| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Amgen Institute/Ontario Cancer Institute, Departments of Medical Biophysics and Immunology, University of Toronto, Ontario, Canada
Correspondence: Josef Penninger, Amgen Institute/Ontario Cancer Institute, Departments of Medical Biophysics and Immunology, University of Toronto, 620 Toronto Ave., Ontario, Canada M5G 2C1. E-mail: jpenning{at}amgen.com
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Key Words: lymphocytes • actin • Vav • cytoskeleton • T-cell activation • immune synapse
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Lymphocyte development, activation, and function are reliant on a dynamic cytoskeleton. During lymphocyte maturation, developing thymocytes and B-cell precursors are in continuous contact with stromal cells of the microenvironment, and these physical contacts are crucial prerequisites for lymphocyte maturation and selection [1 ]. Mature lymphocytes migrate through blood and lymph vessels, attach to endothelial cells, transmigrate through blood vessel walls, and extracellular matrix spaces, home into secondary lymphoid organs, interact with antigen-presenting cells (APCs) and stromal cells, and adhere to target cells. These extracellular interactions relay intracellular signals, which influence lymphocyte homeostasis, inducing proliferation, activation, survival, apoptosis, and/or unresponsiveness. Precise lymphocyte responses rely on effective signal transduction cascades, which are influenced by cytoskeletal remodeling. Many signaling molecules are associated with cytoskeletal scaffolds, and the cytoskeletal structure and scaffold geometries can directly regulate molecular dynamics of signaling and biochemical responses [2 ]. Similarly, cytoskeletal reorganizations and subcellular distribution of signaling molecules and scaffolded signaling modules help to fine-tune cellular responses. Moreover, it has been shown that under certain experimental conditions cell shape can regulate cell fate decisions for survival or apoptosis [3 , 4 ]. Compartmentalization of signaling cascades through protein-protein or protein-lipid interactions is a fundamental mechanism for the initiation, fine-tuning, and propagation of effective signal transduction and cellular responses [5 , 6 ]. In lymphocytes, a highly organized supramolecular activation complex (SMAC) has been described that forms at the site of contact between an APC and a T cell. Based on the structural similarity to synapses in the nervous system, SMACs have also been termed the "immune synapse." SMAC formation is dependent on the actin cytoskeleton and can be induced by costimulatory signals [7 ].
|
CYTOSKELETAL HARD-WIRING AND METAMORPHOSIS |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Although hard-wired actin cables and tension-regulated cell shape changes can contribute to biochemical responses, an extraordinary feature of the cytoskeleton is that it is also highly adaptable and changes rapidly in a dynamic fashion. After biochemical or mechanical stimulation, various reversible and rapid changes occur in the actin cytoskeleton: monomeric, globular G-actin polymerizes into filamentous F-actin (microfilaments) constructed of identical strings of actin monomers. Strings of F-actin assemble into F-actin bundles and form higher-order structures such as stress fibers, which allow cultured cells to attach to surfaces. Actin filaments also associate with the plasma membrane and determine cell shape, motility, or formation of filopodias and lamellipodias. Similarly, intermediate myosin filaments together with F-actin form highly organized structures such as sarcomeres in muscles. Moreover, polymeric microtubules are structures that are in a dynamic balance with monomeric subunits. It is important to note that cytoskeletal organization can be different in distinct compartments of the same cells [13 , 14 ]. Because many cytoplasmic molecules, including signaling molecules, are associated with the cytoskeleton, and the cytoskeleton provides a scaffold on which organelles and proteins bind, subcellular differences in the actin cytoskeleton have profound functional consequences for a cell.
|
LYMPHOCYTE FUNCTION AND ACTIVATION: IMPLICATIONS OF THE CYTOSKELETON |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
(PLC)
activation and Ca2+ flux, activation of protein kinase C
(PKC) isoforms, MAPK, and SAPK/JNK signaling pathways, and
translocation of transcription factors into the nucleus
[15
16
17
] (Fig. 1
). Activation of antigen receptors in lymphocytes also leads to
changes in the actin cytoskeleton defined by actin polymerization,
dynamic reorganization of the actin matrixes, membrane ruffling, or
formation of filopodias and uropods [18
,
19
]. In addition, antigen-receptor engagement leads to
reorientation of the microtubule organizing centers (MTOCs) and
membrane-associated cytoskeleton in order to accomplish unidirectional
killing of target cells by cytotoxic T cells and natural killer (NK)
cells or during physical interactions between T helper cells and B
cells to regulate polar secretion of cytokines [20
].
Moreover, in T cells the Golgi apparatus is reoriented toward
antigen-presenting cells. Butanedione monoxime can prevent CD43-induced
polarization and homotypic cell aggregation indicating the involvement
of the myosin cytoskeleton in these phenomena [21
].
Engagement of antigen receptors on T cells also leads to the
recruitment of signaling molecules and TCR/CD3 complexes into so-called
membrane rafts [22
23
24
25
26
]. Rafts are defined anatomical
microdomains in the cell membrane that are enriched in
glycosphingolipids and cholesterol [22
,
26
].
|
chain [27
]. However, approximately
30–40% of total TCR molecules on the surface of resting T cells
are already associated with the actin cytoskeleton [28
].
