| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Immunology Program, H. Lee Moffitt Cancer Center, Department of Interdisciplinary Oncology, University of South Florida College of Medicine, Tampa, FL 33612;
Center d’Immunologie Institut National de la Santé et de la Recherche Médicale/Centre National de la Recherche Scientifique de Marseille-Luminy, Marseille, France; and
Biosciences Division, Los Alamos National Laboratory, Los Alamos, NM 87545
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, and Fc
RI
adaptors that contain
immunoreceptor tyrosine-based activatory motifs to mediate NK direct
lysis of tumor cells via Syk tyrosine kinase. NK cells may also use
DAP10 to drive natural cytotoxicity through phosphoinositide 3-kinase
(PI3K). In contrast to our recently identified PI3K pathway controlling
NK cytotoxicity, the signaling mechanism by which Syk associates with
downstream effectors to drive NK lytic function has not been clearly
defined. In NK92 cells, which express DAP12 but little DAP10/NKG2D, we
now show that Syk acts upstream of PI3K, subsequently leading to the
specific signaling of the PI3K
Rac1PAK1mitogen-activated
protein/extracellular signal-regulated kinase (ERK) kinaseERK
cascade that we earlier described. Tumor cell ligation stimulated DAP12
tyrosine phosphorylation and its association with Syk in NK92 cells;
Syk tyrosine phosphorylation and activation were also observed.
Inhibition of Syk function by kinase-deficient Syk or piceatannol
blocked target cell-induced PI3K, Rac1, PAK1, mitogen-activated
protein/ERK kinase, and ERK activation, perforin movement, as well as
NK cytotoxicity, indicating that Syk is upstream of all these signaling
events. Confirming that Syk does not act downstream of PI3K,
constitutively active PI3K reactivated all the downstream effectors as
well as NK cytotoxicity suppressed in Syk-impaired NK cells. Our
results are the first report documenting the instrumental role of Syk
in control of PI3K-dependent natural
cytotoxicity. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human NK cells express at least three families of receptors, killer-inhibitory receptors (KIR), NKG2, and natural cytotoxicity receptors (NCR), to mediate cytotoxicity. At present, the NCR family includes NKp30, NKp44, NKp46, and NKp80, which are, to date, known to be expressed only on resting or activated NK cells (9, 10, 11, 12). Their ligands are yet unknown, except for NKp44 and NKp46, which appears to recognize viral hemagglutinin and hemagglutinin-neuraminidase (13). The KIR family constitutes MHC class I receptors that either inhibit or activate NK cytotoxicity and are shared with T cells (14, 15). Those KIR isoforms that contain immunoreceptor tyrosine-based inhibitory motifs can recruit and activate protein tyrosine phosphatases, leading to the down-regulation of NK cytotoxicity (16, 17, 18). In contrast, the KIR isoforms lacking immunoreceptor tyrosine-based inhibitory motifs serve as activatory receptors that recognize specific classical MHC class I molecules on target cells (19, 20, 21). The NKG2 family represents C-type lectin-like receptors and also exists in inhibitory and activatory isoforms that require association with a common CD94 chain for surface expression (22, 23). One exception is NKG2D, which does not associate with CD94, and its ligands are diverse, including stress and viral-induced nonclassical MHC class I molecules, MICA/B and ULBP, as well as HL60 and retinoic acid-inducible early gene-1 like proteins (24, 25, 26, 27). Target cell lysis represents the final balance between positive signals and negative messages from these receptors (14).
NK-activatory receptors do not possess intrinsic kinase property, but
use transmembrane adaptor proteins to transduce signals leading to
lytic function. DAP10 is one such adaptor that associates with NKG2D
and recruits the p85 subunit of PI3K, providing for NKG2D-dependent
signal transduction (22). Although direct evidence is not
yet available, it is likely that DAP10 uses the specific
PI3KRac1PAK1MEKERK signaling cascade that we have described
earlier to mediate NK lysis (7). Upon target ligation, we
discovered that PI3K is rapidly activated in NK cells, triggering a
sequential activation of Rac1, PAK1, MEK, and ERK. The result of this
signal cascade is the mobilization of lytic granules containing
perforin and granzyme B toward the ligated tumor cell, ending in target
lysis. Another key adaptor protein for NK-activatory receptors,
KARAP/DAP12, is a disulfide-bonded homodimer that associates with
several NK-activating receptors, KIR2DS, NKG2C, and NKp44, to mediate
tumoricidal function (10, 28, 29, 30). Upon triggering of
these NK-activating receptors, the tyrosine residues in the
immunoreceptor tyrosine-based activatory motif (ITAM) of DAP12
cytoplasmic domain become phosphorylated and recruit and activate
Syk/Zap70, which then leads to NK activation (10, 28, 29, 30, 31).
