The Src family kinase Lck is essential for T cell Ag receptor-mediated signaling. In this study, we report the effects of acute elimination of Lck in Jurkat TAg and primary T cells using RNA interference mediated by short-interfering RNAs. In cells with Lck knockdown (kd), proximal TCR signaling was strongly suppressed as indicated by reduced ζ-chain phosphorylation and intracellular calcium mobilization. However, we observed sustained and elevated phosphorylation of ERK1/2 in Lck kd cells 30 min to 2 h after stimulation. Downstream effects on immune function as determined by activation of a NFAT-AP-1 reporter, and TCR/CD28-stimulated IL-2 secretion were strongly augmented in Jurkat and primary T cells, respectively. As expected, overexpression of SHP-1 in Jurkat cells inhibited TCR-induced NFAT-AP-1 activation, but this effect could be overcome by simultaneous kd of Lck. Furthermore, acute elimination of Lck also suppressed TCR-mediated activation of SHP-1, suggesting the possible role of SHP-1 in a negative feedback loop originating from Lck. This report underscores Lck as an important mediator of proximal TCR signaling, but also indicates a suppressive role on downstream immune function.
Engagement of the TCR with its ligand, peptide-MHC molecules on APCs, initially leads to full spatial and catalytic activation of Lck (1). This Src family kinase, in turn, phosphorylates ITAMs found within the TCR-associated CD3 and ζ-chains (2) allowing binding of Zap-70, which subsequently phosphorylates the linker for activation of T cells (LAT)3 (3, 4, 5). LAT thereby directly or indirectly provides docking sites for an array of proteins such as Gads, Grb2, PLCγ1, SLP-76, and Vav (6). Recruitment and activation of these signaling proteins enables the mobilization of downstream messengers such as Ca2+ and Ras, leading to activation of the transcription factors NFAT, AP-1, and NF-κB, with subsequent alterations in transcription of a pleiotropic set of genes that constitute the T cell response to stimulation. Most notable is the production of IL-2, a cytokine that promotes long-term proliferation of activated T cells.
In agreement with the paradigm presented above, the JCaM1 line, an Lck-deficient variant of the Jurkat T cell line, is unable to respond to TCR ligation by anti-CD3 treatment (7, 8). Furthermore, a murine Lck null mutation created by homologous recombination shows pronounced thymic atrophy, severe block in thymocyte development, and a dramatic reduction in the number of peripheral T cells (9). Similar findings have been reported with transgenic mice expressing a dominant-negative version of Lck (10). Altogether, these results demonstrate the crucial role of Lck in T cell development and TCR signaling.
Interestingly, a suppressive role has also been attributed to Lck. Not only is Lck important for TCR desensitization by internalization (11, 12), it was also recently reported that IL-2 production in JCaM1 cells triggered by staphylococcal super Ags was elevated, suggesting that activated Lck can trigger signals that negatively regulate T cell activation (13). A potential candidate for a negative feedback loop initiated by Lck is the Src homology protein tyrosine phosphatase-1 (SHP-1), which on numerous occasions has been shown to negatively regulate TCR signaling (reviewed by Zhang et al. in Ref.14). The activity of SHP-1 is augmented by engagement of its Src homology 2 domains by tyrosine phosphorylated ligands (15), and it was recently shown that CD22 may be such a ligand in T cells (16). Additionally, SHP-1 is phosphorylated directly by Lck on Y536 and Y566, which increases its phosphatase activity further (17, 18). Conversely, Lck itself may be dephosphorylated at Y394 and thus inhibited by SHP-1 (19). SHP-1-mediated dephosphorylation also occurs for many other important mediators of T cell signaling such as Zap-70, Vav, Grb2, and SLP-76 (20, 21, 22, 23, 24). This puts SHP-1 in an interesting position, where it can be envisioned that SHP-1 is activated by Lck, both directly and indirectly, and thereby functions as a gatekeeper for T cell activation. Indeed, it has been proposed that low-affinity ligands for the TCR may trigger a negative feedback loop through the binding of the SHP-1 to Lck. This negative feedback may be circumvented by stronger ligands through phosphorylation of S59 in Lck in an ERK-dependent manner (25).