The cytoskeletal-linked TCR chains display distinct patterns of
phosphorylation after TCR activation and could play a negative role in
signal transduction [29
]. In addition, various
cytoskeleton-organizing molecules such as ezrin, talin, vinculin,
paxillin, fimbrin, the focal adhesion kinase (FAK), its homologue Pyk2
(CAKß or RAFTK), and the myosin light chain have been shown to be
phosphorylated and activated after antigen receptor activation
[30
]. These studies suggest that TCR-mediated signal
transduction cascades lead to cytoskeletal remodeling, which is
essential for TCR signal propagation and T-cell activation. It should
be noted that experiments performed on the Jurkat T cell line using
pharmacological inhibitors revealed contradictory evidence that a
dynamic actin and microtubule cytoskeleton were indeed necessary for
proximal T cell signaling events. TCR-induced tyrosine phosphorylation
in Jurkat T cells after a 5-min incubation with anti-CD3 antibody was
not affected by cytochalasin D and colchicine, which disrupt actin
polymerization and microtubule polymerization, respectively
[31
]. Although this study suggests that immediate
TCR-induced tyrosine phosphorylation events may not require actin and
tubulin polymerization, it does not indicate that the signaling events
induced by the TCR in the presence of cytoskeletal-disrupting
pharmacological agents are sufficient to induce T-cell proliferation,
activation, and cytokine production. Recent studies have demonstrated a
correlation between induction of cytoskeletal remodeling and sustenance
of protein tyrosine phosphorylation, both events crucial for full
T-cell activation [7
, 32
]. In naive T
cells, these events require costimulatory signals and are evident 5–15
min after receptor engagement. It is therefore plausible that, after
TCR stimulation, cytoskeletal reorganization inhibitors would abrogate
sustained tyrosine phosphorylation rather than immediate tyrosine
phosphorylation events and result in a failure to induce sufficient
signal transduction events resulting in full T-cell activation. Actin
cytoskeletal reorganization has been shown to be necessary for T-cell
proliferation and cytokine production, however, precise contribution of
the microtubule cytoskeleton to propagation of TCR signaling events is
not certain. Although microtubule blocking agent colchicine does not
appear to have an effect on immediate proximal TCR signals, its
relevance in sustained TCR signaling is unknown [31
].
However, tubulin has been shown to interact with and be a substrate of
ZAP70, a tyrosine kinase important in TCR signaling
[33
]. In addition, tubulin was detected in
co-immunoprecipitates with Vav1, a crucial mediator of cytostructure
remodeling events after TCR signaling [34
]. Finally, as
mentioned, reorientation of the MTOC is an early activation event after
signals that induce full T-cell activation. Collectively, both tubulin
and actin polymerization appear important to cellular reorganization
after TCR signaling to induce, sustain, and propagate effective signal
transduction cascades sufficient for T-cell activation. More recently, the necessity of a dynamic cytoskeleton for lymphocyte activation and function was solidified through the use of genetically targeted mice. Mutations that interfere with antigen receptor-mediated cytoskeletal reorganization, including mutations in the guanosine nucleotide exchange factor Vav1 [35 , 36 ] Rho-family proteins [19 ], or the Wiscott Aldrich syndrome protein (WASP) [37 ] severely impair lymphocyte activation and lymphocyte development. The studies on genetically targeted mice have not only identified important molecular links between signaling cascades and induction of cytoskeletal remodeling, but have allowed the dissection of cytoskeletal-dependent and -independent signaling cascades downstream of the TCR.
Elegant studies have revealed an essential role of actin remodeling on T-cell activation. Most cell surface receptors that regulate cell growth or death form supramolecular activation clusters. In lymphocytes, antigen receptor stimulation leads to the formation of antigen receptor Caps, which are asymmetrical aggregations of antigen receptors at the site of receptor engagement [38 ]. Engagement of the TCR in the presence of costimulation induces the formation of the immunological synapse, which is an organized assembly of antigen receptor clusters, adhesion molecules, co-receptors, membrane rafts, and intracellular signaling molecules located at the site of receptor engagement [39 ]. The immune synapse is further organized into structurally discrete SMACs [40 ]. The process of receptor and lipid raft recruitment and SMAC formation is dependent on dynamic reorganization of the actin and myosin cytoskeleton [39 ]. For a more complete and detailed review of the reorganization events observed during the formation of the immune synapse and the contribution of lipid rafts to TCR signaling, please see previously published reviews [40 41 42 ].
|
THE IMMUNE SYNAPSE |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
In resting T and B cells, antigen receptors appear to be evenly distributed throughout the cell surface. Activation of antigen receptors on T lymphocytes leads to the organization of SMACs at the interfaces of physical contact between T cells and antigen-presenting cells [46 ]. SMACs constitute highly organized structures that assemble antigen receptors, adhesion and co-receptors, and signaling molecules within a focal interaction site (Fig. 2 ). Whereas by definition SMACs form at sites of antigen recognition, antibody cross linking of antigen receptors can also trigger the formation of asymmetric and polar receptor patches and caps. Antibody-induced patches and caps might be similar in their functional relevance to SMACs, although important structural differences must exist due to the nature of these interactions. Formation of SMACs and Caps are relatively late events in TCR activation and depend on a functional actomyosin cytoskeleton. The reorganized cytoskeleton probably pulls the antigen receptors in the lipid membranes to form Caps and SMACs and provides a scaffold for the focal assembly for structural and signaling molecules. Antigen receptor activation also leads to modulation of integrin affinity, a phenomenon known as inside-out signaling [47 ]. Induction of high-affinity integrin binding enhances the overall avidity of interactions between lymphocyte and antigen-presenting cells and further links cell surface receptors to the cytoskeleton. Actin-cytoskeleton-dependent clustering of antigen receptor together with other membrane and cytosolic proteins is critical for sustained TCR signaling [48 , 49 ]. Full T-cell activation requires not only a TCR-mediated signal but also a costimulatory signal. It is intriguing that costimulatory signals have been shown to be required for both sustained TCR-induced protein tyrosine phosphorylation and SMAC formation [7 ]. Thus, if costimulation is essential for SMAC formation, SMAC formation may coordinate the molecular mechanisms that regulate immunological tolerance.
