Report

Requirement for the SLP-76 Adaptor GADS in T Cell Development

See allHide authors and affiliations

Science  09 Mar 2001:
Vol. 291, Issue 5510, pp. 1987-1991
DOI: 10.1126/science.1057176

Abstract

GADS is an adaptor protein implicated in CD3 signaling because of its ability to link SLP-76 to LAT. A GADS-deficient mouse was generated by gene targeting, and the function of GADS in T cell development and activation was examined. GADSCD4CD8 thymocytes exhibited a severe block in proliferation but still differentiated into mature T cells. GADS thymocytes failed to respond to CD3 cross-linking in vivo and were impaired in positive and negative selection. Immunoprecipitation experiments revealed that the association between SLP-76 and LAT was uncoupled in GADS thymocytes. These observations indicate that GADS is a critical adaptor for CD3 signaling.

The development and function of T cells are regulated by signaling through the CD3 complex, which serves both the pre–T cell receptor (pre-TCR) and the TCR [(1) and references therein]. Cross-linking of CD3 induces protein tyrosine phosphorylation in a wide range of proteins. Among these phosphorylation targets are two adaptor proteins, LAT and SLP-76, which function in a coordinated fashion to activate a diverse set of signaling proteins (2–5). The critical function of SLP-76 and LAT is supported by the observation that mice lacking SLP-76 or LAT exhibit an absolute block in early thymocyte development (6–8).

The function of SLP-76 is dependent on its association with LAT (9–13). This association is mediated by an adaptor known as GADS, which contains two SH3 domains flanking a SH2 domain and a linker region. GADS associates constitutively with SLP-76 through the binding of the GADS SH3 domain, and is recruited to LAT through binding of its SH2 domain to phosphotyrosine motifs on LAT upon TCR activation (9–11, 13). Besides GADS, Grb2 and possibly Grap are also implicated as adaptors for SLP-76 (4). Because mutant T cells or primary mast cells lacking LAT demonstrate reduced phosphorylation of SLP-76 upon receptor activation (14–16), the association of SLP-76 with LAT may also couple to the induction of SLP-76 phosphorylation.

To address the function of GADS in vivo, GADS-deficient mice were generated by a gene-targeting approach (Fig. 1A) (17). In contrast to SLP-76 mice, which frequently succumb to severe systemic hemorrhage at an early perinatal stage (7,18), GADS mice were grossly healthy.

Figure 1

Defects in T cell development in GADS mice. (A) Deletion of exons encoding the SH2, linker, and N-SH3 region. Details of the targeting strategy and procedure have been described (17). (B) Absence of GADS protein in GADS homozygotes. Total proteins from splenocytes were blotted with a GADS antiserum specific for the linker region. The lower band in +/+ lysate (GADS/pp) represents a caspase-mediated proteolytic product (25). The weak band in the GADS lysate is a nonspecific protein species. (C) GADS mice have reduced thymocyte numbers. (D) Total thymocytes from +/+ and GADS mice were analyzed for CD4 and CD8 expression by flow cytometry. (E) A severe reduction of TCRhigh thymocytes was detected in GADS mice. (F) Reduction of the peripheral T cell population in GADS mice (2-week-old littermates).

GADS mice had approximately fourfold fewer thymocytes than did wild-type littermates (Fig. 1C). Fluoresence-activated cell sorting (FACS) analysis for CD4, CD8, and TCR-β chain expression revealed that GADS thymocytes exhibited several developmental defects (Fig. 1, D and E). GADS mice had a modest increase in the percentage of CD4CD8double negative (DN) thymocytes and a decrease in the percentages of CD4 and CD8 single positive (SP) thymocytes (Fig. 1D). TCR-β chain staining revealed a reduction in mature cells expressing high levels of TCR (Fig. 1E). In the peripheral T cell pool, CD4 T cells were most affected, with a reduction of up to 10-fold, whereas CD8 T cells were reduced by about twofold (Fig. 1F). However, the total number of peripheral CD4 cells did accumulate over time, indicating that GADS is not critical for the survival of mature T cells. In contrast to its effect on T cell development, GADS deficiency did not influence the development of B cells or other cell lineages (17).

To characterize the developmental defect in the DN thymocytes of GADS mice, we dissected the DN compartment using FACS analysis of CD44 and CD25 expression (19). As shown in +/+ mice, approximately 26% of the DN cells had progressed into the CD44CD25 (pro-T4 stage) (Fig. 2A). In contrast, GADSthymocytes exhibited a severe developmental arrest at the CD44CD25+ (pro-T3) stage, the point at which pre-TCR signaling is first required (20,21). Unlike RAG-2 thymocytes, which are arrested in the DN stage, GADS thymocytes were able to differentiate into DP cells (Fig. 1C). This result suggests that GADS deficiency does not severely compromise the differentiation capacity of the pre-TCR.