Although numerous molecules have been described to be activated in NK
cells along with Syk, there has been no systematic analysis of the
signaling cascade governed by Syk that modulates target lysis. We thus
investigated the downstream signaling events of Syk in NK cells and
took advantage of the NK92 cell line, which expresses high levels of
DAP12, but extremely low levels of NKG2D and DAP10 (L. L. Lanier
(University of California, San Francisco, CA), unpublished
observations).
Using an approach similar to that which we used to unravel the signal events linked to the PI3K-dependent NK lytic process, we found that Syk was coupled to PI3K. The use of recombinant vaccinia virus-mediated gene transfer of kinase-deficient Syk (SykT) (32), constitutively active catalytic subunit of PI3K (Myc-P110*) (33), constitutively active Rac1 (7), in combination with the specific pharmaceutical inhibitors for Syk, PI3K, and MEK, allowed us to carefully delineate the exact sequence of the signaling events. Our results clearly demonstrate that Syk tyrosine kinase critically controls target cell-induced activation of PI3K and its downstream signaling cascade, which ends in the mobilization and polarization of lytic granules toward the engaged tumor cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mouse mAbs to Syk, PI3K, and phosphotyrosine (clone 4G10), as well as Rac activation assay kit, were purchased from Upstate Biotechnology (Lake Placid, NY). Monoclonal anti-human perforin was purchased from Endogen/T Cell Sciences (Woburn, MA). Rabbit anti-phospho-ERK (Thr202/Tyr204) and anti-phospho-MEK (Ser217/221) were obtained from Cell Signaling Technology (Beverly, MA). Mouse anti-pan-ERK and anti-pan-MEK mAbs were obtained from Transduction Laboratories (San Diego, CA). Rabbit Abs to ERK1, ERK2, MEK1, MEK2, PAK1, Syk, P85, and P110 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit antisera to KARAP/DAP12 have been described previously (34). L-phosphatidylinositol-4,5-bisphosphate (PI(4, 5)P2) was from Biomol Research Laboratories (Plymouth Meeting, PA). Piceatannol, PD98059, and LY294002 were purchased from Calbiochem (La Jolla, CA).
The culture of NK92 cells as well as Raji and K562 tumor cells has been described previously (7). Large granular lymphocytes (LGL) possessing high NK activity were freshly isolated from normal donor blood by Percoll gradient centrifugation, as previously described, and cultured in IL-2-containing medium for 60–72 h before their in vitro assays (5).
Cytotoxicity assay
A 4-h 51Cr release assay was performed
using Raji and K562 tumor cells as targets for the NK92 cell line and
peripheral blood NK-LGL effector cells, respectively (5).
The percentage of specific 51Cr release was
determined by the following equation: ((experimental cpm -
spontaneous cpm)/total cpm incorporated) x 100. All
determinations were done in triplicate, and the SEM of all assays was
calculated and was typically 
5% of the mean or less.
Immunostaining
NK92 cells, untreated or treated with 50 µM piceatannol, 50 µM PD098059, 50 µM LY294002, or DMSO for 30 min at 37°C, were added to Raji cells. The cells were spun at 1000 rpm and incubated for 0 or 10 min at 37°C, then centrifuged onto a microscope slide and fixed with methanol/acetone (3/1) for 20 min (7). Cell-Tracker-Orange (Molecular Probes, Eugene, OR) was used to differentiate Raji lymphocytic cells from NK92 cells. Monoclonal anti-human perforin was used to detect the mobilization of lytic granules inside NK92 cells. Antiperforin, diluted 1/200, was applied to the slides. The slides were then washed in PBS and incubated with goat anti-mouse Ig FITC (Sigma-Aldrich, St. Louis, MO). After washing in PBS, the slides were covered with coverslips in mounting medium of antifade and 4',6'-diamidino-2-phenylindole. Immunofluorescence was observed with a Leitz Orthoplan 2 microscope, and images were captured by a charge-coupled device camera with the Smart Capture Program (Vysis, Downers Grove, IL). On each slide, 100 NK92 and Raji conjugates were evaluated for perforin mobilization and were performed in a blind fashion. Necessary controls were performed; i.e., NK92 cells alone or Raji tumor cells alone stained only with FITC-labeled goat anti-mouse Ig to check for nonspecific binding of the secondary Abs. Nonspecific binding was not detected, and the results were omitted from the figures for clarity.
Vaccinia virus construction and gene delivery
The plasmid containing P110* mutant, which is a Myc-tagged constitutively active component of PI3K, was kindly provided by A. Klippel (Chiron, Emeryville, CA) (33), and recombinant vaccinia virus encoding Myc-P110* was constructed using pSC11 vector, in combination with the WR strain of vaccinia. Vaccinia virus containing constitutively active Rac1 (V12Rac1) has been described previously (7). Vaccinia virus encoding kinase-deficient SykT, which is a truncated form of Syk with a lower molecular mass of 50 kDa, was kindly provided by A. M. Scharenberg (Harvard Medical School, Boston, MA) (32).
