Activation of the Ancestral Polarity Regulator Protein Kinase Cζ at the Immunological Synapse Drives Polarization of Th Cell Secretory Machinery toward APCs

T helper cells are known to polarize their secretory machinery toward the immunological synapse (IS) formed with cognate APCs (1–3).Three major features characterize Th cell polarization toward APCs. First, this event is a rapid phenomenon taking place before the large-scale molecular segregation characteristic of a stable IS (4–6). Second, T cell polarization is a low-threshold response. Indeed, Th cells exhibit polarization of their Golgi apparatus even in conditions in which IS formation is blocked (7). Accordingly, cytotoxic T cells can polarize their lytic machinery toward target cells offering weak antigenic stimuli, insufficient to induce the formation of stable IS (8, 9). Third, polarization of T cell secretory machinery is a versatile response that can be rapidly adjusted by changes in the intensity and direction of antigenic stimulation (5, 6).

Although several signaling molecules (including ζ-chain–associated protein kinase 70, linker for activation of T cells, Src homology 2 domain-containing leukocyte protein of 76 kDa, VAV-1, and diacylglycerol) and cellular effectors (including histone deacetylase 6 tubulin deacetylase, dynein, proteins of the formin family, and components of the intraflagellar transport) have been implicated in the polarization of Th cell tubulin cytoskeleton and Golgi apparatus toward APCs (10–17), the molecular guides linking TCR early signaling to selective T cell polarization are still not well characterized.

A second unresolved issue concerns the impact that the rapid establishment of T cell polarity during Ag recognition might have on APC biology. After the initial observation that B cells facing Th cell polarization are the ones preferentially activated to proliferation (18), the effect of dedicated Th cell polarization on APC function has not been further investigated.

In the current study, we provide answers to both open issues.

We identify atypical PKCζ, an evolutionary conserved regulator of cellular polarity and of asymmetric cell function (19), as a key molecular guide for Th cell polarization toward dendritic cells (DCs) and show that Th cell polarization has an impact on DC biology. We show that 1) in primary human T cells, PKCζ is rapidly and selectively activated at the Th cell/DC IS; 2) a functional PKCζ is required for the polarization of the Th cell secretory machinery toward DCs; and 3) PKCζ-dependent polarization allows a dedicated secretion of CD40L and IFN-γ and is required for activation of cognate DCs to IL-12 production.

All in all, our results show that the PKCζ pathway couples TCR triggering to Th cell polarization responses that in turn activate DCs to IL-12 production.

T cells and DCs were isolated from whole blood of healthy donors (Centre Hospitalier Universitaire Purpan, Toulouse, France). The CD4⁺ T cell fraction was sorted from whole blood using RosetteSep (StemCell Technologies, Vancouver, British Columbia, Canada). CD4⁺Vβ2⁺ T cells were isolated by positive selection using an anti-Vβ2 Ab (clone MPB2D5; Beckman Coulter, Fullerton, CA) and goat anti-mouse IgG microbeads (Miltenyi Biotec, Auburn, CA). Cell purity was assessed by FACS analysis (FACSCalibur; BD Biosciences, San Jose, CA) using PE-labeled anti-CD4 mAb (clone RPA-T4; BD Pharmingen, San Diego, CA) and FITC-labeled anti-Vβ2 mAb (clone MPB2D5; Beckman Coulter). CD4⁺Vβ2⁺ fractions were routinely 92–98% pure. Freshly isolated CD4⁺Vβ2⁺ T cell populations were cultured in RPMI 1640 medium, 5% human serum, and IL-2 (250 U/ml) in the presence of anti-CD3/CD28 mAb–coated Dynabeads (Invitrogen, Carlsbad, CA) at a ratio of one bead for one T cell. Lines were used between 2 and 4 wk of culture.