|
. The peripheral SMAC regions are
enriched in LFA-1 and Talin. Segregation of MHC class II molecules and
the LFA-1 ligand ICAM1 in the contact zone of the antigen-presenting
cell mirrors SMAC segregation on T cells. Based on physical dimensions
and molecular interactions, it has been proposed that CD4, CD8, CD28,
and CD2 molecules cosegregate with the central SMAC areas, whereas CD45
and CD43 might also be excluded from the central contact zone
[49
]. Spatial segregation of activated receptors and
signal transducing molecules might provide a unique environment for
signal transduction. When T cells were activated with altered peptides
that do not trigger cytokine production or T cell proliferation, CD3
and talin formed clusters but failed to segregate. Thus, it appears
that segregation of receptors and signaling molecules within SMACs is
critical for the initiation of physiological responses. The critical question remains how and why SMACs segregate into two structurally distinct regions. One hint came from the cloning of a novel SH3-domain-containing adaptor molecule called CD2-associated protein (CD2AP). Engagement of the adhesion molecule CD2 initiates a process of protein segregation, receptor clustering, and polarization of the T cell cytoskeleton [50 ]. Similarly, CD2AP can mediate receptor clustering and cytoskeletal polarization. Besides three SH3 domains, CD2AP also contains proline-rich stretches and a sequence that has homology to the monomeric actin-binding protein thymosine-ß4. Because association between CD2 and CD2AP are dependent on T-cell activation, it has been suggested that CD2AP links TCR activation to the adhesion molecule CD2 and subsequent receptor patterning and T-cell polarization. Similarly, CD2AP might scaffold T-cell polarization and receptor clustering. However, CD2 gene-deficient mice have no apparent defects in T-cell activation and lymphocyte development, suggesting that surface receptors other than CD2 can mediate receptor patterning and/or CD2AP can associate with different molecules. The CD2AP knockout mice are reported to develop a severe polycystic kidney disease, however, the effect of a loss of function mutation of CD2AP on formation of the immune synapse has not been reported [51 ].
|
THE MOLECULAR MOTOR OF ANTIGEN RECEPTOR CLUSTERING |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
chain [36
,
54
]. Cap formation is also impaired in the
CDC42-associated WASP-deficient mice [37
]. In addition
to defects in antigen receptor capping, human Wiscott Aldrich syndrome
is characterized by immunodeficiency, thrombocytopenia, and lymphoid
malignancies. The transfection of the gene for a dominant-negative Rac1
mutation also results in a failure of Cap assembly in lymphocytes
[35
]. These experiments indicate that Vav1, Rac1, and
WASP constitute at least part of a signaling pathway that links antigen
receptor engagement to cytoskeletal reorganization, receptor
clustering, and Cap formation. |
RHO-FAMILY GTPASES AND GUANINE NUCLEOTIDE EXCHANGE FACTORS |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Multiple Rac1, CDC42, and/or RhoA target molecules have been described that differ in their functions and cellular distribution [62 ]. For example, Rho can associate with and/or potentially activate PIP5 kinase, which regulates PIP2 production, protein kinase N (PKN), p164 Rho-associated kinase (Rho-K), p160 Rho-associated coiled-coil-containing protein serine/threonine kinase (ROCK), PLD1, or the myosin light chain (MLC) phosphatase. Rac associates with the p21-activated kinase PAK, PLD1, POSH, POR1, MLK2, MLK3, p60 S6 kinase, p67phox, IQGAP1/2, and possibly PIP5 kinase. Interplay between Rac1 and Rac2 and PKB/Akt has been reported. Rac2-deficient mast cells have a defect in PKB/Akt activation, whereas Rac1 has been implicated as a regulator of PKB/Akt activation [63 64 65 ]. CDC42 interacts with WASP, p21-activated kinase PAK, MLK2, MLK3, IQGAP1/2, or p60 S6 kinase. This list of Rho, Rac, and CDC42 binding partners and molecular targets is expanding [66 ].
Although many of the described Rho, CDC42, and Rac targets might be involved in the regulation of the cytoskeleton, production of PIP2 via PIP5 kinase activation might be a crucial link between surface receptor engagement, Rho-GTPases, and reorganization of the cytoskeleton. PIP2 can bind to the actin capping molecules gelsolin and profilin. This association releases the capping molecules from actin and thus promotes polymerization [67 , 68 ]. PIP2 can also associate with and unfold vinculin and regulate vinculin/talin/actin interactions in focal sites of adhesion [69 , 70 ]. PIP2 has also been implicated in the translocation and association of the chromatin remodeling complex SWI/SNF (or BAF complex) to the chromatin in T cells [71 ]. In addition to PIP2, it has been proposed that the Rho-regulated MLC phosphatase regulates actin bundling and stress fiber formation and links Rho-activation to the myosin-actin cytoskeleton [72 ]. Moreover, PAK1, which is expressed in lymphocytes, has been shown to regulate actin polymerization in mammalian cells [73 ].