NK92 cells were incubated with recombinant vaccinia viruses encoding kinase-deficient SykT, or constitutively active P110*, or V12Rac1 for 1.5 h at 37°C at a multiplicity of infection of 5 (7). CD56-expressing vaccinia viral vector was used as a control. The NK cells were then washed three times and starved in serum-free medium containing 0.5% BSA for another 2.5 h before tumoricidal assay, Western blotting, and in vitro kinase assays. For those experiments combining pharmaceutical inhibitors and vaccinia virus infection, NK92 cells were pretreated for 30 min at 37°C with 50 µM piceatannol, 50 µM LY294002, or 50 µM PD98059 before virus infection, as described above, except that the inhibitors were added back for the last 2.5 h of infection before the evaluation of tumoricidal and effector function.
Detection of protein phosphorylation and protein kinase activities
Whole cell lysates were prepared from IL-2-rested NK cells treated by the pharmacological reagents or infected with the relevant recombinant vaccinia virus, and then challenged with paraformaldehyde-fixed target tumor cells for 0–30 min at 37°C. The phosphorylated forms of ERK and MEK were detected with respective phosphospecific Abs by Western blotting analysis. The immunoprecipitation of DAP12, Syk, PI3K, PAK1, MEK, and ERK, and in vitro kinase assays using myelin basic protein (MBP) as the substrate were conducted as previously described (7, 34). After SDS-PAGE, the phosphorylated forms of DAP12, Syk, P85, and P110 were detected with anti-phosphotyrosine Ab, 4G10. The in vitro kinase reaction was stopped by adding protein-loading buffer, and the mixture was separated by gel electrophoresis. The relative amounts of incorporated radioactivity were determined by autoradiography and quantitated with a PhosphorImager (Molecular Dynamics, Piscataway, NJ). Equal loading was measured by reblotting the same filter with anti-Syk, anti-P85, anti-P110, anti-pan-MEK, or anti-pan-ERK. In Western blotting, fixed Raji cells did not contribute to any of the detected kinases because paraformaldehyde fixation kept the dead target cells intact without release of intracellular components and prevented detergent lysis of these target cells.
Detection of PI3K phosphatidylinositol kinase activities
NK cells were lysed and immunoprecipitated with rabbit
polyclonal anti-p85 Ab at 4°C overnight, then incubated with
protein A agarose beads for 2 h at 4°C. The immunoprecipitates
were washed four times with ice-cold lysis buffer, then twice with
kinase reaction buffer (35). The activity of PI3K in
immunoprecipitates was analyzed by mixing the beads with reaction
buffer containing 100 µM ATP, 10 µCi
[-32P]ATP, and 20 µg
L-phosphatidylinositol-4,5-bisphosphate
(PI(4, 5)P2), 15 mM HEPES (pH 7.4), and 15 mM
MgCl2, for 25 min at 25°C. The reactions were
stopped by adding 150 µl 1 M HCl. The reaction mixture was extracted
with CHCl3-MeOH. The phosphorylated inositol was
differentiated by thin-layer chromatography (35), and the
conversion of PI(4, 5)P2 to
L-phosphatidylinositol-3,4,5-trisphosphate was
determined by autoradiography and quantitated with a PhosphorImager
(Molecular Dynamics).
Detection of Rac1 activity by PAK1 PBD affinity assay
NK92 cells were lysed with Mg2+ lysis/wash buffer provided in the Rac activation commercial kit (Upstate Biotechnology). The active form of Rac1, Rac1-GTP, was pulled down or affinity precipitated by incubation with 10 µg PAK1 PBD agarose from the whole cell lysates, washed three times with wash buffer, then subjected to 12.5% SDS-PAGE. The activated Rac1, Rac1-GTP that bound with PAK1 PBD, was examined by Western blotting with monoclonal anti-Rac provided in the kit and detected by ECL according to the protocol included in the kit.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To analyze Syk-mediated signaling events that lead to NK
tumor lysis, we used the DAP12-expressing NK92 cell line. We first
tested whether DAP12 was phosphorylated in NK92 cells upon ligation
with Raji tumor cells and whether phosphorylated DAP12 could bind Syk,
as has been reported by others (29, 30, 31). DAP12
immunoprecipitates from NK92 cells showed rapid tyrosine
phosphorylation within 5 min of contact with tumor cells (Fig. 1
A). To examine whether DAP12
phosphorylation leads to DAP12 association with Syk, we analyzed
Syk-DAP12 interaction by coimmunoprecipitation and Western blotting.
