The CD3 zeta subunit contains a phosphoinositide-binding motif that is required for the stable accumulation of TCR-CD3 complex at the immunological synapse

Abstract

T cell activation involves a cascade of TCR-mediated signals that are regulated by three distinct intracellular signaling motifs located within the cytoplasmic tails of the CD3 chains. Whereas all the CD3 subunits possess at least one ITAM, the CD3 ε subunit also contains a proline-rich sequence and a basic-rich stretch (BRS). The CD3 ε BRS complexes selected phosphoinositides, interactions that are required for normal cell surface expression of the TCR. The cytoplasmic domain of CD3 ζ also contains several clusters of arginine and lysine residues. In this study, we report that these basic amino acids enable CD3 ζ to complex the phosphoinositides PtdIns(3)P, PtdIns(4)P, PtdIns(5)P, PtdIns(3,5)P(2), and PtdIns(3,4,5)P(3) with high affinity. Early TCR signaling pathways were unaffected by the targeted loss of the phosphoinositide-binding functions of CD3 ζ. Instead, the elimination of the phosphoinositide-binding function of CD3 ζ significantly impaired the ability of this invariant chain to accumulate stably at the immunological synapse during T cell-APC interactions. Without its phosphoinositide-binding functions, CD3 ζ was concentrated in intracellular structures after T cell activation. Such findings demonstrate a novel functional role for CD3 ζ BRS-phosphoinositide interactions in supporting T cell activation.

FIGURE 1

The cytoplasmic tail of CD3 ζ contains multiple clusters of arginine and lysine residues that complex phosphoinositides. A, Amino acid sequence of the cytoplasmic tail of murine CD3 ζ containing clusters of basic amino acids, as indicated by asterisks. Individual biotinylated peptides containing the three BRS sequences (BRS-1, BRS-2, BRS-3) and the three ITAMs (ITAM-1, ITAM-2, ITAM-3) are shown. The tyrosine residues present in the ITAMs are underlined. B, PIP-strips embedded with diverse lipids were probed with the biotinylated peptides. Binding was detected by streptavidin-HRP western blotting, with the different exposure times indicated below the membrane. Data are representative of 4 independent experiments. C, Biotinylated peptides containing ITAM-1, BRS-2, and BRS-3 were used to immunoblot lipid arrays containing serial dilutions of the indicated phospholipids ranging from 100 to 1.56 pmoles, as shown in the left panel. Peptide binding was assessed as in B. The data are representation of at least 4 independent binding assays per peptide.

FIGURE 2

A single polybasic cluster in the cytoplasmic tail of CD3 ζ complexes selected phosphoinositides with high affinity. A, Amino acid sequence of the cytoplasmic domain of wild type CD3 ζ and constructs containing a number of amino acid substitutions at various polybasic cluster positions. These include substitutions of 11 (BRS Sub-A), 9 (BRS Sub-B), and 4 residues (BRS Sub-C), respectively. The substitutions are bolded and underlined. B, PIP-arrays containing serial dilutions of the indicated phosphoinositides were probed with the purified GST-ζ fusion proteins. Binding was detected by anti-GST western blotting assays. Binding assays with the substituted constructs are shown with 2 min (BRS Wild Type) and 30 min exposures (BRS Sub-A, -B, -C). The data are representation of 3 independent experiments. C, Sucrose-loaded liposomes, consisting of the indicated phospholipids, were incubated with GST (lanes 1-4), or GST-ζ Wild Type (lanes 5-8), or GST-BRS Sub-C BRS (lanes 9-12). Proteins retained in the supernatant (S) or pellet (P, liposome binding fraction) were detected by SDS-PAGE analysis using Coomassie Brilliant Blue staining. The percent of GST fusion protein detected in the supernatant and pellet are listed under the appropriate lanes. These results were obtained in two independent experiments.