To further assess the impact of Lck on proximal and distal T cell signaling, we used short-interfering RNA (siRNA)-mediated RNA interference (RNAi) to specifically and acutely knock down Lck. By using this technology, we provide biochemical and functional evidence of the requirement of Lck for production of classical proximal T cell activation upon stimulation with anti-CD3. At the same time, Lck removal paradoxically augments downstream signaling, and we report in this study for the first time that this phenotype occurs in mature human peripheral T cells from healthy blood donors. We also provide mechanistic data which underscores that SHP-1 activity is dependent on the presence of Lck, thus suggesting a possible role for SHP-1 in the observed paradoxical downstream effects in cells with Lck knockdown (kd).
Materials and Methods
Abs and reagents
Abs for Lck, Csk, NFAT, PLCγ1, CD3-ζ, and anti-CD3-ζ Abs coated on beads were obtained from Santa Cruz Biotechnology (catalog nos. sc-433, sc-286, sc-1151, sc-81, sc-1239, and sc-1239AC). Abs for ERK, LAT, and SHP-1 were obtained from Upstate Biotechnology (catalog nos. 06-182, 06-807, and 07-419). Abs for Fyn, protein kinase C (PKC)α, and anti-CD3-FITC (SK7) were obtained from BD Transduction Laboratories (610164, 610108, and 349201). Ab for pERK1/2 (MAPK p44/p42 (p-Thr202/Tyr204)) was obtained from Cell Signaling Technology (9101). Anti-CD3/anti-CD28, anti-CD14, and anti-CD19 Abs coated on paramagnetic beads were from Dynal Biotech (111.31, 111.12, and 111.04). Monoclonal anti-CD3-ε (OKT3) for cell stimulation and anti-phosphotyrosine mAbs (4G10) were affinity-purified from supernatants of hybridoma cell lines obtained from American Type Culture Collection. Peroxidase-conjugated secondary Abs were obtained from Jackson ImmunoResearch Laboratories. U0126 was obtained from Sigma-Aldrich (U-120).
Four 21-nt siRNA duplexes targeting different positions within human Lck mRNA were designed (26) and synthesized in-house (Table I). A siRNA targeting an irrelevant gene, murine Tissue Factor (mTF223), and a mismatched siRNA for Csk (Csk2033M3) described previously (27) were included as controls in initial screenings. A control siRNA for Lck siRNA was designed by switching three G/C nucleotides (Lck232M3), and this oligo pair was used in all subsequent experiments (Control). The sequences of all siRNA strands used in this report are shown in Table I. All oligos are named according to the position of the 5′ nucleotide of the sense strand relative to the reference sequences of the respective target mRNAs (Lck, NM_005356; Csk, NM_004383; mTF, M26071; Fyn, NM_002037).
Cell culture, T cell purification, and transfections
HaCaT cells were cultured in Keratinocyte-SFM (Invitrogen Life Technologies) supplemented with 0.2 ng/ml human recombinant epidermal growth factor, 30 μg/ml bovine pituitary extract, and penicillin/streptomycin. Transfections were conducted at 40–50% cell confluency and performed with siRNA (10–100 nM) and/or cDNA in complex with 5:2 v/w lipofectamine2000 (Invitrogen Life Technologies). The Jurkat TAg T cell line, a derivative of Jurkat cells stably transfected with SV40 large T Ag, was cultured in RPMI 1640 (Invitrogen Life Technologies) supplemented with 10% FCS, 1:100 penicillin, 1:100 pyruvate, and 1:100 nonessential amino acids (complete medium). Transfections were conducted at a cell concentration of 5–8 × 105 cells/ml. Cells (20 × 106) were washed and resuspended in 400 μl of OptiMEM (Invitrogen Life Technologies), mixed with siRNA and/or cDNA, and electroporated at 250V/950 μF in 4-mm cuvettes. Thereafter, the cells were incubated for 48 h in complete medium before harvesting. Primary CD3+ T cells from healthy Norwegian blood donors were purified by negative selection as described elsewhere (28). Briefly, PBMCs were subjected to density gradient (Lymfoprep; Nycomed-Nyegaard) centrifugation, and leukocytes were acquired and subjected to negative selection using beads coated with anti-CD14 and anti-CD19. Purified T cells were transfected with siRNA in accordance with the manufacturer’s instructions using the Amaxa nucleofector and kit (catalogue no. VPA-1002). The cells were then incubated for 48 h in complete medium before harvesting.
