|Home | About | Journals | Submit | Contact Us | Français|
SLP-76 (SH2 domain-containing leukocyte phosphoprotein of 76 kDa) organizes signaling from immunoreceptors, including the platelet collagen receptor, the preTCR and TCR, and is required for T cell development. Here we examine a mouse in which wildtype SLP-76 is replaced with a mutant constitutively targeted to the cell membrane. Membrane-targeted SLP-76 (MTS) supports ITAM signaling in platelets and from the preTCR. Signaling from the mature TCR, however, is defective in MTS thymocytes, resulting in failed T cell differentiation. Defective thymic selection by MTS is not rescued by a SLP-76 mutant whose localization is restricted to the cytosol. Thus, fixed localization of SLP-76 reveals differential requirements for the subcellular localization of signaling complexes downstream of the preTCR versus mature TCR.
Thymocyte development is defined by a sequence of differentiation stages characterized by expression of cell surface molecules. CD4/CD8 double negative (DN) thymocytes progress through DN1 (c-kit+CD25−), DN2 (c-kit+CD25+) and DN3 (c-kit−CD25+) stages independently of the TCR (1). At the DN3 stage, signaling from the preTCR results in transition through the DN4 stage (c-kit−CD25−) to CD4/CD8 double positive (DP) cells in a process known as β-selection (2). DP thymocytes express the mature αβTCR and both CD4 and CD8. Moderate signals from the TCR promote positive selection, whereas strong TCR signaling results in negative selection (3). How signals from the preTCR and then the TCR direct maturation and selection in thymocytes and activation of mature T cells is an area of intense investigation. Although common signaling elements direct each response, how signals are integrated and translated into the appropriate outcome is not fully understood.
SH2 domain-containing leukocyte phosphoprotein of 76 kDa (SLP-76)4 is an adapter protein essential for coordinating signals downstream of immunoreceptors, including the preTCR and mature αβTCR (4). Mice deficient in SLP-76 fail to produce peripheral T cells due to a complete block at the β-selection checkpoint (5). SLP-76 resides in the cytosol of resting cells, constitutively bound to the Grb-2-related adapter protein downstream of Shc (Gads) (6). Following ligation of the TCR, activated protein tyrosine kinases (PTKs) phosphorylate the transmembrane adapter and lipid raft resident linker of activated T cells (LAT) and recruit SLP-76 to phosphorylated LAT via Gads (4). The importance of SLP-76 relocalization to LAT for TCR signaling was first demonstrated in Gads−/− mice by the absence of efficient T cell development (7). Transgenic reconstitution of SLP-76−/− mice with a SLP-76 mutant that fails to associate with Gads, and therefore cannot be recruited to LAT, fails to rescue TCR signaling or support T cell development (8).
Given the critical role of SLP-76 relocalization to LAT, we hypothesized that SLP-76 targeted to LAT would support TCR signaling. We generated a chimeric molecule containing the membrane and lipid raft-targeting sequences of LAT fused to full length SLP-76. Membrane-targeted SLP-76 (MTS) localized to membrane lipid raft domains and restored TCR signaling in SLP-76-deficient Jurkat T cells. We predicted that MTS would also support T cell development and function in a genomic knockin MTS mouse in which LCP2 (gene encoding SLP-76) is modified with the LAT membrane-targeting sequence. Surprisingly, MTS did not partition into lipid rafts in thymocytes, despite localization to cell membranes. Unlike SLP-76−/− thymocytes that arrest at the DN3 stage, MTS thymocytes successfully different into DP cells. Development is impaired at the DP to SP transition, and failure of the MTS to support generation of SP cells is associated with defects in TCR signaling. In contrast, signaling via the ITAM-containing platelet collagen receptor is completely intact in MTS mice. These studies demonstrate a genetic approach to separate the requirements for SLP-76 localization during thymic development and illustrate how different cell types may utilize adapter protein localization to direct biological responses.
