T cell activation through the Ag receptor (TCR) requires sustained signaling from signalosomes within lipid raft microdomains in the plasma membrane. In a proteomic analysis of lipid rafts from human T cells, we identified stomatin-like protein (SLP)-2 as a candidate molecule involved in T cell activation through the Ag receptor. In this study, we show that SLP-2 expression in human primary lymphocytes is up-regulated following in vivo and ex vivo activation. In activated T cells, SLP-2 interacts with components of TCR signalosomes and with polymerized actin. More importantly, up-regulation of SLP-2 expression in human T cell lines and primary peripheral blood T cells increases effector responses, whereas down-regulation of SLP-2 expression correlates with loss of sustained TCR signaling and decreased T cell activation. Our data suggest that SLP-2 is an important player in T cell activation by ensuring sustained TCR signaling, which is required for full effector T cell differentiation, and point to SLP-2 as a potential target for immunomodulation.
The CD4+ T cell activation involves the formation of a highly organized interface between the T cell and the APC known as the immunological synapse (IS)5 (1, 2, 3, 4). This process requires differential polarization of surface and intracellular molecules to the IS-proximal pole (5) or to the antipodal pole of the T cell (6, 7), in a cytoskeleton-dependent manner (8, 9, 10). It also affects cellular organelles, as recently shown by two elegant reports of the translocation of mitochondria to the IS during T cell activation (11), and to the uropod during T cell locomotion (12). In addition, T cell polarization is highly dynamic, as illustrated by the early detection of signaling TCR microclusters or signalosomes in the outer ring of the synapse (13) and their subsequent movement to the center of the synapse, where they are internalized and destined to degradation (2, 3, 14), thus ensuring signal down-regulation (15).
To identify molecules that act as a functional bridge between TCR signalosomes and the cytoskeleton and cellular organelles, we performed a proteomic analysis of lipid raft microdomains from human T cells undergoing activation through their TCR. With this approach, we isolated stomatin-like protein (SLP)-2 (also known as STOML2 or EPB72). This finding has been corroborated by other groups using lipid rafts and/or mitochondrial fractions from different cell lineages of several species (16, 17).
SLP-2 is a member of the highly conserved family of stomatin proteins whose homologues span from archae to humans and include stomatin, SLP-1, and SLP-3 (18, 19, 20, 21, 22). SLP-2 shares with the other stomatins a central stomatin/prohibitin/flotillins/HflK-HflC (SPFH) domain that may mediate interactions with cell membranes (23, 24, 25). However, SLP-2 is unique among stomatins in that it does not have a putative transmembrane domain, but has six myristoylation/palmitoylation sites and an N-terminal mitochondria targeting sequence (25, 26). These features most likely determine its detection in mitochondria and plasma membrane.
The function of stomatins, including SLP-2, is currently unknown. It has been suggested that they may be involved in the organization of the peripheral cytoskeleton, and in the assembly of multichain receptors, such as ion channels (25, 27, 28, 29, 30) and mechanosensation receptors (31, 32, 33, 34, 35, 36, 37). In this context, one would expect that SLP-2 be involved in the regulation of signaling from these receptors. In this study, we report that SLP-2 plays an important role in human T cell activation by contributing to sustain TCR signaling. This effect correlates with the interaction of SLP-2 with TCR signalosome components, and with polymerized actin. More important, modulation of SLP-2 levels translate into changes in effector T cell responses, suggesting that SLP-2 may be a useful immunotherapeutic target.
Materials and Methods
Human PBMC were isolated from heparinized whole blood of normal donors using Ficoll-Hypaque gradients (Amersham Biosciences). Cells were washed in supplemented RPMI 1640 and resuspended at 1 × 106 cells/ml. PBMC blasts were generated by culturing PBMC with PMA (1 ng/ml) and ionomycin (100 ng/ml) for 72 h at 37°C, 5% CO2. T cell blasts were rested 48 h before use. Primary T cells were isolated from PBMC using a Pan T Cell Isolation Kit (Miltenyi Biotec). Jurkat T cells (E6.1) were obtained from American Type Culture Collection and cultured in supplemented RPMI 1640 medium. The B lymphoblastoid cell line LG2, used as APC in some of these experiments, was provided by E. Long (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, MD) and cultured in standard supplemented RPMI 1640 medium.
