn adhesion in HT1080,
MDA-MB-231, and Skov-3 cells. Because the data from these
cell lines are all very similar, we only show results with HT1080
cells (Fig. 5 B). Expression of high levels of endogenous uPAR
leads to the expected phenotype when cells are plated on Fn:
RGD resistance and ?1 peptide susceptibility. siRNA suppression
of uPAR in HT1080 cells, like prior studies with transfected epi-
thelial cells, switched the Fn phenotype to RGD sensitive and ?1
peptide resistant (Fig. 5 B). Furthermore, the noninvasive breast
cancer cells MCF-7 and T47D, expressing little uPAR, showed
only RGD-sensitive Fn adhesion (Fig. S2, available at http://
www.jcb.org/cgi/content/full/jcb.200404112/DC1). These data
support the conclusion that the mechanism of Fn adhesion among
tumor cells depends upon uPAR expression level.
uPAR expression promotes ?5?1-
mediated cell adhesion, migration, and
Fn matrix assembly
When cells are plated on an Fn-coated surface (5 ?g/ml) for 1 h,
the difference in Fn adhesion between cells with and without
surface uPAR is marginal (Fig. 3 A, Fig. 4 C, Fig. 5 B). How-
ever, when cells are seeded onto lower amounts of Fn (0.2–5
?g/ml) for shorter periods of time (20 min), adhesion of uPAR-
expressing HT1080 cells to Fn was obviously more robust (Fig.
6 A). We repeated similar experiments using MDA-MB-231
(Fig. S1, available at http://www.jcb.org/cgi/content/full/
jcb.200404112/DC1) and Skov-3 cells with or without uPAR
suppression and found similar results. These findings indicate
that uPAR expression not only changes the conformation of
?5?1 and how it engages Fn, but together these changes might
promote cell adhesion and migration.
To explore activation of ?5?1 by uPAR further, kidney
epithelial cells expressing uPAR under a tetracycline (Tet)-
inducible promoter were established. In the absence of Tet
(?Tet) the cells had little uPAR surface expression, whereas
Tet-induced cells (?Tet) showed robust surface uPAR (Fig.
6 B). Initial experiments verified that induction of uPAR pro-
duced the same switch in Fn adhesion phenotype observed
for stable clones and tumor cells examined above. Although
the mechanism of ?5?1-mediated Fn attachment changed in
the presence of uPAR, immunofluoresence of fixed cells
plated on Fn did not reveal discernible changes in the distri-
bution of ?5?1 by uPAR induction alone (unpublished data).
Nonetheless, ?Tet cells produced more cell-associated and
deoxycholate-insoluble Fn (Fig. 6 B), indicating that this
pool of Fn is much more organized into matrix fibrils in the
presence of uPAR.
To test whether uPAR association with ?5?1 affects
cell migration, we performed wound assays on tumor cells
(HT1080 and MDA-MB-231) and the Tet-responsive cells
with or without uPAR induction using RGD and ?1 peptides
(?1P1). The cells were seeded onto Fn-coated wells and al-
lowed to form a monolayer before wounding. Preliminary
experiments indicated that as little as 20 ?M of either the
RGD-containing or the ?1P1 peptide suppressed migration. As
shown in Fig. 7, both RGD and ?1P1 alone blocked wound
closure of HT1080 cells, and the combination of both peptides
had a statistically significant greater effect. Similar results were
obtained from MDA-MB-231 cells. More importantly, Tet-
inducible cells with uPAR expression (?Tet) migrated faster in
this assay and the migration was blocked by both RGD and ?1
peptides, whereas cells without uPAR (?Tet) had little migra-
tion (unpublished data). Together, these data confirm our find-
ings that uPAR, through its interaction with ?1 integrin(s),
promotes cell motility and that this function can be specifically
blocked by the ?1-chain peptides identified here.
adhesion to low concentration of Fn. siRNA uPAR or control cells were
seeded to Fn (0.2–5 ?g/ml)-coated wells and incubated for ?20 min.
The adhesion was quantified as described in Materials and methods.
