| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
* Pulmonary Cell Research, and
Cardiovascular Research-Signal Transduction, Department of Research, University Hospital Basel, Basel, Switzerland; and
The Woolcock Institute for Medical Research, Camperdown, Australia
1Correspondence: Pulmonary Cell Research, Department of Research, Basel University Hospital, Hebelstrasse 20, CH-4031 Basel, Switzerland. E-mail: michelr{at}med.usyd.edu.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: uPAR • signaling • FAK • SHP2 • cell migration
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. UPAR is a glycosylphosphatidylinositol (GPI) -anchored extracellular protein that lacks transmembrane and cytoplasmic domains, and its ability to transduce signals is presumed to be mediated through formation of signaling complexes with other transmembrane proteins such as the integrins (2)
. These signaling complexes may be organized through selective inclusion within (or exclusion from) membrane domains. Classification of such domains remains contentious; terms such as membrane rafts and caveolae are used depending on the preferences of each research group. Lipid domains that contain uPAR and other GPI-anchored proteins are characterized by ordered lipids, a predominance of saturated acyl chains, enrichment of sphingolipids on the external membrane leaflet and of phosphatidyl ethanolamine on the internal membrane leaflet, and high cholesterol content within the membrane. Cholesterol is an important component of ordered lipid domains and is closely associated with caveolin. There may be a continuum of ordered lipid domains from caveolin-free rafts to the classic flask-shaped caveolae. The high packing order of these domains tends to exclude most membrane proteins, which partition to more unsaturated and fluid membrane areas. Several paradigms for functional clustering of membrane proteins in lipid domains have been formulated, including clustering of immunoglobulin receptors for immune signaling, Ephrin-related signaling in cell guidance, and integrin-related signaling in cell adhesion. A general scheme is that ligand binding causes clustering of GPI-anchored receptor proteins, which then provide a favorable environment for other transmembrane proteins and on the inner cytoplasmic leaflet for palmitoylated proteins, such as Src family members.
Cell migration requires coordination of a suite of proteins that establish the physical link between the ECM and the cytoskeleton. These focal adhesions or focal contacts contain integrins as receptors of ECM proteins whereas actin chains are anchored on the cytoplasmic side through vinculin, paxillin, and talin (3)
. Migrating cells must break down established areas of ECM attachment and reform them at the leading edge of the cell. Thus the change to a migrating phenotype involves loss of large focal adhesions and formation of numerous small and short-lived focal contacts (4)
. A central molecule in coordination of adhesion stability is focal adhesion kinase (FAK).
Our previous results (5)
showed that urokinase alone did not stimulate ERK in quiesced HASM cells and was unable to initiate cell migration. However, urokinase potentiated ERK activation in PDGF-stimulated cells and enhanced PDGF chemotaxis. This observation raised the question of how urokinase and uPAR signal and which signaling proteins would control and enhance chemokine-dependent cell migration. Here we show that urokinase and uPAR signal through alterations in the localization of membrane-associated phosphatases and kinases and that these effects are paralleled by changes in the membrane association and phosphorylation state of FAK.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Antibodies for FAK (354-533), phospho-FAK (pY397), Grb2, and caveolin 1 were from BD (Franklin Lakes, NJ, USA). SHP2 and v-Src antibodies were from Sigma. Anti-G
s, anti-phosphotyrosine, and anti-uPAR were from Calbiochem, Santa Cruz (Santa Cruz, CA, USA), and American Diagnostica (Salamander Bay NSW, Australia), respectively.
Human airway smooth muscle (HASM) cell culture
Primary cultures of HASM cells were established as reported previously (6)
. Briefly, macroscopically normal human lung tissue was obtained from patients undergoing partial resection. Large bronchi (5–15 mm internal diameter) were dissected from the surrounding parenchyma and the epithelial layer was removed. ASM bundles were dissected from the bronchi and placed in tissue flasks containing DMEM supplemented with 10% fetal bovine serum (10% FBS). The smooth muscle cells grew to confluence in a humidified CO2 incubator (37°C) in 
28 days and were passaged into 75 cm2 flasks at 7 day intervals. Pure populations of smooth muscle cells were confirmed by the presence of positive staining for -smooth muscle actin.
