Report

Positive Regulation of T Cell Activation and Integrin Adhesion by the Adapter Fyb/Slap

See allHide authors and affiliations

Science  21 Sep 2001:
Vol. 293, Issue 5538, pp. 2260-2263
DOI: 10.1126/science.1063397

Abstract

The molecular adapter Fyb/Slap regulates signaling downstream of the T cell receptor (TCR), but whether it plays a positive or negative role is controversial. We demonstrate that Fyb/Slap-deficient T cells exhibit defective proliferation and cytokine production in response to TCR stimulation. Fyb/Slap is also required in vivo for T cell–dependent immune responses. Functionally, Fyb/Slap has no apparent role in the activation of known TCR signaling pathways, F-actin polymerization, or TCR clustering. Rather, Fyb/Slap regulates TCR-induced integrin clustering and adhesion. Thus, Fyb/Slap is the first molecular adapter to be identified that couples TCR stimulation to the avidity modulation of integrins governing T cell adhesion.

Engagement of the TCR by antigen initiates intracellular signaling cascades that result in T cell activation, differentiation, acquisition of effector function, or apoptosis (1). Dramatic changes to the cytoskeleton and transcription are the outcomes of these cascades, but the molecular mechanisms that link TCR stimulation to downstream pathways are not well understood. Signals from cell surface receptors are coupled to distal signaling pathways by adapter proteins, molecules that contain multiple protein-protein interaction domains. Several adapter molecules have been identified that have positive effects on TCR signaling, including Grb2, LAT, GADS, and Slp-76 (2, 3). Others, such as c-Cbl and Cbl-b, have a negative regulatory function.

Fyb/Slap is an adapter protein expressed in T cells and myeloid cells that contains a proline-rich region, tyrosine phosphorylation sites, an EVH1-binding domain,and an SH3-like domain (4–6). It is expressed as two isoforms of 120 kD and 130 kD, which differ in expression pattern between thymocytes and mature T cells (7). Fyb/Slap binds to the Src family kinase Fyn, the adapter protein Slp-76, which is required for thymocyte development, TCR-dependent mitogen-activated protein kinase (MAPK) activation and calcium flux (8–10), and Ena/VASP proteins (6), which are modulators of the cytoskeleton. The functional role of Fyb/Slap has remained controversial because overexpression studies have shown Fyb/Slap to be either a positive regulator (4, 7, 11) or a negative regulator (5, 12) of interleukin-2 (IL-2) production, depending on the experimental system used.

To elucidate Fyb/Slap's biological role in T cell function, we used Rag2 −/− blastocyst complementation to generate gene-targeted chimeric mice lacking Fyb/Slap expression in mature T and B lymphocytes (13).Fyb/slap −/− chimeras displayed normal numbers and differentiation of thymocytes and B cells; however, the relative numbers of CD4+ T cells in the lymph nodes and CD4+ and CD8+ T cells in the spleen were reduced in fyb/slap −/− chimeras as compared tofyb/slap +/− chimeric controls (13). These results indicate that lymphocyte development is normal in the absence of Fyb/Slap but that this adapter may be required for peripheral T cell homeostasis in vivo.

To determine whether Fyb/Slap positively or negatively influences lymphocyte function, we measured the proliferative responses of purified lymphocytes from fyb/slap +/− andfyb/slap −/− chimeric mice in vitro (14). Consistent with the lack of expression of Fyb/Slap in B cells (4), B cell proliferation was comparable between Fyb/Slap-deficient and control animals. Purified T cells from fyb/slap +/− andfyb/slap −/− chimeras responded equally to phorbol 12-myristate 13-acetate (PMA) and calcium ionophore (Fig. 1, A and B). However, stimulation with monoclonal antibody (mAb) to CD3ɛ (anti-CD3ɛ) alone, or anti-CD3ɛ plus antibody to CD28 (anti-CD28) mAbs, revealed a dramatic decrease in the proliferation offyb/slap −/− T cells compared tofyb/slap +/− T cells at both 24 hour and 48 hour time points (Fig. 1, A and B). Using the same stimulation conditions, IL-2 (Fig. 1C) and interferon-γ (IFN-γ) (Fig. 1D) production were also strongly reduced in fyb/slap −/− T cells. In addition, the up-regulation of CD25 (IL-2 receptor α chain) and CD69 (Fig. 2E) was defective in Fyb/Slap-deficient T cells. Consistent with the impaired expression of the high affinity IL-2R, addition of exogenous IL-2 only partially rescued the proliferative defect infyb/slap −/− T cells. These data demonstrate that Fyb/Slap is an important positive regulator of T cell activation and proliferation and of cytokine production.