Plasmids, small interfering RNA (siRNA), and T cell transfectants
Human SLP-2 cDNA was subcloned into the pEGFP-N1 expression vector (BD Clontech) to create an in-frame translational fusion of SLP-2 and gfp at the 3′ end. Subsequently, the SLP-2-gfp was placed into the doxycycline-inducible pBig2i vector (38). Stable transfectants were generated by electroporating linearized plasmid into Jurkat E6.1 T cells and screened for stable expression. Doxycycline (Sigma-Aldrich) was added in culture overnight at 1 μg/ml to induce SLP-2-gfp expression. Expression of SLP-2-gfp was monitored by direct flow cytometry (BD Biosciences). RNA interference targeting SLP-2 was performed using siRNAs obtained from Ambion (catalogue no. 20643 and 20467) and nucleofected into Jurkat T cells, PBMC, or PBMC blasts using the Nucleofector kit for either cell lines or human primary T cells (Amaxa). As controls for SLP-2 siRNA, we used the scrambled siRNAs provided as a control by the commercial supplier (Ambion). As per manufacturer’s tests, these control siRNAs have no significant similarity to any known gene sequences from mouse, rat, or human, and no toxicity to cells; have been shown to lack any significant effect in cell proliferation and apoptosis assays; and do not modulate the mRNA levels of housekeeping genes (18S rRNA, GAPDH, and cyclophilin) up to 48 h after transfection. siRNAs that target different exons of SLP-2 were mixed to establish whether better knockdown of SLP-2 could be achieved. Through this process, it was established that combining two siRNAs (catalogue no. 20643 and 20467) produced the highest level of knockdown. GFP cDNA in the targeting vector was used as a control for nucleofection efficiency. With the protocols supplied by Amaxa, efficiency of nucleofection for Jurkat T cells was between 40 and 65%. After nucleofection, T cells were rested for 24 h until optimum down-regulation of SLP-2 was observed before proceeding with any functional assays.
An antiserum against human SLP-2, generated by immunization of rabbits with a peptide spanning aa 343–356 (ProSci), was used for immunoprecipitation of SLP-2. A preimmunization serum was used as control for rabbit antisera immunoprecipitations (ProSci). Commercially available Abs against SLP-2 were purchased from Protein Tech Group. Antisera against Lck, linker for activation of T cells (LAT), and phospho-LAT were obtained from Upstate Biotechnology. Anti-ZAP70 and anti-phospho-ZAP70 (Y292) antisera, and anti-Ras-GAP Ab were obtained from Santa Cruz Biotechnology. Immunoblotting for ERK-1/2 was done using a rabbit polyclonal immunoaffinity-purified antiserum (StressGen Biotechnologies). Anti-phospholipase C (PLC)-γ1, anti-phospho-PLC-γ1, anti-active ERK-1 and ERK-2 E10, and anti-caspase 3 antisera were obtained from Cell Signaling Technology. CD45 was immunoblotted using a mouse mAb (clone 69; BD Biosciences). Actin was immunoblotted using an affinity-purified goat polyclonal antiserum (Santa Cruz Biotechnology), and GAPDH was immunoblotted with a mouse mAb from Chemicon International. To ensure that the Western blot signals were within the linear range of detection, we performed signal titration over multiple exposures over a range of time periods from 30 s to 30 min. Signal saturation was detected by image acquisition software (Alpha Innotech). Densitometry and m.w. calibration were done with the Phoretix 1D Database package.
Lipid raft isolation
Jurkat T cells or blast PBMC were stimulated with staphylococcal enterotoxin E superantigen (SEE) and LG-2 cells as APC (at a 5:1 ratio; 60 × 106 T cells with 12 × 106 APC). Whole cell lysates were prepared with 0.5% Triton X-100, and lipid rafts were isolated by sucrose gradient ultracentrifugation, as described (39). Lipid rafts from the same cell equivalents per sample were pelleted by centrifugation of the 1-ml raft fraction for 1 h at 14,000 rpm and 4°C, and were resuspended in lysis buffer and sample buffer for biochemical analysis.
