| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
-Associated Protein-70 (ZAP-70) Tyrosine Kinase: Relevance for Invasion and Migration of a T Cell Hybridoma1
Division of Cell Biology, The Netherlands Cancer Institute, Amsterdam, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-associated protein
(ZAP)-70. The truncated ZAP-70 also blocked LFA-1-dependent chemotaxis
through ICAM-1-coated filters induced by 1 ng/ml stromal cell-derived
factor-1, but not LFA-1-independent chemotaxis induced by 100 ng/ml
stromal cell-derived factor-1. This suggested that LFA-1 engagement
triggers a signal that amplifies a weak chemokine signal and that
dominant-negative ZAP-70 blocks this LFA-1 signal. Here we show that
cross-linking of part of the LFA-1 molecules with Abs causes activation
of free LFA-1 molecules (not occupied by the Ab) on the same cell,
which then bind to ICAM-2 on other cells. This causes cell aggregation
that was also blocked by dominant-negative ZAP-70. Thus, an LFA-1
signal involving ZAP-70 activates other LFA-1 molecules, suggesting
that the chemokine signal can be amplified by multiple cycles of LFA-1
activation. The chemokine and the LFA-1 signal were both blocked by a
phospholipase C inhibitor and a calpain inhibitor, suggesting that one
of the amplified signals is the phospholipase C-dependent activation of
calpain. Finally, we show that both Src-homology 2 domains are required
for inhibition of invasion, chemotaxis, and aggregation by the
truncated ZAP-70, suggesting that ZAP-70 interacts with a
phosphorylated immunoreceptor tyrosine-based activation motif (ITAM)
sequence. Remarkably, this is not an ITAM in the TCR/CD3 complex
because this is not expressed by this T cell
hybridoma. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adhesion mediated by LFA-1 is tightly regulated (1). LFA-1 is usually not active, i.e., LFA-1-expressing leukocytes do not spontaneously adhere to LFA-1 ligands. Physiologically relevant signals that activate LFA-1 are triggered by the TCR/CD3 complex upon interaction with MHC molecules on APCs or on target cells of cytotoxic T cells (11). Furthermore, LFA-1 and other integrins are activated by cytokines such as IL-2 (12) and by chemokines and other inflammation mediators such as platelet-activating factor, relevant for leukocyte influx into inflamed tissues (13, 14, 15). This phenomenon has been termed "inside-out" signaling (16) and can be mimicked by binding of Abs against CD3 (17, 18) and against several other surface molecules, including CD2 (18), CD28 (19), and CD44 (20).
Upon ligand engagement, LFA-1 not only provides physical attachment but it also generates signals that, depending on the circumstances, may be required for proliferation (21) or prevention of apoptosis (22). Thus, binding to either the physiological ligand ICAM-1 or to Abs provides costimulatory signals for activation of T cells (21, 23, 24). This phenomenon is termed "outside-in" signaling (16). Some Abs directed against LFA-1 induce aggregation of cells. In part, these mAbs are directed against "activation epitopes" and may act by stabilizing the "active" conformation of the integrins (25, 26, 27, 28). However, aggregation can also be induced by mAbs against epitopes that are not activation sensitive (29, 30). This suggests that outside-in signals generated by Ab binding trigger inside-out signals that lead to integrin activation. These mAbs often block the interactions of their cognate integrins with ligand and it was thus thought that the aggregation was mediated by other integrins or nonintegrin adhesion molecules. Indeed, triggering of one integrin subtype by another does occur (31). However, Koopman et al. (29) showed that aggregation induced by an LFA-1 mAb occurred at low but not high mAb concentrations and was mediated by LFA-1 itself. This suggested that triggering of part of the LFA-1 molecules, at a subsaturating mAb concentration, led to activation of other LFA-1 molecules that were not occupied by the mAb. In this report, we demonstrate that this is in fact true for aggregation of mouse T cell hybridoma cells induced by the M17/4 and GAME-46 LFA-1-blocking mAbs.
Recently, we showed that the kinase activity of the -associated
protein-70 (ZAP-70)4 tyrosine
kinase is essential for LFA-1-dependent in vitro invasion and in vivo
dissemination of T cell hybridomas (32). Invasion and
dissemination are also blocked by pertussis toxin (33),
indicating that Gi protein-coupled receptors such
as chemokine receptors are involved, suggesting that invasion is
triggered by a chemokine. A potentially involved chemokine is stromal
cell-derived factor (SDF)-1 (34), which is a potent
chemoattractant for T cells (35) and also for these T cell
hybridomas. SDF-1 is present in many noninflamed tissues and may
therefore be involved in the migration of the T cell hybridoma cells
into these tissues. We found that LFA-1 is required for migration
through ICAM-1-coated filters induced by a low concentration (1 ng/ml)
of SDF-1, whereas migration induced by a high SDF-1 concentration (100
ng/ml) was independent of LFA-1. ZAP-70 was essential for the
LFA-1-dependent chemotaxis but not when LFA-1 was not required.