Northern blot analysis
HaCaT cells were harvested 24 h posttransfection for isolation of mRNA using Dynabeads oligo(dT)25 (Dynal Biotech). Isolated mRNA was fractioned by electrophoresis for 16–18 h on 1.3% agarose/formaldehyde (0.8 M) gels, transferred onto nylon membranes (MagnaCharge; Micron Separations), and hybridized with Lck (1.2-kb XcmI/BamHI fragment) and GAPDH cDNA probes in PerfectHyb hybridization buffer (Sigma-Aldrich). Hybridized membranes were scanned using a Storm 860 scanner and analyzed with ImageQuant software (Molecular Dynamics). GAPDH was used as a loading and normalization control.
Cell stimulation, lysis, and immunoprecipitations (IPs)
Before all stimulations, cells were preincubated at 37°C for 5 min. Jurkat cells (5 × 107 cells/ml in RPMI 1640) were stimulated with anti-CD3ε Abs (OKT3 at 1500 ng/ml, if not otherwise stated) for the indicated periods of time at 37°C. Cells were lysed in ice-cold lysis buffer (50 mM HEPES (pH 7.4), 100 mM NaCl, 1% Triton X-100, 5 mM EDTA, 1 mM PMSF, 1 mM Na3VO4, 50 mM NaF, and 10 mM NaPPi). After 30-min incubation on ice and vortexing, the lysates were centrifuged at 13,000 × g for 10 min before subjection to either SDS-PAGE/immunoblotting or IP. For SHP-1 IPs, the cells were lysed in SHP-1-IP-buffer, 20 mM Tris (pH 8.0), 1% Triton X-100, 137 mM NaCl, 2 mM EDTA, 1 mM Na3VO4, 5 mM NaF, and 1 mM PMSF. For IPs, Abs or Ab-coated beads were added and incubations continued at 4°C overnight, then protein A/G Sepharose beads (Santa Cruz Technology) were added if necessary, followed by additional 60-min incubation. Thereafter, immune complexes were washed three times in lysis buffer and subjected to SDS-PAGE/immunoblotting.
The NFAT and AP-1 elements of the proximal IL-2 promoter were inserted into a firefly luciferase reporter construct. Cells were cotransfected with this construct and TK-Renilla-luciferase (Promega) at a ratio of 10:1 in the presence of siRNA. Forty-eight hours posttransfection, the cells were stimulated for different periods of time with OKT3 (different concentrations). Thereafter, cells were lysed and assayed for dual luciferase activity according to the manufacturer’s instructions (Promega; catalog no. E1960). NFAT-AP-1 activation in each sample was normalized against TK Renilla luciferase activity.
Primary T cells 48 h posttransfection were stimulated with anti-CD3/anti-CD28 coated beads (variable numbers of beads per cell) for 20 h, thereafter supernatants were harvested and the concentration of IL-2 assessed by ELISA according to the manufacturer’s instructions (R&D Systems; catalog no. D2050).
Intracellular Ca2+ measurements
Jurkat T cells (48 h posttransfection) were resuspended to 2.5 × 106 cells/ml in RPMI 1640 supplemented with 2% FCS, 4 μM Fluo-4 (Molecular Probes), and 0.5% pluronic acid (Sigma-Aldrich), and incubated at 37°C for 45 min. The cells were subsequently washed twice in HBSS (Invitrogen Life Technologies) w/2.5 mM Probenecid (Sigma-Aldrich), and 5 × 106 cells/ml were allowed to settle for 30 min in the dark at room temperature. Cells (5 × 105; 100 μl) were added to a 96-well plate, and intracellular calcium concentrations were measured using a Fluoroscan Ascent FL (Labsystems) after stimulation with 1500 ng/ml OKT3.
SHP-1 phosphatase assay
Jurkat T cells were transfected with siRNA (100 nM), and 48 h posttransfection the cells were stimulated for different periods of time with OKT3 (1500 ng/ml) and lysed in SHP-1-IP-buffer (see above) for 20 min. The cell lysates were centrifuged for 10 min at 13,000 × g, precleared with protein A/G Sepharose beads for 30 min, and subjected to IP with anti-SHP-1 coupled to agarose (Santa Cruz Biotechnology) for 4 h. The IPs were washed three times in lysis buffer and split in two. One part was subjected to SDS-PAGE/immunoblotting to ensure equal loading of SHP-1 in the immunoprecipitates, the other part was washed twice in SHP-1 assay buffer (5 mM EDTA and 50 mM HEPES; pH 7.4) and assayed for phosphatase activity in duplicates. Each sample was incubated in 200 μl of SHP-1 assay buffer supplemented with 10 mM pNPP (Sigma-Aldrich) and 10 μM DTT (Sigma-Aldrich) at 30°C with shaking for 1 h. Reactions were stopped by addition 50 μl of 3 M NaOH, and OD was measured at 405 nm. The readings were corrected for background and normalized against Western blots showing SHP-1 immunoprecipitates for each experiment (quantified using Scion Image for Windows; Scion).
