In humans and mice, offspring of allergic mothers are predisposed to development of allergy. In mice, allergic mothers have elevated β-glucosylceramides (βGlcCers) that are transported to the fetus via the placenta and to offspring via milk. The elevated βGlcCers increase the number of fetal liver CD11c+CD11b+ dendritic cells (DCs) and offspring allergen-induced lung eosinophilia. These effects are modifiable by maternal dietary supplementation with the plant-derived lipids α-tocopherol and γ-tocopherol. It is not known whether βGlcCers and tocopherols directly regulate development of DCs. In this study, we demonstrated that βGlcCers increased development of GM-CSF–stimulated mouse bone marrow–derived DCs (BMDCs) in vitro without altering expression of costimulatory molecules. This increase in BMDC numbers was blocked by α-tocopherol and potentiated by γ-tocopherol. Furthermore, βGlcCers increased protein kinase Cα (PKCα) and PKCδ activation in BMDCs that was blocked by α-tocopherol. In contrast, γ-tocopherol increased BMDC PKCα and PKCδ activation and enhanced the βGlcCer-induced increase in PKCδ activation in a DC subset. Ag processing per DC was minimally enhanced in βGlcCer-treated BMDCs and not altered ex vivo in lung DCs from pups of allergic mothers. Pups of allergic mothers had an increased proportion of CD11b+CD11c+ subsets of DCs, contributing to enhanced stimulation of T cell proliferation ex vivo. Thus, βGlcCer, which is both necessary and sufficient for development of allergic predisposition in offspring of allergic mothers, directly increased development and PKC activation in BMDCs. Furthermore, this was modifiable by dietary tocopherols. This may inform design of future studies for the prevention or intervention in asthma and allergic disease.
The prevalence of allergic diseases and asthma has dramatically increased from 1950 to the present (1–3). In animals and humans, the offspring of allergic mothers have increased responsiveness to allergen (4–13). Reports for child asthma most often associate with maternal rather than paternal asthma (4), indicating that in addition to a genetic component, there may be contributions from the in utero environment (14, 15). In humans (4) and mice (5–12), the allergens that stimulate allergic responses in the mother are not necessarily the allergens that stimulate allergic responses in the offspring, rather what is observed in the offspring is a more generalized hyperresponsiveness to multiple allergens (5, 13, 16). This heightened response to allergen is mediated by alterations in dendritic cells (DCs) of offspring of allergic mothers (4, 12, 17). Spleen DCs of offspring of allergic mothers have enhanced allergen presentation function (17) and are sufficient for the allergen responsiveness in mice. Specifically, transfer of spleen DCs (but not macrophages or CD4+ T cells) from nonsensitized neonates of allergic mothers to recipient neonates from nonallergic mothers transfers the enhanced allergic responsiveness to the recipient neonates (4, 17). Also, in adult mice, transfer of CD11c+CD11b+ DCs, but not CD11b−CD103+ DCs, from allergic mice to nonallergic mice is sufficient to induce airway allergy (18). We have shown that the fetal livers and lungs of neonates of allergic mothers also have increased numbers of distinct subsets of CD11b+CD11c+ DCs without altering CD103+ DCs (19, 20). Allergen, IgE, IL-4, and IL-13 of the mother do not pass to the fetus from the mother (21–23). Instead, maternal lipids are transported to the fetus and offspring that regulate offspring development of DCs in vivo (19, 20, 24). These lipids include at least dietary isoforms of vitamin E and endogenous maternally synthesized β-glucosylceramide (βGlcCer) (19, 20, 24).
The allergic response of offspring of allergic mothers is modified by maternal supplementation with isoforms of vitamin E, specifically the two most abundant isoforms, α-tocopherol (αT) and γ-tocopherol (γT) (19, 20). Supplementation of allergic female mice with αT inhibited neonatal development of allergic inflammation and Th2 cytokines and chemokines (19), whereas maternal γT supplementation elevated the development of these allergic responses in offspring (20). Mammals consume tocopherols in food and cooking oils and cannot synthesize tocopherols or interconvert the isoforms. Thus, dietary intake influences the tissue levels of each isoform. Interestingly, countries that have reports of high human average plasma γT levels tend to have the highest prevalence rate of asthma (25–31). We have demonstrated that in adult humans (male and female), plasma αT associates with better lung spirometry, and γT associates with lower lung spirometry by age 21 y (32), suggesting that tocopherol isoforms regulate development and lung responses to environmental pollutants, allergens, or infections.