Furthermore, we showed that ZAP-70 is involved in the tyrosine
phosphorylation induced by LFA-1 engagement. Thus, the combination of
low SDF-1 levels and an ICAM-1-coated filter appears to provide a
suitable model for the in vivo migration of the T cell hybridoma,
because both the in vivo and in vitro migration require the activity of
the pertussis toxin-sensitive Gi proteins, LFA-1
and ZAP-70. The results suggested that the signal, initiated by the few
SDF-1 receptors triggered at low SDF-1 levels, activates a limited
number of LFA-1 molecules, which can then bind to ICAM-1. This then
leads to ZAP-70 activation and amplification of the signal. We show
here that ZAP-70 activity triggered by LFA-1 engagement leads to
adhesion mediated by unengaged LFA-1 molecules. This combination of
outside-in and inside-out signals may be responsible for the
propagation and amplification of the chemokine signal.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The Syk/ZAP-70-specific inhibitor piceatannol was obtained from Boehringer Mannheim (Mannheim, Germany); and the protein kinase C (PKC) inhibitor Ro 31-8220 and the calpain inhibitor calpeptin were from Calbiochem (La Jolla, CA). The phospholipase C (PLC) inhibitor U-73122 (1-{{6-{[(17ß-3-methoxestra-1,3,5(10)trien-17-yl]amino}hexyl}}-1H-pyrrole-2, 5-dione) and its inactive congener, U-73343 (1-{{6-{[17ß-3-methoxestra-1,3,5(10)trien-17-yl]amino}hexyl}}-2,5-pyrrolidine-dione) were from Biomol (Plymouth, MA). Both U-73122 and U-73343 were dissolved in chloroform and dried under a nitrogen stream. Just before use, the dried film was dissolved in DMSO to a concentration of 5 mM. Ro 31-8220 was dissolved in DMSO to 1 mM, aliquoted, and kept at -80°C. Calpeptin was dissolved in DMSO at 10 mg/ml and kept at -20°C. Piceatannol was dissolved in 30% DMSO at 2.4 mg/ml and kept at -20°C.
Cell lines and culture conditions
Murine TAM2D2 T cell hybridoma cells (8) and rat embryo fibroblasts were cultured as described previously (36).
Antibodies
Hybridomas producing rat anti-murine CD11a mAb M17/4,
anti-ICAM-1 mAb YN1/1.7, and CD44 mAb KM201 were obtained from the
American Type Culture Collection (Manassas, VA). Two new rat
anti-mouse ß2 (CD18) hybridomas, GAME-46 (IgG1) and -245 (IgG2a),
were generated previously by us using purified LFA-1 as immunogen
(37). The GoH3 mAb against CD49f, the
6ß1 integrin
subunit, was kindly donated by A. Sonnenberg. Rat anti-mouse ICAM-2
mAb MIC2/4 (3C4) was purchased from PharMingen (San Diego, CA), and rat
IgG was purchased from Nordic Immunology (Tilburg, The Netherlands).
All mAbs were affinity purified. Fab fragments were generated using
immobilized papain according to the manufacturer’s protocol (Pierce,
Rockford, IL). F(ab')2 fragments were generated
with pepsin and concentrated using a Centricon 100 (Millipore, Bedford,
MA) to remove the Fc parts. The mAb 2F3.2 against human ZAP-70
(38) was kindly supplied by Dr. A. Weiss.
Aggregation assay
To induce aggregation, M17/4 mAb was added to TAM2D2 cells (2 x 106/ml in HBSS, pH 7.0, supplemented with 20 mM HEPES, 0.35 g/l NaHCO3, 1 mM CaCl2, and 1 mM MgCl2) to a final concentration of 0.2 µg/ml or as indicated. The total volume was 0.5 ml in 10-ml tubes. Alternatively, cells were incubated for 20 min on ice with different concentrations of M17/4 Fab fragments as indicated. Subsequently, rabbit anti-rat Ig Abs (RARa/7S, dilution: 1/500; Nordic Immunology) were added. The cells were incubated in a waterbath at 37°C for 2 h in an upright position and shaken at low speed. By mild agitation with a wide-bore pipette, the suspensions were then dispersed, and 100-µl samples were transferred to a flat-bottom dish and photographed using an inverted microscope. To test the effects of Abs or inhibitors, the cells were preincubated with the mAbs for 15 min or with inhibitors for 30 min at room temperature before addition of the M17/4 mAb.
