Mast cell activation is enhanced by Tim1:Tim4 interaction but not by Tim-1 antibodies

Antibodies and reagents

Monoclonal anti-dinitrophenyl (DNP), IgE isotype, clone SPE-7 (Cat No. D8406), DNP₃₂-HSA (Cat No. A6661), anti-FLAG M2 antibody (Cat No. F1804), cyclosporine A (Cat No. C1832), and 4-Nitrophenyl N-acetyl-β-D-glucosaminide (pNAG) (Cat No. N9376) were purchased from Sigma-Aldrich (St. Louis, MO). DNP₅-BSA (Cat No. D5050) was from Biosearch Technologies (Petaluma, CA). Monoclonal antibodies to murine Tim-1 (3B3, RMT1-10, 5G5, 5F12) and purified Tim4-Fc (the latter consists of the IgV and mucin domains of murine Tim-4 fused to the constant region of hIgG1) were obtained from Vijay Kuchroo (Harvard Medical School). Human IgG as isotype control for Tim4-Fc was purchased from Jackson ImmunoResearch Laboratories (Cat No. 009-000-003, West Grove, PA). Monoclonal antibodies to BALB/c Tim-1 (1H9, 3A2, 4A2, 4B2, 4G8, 5D1) were originally developed at Biogen and were obtained under an MTA with CoStim Pharmaceuticals (Cambridge, MA). Purified rat IgG2a as isotype control for Tim-1 antibodies was purchased from eBioscience (Cat No. 14-4321-85, San Diego, CA). Phospho-specific antibody to ERK (T202/Y204) was from BD Biosciences (20 μl per test), (Cat No. 612566, San Jose, CA). Phospho-specific antibodies to Syk (Y519/520) (1:400, Cat No. 2710), and S6 (S235/236), (1:150, Cat No. 4851) were obtained from Cell Signaling Technology (Danvers, MA). Anti-mouse Fc block (2.4G2) was purchased from BD Biosciences (Cat No. 553141, San Jose, CA). IL-6 luciferase reporter constructs were obtained from Sarah Gaffen (University of Pittsburgh), originally from Oliver Eickelberg (Helmholtz Zentrum Munchen).

Mice

All studies were performed in accordance with University of Pittsburgh Institutional Animal Care and Use Committee procedures. Specifically, mice were housed four males or five females per polycarbonate cage in a 12-hour light/dark cycle. Food and water were provided ad libitum. Cages and bedding were changed every seven days. Mutant mice lacking Tim-1 mucin domain (Tim-1^Δmucin) were obtained from David Rothstein (University of Pittsburgh), and were originally from Vijay Kuchroo (Harvard Medical School). Age-and sex-matched wild-type C57BL/6 were purchased from the Jackson Laboratory (Bar Harbor, ME) as control.

BMMC and mast cell line culture

Bone marrow cells from C57BL/6 Tim-1 wild-type (WT) and mutant (Tim-1^Δmucin) were generated as described previously¹⁸. MC/9 mast cells were cultured in DMEM supplemented with 10% BGS, 2-ME, Pen/Strep with Glutamine, and 10% IL3-conditioned media.

BMMC stimulation, cytokine secretion, and phospho-flow cytometry

BMMCs (1 × 10⁵ or 1 × 10⁶ cells for cytokine measurement and 3 × 10⁵ cells for phospho-flow) were sensitized overnight with 1 μg/ml IgE without IL-3 conditioned media. Cells were stimulated with DNP₃₂-HSA or DNP₅-BSA. Six or twenty-four hours post stimulation, supernatants were collected and assayed for murine IL-6 and TNF-α by ELISA (BioLegend). For phospho-flow staining, stimulated cells were fixed in 1.5% paraformaldehyde for ten minutes at room temperature and permeablized with ice cold methanol for thirty minutes. Incubation with phospho-specific antibodies were performed per manufacturer’s instructions. Flow acquisition was performed on Fortessa or LSRII (BD Biosciences), and data were analyzed using FlowJo software version 8.7 (Tree Star).