Cell lysate preparation
Jurkat T cells or PBMC were stimulated with superantigens at final concentration of 1 μg/ml, and APC at a 5:1 ratio, at 37°C, for 1, 5, 15, 30, and 60 min (40, 41). Cells were pelleted in PBS containing sodium o-vanadate (400 μM) and EDTA (400 μM), and lysed in lysis buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris (pH 7.6), 5 mM EDTA, 1 mM sodium o-vanadate, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 25 μM p-nitrophenyl-p′-guanidinobenzoate) at 4°C for 30 min. Lysates were cleared of debris (14,000 rpm, 4°C, 10 min), and used for immunoprecipitation of target molecules using Ab-coated protein A or G agarose beads. Whole cell lysates from human tissues were obtained from ProSci.
Protein A agarose beads were coated with appropriate Abs at 4°C overnight. Beads were washed four times with room temperature lysis buffer and four times with PBS. Immunoprecipitating Abs were cross-linked to the beads in 1 mg/ml dithiobis[succinimidyl propionate] (Pierce) in PBS, at room temperature rotating for 30 min. The cross-linked Ab beads were neutralized with 1 M Tris (pH 8.0) at room temperature for 5 min, washed once with lysis buffer and four times with PBS, and then used for immunoprecipitation. After immunoprecipitation, beads were pelleted and resuspended in sample buffer without 2-ME. Samples were boiled, pelleted, and supernatant collected; run in SDS-PAGE; and immunoblotted with the indicated Abs.
Tonsil B cells
Highly purified human tonsil B cells by RosetteSep B cell enrichment mixture (StemCell Technologies) were fractionated by a seven-step Percoll gradient. With this procedure, four fractions were obtained (labeled f1-f4). All fractions were phenotyped for naive B cells (based on expression of IgM and IgD) and memory B cells (CD20bright and CD27+). The proportion of naive resting B cells increases from fraction 1 to 4 (from 11 to 66% on average), whereas the proportion of memory/activated B cells decreases from fraction 1 to 4 (from 70 to 3% on average). In addition, each fraction was cultured in vitro to determine spontaneous Ig production. Only fractions 1 and 2 were able to spontaneously secrete Ig (data not shown). Whole cell lysates from B cells in each fraction were prepared and immunoblotted with appropriate specific Ab.
T cell functional assays
SLP-2-gfp, SLP-2 siRNA, or control transfected T cells (0.2 × 106 cells/group) were plated in triplicate on 96-well plates with the B lymphoblast LG2 (0.1 × 106 cells/group) as APC and SEE in the presence or absence of 1000 ng/ml doxycycline, at 37°C for 24 h. Supernatants were collected, and measurement of IL-2 by ELISA was performed following manufacturer specifications (BD Biosciences). Results are presented as mean ± SD, and statistical significance was determined by two-tailed t tests for each given time point.
Fragments of thymi and lymph nodes were fixed with 10% formalin, embedded in paraffin, and stained with H&E. Immunohistochemical analysis for SLP-2 was performed with an antisera against SLP-2 at 1/1000 dilution following the streptavidin-biotin peroxidase method. CD5 and CD20 (mouse anti-human mAbs; DakoCytomation) stains were also performed to identify T and B lymphocytes, respectively, and correlated with SLP-2 staining.
Expression of SLP-2 in human lymphoid tissues
To explore the involvement of SLP-2 in lymphocyte activation, we first determined its expression in human lymphoid tissues using whole cell lysates from commercial sources. We detected SLP-2 mostly in lymph node and thymus lysates, and in lower amount, in tonsil lysates (Fig. 1 A). Little expression was detected in lysates from unfractionated spleen and resting peripheral blood leukocytes. Such differences were not due to differences in protein loading, as illustrated by immunoblotting of the same samples for GAPDH.