Determination of number of free LFA-1 molecules by FACScan analysis
TAM2D2 cells were incubated with the M17/4 mAb or the M17/4 F(ab')2 or Fab fragments at different concentrations, ranging from 0.02 to 10 µg/ml. After 2 h of incubation, aggregation was scored as described above. A sample of the cells (2 x 105) was removed, vigorously pipetted to disperse the aggregates, washed with FACS buffer (PBS with 0.5% BSA and 0.02% azide), and subsequently incubated with FITC-labeled M17/4 (10 µg/ml) for 30 min at 4°C. The sample was washed twice and fluorescence was measured on a FACScan (Becton Dickinson, San Jose, CA) using the lysis II program. Cells that had not been exposed to Abs were used as a negative control. The median fluorescence of cells that had not been preincubated with unlabeled M17/4 mAb at all but only with the FITC-labeled mAb minus the median background fluorescence of control cells (F) was set at 100%. The median fluorescence of the samples minus the background fluorescence as a percentage of F is a measure of the percentage of LFA-1 molecules that was not occupied by the nonlabeled M17/4 mAb.
Generation and transduction of DNA constructs
Both the construction and the transduction of constructs have been described previously (32). Briefly, the cDNAs encoding the truncated human ZAP-701–276(1–276) (38) or the full-length wild-type (WT) ZAP-70 protein, cloned in the retroviral vector pMFG-IRES-geo (geo = fusion protein of lacZ and neoR), were transfected into the BOSC23 packaging cell line. TAM2D2 cells were infected with the retrovirus by coculture or incubation with BOSC23 supernatant, and clones with homogeneous high and stable lacZ expression were selected by subcloning and FACS sorting. The internal ribosome entry site (IRES) in the vector allows the translation of both the ZAP-70 and geo protein from one bicistronic mRNA (39). High lacZ activity correlated with high expression of ZAP-70, which was at least 10-fold higher than endogenous ZAP-70 in the clones used in this study, DN22 and DN38. As controls, we generated cells expressing similarly high levels of the WT full-length ZAP-70 protein. Truncated ZAP-70 constructs with a point mutation in either the N-terminal Src-homology domain 2 (SH2) (R37K) or the C-terminal SH2 domain (R190K) (40) were transduced similarly. The N-terminal SH2 domain mutant was generated by PCR, and the C-terminal SH2 mutant was kindly supplied by Dr. A. Weiss. The presence of the mutations was verified by sequencing.
Immunoblotting
SDS-PAGE-separated cell lysates were blotted to nitrocellulose, which was then blocked with 3% BSA and 0.4% Tween 20. The membranes were incubated for 1 h with the mAb 2F3.2 against human ZAP-70 at room temperature, followed by incubation with sheep anti-mouse HRP-coupled Ig (Amersham, Little Chalfont, U.K.). Stained proteins were visualized by chemiluminescence.
Adhesion to anti-LFA-1 mAb-coated substrates
Wells in 96-well plates were coated with 100 µl of Ab solution (5 µg/ml) overnight at 4°C, followed by blocking for 2 h at room temperature with 0.5% OVA (Sigma, St. Louis, MO). Cells were allowed to adhere for 10 min in 20 mM Tris buffer, pH 7.2 (containing 150 mM NaCl, 6 mM KCl, 1 mM CaCl2, 1 mM MgCl2, and 5 mM D-glucose). For rapid warming up of the plate, it was placed on a prewarmed metal block in an incubator at 37°C, 5% CO2. Nonadherent cells were washed off, and the number of adherent cells was determined by assaying hexosaminidase activity using known numbers of cells as standard essentially as described by Landegren et al. (41).
Invasion and migration assays
Invasion assays were performed as described (36).
Briefly, TAM2D2 cells or TAM2D2 transfectants were added to confluent
rat embryo fibroblast monolayers in serum-free medium. After 1 h
at 37°C and 5% CO2, the monolayers were
extensively washed and then paraformaldehyde-fixed. The invaded cells
were counted using phase-contrast microscopy, and the percentage of
invaded cells was calculated. Migration was assayed in Transwells, as
described (32). Briefly, Transwells with 8-µm pore
filters, coated with 1 µg/ml recombinant monomeric truncated soluble
mouse ICAM-1 (without transmembrane and cytoplasmic domains) were used.
The lower chamber was filled with 250 µl RPMI 1640 containing 0.1%
OVA and either 1 or 100 ng/ml SDF-1 (Pepro Tech, Rocky Hill, NJ).
The Transwell was placed on top, and 150 µl medium with
105 cells was inserted into the upper chamber.