Syk immunoprecipitates did not show DAP12 association at 0 min, but
contained significant levels of DAP12 after 5 min of target ligation
(Fig. 1B). Moreover, upon binding to tumor cells,
significant tyrosine phosphorylation of Syk appeared in NK92 cells at 2
min, continued to 5 min, and dropped back to the basal level at 15 min
(Fig. 1C). Thus, Syk association with DAP12 was accompanied
by Syk tyrosine phosphorylation.
|
D). In contrast, the DMSO control solvent or the
expression of CD56, an irrelevant gene, had no deleterious effect on NK
cell function (Fig. 1D). An in vitro kinase assay with Syk
immunoprecipitates using MBP as the substrate ensured that piceatannol
treatment and SykT expression effectively inhibited Syk activation
triggered by target engagement in the same NK92 cells (Fig. 1, E and F). It should be noted that virally
delivered mutant SykT appears at a much lower molecular mass (50
kDa) than endogenous Syk (which is 72 kDa) because of its truncated
nature lacking the kinase domain (Fig. 1F, top
panel, last two lanes). Treatment with piceatannol and
SykT expression had no adverse effect on NK92 cell viability, as
measured by annexin V apoptosis assay and by trypan blue exclusion
(data not shown). Blockade of NK lytic granule movement and polarization by inhibiting Syk function
We next examined whether the inhibition of Syk could hinder the
movement of perforin toward the conjugated target cells. Immunostaining
with FITC-conjugated antiperforin indicated that, at 0 min of NK
binding with Cell-Tracker-Orange-labeled Raji cells, perforin was
evenly distributed in the cytoplasm of NK92 cells (Fig. 2A). At 10 min, perforin
granules were rapidly redistributed and polarized toward the contact
site with the conjugated target cells (Fig. 2A, top
panels), but this movement was blocked in piceatannol-treated NK92
cells (Fig. 2A, bottom panels). Thus, piceatannol
had the same inhibitory effects on perforin mobilization as LY294002,
which inactivates PI3K, or PD98059, which inhibits ERK, as we have
reported earlier (7). We thus asked whether Syk might
function in the same track as the earlier identified
PI3KRac1PAK1MEKERK pathway (7), or follow a
separate signaling cascade to mediate NK cytotoxicity.
|
Our earlier studies indicated that ERK was critical in the control
of lytic granule movement in NK cells (5, 8). We thus
examined whether Syk inactivation could interfere with target
cell-triggered ERK activation in NK cells. MEK and ERK phosphorylation
and activation occurred 5 min following target ligation, confirming our
earlier report (8); however, such activation was blocked
in piceatannol-treated NK92 cells (Fig. 2, B and
C, left panels). To exclude the possibility that
piceatannol might nonspecifically affect other signal transducers,
kinase-deficient SykT was used instead of piceatannol to down-regulate
Syk function. SykT expression in NK92 cells significantly suppressed
the appearance of MEK and ERK phosphorylation (Fig. 2B,
right panels) as well as kinase function (Fig. 2C, right panels), while control CD56 expression
did not. These results therefore indicate that Syk functions upstream
of MEK and ERK to regulate NK cytotoxicity.
Inhibition of PI3K activation triggered by target cell ligation by SykT or piceatannol
To query whether Syk also couples to the PI3K signaling that
specifically controls target cell-induced MEK and ERK activation
(7), we first examined tyrosine phosphorylation of P110
and P85, the catalytic and regulatory subunits of PI3K, which is a
necessary prerequisite for PI3K function (Fig. 3A). Both subunits were
phosphorylated upon 5 min of target ligation, but this effect was lost
by piceatannol treatment or SykT expression, implicating the
involvement of Syk in target cell-induced PI3K activation. We then
inspected the effects of such treatments on the lipid kinase activity
of PI3K. Both piceatannol treatment and SykT virus infection markedly
suppressed the phosphatidylinositol kinase activity of PI3K induced by
tumor cell engagement, as demonstrated by the capability of PI3K
immunoprecipitated from the same NK92 cells to phosphorylate
PI(4, 5)P2 (Fig. 3B). Taken together,
these results indicate that PI3K is under the control of Syk for NK
cytotoxicity.
|
It is likely that Syk operates via the PI3K pathway documented
earlier that involves Rac1 (7), as Rac1 has been shown to
be stimulated by the engagement of NK target cells and play important
roles in natural cytotoxicity (4, 36); in addition, Vav,
the upstream activator for Rac1, is activated by
L-phosphatidylinositol-3,4,5-trisphosphate, the
product of PI3K (37). But an alternative pathway could
also be taken by Syk that is Rac1 independent. If Syk should lie in the
exact pathway as the PI3K/Rac1 cascade described earlier, we predicted
that inhibition of Syk function by either SykT or piceatannol should
impair target-induced Rac1 activation. Indeed, both SykT virus
infection and piceatannol treatment remarkably reduced target-induced
Rac1 activation, as demonstrated by the association of Rac1-GTP, the
active form of Rac1, with PAK1 PBD in an affinity pull-down assay (Fig. 4A). The inactive form,
Rac1-GDP, is unable to associate with PAK1. Thus, these results showed
that Syk works upstream of Rac1 during NK lytic process.
|
B). However, rV12Rac1 markedly restored the impaired
cytotoxicity in piceatannol-treated NK92 cells to the same extent as
that seen in LY294002-treated NK92 cells (Fig. 4B).