In lymph nodes, SLP-2 was mostly detected in the germinal centers (B cell area) and in the paracortical (T cell) area (Fig. 1,B), with a staining pattern of intracellular punctae and peripheral layer consistent with its targeting to mitochondria and plasma membrane association. In the thymus, SLP-2 was detected both in the cortex and in the medulla, although the expression seemed to be higher in the cortex (Fig. 1,B). In those individuals in which SLP-2 was detectable in PBMC, it was expressed by monocytes, and to less extent, by T and B lymphocytes (Fig. 1 C).
Up-regulation of SLP-2 expression in T and B lymphocytes upon activation
The high expression of SLP-2 in sites where lymphocyte signaling and activation take place (i.e., Ag activation in lymph nodes and tonsils, positive and negative selection in the thymus) prompted us to examine the effect of activation on SLP-2 expression. Although, in most volunteers tested, peripheral blood T cells express low levels or no SLP-2 under resting conditions, activation of these cells with bacterial superantigens led to a consistent and significant up-regulation of its expression after 36 h (Fig. 2,A). Such an up-regulation was already detectable after 2 h of stimulation through the TCR, as observed when cross-linked Abs against CD3 were used to activate T cells (data not shown). Similarly, in B cell preparations from human tonsils fractionated according to their activation status, SLP-2 was mostly detected in the fraction corresponding to activated/memory (CD20bright, CD27+) B cells (Fig. 2 B). Seventy-three percent of cells in this fraction (f1) were activated/memory B cells (as indicated by expression of CD27), and only 8% of cells in this fraction were naive B cells. In contrast, 60% of the cells in fraction 4 were naive B cells and only 11% were activated/memory B cells. Fractions 2 and 3 had progressively decreasing numbers of activated/memory B cells (52 and 32%, respectively) and increasing numbers of naive B cells (17 and 45%, respectively). Together, these data led us to conclude that SLP-2 expression is up-regulated by in vivo and ex vivo lymphocyte activation.
SLP-2 interacts with components of TCR signalosomes and with polymerized actin during T cell activation
To study the contribution of SLP-2 to T cell activation, and because we had isolated SLP-2 from lipid rafts of activated T cells, we first analyzed the partitioning of SLP-2 within these microdomains following activation. To do this, we used Jurkat T cells that express SLP-2 under resting conditions. As shown in Fig. 3,A, SLP-2 was detected in lipid rafts and its partitioning to these microdomains increased soon after activation, subsequently returning to basal levels. Such a profile of redistribution into lipid rafts was consistently documented in multiple experiments (Fig. 3,B), and in primary human T cells at discrete time points (Fig. 3 C). We failed to see a reciprocal change in the levels of SLP-2 in the soluble fractions in comparison with the increase in SLP-2 levels in the lipid raft fractions, either because the SLP-2 that redistributes into rafts came from the insoluble cytoskeletal pellet or because the amount of SLP-2 that repartitions to the lipid raft fraction from the soluble fraction is small in comparison with the total amount of soluble SLP-2.
Next, we tested whether SLP-2 interacted with the components of TCR signalosomes during activation for up to 60 min, a time window that covers the formation of TCR signaling microclusters and mature IS (Fig. 4,A). In complementary coimmunoprecipitation studies, we observed that SLP-2 steadily associated with the CD3-ε chain of the TCR complex under resting conditions and during the 60 min of stimulation. Detailed quantitation studies of this association indicated that ∼0.09% of the cellular SLP-2 associated with the TCR complex (data not shown). In these studies, we also found that SLP-2 interacted with Lck, ZAP70, LAT, and PLC-γ1 during the 30-min period following stimulation in vitro. Of interest, the magnitude of LAT interacting with SLP-2 was comparable to that of phospho-LAT interacting with SLP-2, suggesting that this interaction involves a single functionally active pool of LAT. In contrast, the interaction between SLP-2 and ZAP70 or PLC-γ1 was steady during the early times of stimulation, whereas tyrosine phosphorylation of these pools only occurred at discrete time points, suggesting the presence of different SLP-2-interacting pools of ZAP and PLC-γ1 in the cell. Such profile of interactions was only observed in specific SLP-2 immunoprecipitations with cross-linked Abs, but not with nonspecific control Abs. Also, these interactions were preserved under stringent lysis conditions, and were selective because no association of SLP-2 with other surface receptors (CD45) or with control intracellular molecules (Ras-GAP, caspase 3) was detected (Fig. 4 A). In addition, such interactions were observed using a different detergent (octylglucoside, 2%), and in primary human T cells (data not shown).