The data presented are the percentages of added cells that have been
collected from the lower chamber after 2 h at 37°C, 5%
CO2.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TAM2D2 T cell hybridoma cells express the adhesion molecules
LFA-1, 6ß1, CD44, and
ICAM-2 but not ICAM-1 (9). LFA-1 is the only
ß2 integrin present. We found that the M17/4
mAb against the subunit of LFA-1 induced aggregation of TAM2D2
cells. Aggregation was noticeable within 30 min, but the aggregates
reached their maximum size usually after 2 h of incubation with
0.2 µg/ml M17/4 mAb. To determine which adhesion molecules were
involved, we preincubated the cells with inhibitory mAbs against
ICAM-2, the 6 integrin subunit, and CD44, as
well as rat IgG as a negative control. The anti-murine ICAM-2 mAb
MIC2/4 prevented aggregation almost completely (Fig. 1
), whereas rat IgG and the blocking mAbs
GoH3 against 6 (Fig. 1) and KM201 against CD44
(data not shown) had no effect. This shows that aggregation is mediated
by the binding of LFA-1 to ICAM-2 and is not caused by agglutination of
the cells by the M17/4 mAb. Indeed, two anti-ß2 (CD18) mAbs
generated by us, GAME-46 and -245 (37), blocked
aggregation completely (Fig. 1 and data not shown). By using flow
cytometry with FITC-labeled M17/4 mAb, we established that the GAME-46
and -245 mAb did not inhibit binding of M17/4 to LFA-1 (data not
shown), so that inhibition is not due to interference with the binding
of M17/4 to LFA-1.
|
The induction of LFA-1/ICAM-2-dependent aggregation by the M17/4
Ab was surprising because M17/4 is a function-blocking mAb. We found,
however, that aggregation only occurred at M17/4 concentrations lower
than the saturating concentration of 10 µg/ml and most optimally at
0.2–0.5 µg/ml. To determine the number of free LFA-1 molecules at
this subsaturating concentration, we used FACS analysis on cells
treated with different concentrations of unlabeled M17/4, and
subsequently, after 2 h, with 10 µg/ml FITC-labeled M17/4. The
results are shown in Fig. 2A.
We found that 
25% of LFA-1 molecules were occupied by M17/4 mAb at
0.2 µg/ml and 50% at 0.5 µg/ml. Aggregation was reduced at 1
µg/ml, with 80% of LFA-1 occupied. F(ab')2
fragments induced aggregation at similar concentrations (data not
shown), whereas Fab fragments did not induce aggregation at similar
occupation percentages (determined similarly, see Fig. 2B).
We conclude that signals triggered by cross-linking of LFA-1 molecules
by the M17/4 Ab cause the unoccupied LFA-1 molecules to bind to ICAM-2,
leading to aggregation of the cells.
|
40%), but between 50 and 80% occupation aggregation
was more extensive, with larger aggregates (Fig. 2
B). This
may be due to more efficient cross-linking. Indeed, aggregation by
intact M17/4 was also enhanced by anti-rat Ig Abs: it started at
slightly lower M17/4 concentrations and the aggregates were very large
at the optimal concentration (Fig. 2A). Incubation with
anti-rat Ig Abs alone did not induce aggregation (data not
shown). Aggregation induced by LFA-1 engagement is blocked by dominant-negative ZAP-70
To study the signal pathways involved in the M17/4-induced LFA-1
activation, we tested several inhibitors. Piceatannol, an inhibitor of
the Syk tyrosine kinase (42), blocked aggregation
completely (Table I). We have shown
previously that the Syk homologue ZAP-70, but not Syk, is expressed by
TAM2D2 cells. Furthermore, we showed that piceatannol inhibits ZAP-70
activity with the same dose dependence as reported for Syk with a
complete block at 50 µg/ml (32).
|
and Fig. 4), except for the occasional formation of a few small cell
clumps. Cross-linking of Fab fragments with anti-Ig Abs did not
induce aggregation of these cells either (data not shown). In contrast,
WT cells expressing similarly high levels of the full-length WT human
ZAP-70 protein (32), as well as cells transduced with
empty vector, aggregated as extensively as the parental TAM2D2 cells
(Table I). Furthermore, we recently generated cells that express a
truncated ZAP-70 with a point mutation in either the C-terminal SH2
domain (R190K) or the N-terminal SH2 domain (R37K) at similar levels as
the nonmutated truncated ZAP-70 (Fig. 3).