Measurement of beta-hexosaminidase release

BMMCs (2.5 × 10⁵ cells) were sensitized and subsequently stimulated for thirty minutes in Tyrode’s buffer (135mM NaCl, 5mM KCl, 5.6mM glucose, 1.8mM CaCl₂, 1mM MgCl₂, 20mM HEPES, and 0.5mg/ml BSA). Supernatants (stimulated release) were collected and cells were lysed (content) with 0.5% Triton-X100 in PBS for fifteen minutes on ice. Content and stimulated release fractions were incubated with 1mM pNAG substrate for 1 hour at 37°C. Carbonate buffer (0.1M, pH 9.0) was added to stop reaction and absorbance was obtained at 405nm on a plate reader (BioTek ELx808). Percentage of beta-hexosaminidase release was calculated as (release^stimulated/content^total) × 100.

Mast cell degranulation assay by Annexin V-based flow cytometry

Degranulation was measured as described previously^18,38. Briefly, BMMCs were loaded with 0.1μM of Lysotracker Deep Red (Invitrogen) for thirty minutes at 37°C prior to sensitization with 0.5 μg/ml IgE for 1 hour. Ninety minutes after antigen cross-linking by DNP₃₂-HSA, cells were collected and stained with Annexin V (BioLegend). %degranulation was determined as percentage of BMMCs that is AnnexinV⁺Lysotracker^lo. Flow acquisition was performed on Fortessa or LSRII (BD Biosciences), and data were analyzed using FlowJo software version 8.7 (Tree Star).

Transcriptional reporter luciferase assays

MC/9 mast cells (15 × 10⁶) were transfected with 15 μg of IL6-luc, NF-κB-luc, or NFAT/AP1-luc with 5 μg of pCDEF3 (empty vector), FLAG-tagged Tim-1 full length (FL), Tim-1 cytoplasmic deletion (Δcyto), or Tim-1 tyrosine mutant (Y276F). Electroporation was performed at 290V, 950 μF using a Gene Pulser II apparatus (Bio-Rad). Cells were collected twenty-four hours post-transfection and stimulated with 0.5 μg/ml IgE and 30 ng/ml or 100 ng/ml of DNP₃₂-HSA as antigen for six hours. Luciferase assays were performed as previously described³⁹.

Statistical analysis

All statistical analyses was performed using Prism version 6.0 (GraphPad Software). Paired, unpaired, two-tailed Student’s t-test, one-way and two-way ANOVA with multiple comparison were used for data analysis and calculation of p values, as appropriate.

Tim1-Tim4 interaction enhances mast cell cytokine production without affecting degranulation

Mast cells constitutively express surface Tim-1 and Tim-3, but not Tim-2 or Tim-4¹⁶. Tim-1 cross-linking by Tim4-Fc was shown to enhance cytokine production in a dose-dependent manner without affecting degranulation¹⁶. We first confirmed that Tim4-Fc could bind to our cultures of bone marrow derived mast cells (BMMC) from C57BL/6 mice (Figure 1A). We observed similar effects of Tim4-mediated IL-6 production in IgE/Ag-stimulated BMMCs (Figure 1B). In addition, we quantified the degranulation response by measuring beta-hexosaminidase release as well as with a flow cytometry-based assay that tracks both Annexin V binding to exposed PS and loss of Lysotracker staining, due to granule release (AnnexinV+Lysotracker^lo). The latter method has the advantage of a high signal-to-noise ratio, and as such has been a robust assay for evaluating IgE/Ag-stimulated mast cell degranulation¹⁸. Thus, we showed that varying the concentration of Tim-4, along with suboptimal antigen concentration to observe potential co-stimulation, did not have an effect on the immediate degranulation response (Figure 1C–D). These results suggest that Tim-1 ligation by Tim-4 can exert differential effects on the immediate degranulation response, vs. late-phase cytokine production.