In addition to tocopherol isoforms, the proinflammatory lipid βGlcCer, which is endogenously synthesized by allergic mothers, is transported to the fetus in utero and to the pup in maternal milk. This maternally derived βGlcCer modulates offspring development of DCs and responses to allergen (24). There is specificity for βGlcCer among the glycosphingolipids because in offspring of allergic mothers, there is no increase in ceramides, sphingomyelins, sphingosines, sphingosine 1-phosphate, dihydrosphingosine, and dihydrosphingosine 1-phosphate (24). Although maternal βGlcCer was necessary and sufficient for heightened development of DCs and allergic responses in offspring, it is not known whether βGlcCers directly regulate hematopoiesis of DCs or DC function. It is also not known whether αT and γT modify DC signaling or function or βGlcCer regulation of DCs. We have previously shown that tocopherol isoforms modify protein kinase C (PKC) activity by binding to the diacylglycerol-binding C1A regulatory domain. PKC activity is inhibited by αT and enhanced by γT (33–35). Therefore, we determined in vitro whether βGlcCer modifies DC development, function, and PKC activation with and without tocopherol isoforms. We also examined ex vivo the function of DCs of offspring of allergic mothers with and without dietary consumption of tocopherol isoforms.
In this report, we show that βGlcCer, at ratios found in vivo, augmented GM-CSF–stimulated generation of DCs and enhanced PKC activation in vitro. DCs isolated from offspring of allergic mothers had an increased proportion of monocyte-derived DCs (mDCs) and resident-phenotype DCs (rDCs) and demonstrated enhanced activation of T cells ex vivo. DC Ag processing or presentation per cell was minimally altered ex vivo or in vitro in these studies. The increase in development of DCs and activation of DC PKCs by βGlcCers in vitro were augmented by γT and blocked by αT in vitro. Moreover, exposure of offspring to elevated levels of βGlcCers and tocopherols from allergic mothers in vivo regulated the development of DCs, and this increase in numbers of DCs augmented offspring lung DC activation of CD4+ T cells. These data demonstrate a direct regulatory function for these lipids in PKC signaling and development of DCs.
Materials and Methods
Adult C57BL/6 and OT II transgenic female and male mice were from Jackson Laboratory (Bar Harbor, ME). The studies were approved by the Indiana University Institutional Review Committee for animals.
Bone marrow–derived DC culture
Bone marrow was obtained from tibia and femurs of 5- to 7-wk-old C57BL/6 female mice by isolating the bones, removing the epiphysis, and either flushing the bones or spinning bones in a tabletop centrifuge × 30 s. Bone marrow cells were then resuspended in DC media composed of RPMI-1640, glutamine, HEPES, gentamicin sulfate, 10% FBS, and 5 μM 2-ME, and then passed through a 70-μm filter. Nucleated cells were enumerated on hemocytometer and cultured in six-well dishes at a concentration of 0.8 × 106/ml in each well. The bone marrow cells were treated with the solvent control (DMSO), d-αT (80 µM), or d-γT (2 µM) with and without the βGlcCer mix as described previously (19, 20, 25). The cells were also stimulated with 5 ng/ml rGM-CSF for DC differentiation. Half of the culture media was refreshed with corresponding lipid treatments and fresh GM-CSF on days 3 and 6.
For experiments involving exposure to Ag, on day 7 the cells were either stimulated with 20 µg/ml house dust mite (HDM) extract or nonstimulated. On day 8 the cells were collected, counted, immunolabeled with the markers described later for DCs, and examined by flow cytometry.
For PKC isoform phosphorylation experiments, cells were collected in the presence of the phosphatase inhibitor okadaic acid at a concentration of 500 nM in PBS-EDTA.
Assessing Ag processing by bone marrow–derived DCs
On day 7 of the bone marrow–derived DC (BMDC) culture, wells were treated with 15 µg/ml HDMs (Greer) or nontreated and incubated overnight. The next day, cells were harvested, enumerated, and plated at a concentration of 1 × 106/ml. Cells were incubated with DQ-OVA (Molecular Probes), a self-quenched conjugate of a model allergen OVA, at a concentration of 10 µg/ml for 30 or 90 min at 37°C in 5% CO2. Ag processing by proteolytic cleavage of DQ-OVA into peptides relieves the self-quenching, thereby increasing fluorescence and detection by flow cytometry (36–38). Negative controls included cells without DQ-OVA or cells incubated with DQ-OVA on ice. On DQ-OVA processing, cells were extensively washed with PBS and fixed with 2% paraformaldehyde (PFA) on ice for 15 min. These cells were washed and stained for surface markers for flow cytometry analysis.