Two independently generated clones of each (C-SH2.7, C-SH2.16 and
N-SH2.10, N-SH2.15) were examined. Aggregation of these cells was
comparable to that of TAM2D2 cells (Table I and Fig. 4), showing that both SH2 domains are
required for the dominant-negative effect and therefore both are
essential for the participation of ZAP-70 in LFA-1 function. The
ZAP-70-transduced cells, including WT cells that overexpress the intact
ZAP-70, did not aggregate spontaneously.
|
|
A potential downstream effector of ZAP-70 is PLC-
. To
investigate its possible involvement, we used the PLC inhibitor
U-73122. This inhibited aggregation completely, whereas the inactive
structural analogue U-73343 had no effect (Table I). PLC generates two
products: diacylglycerol, which activates PKC, and
inositol-trisphosphate which releases Ca2+. The
PKC inhibitor Ro 31-8220 did not affect aggregation (Table I), and PKC
is therefore probably not involved. One of the possible effectors
activated by Ca2+ is the calcium-dependent
protease calpain, which was recently proposed to be involved in LFA-1
activation (43). In agreement with this, the calpain
inhibitor calpeptin inhibited aggregation completely (Table I).
Adhesion to immobilized LFA-1 mAb: roles of ZAP-70, PLC, and calpain
To further study the effects of signals induced by the binding of
the M17/4 anti-LFA-1 mAb, we allowed cells to adhere to an
M17/4-coated substrate. This adhesion is mediated by LFA-1, because
preincubation of cells with M17/4 Abs prevented adhesion completely,
whereas control Abs did not (data not shown). TAM2D2 cells readily
adhered to these surfaces (Fig. 5). In
contrast, the DN22 and DN38 clones, overexpressing dominant-negative
ZAP-70, did not adhere at all, i.e., did not adhere strongly enough to
resist the shear forces of the washing steps. Furthermore, piceatannol
completely blocked adhesion of TAM2D2 cells (Fig. 5A). In
contrast, truncated ZAP-70 with a mutation in either of the two SH2
domains had no effect at all (Fig. 5B). In addition,
adhesion was completely blocked by U-73122 but not U-73343 and
calpeptin but not Ro 31-8220 (Fig. 5C).
|
In addition to M17/4, which reacts with the L subunit, the
blocking GAME-46 Ab against the ß2 subunit induced aggregation at
subsaturating concentrations (data not shown). In contrast, the
nonblocking M18/2 Ab against the ß2 subunit did not. Indeed, TAM2D2 T
cell hybridoma cells adhered extensively to and spread on the
immobilized blocking Abs M17/4 and GAME-46, but not the nonblocking Ab
M18/2 (Figs. 6 and
7). This was not due to a difference in
affinity, which was similar for all three Abs. As determined by FACS
analysis (similarly as in Fig. 2), half-maximal binding occurred at
0.5 µg/ml (M17/4, 0.48 µg/ml; GAME-46, 0.56 µg/ml; and M18/2,
0.54 µg/ml).
|
Invasiveness of cells was tested in rat embryo fibroblast cultures
(36). TAM2D2 cells invaded the monolayers within 1 h.
As shown previously (32), invasion was completely
inhibited by piceatannol at 50 µg/ml, and both DN22 and DN38 cells
expressing truncated dominant-negative ZAP-70 had lost invasive
capacity (Fig. 8A). Here we
show that the cells expressing the truncated ZAP-70 with a mutation in
either of the two SH2 domains invaded normally (Fig. 8A).
Furthermore, we show that invasion is blocked by U-73122 but not
U-73343 and by calpeptin but not by Ro 31-8220 (Fig. 8B).
|
Similarly as described before (32), 1 ng/ml SDF-1 induced
substantial LFA-1-dependent migration through ICAM-1-coated filters.