(B) Top: FACS analysis of uPAR inducible clone Tet-uPAR cells. Both Tet-
treated (2 ?g/ml; ?Tet) and nontreated cells (?Tet) were stained with
FITC-conjugated uPAR antibody. n ? 8. Bottom: Fn in Tet-uPAR cells. Cells
without (?Tet) or with (?Tet) tetracycline induction were lysed with 3% Triton
X-100 and centrifuged. Triton-insoluble pellets were then extracted with
2% deoxycholate (DOC) and centrifuged. The insoluble and soluble
fractions on different membranes were analyzed by Western blotting using
anti–human Fn antibodies. The same samples were blotted for ?-actin to
normalize the loading. n ? 3. White line indicates that intervening lanes
have been spliced out.
uPAR overexpression enhances Fn fibril formation. (A) HT1080
JCB • VOLUME 168 • NUMBER 3 • 2005508
Here, we report evidence that uPAR directly associates with
the head domains of integrin ?5?1, modifies ?5?1 conforma-
tion, and creates an additional binding site for Fn, likely
within the second Fn heparin-binding domain. Complexes of
uPAR and ?5?1 are functionally relevant because uPAR bind-
ing promotes ?5?1-dependent Fn matrix assembly and migra-
tion. Importantly, our observations are not based strictly on
transfected cells because these features of uPAR–?5?1 inter-
action could be demonstrated in several tumor cell lines ex-
pressing endogenous uPAR. Collectively, these functional
changes imply ?5?1 activation by uPAR binding, as sug-
gested previously by Aguirre-Ghiso et al. (1999). Our studies
solidify ?5?1 as a binding partner of uPAR and further define
the uPAR-binding region on the integrin. The positioning of
the uPAR-binding site near the integrin RGD-binding site also
reveals a potential mechanism whereby uPAR–?5?1 complex
formation empowers PAI-1–dependent cell detachment from
Fn: while the integrin RGD-binding site remains intact in
uPAR–?5?1 (Fig. 4), concurrent binding of urokinase and
PAI-1 to uPAR now displaces intact Fn or the Fn cell–binding
domain (containing RGD) from the integrin (Fig. 1), presum-
ably by steric hindrance.
The consequences of uPAR–?5?1 complex formation
contrast with that of other pathways of integrin activation.
Available evidence indicates that integrin activation involves a
global change in integrin conformation, at least part of which is
a change in the orientation of the ? and ? head domains to bet-
ter accommodate ligand binding. Several lines of evidence also
support a model in which the very “bent” integrin conforma-
tion found in the ?V?3 crystal structure extends to point the
head domains away from the cell under activating conditions
(Takagi et al., 2002). However, the full range of conforma-
tional changes a ligand-bound integrin may assume is uncertain
(Mould and Humphries, 2004). It is especially difficult to envi-
sion a fully extended conformation of activated integrins acting
in cis to engage the much smaller GPI-anchored uPAR at the
integrin upper surface. Rather, our findings suggest uPAR–
?5?1 complexes exhibit an activation state involving a modified
bent integrin with distinct functional properties. Similarities and
differences between models of “extended” integrin activation
and a model consistent with results reported here are summa-
rized in Fig. 8. The model raises the more general possibility
that some version of an angled integrin configuration, rather
than being inactive, actually functions to promote integrin
binding to cis-acting membrane ligands, such as uPAR, which
coordinate integrin function with specific cellular needs.
The notion that uPAR activates and stabilizes ?5?1 has
been previously proposed by Ossowski and colleagues based
on their studies of human epidermoid carcinoma cell lines. En-
hanced adhesion to Fn of tumorigenic (T-HEp-3) over dormant
(D-HEp-3) epidermoid cells was directly related to uPAR lev-
els (Aguirre-Ghiso et al., 1999). The high levels of uPAR in hu-
man epidermoid carcinoma cells (Hep-3) resulted in increased
?5?1-dependent signaling to ERK as well as increased forma-
tion of Fn fibrils (Aguirre-Ghiso et al., 2001). Both ?5?1-depen-
dent ERK signaling and Fn matrix assembly were decreased in
the presence of a uPAR-binding peptide, P25, which blocks
uPAR–integrin association, indicating that in these cells liga-
tion of uPAR with P25 inhibited ?5?1 function. Similar to
these findings, we observed that induction of uPAR expression
with Tet in an epithelial cell line increased Fn fibril formation
(Fig. 6 B). Conversely, suppression of uPAR expression by
RNA interference in many tumor cell lines decreased adhesion
(Fig. 6 A, Fig. S1). Additional observations reported here help
provide a physical rationale for these and prior functional studies.
The current data indicate that uPAR directly binds and changes
the conformation and matrix-binding properties of ?5?1. The
capacity of short ?1-chain peptides to block uPAR–?5?1 func-
tions without affecting ?5?1-mediated Fn binding itself point
to important conformational differences between free and
uPAR-bound ?5?1. The finding that ?1P1 does not simply
convert the function of uPAR–?5?1 to that of free ?5?1 by
dissociating uPAR also suggests that the conformational
change incurred by complex formation with uPAR is distinct
and perhaps not readily reversible. This is consistent with the
hypothesis that bent and extended conformations of ?5?1 can
function as distinct activation states (Fig. 8).