The Rho family of GTPases has been most characterized as crucial regulators of cytoskeletal remodeling. However, the complex regulation of cytoskeletal architecture is not under exclusive control by the Rho family GTPases. ARF6, a member of the ARF family of Ras-related GTPases, redistributes to the cortical cytoskeleton [74 ] and can induce cortical actin polymerization possibly by regulating Rac1 distribution [75 ]. ARF1 potentiates Rho-induced assembly of actin stress fibers and formation of paxillin-rich focal adhesions [76 ]. As well, overexpression of the ARF-activating protein, ASAP, in cell lines alters the cellular morphology and influences cytoskeletal remodeling in response to PDGF stimulation [77 ]. Gcs1p, an Arf guanosine triphosphatase-activating protein in Saccharomyces cerevisiae, binds actin filaments and stimulates actin polymerization in vitro [78 ].
Lymphocytes express many small G proteins including p21Ras, Rap1, and
the Rho-family GTPases RhoA, Rac1, Rac2, and CDC42 [19
].
All of these molecules have a crucial role in the development and
function of lymphocytes. For example, a functional Ras signaling
pathway is required in B cell development before formation of the pre-B
cell receptor and Ras signaling is important for TCR-mediated positive,
but not negative selection [79
]. Rap1 has been
implicated in the induction of clonal anergy and regulation of
interleukin-2 (IL-2) production in Jurkat cells [80
].
Early thymocyte differentiation requires Rho for the survival of pro-T
cells and for the cell cycle progression of late pre-T cells
[81
, 82
]. Constitutive activation of Rac2,
a hematopoietic-specific small GTPase, or Cdc42 in thymocytes or T cell
lines can induce apoptosis [83
]. Cdc42 controls T cell
polarity toward antigen-presenting cells and is required for IL-2
production [84
, 85
]. Rac1 also has a role
in antigen-receptor and IL-2 receptor-mediated membrane ruffling
[86
] and, importantly, TCR-mediated cap formation
[87
]. Moreover, Rac1 is important in NK cells to
reorient cytotoxic granules toward their target cells and abrogation of
Rac1 activity in NK cells activity impairs cytotoxic activity
[88
]. In mast cells, Rac and CDC42 activity has also
been implicated in Fc
RI-mediated phagocytosis and kit-mediated
proliferation. CDC42 is also a crucial regulator of dendritic cell (DC)
endocytosis. Immature DCs actively internalize antigen and express
active CDC42, whereas active CDC42 is undetectable in poorly endocytic
mature DCs [89
]. In short, Rho-like GTPases play a
crucial role in orchestrating physiological responses in hematopoietic
cells and T and B lymphocytes in response to receptor engagement. The
challenge for the future of lymphocyte activation will be to dissect
expression and complex formation of different Rho-GTPases with
different downstream targets.
Similar to Ras, Rho-like GTPases are being activated when bound GDP is exchanged for GTP, a process catalyzed by guanine nucleotide exchange factors (GEF; or GDP-dissociation stimulator GDS proteins). GEFs are a growing family of cellular second messengers characterized by the canonical coupling of a Dbl-homology (DH) domain, which confers guanine-nucleotide exchange activity, to an adjacent pleckstrin-homology (PH) domain. The DH domain of the proto-oncogene Dbl confers the capacity for guanine nucleotide exchange for the Rho family of small GTPases [59 , 62 , 73 , 90 91 92 ]. Besides prototypic Dbl-homology family GEFs, an additional family of GEFs, the PIX (PAK-interacting exchange protein) family, which exhibit catalytic activity toward Rac1 have been isolated by virtue of binding to the PAK kinase family of Rac and Cdc42 effectors [93 ]. In yeast, binding of the GEF Cdc24 to the Cdc42 effector Ste20, which is a PAK homologue, regulates the growth of filamentous actin. In mammalians, GEFs exhibit a multitude of different functions. For example, the GEF Tiam1 and Tiam1-regulated Rac1 signaling regulate membrane ruffling and invasiveness of T cell lymphomas [94 ]. On the other hand Tiam1/Rac1 signaling can inhibit invasion of epithelial cells and can restore E-cadherin-mediated adhesion in Ras-transformed MDCK cells [95 ], suggesting that Tiam1-Rac1 can promote or suppress invasion of tumor cells depending on the cellular signals and cell types.
|
VAV1: SIGNALING AND THE ACTIN CYTOSKELETON |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Recently two structural Vav1 homologues have been identified, Vav2 and Vav3 [101 , 102 ]. Both Vav2 and Vav3 are expressed in T cells and could potentially function in a similar signaling pathway as Vav1. However, in in vitro experimental systems, it appears that at least Vav3 has different substrate specificity for Rho-family GTPases as compared to Vav1 [102 ]. Moreover, it is possible that different receptors and/or upstream signaling pathways differentially activate Vav1, Vav2, or Vav3.
Vav1 was originally implicated in Ras and MAPK signaling [98 , 103 , 104 ]. However, recent biochemical in vitro experiments and genetic studies have established that Vav1 functions as a GDP/GTP-exchange factor (GEF) for the Rho-like small GTPases RhoA, Rac1, and CDC42 [53 , 60 , 62 , 92 , 105 ]. Moreover, Vav1 expression can complement the temperature-sensitive Cdc24ts yeast strain, but Vav1 does not complement the Cdc25ts strain, which is the yeast exchange factor for Ras. In intact cells, it is, however, not known whether Vav1 activates all Rho-family GTPases or exhibits substrate specificity toward a selective member of the Rho-family GTPases [53 ]. Because RhoA, Rac, and CDC42 can interact and regulate distinct downstream signaling pathways, elucidation of the exact in vivo GTPase target of Vav1 is of crucial importance to identify specific effectors of Vav1-regulated signaling pathways.