Figure 1

Fyb/Slap is a critical regulator of T cell activation. (A and B) Proliferation of purified lymph node T cells. Cells were stimulated with plate-bound anti-CD3ɛ (0.75 μg/ml or 0.25 μg/ml), anti-CD3ɛ (0.25 μg/ml) and soluble anti-CD28 (1 μg/ml), or with PMA (50 ng/ml) plus ionomycin (100 ng/ml). Proliferation was determined by [3H]thymidine incorporation after 24 hours (A) or 48 hours (B). (C andD) Cytokine production. T cells were stimulated as in (A) for 48 hours (C) or 24 hours (D). Levels of IL-2 (C) and IFN-γ (D) production were determined by ELISA. (E) Activation marker up-regulation. Purified fyb/slap +/− (solid line) and fyb/slap −/− (dotted line) T cells were stimulated as indicated for 48 hours (CD25, top) or 24 hours (CD69, bottom) and analyzed by flow cytometry. (F) Impaired response to TD antigen. Fyb/slap +/− (solid squares) and fyb/slap −/− (open squares) chimeric mice (two mice per genotype) were immunized with NP-OVA intraperitoneally (ip), and serum anti-NP IgG1 titers were determined by ELISA. Each curve represents one individual mouse. One result representative of two independent experiments is shown. (G) Normal response to TI antigen.Fyb/slap +/+ (X),fyb/slap +/− (solid squares) andfyb/slap −/− (open squares) chimeric mice (two mice per genotype) were immunized with TNP-Ficoll ip, and IgG3 titers were determined by ELISA. Each curve represents one individual mouse.

Figure 2

Normal TCR proximal signaling but impaired integrin clustering in the absence of Fyb/Slap. (A) Calcium mobilization. Freshly isolated INDO-1 loaded peripheral T cells from fyb/slap +/− andfyb/slap −/− mice were stimulated with anti-CD3ɛ. Ca2+ flux was measured by flow cytometry. The x axis shows real time Ca2+release followed for 180 s, and the y axis shows the intracellular Ca2+ concentration. Arrowheads indicate the time of addition of cross-linking antibody. (B) Activation of Erk1 and Erk2. Purified fyb/slap +/−and fyb/slap −/− T cells were activated with hamster anti-CD3ɛ alone or in conjunction with hamster anti-CD28 for 2, 5, (shown) or 10 min. Active Erk1 and Erk2 were detected using a phosphospecific Ab (p-ERK). Levels of total Erk1 and Erk2 protein are shown in the lower panel. Normal Erk activation was confirmed using in vitro kinase assays. (C) TCR-induced CD3 clustering, actin polymerization, and LFA-1 clustering. T lymphocytes were incubated with anti-CD3ɛ and then with biotinylated cross linker at 4°C (control) or 37°C. Cells were cytospun onto slides and fixed with 3.7% PFA. TCR-CD3 clustering was visualized using fluorescein isothiocyanate (FITC)–labeled streptavidin and fluorescence microscopy. F-actin was visualized with tetramethyl rhodamine isothiocyanate (TRITC)–phalloidin, and LFA-1 with a FITC Ab to LFA-1. Representative images after activation for 30 min at 37°C are shown. (D) TCR-induced CD3 clustering, F-actin polymerization, and LFA-1 clustering. Cells were stimulated as in (C). Clustering was quantified using a fluorescent microscope (10 fields counted per treatment). Cells were considered to have clustered receptors if the staining pattern showed receptors polarized to one side of the cell.