This hypothesis is supported by our results, which reveal
that suppressing uPAR expression induces ligand-induced
binding site (LIBS) epitopes in HT1080, MDA-MB-231, and
Skov-3 cells (Fig. 5 A). A recent report also documented in-
Serum-starved HT1080 monolayers were wounded and incubated with
different peptides (RAD, RGD, sc?1P1, and ?1P1; 20 ?M) in DME/0.1%
BSA. The wounded areas were imaged at 0 and 24 h using a bright-field
imaging system (Spot camera). The migration of HT1080 cells was quanti-
fied using SimplePCI software. The percent wound closure of each peptide-
treated cell is shown on the right. n ? 3. Marked pair (*) shows significant
difference by t test (P ? 0.006).
RGD and ?1P1 peptides inhibit HT1080 cell wound healing.
UROKINASE RECEPTOR REGULATES ?5?1 INTEGRIN • WEI ET AL.509
creased ?1-chain LIBS epitopes on human skin fibroblasts ex-
posed to a peptide that disrupts uPAR–integrin interactions
(Wei et al., 1996; Monaghan et al., 2004). The LIBS antibodies
(HUTS-21, 9EG7) used here map to sites near the hinge region
of the integrin but far away from the uPAR interaction site. Al-
though these LIBS antibodies are thought to recognize the “ac-
tive” conformational state of the ?1 subunit that can be induced
by ligand binding (e.g., Fn, RGD peptides, by activating anti-
body TS2/16, or Mn2?), their binding is more sensitive to con-
formational changes in the hinge, knee, or leg domains than
changes near the ligand-binding pocket (Bazzoni et al., 1995;
Luque et al., 1996; Mould and Humphries, 2004). We postulate
that in uPAR-expressing cells the lower LIBS antibody binding
reflects integrin angulation resulting from uPAR–?5?1 com-
plex formation. This occurs in spite of “activation” of the inte-
grin as judged by enhanced adhesion and Fn matrix assembly,
further supporting the idea that activated integrins could exist
in grossly different conformational states depending on the na-
ture of the ligand (Fig. 8).
Previous studies have shown that initial cell attachment
and spreading on Fn is mediated by the interaction of the RGD-
containing Fn cell-binding domain (type III repeats 9–10) with
?5?1 (Mould et al., 2000; Redick et al., 2000; Takagi et al.,
2003), but that further progression of the cytoskeletal response
requires additional signals (Hocking et al., 1998; Tarui et al.,
2003). Additional binding sites for cells on Fn provide the nec-
essary signals. For example, interaction of cells with the Fn
NH2-terminal region can trigger integrin-mediated intracellular
signals that are distinct from those generated in response to li-
gation with the RGD sequence (Forsyth et al., 2002). However,
in our assays, adhesion of epithelial cells to this 70-kD frag-
ment was not influenced by uPAR expression and not inhibited
by ?5?1 blocking antibodies (unpublished data). Signals for
cytoskeletal reorganization may also be provided by the inter-
action of Fn fragments containing the heparin-binding domain
(Hep II) (type III repeats 12–14) with cell surface proteogly-
cans (Huang et al., 2001). In fibroblasts, this response requires
two cooperative signals provided by interactions of the RGD
sequence with ?5?1 integrin and the heparin-binding domain
with syndecan-4 (Kim et al., 2001). Our data show that both
cells with uPAR or without uPAR adhere to Fn III 9-11 in an
RGD-dependent manner, whereas only cells bearing uPAR ad-
here to Fn III 12-15. The latter cannot be blocked by RGD pep-
tides, but can be blocked by ?1 peptides that disrupt uPAR–?1
integrin interaction (Fig. 4 C). In most uPAR-expressing cells
there are likely to be pools of ?5?1 both free and bound to
uPAR, suggesting that the incorporation of the heparin binding
domain into the uPAR–?5?1 complex results in distinct sig-
nals that lead to enhanced integrin function, as our data show
(Fig. 6). We cannot be sure whether uPAR–?5?1 complexes
possess both Fn-binding sites or binding to both sites in Fn re-
quires free and uPAR-complexed integrin. Future studies may
distinguish between these possibilities.
We have previously reported that uPAR expression in
kidney embryonic 293 cells both promotes Vn adhesion
through association of uPAR with ?3?1 and impairs Fn adhe-
sion mediated by ?5?1 (Wei et al., 1996, 2001). Impairment of
Fn adhesion in 293 cells appears anomalous with respect to all
other cells expressing uPAR examined here and by others
(Aguirre-Ghiso et al., 1999). Consistent with this difference,
expression of uPAR in 293 cells did not decrease binding of
HUTS-21 and 9EG7 antibodies (unpublished data), implying
that for some reason uPAR interacts, but not in the same man-
ner, with ?5?1 in 293 cells as that seen in other transformed
cells. The molecular basis for the anomalous behavior of 293
cells remains to be defined.