For Vav1 to act as a GEF for Rac, RhoA, or CDC42, PI3’ kinase activity
and the binding of PI3’K-generated phospholipid products to the PH
domain of Vav1 are required [106
]. PH domains mediate
interactions with specific charged headgroups of phosphoinositides and
regulate subcellular redistribution of signal molecules to regions of
localized signaling at the plasma membrane as well as the activity of
the target proteins. For instance, the exchange activity of Vav1 toward
Rac1 is enhanced in the presence of the PI3’K products
phosphatidylinositol-3-4-phosphate [PtdIns(3,4)P] or
phosphatidylinositol-3-4-5-phosphate [PtdIns(3,4,5)P] but inhibited
in the presence of the PI3’K substrate PI(4,5)P2 [106
].
Thus it appears that PI3’ kinases may strongly stimulate Vav1 activity
by converting inhibitory regulators of Vav1 to positive activators. In
addition to the regulation by phospholipids, full activation of Vav1 is
dependent on TCR-mediated tyrosine phosphorylation, which is mediated
via p56lck and/or ZAP70 kinases [53
, 107
,
108
]. These experiments suggest that the catalytic
activity of Vav1 is regulated by a combination of enzymatic and
positional signals and Vav1 catalytic activity requires two input
signals originating from activated Lck/ZAP70 and membrane
phosphoinositides generated by activated PI3’K. The differences in the
levels of the two signals may contribute to the fine tuning of cellular
responses to antigen-receptor signals via Vav1 activity. In addition,
GTPases activated by Vav1 may also be located in or recruited to the
microenvironment of the activated receptor. This prediction is in line
with findings that RhoA, Rac1, and CDC42 colocalize with the TCR/CD3
complex to receptor patches and caps [36
]. It is
interesting that Vav1 may initiate a negative feedback loop via
activation of Rac and Rho, both of which have been shown to bind to and
activate phosphoinositide-4-phosphate 5 kinase, which catalyzes
production of the Vav1 inhibitor PI(4,5)P2 (PIP2)
[109
]. PIP2 in cell membranes is highly rate
limiting, and catalysis of PIP2 into IP3 and
diacylglycerol by PLC is a crucial prerequisite for the opening of
intracellular Ca2+ channels and activation of various PKC
isoforms. Moreover, PIP2 can promote actin polymerization
and stabilizes focal adhesion complexes.
In addition to the GEF activity, Vav1 appears to function as a
molecular scaffold with adaptor molecules such as Slp-76 or Nck, which
participate in the assembly of multimolecular signaling complexes
downstream of activated receptors [110
]. Accordingly, it
has been reported that Vav1 associates with a large variety of
molecules via its SH2 and two SH3 domains. Whereas SH3 domains can be
also found in two other mammalian GEFs, Dbs, and PIX, each of which
contain one SH3 domain, Vav1 was the first known GTPase activator that
contains an SH2 domain. The Vav1 SH2 domain binds Shp-1, Syk, Zap70,
Slp-76, and c-Cbl [108
, 111
112
113
114
]. Vav1
itself becomes rapidly phosphorylated on tyrosine residues upon
activation by a variety of cell-surface receptors, and becomes a target
of SH2-containing proteins, which include Syk, Fyn, PLC, p120
RasGAP, Shc, and the p85 subunit of PI3’K. In Jurkat T cells, Vav1 and
the Vav1-associated phosphoprotein Slp-76 synergize with Calcineurin to
induce NF-AT transactivation and IL-2 production [112
,
114
, 115
]. Slp-76 expression is fairly
restricted to T cells and gene ablation of Slp-76 leads to a complete
block in early thymocyte development [116
].
The Vav1-SH3 domain interacts with a variety of quite unrelated proteins such as the heterogenous ribonucleoprotein K (hnRNPK), which has been proposed as a docking platform for the assembly of signaling modules, Ku-70, which forms part of the DNA-binding subunit of the nuclear DNA-dependent subunit of protein kinase DNA-PK, and Cbl-b, a member of the Sli/c-Cbl protein family implicated in negative regulation of signal transduction [117 ]. Moreover, the molecular adapter Grb2 associates Vav1 via the Grb2 SH3 domain. In B cells, it has been shown that Vav1 associates with Btk via the Btk-SH3 domain [118 ]. Intriguingly, hnRNPK, which associates with the Vav1-SH3 domain and WASP, also binds to the SH3 domain of Btk [119 ]. WASP is a cytoskeletal protein and effector of CDC42 required for IL-2 production and proliferation in T cells and TCR capping [37 ]. Moreover, the Vav1 carboxy-terminal SH3 domain has been shown to associate with the cytoskeletal protein zyxin.
To analyze the role of Vav1 on in vivo T cell development
and activation, we and others have targeted the vav1 gene by
homologous recombination [35
, 36
].
vav1-/- mice are viable and healthy. However, in
the absence of Vav1, early thymocyte development at the pre-TCR stage,
positive and negative thymocyte selection, peptide/MHC- and antigen
receptor-mediated thymocyte apoptosis, antigen receptor-induced cell
cycle progression, calcium mobilization, and the activation of immune
response genes such as the IL-2 gene are all impaired
[35
, 36
, 54
,
120
]. Although the TCRs do not cluster in response to
antigenic stimulation, principal signaling pathways such as overall
tyrosine phosphorylation, MAPK and SAPK activation can still operate.