To investigate whether the requirement for Fyb/Slap in T cell activation and proliferation in vitro translated into a defective immune response in vivo, we immunized the chimeric mice with the T cell–dependent (TD) antigen 4-hydroxy-3-nitrophenylacetyl–ovalbumin (NP-OVA) (15). Whereas fyb/slap +/− chimeras produced high titers of NP-specific immunoglobulin G1 (IgG1) Ab, the IgG1 response was completely absent infyb/slap −/− chimeras (Fig. 1F). Moreover, in mice infected with vesicular stomatitis virus (VSV), virus-specific IgG titers showed a gene-dosage dependent decrease infyb/slap +/− andfyb/slap −/− chimeras as compared tofyb/slap +/+ controls (15). The block in antibody production was due to a defect in T cells and not B cells, because fyb/slap −/−,fyb/slap +/−, andfyb/slap +/+ mice produced comparable 2,4,6-trinitrophenyl (TNP)–specific IgG3 after immunization with the T cell–independent (TI) polyvalent antigen TNP-Ficoll (Fig. 1G). We conclude that Fyb/Slap is essential for the delivery of effective T cell help and for the generation of functional immune responses to TD antigens in vivo.

To identify the molecular mechanism that accounts for the functional defects in fyb/slap −/− T cells, we analyzed signaling pathways downstream of TCR stimulation.Fyb/slap +/− andfyb/slap −/− T cells exhibited comparable calcium fluxes after anti-CD3ɛ crosslinking (Fig. 2A) (14). Moreover, stimulation offyb/slap +/− andfyb/slap −/− thymocytes or peripheral T cells with anti-CD3ɛ with or without anti-CD28 revealed no apparent differences in total tyrosine phosphorylation or in expression levels or tyrosine phosphorylation of LAT, PLCγ1, Slp-76, c-Cbl, Vav1, Fyn, or Lck (14). Both Fyn and Lck kinase activities were normal. Lastly, the phosphorylation kinetics and activities of Erk1/Erk2, SAPK/JNK, and p38 were also comparable betweenfyb/slap +/− andfyb/slap −/− peripheral T cells (Fig. 2B) (14). Thus, despite the requirement of Fyb/Slap for T cell activation and proliferation, the major signaling pathways downstream of the TCR appear intact in the absence of Fyb/Slap.

Overexpression studies have suggested a role for Fyb/Slap in cytoskeletal rearrangement downstream of the TCR (6). In addition, Fyb/Slap co-localizes with Ena/VASP proteins that have been shown to control cytoskeletal reorganization, cell motility, and cell adhesion (6,16–18). Clustering of the TCR/CD3 chains after T cell stimulation is required for efficient T cell activation and is dependent on the actin cytoskeleton (19). To further explore the molecular mechanism by which Fyb/Slap regulates T cell activation and immune responses, we investigated whether Fyb/Slap was involved in TCR-induced CD3 clustering or actin polymerization (14). The extent of TCR-mediated antigen receptor clustering was comparable betweenfyb/slap +/− andfyb/slap −/− T cells after anti-CD3ɛ stimulation (Fig. 2, C and D). TCR-induced F-actin clustering was also comparable between fyb/slap +/− andfyb/slap −/− T cells (Fig. 2, C and D). Thus, Fyb/Slap is not required for TCR-triggered antigen receptor clustering and actin polymerization.

In addition to clustering of the TCR itself, a second cytoskeleton-dependent process triggered by the TCR is the clustering of integrin receptors leading to T cell adhesion (20). Adhesion of T cells via integrin receptors is critical for T cell development, migration, and activation, and integrins such as LFA-1 have been demonstrated to be important for in vitro T cell proliferation (21, 22). In resting T cells, integrins are maintained in a relatively inactive state. TCR ligation initiates a poorly understood process of “inside-out” signaling that induces ligand binding (23, 24). Inside-out signaling and integrin binding are required to stabilize the interaction between T cells and antigen-presenting cells and to form the immune synapse (25). TCR-induced clustering of the β2 integrin LFA-1 in the cell membrane increases integrin avidity for its ligand, intercellular adhesion molecule−1 (ICAM-1), and is an important mechanism for the induction of lymphocyte adhesion (20). To examine whether Fyb/Slap plays a role in integrin clustering, we assessed LFA-1 distribution on the cell surface after CD3ɛ stimulation (14). The proportion offyb/slap −/− T cells that exhibited LFA-1 clusters after stimulation was markedly reduced compared tofyb/slap +/− T cells (Fig. 2, C and D). The clustering defect in Fyb/Slap-deficient cells is not due to a change in clustering kinetics, because fyb/slap −/− T cells also exhibit defective clustering at 15 min or 45 min, whereas maximal clustering of fyb/slap +/− T cells is seen at 30 min. Thus, Fyb/Slap plays a specific role in coupling TCR stimulation to LFA-1 lateral mobility in the membrane.