The discovery of the capacity of ?5?1 to undergo a pheno-
typic switch (i.e., RGD vs. ?1P1 dependent; Fig. 8), in Fn at-
tachment may be relevant to attempts to regulate inflammation
or tumor progression through integrin inhibition in vivo. uPAR
is up-regulated in both inflammatory cells and many tumor cells
with a metastatic phenotype. Indeed, uPAR expression is an in-
dependent risk factor for tumor metastasis in several clinical
studies. RGD-based compounds or peptides have been shown to
inhibit integrin function in vivo, but our data imply that one lim-
itation in their use is the complete resistance of ?1 integrins
complexed with uPAR from RGD-dependent ligand binding. Vn
adhesion mediated by uPAR–?3?1 complexes is also RGD-
resistant (Wei et al., 1994). Instead, uPAR-bound ?1 integrins are
sensitive to ?1 peptides that map to the region of uPAR–integrin
interaction. As these ?1 peptides block cell adhesion (Fig. 5 B)
and migration of various tumor cells (Fig. 7), it is possible that
these reagents, perhaps coupled with RGD-based compounds,
have therapeutic potential for suppression of tumor progression.
by uPAR binding. Model proposes three basic forms of ?5?1 exist on the cell
surface: (1) A bent inactive form in the absence of integrin ligand; (2) an
extended, active form induced by Fn binding in the absence of uPAR; and
(3) a modified bent active form stabilized by uPAR binding, and potentially
other cis-acting membrane proteins. In the case of uPAR, the modified bent
form engages Fn differently as judged by Fn fragment III 9-11 and III 12-15
binding, altered ?1P1 peptide sensitivity, the reversal of Fn binding by the
presence of uPA–PAI-1 complexes, and enhanced Fn matrix assembly.
Model for regulation of ?5?1 integrin conformation and function
JCB • VOLUME 168 • NUMBER 3 • 2005510
Materials and methods
Reagents and antibodies
Active human uPA and uPAR mAbs were purchased from American Diag-
nostica. Active PAI-1 was a gift from Dr. Dan Lawrence (American Red
Cross, Rockville, MD). suPAR was supplied by Dr. Gary Deng (Berlex Bio-
sciences). Fn, Vn, and peptides GRGDSPK and GRADSPK were pur-
chased from Sigma-Aldrich. Fn fragment III 9-11 was provided by Dr.
Mark H. Ginsberg (The Scripps Research Institute, San Diego, CA). 804G
supernatant rich in Ln-5 was a gift from J.C. Jones (Northwestern Univer-
sity Medical School, Chicago, IL). Peptides ?1P1 (NLDSPEGGF), sc?1P1
(EDGLFNPSG), ?1P2 (FHFAGDGKL), sc?1P2 (KDGLFAHFG), and
?1P1SA (NLDAPEGGF) were synthesized at University of California, San
Francisco Biomolecular Resources Center and purified by HPLC. The anti-
bodies against active conformation of the ?1 integrins (HUTS-21 and
9EG7), mouse ?5-blocking antibody 5H10-27, and Fn mAb were pur-
chased from BD Biosciences. Integrin ?1 mAb (JB1A) was a gift from Dr.
John Wilkins (University of Manitoba, Winnipeg, Manitoba, Canada).
Blocking antibodies to human integrin ?3 (P1B5) and ?5 (P1D6) and ?5
pAb were purchased from CHEMICON International. HRP-conjugated
anti–human Fc? antibody was purchased from Jackson ImmunoResearch
Laboratories, Inc. uPAR mAb for blotting (R2) was a gift from G. Hoyer-
Hansen (Finsen Lab, Copenhagen, Denmark).
Mouse kidney epithelial cells expressing wt ?3 (a gift from Dr. Jordan A.
Kreidberg, Harvard Medical School, Boston, MA) or mut ?3 and their uPAR
cotransfected cells were cultured in DME as described previously (Wang et
al., 1999; Zhang et al., 2003). HEK293, human fibrosarcoma HT1080,
breast cancer MCF-7, T47D, and MDA-MB-231 Skov-3 cell lines were ob-
tained from American Type Culture Collection (Rockville, MD) and grown in
DME. The Tet-inducible uPAR cells (Tet-uPAR) were maintained in DME sup-
plemented with zeocin, hygromycin, and blasticidin (5 ?g/ml). Additional
Tet (2 ?g/ml) was added to induce uPAR expression. Modified Skov-3 cell
line was a gift from Dr. Ernest Lengyel (University of Chicago, Chicago, IL).