Because polarization of signaling molecules and organelles to the IS is cytoskeleton dependent, we examined the association of SLP-2 with actin. We found that SLP-2 interacted with actin under resting conditions and upon TCR stimulation (Fig. 4,B). Such an association was not observed using a control preimmune serum, ruling out a nonspecific interaction, and mostly involved T cell actin given the T:APC ratio (5:1) used in these experiments. Furthermore, the association between SLP-2 and β-actin during T cell activation seems to involve mostly the polymerized form of actin because cytochalasin D, an inhibitor of actin polymerization, completely prevented this interaction (Fig. 4,B). We observed a lower level of association between SLP-2 and actin at time 0 in T cells treated with cytochalasin D, most likely reflecting a basal level of interaction between these molecules under steady-state conditions. It is important to note that T cell activation in this experimental system is dependent on actin polymerization, because inhibition of actin polymerization with cytochalasin D led to inhibition of IL-2 production in a dose-dependent manner (Fig. 4 C).
SLP-2 contributes to sustain T cell signaling
Next, we looked at the functional correlates of SLP-2 expression and its interactions with the components of the TCR signalosome. We reasoned that if SLP-2 played a role in signaling from the TCR signalosomes, then knocking down SLP-2 would decrease TCR signaling. To test this hypothesis, we knocked down SLP-2 expression by RNA interference using two siRNAs against SLP-2. With this technique, we achieved between 65 and 85% reduction of SLP-2 levels. Under these conditions, we found that SLP-2 down-regulation correlated with a significantly shorter timeline of TCR signaling (Fig. 5,A). That is, at equal levels of early TCR signaling (as measured by similar levels of initial ERK-1/-2 activation), down-regulation of SLP-2 caused a remarkably shorter duration of ERK activation in response to TCR stimulation, already apparent at 5 min and still significant at 10 min of stimulation. The effect of SLP-2 down-regulation on TCR-dependent signaling was also observed for the kinetic profile of activation of PLC-γ1 (Fig. 5,A), and was statistically significant across multiple experiments looking at the profile of ERK activation in response to TCR signaling using Jurkat T cells and primary T cells (Fig. 5 B).
Modulation of IL-2 responses by SLP-2
Because SLP-2 down-regulation prevented sustained signaling, and because sustained TCR signaling is required for full T cell activation, we predicted that changes in the levels of SLP-2 expression would translate in changes in an effector T cell response such as IL-2 production. We tested this hypothesis in human T cell lines and primary T cells. To determine the effect of SLP-2 overexpression on IL-2 production, we transfected Jurkat T cells with a doxycycline-inducible SLP-2-gfp cDNA. In these cells, SLP-2 expression is increased due to basal leakiness of the vector promoter. Such an expression can be further up-regulated with the addition of doxycycline to the culture, as demonstrated by FACS (Fig. 6,A). We found that de novo overexpression of SLP-2 (as corroborated by FACS) significantly increased IL-2 production of these T cells in response to SEE and APC, compared with the parental E6.1 T cell line expressing low levels of SLP-2 (Fig. 6 A).
To test whether knocking down the expression of SLP-2 decreased IL-2 responses, we nucleofected Jurkat T cells with two different SLP-2 siRNA constructs. Down-regulation of SLP-2 upon RNA interference was documented by Western blotting, as shown in Fig. 6,B. On average, with a reduction of SLP-2 levels between 70 and 80%, we observed that the IL-2 response to APC and SEE significantly decreased compared with the parental SLP-2-expressing Jurkat T cell line (Fig. 6 B).
