Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, Canada
Correspondence: Hanne L. Ostergaard, Dept. of Medical Microbiology & Immunology, University of Alberta, Edmonton, Alberta, Canada T6G 2S2. E-mail: hanne.ostergaard{at}ualberta.ca
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: extracellular matrix protein • T cell activation • adhesion
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ß chain [2
]. Multiple integrins are expressed by T lymphocytes. ß1 integrins are expressed in murine CD4+ [3
] and CD8+ [4
] T cell clones and human CD4+ T cells [5
]. ß1 integrins mediate adhesion to fibronectin in CD4+ [3
] but not in CD8+ cells [4
, 6
]. ß3 integrins are expressed in murine cytolytic T lymphocyte (CTL) clones [4
] and ex vivo T cells [7
] and mediate binding to fibronectin and vitronectin. Integrins can generate downstream signals ("outside-in" signaling) but can also respond to cytosolic signals by altering their avidity for ligand ("inside-out" signaling) [1 , 2 ]. Integrin ligation can enhance or modulate the effects of other receptors, including growth factor receptors [8 ] and the T cell receptor (TCR) [6 , 9 10 11 ]. The nature of integrin-mediated signals is not well characterized. In fibroblasts and lymphocytes, contact with integrin ligands results in tyrosine phosphorylation of the nonreceptor protein tyrosine kinase focal adhesion kinase (FAK) [4 , 12 13 14 ]. In fibroblasts, this leads to activation of mitogen-activated protein (MAP) kinases [15 16 17 18 ]. Tyrosine phosphorylation of FAK has been shown to create binding sites for the adaptor protein Grb2 [16 ]. Grb2 can recruit the Ras guanylnucleotide-exchange factor SOS, which promotes release of guanosine 5'-diphosphate from Ras in exchange for guanosine 5'-triphosphate (GTP). The GTP-bound form of Ras recruits the kinase Raf to the membrane proximal region where Raf is activated. Activated Raf phosphorylates the MAP kinase kinase (MEK), which in turn, phosphorylates and activates the MAP kinases extracellular signal-regulated kinase (ERK)1 and ERK2. However, expression of a dominant-negative form of FAK does not block ß1 integrin-mediated activation of MAP kinase, indicating that other pathways to MAP kinase activation exist in fibroblasts [19 ]. The GTPase Rho may be involved in integrin-mediated Ras activation in certain cell types [20 ].
The avidity of integrins for ligand can be regulated independently of changes in expression level [1 ]. In T lymphocytes, TCR stimulation triggers binding of the ß2 integrin lymphocyte function-associated antigen-1 (LFA-1) to its ligand intercellular adhesion molecule-1 (ICAM-1) [1 ] and of the ß1 [very late antigen (VLA)-4 and VLA-5] and ß3 integrins to fibronectin [3 4 5 , 10 ]. This process is referred to as integrin activation. The TCR-to-integrin signals responsible for integrin activation are not well understood, although phosphatidylinositol-3 kinase (PI-3K) and the tyrosine kinase Itk have been shown to play a role in the activation of ß1 integrins [21 ], and mice lacking Slap-130/Fyb are defective in TCR-triggered binding to ICAM-1 or fibronectin [22 , 23 ]. In addition, there is evidence that MAP kinases play a role in the activation of LFA-1. Transfection of a thymocyte cell line with a dominant-negative form of p21ras was seen to inhibit TCR-triggered adhesion of LFA-1 to ICAM-1, and transfection with constitutively active ras enhanced adhesion [24 ]. Therefore, MAP kinases have been implicated in inside-out and outside-in integrin signaling.