FIGURE 3

The tyrosine phosphorylation of the CD3 ζ ITAMs is uncoupled from the lipid-binding functions of ζ. A, HEK 293 cells were co-transfected with Lck and control vector (lane 1), or the various CD3 ζ constructs including wild type CD3 ζ (lane 2), BRS Sub-A (lane 3), BRS Sub-B (lane 4), and/or BRS Sub-C (lane 5). The cells were processed for immunoblotting with anti-phosphotyrosine. B and C, CD3 ζ was immunoprecipitated from the lysates prepared in A, and immunoblotted with anti-phosphotyrosine (B) and anti-CD3 ζ mAbs (C). D-F, The cells were transfected as in A along with an expression vector containing the tandem SH2 domains of ZAP-70. D, Samples were processed for immunoblotting of total cell lysates with anti-phosphotyrosine mAbs. E-G, CD3 ζ was immunoprecipitated from the cell lysate and blotted with mAbs against phosphotyrosine (E), CD3 ζ (F), and ZAP-70 (G). The results were confirmed in 5 independent experiments.

FIGURE 4

TCR-mediated activation of early TCR signaling processes is independent of the CD3 ζ-lipid binding interactions. CD3 ζ-deficient T cells (BWδ) were reconstituted with wild type ζ or the BRS Sub-C construct. Clones, selected for equivalent TCR expression, were left untreated or stimulated with anti-CD3 mAbs for the indicated times (0, 3, 10, and 30 min). Total cell lysates (A) or immunoprecipitates of ZAP-70 (B) and SLP-76 (C) were prepared for immunoblotting with anti-phosphotyrosine. In addition, anti-phospho-specific antisera and antibodies against a number of additional proteins were used to immunoblot total cell lysates. These included anti-phospho-PLC γ and anti--PLC γ (D), anti-phospho-PKC θ (E), anti-phospho-AKT and anti-AKT (F), anti-phospho-Erk and anti-Erk (G), and anti-actin (H) and anti-CD3 ζ (I). Data are representative of three independent experiments.

FIGURE 5

The spatiotemporal accumulation of CD3 ζ at the immunological synapse is partially regulated by its phospholipid-binding properties. 5C.C7 (A-C) or D011.10 (D, E) TCR transgenic T cells were retrovirally transduced with constructs encoding the wild type CD3 ζ or BRS Sub-C, both linked to GFP. The GFP-positive cells were imaged after incubations with 10 μM peptide loaded APCs (MCC). A, Representative interaction of CD3 ζ Sub-C-GFP 5C.C7 T cells with the APC is shown at the indicated time points (sec) relative to the time of formation of a tight cell couple. Differential interference contrast (DIC) images are shown in the top row, with top-down, maximum projections of 3-dimensional ζ Sub-C-GFP fluorescence data in the bottom row. The fluorescence intensity is displayed in a rainbow-like false-color scale (increasing from blue to red). A movie covering the entire time frame is available in supplemental movie S1. B-C, The graphs display the percentage of cell couples from experiments similar to (A) with accumulation of wild type CD3 ζ-GFP (B), Sub-C-GFP (C). Any accumulation refers to ζ-GFP molecules that are grouped at the entire interface formed between the T cell and APC. Central accumulation refers to the clustering of ζ -GFP at the very center of the T cell-APC interface. Peripheral accumulation is for the clustering of ζ-GFP molecules at the outside ring of the T cell-APC interface. Finally, distal accumulation is evident when there is a clustering of ζ -GFP at the distal pole of the T cell, opposite to the area of T cell-APC contact. These patterns are described in detail elsewhere (15, 24). (D, E) DO11.10 T cells were mixed with A20 B cell lymphoma APCs in the presence of 10 μM Ova peptide. The graphs display the percentage of cell couples with accumulation of wild type ζ-GFP (D) or BRS Sub-C-GFP (E) as described in (A-C). A total of 45, 90, 85, 50 cell couples were analyzed in B, C, D, and E, respectively. Differences in accumulation in any interface pattern between wild type and BRS Sub-C were significant (p < 0.001) at all time points ≥ 100s; differences in central accumulation were significant (p < 0.005) at all time points ≥ 40 s. A movie covering the entire time frame for the OTII cells is available in supplemental movie S2.