However, the level of signal transduction achieved is clearly not
sufficient to produce a complete T cell response. Vav1 was also found
to bind constitutively to the cytoskeletal membrane anchors talin and
vinculin and in the absence of Vav1, phosphorylation Slp-76 and
vinculin, Slp-76/talin interactions, and recruitment of the actin
cytoskeleton to the CD3 chain were impaired. Consistent with a role
for Vav1 in transducing TCR signals to the actin cytoskeleton,
vav1-/- mice T cells displayed impaired actin
polymerization in response to antigen receptor activation and exhibit
defective clustering (patching and capping) of the TCR. Functionally,
loss of Vav1 expression results in impaired, albeit not abolished,
in vivo B and T cell immunity in response to multiple virus
infections and challenges with protein antigens [121
].
vav-/- B cells develop normally and can be
activated by repetitive, polyvalent T cell-independent viral antigens
that effectively cross link B cell receptors, but not by non-repetitive
hapten antigens or anti-IgM antibody treatment [122
].
Thus, it appears that, similar to in vitro T-cell
activation, Vav1 is required for efficient antigen receptor clustering
in vivo and viral antigens that can provide strong antigen
receptor cross linking can overcome the functional defect.
The data from vav1-/- T cells demonstrates that the activation of non-clustered T cell receptors can still induce activation of principal signaling pathways, but are not sufficient to confer the full T cell response invoked by multimerized T cell receptors. Thus, we proposed that antigen receptor activation may lead to the organization of focal actin-scaffolded signaling highways [123 ]. These results may offer a biochemical rationale for receptor reorganization and clustering, i.e, actin-dependent receptor clustering results in signal transduction at regionally organized focal points as a prerequisite for the induction of physiological responses. These genetic data provided the first genetic evidence for a role for Vav1 in cytoskeletal reorganization in vivo and showed that Vav1 is a crucial and specific regulator of TCR-mediated cytoskeletal reorganization and receptor clustering and calcium flux required for T cell maturation, IL-2 production, and cell cycle progression.
Similarly to Vav1 mutant mice, deletion of the WASP gene severely impairs T cell proliferation and cytokine production [37 , 124 ]. At the molecular level, wasp-/- T cells display defective antigen receptor clustering after antigen receptor stimulation. Antigen receptor- and CD28-mediated activation of ERK1/ERK2, SAPKs/JNKs, as well as Ca2+-mobilization is normal in wasp-/- T cells. Moreover, WASP associates with activated CDC42 and WASP-regulated cytoskeletal changes require binding of WASP to the active form of CDC42 [125 ]. Thus, in T cells, the Rho-family GDP/GTP exchange factor Vav1, the Rho-family GTPases Rac1, CDC42, and RhoA, and the CDC42-associated WASP constitute a signaling pathway that links antigen receptor engagement to receptor clustering. Genetically Vav1, Rac1, and WASP regulated cytoskeletal reorganization and receptor clustering are essential prerequisites for lymphocyte maturation, cell cycle progression, IL-2 production, and the induction of physiological T cell responses.
Additional clues can be found in studies of thymocytes and mature peripheral T cells treated with cytochalasin D (CytD) and latrunculin, two inhibitors of actin polymerization. These drugs inhibit IL-2 production, proliferation, TCR capping, and peptide/MHC-mediated thymocyte apoptosis but do not impair known signaling pathways with the exception of Ca2+ mobilization [35 , 54 ]. Thus, the effects of actin polymerization blockers on T cell signaling and T cell functions mimic an absence of Vav1 or WASP or the overexpression of dominant-negative Rac1. Moreover, it has been reported that Nck, Slp-76, SLAP/Fyb contribute to the scaffolding and formation of the molecular motor complex that ultimately links to Arp2/3, Ena/VASP, and actin polymerization [126 ] (Fig. 3 ).
|
|
MEMBRANE RAFTS |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Functionally, it has been suggested that rafts might play a role
in morphogenesis, membrane trafficking, and/or signal transduction
[41
, 42
]. For example, rafts are enriched
for a variety of important signaling proteins, including
glycosylphosphatidylinositol (GPI)-linked surface molecules, src-family
and receptor tyrosine kinases, heterotrimeric and Ras-like G proteins,
molecules involved in Ca2+ mobilization, including an
IP3-responsive Ca2+ channel and a
Ca2+ ATPase [25
]. Moreover, high
concentrations of phosphatidylinositol bisphosphate (PIP2)
can be found in rafts, suggesting that rafts play a role in
phospholipid signaling [136
]. Recent experiments in
mouse and human T cells have demonstrated that several molecules
localize to the detergent-insoluble fraction of T cells including the
CD4 co-receptor, GPI-linked molecules Thy-1 and ThB, the src-family
kinases p56Lck and p59fyn, the phospholipid PIP2, the
molecular adapter c-Cbl, the Syk tyrosine kinase, or the small G
protein Ras and Rho-family proteins [128
,
130
, 137
]. By contrast, it appears that the
protein tyrosine phosphatase CD45 [137
], paxillin, Grb2,
Raf1, Rab5, 
-tubulin, and the transferrin receptor CD71 are excluded
from the detergent-insoluble fraction [138
].