This was completely blocked by 50 µg/ml piceatannol, and
dominant-negative ZAP-70-expressing DN22 and DN38 cells did not migrate
at all (Fig. 9A). The
truncated ZAP-70 SH2 domain mutants did not affect migration (Fig. 9B). Furthermore, migration was blocked by U-73122 but not
U-73343 and by calpeptin but not by Ro 31-8220 (Fig. 9C). In
contrast, migration induced by 100 ng/ml SDF-1, which was completely
blocked by pertussis toxin, did not depend on LFA-1 because it was not
inhibited by LFA-1 mAb and occurred to the same extent through
ICAM-1-coated and uncoated filters (32). This migration
was not at all affected by piceatannol or by the truncated ZAP-70 (Fig. 9A). However, U-73122 did inhibit this LFA-1-independent
migration induced by 100 ng/ml SDF-1 (Fig. 9C), suggesting
the involvement of PLC-ß, triggered by SDF-1 receptor-coupled
G-proteins. Furthermore, the calpain inhibitor calpeptin completely
blocked this migration (Fig. 9C), suggesting that calpain is
also required for LFA-1-independent migration.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(44) and the
calcium-dependent protease calpain, as suggested by the effects of
inhibitors. ZAP-70 is also required for LFA-1-dependent migration
through ICAM-1-coated filters and for LFA-1-dependent invasion into
fibroblast monolayers, and these processes are also blocked by PLC and
calpain inhibitors. This indicates that similar signals are triggered
by cross-linking of LFA-1 with blocking Abs as by binding of LFA-1 to
its ligand ICAM-1, in line with previous reports on LFA-1 as well as
other integrins (45, 46, 47). Thus, the outside-in signals
(48) induced by interaction of LFA-1 with ICAM-1 (or
ICAM-2) lead to inside-out signals that promote binding of other LFA-1
molecules to additional ligand molecules. This binding may lead to a
new cycle of outside-in and inside-out signals. This repetitive
process, illustrated in Fig. 10, seems
perfectly suitable for the propagation of signals triggered by the
sequential contacts of cell protrusions with ligand molecules when a
cell moves on a surface. However, because the T cell hybridoma cells do
not bind spontaneously to ICAM-1, a different signal is required to
initiate the process, which is provided by low concentrations of a
chemokine such as SDF-1. Migration induced by higher chemokine
concentrations does not require the LFA-1 signal and is not blocked by
dominant-negative ZAP-70. Therefore, this does not seem a relevant
model for in vivo migration, because both the in vitro invasion into
fibroblast monolayers and the in vivo invasion of many different
tissues by the T cell hybridoma cells requires LFA-1 and is completely
blocked by the dominant-negative ZAP-70.
|
Signals amplified via LFA-1 and ZAP-70 may include PLC and calpain
LFA-1-independent chemotaxis induced by high SDF-1 levels was
blocked by U-73122, which inhibits PLC (3), as well as by
calpeptin, which inhibits calpain. This suggests that PLC-ß activated
by the chemokine receptor is involved and that calpain activated by
calcium elicited by the PLC product inositol-trisphosphate is also
required. However, because these drugs are unlikely to be specific for
only these targets, the involvement of PLC and calpain will have to be
confirmed by more direct approaches. Ab-induced aggregation is also
blocked by U-73122 and calpeptin, suggesting that LFA-1 engagement
triggers in part similar signals that are thus amplified. A likely
possibility again is that PLC is involved, but rather than the PLC-ß
activated by the chemokine receptor, it would be PLC-, which is
activated by ZAP-70 (50), and by Ab engagement of LFA-1
(44). Thus, chemokines and integrins may collaborate in
providing sufficient PLC activity by activation of the two distinct
phospholipases.
Induction of LFA-1-mediated adhesion: possible mechanisms
Adhesion triggered by inside-out signals has been ascribed either
to changes in the conformation of integrins that increase the affinity
for ligand or to clustering of integrins that leads to increased
avidity (1, 16, 43, 48). However, the same inhibitors that
block LFA-1-dependent adhesion during invasion and migration
(dominant-negative ZAP-70, piceatannol, U-73122, and calpeptin) also
inhibit adhesion of T cell hybridoma cells to immobilized Abs. In the
latter case a change in affinity or avidity for the ligand, which is an
Ab, cannot explain the enhanced adhesion, because the Ab binds to
nonactivated LFA-1 with an affinity that is not increased when the
active conformation is induced by Mn2+ (data not
shown). In this respect, "activation" of LFA-1 by Ab is comparable
with activation by PMA, which does not involve a change in integrin
conformation either (51, 52). An alternative explanation,
as suggested above, is that clustering plus binding of a single ICAM-1
is sufficient to trigger the required signals. The role of calpain in
LFA-1 function has been described (43), and it was
proposed that calpain releases a cytoskeletal constraint and thus
facilitates the redistribution of LFA-1 molecules to the contact area.
Indeed, it was recently shown that activation of leukocytes leads to
proteolysis of talin that links the ß2-chain to the cytoskeleton and
this is blocked by calpeptin (53). It is therefore
conceivable that the outside-in signal leads to increased lateral
mobility and consequently to accumulation of LFA-1 molecules in the
area of contact with the Ab-coated substrate. It should be noted,
however, that calpeptin affects the migration induced by high SDF-1
concentrations as well, and that this is independent of LFA-1 (see Fig. 9C). If this is also due to calpain inhibition, calpain
should have other relevant effects than only the release of LFA-1
constraints. A potential alternative explanation is that the elicited
signals cause the cells to spread on the substrate (see Fig. 6). This
would increase the contact area and thereby the number of LFA-1-ligand
interactions. This possible explanation is supported by a recent report
on NIH 3T3 cells overexpressing the endogenous calpain inhibitor
calpastatin in which the capacity to spread was greatly reduced
(54). Other explanations are not excluded. For
instance, it is conceivable that the outside-in signals induce the
interaction of LFA-1 molecules with cytoskeletal or signaling
complexes.
