Immunoregulation of Dendritic Cells by the Receptor T cell Ig and Mucin Protein-3 via Bruton’s Tyrosine Kinase and c-Src

T cell Ig and mucin protein-3 (TIM-3) was originally discovered as a Th1-specific surface protein that negatively regulates Th1 responses (1–3). However, the role for TIM-3 in innate immunity has now been established (4–6). Indeed, many innate immune cells, including dendritic cells (DCs), express TIM-3 (5). In DCs, TIM-3 plays multiple roles. For example, TIM-3 on DCs impedes antitumor immunity (7). Furthermore, TIM-3 serves as a phagocyte receptor for apoptotic cells and thus may contribute to the induction of peripheral tolerance (8). These studies suggest that TIM-3 influences the outcome of an immune response, possibly by regulating DC function. However, the reports on DC immunoregulation by TIM-3 are limiting and contradictory. Previously, TIM-3 was proposed to positively regulate DC activation (5). In contrast, an inhibitory role for TIM-3 in DC activation has been implicated by another report showing that TIM-3–deficient DCs, upon stimulation with nucleic acids, secrete more proinflammatory cytokines and type I IFNs than do wild-type DCs (7). Therefore, how TIM-3 regulates DC activation and maturation needs to be clarified, which should be determined by triggering TIM-3 signaling in DCs.

TIM-3 is a type I transmembrane glycoprotein receptor that contains an N-terminal IgV domain, followed by mucin, stalk and transmembrane domains, and a cytoplasmic tail (9). In T cells, phosphorylation of tyrosine residues in the cytoplasmic domain of TIM-3 leads to the recruitment of SH2 (Src homology 2) domain–containing proteins Fyn and PI3K (10). PI3K and its downstream effector kinase Akt are also activated in macrophages by TIM-3 signaling (6). Nevertheless, the TIM-3 signaling pathway remains largely undefined.

Previously, we have demonstrated a role for the SH2 domain–containing signal transducers Bruton’s tyrosine kinase (Btk; a member of Tec nonreceptor tyrosine kinase family) and c-Src in immunoregulation of DCs (11, 12). However, it remains unknown whether Btk and c-Src participate in TIM-3 signaling. Accordingly, we investigated how TIM-3 signaling influences DC activation and maturation, and also determined the role for Btk and c-Src in this process. We found that TIM-3, upon Ab-mediated crosslinking, induced DC suppression by blocking the NF-κB pathway, which regulates expression of several genes involved in DC activation and maturation (13, 14). In addition, Btk and c-Src served a pivotal role in mediating this immunoregulatory effect. Thus, our results provide an insight into the molecular mechanisms of immunoregulation of DCs by TIM-3.

BALB/c and C57/BL6 (B6) mice were maintained at the Institute of Microbial Technology animal facility and used at 8–12 wk of age. Use of mice was approved by the Institutional Animal Ethics Committee of the Institute of Microbial Technology (Chandigarh, India).

Bone marrow–derived DCs (BMDCs) and splenic DCs (sDCs) were prepared from male or female BALB/c or B6 mice as described (11, 13, 15).

To prevent Fc receptor binding, DCs (5 × 10⁶ cells per well) were incubated with 10 μg/ml anti-mouse CD16/CD32 mAb (BD Biosciences) for 30 min at 37°C. DCs were then treated for specified times with 10 μg/ml rat anti-mouse TIM-3 mAb (clone 215015; R&D Systems) or isotype control Ab (rat IgG2a,κ; clone RTK2758; BioLegend). Alternatively, DCs were treated with goat anti-mouse TIM-3 polyclonal Ab (pAb) (10 μg/ml; R&D Systems) or isotype control Ab (goat IgG; 10 μg/ml; R&D Systems). In some experiments, cell-free supernatants were collected after treating DCs (3 × 10⁶ cells per milliliter) with isotype control Ab or anti–TIM-3 mAb for 6 h. Supernatants were incubated for 2 h with protein A/G PLUS-Agarose beads (Santa Cruz Biotechnology) to remove free Abs and were subsequently used for treatment of fresh DCs. DCs were then washed, resuspended in complete RPMI 1640 medium (10% FBS, penicillin/streptomycin, l-glutamine, and 2-ME), and stimulated with LPS (500 ng/ml; Sigma-Aldrich) or TNF-α (20 ng/ml; R&D Systems). The viability of DCs was assessed using LIVE/DEAD Fixable Far Red Stain (Invitrogen).