Cell detachment assay
Microtiter plates were coated with 5 ?g/ml Fn or Ln-5 supernatant (1:100)
for 18 h at 4?C. The cell detachment assay was performed as described
previously (Czekay et al., 2003). In brief, cells attached were acid washed,
resuspended in incubation buffer (RPMI, 20 mM Hepes, and 0.02% BSA),
and then incubated in the absence or presence of active uPA followed by
PAI-1. After wash, the remaining adherent cells were fixed and stained. The
amount of extracted stain was quantified by absorbance at 590 nm.
Cell adhesion assay
The cell adhesion assay was performed as described previously (Wei et
al., 2001). In brief, cells were seeded onto Fn (5 ?g/ml) or Fn fragment
(10 ?g/ml)–coated plates and incubated in DME/0.1% BSA with or with-
out RGD or ?1 peptides for 1 h at 37?C. After washing, attached cells
were fixed and stained with Giemsa. The data were quantified by measur-
ing absorbance at 550 nm.
Biotinylation of suPAR, III 9-11, and RGD peptides
Human suPAR, Fn fragment III 9-11, or RGD peptides were biotinylated at
0.25 mg/ml using FluoReporter Biotin-XX Protein Labeling Kit (Molecular
Probes, Inc.) following the manufacturer’s instructions.
Purification of ?5?1-Fc and ?5?1SA-Fc integrins
Integrin ?5-Fc and ?1-Fc or Ser227 to Ala mutant ?1SA-Fc constructs (?5/
pEE12.2hFc and ?1/pV.16hFc or ?1SA/pV.16hFc; Coe et al., 2001)
were transfected into 293 cells. Culture supernatant was harvested after
48–72 h and passed through a Protein A–agarose column. Soluble inte-
grin was eluted using 0.1 M glycine, pH 3.0, and neutralized in 1 M Tris-
HCl, pH 8.0. Protein-containing fractions were dialyzed, concentrated,
and identified by SDS-PAGE.
Purified protein binding assay
Nunc high binding microtiter plates were coated with 20 ?g/ml Fn and
blocked with 1% BSA. 20 ?g/ml purified recombinant ?5?1-Fc with or with-
out 20 nM purified suPAR was added to each well in PBS with 1 mM MnCl2,
and the plates were incubated for 1 h at 25?C. For different purposes, RGD
peptides or uPA–PAI-1 mixture may be added together with suPAR. After
washing, bound ?5?1-Fc was detected by protein A–HRP and quantified by
measuring absorbance at 490 nm. Data were expressed as specific binding
(i.e., total binding minus the binding to wells coated with BSA alone).
To test biotinylated RGD peptides binding to Fn–?5?1–suPAR,
Nunc microtiter plates were coated with Fn and incubated with ?5?1-Fc
with or without suPAR as above. 0.5 mM biotin-RGD was then added to
each well for another hour. After washing, avidin peroxidase was added
and the bound biotin-RGD was quantified as described above. To test
specificity of binding, 10-fold molar excess nonbiotinylated RGD peptides
were added. Biotinylated suPAR binding assay was performed similarly to
confirm Fn–?5?1–suPAR complex formation.
uPAR RNA interference
HT1080 cells were transfected with siRNAs that specifically target the
uPAR gene or nonsilencing control and used within 48–72 h. siRNA du-
plexes were synthesized by in vitro transcription. The sequence of the
DNA targeting uPAR is 5?-GGTGAAGAAGGGCGTCCAA-3?. A nonsi-
lencing siRNA 5?-AACCTGCGGGAAGAAGTGG-3? was used as a con-
trol (Vial et al., 2003). Synthetic siRNA oligonucleotides were purified
with Microspin G-25 columns from Amersham Biosciences.
Cells with or without siRNA uPAR transfection were incubated with pri-
mary antibody to active form ?1 integrin (HUTS-21, 9EG7) or to total ?1
integrin (JB1A) and secondary FITC-conjugated anti–mouse IgG or anti–rat
IgG (for 9EG7; Sigma-Aldrich) and analyzed on a flow cytometer (FACS-
Caliber; BD Biosciences). uPAR was detected by a mAb to uPAR.
Biotin-III 9-11 binding assay
All the procedures were done at 4?C. HT1080 cells were acid washed
and incubated without or with uPA followed by PAI-1. The cells were then
incubated with 50 nM biotinylated Fn fragment III 9-11 in RPMI/0.02%
BSA for 1 h. After washing, the cells were lysed and the total protein sep-
arated by SDS-PAGE. The bound biotin-III 9-11 was detected by avidin-
HRP. The bands were quantified and analyzed by densitometry.