To examine whether the defect in LFA-1 clustering seen infyb/slap −/− T cells has functional relevance for integrin-mediated adhesion, we stimulatedfyb/slap +/− andfyb/slap −/− T cells with anti-CD3ɛ and assayed adhesion to plates coated with recombinant mouse ICAM-1 (mICAM-1) and human ICAM-2 (hICAM-2) (26). Adhesion to hICAM-2 was performed using a parallel plate flow chamber that produces a reproducible force on adherent cells and quantifies the strength of cell adhesion. Fyb/slap −/− T cells stimulated with anti-CD3ɛ were defective in adhering to mICAM-1 and hICAM-2 (Fig. 3, A and B). TCR-induced adhesion mediated by other integrins was also defective in fyb/slap −/− T cells. This was the case with adhesion to fibronectin, mediated by several integrins including α5β1 and αvβ3, and adhesion to mouse vascular cell adhesion molecule–1 (mVCAM-1), mediated by α4 integrins (Fig. 3, C and D). The cell surface expression of LFA-1 (CD11a), CD18, CD11c, α4 integrin, LFA-2, ICAM-1, L-selectin, platelet endothelial cell adhesion molecule–1 (PECAM-1), and CD30 was not reduced infyb/slap −/− T cells compared tofyb/slap +/− controls. In all cases, treatment with PMA, magnesium, or manganese induced comparable adhesion to controls, indicating that the defect in TCR-induced adhesion is due to a block in inside-out signaling between the TCR and integrins and is not due to an intrinsic defect of the integrins (Fig. 3, A through D). Thus, in the absence of Fyb/Slap, communication between the antigen receptor and multiple integrins is impaired, inhibiting the coupling of TCR signaling to T cell adhesion.

Figure 3

Fyb/Slap is required for TCR-induced integrin-mediated adhesion. (A) T cell adhesion to mICAM-1. Purified fyb/slap +/− andfyb/slap −/− T cells were cultured on plates coated with mICAM-1. Adherent cells were counted by trypan blue exclusion. Adhesion of untreated samples to mICAM-1 was 10% of cells plated for both control and fyb/slap −/− cells, indicating that loss of Fyb/Slap did not affect baseline adhesion. (B) Adhesion of T cells to hICAM-2 under shear stress. Purifiedfyb/slap +/− andfyb/slap −/− T cells were treated as indicated and allowed to settle on hICAM-2 in a parallel plate flow chamber for 10 min at 37°C. Shear stress was then applied and the percentage of cells remaining attached after each shear stress interval determined. For both (B) and (D), adhesion at 10 dynes/cm2 of shear stress is shown, and comparable results were obtained at 4 dynes/cm2 of shear force. (C) T cell adhesion to fibronectin. Purified fyb/slap +/− andfyb/slap −/− T cells were cultured on plates coated with fibronectin or BSA (control) and adhesion assays carried out as in (A). Adhesion of untreated cells to fibronectin was 10% of cells plated for both groups. (D) Adhesion of T cells to mVCAM-1 under shear stress. Adhesion assays were performed as described in (B).

In conclusion, we have shown both in vitro and in vivo that Fyb/Slap is a critical positive regulator of T cell activation and function. Fyb/slap −/− T cells are defective in TCR-mediated activation, proliferation, and cytokine production. In vivo,  fyb/slap −/− chimeric mice display impaired immunity to TD antigens. Functionally, Fyb/Slap has no apparent role in the activation of known TCR signaling pathways, TCR clustering, or F-actin formation. Rather Fyb/Slap controls communication between the TCR and integrins and regulates integrin clustering and integrin-mediated adhesion in response to TCR stimulation. Thus, Fyb/Slap is the first molecular adapter to be identified that couples TCR signaling to integrin activation, inside-out signaling, and T cell adhesion. These data also demonstrate that integrin clustering can be genetically uncoupled from actin reorganization and define a previously unknown Fyb/Slap-regulated signaling pathway that specifically links TCR stimulation to changes in integrin avidity. On the basis of these studies and other work implicating Fyb/Slap in vasoactive mediator release (27), we propose that Fyb/Slap be redesignated Adhesion and Degranulation promoting Adapter Protein (ADAP).

  • * To whom correspondence should be addressed. E-mail: jpenning{at}amgen.com

REFERENCES AND NOTES

View Abstract

Navigate This Article