Nuclear and cytoplasmic extracts were prepared as described (16). EMSA was performed using a [³²P]-labeled DNA probe containing NF-κB binding sites derived from MHC-I H2K promoter, 5′-CAGGGCTGGGGATTCCCCATCTCCACAGTTTCACTTC-3′; or an OCT-1 DNA probe (internal control), 5′-TGTCGAATGCAAATCACTAGAA-3′ (12). Bands were visualized using a phosphoimager (Fujifilm FLA-9000).

Immunoblot assays were done as described (15). Blots were probed with Abs specific for: IκBα, IκB kinase (IKK) α, PI3K p110β, PI3K p110δ, and TIM-3 (Santa Cruz Biotechnology); phosphorylated IκBα, phosphorylated IKKα-IKKβ (Ser¹⁸⁰/Ser¹⁸¹), phosphorylated JNK (Thr¹⁸³/Tyr¹⁸⁵), phosphorylated p38MAPK (Thr¹⁸⁰/Tyr¹⁸²), phosphorylated Erk1-Erk2 (Thr²⁰²/Tyr²⁰⁴), phosphorylated Btk (Tyr²²³), phosphorylated c-Src (Tyr⁴¹⁶), c-Src, JNK, p38MAPK, Erk1-Erk2, Btk, PI3K p110α, PI3K p85α, and IKKβ (Cell Signaling Technology); phosphorylated tyrosine (BD Biosciences); and β-actin (Sigma-Aldrich). Binding of secondary HRP-conjugated goat anti-rabbit or anti-mouse (Santa Cruz Biotechnology) Abs was analyzed using femtoLUCENT PLUS-HRP (G-Biosciences). Densitometric analysis was performed using ImageJ software (National Institutes of Health). In some experiments, membrane proteins or total lysates (prepared as described below) of DCs were treated with a Protein Deglycosylation Mix (New England Biolab) for 4 h prior to immunoblot analysis.

DCs were transfected with 60 nM scrambled control small interfering RNA (siRNA) or siRNA targeting Btk, c-Src, or TIM-3 (Dharmacon) using Lipofectamine RNAiMAX (Invitrogen).

Whole-cell lysates were prepared from DCs using the cell lysis buffer (Cell Signaling Technology). Membrane proteins from DCs were isolated using the FOCUS Membrane Proteins kit (G-Biosciences). Cell lysates or membrane proteins were immunoprecipitated with anti–TIM-3 Ab (clone 8B.2C12; eBioscience) or rat IgG1,κ (clone RTK2071; BioLegend), as described (11), and were analyzed by immunoblotting. Glycosylation of immunoprecipitated TIM-3 was checked on SDS-PAGE gel using a Pro-Q Emerald 488 Glycoprotein Stain Kit (Molecular Probes).

Akt kinase activity was assessed using an Akt Kinase Assay Kit (Cell Signaling Technology).

DCs (1 × 10⁶ cells per milliliter) were treated with isotype control Abs or anti–TIM-3 mAb or pAb for specified times, washed, and then stimulated with LPS for 24 h. In some experiments (for IL-10 secretion), DCs were treated only with isotype control Ab or anti–TIM-3 mAb for specified times. IL-12p70, TNF-α, and IL-10 in supernatants were measured in triplicate with ELISA kits (BD Biosciences).

The following mAbs were used for immunostaining: FITC-conjugated anti-CD40, anti-CD80, or anti-CD86 (all from BD Biosciences); and PE-conjugated anti-TLR4 (eBioscience), anti-TNFR1 (BioLegend), anti-TNFR2 (BioLegend), mouse IgG1 (eBioscience), or hamster IgG (BioLegend). Data were collected with a C6 Accuri flow cytometer (BD Biosciences) and analyzed with FlowJo software (TreeStar).