Generation of inducible uPAR clones (Tet-uPAR)
Wt ?3 epithelial cells were transfected with Tet repressor (pcDNA6/TR,
blasticidin; Invitrogen) and Tet on an expression construct containing full-
length uPAR (pcDNA5/TO, hygromycin; Invitrogen) with a ratio of 6:1.
After antibiotic selection, 2 ?g/ml Tet was added and uPAR-expressing
clones were selected by cell sorting. Tet was removed and the cells re-
sorted for nonexpressing clones. These cells were diluted to select single
clones. The data presented in this paper were obtained from one of the
Detection of ECM-associated Fn
Fn fibrils were detected as described previously (Aguirre-Ghiso et al.,
2001). In brief, Tet-uPAR cells without or with Tet induction were lysed
with 3% Triton X-100 buffer. Triton-insoluble pellets were treated with
DNase and then extracted with 2% deoxycholate buffer. The insoluble
and soluble fractions were mixed with sample buffer and analyzed by
SDS-PAGE and Western blotting using anti–human Fn antibodies and an
antibody to ?-actin.
HT1080 cells were grown to confluence on Fn-coated surface. Medium
was replaced with DME/0.1% BSA 6 h before wounding. The wound was
made using a 1-ml pipet tip. The detached cells were removed by washing
and the wounded cells were incubated without or with RGD or ?1 pep-
tides (20 ?M) for 24 h. Cells were imaged at 0 and 24 h by phase-con-
HT1080 cells were lysed in Triton lysis buffer (50 mM Hepes, pH 7.5,
150 mM NaCl, and 1% Triton X-100) supplemented with protease inhib-
itors and 1 mM PMSF. Clarified lysates were incubated with or without
400 ?M peptide ?1P1 and immunoprecipitated with antibody to inte-
grin ?5 (P1D6). The immunoprecipitates were blotted for uPAR (R2) or
integrin ?5 (pAb).
Online supplemental material
The supplemental material (Figs. S1–S3) is available at http://
www.jcb.org/cgi/content/full/jcb.200404112/DC1. Fig. S1 shows that
suppression of uPAR expression by RNA interference induces LIBS epitope
and changes ?5?1-mediated Fn binding in MDA-MB-231 cells. Fig. S2
shows the effect of uPAR expression level on Fn adhesion of different
breast cancer cell lines. Fig. S3 shows the RGD-resistant adhesion of
uPAR-expressing cells on different concentrations of Fn.
UROKINASE RECEPTOR REGULATES ?5?1 INTEGRIN • WEI ET AL. 511
We thank Dr. D. Lawrence for PAI-1, Dr. G. Deng for suPAR, Dr. M.H. Gins-
berg for Fn fragment III 9-11, Dr. J.C. Jones for Ln-5, Dr. J.A. Wilkins for ?1 in-
tegrin antibody, Dr. G. Hoyer-Hansen for R2 antibody, Dr. E. Lengyel for
Skov-3 cells, and Dr. J.A. Kreidberg for the murine epithelial cells.
This work was supported by National Institutes of Health grants
HL44712 (to H.A. Chapman) and HL 31950 (to D.J. Loskutoff).
Submitted: 20 April 2004
Accepted: 7 December 2004
Aguirre-Ghiso, J.A., K. Kovalski, and L. Ossowski. 1999. Tumor dormancy in-
duced by downregulation of urokinase receptor in human carcinoma in-
volves integrin and MAPK signaling. J. Cell Biol. 147:89–104.
Aguirre-Ghiso, J.A., D. Liu, A. Mignatti, K. Kovalski, and L. Ossowski. 2001.
Urokinase receptor and fibronectin regulate the ERKMAPK to p38MAPK ac-
tivity ratios that determine carcinoma cell proliferation or dormancy in
vivo. Mol. Biol. Cell. 12:863–879.
Aguirre-Ghiso, J.A., Y. Estrada, D. Liu, and L. Ossowski. 2003. ERKMAPK ac-
tivity as a determinant of tumor growth and dormancy; regulation by
p38SAPK. Cancer Res. 63:1684–1695.
Ahmed, N., K. Oliva, Y. Wang, M. Quinn, and G. Rice. 2003. Downregulation of
urokinase plasminogen activator receptor expression inhibits Erk signal-
ling with concomitant suppression of invasiveness due to loss of uPAR-
?1 integrin complex in colon cancer cells. Br. J. Cancer. 89:374–384.