Recent experiments in human Jurkat T cell lines and mouse
thymocytes have shown that T cell raft integrity is required for
effective T cell receptor signal transduction [22
]. It
is intriguing to note that antigen receptor engagement triggers the
dynamic recruitment of TCR/CD3 complexes to the rafts. Moreover, Vav1,
the high (p85) and low (p55) PI3’K regulatory subunits, PLC1,
p36LAT, the TCR chain, or SHC redistribute to rafts after TCR/CD3
activation. Redistribution of the TCR/CD3 complex and signaling
molecules to the rafts is accompanied by the accumulation of a series
of tyrosine-phosphorylated substrates, in particular
hyperphosphorylated ZAP70, PLC1, LAT, and p23 TCR chains, and
increased p56lck kinase activity and these molecules appear
to form higher-order supramolecular structures. Recruitment of TCR/CD3
molecules to the rafts depends on the activity of Src family kinases
[139
]. It is important to note that recruitment of
TCR into rafts could also be observed in peptide-MHC
complex-stimulated thymocytes from female transgenic mice that
expressed a transgenic TCR recognizing the male H-Y antigen presented
by MHC class I molecule H-2Db [139
]. Conditions that
reversibly disrupt raft structure either by dispersing their contents
or by forcing their internalization reversibly disrupt the earliest
steps of T-cell activation, including decreased phosphorylation of
TCR and PLC1 and decreased Ca2+ flux after anti-CD3
stimulation [140
].
The question remains whether these rafts and localization of
signal transduction intermediates to these anatomically distinct
microdomains are regulated by TCR-triggered reorganization of the
cytoskeleton [140
]. For example, 30–40% of total
TCR protein localizes to the insoluble fraction in resting T cells
and around 90% of these TCR chains are associated with insoluble
fractions, rafts, and/or polymerized actin filaments after antigen
receptor activation [28
, 141
]. Moreover,
TCR-engagement triggers association of phosphorylated TCR chains
with actin microfilaments, and inhibition of this association abrogates
TCR/CD3-mediated cytokine production [27
,
28
]. The formation and stabilization of supramolecular
activation clusters at contact zones between T and B cells depends on a
functional actin cytoskeleton [87
]. Actin
cytoskeleton-dependent recruitment of TCR/CD3 complexes to membrane
rafts and receptor clustering are strategies utilized by T cells to
ensure organization of higher-order molecular clusters of signaling
proteins that concentrate second messenger molecules and adaptor
proteins and exclude negative regulators such as the CD45, leading to
effective TCR signaling and lymphocyte activation [49
,
137
, 140
, 142
]. It is
conceivable that anatomically and biochemically different raft
structures exist that mediate distinct functions. The question also
arises whether recruitment of antigen receptors and signaling molecules
into rafts and organization of actin-scaffolded supramolecular
activation complexes are regulated via the same signaling molecules
such as Vav1 and Rho-family GTPases?
|
FUNCTIONAL IMPLICATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
The studies reviewed above provide compelling evidence that clustering of signaling molecules, possibly in specific architectures, is an essential step in lymphocyte activation. Antigen receptor activation along with costimulation may lead to the organization of focal actin-scaffolded signaling highways whose function is to sustain TCR signaling and coordinate downstream signaling events such that complete activation is achieved and late events such as proliferation and cytokine secretion can occur. Clustering of receptors could favor sustained signaling in three ways: (1) by increasing the likelihood of contacts between the TCR and the MHC-bound ligand, the increased concentration of TCR molecules in a high density zone would allow low-affinity receptors to initiate and maintain signals; (2) by increasing the concentration of cytosolic signaling molecules and second messengers at regionally organized focal points in the proximity of the TCRs; and (3) by excluding negative regulatory molecules such as phosphatases from the zone of antigen receptor signaling.
T cells display 
30,000–40,000 TCR on the cell surface but only
50–100 peptide/MHC complexes on APCs are required for T-cell
activation [147
]. Moreover, it appears that survival of
naive peripheral CD4 and CD8 T cells is dependent on continuous
recognition of self-peptide/MHC complexes by the TCR
[148
]. Why are these T cells not constantly activated?
Ligand-induced formation of Caps and SMACs could introduce an
additional level that regulates lymphocyte effector functions. Thus,
TCR-mediated MAPK, SAPK/JNK, or NF-
B induction in the absence of
Caps and SMACs might regulate cell survival. However, formation of
higher-order clusters appears necessary to induce immune responses such
as proliferation and expression of regulatory cytokines. In this
scenario, recruitment and activation of a Cap- and SMAC-dependent
signaling cascade is probably the crucial and limiting step required
for the induction of lymphocyte effector functions.
A role for an actin-dependent clustering of signaling molecules is in line with fundamental regulatory mechanisms in signaling that might be stated in chemical terms as effective molarity, encompassing concepts of both proximity and orientation. The widespread use of dimerization domains, localization domains, and protein-protein interaction domains, as well as scaffolding proteins probably reflects the importance of effective molarity as a regulatory mechanism in intracellular signaling. Most steps in intracellular signaling can be regulated by induced proximity in specific architectures and actin-dependent clustering of signaling molecules could well serve the purpose of controlling effective molarity. Future understanding is likely to depend on the identification of the biochemical pathway(s) that are defective in the absence of Caps and SMACs and novel strategies to test the function of the molecules involved in these signaling clusters.