One-way ANOVA (SigmaPlot 11.0 program) was used for all statistical analyses. A p value < 0.05 was considered significant.

Because the role for TIM-3 in immunoregulation of DCs remains ambiguous (5, 7), we analyzed the modulatory effect of TIM-3 on DC activation and maturation. However, because TIM-3–specific ligand (or ligands) is (or are) currently unknown, we first tested whether TIM-3 phosphorylation could be triggered in DCs by crosslinking TIM-3 using anti-TIM-3 Ab. Unlike isotype control Ab, treatment of BALB/c BMDCs with anti–TIM-3 mAb induced Tyr (tyrosine) phosphorylation of TIM-3 with a molecular mass of ∼42 kDa (Fig. 1A). We noted this phenomenon regardless of whether we immunoprecipitated TIM-3 from whole-cell lysates (Fig. 1A, left panel) or membrane fraction of BMDCs (Fig. 1A, right panel). The Tyr phosphorylation of TIM-3 (42 kDa) was also induced after TIM-3 crosslinking by anti–TIM-3 pAb (anti–TIM-3 pAb; Fig. 1B). Given that the observed size of the TIM-3 band (42 kDa) was higher than the predicted molecular mass (30.934 kDa), the 42-kDa band possibly represented a glycosylated form of TIM-3. Indeed, this possibility was confirmed by the observations that enzymatic deglycosylation reduced the mass of TIM-3 from 42 kDa to ∼31 kDa (Fig. 1C) and that the 42-kDa TIM-3 protein band was effectively stained with glycoprotein detection reagent (Supplemental Fig. 1). Overall, these results suggested that TIM-3 phosphorylation could be induced by Ab-mediated crosslinking. Accordingly, we used these Abs to crosslink TIM-3 in subsequent experiments.

Next, we investigated whether crosslinking of TIM-3 influenced DC activation. For this, we pretreated DCs with anti–TIM-3 Abs and examined the capacity of these DCs to secrete proinflammatory cytokines after stimulation with LPS. BALB/c BMDCs treated only with LPS secreted more IL-12p70 than did untreated BMDCs (Fig. 2A). However, LPS-stimulated IL-12p70 secretion was progressively inhibited with increasing time of BMDC pretreatment with anti–TIM-3 mAb, but not isotype control Ab (Fig. 2A). LPS-induced IL-12p70 secretion was also inhibited in BALB/c sDCs and B6 BMDCs after pretreatment with anti–TIM-3 mAb (Fig. 2B, 2C). The inhibitory effect of anti–TIM-3 mAb pretreatment on LPS-stimulated IL-12p70 secretion, however, was blocked after silencing of TIM-3 expression by siRNA (Fig. 2D, 2E). This finding ruled out the possibility that the inhibition of LPS-stimulated IL-12p70 secretion occurred owing to off-target binding of anti–TIM-3 mAb. Like anti–TIM-3 mAb, anti–TIM-3 pAb also reduced LPS-induced IL-12p70 secretion from BALB/c BMDCs (Supplemental Fig. 2A). Compared with IL-12p70, LPS-stimulated TNF-α secretion was marginally affected by BMDC pretreatment with anti–TIM-3 mAb (Fig. 2F). We further determined the effect of TIM-3 crosslinking on LPS-stimulated upregulation of costimulatory molecule expression on DCs. TIM-3 crosslinking by anti–TIM-3 mAb inhibited LPS-stimulated upregulation of CD40, CD80, and CD86 expression on BMDCs (Fig. 2G). Collectively, these results suggest that TIM-3 triggering suppresses activation and maturation of DCs.