For DC preparation, CD14⁺ monocytes were isolated by positive selection using anti-CD14–coated beads (Miltenyi Biotec, Bergisch Gladbach, Germany). Monocytes were cultured in RPMI 1640 medium and 5% FCS supplemented with IL-4 (100 U/ml) and GM-CSF (750 U/ml; both from R&D Systems, Minneapolis, MN). The percentage of CD11c⁺ cells in monocyte-derived DCs was checked by FACS analysis after 5 d of culture using anti-CD11c mAb (clone B-ly6; BD Pharmingen) and was routinely 95–98% pure. EBV-transformed B cells (LG2) were cultured as described previously (5). Blood samples from healthy donors were obtained following standard ethical procedures (Helsinki protocol) and with the approval of the concerned Internal Review Boards.

DCs were either unpulsed or pulsed with 10 or 0.1 ng/ml of the bacterial superantigen toxic shock syndrome toxin-1 (TSST-1) (Toxin Technology, Sarasota, FL) for 1 h at 37°C in RPMI 1640 medium and 5% FCS. During the last 15 min of pulsing, DCs were stained with 5-(and-6)-(((4-chloromethyl)benzoyl)amino)tetramethylrhodamine (CMTMR)-Orange (red; Molecular Probes, Eugene, OR) or BODIPY 630 (blue; Molecular Probes) or 5-chloromethylfluorescein diacetate (CMFDA)-green (green; Molecular Probes). Alternatively, DCs were loaded for 45 min with indo-1-AM (cyan; Invitrogen). In some experiments, EBV-B cells were used as APCs and were either unpulsed or pulsed with 100 ng/ml of the bacterial superantigen TSST-1 for 1 h at 37°C in RPMI 1640 medium and 5% FCS. After washing, cells were conjugated by 1 min of centrifugation with CD4⁺Vβ2⁺ T cells at a ratio of one T cell for two DCs. At different times after conjugation, cells were fixed, permeabilized with 0.1% saponin (in PBS 3% BSA/HEPES), and stained with the following primary Abs: anti-phosphotyrosine (p-Tyr) mAb (clone sc-7020; Santa Cruz Biotechnology, Santa Cruz, CA), anti–α-tubulin mAb (clone DM1A; Sigma-Aldrich, St. Louis, MO), anti–IL-12p40 mAb (clone C11.5; BD Biosciences), anti–phospho-PKCζ (p-PKCζ) rabbit Ab (sc-12894-R; Santa Cruz Biotechnology), anti-GM130 mAb (clone 35/GM130; BD Biosciences), anti–IFN-γ (clone B27, BD Biosciences), anti-CD3ζ (clone 6B10.2; Santa Cruz Biotechnology), or anti-CD40L mAb (clone TRAP1; BD Biosciences). Primary Abs were followed by goat anti-mouse isotype-specific Ab or goat anti-rabbit labeled with Alexa 488, Alexa 633, or Alexa 647 (Molecular Probes). In some experiments, T cells were pretreated with a myristoylated PKCζ pseudosubstrate peptide (PKCζ-PS) at 10 μM (Invitrogen) during 1 h at 37°C and washed. In additional experiments, CD4⁺Vβ2⁺ T cells were cotransfected with the pEGFPmax vector (Amaxa, South San Francisco, CA) and either T410A PKCζ vector or empty vector using an Amaxa nucleoporator. The samples were mounted in 90% glycerol-PBS containing 2.5% 1,4-diazabicyclo[2.2.2]octane (Fluka, Geneva, Switzerland) and examined using either a Zeiss LSM 510 or a Zeiss 710 confocal microscope (Zeiss, Oberkochen, Germany) with a ×63 Plan-Apochromat objective (1.4 oil), as described previously (5). For each pair of Abs used, standardized conditions for pinhole size, and for gain and offset (brightness and contrast), were used for image capture. For the given settings, staining with secondary Ab only did not result in detectable fluorescence.

