This study investigated the mechanism by which distinct biochemical pools of ADAP control integrin activation and NF-κB signaling following TCR stimulation of primary murine T lymphocytes. We tested the hypothesis that the physical association of ADAP with SKAP55 not only regulates integrin activation following TCR stimulation but also attenuates NF-κB activation by controlling the ability of ADAP to interact with the CBM complex. We have identified an arginine motif (R131) in the PH domain of SKAP55 that is required for the recruitment of the ADAP:SKAP55 complex to LFA-1 integrins and subsequent ADAP-dependent T:APC conjugate formation, and restrains efficient NF-κB activation ().
Although the constitutive association of ADAP and SKAP55 is critical for efficient TCR-mediated activation of integrins (18
), the precise mechanism by which this scaffolding complex triggers changes in integrin function remains unclear. SKAP55 is obligately associated with ADAP(29
), while studies with both Jurkat T cells and ADAP−/−
T cells have shown that there is a biochemical and functional pool of ADAP in T cells that is not associated with SKAP55 and regulates NF-κB signaling (22
). In this study, we utilized the fact that ADAP−/−
T cells lack expression of both endogenous ADAP and SKAP55 in order to define the mechanism by which the ADAP:SKAP55 complex specifically regulates integrin activation but not NF-κB activation.
We developed a SKAP/ADAP chimeric molecule consisting of the N-terminus of SKAP55 fused in frame to the C-terminus of ADAP. This chimera was efficiently expressed and did not induce upregulation of endogenous SKAP55, allowing us to analyze T cells where all of the ADAP was constitutively associated with SKAP55. Importantly, the chimera efficiently rescued conjugate formation defects in ADAP−/−
T cells, demonstrating that these minimal domains of SKAP55 and ADAP are sufficient for integrin activation in a system where neither protein is endogenously expressed. It has been speculated by several investigators that SKAP55 contributes key integrin regulatory information to the ADAP:SKAP55 complex (18
). The SKAP/ADAP chimera allowed us to define a novel function for the PH domain of SKAP55 in regulating conjugate formation by controlling the localization of the ADAP:SKAP complex to integrins (). We focused on the SKAP55 PH domain because of recent crystallographic studies of the SKAP55 homolog SKAP-HOM (SKAP55R; SKAP2) (21
). This work demonstrated that mutation of R140 in the PH domain of SKAP-HOM to methionine abrogated the binding of the SKAP-HOM PH domain to PI[3,4,5]P3
and the recruitment of SKAP-HOM to actin-rich membrane ruffles (21
). These findings are consistent with the known role of other PH domains in recruiting proteins to sites of phosphoinositide synthesis (36
) and are particularly intriguing given the role of PI3K in inside-out integrin activation (36
). Interestingly, the PH domain in SKAP-HOM is atypical in that it folds into a novel intramolecular interaction with another domain within SKAP-HOM that controls SKAP-HOM dimerization, resulting in constitutive dimers of SKAP-HOM that contain masked PI[3,4,5]P3
binding sites (21
). High concentrations of PI[3,4,5]P3
release the PH domain from this intramolecular association, allowing for SKAP-HOM recruitment to membranes rich in PI[3,4,5]P3
. Although this has yet to be demonstrated, the high sequence identity and homology between the PH domains of SKAP-HOM and SKAP55 suggest that SKAP55 may also undergo similar PI[3,4,5]P3
dependent conformational switching and membrane recruitment via its PH domain (21
Consistent with this model, we found that removal of either the entire PH domain or mutation of the conserved arginine at position 131 in the SKAP/ADAP chimera to methionine severely impaired the rescue of T:APC conjugate formation in ADAP−/− T cells, likely tracing to the impaired accumulation of the SKAP/ADAP chimera with LFA-1 adhesion complexes observed following CD3/28 simulation (). Our results provide direct structural and functional demonstration of an integrin-regulatory domain within SKAP55 that is distinct from the ADAP-binding SH3 domain. In addition, our findings establish in primary T lymphocytes a model where the SKAP55 PH domain senses activated T cell membranes and thereby targets the ADAP:SKAP55 complex to LFA-1 integrins.