Because DC activation and maturation are largely regulated by NF-κB (13, 14), we analyzed the effect of TIM-3 crosslinking on activation of the NF-κB signaling in DCs. We pretreated BALB/c BMDCs with isotype control Ab or anti–TIM-3 mAb for varying times and stimulated with LPS. We assessed NF-κB DNA binding activity and degradation of IκBα via EMSA and immunoblot analysis, respectively. We found that LPS-stimulated NF-κB DNA binding and IκBα degradation were inhibited as early as 1 h, with maximum inhibition occurring 6 h after pretreatment of BMDCs with anti–TIM-3 mAb, but not isotype control Ab (Fig. 3A). LPS-induced NF-κB DNA binding and IκBα degradation were similarly impaired in BALB/c sDCs and B6 BMDCs pretreated with anti–TIM-3 mAb for 6 h (Fig. 3B, 3C). Like anti–TIM-3 mAb, anti–TIM-3 pAb also inhibited LPS-induced NF-κB DNA binding and IκBα degradation (Supplemental Fig. 2B). Furthermore, pretreatment with anti–TIM-3 mAb inhibited LPS-induced IκBα phosphorylation and activation of the upstream IKK (Fig. 3D). The latter was determined by measuring phosphorylation of IKKα and IKKβ (Fig. 3D). Similar to LPS, TNF-α–stimulated NF-κB DNA binding and IκBα degradation were substantially inhibited owing to pretreatment of BALB/c BMDCs with anti–TIM-3 mAb (Fig. 3E). However, neither DC viability nor the expression levels of TLR4 and TNF receptors (TNFR1 and TNFR2) on DCs were affected by anti–TIM-3 mAb treatment (Fig. 3F, 3G). These data negate the possibility that the hyporesponsiveness of DCs to LPS or TNF-α after anti–TIM-3 Ab treatment was due to downregulation of the cognate receptors or loss of DC viability. Taken together, the above results indicate that TIM-3 crosslinking on DCs results in blockade of multiple pathways that converge to activate NF-κB. Unlike NF-κB signaling, LPS-induced activation of the MAPK pathway remained unaffected despite anti–TIM-3 mAb pretreatment (Fig. 3H). Thus, TIM-3 triggering inhibits activation of the NF-κB but not other signaling pathways controlling DC activation and maturation.

Next, we tested whether the inhibitory effect of TIM-3 crosslinking on the NF-κB pathway could be withdrawn by the silencing of TIM-3 in DCs. We found that TIM-3 silencing efficiently blocked anti–TIM-3 mAb–induced inhibition of NF-κB DNA binding, IκBα degradation, and IKK phosphorylation in BALB/c BMDCs stimulated with LPS (Fig. 3I). Collectively, these results suggest that TIM-3 negatively regulates NF-κB signaling in DCs.

To determine the mechanism by which TIM-3 suppressed NF-κB activation in DCs, we initially directed our effort to identifying the signaling events triggered by TIM-3 in DCs. A study has reported that two Src kinase family members, Lck and Fyn, mediate TIM-3 phosphorylation in T cells (10). However, the role of c-Src (another member of the Src family) in TIM-3 signaling remains unrecognized. In addition, an inhibitory role for c-Src in NF-κB activation in DCs has previously been demonstrated (12). Accordingly, we tested the involvement of c-Src in TIM-3–induced signaling in DCs. Kinetic analyses showed that c-Src activation, as measured by Tyr⁴¹⁶ phosphorylation of c-Src (12), was induced within 10 min and continued up to 6 h after treatment of BALB/c BMDCs with anti–TIM-3 mAb, but not isotype control Ab (Fig. 4A). TIM-3 crosslinking by anti–TIM-3 mAb induced c-Src phosphorylation in BALB/c sDCs also (Fig. 4B). Similar to anti–TIM-3 mAb, anti–TIM-3 pAb also stimulated phosphorylation of c-Src in BALB/c BMDCs (Supplemental Fig. 2C). However, silencing of TIM-3 blocked this TIM-3–driven c-Src phosphorylation (Fig. 4C). These data strongly suggest that TIM-3 triggering promotes activation of c-Src in DCs.