To evaluate the polarization of Th cell secretory machinery toward DCs, T cell/APC conjugates were randomly selected and distances of microtubule-organizing center (MTOC), CD40L, IFN-γ, or Golgi apparatus from the center of the T cell/APC contact site were measured, on unprocessed images, using the Profile function of the Zeiss software (20). To score the intensity of p-PKCζ and CD3ζ at the IS, unprocessed images were analyzed using the Linescan function of the MetaMorph software (Universal Imaging, Downingtown, PA) as described previously (5). Briefly, a reference line was drawn at the center of the contact site. To exclude p-PKCζ staining coming from the DCs, lines were drawn from the cell–cell contact toward the inside of the T cell (see schemes in Figs. 1, 6). The software calculates the mean green fluorescence intensity all along the reference lines for 25 pixels of width (12 and 12 laterally to reference line) and plots the measurements. The integral of the curves obtained are used to define the intensity of p-PKCζ or CD3ζ staining. Experiments presented in Figs. 1 and 6 and Supplemental Fig. 5 were performed using either a Zeiss LSM 510 (Fig. 1) or a Zeiss LSM 710 (Fig. 6, Supplemental Fig. 5). Even though the basal values in p-PKCζ staining detected by the two microscopes were different, the two analyses showed similar increments in p-PKCζ in TSST-1–stimulated versus unstimulated T cells.

DCs or EBV-B cells were either unpulsed or pulsed with 10 or 100 ng/ml TSST-1, respectively, for 1 h at 37°C in RPMI 1640 medium and 5% FCS. During the last 15 min of pulsing, APCs were stained with CMTMR-Orange (red; Molecular Probes). After washing, cells were conjugated by 1 min of centrifugation with CD4⁺Vβ2⁺ T cells at a ratio of one T cell for two DCs. After 30 min of conjugation, conjugates were broken in PBS-EDTA, and cells were fixed, permeabilized with 0.1% saponin (in PBS 3% BSA/HEPES), and stained with anti–phospho-PKCζ (p-PKCζ) rabbit Ab (sc-12894-R; Santa Cruz Biotechnology), followed by an isotype-matched FITC secondary Ab. For CD40L detection by FACS analysis, T cells, either untreated or pretreated with a PKCζ-PS at 10 μM during 1 h at 37°C, were conjugated with DCs either unpulsed or TSST-1 pulsed. In some experiments, brefeldin A (Sigma-Aldrich) was added to the culture at 10 μg/ml. After 6 h of culture, cells were fixed, permeabilized with 0.1% saponin (in PBS 3% BSA/HEPES), and stained with anti-CD40L mAb (clone TRAP1; BD Biosciences), followed by an isotype matched FITC labeled secondary Ab.

DCs were either unpulsed or pulsed with 10 ng/ml TSST-1 for 1 h at 37°C in RPMI 1640 medium and 5% FCS. During the last 15 min of pulsing, DCs were loaded with different concentrations of CMFDA-green (Molecular Probes) to allow their discrimination by FACS analysis. After washing, cells were cocultured with CD4⁺Vβ2⁺ T cells at a ratio of one T cell for two DCs for 16 h. Cells were stained with anti-CD40 mAb (clone 5C3; BD Biosciences), anti-CD54 mAb (clone HA58; BD Biosciences), or anti-CD80 mAb (clone L307.4; BD Biosciences). Primary Abs were followed by a goat anti-mouse isotype-specific Ab labeled with PE (Molecular Probes). In some experiments, DCs were left unstained and conjugated at a 1:2 T cell/DC ratio with CD4⁺Vβ2⁺ T cells (either untreated or treated with 10 μM PKCζ-PS during 1 h at 37°C, followed by washing).

CD4⁺Vβ2⁺ T cells and DCs (either unpulsed or pulsed with TSST-1) were conjugated at a ratio of one T cell for two DCs in 200 μl culture medium in 96-well flat-bottom microplates. In some experiments, T cells were pretreated during 1 h with 10 μM PKCζ-PS or 10 μg/ml anti-CD154 Ab (clone TRAP1; BD Biosciences) or isotype control Ab (BD Pharmingen) was added to the culture medium. After 16 h, the concentration of IFN-γ and IL-12p70 was measured in the cell supernatant using a cytometric bead array (CBA), according to the manufacturer’s procedure (BD Biosciences).

[Ca²⁺]_i and TCR downregulation were measured in T cells conjugated with DCs, respectively, 30 min and 4 h after conjugate formation as described previously (8).