Ex vivo lung DC isolation
To generate allergic and nonallergic moms, we sensitized C57BL/6 female mice by i.p. injection OVA grade III (20 μg)/alum (0.4 mg) or saline/alum (0.4 mg) in 200 µl saline on days 0 and 7 (4–6). The mice were challenged with nebulized saline or 3% (w/v) OVA in saline for consecutive days at 8, 12, and 16 wk of age and then mated. Pups born from allergic and nonallergic moms were sensitized suboptimally with a single i.p injection of 5 µg OVA/1 mg alum in 50 µl saline on day 5 and then challenged with 3% OVA daily for 3 d starting postnatal day 10. At 24 h after last OVA challenge, pups were sacrificed, and tissues were collected for analysis of DC function. For isolation of DCs for ex vivo analysis of function, whole lungs were excised and homogenized in lung digest buffer containing collagenase IV (0.875 mg/ml) and DNase I (0.1 mg/ml) for 30 min at 37°C using gentleMacs Octo dissociator (MACS). Postdigestion, lung homogenates were filtered with 70-μM filter and washed with EasySep buffer (StemCell). CD11c+ lung cells were isolated using magnetic CD11c+ positive selection kit from StemCell according to the manufacturer’s protocol with slight modifications. In brief, lung homogenates were incubated with isolation Ab mixture and CD11c PeCy7 (N418) to label the CD11c+ cells for analysis of purity. Isolated cells were resuspended in 1 ml DC media described for the BMDC cultures and counted on a hemocytometer. Cell concentrations were adjusted to 0.5 × 106/ml for further culture. A small pool of samples was used to assess purity of the isolated DCs, which was >78% CD11c+ as determined by immunolabeling and flow cytometry. Equal numbers of CD11c+ cells were plated for DC–T cell coculture and DQ-OVA assay to assess DC-mediated T cell activation and Ag processing capacity, respectively.
Ag processing assay of ex vivo lung DC culture
CD11c+ isolated lung DCs were plated at 0.5 × 106 in 1 ml of DC media per well of a 24-well plate to assess the Ag processing capacity of the cells. In brief, 10 µg/ml DQ-OVA conjugate was added to the well for 15 min at 37°C in 5% CO2. Meanwhile, also included were negative controls of cells without DQ-OVA and cells with DQ-OVA on ice. After 30 min, cells were washed and fixed with 2% PFA before immunolabeling DC subsets for flow cytometry.
Ex vivo lung DC activation of T cells
Splenocyte suspensions were prepared from sex-matched OT II mice, and CD4+ T cells were isolated using a CD4+ T cell negative selection kit (catalog #19852; STEMCELL Technologies). Isolated lymphocytes were enumerated using a hemocytometer, and cell concentrations were adjusted to 1 × 106 cells/ml for CFSE staining. A portion of unstained cells was used as negative controls for gating in flow cytometry analysis. CFSE staining was performed with 5 µM CFSE (Invitrogen) in PBS for 15 min at 37°C in the dark. Poststaining, cells were washed and resuspended in culture media. CD11c+ DCs were plated in a 96-well plate at 5 × 104 cells/well and preincubated with 1 µg/ml OVA peptide 329–343 for 2 h. Then 25 × 104 CFSE-stained CD4+ T cells were added to the DCs. As controls, additional wells included DCs that received no Ag control or no T cells, or wells with T cells without DCs. Additional wells were set up for controls for immunolabeling for analysis of fluorescence minus one control. DC–T cell cocultures were incubated for 72 h. After 3 d of culture, cells were collected and fixed with 2% PFA before staining for surface markers for flow cytometry analysis. Cells from four replicate wells of the 96-well plate were pooled to generate each sample.