Next, we sought to identify the signaling mediator, which might link TIM-3 with c-Src. One potential candidate is Btk, which is known to regulate c-Src activation (11). Therefore, we verified whether crosslinking of TIM-3 activates Btk. We assessed Btk activation by measuring Tyr²²³ phosphorylation of Btk (11). Phosphorylation of Btk was induced in BMDCs within 2.5 min after anti–TIM-3 mAb treatment, reached maximum at 5 min, and dropped back to basal level by 10 min (Fig. 4D). We observed a similar induction of Btk phosphorylation after treating BALB/c sDCs and BMDCs with anti–TIM-3 mAb and pAb, respectively (Fig. 4E, Supplemental Fig. 2D). However, TIM-3 silencing blocked this TIM-3–triggered Btk phosphorylation (Fig. 4F). On the contrary, TIM-3 phosphorylation induced by anti–TIM-3 mAb was not affected by the silencing of Btk (Fig. 4G, 4H). We further determined whether Btk or c-Src interacted with TIM-3 and found that Btk and c-Src were immunoprecipitated with TIM-3 in BMDCs treated with anti–TIM-3 mAb, but not isotype control Ab (Fig. 4I). Because activation of Btk occurred earlier than c-Src (Fig. 4A, 4D), we examined whether Btk acted as an upstream activator for c-Src. Whereas silencing of Btk abrogated TIM-3–triggered c-Src phosphorylation (Figs. 4G, 5A), c-Src silencing had no effect on Btk phosphorylation induced by TIM-3 (Fig. 5B, 5C). Furthermore, we did not detect any c-Src–TIM-3 interaction in Btk-silenced BMDCs despite TIM-3 crosslinking (Fig. 5D). These findings placed Btk upstream of c-Src in the TIM-3 signaling pathway. Together the above data suggest that TIM-3 triggering activates Btk and c-Src in DCs.

To test a functional role for Btk and c-Src in mediating the TIM-3–triggered responses in DCs, we initially determined whether Btk and c-Src contributed to TIM-3–induced inhibition of NF-κB signaling. Pretreatment with anti–TIM-3 mAb, but not isotype control Ab, inhibited LPS-induced activation of NF-κB signaling in BALB/c BMDCs transfected with control siRNA (Fig. 6A, 6B). However, silencing of Btk or c-Src blocked this inhibitory effect of anti–TIM-3 mAb (Fig. 6A, 6B). Btk or c-Src silencing similarly attenuated TIM-3–induced inhibition of NF-κB activation in BALB/c sDCs and B6 BMDCs (Supplemental Fig. 3). We further determined whether Btk and c-Src were involved in TIM-3–induced DC inhibition. Ligation of TIM-3 with anti–TIM-3 mAb markedly reduced LPS-stimulated IL-12p70 secretion and upregulation of costimulatory molecule expression in control BMDCs (i.e., BMDCs left untransfected or transfected with control siRNA; Fig. 6C–F). However, these TIM-3–induced inhibitions of DCs were vastly reduced after Btk or c-Src silencing (Fig. 6C–F). Thus, TIM-3 induces DC suppression through Btk and c-Src.

Previous work by our group demonstrated that Btk–c-Src signaling mediates hepatocyte growth factor (HGF)–induced inhibition of DCs by promoting autocrine IL-10 secretion (11). Therefore, we wondered whether TIM-3 crosslinking induced IL-10 secretion from DCs. Strikingly, we did not detect notable amounts of IL-10 after treating BALB/c BMDCs with anti–TIM-3 mAb (Fig. 7A). This finding ruled out the involvement of IL-10 in TIM-3–induced DC suppression. We then examined whether the supernatants of BMDCs cultured in the presence of anti–TIM-3 Ab for 6 h were able to inhibit LPS-induced activation of fresh BMDCs. We found that preincubation of fresh BMDCs with culture supernatants from anti–TIM-3 Ab–treated BMDCs, but not isotype control Ab–treated BMDCs, inhibited LPS-stimulated activation of NF-κB signaling, IL-12p70 secretion, and upregulation of costimulatory molecule expression (Fig. 7B–D). However, supernatants of Btk- or c-Src–silenced BMDCs treated with anti–TIM-3 mAb did not show such inhibitory effects (Fig. 7E–G). These results indicate that TIM-3–driven Btk–c-Src signaling promotes secretion of some inhibitory factor (or factors) (other than IL-10) that induces (or induce) DC suppression.