T cell hybridoma cells do not adhere to immobilized nonblocking LFA-1 Ab
The cells adhered to the blocking Abs M17/4 and GAME-46 directed
against the murine LFA-1 L and
ß2 subunits, respectively. In contrast, the
cells did not bind to the nonblocking mAb M18/2, which has similar
affinity for the ß2-chain as GAME-46. Furthermore, both blocking Abs
induced aggregation, whereas M18/2 did not. This is in line with the
report by Miyamoto et al. (47) that cross-linking of the
integrin 5ß1 by
blocking Abs leads to formation of large signaling complexes, similar
to binding to the 5ß1
ligand fibronectin, whereas cross-linking by nonblocking Abs does not.
Blocking Abs usually bind at or near the ligand-binding site, although
they may also affect the binding site conformation allosterically. For
the Abs used in this study, the binding sites have not been mapped.
However, assuming that these are located near the binding site, it may
affect integrin conformation in a manner comparable with ligand
binding. This may be required, in addition to cross-linking, to mimic
the effect of ligand engagement. The unexpected complete lack of
adhesion to immobilized M18/2 may thus be explained by the inability of
this nonblocking Ab to trigger the appropriate outside-in signals.
ZAP-70 activation depends on SH2 domains but not on the TCR or CD3
The T cell hybridoma cells used in this study express ZAP-70 but
not the homologous Syk tyrosine kinase (32). Our
conclusion that ZAP-70 is involved in LFA-1-dependent migration,
invasion, and metastasis, as well as Ab-induced aggregation and
adhesion to immobilized LFA-1 Abs, is based on two considerations.
First, these processes are blocked by the inhibitor piceatannol with
the exact same dose dependence as inhibition of ZAP-70 activity as we
showed previously (32) and, second, these processes are
blocked in cells overexpressing a truncated dominant-negative ZAP-70.
This truncated protein consists mainly of the two SH2 domains.
Triggering of ZAP-70 by the ß2 integrin LFA-1
is in line with the previously reported activation of Syk upon
engagement of ß1 and ß3
integrins (55, 56, 57, 58). However, overexpression of a truncated
Syk containing the SH2 domains did not affect the integrin-induced
activation of Syk, whereas it blocked Fc-receptor-induced Syk
activation (59). This led to the conclusion that the SH2
domains are not required for the signals triggered by the
ß1 and ß3 integrins. In
contrast, we show here that inhibition of LFA-1-mediated processes by
the truncated ZAP-70 depends on both SH2 domains, because the
inhibitory activity was blocked by a mutation in either domain that
impairs the binding to tyrosine-phosphorylated proteins
(40). Individually, the two SH2 domains bind very poorly
to their tyrosine-phosphorylated target sequences and with distinct
preferences, and the affinity of the tandem SH2 domains for doubly
phosphorylated immunoreceptor tyrosine-based activation motif (ITAM)
sequences is 1000-fold higher (60, 61). Hence, it is
most likely that a tyrosine-phosphorylated protein, containing two
phosphotyrosines in an ITAM sequence comparable with those in the
TCR- and the CD3 chains (59), is involved in
LFA-1-induced ZAP-70 activity. This protein is distinct from the TCR
and CD3 chains, however, because these are not expressed by the T cell
hybridoma used in this study. This protein remains to be
identified.
The ITAMs in the TCR/CD3 complex are phosphorylated by Src family kinases (62). However, these kinases do not appear to be involved in invasion and migration of the T cell hybridoma because Src inhibitors such as herbimycin (32) and overexpression of a kinase-dead Lck mutant (R. Soede et al., unpublished observations) had no effect. This indicates that the ITAM involved is phosphorylated by a non-Src-like kinase. It is noteworthy that ZAP-70 becomes associated with the focal adhesion kinase (FAK) (3) upon integrin signaling in NK cells (63), and that in a complex containing paxillin, FAK, and ZAP-70, the latter two become highly phosphorylated upon chemokine signaling in T cells (64). Because the FAK homologue Pyk2 also binds paxillin (65), it is likely to form part of such complexes. How ZAP-70 binds to the complex is unknown but the presence of an ITAM-containing protein seems a likely option. If so, FAK and Pyk2 are likely candidates for the ITAM-phosphorylating kinase.