Fibronectin can synergize with suboptimal TCR stimuli to trigger activation of CD4+ T cells [9 ] and CTL [6 , 11 ]. CTL adhesion to fibronectin is mediated by ß3 integrin [4 ]. However, the exact mechanism by which fibronectin binding promotes CTL activation is unknown. We previously found that fibronectin, vitronectin, or monoclonal antibodies (mAb) to the ß1 or ß3 integrins induce phosphorylation of the nonreceptor protein tyrosine kinases FAK and proline-rich tyrosine kinase 2 (PYK2) in murine CTL clones [4 ]. Therefore, we examined whether MAP kinase signaling occurred in these CTL downstream of integrin ligation or conversely, whether triggered binding of CTL to extracellular matrix (ECM) protein required MAP kinase activation.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) was obtained from the American Type Culture Collection (Manassas, VA). The hybridoma producing PY72 (antiphosphotyrosine) was provided by Dr. Bartholomew Sefton (The Salk Institute, La Jolla, CA). Anti-PYK2 antisera was generated in our laboratory [26
]. Anti-MAP kinase (ERK1 and ERK2) mAb was purchased from Zymed (San Francisco, CA). Phospho-p44/42 MAP kinase (Thr202/Tyr204) Ab 9101 was purchased from New England Biolabs (Beverly, MA). Anti-CD29 (ß1) and anti-CD61 (ß3) mAb were purchased from BD PharMingen (San Diego, CA). Anti-Grb2 mAb was purchased from Transduction Laboratories (Lexington, KY). The MEK inhibitor PD98059 was purchased from Calbiochem (San Diego, CA). Human cellular fibronectin was purchased from Upstate Biotechnology Inc. (Lake Placid, NY).
Protein immobilization
Various concentrations of 145-2C11, alone or in combination with 15–20 µg/ml fibronectin, were diluted in phosphate-buffered saline (PBS) and incubated overnight at 4°C in 96-well flat-bottom Nunc Maxisorp immuno-plates (Rochester, NY). Prior to use, wells were washed twice with PBS, blocked with 2% bovine serum albumin (BSA)/PBS for 30–60 min at 37°C, and then washed twice with PBS.
Cell stimulation
CTL clones were harvested and washed in Dulbecco’s PBS (D-PBS; Life Technologies, Rockville, MD). For cross-linking experiments, 145-2C11 was added at a concentration of 10 µg/ml, and all cells were incubated for 15 min on ice. Cells were pelleted by brief centrifugation and resuspended in D-PBS. Rabbit anti-hamster Ab (5 µg/ml) was added for cross-linking. Cells were aliquoted to the BSA-blocked microtiter plates at a final concentration of 1 x 105 cells/well. Cells were lysed at the indicated times by the addition of 2x Laemmli-reducing sample buffer and were boiled for 3 min.
Western blotting/MAP kinase mobility shift assay
Whole cell lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). For MAP kinase mobility shift assays, a 15% low N,N'-methylene-bis-acrylamide (175:1 acrylamide:bis) gel was used. For all other assays, standard 7.5% or 10% SDS-PAGE gels were used. Proteins were transferred to Immobilon P (Millipore Corporation, Bedford, MA). Western blotting was performed using anti-mouse horseradish peroxidase (HRP) or protein-AHRP and visualized by enhanced chemiluminesence (NEN, Boston, MA).
Degranulation assay
Degranulation was measured as described previously [25
]. Briefly, 1.5 x 105 cells in 150 µl RPMI 1640 with 2% newborn calf serum was added to each well of a 96-well microtiter plate, prepared as described above. After 4–5 h at 37°C, 25 µl supernatant was tested for benzylocarbonil-L-lysine thiobenzyl ester esterase activity. Results are the average of triplicates.
Immunoprecipitation
Cells were stimulated as described above except that fibronectin and/or 145-2C11 mAb were immobilized on Falcon 1147 nontissue culture-treated 24-well plates. Cells (1.5x106) were stimulated for each condition and lysed by the addition of an equal volume of 2x lysis buffer [2% Nonidet P-40 (NP-40), 20 mM Tris, 150 mM NaCl]. Postnuclear lysates were incubated with PYK2 antisera for 15 min on ice followed by 1–2 h incubation with protein A sepharose at 4°C to capture immune complexes. Precipitated material was washed four times in 0.5% NP-40 lysis buffer, resuspended in Laemmli-reducing sample buffer, and boiled for 3 min. Samples were analyzed by Western blotting as described above.