Earlier, TIM-3 had been shown to promote a proinflammatory response of DCs upon treatment with galectin-9, which used to be considered a TIM-3 ligand (3, 5). However, a recent report has confirmed that galectin-9 is not a ligand for TIM-3 (17). In fact, galectin-9 binds to other cell surface molecules, such as CD44 and CD40 (18, 19). Notably, CD44 and CD40 are also expressed on DCs (20, 21). These reports reinforced the need to further investigate how DC activation and maturation are regulated specifically by TIM-3 signaling. In the current study, we therefore triggered TIM-3 signaling via Ab-mediated crosslinking and have shown that TIM-3 inhibited activation and maturation of DCs, regardless of whether mAb or pAb was used to crosslink this receptor. Our findings are in agreement with a report (7) demonstrating that TIM-3 deficiency results in an increased production of proinflammatory cytokines and type I IFNs by BMDCs after stimulation with nucleic acids (agonists of TLRs). However, in this report, Chiba et al. (7) have not examined the direct effect of TIM-3–triggered signaling on DC activation and maturation. Furthermore, the TIM-3 signaling events that contribute to immunoregulation of DCs remain undefined. Our study, in addition to demonstrating a role for TIM-3 in inducing DC suppression, has also identified the proximal components of TIM-3 signaling that mediate this effect.

While exploring the signaling pathway transduced by TIM-3 in DCs, we found that Btk was activated within 2.5 min after TIM-3 crosslinking. Because of its early activation, Btk may therefore be considered a proximal mediator of TIM-3 signaling. Unlike Itk (IL-inducible T cell kinase; another Tec family member), which mediates phosphorylation of TIM-3 (22), Btk activation was dispensable for TIM-3 phosphorylation induced by anti–TIM-3 Ab. In contrast, TIM-3 crosslinking was necessary for activation of Btk in DCs. In addition, Btk did not interact with TIM-3 in untreated DCs but bound to TIM-3 when DCs were treated with anti–TIM-3 Ab. These data suggest that Btk interacts with TIM-3 and that this interaction between TIM-3 and Btk is triggered by TIM-3 crosslinking. Whether Btk interacts with TIM-3 directly or requires any other adaptor protein (or proteins) for this interaction needs further investigation. Nevertheless, our data have provided the evidence that Btk participates in the proximal signaling events induced by TIM-3 in DCs.

Currently, there are conflicting reports regarding the role of Btk in the immunoregulation of DCs. Using DCs from patients with Btk-deficient X-linked agammaglobulinemia, some studies have suggested a redundant role for Btk in LPS-induced DC activation and maturation (23, 24). In contrast, our group and others have previously reported on the inhibitory role for Btk in DC activation and maturation induced by LPS (11, 25, 26). In this article we have provided additional evidence for the inhibitory role of Btk by demonstrating that Btk was required for TIM-3–mediated inhibition of DCs. Btk mediated this suppressive effect of TIM-3 by blocking activation of the NF-κB pathway. Indeed, crosslinking of TIM-3 led to inhibition of NF-κB activation in both BMDCs and sDCs. Our finding differs from an earlier report demonstrating that TIM-3 induces NF-κB activation in a DC cell line (5). One possible explanation for this discrepancy is that the type of DCs used in the above-mentioned report (DC cell line; Ref. 5) is quite different from those used in the current study (primary DCs). Furthermore, our observations are in agreement with recent reports showing the impairment of NF-κB activation in HEK293 cells (human embryonic kidney cells) and RAW264.7 cells (mouse macrophage cells) upon TIM-3 overexpression and TIM-3 stimulation, respectively (6, 7). So far, Btk is reported to function as a positive regulator of NF-κB signaling in B cells stimulated with BAFF (B cell–activating factor) and in a human monocyte cell line stimulated with LPS (27, 28). However, consistent with our previous report (11), we found in this study that TIM-3–triggered Btk activation inhibited the NF-κB pathway in DCs. Similarly, another study on CD303-mediated inhibition of plasmacytoid DCs has indicated the possibility that Btk may negatively regulate NF-κB activation (29). It is reported that Btk interacts with many signaling mediators (30). The downstream effects of Btk may therefore vary depending on the nature of the signaling molecules interacting with Btk and the context of other signaling events triggered by a given receptor.