Andreasen, P.A., L. Kjoller, L. Christensen, and M.J. Duffy. 1997. The uroki-
nase-type plasminogen activator system in cancer metastasis: a review.
Int. J. Cancer. 72:1–22.
Arnaout, M.A., S.L. Goodman, and J.P. Xiong. 2002. Coming to grips with inte-
grin binding to ligands. Curr. Opin. Cell Biol. 14:641–651.
Bazzoni, G., D.T. Shih, C.A. Buck, and M.E. Hemler. 1995. Monoclonal antibody
9EG7 defines a novel ?1 integrin epitope induced by soluble ligand and
manganese, but inhibited by calcium. J. Biol. Chem. 270:25570–25577.
Chapman, H.A., and Y. Wei. 2001. Protease crosstalk with integrins: the uroki-
nase receptor paradigm. Thromb. Haemost. 86:124–129.
Coe, A.P., J.A. Askari, A.D. Kline, M.K. Robinson, H. Kirby, P.E. Stephens,
and M.J. Humphries. 2001. Generation of a minimal ?5?1 integrin-Fc
fragment. J. Biol. Chem. 276:35854–35866.
Cukierman, E., R. Pankov, D.R. Stevens, and K.M. Yamada. 2001. Taking cell-
matrix adhesions to the third dimension. Science. 294:1708–1712.
Czekay, R.P., K. Aertgeerts, S.A. Curriden, and D.J. Loskutoff. 2003. Plasmino-
gen activator inhibitor-1 detaches cells from extracellular matrices by in-
activating integrins. J. Cell Biol. 160:781–791.
Degryse, B., S. Orlando, M. Resnati, S.A. Rabbani, and F. Blasi. 2001. Uroki-
nase/urokinase receptor and vitronectin/?v?3 integrin induce chemotaxis
and cytoskeleton reorganization through different signaling pathways.
Forsyth, C.B., J. Pulai, and R.F. Loeser. 2002. Fibronectin fragments and block-
ing antibodies to ?2?1 and ?5?1 integrins stimulate mitogen-activated
protein kinase signaling and increase collagenase 3 (matrix metallopro-
teinase 13) production by human articular chondrocytes. Arthritis
Hocking, D.C., J. Sottile, and P.J. McKeown-Longo. 1998. Activation of dis-
tinct ?5?1-mediated signaling pathways by fibronectin’s cell adhesion
and matrix assembly domains. J. Cell Biol. 141:241–253.
Huang, W., R. Chiquet-Ehrismann, J.V. Moyano, A. Garcia-Pardo, and G.
Orend. 2001. Interference of tenascin-C with syndecan-4 binding to fi-
bronectin blocks cell adhesion and stimulates tumor cell proliferation.
Cancer Res. 61:8586–8594.
Jin, H., and J. Varner. 2004. Integrins: roles in cancer development and as treat-
ment targets. Br. J. Cancer. 90:561–565.
Kim, J., I. Han, Y. Kim, S. Kim, and E.S. Oh. 2001. C-terminal heparin-binding
domain of fibronectin regulates integrin-mediated cell spreading but not the
activation of mitogen-activated protein kinase. Biochem. J. 360:239–245.
Luque, A., M. Gomez, W. Puzon, Y. Takada, F. Sanchez-Madrid, and C. Cabanas.
1996. Activated conformations of very late activation integrins detected by
a group of antibodies (HUTS) specific for a novel regulatory region (355-
425) of the common ?1 chain. J. Biol. Chem. 271:11067–11075.
Monaghan, E., V. Gueorguiev, C. Wilkins-Port, and P.J. McKeown-Longo.
2004. The receptor for urokinase-type plasminogen activator regulates
fibronectin matrix assembly in human skin fibroblasts. J. Biol. Chem.
Mould, A.P., and M.J. Humphries. 1991. Identification of a novel recognition
sequence for the integrin ?4?1 in the COOH-terminal heparin-binding
domain of fibronectin. EMBO J. 10:4089–4095.
Mould, A.P., and M.J. Humphries. 2004. Regulation of integrin function
through conformational complexity: not simply a knee-jerk reaction?
Curr. Opin. Cell Biol. 16:544–551.
Mould, A.P., J.A. Askari, and M.J. Humphries. 2000. Molecular basis of ligand
recognition by integrin ?5?1. I. Specificity of ligand binding is deter-
mined by amino acid sequences in the second and third NH2-terminal re-
peats of the ? subunit. J. Biol. Chem. 275:20324–20336.
Pluskota, E., D.A. Soloviev, and E.F. Plow. 2003. Convergence of the adhesive
and fibrinolytic systems: recognition of urokinase by integrin ?M?2 as
well as by the urokinase receptor regulates cell adhesion and migration.