Correspondingly, V12Rac1 restored MEK and ERK phosphorylation and
kinase activities in piceatannol-treated NK92 cells as effectively as
in LY094002-treated NK92 cells (Fig. 4, C and D).
These results indicate that Syk and PI3K do not function downstream
of Rac1. Suppression of target-triggered PAK1 activation by SykT or piceatannol, and the rescue of this inhibition by V12Rac1
Our previous work has pointed out that PAK1 was regulated by PI3K
and Rac1 in NK cytotoxicity (7). Examination of PAK1
demonstrated that SykT expression or piceatannol treatment
significantly suppressed target-triggered PAK1 activation in NK92 cells
(Fig. 5A). Moreover, V12Rac1
restored PAK1 activation impaired in piceatannol-treated NK92 cells as
effectively as in LY094002-treated NK92 cells (Fig. 5B).
These data indicate that Syk and PI3K act upstream of PAK1.
|
To date, Syk is the only signal molecule of those we
have tested that appears to lie upstream of PI3K. To confirm that Syk
cannot act downstream of PI3K in this cascade, we constructed a
vaccinia viral vector expressing constitutively active catalytic
subunit of PI3K, Myc-P110* (33). For PI3K to act
downstream of Syk, Myc-P110* should be able to counteract the
suppressive effects of piceatannol on the downstream effectors and
significantly rescue the impaired NK cytotoxicity in
piceatannol-treated NK92 cells. However, Myc-P110* should not be able
to restore PD98059-treated NK92 cells. To test this hypothesis, we
performed the following assays: NK92 cells, pretreated with piceatannol
or PD98059, were infected with vaccinia virus carrying Myc-P110* or
control CD56. The medium, DMSO plus CD56, and DMSO plus P110* NK92
control groups exhibited high lytic capacity against target cells, but
piceatannol- or PD98059-treated NK92 cells displayed significantly
suppressed NK lysis (Fig. 6A).
Furthermore, Myc-P110* expression, which was easily detected by
anti-P110 at a significantly high level, effectively recovered the
suppressed natural cytotoxicity in piceatannol-treated, but not
PD98059-treated NK92 cells (Fig. 6A). Correspondingly,
Myc-P110* rescued MEK/ERK phosphorylation and kinase activities in the
same NK92 cells that were inhibited by piceatannol, but not PD98059
treatment (Fig. 6, B and C). The same rescue of
PAK1 kinase activity was obtained by Myc-P110* expression in
piceatannol-treated NK92; however, because PAK1 is upstream of MEK,
PD98059 treatment did not impair PAK1 activation in NK92 cells and
subsequently, Myc-P110* expression had no effect (Fig. 6D).
A critical observation was that P110* expression could not restore the
activation of Syk in piceatannol-treated NK92 cells, unequivocally
demonstrating that Syk is upstream of PI3K (Fig. 6D). These
results, taken together, add up with earlier experiments to demonstrate
that Syk operates via PI3K in NK cells to effect lysis.
|
To ensure that this newly defined Syk-regulated lytic pathway does
not only exist in the NK92 cell line, but also functions in human NK
cells, we extended our research and applied the same P110*-rescue
strategy on fresh LGL isolated from normal donor peripheral blood. LGL
cells, similarly treated by piceatannol (50 µM) or PD98059 (50 µM)
at 37°C for 30 min, were infected with vaccinia virus carrying
Myc-P110* or CD56 as in Fig. 6A. Medium, DMSO plus CD56, and
DMSO plus P110* control exhibited high lytic capability against K562
tumor cells, and piceatannol or PD98059 treatment significantly
suppressed it (Fig. 7A).
Infection of Myc-P110* virus, but not that of CD56, significantly
restored the impaired cytotoxicity in piceatannol-treated, but not
PD98059-treated LGL (Fig. 7A). Correspondingly, P110*
markedly reelevated MEK and ERK activation suppressed in
piceatannol-treated, but not PD98059-treated, LGL cells (Fig. 7B). Thus, this Syk-directed PI3K signaling cascade also
appears to be physiologically relevant in freshly isolated human NK
cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, we have addressed this issue and have obtained firm
biochemical and molecular evidence demonstrating that Syk acts via the
PI3K-dependent pathway that we have earlier uncovered to control the
lytic process in NK cells. Target ligation rapidly caused DAP12
phosphorylation and association with Syk in NK cells. Inhibition of Syk
activation in NK cells triggered by target engagement led to the
down-regulation of target-stimulated PI3K, Rac1, PAK1, MEK, and ERK
activation, and consequently, the suppression of NK lytic capability,
indicating that Syk is upstream of all these effectors (
Figs. 1–5).