Our findings further demonstrated that TIM-3–induced Btk blocked the NF-κB pathway and subsequent DC activation and maturation through c-Src. Upon crosslinking, TIM-3 recruited and activated c-Src in a Btk-dependent manner. Our finding that Btk functioned upstream of c-Src in TIM-3–induced signaling is similar to what we have shown previously in the case of HGF-stimulated signaling pathway (11). How c-Src blocked the NF-κB pathway is currently under investigation. Our preliminary results have revealed that TIM-3, at 10 min after its crosslinking, interacted with the PI3K complexes p85α-p110α and p85α-p110δ, and induced Akt kinase activity from 15 min onward after stimulation (Supplemental Fig. 4). The inhibitory roles for these PI3K complexes and Akt in NF-κB signaling and their functions as downstream effectors of c-Src have previously been reported by our group (12). Furthermore, TIM-3–mediated activation of the PI3K–Akt pathway has been shown to negatively regulate NF-κB activation in RAW264.7 cells (6). This report, coupled with our own data, indicates the possibility that TIM-3–activated c-Src prevented NF-κB signaling in DCs via the PI3K–Akt pathway.

In an effort to determine whether TIM-3–driven Btk–c-Src signaling inhibited the NF-κB pathway through autocrine action of any inhibitory factor (or factors) produced by DCs, we analyzed the ability of TIM-3 to stimulate IL-10 secretion from DCs. Surprisingly, we could not detect IL-10 production by DCs despite TIM-3 crosslinking. These data suggest that IL-10 is not involved in TIM-3–induced DC suppression. However, our findings using supernatants from anti–TIM-3 Ab–treated DCs indicate that TIM-3–triggered Btk–c-Src signaling induces secretion of some other inhibitory factor (or factors) from DCs, which in turn impedes (or impede) DC activation and maturation events by inhibiting the NF-κB pathway. At present, we are uncertain about the inhibitory factor (or factors) that mediated TIM-3–induced DC suppression, and we are currently investigating this aspect. In addition, it is not clear to us why Btk–c-Src signaling failed to elicit IL-10 secretion from DCs after TIM-3 crosslinking, although the same signaling pathway mediates HGF-induced IL-10 production by DCs (11). It is possible that the ability of Btk–c-Src signaling to stimulate IL-10 production may differ in the context of other signaling events induced by a given receptor.

Overall, we have demonstrated that TIM-3–triggered signaling inhibits activation and maturation of DCs. We have also identified Btk as a proximal component of TIM-3 signaling, which, by stimulating the secretion of inhibitory factor (or factors) from DCs through c-Src, blocks the NF-κB pathway and consequently induces DC suppression. Our findings therefore have documented a new role for Btk and c-Src in TIM-3–mediated immunoregulation of DCs. Notably, TIM-3 has been shown to interfere with the activation of tumor-associated DCs, thereby attenuating the antitumor efficacy of DNA vaccines and chemotherapy (7). Thus, blockade of the inhibitory effects of TIM-3–transduced signaling on DC activation by targeting Btk or c-Src may provide an approach to improve the efficacy of antitumor therapies. Furthermore, our study will contribute to better understanding of the role of Btk and c-Src in the regulation of innate immune responses by TIM-3.

We thank the Institute of Microbial Technology animal house facility for providing mice required for experiments.

The authors have no financial conflicts of interest.