PKCζ is activated by phosphorylation in a regulatory threonine (Thr⁴¹⁰) (21). Staining for the phosphorylated form of PKCζ can be therefore used to detect PKCζ activation. Conjugates were formed between human CD4⁺Vβ2⁺ T cells and human DCs either unpulsed or pulsed with the bacterial superantigen TSST-1. After 30 min, fixed and permeabilized Th cell/DC conjugates were stained with an Ab directed against p-Tyr (an early marker of signaling at the IS) (5) and with an Ab directed against p-PKCζ (PKCζ phosphorylated in the regulatory Thr⁴¹⁰). As shown in Fig. 1A, a localized phosphorylation of PKCζ was observed at the IS in T cells interacting with cognate DCs (in parallel with p-Tyr staining) but not in unspecific conjugates. Measurement of p-PKCζ staining intensity at the IS showed increased PKCζ phosphorylation in a significant number of T cell/DC-specific conjugates as compared with unspecific conjugates (Fig. 1B). Similar results were obtained in Th cells interacting with EBV-transformed B cells (Supplemental Fig. 1A). In agreement with morphological data, FACS analysis showed an increased intensity of PKCζ phosphorylation in TSST-1–stimulated T cells (Supplemental Fig. 1B).

Taken together, the above results show that TCR triggering results in activation of PKCζ at the IS, raising the question of whether PKCζ might play a role in T cell polarization responses.

To investigate the role of PKCζ in T cell polarization toward APCs, we used a myristoylated pseudosubstrate peptide, which functions as a highly specific and cell-permeable inhibitor of PKCζ function (PKCζ-PS) (22). T cells were pretreated for 1 h with 10 μM PKCζ-PS. After washing, the capacity of these cells to polarize their MTOC toward IS was analyzed by confocal microscopy. As shown in Fig. 2A, in T cells treated with PKCζ-PS, MTOC polarization toward the IS formed at the contact site with TSST-1–pulsed DCs was inhibited, indicating that PCKζ function is required for T cell polarization toward the APCs. To quantify this inhibition, we measured in individual T cell/DC conjugates the distance between the MTOC and the center of the IS. This distance was significantly increased when PKCζ was inhibited in T cells (Fig. 2B). Similar results were obtained when the impact of PKCζ-PS treatment on Th cell Golgi apparatus polarization was investigated (Supplemental Fig. 2).

The role of PKCζ in T cell polarization was also investigated using a complementary approach based on the overexpression in Vβ2 Th cells of the activation loop Thr⁴¹⁰-mutant PKCζ (that keeps its ability to bind molecular partners but cannot be activated by phosphorylation) acting as a dominant-negative (23). As shown in Fig. 2C and 2D, cotransfection of an empty plasmid and a GFP-encoding plasmid had no effect on MTOC polarization. Conversely, transfection with a plasmid encoding dominant-negative PKCζ randomized MTOC positioning in T cells (Fig. 2C, 2D).

Taken together, the above results show that the PKCζ pathway is a key component of the Th cell polarization machinery toward APCs.

We next asked whether inhibition of Th cells PKCζ pathway might have an impact on APC function. In a first set of experiments, TSST-1–pulsed DCs exhibited similar levels of CD40, CD54, and CD80 upregulation after overnight interaction with T cells irrespectively of whether Th cells were previously treated or not with PKCζ-PS (Fig. 3A), indicating that Th cell-driven DC upregulation of costimulatory molecules is uncoupled from Th cell polarization. In a second set of experiments, we investigated whether the inhibition of the PKCζ pathway in T cells could inhibit IL-12 production by DCs, a function known to require the physical contact with Th cells (24, 25). This analysis showed that DCs cultured with PKCζ-PS–pretreated Th cells released lower amounts of IL-12 when compared with DCs interacting with untreated T cells (Fig. 3B).