Antigen valency and concentration have been shown to control the outcomes of FcεRI engagement in not just quantitative, but also qualitative, fashions¹⁹. Specifically, low antigen concentration or valency will activate only positive FcεRI signaling pathways, while high antigen valency or concentration will preferentially engage negative signaling components downstream of FcεRI. This is due to activity of the Src family kinase Lyn as a positive and/or negative regulator of FcεRI signaling at low or high antigen valency, respectively¹⁹. We stimulated BMMCs with low (DNP₅-BSA) or high (DNP₃₂-HSA) potency antigens in the presence of Tim4-Fc or isotype control and assessed whether Tim-1 ligation could contribute to receptor signaling intensity. Thus, at high antigen valency, which also activates negative feedback of antigen receptor signaling, Tim-4 was able to maintain high IL-6 production but not at low antigen valency that induces robust antigen receptor signaling (Figure 1E–F). Specifically, increasing the amount of Tim-4 further promoted cytokine secretion under high valency antigen stimulation, i.e. under conditions where the negative signaling loop is triggered. To exclude the possibility that Fcγ receptor binding may interfere with Tim1-Tim4 interaction, we compared IL-6 release by BMMCs, with or without addition of Fc blocking antibody, and found no differences, at least at the concentration of Tim4-Fc used throughout this study (5 μg/ml). These findings indicate that Tim-1 ligation can modulate the intensity of the antigen-induced positive FcεRI signaling pathways and may be able to bypass the negative feedback signaling loop controlled by Lyn.

Tim-1 antibodies do not have significant effects on IgE/Ag-mediated mast cell activation

Several antibodies against Tim-1 have been tested and showed no effects on either IgE/Ag-induced mast cell degranulation or cytokine production¹⁶. Of significance are the mAbs 3B3 and RMT1-10, which have been termed “agonistic” and “antagonistic” antibodies, respectively, due to their ability to enhance or inhibit effector T cell activation²⁰. Similarly, we did not observe any dose-dependent effects of 3B3, RMT1-10 or the IgV domain-binding mAb 5F12 on levels of IL-6 and TNF-α secreted by IgE-sensitized and Ag-stimulated BMMCs (Figure 2A–D). We examined several other antibodies, generated against the BALB/c allele of Tim-1, which were shown to either exacerbate or ameliorate Th2-dependent OVA-induced lung inflammation in mice⁷. Specifically, mAbs 1H8 and 3A2 bind to distinct epitopes near an N-linked glycosylation site in the stalk region and have agonistic and antagonistic activity, respectively. The mAb 4A2 binds to the IgV domain of Tim-1 and reduces lung inflammation and pathology⁷. However, aside from 4G8, we did not detect significant binding of these antibodies to BALB/c Tim-1. While 4G8 recognized surface Tim-1, it did not alter Ag-mediated cytokine production in mast cells (Figure 2E). Even though these antibodies were raised against the BALB/c allele of Tim-1, we also observed binding of 4G8 to C57BL/6 Tim-1 (Figure 2F). Again, there was no detectable increase in IL-6 production when these antibodies were used in co-stimulation with FcεRI crosslinking by IgE/Ag, on BMMC derived from C57BL/6 mice (Figure 2F).

Figure 2. Tim-1 antibodies did not significantly alter IgE/Ag-mediated mast cell cytokine production compared to ligation by Tim4-Fc.

BL/6 or BALB/c BMMCs (1 × 10⁶) as indicated were sensitized with IgE overnight and stimulated with DNP₃₂-HSA in the presence of isotype control or monoclonal antibodies against Tim-1 3B3 (A–B), RMT1-10 (C–D), 3A2, 4A2, 4B2, 4G8, 1H9 (E) for six hours. The indicated antibodies were incubated with BALB/c BMMCs for 30 minutes on ice followed by anti-rat IgG-Alexa-647 secondary antibody, prior to flow cytometry analysis (E). Culture supernatants were analyzed for IL-6 and TNF-α by ELISA. BL/6 BMMCs were incubated with 4G8, 4A2 and 5D1 antibodies followed by anti-rat IgG-Alexa 647 secondary antibody prior to flow cytometry analysis (F). BL/6 BMMCs sensitized with IgE overnight and stimulated with DNP₃₂-HSA together with either isotype control or the indicated antibodies for six hours prior to IL-6 measurement by ELISA (F). Results are representative of three (A–D) and two (E–F) independent experiments performed in duplicates.