Conclusion
In conclusion, we have provided evidence that signals induced by
engagement of the integrin LFA-1 promote adhesion mediated by unengaged
LFA-1 molecules. The effector molecules triggered by these signals are
ZAP-70 and probably PLC- and calpain. This LFA-1 signaling cascade
may propagate and amplify the migration-initiating signals induced by
low chemokine levels. Gi proteins, LFA-1, and
ZAP-70 are required for T cell hybridoma invasiveness in vitro as well
as for dissemination of these cells in vivo, as we have shown
previously. These three components are also essential for
LFA-1-dependent migration initiated by low chemokine concentrations,
which apparently involves amplification of the chemokine signal. LFA-1
to LFA-1 signaling may therefore be important for T cell migration
in vivo.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Medical Research Council Laboratory for Molecular Biology, University College London, London, U.K.
3 Address correspondence and reprint requests to Dr. E. Roos, Division of Cell Biology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands. E-mail address:
4 Abbreviations used in this paper: ZAP, -associated protein; FAK, focal adhesion kinase; PKC, protein kinase C; PLC, phospholipase C; SDF, stromal cell-derived factor; WT, wild type; SH2, Src-homology domain 2; ITAM, immunoreceptor tyrosine-based activation motif.
Received for publication March 1, 1999. Accepted for publication August 6, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and MALA-2. Transplantation 52:842.[Medline]
and ß chains in human T-lymphocyte activation. Nature 333:850.[Medline]
-chain of human leukocyte function-associated antigen 1 induces homotypic cell-cell interactions. J. Immunol. 140:1393.[Abstract]
(CD11a): differential effects on resting and activated T cells. Eur. J. Immunol. 22:1851.[Medline]
5ß1 integrin activates the 2ß1 receptor via the 5 subunit cytoplasmic domain and protein kinase C. J. Immunol. 153:2222.[Abstract]
-associated protein 70 inhibits T cell antigen receptor signaling. J. Exp. Med. 183:611.
R1-mediated signaling and effector function by the Syk-selective inhibitor, piceatannol. J. Biol. Chem. 269:29697.1 activation. Proc. Natl. Acad. Sci. USA 90:7099.1 in an SLP-76-deficient T cell. Science 281:413.Lß2 (LFA-1, CD11a/CD18) cytoplasmic domain/cytoskeleton interaction. J. Exp. Med. 181:315.IIbß3. J. Biol. Chem. 269:28859.IIbß3. EMBO J. 16:6414.[Medline]
This article has been cited by other articles:
|
|
 |
|
  O. D. Perez, D. Mitchell, G. C. Jager, and G. P. Nolan LFA-1 signaling through p44/42 is coupled to perforin degranulation in CD56+CD8+ natural killer cells Blood, August 15, 2004; 104(4): 1083 - 1093. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Hogg, M. Laschinger, K. Giles, and A. McDowall T-cell integrins: more than just sticking points J. Cell Sci., December 1, 2003; 116(23): 4695 - 4705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Barat and M. J. Tremblay Treatment of Human T Cells with Bisperoxovanadium Phosphotyrosyl Phosphatase Inhibitors Leads to Activation of Cyclooxygenase-2 Gene J. Biol. Chem., February 21, 2003; 278(9): 6992 - 7000. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Katakai, T. Hara, M. Sugai, H. Gonda, Y. Nambu, E. Matsuda, Y. Agata, and A. Shimizu Chemokine-independent Preference for T-helper-1 Cells in Transendothelial Migration J. Biol. Chem., December 20, 2002; 277(52): 50948 - 50958. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ticchioni, C. Charvet, N. Noraz, L. Lamy, M. Steinberg, A. Bernard, and M. Deckert Signaling through ZAP-70 is required for CXCL12-mediated T-cell transendothelial migration Blood, May 1, 2002; 99(9): 3111 - 3118. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Sirim, L. Zeitlmann, B. Kellersch, C. S. Falk, D. J. Schendel, and W. Kolanus Calcium Signaling through the beta 2-Cytoplasmic Domain of LFA-1 Requires Intracellular Elements of the T Cell Receptor Complex J. Biol. Chem., November 9, 2001; 276(46): 42945 - 42956. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L. Woods and Y. Shimizu Signaling networks regulating {beta}1 integrin-mediated adhesion of T lymphocytes to extracellular matrix J. Leukoc. Biol., June 1, 2001; 69(6): 874 - 880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. M. Soede, I. S. Zeelenberg, Y. M. Wijnands, M. Kamp, and E. Roos Stromal Cell-Derived Factor-1-Induced LFA-1 Activation During In Vivo Migration of T Cell Hybridoma Cells Requires Gq/11, RhoA, and Myosin, as well as Gi and Cdc42 J. Immunol., April 1, 2001; 166(7): 4293 - 4301. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