To further define the impact of Th cell secretory machinery polarization on DC activation to IL-12 production, we investigated whether IL-12 production selectively occurs in cognate T cell/DC conjugates or can also be detected in bystander noncognate DCs. CD4⁺Vβ2⁺ T cells were simultaneously cocultured with TSST-1–pulsed and –unpulsed DCs (stained with different dyes to allow their discrimination by confocal microscopy). After 16 h, cells were fixed, permeabilized, and stained with an Ab directed against the IL-12/IL-23 p40 common chain and analyzed by confocal microscopy. The analysis was performed by randomly identifying p40-positive DCs in the fields and then verifying whether they were pulsed or unpulsed. This analysis showed that ∼90% of p40-positive cells were TSST-1–pulsed DCs (Fig. 4A, 4B). In IL-12⁺ DCs, p40 staining colocalized with Golgi apparatus as visualized by staining for the Golgi marker GM130 protein (Fig. 4C). Under similar conditions, both cognate DCs and bystander ones upregulated CD40, CD54, and CD80 after overnight cocultures with Th cells (Fig. 4D). These results indicate that only cognate DCs and not bystander ones are activated to IL-12 production, supporting the notion that PKCζ-dependent polarization of T cell secretory machinery is required for activation of DCs to IL-12 production.

The above results raised the question of which molecular mechanisms might be implicated in linking PKCζ-dependent Th cell polarization to IL-12 production by cognate DCs. To address this question, we focused on IFN-γ and CD40L because these two signals are known to be major Th-derived signals for activation of human DCs to IL-12 production (26).

To define whether PKCζ inhibition in Th cells might interfere with the delivery of these signals to DCs, we initially investigated whether the PKCζ pathway might be required for productive TCR engagement and signaling in Th cell/DC conjugates. We investigated several activation parameters in Th cells pretreated or not with PKCζ-PS before conjugation with DCs. As shown in Supplemental Fig. 3A and 3B, treatment with PKCζ -PS did not affect sustained [Ca²⁺]_i increase and TCR downregulation in T cells interacting with DCs. Moreover, as shown in Supplemental Fig. 3C and 3D, treatment with PKCζ-PS did not affect the enrichment of TCR/CD3ζ complexes into the IS formed at the contact site between Th cells and DCs, suggesting that TCR delivery to the IS uses a PKCζ-independent mechanism.

We next measured IFN-γ production and CD40L expression in Th cells, either untreated or pretreated for 1 h with PKCζ-PS. As shown in Fig. 5A, PKCζ-PS treatment did not affect IFN-γ secretion by Th cells as detected by CBA analysis in culture supernatants. For detection of CD40L upregulation in Th cells, conjugates were formed between DCs and Th cells in the presence or absence of brefeldin A (to induce intracellular accumulation of CD40L). Cells were fixed, permeabilized, and stained with anti-CD40L Abs, followed by isotype-matched secondary Abs. As shown in Fig. 5B and 5C, treatment with PKCζ-PS did not affect CD40L expression neither in the absence nor in the presence of brefeldin A. The increased staining for CD40L observed in brefeldin A-treated Th cells indicated that in our cell system, a significant fraction of CD40L was secreted, in agreement with previously reported data (27, 28). Accordingly, surface expression of CD40L in nonpermeabilized cells was barely detectable at different times after stimulation (data not shown). Nevertheless, CD40/CD40L interaction was required for IL-12 production by DCs. Indeed, block of CD40L/CD40 interaction inhibited IL-12 production to an extent similar to that obtained by inhibiting Th cell polarization (compare Fig. 3B with Supplemental Fig. 4).

These results indicate that the activation of the PKCζ function is dispensable for productive TCR engagement and signaling and for IFN-γ and of CD40L induction in Th cells.

We finally tested whether inhibition of PKCζ-dependent Th cell secretory machinery polarization could lead to a defective polarization of IFN-γ and CD40L toward cognate DCs. To this end, IFN-γ and CD40L were detected by confocal microscopy in fixed and permeabilized Th cell/DC conjugates. As shown in Fig. 5D and 5F, CD40L and IFN-γ polarization toward the IS was defective in PKCζ-inhibited Th cells. Measurement of the distance between IFN-γ or CD40L intracellular pools and the center of the IS showed a defect of polarization toward the cognate DCs in a significant fraction of PKCζ-inhibited Th cells (Fig. 5E, 5G).