Preparation of cell extracts
For membrane fractions and immunoprecipitation, cells were grown to confluence (6 days) in 10 cm diameter culture dishes, then quiesced for 24 h in 1% FBS. For whole-cell extracts, cells were grown under the same conditions in 6-well culture plates. After stimulation, dishes were immediately cooled on ice and rinsed twice with ice-cold PBS. For whole-cell extracts, 200 µL of 2 x electrophoresis sample buffer (1% SDS; 3.75% glycerol; 15.625 mM Tris-HCl, pH 6.8; 0.001% bromophenol blue, 100 mM DTT) was added directly to the wells, lysed cells were collected by scraping from the wells with the pipette tip, and the wells were washed with a further 200 µL of water. For membrane extraction, cells were scraped off with a cell scraper into 2 mL of extraction buffer (20 mM Tris pH 7.6, 1 mM EDTA, 150 mM NaCl, 1 mM PMSF, 5 µg/mL leupeptin, 5 µg/mL aprotinin). Cells were left to swell at 4°C for 10 min, then homogenized by repeated pipetting through a fine tip. Homogenates were cleared of cell debris by centrifugation at 2000 x g for 5 min. Supernatants were centrifuged at 18,000 x g for 60 min to sediment membrane particles (7)
. The supernatant was kept as a cytosolic fraction and the pellet was washed with extraction buffer, then resuspended in 50 µL of membrane extraction buffer (containing 0.1% Triton X-100). Extracts were mixed on a turntable at 4°C for 1 h, and stored at –20°C.
Immunoprecipitation
For immunoprecipitation, cells were treated as for membrane extraction but extracted in buffer containing 0.1% Triton X-100. After homogenization and low-speed centrifugation to remove cell debris, supernatants were incubated with primary antibodies (anti-uPAR or anti-SHP2) for 2 h on a turntable at 4°C. Protein G agarose was then added and mixing continued for another hour. Agarose was sedimented by centrifugation, the supernatant was removed, and immunoprecipitated complexes were washed 3x in 10 bed volumes of PBS. The agarose was then extracted in 1 bed volume of electrophoresis sample buffer, which was removed and stored at –20°C.
Electrophoresis and Western blots
Extracts were mixed with an equal volume of electrophoresis sample buffer, boiled for 5 min, then subjected to standard SDS PAGE and electrotransfer onto PVDF membranes. Protein loading equivalence was routinely verified after electrotransfer to PVDF membranes by staining with Ponceau S and immunoblotting for Gs subunit in the case of membrane and whole-cell lysate samples. 12.5% gels were used for small proteins (10–40 kDa), 10% gels for mid-range (40–80 kDa), and 7.5% gels for larger proteins (80–150 kDa). For immunoblotting, PVDF membranes were blocked overnight in 5% skim milk powder solution and incubated with primary and secondary antibodies according to the suppliers’ recommendations. Specific staining was detected by HRP-catalyzed chemiluminescence and exposure of photographic film.
Migration assay
Cells from passages 4–7 were seeded into 6-well plates at a density of 6 x 104 cells/cm2 in 10% FBS and allowed to settle overnight. For the next 24 h cells were quiesced in DMEM containing 1% FBS. Cell culture membrane inserts with 8 µm pore size in 24-well plate format (Falcon, Oxnard, CA, USA) were collagen coated by incubation with collagen solution (0.1 mg/mL collagen type 1 in DMEM diluted from 1 mg/mL stock in 0.1 M acetic acid) for 1 h. Coated inserts were air dried for 1 h, then washed with DMEM containing 0.1% BSA. Cells were lifted with minimal trypsinization for the chemotaxis assay (terminated with 10% FBS), resuspended in DMEM containing 0.1% BSA, and added to the upper compartment of the transwells at 2 x 104 cells/mL. Migration assays were carried out in 1 mL DMEM containing 0.1% BSA. Cells were left to settle for 30 min in a CO2 incubator before treatments were added to the top or bottom compartments. At the end of the assay (4 h), membranes were fixed in ice-cold methanol. The upper membrane surface was scraped clean of cells and cells that migrated through to the membrane underside were stained with hematoxylin. Membranes were then mounted on slides and the cells were manually counted as the number of nuclei in a standard central field.