The ability of constitutively active PI3K to activate NK cytotoxicity
and PAK1, MEK, ERK kinase functions in piceatannol-pretreated NK cells
confirmed that Syk does not act downstream of PI3K and that PI3K is
upstream of PAK1, MEK, and ERK (Fig. 6). Constitutively active Rac1
activated PAK1, MEK, and ERK not only in piceatannol-treated NK cells,
but also in LY094002-treated NK cells, demonstrating that Syk and PI3K
do not act downstream of Rac1; however, PAK1, MEK, and ERK are
controlled by Rac1 (Figs. 4 and 5). With such combined use of specific
inhibitors of Syk, PI3K, and MEK, together with constitutively active
PI3K and Rac1 constructs, we believe that we have clearly demonstrated
that Syk critically controls the PI3KRacPAKMEKERK signaling
cascade that is required to mobilize lytic granules to effect tumor
lysis. This pathway was validated in freshly isolated human LGL,
implicating its biological relevance (Fig. 7).
Numerous cell surface receptors use ITAM-containing transmembrane
adaptor proteins as intermediates to signaling via Syk (38, 39). In response to immune receptor stimulation, the tyrosine
residues in the ITAMs become phosphorylated, creating binding sites for
the Src homology 2 domain of Syk, leading to Syk activation. The
ability of Syk to interact with the ITAMs of Ig
, CD3, CD3,
FcRI
, and FcRI reveals the fact that Syk essentially
participates in a variety of signaling process by a wide spectrum of
receptors in hemopoietic cells. For example, Syk is involved in T cell
development and TCR signaling, B cell activation and differentiation,
mast cell degranulation and macrophage phagocytosis, as well as
platelet activation (38). Thus, it is of obvious
significance to clarify the mechanisms by which Syk transduces the
signals originated from the membrane-anchoring receptors to the
intracellular effectors. In the case of NK cells, our identification of
the specific Syk-dependent PI3K cascade helps to distinguish at least
one pathway that contributes to natural cytotoxicity.
NK cells possess several ITAM-containing adaptor proteins that
associate with Syk. DAP12 forms homodimers and binds Syk or Zap70 via
its ITAMs and interacts with a wide range of NK receptor partners that
include KIR2DS2, KIR3DS, NKG2C/CD94, and NKp44 in humans and LY49D and
LY49H in mice, whereas CD3 and FcRI are present as either
disulfide-bonded homodimers or heterodimers (10, 28, 29, 30, 40). In human NK cells, CD3 and FcRI associate with
CD16, the low-affinity FcR for IgG that is responsible for Ab-dependent
cellular cytotoxicity. Cross-linking of CD16 results in tyrosine
phosphorylation of ITAMs in CD3 and FcRI, and the activation
of Syk/ZAP70 in NK cells (41, 42, 43). Thus, CD16
signaling-mediated Ab-dependent cellular cytotoxicity may also use the
SykPI3K pathway described in this work. Moreover, the recently
cloned NCRs also appear to use the same ITAM-containing adaptors. NKp44
has been reported to signal via DAP12 (10), NKp30 couples
to CD3 (11), while NKp46 couples to CD3 or
FcRI (44). Remarkably, NKp30 and NKp46, which
require Syk-recruiting ITAM+ adaptor proteins,
are reported to require PI3K for signaling (45), thus
providing support for a unique SykPI3K linkage in lytic function.
Whether all these receptors modulate NK function via the Syk-directed
PI3K pathway described in this work remains to be investigated.
Supporting literature validates the cross talk between Syk and PI3K in
other systems. Syk regulated PI3K and AKT activation following B
cell receptor (BCR) engagement, which was absent in Syk-deficient B
cells (46, 47). In FcRI-stimulated mast cells, Syk
contributed to the phosphorylation and enzymatic activation of AKT that
regulated the production of a variety of proinflammatory mediators,
such as IL-2 and TNF-, implicating the SykPI3K signaling in
cytokine production in mast cells (48). In addition, Syk
was also critical in
IIb3
integrin-mediated AKT activation in platelets and megakaryocytes
(49). Furthermore, PI3K has been implicated in a
Syk-dependent pathway controlling FcR-mediated phagocytosis in
macrophages (50), while TCR-induced Syk-dependent ERK
activation required the recruitment and activation of PI3K
(51). These findings strongly support that Syk associates
with PI3K to control a variety of functions in different cell
types.