TCR internalization assay
Forty-eight hours posttransfection, Jurkat T cells were stimulated for different periods of time with OKT3 (1500 ng/ml), washed twice in ice-cold PBS, fixed in 4% paraformaldehyde (Sigma-Aldrich) for 5 min at 37°C, washed twice in PBS, and resuspended in PBS w/1% BSA and anti-CD3-FITC for 30 min at 4°C. Thereafter the cells were washed three times in PBS w/1% BSA and subjected to flow cytometric analysis for detection of viable CD3-positive cells (analyzed using FlowJo software; Three Star).
Selection of transfected cells
Jurkat TAg T cells were cotransfected with 2 μg of pMACS Kk.II selection vector encoding a truncated murine H-2Kk surface marker, siRNA (100 nM) and NFAT-AP-1 luciferase construct (10 μg), and subjected to a MACSelection procedure according to the manufacturer’s instructions. Briefly, paramagnetic microbeads coupled to mAbs toward the surface marker were used to label the cells 24 h posttransfection. The labeled cells were separated magnetically on a column from the nonlabled nontransfected cells, and grown for another 24 h before being used in additional experiments. Cells from the pooled original, negatively and positively selected fractions were labeled with MACSelect control FITC Abs and analyzed by flow cytometry for determination of the enrichment rate.
We have studied the effects of acute elimination of Lck in Jurkat and primary T lymphocytes using siRNA-mediated RNAi. HaCaT keratinocytes were chosen for screening of the duplexes because siRNA transfection is consistently high in this cell line. Northern blot analysis demonstrated that all of the designed duplexes were highly effective in depleting Lck mRNA (Fig. 1,A). We further showed that all four duplexes were capable of reducing the protein levels of both cotransfected Lck-wild type (wt) (Fig. 1,B) and endogenous Lck (Fig. 1,C) in HaCaT cells, demonstrating the effectiveness of these oligos. To validate specificity, we designed a mutated version of the most effective siRNA in which three G/C pairs in position 1, 7, and 13 relative to the 5′ end of the sense strand were inverted. We next tested the control duplex (Lck232M3) and the two most effective duplexes (Lck232 and Lck377) in Jurkat T cells. Whereas the mutated control siRNA was completely devoid of effect, both Lck-specific siRNAs were able to reduce Lck protein levels while the expression of other proteins was essentially unaffected (Fig. 1 D). The kd was substantial after 24-h incubation, but no additional nor diminished effect was seen in prolonged incubations for up to 72 h (data not shown). siRNA concentrations as low as 10 nM resulted in a significant kd in Jurkat T cells. Conversely, a threshold for Lck kd was reached at ∼100 nM of siRNA, whereas siRNA concentrations as high as 1500 nM did not affect the expression of other proteins (data not shown). The use of siRNA thereby produced a specific, potent, and acute depletion of Lck in a time- and concentration-dependent manner. Unless otherwise stated, all experiments in the rest of this study were performed using the optimized conditions, 100 nM Lck232M3 as control and 100 nM Lck232 for Lck kd with 48 h of incubation after transfection.
We next investigated how proximal TCR signaling in Jurkat T cells was affected by the elimination of Lck. As is evident from Fig. 2, A and B, both overall tyrosine and specifically ζ-chain phosphorylation were strongly reduced upon stimulation of the TCR. Similarly, both basal and TCR-induced intracellular Ca2+ levels were suppressed (Fig. 2,C), and the phosphorylation of ERK1/2 was acutely reduced (Fig. 2 D). Taken together, these data accentuates the requirement of Lck to produce the classical features of proximal T cell activation upon TCR ligation with anti-CD3 Abs.