Statistical analysis
All experiments were performed on at least three independent occasions. Results are given as mean ± SD. Statistical significance was evaluated using Student’s paired t test; a P value of <0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, membrane domains where uPAR is located. For this reason, we tested whether SHP2 localization in HASM cells was affected by migratory stimuli. SHP2 partitioned differentially in the membrane fraction depending on the type of stimulus (Fig. 1A
). PDGF stimulation for 30 min increased the amount of membrane-associated SHP2 3-fold above unstimulated levels (P<0.05, n=6); costimulation with PDGF and urokinase had a similar effect, with no further significant increase compared with PDGF alone. In contrast, stimulation with urokinase alone significantly decreased membrane-associated SHP2 to 0.7-fold that of unstimulated levels (P<0.05, n=6).
|
To test whether uPAR was acting through physical association with SHP2, whole-cell extracts were immunoprecipitated with anti-uPAR antibody after stimulation of HASM cells with urokinase for up to 40 min. Protein staining of extracts resolved on 10% SDS gels showed a number of proteins in the immunoprecipitates; Western blots of the immunoprecipitates showed that SHP2 associates with uPAR differentially during a time course of 40 min of urokinase exposure (Fig. 1B
). SHP2 coprecipitated with uPAR in unstimulated cells, but stimulation with urokinase caused a rapid (significant within 2.5 min) increase in SHP2. After 20 min, the amount of uPAR-associated SHP2 decreased, reaching 60% of control levels by 40 min. We tested the association of uPAR with SHP2 by converse immunoprecipitation with SHP2 antibody and found uPAR in Western blots, although the amount of associated uPAR did not vary during urokinase stimulation (Fig. 1C
). Total cell uPAR and SHP2 levels did not alter significantly over the same period (Fig. 1D
).
Membrane fractions
We extended the protein analysis to membrane fractions from urokinase-stimulated HASM cells. In Western blots of membrane fractions, FAK bands appeared at two molecular weights: a whole FAK at 120 kDa and a smaller band at 85 kDa. 120 kDa FAK was readily detectable in whole-cell extracts and the cytoplasmic fraction of cell extracts (not shown), which indicates that the membrane-associated FAK represents only a fraction of total cellular FAK. In contrast, the 85 kDa FAK band was undetectable in whole-cell or cytoplasm fractions, and so appeared to be unique to the membrane fraction, where it constituted a major fraction of FAK. However, the relative proportion of the 85 kDa FAK was variable between individual cell cultures. The 120 kDa band decreased in the membrane fraction in response to urokinase whereas the smaller band increased (Fig. 2
A).
|
The phosphorylation status of FAK in membranes was determined using anti-phospho FAK antibodies that recognize Tyr 397, which is contained in 120 and 85 kDa proteins (19)
. In response to urokinase, the proportion of phosphorylated FAK significantly increased to 3-fold in the whole 120 kDa protein (Fig. 2B
). Although we noted that the phosphorylation state of the smaller 85 kDa protein decreased marginally within the first 20 min and increased thereafter (Fig. 2B
), these differences were not significant. Changes in phosphorylation status of whole FAK (120 kDa) were readily detectable in whole-cell lysates whereby urokinase induced a 3-fold increase (Fig. 2C
). Using anti-pY397, we were unable to detect 85 kDa in whole-cell lysates, as was the case using anti-FAK antibodies.
Depletion of cholesterol by preincubation of cells with cyclodextrin increased the level of FAK phosphorylation in the cell membrane fraction either with or without subsequent urokinase incubation (Fig. 3
). Inhibition of phosphatase activity with vanadate increased phosphorylation of FAK.
|
Effects of cyclodextrin and vanadate on cell migration
We used two interventions to assess the importance of membrane rafts and SHP2 in urokinase signaling. Cyclodextrin treatment (10 mM, 30 min) depletes membrane cholesterol and disturbs raft formation (9)
. We found that cyclodextrin had no effect on basal (unstimulated) cell migration but significantly potentiated chemotaxis toward PDGF (P<0.05, n=3) (Fig. 4
). Costimulation with urokinase and PDGF did not further increase chemotaxis. Vanadate, which blocks the dephosphorylase action of SHP2, significantly reduced cell migration toward PDGF both with and without urokinase (P<0.05, n=3). Vanadate removed the potentiation of PDGF chemotaxis by urokinase (Fig. 4)
.