Finally, we corroborated the effect of SLP-2 expression on IL-2 responses in primary human T cells from normal volunteers. Human naive T cells express low levels of SLP-2, but the levels are up-regulated significantly upon activation. Thus, to induce high expression of SLP-2 in primary human T lymphocytes, we preactivated these cells with a mitogenic combination of PMA and ionomycin for 3 days, followed by a resting period. At this point, these effector T cells and their naive T cell counterparts from the same donor were nucleofected with SLP-2 siRNA or control siRNA and restimulated with SEE and APC for 24 h, and their IL-2 response was examined. As shown in Fig. 7,A, knocking down the expression of SLP-2 by more than 60% correlated with a significant decrease in the IL-2 response of the effector T cells (∼10–100 times), as illustrated by the shift to the right of the dose-response curve to SEE (p < 0.001). Little effect was observed for SLP-2 siRNA in resting T cells in which SLP-2 expression is almost absent (Fig. 7,A), suggesting that the SLP-2 siRNAs used for these experiments do not have an off-target (non-SLP-2-dependent) effect on the IL-2 response of T cells. SLP-2 siRNA had no effect on the IL-2 response of human T cells to mitogenic stimulation with PMA and ionomycin, which bypasses early TCR signaling events (Fig. 7,B). Together, the findings from Jurkat T cells and from primary T cells demonstrate that T cell activation can be modulated by changing the levels of SLP-2. We documented that, in primed primary human T cells, there is a more sustained profile of TCR-dependent ERK activation compared with naive primary T cells (Fig. 7,C), with stronger phospho-ERK signal still detectable by 30 min postactivation. This finding is consistent with our claim given that primed T cells up-regulate SLP-2 expression, and complements the results shown in Fig. 5 of less sustained ERK activation in Jurkat T cells when SLP-2 was down-regulated by siRNA.
The results presented here show that SLP-2 expression in T and B lymphocytes is up-regulated by activation in vivo and ex vivo. Upon expression in T cells, SLP-2 interacts with molecules involved in early TCR signaling and with polymerized actin, and contributes to sustain TCR-dependent signaling. Consistent with these findings, one can modulate effector T cell responses such as IL-2 production by changing the levels of SLP-2: overexpression of this protein increases IL-2 responses, whereas down-regulation of its expression correlates with decreased IL-2 responses, both in T cell lines as well as in primary human T cells. To our knowledge, these results provide the first biological evidence of a function for SLP-2 in an eukaryotic cell type, that of sustaining TCR signaling and enhancing T cell activation.
The findings of SLP-2 partitioning in lipid rafts, its interactions with TCR signalosome components and polymerized actin, and its contribution to sustain TCR-dependent signaling are consistent with the emerging role assigned to proteins containing an SPFH domain. This domain has been linked to protein interactions with cell membranes (23, 24, 25). In addition, proteins containing such a domain are enriched in lipid rafts, and have been previously implicated in the regulation of cell signaling (42, 43). For example, prohibitin is indispensable for the activation of the Ras-ERK signaling pathway (44), and a similar role has been proposed for the flotillins (45, 46, 47). Together this evidence positions SLP-2 in the proper subcellular compartment to participate in TCR-dependent signaling.
The structural basis of the molecular interactions identified for SLP-2 is currently under study. The primary sequence of SLP-2 predicts that there are at least two cellular pools of this protein: one in mitochondria (25) and the other associated with the plasma membrane (16, 48, 49, 50). Preliminary data from our laboratory confirm these two pools and suggest that both pools of SLP-2 coalesce in the peripheral area of the IS, a region in which signaling TCR microclusters accumulate (14) (M. Kirchhof, C. Lemke, and J. Madrenas, preliminary observations). Such an observation is also consistent with the recent report of mitochondrial polarization to the IS during T cell activation (11). The primary sequence of SLP-2 does not stand out with an enzymatic domain or with a conventional protein-protein interaction domain that could explain its direct interactions with signaling molecules. It is therefore plausible to assume that the interactions of SLP-2 with TCR signalosomes and with polymerized actin involve indirect associations between SLP-2 and other molecules, with SLP-2 potentially acting as a scaffolding/assembling protein.