To our surprise, the effects of Lck kd on downstream T cell effector function were contrary to the initial events reported above. Anti-CD3-induced activation of the proximal IL-2 promoter (containing NFAT and AP-1 elements) was greatly accelerated and augmented in the kd state, >4-fold increased over control after stimulation for 6 h with 1500 ng/ml OKT3 (4.6 ± 0.5-fold increase, mean ± SEM; n = 8; Figs. 3, A and B). These results were also confirmed using another Lck-specific siRNA as a control for possible off-target siRNA effects (Lck377; data not shown). Moreover, endogenous dephosphorylation of NFAT was surprisingly comparable in cells with Lck kd and control cells as depicted in Fig. 3,C. Fig. 3 D is provided as a control to rule out the possibility of delayed expression or degradation of Lck and other proteins throughout the time course of the experiments (0–6 h of stimulation with OKT3).
To validate the unexpected data from the NFAT-AP-1 luciferase assays, we coexpressed a selection marker and performed a selection procedure to sort out transfected cells (see Materials and Methods). We obtained >90% purity in the positively selected fraction (Fig. 4,A). As can be seen in Fig. 4,B, however, the relative kd efficiency was not higher in the selected cells. This is not surprising given the already very potent kd of Lck in the pooled population, and the fact that siRNAs are transfected 3-fold more efficiently than cDNAs in suspension immune cells (29, 30). As a consequence, there seems to be a threshold of siRNA-mediated kd efficiency, which cannot be improved by selection of transfected cells. In the subsequent NFAT-AP-1 reporter assays, the ratio of luciferase activity in Lck kd cells vs control cells was similar between the pooled and selected populations. The only difference was the relative amplitude of the luciferase counts, which was ∼30-fold higher in the selected fraction, suggesting that cells containing NFAT-AP-1-luciferase constructs had been effectively coselected (Fig. 4,B). These data indicate that the pooled population of cells can be used as a valid model for the Lck kd phenotype, and supports the data presented in Fig. 3.
Knowing that Jurkat T cells are of leukemic origin and therefore have molecular signaling defects (31), we next evaluated the effect of Lck kd in primary T cells using the Amaxa Nucleofector kit. The kd efficiency in primary T cells was comparable to that of Jurkat cells, without affecting the expression of other proteins (Fyn, Csk, PKCα, PLCγ1, LAT, FAK, and Pyk2; Fig. 5,A and data not shown). The impact of Lck kd on downstream signaling in primary T cells was then assessed. As seen in Fig. 5 B, IL-2 secretion in cells with Lck kd was significantly augmented upon stimulation using beads coated with anti-CD3 and anti-CD28. With a 1:1 bead-to-cell ratio, Lck kd cells secreted 50% more IL-2 than control transfected cells (53.4 ± 11.5%, mean ± SEM; n = 4; p = 0.01).
We reasoned that the observed paradoxical effects of Lck kd on downstream TCR signaling could be attributed to either altered receptor expression/turnover or dysregulation of molecules normally inhibiting TCR signaling. We first investigated whether TCR internalization was affected by the acute removal of Lck. It has been shown that TCR internalization is dependent on Lck (11, 12), and decreased endocytosis or altered redistribution of the receptor complex could account for the phenotype we report. However, we failed to detect any significant disparity of TCR internalization in the Lck kd state compared with control, neither did we observe any redistribution of the receptor throughout the time course of the experiment (Fig. 6). It is currently unknown whether the internalized receptor complex may continue to signal from endosomes.
Because the NFAT-AP-1 luciferase assay lasts for 6 h, we examined whether the phosphorylation of CD3-ζ, general tyrosine phosphorylation, or the phosphorylation of the MAPKs ERK1/2 was altered when observed for hours and not only during the initial minutes following TCR activation. As can be seen in Fig. 7,A, the phosphorylation of CD3-ζ was suppressed in Lck kd cells also after prolonged stimulation, but the general tyrosine phosphorylation as determined by 4G10 blot was comparable in control and Lck kd cells (Fig. 7,B). Interestingly, the phosphorylation of the MAPKs ERK1/2 was augmented in Lck kd cells, 2–3-fold over control (2.70 ± 0.36-fold after 1 h of stimulation, mean ± SEM; n = 7; Fig. 7,C), and it remained higher from 30 min to 2 h of stimulation (Fig. 7,C). This indicates a role for ERK1/2 in producing the hyperactive phenotype of Lck kd cells. ERK1/2 is known to activate c-Fos and c-Jun, which constitute the transcription factor AP-1, which is part of the proximal IL-2 promoter. This notion was further accentuated by using the pharmacological MEK1/2-inhibitor U0126. Treatment of the cells with U0126 strongly suppressed the previously observed increase in NFAT-AP-1 reporter activity in Lck kd cells (Fig. 7,D). kd of Fyn and concomitant kd of Lck and Fyn, in contrast, had only a modest inhibitory influence on NFAT-AP-1 activity (Fig. 7 D), indicating that Fyn does not play a significant role in producing the hyperactive phenotype.