|
Other associated proteins
Further Western blot analysis of the anti-uPAR immunoprecipitates from whole-cell extracts was performed to determine whether other proteins implicated in focal adhesion signaling were physically linked to uPAR. Caveolin 1 was detected in anti-uPAR immunoprecipitates, and its association with uPAR decreased with the time of urokinase exposure (Fig. 5
A). Levels of total cell caveolin did not change throughout the 40 min time course of urokinase exposure. Detection of Src in these immunoprecipitates was not possible due to the presence of bands at the same molecular mass (55 kDa) from the immunoprecipitate antibodies. Although Src was not measurable in immunoprecipitates for technical reasons, it was readily detectable in membrane fractions, where it accumulated 6-fold in response to urokinase stimulation (Fig. 5B
). Urokinase-stimulated accumulation of Src in the membrane fraction was completely blocked after blockade of phosphatase activity by preincubation with vanadate. Levels of whole-cell Src did not change with urokinase stimulation (Fig. 5D
).
|
We tested for the presence of the adaptor protein Grb2, which has been implicated in signaling of tyrosine-phosphorylated proteins (10)
and SHP2 (11)
. We were unable to detect Grb2 in anti-uPAR immunoprecipitates. Grb2 was readily detectable in membrane fractions, but found no significant change during incubation with urokinase or PDGF (Fig. 5C
). Levels of whole-cell Grb2 did not change (Fig. 5D
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Urokinase did, however, potentiate cell migration and MAP kinase activation when PDGF, a primary migratory stimulus, was present. These observations were consistent with a role for urokinase in reorganizing signal transduction molecules to facilitate migration. Although numerous reports have described signal transduction somewhat downstream from uPAR, the more proximal signaling intermediates are less well characterized. The present study provides direct evidence for the reorganization of signal transduction molecules; the urokinase signaling we observed is consistent with a model of functional clusters of proteins on the plasma membrane.
Cells attach to a protein matrix via integrins, which become clustered and connected to the actin cytoskeleton through focal adhesions (reviewed by Giancotti and Rouslahti, refs 12
, 13
). Signaling and linking proteins in these adhesion complexes then generate intracellular signals that "read" the ECM, the starting state for HASM cells in our experiments. In airway smooth muscle cells, the ECM provides a strong survival signal (14)
; the local composition of the ECM seems to control cell response, as cells plated onto collagen have low levels of spontaneous migration. Focal adhesions (or contacts) are central to cell migration as they must be newly formed at the advancing edge of the cell and disassembled in the trailing part. uPAR has been shown to associate with focal adhesions by colocalization studies (15)
and by coimmunoprecipitation with ß1-integrins (16)
. This association was shown to be functional (16)
through expression of various deletion forms of the proteins. Various schemes for the composition of focal adhesions have been proposed, but a general one, as envisaged by Frame et al. (3)
, involves FAK as a central module that physically links the cytoskeleton to the ECM via paxillin and talin (internal) and integrins (external). The ECM protein fibronectin can be a promigratory stimulus in some cell types (17
, 18)
; when HEK293 cells attach to a fibronectin matrix, they respond with phosphorylation of FAK and Src.
In this study we found that urokinase stimulation of HASM cells affected FAK by changing its localization, phosphorylation, and breakdown. Urokinase stimulation caused a progressive loss from the membrane fraction with a concomitant increase in FAK phosphorylation, and appearance of an 85 kDa fragment. The antibodies we used to detect FAK and phospho-FAK recognize the same FAK region—residues 354-533 and tyrosine 397, respectively, which are in the central kinase region on the N-terminal side. Similar-sized cleavage products of FAK have been described and attributed to the actions of several proteases. Gervais et al. (19)
described cleavage of FAK after apoptotic stimuli and found several cleavage sites, including one at VSWD704 that is preferentially cleaved by caspase 6 and generates an 85 kDa N-terminal fragment. Other authors have described an 85 kDa FAK fragment produced by caspase 3 (20
, 21)
. A suite of smaller fragments has been described in apoptosis of renal epithelial cells (22)
. FAK can also be cleaved by calpain (23
, 24)
; using an inducible form of the constitutively active vSrc, it was shown that Src activity causes calpain cleavage of FAK accompanied by focal adhesion disassembly. Our observations are consistent with phosphorylation of FAK by Src and subsequent cleavage by a protease. Urokinase stimulation caused a 6-fold increase of Src in the membrane fraction. FAK has motifs for attachment to the integrin part of focal contacts at its N-end and cytoskeletal attachments at its C-end, so cleavage between the two ends of the molecule may provide a mechanism for releasing focal adhesions from the cytoskeleton.