The precise mechanism by which Syk activates PI3K is still under investigation. Shc is a good candidate for mediating this activation because it is a direct substrate of Syk (52). Shc cannot only recruit the Ras/MAPK cascade via Grb2/son-of-sevenless, but can also recruit the PI3K pathway. The recent cloning of Gab2 and the discovery of its association with Shc have provided a possible mechanism for propagating PI3K signaling (53). Engagement of the GM-CSF or IL-3R on BAF3 cells results in Shc phosphorylation, which forms a complex with Grb2 and Gab2, and leads to Gab2 tyrosine phosphoryation. It is the phosphorylated Gab2 that possesses a binding site for p85 subunit of PI3K. A similar system also appears to play a role in IL-2 activation, whereby Gab2 recruits and mediates the activation of PI3K in response to IL-2 (53, 54). In our NK cell system, it is convincing that Syk is critical for target cell-stimulated PI3K activation, and one possible outcome might be that Syk phosphorylates Gab2 directly or indirectly through Shc. In fact, we have observed a significant reduction in the tyrosine phosphorylation of Shc induced by target ligation in Syk-impaired NK cells (data not shown). Notably, a novel B cell adaptor protein for PI3K, which mediates BCR-associated protein tyrosine kinase-induced PI3K activation, has been reported recently to participate in BCR-induced PI3K signaling, whose tyrosine phosphorylation mediated by Syk and Btk provides binding site(s) for the p85 subunit of PI3K (55). Thus, it will be tempting to examine whether NK cells use B cell adaptor protein for PI3K-like and/or p36/38-like adaptors to bridge Syk to the PI3K signaling in response to triggering of NK-activatory receptors, as has been uncovered in BCR- and TCR-induced PI3K activation (51, 55).
In summary, our present work demonstrates that Syk is a critical
regulator of the PI3K pathway that we have earlier described to control
NK lytic function. Although our data only connected this specific
Syk-directed cascade to DAP12 activation, there is the likelihood that
other ITAM-bearing transmembrane adaptor proteins, CD3 and
FcRI, which also depend on Syk for signaling, and are known to
associate with CD16, NKp30, and NKp46 (15, 22), may use
the same specific pathway to mediate lysis of target cells. Future
experiments will answer this important question.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 K.J. and B.Z. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Julie Y. Djeu, Immunology Program, H. Lee Moffitt Cancer Center, Department of Interdisciplinary Oncology, MRC 4072, 12902 Magnolia Drive, Tampa, FL 33612. E-mail address: djeu{at}moffitt.usf.edu
4 Abbreviations used in this paper: MAPK, mitogen-activated protein kinase; BCR, B cell receptor; ERK, extracellular signal-regulated kinase; ITAM, immunoreceptor tyrosine-based activatory motif; KIR, killer-inhibitory receptor; LGL, large granular lymphocyte; MBP, myelin basic protein; MEK, mitogen-activated protein/ERK kinase; NCR, natural cytotoxicity receptor; PI3K, phosphoinositide 3-kinase; PI(4,5)P2, L-phosphatidylinositol-4,5-bisphosphate.
Received for publication July 19, 2001. Accepted for publication January 4, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cells by NKG2D via engagement by MIC induced on virus-infected cells. Nat. Immun. 2:255.[Medline]
with KARAP/DAP-12. Eur. J. Immunol. 30:2147.[Medline]
receptor-initiated activation of natural killer cells. J. Biol. Chem. 270:16415. receptors on macrophages. J. Exp. Med. 186:1027. chain-using cytokines. J. Biol. Chem. 275:26959.This article has been cited by other articles:
|
|
 |
|
  X. Chen, F. Bai, L. Sokol, J. Zhou, A. Ren, J. S. Painter, J. Liu, D. A. Sallman, Y. A. Chen, J. A. Yoder, et al. A critical role for DAP10 and DAP12 in CD8+ T cell-mediated tissue damage in large granular lymphocyte leukemia Blood, April 2, 2009; 113(14): 3226 - 3234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Zebedin, O. Simma, C. Schuster, E. M. Putz, S. Fajmann, W. Warsch, E. Eckelhart, D. Stoiber, E. Weisz, J. A. Schmid, et al. Leukemic challenge unmasks a requirement for PI3K{delta} in NK cell-mediated tumor surveillance Blood, December 1, 2008; 112(12): 4655 - 4664. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Tassi, M. Cella, R. Presti, A. Colucci, S. Gilfillan, D. R. Littman, and M. Colonna NK cell-activating receptors require PKC-{theta} for sustained signaling, transcriptional activation, and IFN-{gamma} secretion Blood, November 15, 2008; 112(10): 4109 - 4116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. V. Kondadasula, J. M. Roda, R. Parihar, J. Yu, A. Lehman, M. A. Caligiuri, S. Tridandapani, R. W. Burry, and W. E. Carson III Colocalization of the IL-12 receptor and Fc{gamma}RIIIa to natural killer cell lipid rafts leads to activation of ERK and enhanced production of interferon-{gamma} Blood, April 15, 2008; 111(8): 4173 - 4183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Wahle, K. H. T. Paraiso, R. D. Kendig, H. R. Lawrence, L. Chen, J. Wu, and W. G. Kerr Inappropriate Recruitment and Activity by the Src Homology Region 2 Domain-Containing Phosphatase 1 (SHP1) Is Responsible for Receptor Dominance in the SHIP-Deficient NK Cell J. Immunol., December 15, 2007; 179(12): 8009 - 8015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Pandey, C. M. DeStephan, L. A. Madge, M. J. May, and J. S. Orange NKp30 Ligation Induces Rapid Activation of the Canonical NF-{kappa}B Pathway in NK Cells J. Immunol., December 1, 2007; 179(11): 7385 - 7396. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. K. Epling-Burnette, F. Bai, J. S. Painter, D. E. Rollison, H. R. Salih, M. Krusch, J. Zou, E. Ku, B. Zhong, D. Boulware, et al. Reduced natural killer (NK) function associated with high-risk myelodysplastic syndrome (MDS) and reduced expression of activating NK receptors Blood, June 1, 2007; 109(11): 4816 - 4824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. D. Wiseman, L. L. Ma, K. J. Marr, G. J. Jones, and C. H. Mody Perforin-Dependent Cryptococcal Microbicidal Activity in NK Cells Requires PI3K-Dependent ERK1/2 Signaling J. Immunol., May 15, 2007; 178(10): 6456 - 6464. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-K. Lee, H. M. Lee, H. J. Kim, H.-J. Park, K.-J. Won, H. Y. Roh, W. S. Choi, B. H. Jeon, T.-K. Park, and B. Kim Syk contributes to PDGF-BB-mediated migration of rat aortic smooth muscle cells via MAPK pathways Cardiovasc Res, April 1, 2007; 74(1): 159 - 168. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Wang, C. Lau, S. Wiehler, A. Pow, T. Mazzulli, C. Gutierrez, D. Proud, and C.-W. Chow Syk Is Downstream of Intercellular Adhesion Molecule-1 and Mediates Human Rhinovirus Activation of p38 MAPK in Airway Epithelial Cells J. Immunol., November 15, 2006; 177(10): 6859 - 6870. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Ortaldo, R. Winkler-Pickett, J. Wigginton, M. Horner, E. W. Bere, A. T. Mason, N. Bhat, J. Cherry, M. Sanford, D. L. Hodge, et al. Regulation of ITAM-positive receptors: role of IL-12 and IL-18 Blood, February 15, 2006; 107(4): 1468 - 1475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Parihar, R. Trotta, J. M. Roda, A. K. Ferketich, S. Tridandapani, M. A. Caligiuri, and W. E. Carson III Src Homology 2-Containing Inositol 5'-Phosphatase 1 Negatively Regulates IFN-{gamma} Production by Natural Killer Cells Stimulated with Antibody-Coated Tumor Cells and Interleukin-12 Cancer Res., October 1, 2005; 65(19): 9099 - 9107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. D. Huntington, Y. Xu, S. L. Nutt, and D. M. Tarlinton A requirement for CD45 distinguishes Ly49D-mediated cytokine and chemokine production from killing in primary natural killer cells J. Exp. Med., May 2, 2005; 201(9): 1421 - 1433. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. D. Moon, C. B. Post, D. L. Durden, Q. Zhou, P. De, M. L. Harrison, and R. L. Geahlen Molecular Basis for a Direct Interaction between the Syk Protein-tyrosine Kinase and Phosphoinositide 3-Kinase J. Biol. Chem., January 14, 2005; 280(2): 1543 - 1551. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Zompi, H. Gu, and F. Colucci The absence of Grb2-associated binder 2 (Gab2) does not disrupt NK cell development and functions J. Leukoc. Biol., October 1, 2004; 76(4): 896 - 903. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Inngjerdingen, B. Rolstad, and J. C. Ryan Activating and Inhibitory Ly49 Receptors Modulate NK Cell Chemotaxis to CXC Chemokine Ligand (CXCL) 10 and CXCL12 J. Immunol., September 15, 2003; 171(6): 2889 - 2895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Zhong, K. Jiang, D. L. Gilvary, P. K. Epling-Burnette, C. Ritchey, J. Liu, R. J. Jackson, E. Hong-Geller, and S. Wei Human neutrophils utilize a Rac/Cdc42-dependent MAPK pathway to direct intracellular granule mobilization toward ingested microbial pathogens Blood, April 15, 2003; 101(8): 3240 - 3248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Khurana and P. J. Leibson Regulation of lymphocyte-mediated killing by GTP-binding proteins J. Leukoc. Biol., March 1, 2003; 73(3): 333 - 338. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Resendiz, S. Feng, G. Ji, K. A. Francis, M. C. Berndt, and M. H. Kroll Purinergic P2Y12 Receptor Blockade Inhibits Shear-Induced Platelet Phosphatidylinositol 3-Kinase Activation Mol. Pharmacol., March 1, 2003; 63(3): 639 - 645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Jiang, B. Zhong, C. Ritchey, D. L. Gilvary, E. Hong-Geller, S. Wei, and J. Y. Djeu Regulation of Akt-dependent cell survival by Syk and Rac Blood, January 1, 2003; 101(1): 236 - 244. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