Adhesion assay
Adhesion was measured under static conditions by a standard method [27
]. Briefly, cells were labeled with 51Cr, then treated with PD98059 [or dimethyl sulfoxide (DMSO) carrier control] for 30 min at 37°C. Cell aliquots (100 µl) were added to wells of a 96-well MaxisorpTM plate that had been coated with 145-2C11 and/or fibronectin as described above. Plates were incubated for 1 h and then washed with cold media to remove unbound cells. After removal of all media, 125 µl SupermixTM (Wallac, Finland) was added, and 51Cr counts were measured on a Trilux 1450 MicroB liquid scintillation counter. Results were determined in triplicate, and standard deviation is shown.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (50K): [in a new window] |
Figure 1. Fibronectin is not sufficient to trigger MAP kinase activation. Clone AB.1 cells were aliquoted into microtiter wells that had been coated with 10 µg/ml 145-2C11 (anti-CD3 mAb) or 20 µg/ml fibronectin and were blocked with 2% BSA in PBS. Whole cell lysates were harvested by the addition of 2x Laemmli-reducing sample buffer. p44ERK1 and p42ERK2 MAP kinases were detected by Western blotting. Activated ERK1 and ERK2 were distinguished from the inactive forms by their reduced mobility on low bis 15% SDS-PAGE. The optical density of each band was determined using NIH Image software, and the proportion of activated ERK1 and ERK2 MAP kinases was determined. The percentage of ERK activated by immobilized BSA (open triangles, dotted line), fibronectin (closed squares), or 145-2C11 (closed diamonds) is shown. Data shown are representative of greater than three independent experiments.
|
 View larger version (32K): [in a new window] |
Figure 2. Fibronectin synergizes with CD3 ligation to facilitate MAP kinase activation and CTL degranulation. (A) Plates were prepared with the indicated concentrations of 145-2C11 in the presence or absence of fibronectin, and degranulation was measured as described in Materials and Methods. (B) Cloned AB.1 cells were added to fibronectin-coated microtiter wells in the absence or presence of a substimulatory concentration (0.1 µg/ml) of 145-2C11 (anti-CD3) mAb. ERK1 and ERK2 MAP kinases were detected from whole cell lysates by Western blotting (WB) under conditions in which active MAP kinase can be distinguished by its reduced mobility. Results shown are representative of a minimum of three independent experiments.
|
 View larger version (56K): [in a new window] |
Figure 3. Anti-integrin mAb do not cause MAP kinase activation. Cloned AB.1 cells were added to microtiter wells that had been coated with anti-CD29 (ß1 chain) or anti-CD61 (ß3 chain) mAb in the absence (-) or presence (+) of a substimulatory concentration (0.1 µg/ml) of 145-2C11 (anti-CD3) mAb. ERK1 and ERK2 MAP kinases were detected from whole cell lysates by Western blotting. Activated MAP kinase was detected by mobility shift (upper panel) and by Western blotting with Ab that recognize the activated (dual-phosphorylated) form of ERK1 and ERK2 (lower panel). Data are representative of three independent experiments.
|
 View larger version (46K): [in a new window] |
Figure 4. Phosphorylated PYK2 does not coimmunoprecipitate Grb2. Clone 11 cells were stimulated with the indicated concentrations of fibronectin and/or 145-2C11. PYK2 immunoprecipitates (IP) were analyzed by Western blotting (WB) with PY72 (antiphosphotyrosine) and anti-Grb2 mAb. Preimmunoprecipitation postnuclear supernatant (PNS) was also probed for Grb2 (bottom panel). Result is representative of three independent experiments.
|
 View larger version (27K): [in a new window] |
Figure 5. MAP kinase activation is not essential for triggered adhesion to fibronectin. Wells of a Nunc MaxisorpTM plate were coated with 20 µg/ml fibronectin or the indicated concentrations of 145-2C11 alone or in combination. Cell adhesion was measured in the presence or absence of 5 µg/ml MEK inhibitor PD98059, as described in Materials and Methods. Result is representative of three independent experiments.