Taken together, the above results indicate that a molecular link between the polarization of the Th cell secretory machinery toward cognate DCs and activation of DCs to IL-12 production is ensured by the dedicated delivery of IFN-γ and CD40L at the IS.

The observation that IL-12 production by DCs required polarization of Th cell secretory machinery raised the question of whether PKCζ-dependent polarization might be instrumental to allow Th cells to give their help for IL-12 production in a dedicated fashion to a given DCs. To address this question, we investigated whether PKCζ synaptic activation might correlate with selective Th cell polarization toward a given APCs.

To this end, we initially measured the intensity of PKCζ phosphorylation in Th cells interacting simultaneously with two DCs offering antigenic stimuli of different strength. As shown in Fig. 6A and 6B, in Th cells interacting with an individual DC pulsed with either a high (10 ng/ml) or a low (0.1 ng/ml) TSST-1 concentration, PKCζ was phosphorylated at both concentrations, indicating that, in Th cells, the activation of this ancestral regulator of cellular polarity occurs at a low threshold. When Th cells were simultaneously conjugated with DCs offering a high and low stimulus, PKCζ phosphorylation mostly occurred at the IS formed with the DCs displaying the strongest stimulation, whereas it was inhibited at the contact site with the DCs offering the weaker stimulus (Fig. 6A, 6B).

We next monitored PKCζ phosphorylation in parallel with the position of the MTOC using four-color confocal microscopy on three-cell conjugates in which one T cell was simultaneously in contact with two different DCs during 10 min. This approach showed that MTOC polarized toward the area of PKCζ phosphorylation at the contact site with the DCs offering the strongest stimulus (Fig. 6C). Measurements on three-cell conjugates of p-PKCζ fluorescence intensity and of the distance in micrometers of the MTOC from the two synapses are plotted in Fig. 6D. The figure shows that a short distance of the MTOC from the IS formed with the DCs offering the strongest stimuli (mean distance of 1.43 μm) correlated with a more intense p-PKCζ staining at the IS (mean p-PKCζ intensity 974). Conversely, a longer distance of the MTOC from the IS formed with the DCs providing weaker stimuli (mean distance, 6.02 μm) correlated with a modest p-PKCζ staining at the synaptic area (mean p-PKCζ staining, 440).

We also tested the effect that PKCζ-PS might have on MTOC polarization of Th cells simultaneously in contact with two DCs offering different stimuli. As shown in Supplemental Fig. 5, inhibition of PKCζ function in Th cells randomized MTOC polarization.

Taken together, the above results show that PKCζ phosphorylation at the Th cell synapse is selectively triggered at the contact site with the DCs offering the strongest stimulus and correlates with the “polarization choice” of the Th cell secretory machinery.

In the current study, we show that the atypical PKCζ links TCR triggering in Th cells to polarization of their secretory machinery toward APCs, thus contributing to the selectivity and efficiency of Th cell function.

We provide two lines of evidence supporting the role of the PKCζ pathway in the establishment of polarity in T cells. We show that 1) following TCR triggering, PKCζ is activated at the T cell/APC contact site within minutes after cell/cell conjugation; and 2) PKCζ function is necessary for reorientation of T cell secretory machinery toward APCs. This second observation results from two complementary approaches: treatment of T cells with a selective pseudosubstrate of PKCζ and overexpression of the activation loop Thr⁴¹⁰ PKCζ mutant. Both approaches resulted in the inhibition of T cell MTOC reorientation toward the IS, indicating that T cell polarization responses require both the enzymatic activity of PKCζ (inhibited by PKCζ-PS) and its binding to molecular partners, such as Par6 (inhibited by Thr⁴¹⁰-mutant overexpression).