Next, we investigated the involvement of Lck in the control of negative regulators of TCR signaling. We focused on the possible involvement of the tyrosine phosphatase SHP-1. It has previously been shown that SHP-1 acts as a negative regulator of T cell activation and maturation, and that Lck contributes to the activation of this phosphatase (14). Indeed, an interesting association is that found between SHP-1 and SLP-76 (32, 33). SLP-76 has been shown to interact with Grb2 (34), and may therefore be critical for relaying TCR signaling to the Ras/MAPK pathway (22). Such a notion may explain the sustained pattern of elevated ERK1/2 phosphorylation we report for the Lck kd phenotype, and SHP-1 may thus serve as the negative link between Lck and downstream effector functions. Overexpression of SHP-1 in Jurkat cells inhibited TCR-induced NFAT-AP-1 activation, but this effect could be overcome by simultaneous kd of Lck (Fig. 8,A). Furthermore, the phosphorylation of SHP-1 was reduced upon down-regulation of Lck (Fig. 8,B), and the phosphatase activity of both overexpressed (data not shown) and endogenous SHP-1 immunoprecipitated from whole cell lysates was decreased in the Lck kd state upon stimulation (3.6 ± 1.0-fold decrease in phosphatase activity in Lck kd vs control after 30 min of stimulation with OKT3, mean ± SEM; n = 3; Fig. 8,C). Accordingly, Jurkat TAg T cells cotransfected with SHP-1 Y538F,Y566F, a mutant of SHP-1 that may not be phosphorylated by Lck on the two putative phosphorylation sites Y538 and Y566, displayed augmented activation of NFAT-AP1 compared with wt-transfected cells (Fig. 8,D). Likewise, the phosphorylation of ERK1/2 after prolonged stimulation with anti-CD3 (6 h) was also stronger in cells transfected with SHP-1 Y538F,Y566F compared with control or wt-transfected cells (Fig. 8,E). Altogether, these data suggest that Lck is necessary for the functional activation of SHP-1, and that a negative feedback loop involving SHP-1 is dependent on the presence of Lck (Fig. 8 F).
The Src family kinase Lck is thought of as the main conveyer of the signals occurring after engagement of the TCR with peptide-MHC. The current paradigm of TCR signaling predicts a correlation between proximal and distal signaling events, and accordingly, inhibition of either TCR/CD3 phosphorylation or the mobilization of intracellular calcium should result in diminished activation of the NFAT/AP-1 transcription factors and production of IL-2.
In so far as siRNA-mediated kd of Lck strongly inhibited the classical features of TCR signaling, the data presented in this report are in agreement with the classical role ascribed to Lck. The consequences on downstream signaling events, in contrast, are clearly in opposition to such a model. Lck kd exhibited elevated and prolonged phosphorylation of the MAPKs ERK1/2, the activation of an NFAT-AP-1 reporter was strongly augmented, and likewise secretion of IL-2. This suggests the following: 1) parallel pathways of T cell activation after engagement of the TCR are in operation, a notion that has been proposed by others (35, 36); and 2) Lck plays a dual role in TCR signaling, being important for transferring classical TCR-induced signaling and suppressing the very same signal by initiating one or more negative feedback loops.
We report that T cell activation may indeed occur in the absence of Lck, and that such T cells are hyperactivated after stimulation of the TCR. A possible explanation for the discrepancy between the results presented in this study and similar experiments conducted with T cells from the Lck-deficient JCAM1 clone, which is defective in IL-2 production, may be the large difference in duration under which the cells have endured their respective deficiencies. JCAM1 cells and T cells from Lck−/− mice have been without Lck for long periods of time, during which compensatory and interfering mechanisms may have developed as opposed to cells with acute siRNA-mediated kd. Surprisingly, it is currently unknown whether or to what extent such compensatory mechanisms take place and how they affect the experimental system.