It is currently not known how PI[3,4,5]P3
dependent conformational shifts in the SKAP55 PH domain promote T:APC conjugate formation via association with LFA-1. One likely possibility is that there exist high concentrations of TCR-dependent signaling lipids in membranes that are also rich in integrin heterodimers, bringing the SKAP55:ADAP complex into close proximity with these integrins. Current models of integrin activation propose that recruitment of the small GTPase Rap1 and its effectors RIAM and RapL are critical final steps required for integrin activation (38
). A role for the SKAP55 PH domain in recruiting the SKAP:ADAP complex to integrins is consistent with this model, as both RIAM and RapL have been identified as binding partners for SKAP55 (19
). While no RIAM binding site within SKAP55 has yet been defined, RapL appears to bind the N-terminus of SKAP55 (31
). It is unclear whether RapL binding would affect SKAP55 constitutive homodimerization through this domain, but one scenario is that the opened PH domain exposes part of the N-terminal dimerization domain to trigger RapL binding, further recruiting the ADAP:SKAP complex to PI[3,4,5]P3
rich membranes where RapL can recognize the αL integrin chain of LFA-1 (34
). It is also not known why the ADAP/SKAP55 complex does not fully account for all TCR-dependent LFA-1 activation in T cells, as evidenced by the observation that ADAP−/−
T cells show only a partial reduction in T:APC conjugate formation. Since ADAP appears to be most critical in situations of weak or limiting antigen stimulation (13
), this suggests the existence of ADAP-independent modes of integrin activation downstream of the TCR. One possibility is that since Rap1 activation is normal in the absence of ADAP (18
), this pool of activated Rap1 may positively regulate integrin activation even in the context of the defective Rap1 plasma membrane targeting that has been reported in ADAP−/−
In addition to the targeting information provided by SKAP55, the requirement for the ADAP C-terminus on the SKAP/ADAP chimera in our present work reinforces previous findings pointing to the importance of the ADAP C-terminus towards LFA-1 integrin activation (18
). Since ADAP inducibly associates with the SH2 domain of SLP-76 (40
), it remains likely that phosphorylation of the ADAP C-terminus following TCR simulation couples ADAP to the LAT/SLP-76 complex. In contrast, other studies have placed ADAP within LFA-1 outside-in signaling pathways following TCR and LFA-1 costimulation (43
), suggesting that additional levels of ADAP regulation may be operative in activated T cells. In addition, a number of previously unrecognized phosphorylated tyrosine residues in the ADAP C-terminus as well as novel ADAP interacting proteins have recently been reported, including cytoskeletal regulatory proteins such as Dock2 and Nck (45
). These results are consistent with a proposed role for ADAP in controlling TCR-dependent cytoskeletal rearrangements (48
). Future studies using well controlled systems such as ADAP−/−
cells expressing C-terminal tyrosine or actin-binding mutants of the SKAP/ADAP chimera could help clarify these signaling networks. In addition, the atypical helical SH3 domains within the ADAP C-terminus have been reported to bind negatively charged phospholipids and potentially alter adhesion to immobilized integrin ligands (50
). Future studies will need to address whether lipid- or phosphoinositide- binding cooperativity exists involving the PH domain of SKAP55 and the helical SH3 domains of ADAP.
ADAP has recently emerged as a key regulator of TCR dependent NF-κB activation (27
). This pathway depends on C-terminal amino acid motifs in ADAP that are distinct from the central SKAP55 binding domain in ADAP. Surpisingly, the SKAP/ADAP chimera did not rescue NF-κB activation in ADAP−/−
cells despite containing the sites in ADAP critical for binding to CARMA1 and TAK1. However, the SKAP/ADAP R131M mutant, which is unable to efficiently rescue integrin function when expressed in ADAP−/−
T cells, is able to restore NF-κB activation, suggesting that the intact SKAP55 PH domain plays a previously unappreciated role in excluding bound ADAP from interacting with the CBM complex even in TCR stimulated cells. Since SKAP55 is limiting in T cells, increases in SKAP55 expression might reduce the amount of ADAP able to interact with the CBM signalosome and promote NF-κB activation. Our finding that over-expression of wild-type SKAP55 in wild-type T cells impairs IκB phosphorylation and degradation is consistent with this model. In contrast, over-expression of the SKAP55/R131M mutant did not affect CD3/CD28-mediated activation of NF-κB. Overall, our results are consistent with the hypothesis that the fraction of ADAP that associates with SKAP55 is recruited by the PH domain of SKAP55 to integrins and thus impairs ADAP from interacting with the CBM signalosome and regulating the NF-κB pathway ().
In summary, our results demonstrate a critical function for the SKAP55 PH domain not only in controlling integrin function via recruitment of ADAP/SKAP complexes to integrins, but also in controlling the ability of ADAP to interact with the CBM signalosome and regulate NF-κB (). The ability to modulate NF-κB activation by changing levels of SKAP55 expression in T cells suggests the intriguing possibility that changes in the expression patterns of SKAP55 and/or ADAP during the course of T cell activation may be a mechanism by which T cells modulate this critical transcription factor signaling pathway.