How can SLP-2 contribute to sustain TCR-dependent signaling? We propose two, nonmutually exclusive ways by which SLP-2 can do that. One way is that SLP-2 contributes to the assembly of the multimolecular signalosomes on the transmembrane adapter LAT. Such a facilitating role in signalosome assembly would stabilize signalosomes for sufficient time to deliver sustained signaling. This possible mechanism is supported by the observed sequential interactions between SLP-2 and signaling molecules and polymerized actin, and by the prevention of sustained TCR signaling seen with down-regulation of SLP-2 expression. This possibility is also consistent with the formation of TCR signalosomes within lipid rafts (51, 52, 53), where SLP-2 is enriched, and with the recent claim that SPFH family members bind cholesterol, and contribute to the formation of signaling-permissive protein-lipid complexes (30, 54).
Another way to explain how SLP-2 contributes to sustain TCR-dependent signaling is by claiming that SLP2 modulates mitochondrial function, which is in itself linked to TCR signaling, as indicated by the polarization of mitochondria to the IS. The specific role that SLP-2 plays in mitochondrial function is unknown, but may be linked to the assembly of the electron transport supercomplexes in lipid rafts of the inner mitochondrial membrane in a way analogous to what we propose for the SLP-2 pool in the plasma membrane. According to this hypothesis, polymerization of actin in response to TCR signaling would bring TCR signalosomes and mitochondria to the IS. In this way, the energetic requirements of TCR signaling can be met because it warrants close proximity between the TCR signalosomes and the compartmentalized mitochondria. This possibility is in line with the changes in mitochondria positioning during cell responses through fission and fusion, processes that make use of the cytoskeleton and/or microtubules (12, 55). We cannot rule out that SLP-2 may scaffold mitochondria with the cytoskeleton through its interaction with mitochondrial external membrane proteins (56) such as mitofusin-2 (although we have to date failed to document SLP-2-mitofusin-2 association in T cells). This proposed role for SLP-2 in T cell activation points to a novel regulatory mechanism of signal transduction based on cytoskeleton-dependent interaction between signalosomes and cellular organelles (9), a mechanism that may be applicable to other cell surface receptors in addition to the TCR (our preliminary observations).
We have shown that, following T cell activation, the levels of SLP-2 increase, and this translates into enhanced effector responses (e.g., IL-2 production). In contrast, decreasing SLP-2 expression decreases IL-2 responses. Such observations have, in our opinion, several important biological implications. One is that it may partially explain the long-standing finding that, at equal requirement for sustained signaling, primed T cells show much stronger signal transduction and functional responsiveness than naive T cells. One may argue that this is, in part, due to primed T cells having higher levels of SLP-2 than naive T cells (57, 58). Another implication is that the increased effector phenotype seen with neoplastic transformation may be due, in some cases, to higher SLP-2 expression (59, 60, 61, 62). Finally, the ability to modulate effector T cell responses by regulating SLP-2 expression identifies SLP-2 as a potentially useful target for immunotherapy. This could take the form of down-regulation of SLP-2 expression to decrease T cell reactivity (e.g., in the course of autoimmune disease) or alternatively, enhancement of SLP-2 expression to increase T cell responsiveness (e.g., vaccine development).
We thank Drs. E. H. Ball, E. Cairns, M. Chalfie, T. L. Delovitch, H. M. McBride, B. Singh, and the members of the Madrenas laboratory for many insightful discussions.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by grants from the Canadian Institutes of Health Research, the Kidney Foundation of Canada (Allison Knudsen Research Award), and the Multi-Organ Transplant Program of the London Health Sciences Centre (to J.M.); National Institutes of Health Grant R01-AI43549 (to M.L.D.); National Institutes of Health Training Grant T32-GM07308 (to S.V.); and FONACIT S12002000575 (to I.B.). M.G.K. is the recipient of a Canadian Institutes of Health Research MD/PhD studentship, and J.M. holds a Canada Research Chair in Immunobiology.
Abbreviations used in this paper: IS, immunological synapse; LAT, linker for activation of T cells; PLC, phospholipase C; SEE, staphylococcal enterotoxin E superantigen; siRNA, small interfering RNA; SLP, stomatin-like protein; SPFH, stomatin/prohibitin/flotillins/HflK-HflC.