Redick, S.D., D.L. Settles, G. Briscoe, and H.P. Erickson. 2000. Defining fi-
bronectin’s cell adhesion synergy site by site-directed mutagenesis. J. Cell
Simon, D.I., Y. Wei, L. Zhang, N.K. Rao, H. Xu, Z. Chen, Q. Liu, S. Rosen-
berg, and H.A. Chapman. 2000. Identification of a urokinase receptor-
integrin interaction site. Promiscuous regulator of integrin function.
J. Biol. Chem. 275:10228–10234.
Sitrin, R.G., D.R. Johnson, P.M. Pan, D.M. Harsh, J. Huang, H.R. Petty, and
R.A. Blackwood. 2004. Lipid raft compartmentalization of urokinase
receptor signaling in human neutrophils. Am. J. Respir. Cell Mol. Biol.
Sturge, J., D. Wienke, L. East, G.E. Jones, and C.M. Isacke. 2003. GPI-
anchored uPAR requires Endo180 for rapid directional sensing during
chemotaxis. J. Cell Biol. 162:789–794.
Takagi, J., B.M. Petre, T. Walz, and T.A. Springer. 2002. Global conforma-
tional rearrangements in integrin extracellular domains in outside-in and
inside-out signaling. Cell. 110:599–611.
Takagi, J., K. Strokovich, T.A. Springer, and T. Walz. 2003. Structure of integrin
?5?1 in complex with fibronectin. EMBO J. 22:4607–4615.
Tarui, T., N. Andronicos, R.P. Czekay, A.P. Mazar, K. Bdeir, G.C. Parry, A.
Kuo, D.J. Loskutoff, D.B. Cines, and Y. Takada. 2003. Critical role of
integrin ?5?1 in urokinase (uPA)/urokinase receptor (uPAR, CD87)
signaling. J. Biol. Chem. 278:29863–29872.
van der Pluijm, G., B. Sijmons, H. Vloedgraven, C. van der Bent, J.W. Drijf-
hout, J. Verheijen, P. Quax, M. Karperien, S. Papapoulos, and C. Lowik.
2001. Urokinase-receptor/integrin complexes are functionally involved
in adhesion and progression of human breast cancer in vivo. Am. J.
Vial, E., E. Sahai, and C.J. Marshall. 2003. ERK-MAPK signaling coordinately
regulates activity of Rac1 and RhoA for tumor cell motility. Cancer Cell.
Wang, Z., J.M. Symons, S.L. Goldstein, A. McDonald, J.H. Miner, and J.A.
Kreidberg. 1999. ?3?1 integrin regulates epithelial cytoskeletal organiza-
tion. J. Cell Sci. 112:2925–2935.
Wei, Y., D.A. Waltz, N. Rao, R.J. Drummond, S. Rosenberg, and H.A. Chap-
man. 1994. Identification of the urokinase receptor as an adhesion recep-
tor for vitronectin. J. Biol. Chem. 269:32380–32388.
Wei, Y., M. Lukashev, D.I. Simon, S.C. Bodary, S. Rosenberg, M.V. Doyle,
and H.A. Chapman. 1996. Regulation of integrin function by the uroki-
nase receptor. Science. 273:1551–1555.
Wei, Y., J.A. Eble, Z. Wang, J.A. Kreidberg, and H.A. Chapman. 2001. Uroki-
nase receptors promote ?1 integrin function through interactions with
integrin ?3?1. Mol. Biol. Cell. 12:2975–2986.
Xiong, J.P., T. Stehle, B. Diefenbach, R. Zhang, R. Dunker, D.L. Scott, A.
Joachimiak, S.L. Goodman, and M.A. Arnaout. 2001. Crystal structure
of the extracellular segment of integrin ?V?3. Science. 294:339–345.
Xue, W., I. Mizukami, R.F. Todd III, and H.R. Petty. 1997. Urokinase-type
plasminogen activator receptors associate with ?1 and ?3 integrins of fi-
brosarcoma cells: dependence on extracellular matrix components.
Cancer Res. 57:1682–1689.
Yebra, M., L. Goretzki, M. Pfeifer, and B.M. Mueller. 1999. Urokinase-type
plasminogen activator binding to its receptor stimulates tumor cell mi-
gration by enhancing integrin-mediated signal transduction. Exp. Cell Res.
Zhang, F., C.C. Tom, M.C. Kugler, T.T. Ching, J.A. Kreidberg, Y. Wei, and
H.A. Chapman. 2003. Distinct ligand binding sites in integrin ?3?1 reg-
ulate matrix adhesion and cell–cell contact. J. Cell Biol. 163:177–188.