The PKCζ pathway plays a central role in shaping polarity and in asymmetric functions of different cellular systems (e.g., epithelial cell polarity, asymmetric cell division, and cell migration) (19, 22, 23). Our results show that T cells activate PKCζ cascade following TCR engagement to guide the rapid polarization of their secretory machinery toward APCs. They extend recently reported data showing that 1) functional Par1b, a polarity protein downstream of PKCζ, regulates T cell polarization responses (29); and 2) the scribble complex (another ancestral polarization regulator, linked to the PKCζ pathway) (30), is implicated in T cell polarization responses both in the early and in the late phases of T cell/APC interaction (31, 32).

We also show that, in our cellular system, the blockade of T cell polarization does not affect productive TCR engagement and signaling, TCR/CD3ζ enrichment at the IS and activation of Th cell effector function. Our results therefore indicate that polarization of the Th cell secretory machinery is a consequence and not a prerequisite for signaling in T cells.

These results are in agreement with studies reporting that interference with the polarization machinery does not affect TCR-mediated signaling or IS formation in T cells (10, 13, 29) but are in apparent contrast with others reporting that T cell secretory pathway participates to the assembly of the signaling cascade at the IS (12, 17, 33). The use of different cellular models and the targeting of different molecular steps of the T cell secretory pathway in the reported studies could, at least in part, explain this discrepancy.

Our results show that IL-12 production by DCs, but not upregulation of costimulatory molecules, is inhibited when Th cell polarization is blocked. Accordingly, in Th cells interacting simultaneously with cognate and noncognate DCs, only cognate DCs are activated to IL-12 production, whereas noncognate DCs can upregulate costimulatory molecules in a bystander fashion. These observations reveal that dedicated Th cell polarization has indeed an impact on some but not all DC biological responses. These results extend previously published data showing that coculture of DCs with Th cells, γδ T cells, or NK cells activates DCs to IL-12 production (24, 25, 34, 35), whereas upregulation of costimulatory molecules by DCs can be triggered in the absence of direct T cell/DC contact (25, 24).

What could be the mechanisms underlying this apparent split activation of DCs? It is tempting to speculate that on one hand, the PKCζ-dependent polarization of the Th cell secretory machinery toward the IS favors the targeted delivery of IFN-γ and CD40L into the synaptic cleft in a concentration sufficient to trigger the activation of IL-12 production by cognate DCs (36). On the other hand, the release of cytokines into the extracellular environment may be sufficient to elicit the observed bystander upregulation of costimulatory molecules in noncognate DCs. This view is in line with the demonstration of the coexistence of two main secretory pathways in Th cells: a secretory pathway directed toward the IS (that could be mostly implicated in dedicated APC activation) and a multidirectional secretory pathway (possibly implicated in activation of bystander cells) (37).

Our results also shed new light on the molecular mechanisms of polarization in Th cells interacting with multiple APCs simultaneously. In these conditions, T cell polarization is known to be remarkably versatile, allowing Th cells to dedicate their secretion to the APCs offering, at any given time, the strongest antigenic stimulus (5, 6). Our results identify PKCζ as a possible candidate for the flexibility of Th cell polarization responses. We show that 1) PKCζ activation at the IS is saturated at relatively weak stimuli; and 2) in T cells interacting simultaneously with different DCs, a dominant focus of PKCζ synaptic activation is formed at the contact site with the APCs offering the strongest stimulus and correlates with the “polarization choice” of Th cells while, at the opposing synaptic contact, PKCζ activation is turned off.

All in all, our study indicates that the low threshold and selective activation of PKCζ at the Th cell/APC synapses might serve as a molecular switch allowing, on the one hand, Th cell-sensitive polarization responses and, on the other hand, APC discrimination for dedicated activation of cognate APCs. Further research is required to define whether the observed link between physical Th cell polarization toward DCs and IL-12 production might play a role in modulating the functional Th cell polarization during adaptive immune responses.

We thank Loïc Dupré, Eric Espinosa, Sabina Müller, and Abdelhadi Saoudi for discussion and critical reading of the manuscript. We also thank the operators of the Plateau Technique de Cytométrie, Institut Fédératif de Recherche 150, Toulouse and of the Plateau Technique d’Imagerie Cellulaire, Institut Fédératif de Recherche 150 Toulouse for assistance.

Disclosures The authors have no financial conflicts of interest.