Tim-1 cytoplasmic tyrosine is required for enhancement of transcriptional activation

We previously demonstrated that transient expression of Tim-1 co-stimulated TCR/CD28-mediated transcriptional activation of IL-4 and IFN-γ production and NF-AT/AP1-dependent transcription. This co-stimulatory activity was dependent on tyrosine 276 in the Tim-1 cytoplasmic tail⁶. Similarly, ectopic expression of an N-terminal FLAG-tagged Tim-1 on MC/9 mast cells was able to enhance IgE/Ag-stimulated NF-κB transcriptional activation (Figure 3A). This enhancement was abrogated when the cytoplasmic tail of Tim-1 was deleted or when a tyrosine-phenylalanine mutant (Y276F) was used (Figure 3A). Polymorphisms in Tim-1 have been associated with differential responses to OVA-induced allergic asthma². We found that both isoforms of Tim-1 (BL/6 and BALB/c) could significantly enhance activation of an NF-κB transcriptional reporter to a comparable extent (Figure 3B). Consistent with findings in T cells and effects on mast cell cytokine production, transient expression of Tim-1 also up-regulated NF-AT/AP1 and IL-6 promoter activation, in a tyrosine phosphorylation-dependent manner (Figure 3C–D). Furthermore, using IL-6 promoter deletion constructs, we showed that Tim-1 mediated enhancement of transcriptional activation and subsequent production of IL-6 through activation of NF-κB and AP-1 transcription factors (Figure 3E). Unlike Tim-4 crosslinking of Tim-1, addition of anti-FLAG antibody could not further promote reporter activity, suggesting either that Tim-1 may have other ligands on MC/9 mast cells or that Tim-1 can homo-dimerize after ectopic expression, via its heavily glycosylated mucin domain, leading to downstream signaling.

Figure 3. Tim1-mediated enhancement of transcriptional activation is dependent on a tyrosine on its cytoplasmic tail.

MC/9 mouse mast cells were transfected with empty vector (pCDEF3), BL/6 or BALB/c Tim-1 (full length (FL)), BL/6 Tim-1 lacking the cytoplasmic region (Δcyto) or full length Tim-1 (BL/6) with tyrosine to phenylalanine mutated at tyrosine 276 (Y276). Transfected cells were stimulated with 0.5 μg/ml IgE plus either 30 ng/ml or 100 ng/ml DNP₃₂-HSA, with or without addition of anti-FLAG antibody for six hours. MC/9 mast cells were co-transfected NF-κB (A–B), NFAT/AP1 (C), and IL-6 (D) luciferase reporters. MC/9 mast cells were transfected with full length BL/6 Tim-1 along with indicated IL-6 luciferase reporters and stimulated as described (E). Results are representative of three independent experiments performed in triplicate. *p<0.05, **p<0.005, ****p<0.00005.

Tim-1 mucin domain is not required for Tim4-mediated enhancement of mast cell cytokine production

Our findings thus far suggest that the Tim1-Tim4 interaction augments FcεRI signaling itself, rather than acting through a parallel pathway, since Tim-4 treatment alone does not induce any detectable cytokine production or degranulation. While the Tim1-Tim4 interaction has been attributed primarily to the IgV domain, Tim-4 has also been proposed to bind to the Tim-1 mucin domain¹⁰. Intriguingly, the genetic linkage of Tim-1 to allergies and asthma is associated with polymorphisms in the mucin domain. Using a mutant mouse lacking only the Tim-1 mucin domain (Tim-1^Δmucin)²¹, we determined whether the Tim-1 mucin domain is necessary to relay the co-stimulatory effects of Tim-4 binding. We noted no obvious defects in mast cell development or maturation, when BMMC were generated from Tim-1^Δmucin bone marrow. Thus, WT and Tim-1^Δmucin BMMCs expressed comparable levels of surface Tim-1, based on staining with IgV-binding Tim-1 antibodies (Figure 4A). Tim-1^Δmucin BMMCs exhibited a similar extent of degranulation in response to antigen stimulation, compared to WT BMMCs, which was not altered by Tim-1 ligation (Figure 4B–C). Next, we examined the ability of Tim-1^Δmucin BMMCs to secrete cytokines in response to IgE/Ag. Tim-1^Δmucin mast cells appeared to respond normally to antigen stimulation, and also to Tim4-mediated co-stimulation (Figure 4D). Using different batches of BMMCs over the course of multiple experiments, we were not able to consistently observe any major difference between WT and Tim-1^Δmucin BMMCs. We did observe some variation in the amount of IL-6 produced by different batches of BMMCs, but this appeared to be largely due to the relative maturation status of the cells (Figure 4E–H). These results indicated that the mucin domain of Tim-1 is not required for normal mast cell responses and also that Tim-4-mediated mast cell activation does not appear to involve the mucin domain of Tim-1, but rather its IgV domain.