Bps 1-105 of murine LAT cDNA were fused to LCP2 exon 1 and intron 1 DNA sequences. The LCP2 ATG was mutated, and the resulting fragment was cloned into the pPNT-FRT vector (M. Kahn, University of Pennsylvania) adjacent to the FRT-flanked neoR cassette. The targeting construct was electroporated into R1 ES cells, screened by Southern blot, and injected into B6×129 blastocysts. The neoR cassette was excised by breeding to FLPe mice (Jackson Laboratory).
Fetal liver was retrovirally transduced with MIGR1 (9). Recipient mice were analyzed 6-8 weeks post reconstitution. All mice were housed under pathogen-free conditions at the UPENN Animal Care Facility and used in accordance with NIH guidelines and approved protocols.
Cytosol and membranes were purified as described (10). Thymocytes were sheared and nuclei pelleted by low speed centrifugation. Cytosol was isolated as supernatant following ultracentrifugation, and NP-40 solubilization isolated cell membranes.
Thymocytes were lysed in MES buffer with 1% Triton and resuspended in MES with 40% sucrose. Sucrose solutions of 30% and 5% were overlaid, and lysates were submitted to overnight ultracentrifugation (44,000g at 4°C). 300μl fractions were taken topwise and analyzed.
Thymocytes were loaded with 2 μg/ml Indo-1, coated with biotinylated anti-CD3 (2C11) and anti-CD4 (RMA-4), and stained for CD4 (GK1.5), and CD8 for 30 min at 30°C. Ca2+ flux was triggered with streptavidin (12.5 μg/ml) and ionomycin (2 μg) as a positive control. Ca2+ release was measured by Indo-1 fluorescence.
Thymocytes were stimulated with anti-CD3 (5 μg/ml), lysed, and resolved by SDS PAGE or subjected to immunoprecipitation (IP). For GPVI stimulation, purified platelets were resuspended in Walsh buffer (11) and stimulated with convulxin (CVX). Reactions were stopped with of 2× lysis buffer. Western blotting used Abs to pY (4G10), pPLCγ2 (Y1217), PLCγ2, pSLP-76 (Y128), and actin.
Platelets were isolated as described (11). Degranulation was stimulated with CVX for 20 min at 37°C. CVX or collagen-stimulated aggregation was monitored in an aggregometer with stirring at 37°C by measuring changes in light transmission.
Since MTS partitions into lipid rafts and bypasses TCR signaling requirements for both SLP-76 and LAT in Jurkat cells (12, data not shown), we predicted that MTS would rescue SLP-76 deficiency in vivo. We generated genomic MTS knockin mice by fusing the LAT membrane targeting sequence to LCP2 exon 1. MTS/MTS mice were born at Mendelian ratios. Immunoblot analysis of MTS/+ thymocytes detected two species of SLP-76 protein: WT (76 kDa) and a slower migrating MTS (~80 kDa). Only MTS was detected in MTS/MTS cells (Fig. 1A).
T cell development in the thymus requires SLP-76-dependent signaling through both the pre- and mature TCRs (5, 13). To test if MTS supports these signals, we assessed thymocyte development in MTS mice. Unlike SLP-76−/− cells, MTS thymocytes successfully pass β-selection but fail to develop SP from DP cells (Fig. 1B). The few MTS CD8 SP cells express increased levels of heat stable antigen, indicating that they are immature (not shown). The thymic cellularity of MTS mice was 40% of WT, reflecting diminished DP, CD4 SP and CD8 SP subpopulations. MTS heterozygous animals demonstrated normal thymocyte development and function compared to WT mice.
To investigate why MTS failed to support normal T cell development, we examined MTS localization. MTS was isolated in membrane preparations (Fig. 1C). But in contrast to our findings in Jurkat, MTS failed to partition with LAT in lipid raft preparations of resting thymocytes (Fig. 1D). By direct visualization, GFP-tagged MTS localized to internal membrane structures, failing to colocalize with surface and raft markers (not shown). These findings indicate that peptide motifs may target cytosolic molecules differently in cell lines and primary cells and that SLP-76 adapter function requires different localization during preTCR and mature TCR signaling.