Flow cytometry analyses of DCs and T cells
The cells were immunofluorescence labeled for surface markers of lung DC subsets. In brief, Fc receptors were blocked by incubating the cells in purified rat anti-mouse CD16/CD32 Mouse BD Fc Block (553142; BD Pharmingen) for 15 min on ice. For experiments with intracellular labeling of PKC, cells were collected in the PBS-EDTA + 500 nM okadaic acid (phosphatase inhibitor), washed, and then fixed in 2% PFA. The cells were permeabilized overnight with permeabilization buffer (Invitrogen), labeled with PKCδ phospho-Ser645 (Life Technologies), washed, and then labeled with PE-conjugated goat anti-rabbit IgG secondary Ab (Life Technologies) before labeling for DC markers and p-PKCα/β. For immunolabeling DC markers and p-PKCα/β, an Ab stock was prepared by adding the following volume of Abs per 50 µl of flow cytometry permeabilization buffer: 0.5 µl Alex700-conjugated anti-CD45 (30-F11; Invitrogen), 0.05 µl of PE-CF594–conjugated anti-CD11b (M1/70; BD Biosciences), 0.4 µl allophycocyanin/Cy7-conjugated anti-Ly-6C (HK1.4; BioLegend), 0.2 µl PE/Cy7-conjugated anti-CD11c (N418; BioLegend), 0.2 µl BV711-conjugated anti-MHC class II (MHC II) (M5/114.15.2; BioLegend), 0.2 µl BV421-conjugated anti-CD103 (2E7; BioLegend), 0.5 µl allophycocyanin-conjugated anti-CD317 (927; BioLegend), and 0.5 µl FITC-conjugated anti-PKCα/β phospho-Thr638 (Bioss). Cells were labeled for 45 min on ice at 4°C in the dark and then washed in flow cytometry staining buffer and resuspended for analysis by flow cytometry. Ab labeling of costimulatory molecules was with either 0.2 µl FITC-conjugated anti-CD70 (Biorbyt), 0.5 µl PE-conjugated anti-CD80 (BD Biosciences), and 0.4 µl allophycocyanin-conjugated anti-CD86 (BioLegend), or 0.4 µl allophycocyanin-conjugated anti-OX40L (Tonbo Biosciences), and 0.4 µl PE-conjugated anti-GITRL (BioLegend), or 0.6 µl PE-conjugated anti–4-1BBL (BD Biosciences). All of the Ab stocks were 0.2 mg/ml, except for p-PKCδ and p-PKCα/β at 1 mg/ml and IL-12 at 0.03 mg/ml.
The immunofluorescence-labeled cells were examined with an Aurora (Cytek) flow cytometer. In DC analysis of Ag processing, DQ-OVA–positive cells were assessed using the FITC channel. Analysis was performed using FlowJo VX software (Tree Star). Compensation was done using FlowJo compensation wizard based on single-color control staining of compensation beads (eBioscience). Fluorescence minus one staining controls were used as negative controls to identify gates for populations of interest. The following subpopulations of DCs were analyzed: (1) resident conventional DCs, CD45+CD11b+Ly6C−CD11c+MHC IIhigh; (2) alveolar DCs, CD11b+Ly6C−CD11chighMHC IIintermediate; (3) mDCs, CD11b+Ly6C+CD11c+MHC IIhigh; (4) plasmacytoid DCs, CD11b−Ly6C−CD11clowPDCA-1+MHC II−; and (5) resident CD103+ DCs, CD11b−Ly6C−CD11c+CD103+MHC II−. The flow cytometry gating strategy is as we previously described (19, 20).
For flow cytometry analyses of intracellular cytokines, cells were treated with 2 µl GolgiPlug (BD Biosciences) for 5 h before collection. They were then fixed with 2% PFA in PBS, permeabilized with permeabilization buffer (Invitrogen), stained with 5 µl PE-conjugated anti-IL-12p70 (Miltenyi Biotec) per sample along with DC markers, and examined by flow cytometry.
For T cell staining, T cells were immunolabeled with 50 µl of an Ab mixture per sample. In brief, 0.5 µl BV650-conjugated rat anti-mouse CD4 (RM4-5; BD Biosciences) and 0.4 µl PE-Cy7–conjugated rat anti-mouse CD44 (IM7; BD Biosciences) were used to label CD4+ T cells and activated T cells. CD3+CD4+CD44+CFSEdim cells were quantified as activated T cells. The proliferation modeling function of FlowJo software was used to calculate populations of cells by the number of cell divisions.