Figure 4. Tim-1 mucin domain is not required for Tim4-mediated enhancement of cytokine production in IgE/Ag-stimulated Tim-1^Δmucin BMMCs.

Tim-1 surface expression and maturity of WT and Tim-1^Δmucin BMMCs were determined by FcεRI and c-kit staining (A). WT and Tim-1^Δmucin BMMCs were sensitized overnight with IgE and stimulated with 10 ng/ml, 50 ng/ml, 100 ng/ml DNP₃₂-HSA, and 5 μg/ml Tim4-Fc. Mast cell degranulation was measured by Annexin V and Lysotracker staining (B–C). Results are average of three (B) and two (C) independent experiments. Culture supernatants were collected and analyzed for IL-6 by ELISA (D). Using six separate batches of BMMCs. WT and Tim-1^Δmucin BMMCs were stimulated with indicated amount of antigen with either 5 μg/ml isotype control Fc (iso) or Tim4-Fc for six hours (E–H). Results are comparison between WT and Tim-1^Δmucin BMMCs from six independent experiments using six separate batches of BMMCs. *p<0.05.

Tim-1 enhances Ag-activated phosphorylation of ribosomal protein S6

Next, we assessed the potential signaling pathways utilized by Tim1-Tim4 interaction downstream of FcεRI signaling to upregulate mast cell cytokine production. Using a Nur77GFP reporter mouse, we previously showed that unlike the related family member Tim-3, engagement of Tim-1 did not enhance FcεRI signal intensity, thereby showing that Tim-1 cross-linking does not augment antigen receptor-proximal signaling¹⁸. We moved on to explore signaling pathways both proximal and distal to the FcεRI complex, for any effects of Tim-1. Thus, the Zap70-related kinase Syk is an FcεRIγ-associated activator integral to activation of LAT, SLP76, PLC-γ and other adaptor molecules essential for signal transduction downstream of FcεRI²². Phosphorylation of Syk was not further increased by Tim-4 (Figure 5A). Similarly, we observed robust phosphorylation of ERK and Akt upon IgE/Ag-induced activation that was not affected by Tim-1 engagement (Figure 5B and data not shown). These results demonstrate that MEK/ERK and PI3K/Akt pathways are not involved in enhancement of Ag-mediated mast cell function by Tim-1, even though we observed that phosphorylation of Tim-1 cytoplasmic tail by TCR activation led to recruitment of p85 subunit of PI3K⁸. Nevertheless, we did detect a significant increase in phosphorylation of ribosomal protein S6, an important target of the PI3K/mTOR pathway regulating cell growth, survival, metabolism, and protein synthesis in mast cells²³. Mast cells exhibited robust phosphorylation of S6 (~80% of BMMCs) one hour post-Ag stimulation, which did not increase further with Tim-4 addition. However, Tim-4 treatment was able to maintain a significant percentage of pS6-positive BMMCs for as long as four hours, even as the Ag-triggered signal returned to basal levels (Figure 5C). Overall, these results support a positive role of Tim-1 activation by Tim-4 to sustain mTOR-dependent mast cell metabolism and protein synthesis, leading to enhanced cytokine production.

Figure 5. Tim-1 ligation by Tim4-Fc enhances mast cell cytokine production by sustaining ribosomal protein S6 phosphorylation upon IgE/Ag activation.

BL/6 BMMCs were sensitized overnight with IgE and stimulated with DNP₃₂-HSA in the presence of isotype control or Tim4-Fc for the indicated time. Syk phosphorylation (Y519/520) (A), pErk (T202/Y204) (B), and pS6 (S235/236) (C) were analyzed by phospho-flow cytometry. Results are average of two independent experiments (A–B) or three independent experiments (C). *p<0.05, ***p<0.0005.