Figure 2

Defects in pre-TCR and TCR signaling. (A) Dissection of the DN subsets with CD44/CD25 immunostaining. (B) Reduction of cycling cells in GADS thymocytes as revealed by cell size distribution in the CD44lowCD25+ (pro-T3) population. (C) Purified DN thymocytes were stained with propidium iodide (PI) to reveal DNA content (29). (D) GADS DN thymocytes did not respond to CD3 cross-linking in vivo. Four-week-old female littermates (n = 3) were injected peritoneally with 2C11 antibodies to CD3 (250 μg) and analyzed 3 days afterward for CD44/CD25 profile as described above. (E). Same group of mice as in (D) was analyzed for CD3-induced deletion of CD4+CD8+ thymocytes.

The severe reduction of pro-T4 cells in GADS mice may have resulted from a specific defect in proliferation induced by the pre-TCR. To test this hypothesis, thymocytes gated on the pro-T3 subset were analyzed for cell size distribution. In contrast to wild-type cells, the GADS pro-T3 population contained significantly fewer large cells (Fig. 2B), previously characterized as actively dividing precursors of pro-T4 cells (21). The defect in proliferation of the pro-T3 population in GADS mice was confirmed by comparing the DNA content of CD25+ DN thymocytes (Fig. 2C).

Cross-linking of CD3 in RAG-2 mice induces the differentiation and expansion of DN cells into DP cells, presumably by activating the pre-TCR signaling pathway (22). To examine whether GADS thymocytes had defects in CD3 signaling, we compared the effects of CD3 cross-linking in vivo in wild-type, RAG-2, and GADS mice. Wild-type mice injected with antibody to CD3 exhibited accelerated development of DN thymocytes, as shown by a significant increase in pro-T4 cells (Fig. 2D). Cross-linking of CD3 also significantly promoted the development of RAG-2 thymocytes, resulting in an expanded population of DP thymocytes (200 × 106 cells, n= 3). By contrast, GADS mice treated with the antibody to CD3 exhibited no signficant change in the developmental profile and none showed a significant increase in thymocyte numbers. Consistent with this observation, CD44/CD25 staining and cell size analysis revealed the lack of cellular expansion in the pro-T3 and pro-T4 compartments (Fig. 2D).

Cross-linking of CD3 in wild-type mice also induces cell death predominantly in the DP population, an effect thought to mimic thymic negative selection (Fig. 2E) (23). However, in vivo cross-linking of CD3 in GADS mice did not induce a significant deletion of the DP thymocytes (Fig. 2E) and had no effect on the total number of thymocytes. These results indicate that GADS is an integral part of CD3 signaling. Consistent with this conclusion, CD69 up-regulation was compromised in GADS thymocytes (17).

To test the function of GADS in thymic selection, GADSmice were crossed with mice carrying an αβ TCR transgene specific for the male H-Y antigen (24). In +/+ male HY mice, clonotypic T cells undergo extensive negative selection, resulting in the severe depletion of DP cells and greatly diminished thymocyte numbers. In contrast, GADS HY+ thymocytes did not exhibit negative selection in male mice, as shown by the persistent DP population (Fig. 3A). The ability of GADS thymocytes to undergo positive selection in HY transgenic female mice, which lack the H-Y antigen, was also examined (Fig. 3B). In contrast to the wild-type HY+ thymocytes, which underwent positive selection to become predominantly CD8 T cells, GADS HY+ thymocytes failed to develop into mature CD8 cells (Fig. 3B). Similar defects in the postive selection of CD4 T cells were observed in transgenic mice expressing the DO11.10 TCR (25).

Figure 3

Thymic selection is severely impaired in GADS thymocytes. (A) Failure of negative selection in HY+ GADS male mice. Panels display CD4/CD8 plots of thymocytes gated positive for the transgenic HY TCR. Total thymocyte numbers are indicated. All mice examined were littermates at 8 weeks old. (B) Failure of positive selection in HY+ GADS female mice. (C) Lack of clonotypic peripheral CD8 T cells in GADS mice.