Data in the figures were analyzed by either a one-way ANOVA followed by Tukey’s multiple comparisons test, Dunnett’s multiple comparisons to a control (SigmaStat; Jandel Scientific, San Ramon, CA), or nonparametric Wilcoxon multiple comparisons test (JMP; SAS Institute, Cary, NC). For data with two groups, a t test with two-sample unequal variance was used. Data in the figures are presented as the means ± SEs, and each group had 8–10 mice. Data include both genders because there were no differences in outcomes by gender (data not shown).
βGlcCer induces a dose-dependent increase in GM-CSF–stimulated development of DCs in vitro
Maternal βGlcCers increase offspring responses to allergens and specifically increase numbers of CD11c+CD11b+ DC subsets, including mDCs, rDCs, and alveolar-like DC (alvDC) phenotypes, but not CD11b−CD103+ DC phenotypes, in offspring of allergic mothers (19, 24). Furthermore, transfer of CD11c+ DCs from neonates of allergic mothers to neonates of nonallergic mothers is sufficient to induce airway allergy (4, 17). However, it is not known whether βGlcCers directly affect development of CD11c+CD11b+ DCs. Therefore, to focus on whether βGlcCers can directly regulate development of DCs with the phenotype of mDCs and rDCs, we isolated and stimulated mouse bone marrow with 5 ng/ml GM-CSF in the presence or absence of βGlcCer isoforms in vitro. We did not assess in vitro βGlcCer regulation of CD11b-CD103+ DC subsets because these DC subsets do not significantly impact initiation of allergic disease phenotype in vivo (4, 17, 18) and are not generated in GM-CSF–stimulated BMDC cultures in vitro (39). βGlcCer isoforms have 14–26 carbon length lipid tails with 0–1 double bonds. Of these, four isoforms that were elevated in vivo in allergic mothers (24) are commercially available, including C16:0 βGlcCer, C18:0 βGlcCer, C18:1 βGlcCer, and C24:1 βGlcCer. A dose curve for these βGlcCers was added to the GM-CSF–stimulated bone marrow at the beginning of the culture and every 3 d during media changes. Each of these βGlcCers elevated GM-CSF–stimulated development of BMDCs of the mDC phenotype and rDC phenotype in a dose-dependent manner (Fig. 1). The optimal dose for the four forms of βGlcCer differed but was on the same scale as the amounts of these βGlcCer isoforms in plasma of allergic mothers, and the ratios were consistent with that present in the fetal liver of offspring of allergic mothers (24). Because all the βGlcCer isoforms increased the numbers of DCs and these isoforms coexist in vivo, it was determined whether a mixture of the optimal doses of these βGlcCer forms increased development of DC subsets. The βGlcCer mixture enhanced GM-CSF–stimulated differentiation of mDC, rDC, and alvDC phenotypes in a dose-dependent manner (Fig. 2A–C). βGlcCer treatment did not alter the geometric mean fluorescence intensity (gMFI) for IL-12 in DCs without HDM stimulation, but βGlcCer did slightly increase the gMFI for IL-12 in HDM-stimulated mDCs and rDCs, but not alvDCs (Fig. 2D–F). HDM stimulation increased the proportion of DCs producing IL-12; this was not altered by βGlcCer (Fig. 2G–I). However, βGlcCers stimulated a 3-fold increase in the total number of DCs in the GM-CSF–treated cultures (Fig. 2J). Interestingly, CD11c+CD11b+ DCs increased from 7 ± 1% of the total cells in the GM-CSF–treated cultures to 22 ± 0.5% in the βGlcCer-stimulated GM-CSF–treated cultures (Fig. 2K), indicating that βGlcCers enhance the proportion of CD11c+CD11b+ DCs generated during DC differentiation.
The βGlcCer mixture did not affect the expression of the costimulatory molecules CD80, MHC II, CD86, CD70, 4-1BBL, or GITRL in mDCs, rDCs, and alvDCs (Fig. 3). A slight decrease in OX40L expression was observed in βGlcCer-treated rDCs without HDM, but not in HDM-stimulated DCs (Fig. 3Z). Interestingly, in comparison with mDCs and rDCs, a lower percentage of alvDCs expressed CD80, CD86, CD70, 4-1BBL, and OX40L (Fig. 3D, 3L, 3P, 3T, 3AB), while the alvDCs had an increased percentage of cells expressing GITRL (Fig. 3X). Also compared with mDCs, there was a lower percent of rDCs that expressed 4-1BBL and GITRL but increased OX40L (Fig. 3T, 3X, 3AB).