|
 View larger version (103K): [in a new window] |
Figure 6. PD98059 inhibits triggered cell-spreading on fibronectin (FN). Cloned AB.1 cells were plated in microtiter wells coated with 0.1 µg/ml 145-2C11 (A–D and I–K), 20 µg/ml fibronectin (B–E and J–L), or 10 µg/ml 145-2C11 (F–H, M, and N). Cells were treated with DMSO carrier control (A, B, E, and F) or with 5 µg/ml PD98059 (C and G) or 10 µg/ml PD98059 (D and H). Cells were observed after 1 h of incubation. Examples of cells exhibiting highly spread and flattened morphology are indicated by solid arrowheads. Spread and nonspread cells were readily distinguished, as seen in the magnified views shown in panels I–N. (I–N) Magnified views of panels A–C and E–G, respectively. Additionally, attachment to fibronectin was triggered by stimulation of AB.1 cells in suspension with 10 µg/ml 145-2C11 mAb and anti-hamster immunoglobulin (Ig) cross-linking Ab (XL) before addition of cells to fibronectin-coated plates in the presence of DMSO carrier control (O) or 15 µg/ml PD98059 (P). (O) Highly spread cells are highlighted by arrowheads. Data are representative of three independent experiments. (Q) Cell diameters were measured on equivalently enlarged photographs of AB.1 CTL plated on fibronectin and coimmobilized 0.1 mg/ml 145-2C11 for 60 min in the presence of DMSO solvent control (closed diamonds), 10 µg/ml PD98059 (open squares), or 10 µM U0126 (open triangles). A minimum of 75 cells was measured under each condition.
|
4-integrin-mediated cell-spreading was inhibited by PD98059 but not by another MEK inhibitor U0126 [33
]. In agreement with that report, we found that cell-spreading on fibronectin, triggered by coimmobilized or soluble, cross-linked 145-2C11 anti-CD3 mAb, was not blocked by 10 µM U0126. The average diameter of cells plated on coimmobilized fibronectin and 0.1 µg/ml 145-2C11 was not significantly different between U0126-treated and control-treated cells (Fig. 6Q)
, although ERK activation was blocked (data not shown). Therefore, the effects of PD98059 on triggered cell-spreading cannot be attributed to a requirement for ERK. These data indicate that ERK activation is not required for triggered adhesion or spreading on fibronectin and that triggered adhesion can be uncoupled from increased cell-spreading. |
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Integrin affinity for ligand is not constitutively high but is regulated [1 , 2 ]. Therefore, integrin-mediated MAP kinase signaling might not occur until integrin avidity was first activated by signals from the TCR. Two observations argue against this explanation. First, direct integrin ligation by antibodies did not result in MAP kinase signals (Fig. 3) . Second, the presence of fibronectin or integrin ligation by mAb resulted in tyrosine phosphorylation of FAK and PYK2, indicating that integrins were functionally engaged (Fig. 4 , and additional data not shown) [4 , 11 ].
In Jurkat T lymphoma and PC12 neuronal cells, tyrosine phosphorylation of PYK2 results in recruitment of Grb2 and consequent activation of MAP kinases downstream of Ras [30 31 32 ]. Our present study shows that in cloned CTL, ß1 and ß3 integrins trigger PYK2 tyrosine phosphorylation but are uncoupled from MAP kinase activation, likely at the point of Grb2 recruitment. In CTL, PYK2 might preferentially bind partners other than Grb2. This may be related to the expression of an alternatively spliced form of PYK2 in most hematopoietic cells, which lacks a 42 amino acid portion of the C-terminal domain, whereas Jurkat T cells and nonhematopoietic cells express full-length PYK2 [34 35 36 ].
The Ras-MAP kinase pathway has been reported to be involved in triggering avidity of the ß2 integrin LFA-1 for its ligand ICAM-1 [24 ]. To determine whether this pathway is important for ß3-mediated triggered binding to fibronectin, adhesion assays were performed in the presence or absence of the MEK inhibitor PD98059. This inhibitor only slightly inhibited triggered binding to coimmobilized fibronectin and anti-CD3 mAb (Fig. 5) . This slight decrease might be attributed to reduced cell-spreading in the presence of PD98059 (Fig. 6) . It is interesting that a combination of PD98059 and the PI-3K inhibitor wortmannin is required to completely inhibit the triggered binding of LFA-1 to ICAM-1 in mature CD4+ T cells [24 ].