uPAR is reported to interact physically and functionally with integrins (16
, 25
26
27)
, but how this may contribute to further signaling has not been established. As integrins and uPAR lack intracellular receptors in the formal sense, it seems that signaling must take place through alteration of protein associations within the focal adhesion/contact. The phosphotyrosine phosphatase SHP2 (28)
is often implicated in cell migration; in IGF-stimulated cell migration, for instance, IGF promotes association of a FAK-paxillin-SHP2 complex that results in dephosphorylation of FAK and paxillin (29)
. Miao et al. (30)
found a similar mechanism after ephrin stimulation, leading to weakening of focal adhesions. Chernock et al. (31)
found that overexpression of SHP2 increased chemotaxis whereas a nonfunctional SHP2 caused reduced chemotaxis (29)
. In our study we found that urokinase stimulation rapidly brought SHP2 into association with uPAR, then (>20 min) saw SHP2 disassociate. That SHP2 was involved in general chemotaxis was confirmed by our observation that it was also brought to the membrane fraction by PDGF stimulation. At the functional level, when we blocked the dephosphorylase activity of SHP2 with vanadate, cell migration toward PDGF was inhibited in both the presence and absence of urokinase and the potentiating action of urokinase on PDGF chemotaxis was completely removed. The tyrosine kinase Src was involved in urokinase signaling in our cells, where it increased 6-fold in the membrane fraction. The functional association of Src with uPAR and FAK has been reported in a cancer cell line (25)
, and Src and FAK act in concert in signaling for cell migration (32)
.
Lipid rafts appear to be the location of many of the proteins involved in cell adhesion and migration. When SHP2 was modified to be targeted to raft domains of HEK293 cells, it became constitutively activated and caused phosphorylation of FAK (7)
. Wei et al. described a membrane complex of caveolin, cholesterol, Src, ß1 integrins, FAK, and uPAR (27)
. In their scheme, uPAR associates with and stabilizes the complexes, whereas uPAR binding peptides disrupt the complexes and inhibit cell-ECM adhesion (16)
. Src is reported to interact directly with caveolin by phosphorylating it (33)
. Our results support a model (Fig. 6
) where uPAR associates with caveolin, SHP2, FAK, and Src in membrane domains that can be isolated by immunoprecipitation and disrupted by cholesterol depletion. Stimulation with urokinase causes increased association of SHP2 and Src and increased phosphorylation of FAK, which in turn brings about proteolysis of FAK and breakdown of the membrane domains. Urokinase causes uPAR to disassociate from caveolin. Disruption of the domains by cholesterol stripping prevented urokinase signaling as assessed by Src accumulation and loss of its ability to potentiate chemotaxis toward PDGF.
|
In conclusion, our results support a model of urokinase action in which it promotes cell migration and chemotaxis by breaking down focal adhesion contacts. This explains its functional role in chemotaxis in HASM cells of facilitating cell migration without itself providing a primary migratory stimulus.
Received for publication March 25, 2004. Accepted for publication August 5, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. T. Gerthoffer Migration of Airway Smooth Muscle Cells Proceedings of the ATS, January 1, 2008; 5(1): 97 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Bryer and T. J. Koh The urokinase-type plasminogen activator receptor is not required for skeletal muscle inflammation or regeneration Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R1152 - R1158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Shetty, T. Velusamy, S. Idell, H. Tang, and P. K. Shetty Regulation of urokinase receptor expression by protein tyrosine phosphatases Am J Physiol Lung Cell Mol Physiol, February 1, 2007; 292(2): L414 - L421. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Maquerlot, S. Galiacy, M. Malo, C. Guignabert, D. A. Lawrence, M.-P. d'Ortho, and G. Barlovatz-Meimon Dual Role for Plasminogen Activator Inhibitor Type 1 as Soluble and as Matricellular Regulator of Epithelial Alveolar Cell Wound Healing Am. J. Pathol., November 1, 2006; 169(5): 1624 - 1632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. Hardin and J. Vallejo Caveolins in vascular smooth muscle: Form organizing function Cardiovasc Res, March 1, 2006; 69(4): 808 - 815. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Pulukuri, C. S. Gondi, S. S. Lakka, A. Jutla, N. Estes, M. Gujrati, and J. S. Rao RNA Interference-directed Knockdown of Urokinase Plasminogen Activator and Urokinase Plasminogen Activator Receptor Inhibits Prostate Cancer Cell Invasion, Survival, and Tumorigenicity in Vivo J. Biol. Chem., October 28, 2005; 280(43): 36529 - 36540. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