To investigate whether these defects in thymocyte development and activation are consistent with the adaptor function of GADS, we examined whether SLP-76 phosphorylation or association with LAT was compromised in GADS thymocytes. Cell lysates from total thymocytes activated by CD3 cross-linking were immunoprecipitated for SLP-76 and immunoblotted with an antibody to phosphotyrosine. Upon CD3 cross-linking, GADS thymocytes exhibited a significant increase in SLP-76 phosphorylation, which indicates that GADS is not required to mediate SLP-76 phosphorylation (Fig. 4A). Despite this observation, the association of SLP-76 with a phosphoprotein of approximately 40 kD was specifically reduced. Based on the molecular size and the pattern of induced phosphorylation, this 40-kD species is most likely to be LAT. Unfortunately, a direct demonstration of the LAT identity is restricted by the sensitivity of immunoblotting assay required for the detection of LAT in the trimolecular SLP-76/GADS/LAT complex. Based on the interaction between Grb2 and SLP-76 in vitro (4), it is possible that Grb2 may substitute for GADS in the GADS-deficient cells. We investigated this possibility by immunoprecipitating Grb2 from total thymocytes and immunoblotting for SLP-76 or SoS, but found no evidence of increased Grb2 association with SLP-76 (17).

Figure 4

Adaptor function of GADS in thymocytes. (A) SLP-76 was readily phosphorylated but failed to associate with LAT in activated GADS thymocytes. Purified thymocytes were cross-linked with anti-CD3 biotin at 4°C, followed by activation with streptavidin at 37°C for 2 min. Anti–SLP-76 immunoprecipitates (IP: α-SLP-76) were immunoblotted with an antibody to phosphotyrosine (α-pTyr) (4G10). The middle panel shows an immunoblot of SLP-76 of the same membrane after stripping. The bottom panel shows an immunoblot with 4G10. (B) Impaired PLC-γ1 activation in GADS thymocytes. Lysates from activated thymocytes were immunoprecipitated for PLC-γ1 (IP: α-PLC-γ1) and immunoblotted with the 4G10 antibody to phosphotyrosine. The bottom panel shows an immunoblot of PLC-γ1 of the same membrane after stripping. (C) Defective calcium mobilization in GADS peripheral T cells. Total splenic T cells were stained with an APC-conjugated CD8 antibody and loaded with Fluo-4 to determine calcium flux in a FACS Calibre. Samples were first incubated with a biotinylated antibody to CD3 (αCD3) at 4°C, followed by cross-linking with streptavidin (Sav) at 37°C. Ionomycin (Iono) was added at the end of the experiment as a positive control.

Biochemical studies suggest that ITK bound to SLP-76 is recruited to LAT in order to phosphorylate and activate PLC-γ1 (3,26). In accordance with this model, GADS thymocytes exhibited significantly reduced PLC-γ1 phosphorylation (Fig. 4B), and GADS-deficient T cells failed to flux calcium upon CD3 activation (Fig. 4C). GADS T cells activated with antigen proliferated to an extent comparable to that of the wild type (25), a result that may be due to compensation during thymic development. Inducible gene targeting of GADS is required to further address this possibility. Although GADS is also expressed in the B cell lineage (11), our analysis indicated that GADS is not required for B cell activation (17).

This study provides genetic evidence that GADS is important for pre-TCR and TCR signaling. Mice deficient in GADS exhibit less severe defects, such as the lack of hemorrhage and the presence of DP and SP thymocytes, than do mice deficient in SLP-76. Such a phenotypic difference indicates that not all functions of SLP-76 are mediated through GADS. Despite the importance of GADS in the various aspects of thymocyte development, there is only a four- to sixfold reduction in total thymocyte number. This moderate reduction in thymocyte number may be due to the lack of efficient positive selection, which impairs the maturation of DP thymocytes; combined with the lack of negative selection, which limits the deletion of self-reactive thymocytes. GADS deficiency has a more pronounced effect in the maturation of the CD4 than the CD8 lineage. It has been proposed that the Ras pathway may quantitatively regulate the development of the CD4 and CD8 lineages (27). In this context, it is important to note that Ras activation by the SLP-76/LAT complex is regulated by at least two different pathways: one via Grb2-SoS bound to LAT and the other via RasGRP (28), which is activated by diacylglycerol and in principle lies downstream of PLC-γ1, whose activation is dependent on GADS-SLP-76. It will be interesting to address the differential effects in the maturation of CD4 and CD8 cells caused by these two pathways. Finally, the observation that LAT (14, 15), but not GADS, is required for the optimal phosphorylation of SLP-76 suggests that LAT has a yet-undefined role in SLP-76 phosphorylation. Further studies on the mechanism of SLP-76 phosphorylation may provide insight into this issue.

  • * To whom correspondence should be addressed. E-mail: acheng{at}im.wustl.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article