The clustering of receptors appears to be important for assembling
signaling molecules at focal sites, an event that may promote sustained
signaling necessary for lymphocyte activation. However, if true, why do
the kinetics of tyrosine phosphorylation and the activation of SAPK,
MAPK, and NF-B appear normal in cells that cannot form Caps? One
possible explanation is that the induction of physiological responses
may require that distinct signaling pathways be activated in a
temporally and spatially coordinated fashion. For example, oscillations
of calcium flux within a focal site regulated by receptor clustering
might be a mechanism by which multiple signaling pathways could be
coordinated. Without SMACs or Caps, individual signaling pathways may
fire but the synergy is lost. Alternatively, the formation of SMACs or
Caps may lead to the recruitment and activation of members of a unique
signaling cascade. The possibility that actin-dependent events activate
a novel signaling pathway is suggested by findings that the downstream
effects of Rac1 and CDC42 can be dissociated using different mutant
alleles [61
]. Thus, Rac1 and CDC42 mutants have been
identified that fail to activate PAK and JNK/SAPK but still lead to
cytoskeletal changes and cell cycle progression. The pathways
downstream of these Rac1/CDC42 alleles leading to proliferation and
actin changes remain to be discovered.
Signals that induce T-cell activation facilitate the formation of SMACs and lipid raft aggregation at the site of receptor contact. Negative regulators of T-cell activation could presumably play a crucial role in the regulation of cytoskeletal reorganization that results in immune synapse formation. We have identified Cbl-b, a member of the Cbl family of E3-ubiquitin ligases [149 , 150 ], as a key negative regulator of lipid raft aggregation, immune synapse formation, proliferation, and cytokine production in T cells [32 , 151 , 152 ]. T cells deficient in Cbl-b cluster their T cell receptors and lipid rafts in the absence of costimulatory signals. In addition, cbl-b-/- T cells exhibit sustained protein tyrosine phosphorylation after TCR stimulation alone, a phenomenon that has been shown to require costimulation [7 ]. These findings uncover a suppressive mechanism, mediated at least in part by Cbl-b, that is in place that forces a T cell to rely on additional signals, namely costimulatory signals, for immune synapse formation and full T-cell activation. How costimulatory signals bypass the negative regulation by Cbl-b is unknown. However, Vav1 is phosphorylated in response to TCR stimulation and this phosphorylation is enhanced when a costimulatory signal is delivered. In cbl-b-/- T cells, Vav1 is hyperphosphorylated in response to TCR stimulation alone at levels seen in cbl-b+/- cells when a costimulatory signal is delivered [151 , 152 ]. As discussed, Vav1 is a crucial regulator of cytoskeletal reorganization and perhaps once a level of Vav1 phosphorylation (or other signaling proteins) is achieved, induction of cytoskeletal reorganization is initiated, followed by T-cell activation. Therefore, phosphatases and molecules that facilitate protein degradation could play a crucial role in regulating immune synapse formation and T-cell activation by regulating levels of activated TCR proximal signaling proteins.
Activation of PKC can bypass the functional defects in T cells
deficient for Vav1, Rac1, or WASP function [35
36
37
].
Moreover, PKC activation restores the susceptibility to apoptosis of
vav1-/- thymocytes, and inhibition of a PKC
isoform inhibits TCR-mediated thymocyte apoptosis [54
].
Of the several PKC isoforms, Vav1 was found to associate only with the
Ca2+-independent PKC molecule [54
] and
Vav1 has been placed upstream of PKC activation
[153
]. PKC is highly expressed in the hematopoietic
system, particularly in T cells, and has been shown to cooperate with
calcineurin to induce transcription of the T cell growth factor IL-2
[154
]. This observation is reminiscent of the
coordinated transactivation of the IL-2 gene by calcineurin and
Vav1/Rac1 [35
, 114
]. In addition, PKC
translocates to the central areas of SMACs, whereas the PKC, ß1,
, 
, and isoforms are excluded from the contact site
[46
, 155
]. Thus, PKC is a candidate for
an effector kinase that links Cap and SMAC formation to downstream
signaling pathways involved in the development, proliferation, and
activation of T cells. PKC knock-out mice have been recently
described [156
]. These mice display normal T and B cell
development. However, peripheral T cells have a severe block in antigen
receptor-induced proliferation and IL-2 production. Biochemically, loss
of PKC results in abolished NF-B activation but normal activation
of other signaling pathways. It remains to be tested whether
PKC-deficient T cells can form SMACs and/or whether the stability
and segregation of SMACs is impaired.
In addition to the Vav1-regulated PKC-NF-B pathway,
overexpression of Vav1 activates NF-AT-dependent transcription
[114
], suggesting that NF-AT is an additional nuclear
terminus of Cap- and actin-dependent signaling. Because NF-AT and
NF-B transcription complexes contribute to the transactivation of
the IL-2 promoter, a biochemical pathway can be proposed that relays
TCR engagement to Vav1 activation, actin-polymerization, NF-AT, and
NF-B activation and expression of immune response genes (Fig. 1)
. In
Vav1- and Rac1-mutant as well as cytochalasin D-treated T cells, NFATc1
is still able to translocate from the cytoplasm to the nucleus after
TCR engagement, but the NFAT transcriptional complex is inactive
[35
]. Thus, Vav1- and Cap-dependent signaling activates
either a cascade that induces the expression of an as yet unknown
nuclear subunit of NF-AT complexes, or a unique pathway that mediates
direct NF-AT activation.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
Received September 7, 2000; revised October 22, 2000; accepted October 25, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONCYTOSKELETAL HARD-WIRING AND...LYMPHOCYTE FUNCTION AND...THE IMMUNE SYNAPSETHE MOLECULAR MOTOR OF...RHO-FAMILY GTPASES AND GUANINE...VAV1: SIGNALING AND THE...MEMBRANE RAFTSFUNCTIONAL IMPLICATIONSCONCLUSIONSREFERENCES |
|---|