A more intriguing explanation may be that reduced levels of Lck results in a different phenotype all together than complete removal of the protein. The most potent kd of Lck using siRNA-mediated RNAi still rendered an endogenous pool of Lck left in the cell that could potentially produce incomplete phosphorylation of CD3-ζ and weak mobilization of intracellular Ca2+. We are currently investigating how Lck kd cells are able to produce strong activation of NFAT-AP-1 and enhanced IL-2 production when the conventional prerequisites such as CD3-ζ-chain phosphorylation and mobilization of Ca2+ are defective. One proposed model suggests signaling through Fyn-PLCγ1-DAG with activation of novel-type PKCs and Ras-GRP in a Zap-70-negative Jurkat cell line (35). This concept is intriguing because it may circumvent signaling through the classical pathway of Lck-CD3-ζ-Zap-70-LAT. However, concurrent kd of Fyn and Lck only modestly reversed the hyperactive phenotype. Our data underscore a role for ERK1/2 to which both mSOS and RasGRP signal via Ras. It has been reported that a Grb2-mSOS complex may be recruited to incompletely phosphorylated CD3-ζ, thus producing Ras signaling independently of LAT phosphorylation (36). ERK1/2 displayed sustained activation in Lck kd cells, and the MEK1/2-inhibitor U0126 prevented the strong activation of NFAT-AP-1 seen in these cells. The dephosphorylation of NFAT was similar in control and Lck kd cells, but ERK1/2 activates c-Fos and c-Jun, which constitutes the AP-1 part of the NFAT-AP-1 reporter construct. We therefore hypothesize that the hyperactivity of the NFAT-AP-1 reporter in OKT3-stimulated Jurkat TAg cells with Lck kd was due to prolonged augmented phosphorylation of ERK1/2.
It seems likely that release of inhibition by negative signaling pathways is essential to provide signal amplification in the hyperactive Lck kd phenotype. We observed that general tyrosine phosphorylation was comparable in control and Lck kd cells after prolonged stimulation (30 min to 2 h) as opposed to initial signaling (0–30 min). The tyrosine phosphatase SHP-1 is known to negatively regulate T cell signaling, and overexpression of SHP-1 inhibited the activation of NFAT-AP-1-luciferase in control cells. This inhibition could be overcome by simultaneous kd of Lck, indicating that removal of Lck also removed the stimulus for SHP-1, thus augmenting downstream immune function if parallel pathways of T cell activation exist. Moreover, the phosphorylation and phosphatase activity of SHP-1 was dependent on Lck in control cells and dysregulated in Lck kd cells. The phosphorylation of SHP-1 seemed to be important for signaling to ERK1/2 and NFAT-AP-1 because overexpression of a SHP-1 Y538F,Y566F double mutant resulted in augmented activation of both ERK1/2 and NFAT-AP-1. Altogether, the presented data indicate that SHP-1 may be an amplifier for a negative feedback loop originating via Lck to dampen T cell activation. Acute removal of Lck removes the stimulus for SHP-1 and thereby the suppressive effects of this tyrosine phosphatase, leading in total to an augmented downstream T cell response upon engagement of the TCR.
In summary, we report the effects of acute elimination of Lck in both Jurkat TAg and primary T cells using siRNA-mediated RNAi. Lck kd cells displayed a phenotype with diminished proximal response to TCR stimulation, whereas downstream T cell effector functions were paradoxically enhanced. This underscores a role for Lck both as a mediator and suppressor of T cell activation. Clearly, the signaling occurring in T cells is not completed after the initial minutes of TCR engagement. We report the augmented phosphorylation of ERK1/2 after prolonged stimulation of T cells with Lck kd, and such long-lasting activation can be significant for the final outcome of T cell activity. Whether or to what extent the hyperactive T cell phenotype we report is of physiological relevance is currently unknown. It is intriguing to speculate that such T cell signaling may play a role in the pathogenesis and progression of autoimmune diseases, a theory that remains to be investigated.
We are grateful to Guri Opsahl and Gladys Tjørhom for technical assistance. The SHP-1 Y538F,Y566F construct was a gift from Dr. Frank Böhmer (Institute of Molecular Biotechnology, Jena, Germany). We thank Jens Henrik Norum for help with the calcium measurements.
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 National Programme for Research in Functional Genomics, Research Council of Norway, Norwegian Cancer Society, Novo Nordic Foundation Committee, and European Union (RTD Grant QLK3-CT-2002-02149).
Abbreviations used in this paper: LAT, linker for activation of T cells; SHP-1, Src homology protein tyrosine phosphatase-1; siRNA, short-interfering RNA; RNAi, RNA interference; kd, knockdown; PKC, protein kinase C; IP, immunoprecipitation; wt, wild type.