To understand why MTS failed to restore development of SP thymocytes, we investigated TCR signaling in thymocytes. Signals from the TCR that drive positive selection result in upregulation of TCRβ and expression of the maturation markers CD69 and CD5 (1). Analysis of MTS DP thymocytes revealed fewer TCRβhi and CD69 hi cells and reduced CD5 surface expression compared to WT. The few CD4 and CD8 SP cells that develop in MTS mice also express low levels of surface TCRβ (not shown). To directly test TCR signaling, WT or MTS thymocytes were stimulated by TCR crosslinking and analyzed biochemically. TCR-induced signals upstream of SLP-76, including phosphorylation of ZAP-70 and LAT, were unaffected or only mildly diminished in MTS thymocytes (not shown), yet tyrosine phosphorylation of the MTS itself was abolished (Fig. 2A). Strikingly, in stimulated MTS/+ thymocytes, only WT SLP-76 is phosphorylated, despite equal expression of both MTS and WT proteins.
We have shown that deletion of SLP-76 in DP thymocytes uncouples TCR engagement from PLCγ1 phosphorylation and Ca2+ mobilization (13). MTS thymocytes demonstrated diminished TCR-induced PLCγ1 phosphorylation (Fig. 2B), and Ca2+ flux was dramatically reduced in DP MTS thymocytes (Fig. 2C). TCR signaling to the MAPK pathway was also defective in MTS thymocytes, as measured by phosphorylation of ERK1,2 (Fig. 2C). These data indicate that MTS is hypophosphorylated and fails to couple the TCR to activation of effector molecules in thymocytes.
Our previous studies demonstrating that LAT is dispensable for MTS function in Jurkat (12) suggested that MTS might function independently of LAT in vivo. We thus examined the effect of LAT deletion in MTS mice, anticipating that if MTS could bypass the requirement for LAT, these mice would phenocopy LAT-sufficient MTS mice. LAT−/− mice demonstrate a complete block in thymocyte development at the DN3 stage (14). MTS expression failed to rescue this block at the DN stage (Fig. 3A).
One possible explanation for the defect in the MTS thymus is that upon reaching the DP stage, thymocytes require a cytosolic pool of SLP-76. We co-expressed with MTS a SLP-76 mutant that cannot bind Gads (G2-SLP-76) and therefore cannot be recruited via LAT to the plasma membrane. MTS fetal liver cells were retrovirally transduced with GFP only (vector) or either GFP-tagged G2-SLP76 or WT SLP-76 and were transplanted into lethally irradiated recipient mice. Following transplant, defective development was recapitulated in thymi of hosts reconstituted with MTS fetal liver cells expressing vector, while expression of WT SLP-76 rescued the defect (Fig. 3B). In contrast, expression of G2-SLP-76 failed to restore MTS thymocyte development beyond the DP stage. Thus, defective T cell development in MTS mice does not result from the absence of a cytosolic pool of SLP-76.
The failure of MTS to support TCR signaling in thymocytes raised the question of whether MTS can function in any primary cell type. Therefore, we investigated MTS platelet signaling downstream of the GPVI collagen receptor. Like the TCR, GPVI signals via an ITAM-based PTK pathway, initiated by Src and Syk kinases, and requires LAT, SLP-76 and PLCγ2 (15). Consistent with previous findings, SLP-76−/− platelets fail to upregulate surface expression of P-selectin in response to CVX (9), while WT cells demonstrate a dose-dependent response. GPVI-induced P-selectin expression was completely restored in MTS platelets (Fig. 4A). MTS and WT platelets mounted comparable aggregation responses to CVX or collagen stimulation (Fig. 4B).
We next asked if MTS is phosphorylated upon GPVI stimulation. Treatment of platelets with CVX resulted in inducible phosphorylation of both SLP-76 and MTS, as measured by 4G10 or pY128 SLP-76 immunoblotting (Fig. 4C). Furthermore, phosphorylation of PLCγ2 was induced to equal levels in MTS and WT platelets (Fig. 4D), indicating that SLP-76-dependent function downstream of the GPVI is fully restored by MTS.