The relationship between MAP kinase and cell-spreading was also investigated. MAP kinase activity does not appear to be necessary for the extensive cell-spreading that occurs on high concentrations of immobilized anti-CD3 mAb (Fig. 6 , F–H, M, and N), a result also reported for Jurkat T cells [37 ]. In the case of strong TCR stimulus, the exchange factor Vav is thought to link TCR stimulation to the Rho family kinases Rac, CDC42, and RhoA, which in turn, mediate cytoskeletal changes [38 ]. Lack of ERK involvement in cell-spreading on high-concentration TCR/CD3 ligand is consistent with this model. In contrast, although PD98059 did not block triggered adhesion to fibronectin (Fig. 5) , it did inhibit triggered cell-spreading on fibronectin (Fig. 6) . A similar uncoupling of spreading and adhesion has been observed in REF52 cells adhering to fibronectin in the presence of serum [39 ], where MEK inhibition prevented cell-spreading while not altering net adhesion. It is interesting that ERKs have been observed to colocalize with integrins at points of ECM contact in fibroblasts [17 , 39 ]. However, in agreement with a recent report [33 ], triggered T cell spreading on fibronectin was not inhibited by another MEK inhibitor U0126 (Fig. 6Q) , although ERK activation was blocked by this compound (data not shown). This implies that the effect of PD98059 on cell-spreading is not a result of blockage of ERK activation. PD98059 is thought to bind unphosphorylated MEK and prevent its activation [40 ], whereas U0126 can bind activated MEK and prevent it from phosphorylating ERK1/2 [41 ]. This raises the possibility that phosphorylated MEK may have a role in cell-spreading that is independent of its role in ERK activation, perhaps via protein–protein interactions. However, the possibility that PD98059 has an unanticipated effect on an unknown target that regulates cell morphology cannot be ruled out.
Coimmobilization of anti-CD3 mAb with fibronectin or anti-integrin antibodies facilitated CTL activation by amounts of anti-CD3 mAb that would otherwise be substimulatory. At least three possible mechanisms may account for the synergy between fibronectin and TCR/CD3 ligation: increased TCR/CD3 occupancy, integrin-mediated signals that cooperate with CD3 signaling, or nonintegrin signals that result from cell anchorage that cooperate with CD3 signaling. Increased TCR/CD3 occupancy probably makes a minor contribution. The presence of fibronectin facilitated degranulation at tenfold lower concentrations of anti-CD3 (Fig. 2)
, whereas fibronectin-mediated cell-spreading only increases the surface area contacted by 
2.25-fold (Fig. 6)
. However, fibronectin-mediated adhesion probably also increases the duration of contact with CD3 ligand. Therefore, increased TCR/CD3 occupancy may contribute to CTL activation. The presence of fibronectin can enhance degranulation from CTL stimulated by anti-CD3-coated beads, although the integrin and CD3 ligands are on different surfaces in this situation [11
]. This observation is most consistent with cooperation via signaling rather than cooperation via increased TCR/CD3 occupancy. CD3 and integrin ligation induce phosphorylation of FAK and PYK2 [4
, 14
, 26
], suggesting that synergism can occur at the level of these kinases. Third, it is possible that CD3 signaling is enhanced as an indirect consequence of integrin-mediated cell anchorage. In several systems, it has been demonstrated that cell shape rather than integrin ligation per se is the requisite factor for responsiveness to other stimuli [42
, 43
].
In summary, although stimulation of ß1 and ß3 integrins leads to phosphorylation of FAK and PYK2, this does not result in MAP kinase activation in CTL as it does in other types of cells. Rather integrin ligation facilitated strong ERK activation in response to levels of TCR/CD3 stimulation that would otherwise have been substimulatory. MAP kinase activation was not essential for TCR-triggered, ß3-mediated adhesion to fibronectin. Therefore, in addition to facilitating adhesion to target cells, integrin ligation may promote CTL activation by facilitating cell-spreading and increased target contact and by initiating intracellular signals that cooperate synergistically with TCR/CD3-mediated signals upstream of MAP kinase.
Received April 22, 2002; revised November 7, 2002; accepted December 5, 2002.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4 integrin signaling activates phosphatidylinositol 3-kinase and stimulates T cell adhesion to intercellular adhesion molecule-1 to a similar extent as CD3, but induces a distinct rearrangement of the actin cytoskeleton J. Immunol. 168,696-704
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