We were surprised by the failure of MTS to support thymic development despite rescued TCR signaling in SLP-76−/− and LAT−/− Jurkat T cells (12). We considered that MTS may be unable to function in vivo. However, the near complete rescue of preTCR signaling and the restoration of signaling via GPVI indicated that this was not the case. Given the robust MTS GPVI response in platelets, we next speculated that augmented signaling by MTS could drive thymic negative selection. However, normal development of MTS heterozygotes argues a loss of rather than a gain of function. Our third model for failed thymocyte development was that certain SLP-76-dependent signaling may be restricted to the cytosol and cannot be organized by a membrane-bound protein. Failed reconstitution with G2-SLP-76 demonstrated that this cannot explain the MTS defect.
One hint to the failure of MTS to support thymic development is defective MTS phosphorylation in response to TCR crosslinking. Tyrosine phosphorylation of SLP-76 is required for activation of effector molecules. Following TCR stimulation, ZAP-70 relocalizes to mobile microclusters at the cell surface (17, 18), and if MTS cannot access the surface, its phosphorylation may be diminished. However, normal phosphorylation of the G2-SLP-76 mutant, despite its failure to relocalize to the cell surface, suggests that ZAP-70 may be active in multiple locations within the cell (8). It is possible that DP thymocytes restrict productive TCR signaling to the cell surface, whereas DN thymocytes and platelets may permit the assembly of multimolecular signaling complexes at intracellular sites. This is consistent with described differences in the requirements for these receptors to function. Early studies (19) documented that the preTCR could transduce signals in a ligand-independent fashion and perhaps without trafficking to the cell surface. More recent studies (20) indicate that the preTCR signals within preformed membrane complexes, while the mature TCR must be recruited into these membrane regions. Our work extends these studies by showing that SLP-76, a key integrator of both preTCR and TCR signals, supports signaling from these two receptors at different locations within the cell.
Comparing TCR and GPVI signaling in MTS mice provides additional insight contrasting the requirements for adapter localization in two primary cell systems, both of which utilize ITAMs. Platelets and early thymocytes rely on the Syk PTK for signal generation, whereas the mature TCR relies instead on ZAP-70. Others have shown that the rules for Syk versus ZAP-70 activation differ considerably (21). Concordantly, we found that MTS is inducibly phosphorylated on tyrosines following GPVI engagement on platelets but not in response to TCR ligation in thymocytes. Thus, it is possible that one other difference between Syk and ZAP-70 is where in the cell these two PTKs can act. In preliminary experiments, overexpression of Syk in MTS fetal liver cells resulted in partial restoration of SP thymocyte development and appearance of peripheral T cells.
Analysis of MTS mice has offered new understanding of how adapter protein localization regulates immunoreceptor signals. First, our data caution that targeting protein subcellular localization depends on the cell type. Our data also reinforce the notion that requirements for signaling via the preTCR and the mature TCR overlap but are not identical. Receptors in different lineages, despite shared signaling components, do not possess identical requirements for the subcellular localization of at least some critical elements. Assumptions of cell types based on discoveries in other lineages should be made carefully, despite similar paradigms for signaling. Finally, although T cell development is impaired in the MTS mice, some SP thymocytes and mature T cells arise. In preliminary analysis, rather than remaining inert due to impaired TCR signaling, these peripheral T cells are activated and skew towards production of inflammatory cytokines. How the partial rescue of TCR signaling by MTS may predispose peripheral T cells to specific effector fates is currently under investigation.
We thank the Koretzky laboratory for discussion and Taku Kambayashi for critical review of this manuscript.
1NAB was supported by a Predoctoral Fellowship from the Howard Hughes Medical Institute; RGB by the American Heart Association; MSJ by the Arthritis Foundation and the NIH; and GAK is supported by the NIH and Abramson Family Cancer Research Institute.
4Abbreviations: SLP-76, SH2 domain-containing leukocyte phosphoprotein of 76 kDa; PTK, protein tyrosine kinase; LAT, linker of activated T cells; MTS, membrane-targeted SLP-76; WT, wildtype; CVX, convulxin