βGlcCer-enhanced development of DCs is blocked by αT and potentiated by γT in vitro
We have reported that the increased numbers of CD11c+CD11b+ DCs in offspring of allergic mice are decreased by maternal dietary supplementation with αT and increased by maternal dietary supplementation with γT (19, 20). Also, because maternal βGlcCer is sufficient for the increase in CD11c+CD11b+ DCs in offspring of allergic mothers (24), it was determined whether tocopherols directly modify the in vitro βGlcCer-enhanced development of these DCs. To address this, we added αT or γT to GM-CSF–stimulated bone marrow culture in the presence or absence of the βGlcCer mixture. For the cultures with αT treatments, the mixture of optimal doses for βGlcCer from (Fig. 2A–C was used to determine whether αT could reduce optimal βGlcCer stimulation of the development of DCs. In contrast, for cultures with γT, a 1/10 dilution of the βGlcCer mixture from (Fig. 2A–C was used to determine whether there was a synergistic effect of γT and βGlcCers on development of DCs. Tocopherol doses were at concentrations previously described for in vitro treatment with tocopherol isoforms because these doses increase intracellular concentrations in vitro to the levels in tissues in vivo (19, 20, 25). αT supplementation in vitro blocked the βGlcCer-induced increase in numbers of mDCs, rDCs, and alvDCs (Fig. 4A, 4F, 4K). However, αT did not alter the numbers of DCs in cultures without βGlcCer (Fig. 4A, 4F, 4K). In contrast, γT increased the numbers of these DCs in the absence of βGlcCer (Fig. 4P, 4U, 4Z) and, importantly, enhanced the βGlcCer-induced increase in DCs (Fig. 4P, 4U, 4Z). Thus, the βGlcCer-enhanced bone marrow development of DCs was modifiable by the plant-derived dietary lipids αT and γT.
βGlcCer increases numbers of DCs with active autophosphorylated PKC that is blocked by αT and increased by γT in vitro
GM-CSF signals through PKCβ and PKCδ isoforms during DC differentiation (40–47). We previously reported that αT and γT bind to the C1A regulatory domain of recombinant PKCα and function as an antagonist and agonist, respectively (35). The C1A regulatory domain is also present in PKCβ and PKCδ (48). It is not known whether βGlcCer or tocopherol isoforms regulate activation of PKCα/β or PKCδ in DCs during bone marrow development of DCs. Therefore, it was determined whether tocopherols and βGlcCer regulate PKC activity as measured by autophosphorylation at PKCα/β Thr638 (p-PKCα/β) and PKCδ Ser645 (p-PKCδ) (49). βGlcCer induced an increase in numbers of DCs with p-PKCα/β and p-PKCδ that was blocked by αT and highly elevated by γT (Figs. 4B, 4G, 4L, 4Q, 4V, 4AA, 5A, 5E, 5I, 5M, 5Q, 5U). Because βGlcCer increased the numbers of DCs and numbers of DCs with p-PKCs, it was determined what proportion of each DC subset had p-PKCα/β and p-PKCδ. Almost all of the DCs in each subset had p-PKCα/β (Fig. 4C, 4H, 4M, 4R, 4W, 4AB), although the proportion of alvDCs with p-PKCα/β was increased by βGlcCer (Fig. 4M). In contrast, βGlcCer increased the proportion of DCs with p-PKCδ, and this was blocked by αT (Fig. 5B, 5F, 5J). The proportion of cells with p-PKCδ was increased by γT and the 1/10 dilution of βGlcCer (Fig. 5N, 5R, 5V). However, most notably, the levels of p-PKC per cell were regulated by βGlcCer and tocopherols. In brief, the gMFI for p-PKCα/β and p-PKCδ was significantly increased by βGlcCer and by γT in the mDC, rDC, and alvDC phenotypes (Figs. 4S, 4T, 4X, 4Y, 4AC, 4AD, 5O, 5P, 5S, 5T, 5W, 5X). The βGlcCer-induced increases in gMFI for p-PKCα/β (Fig. 4D, 4I, 4N, 4E, 4J, 4O, 4S, 4T, 4X, 4Y, 4AC, 4AD) and p-PKCδ (Fig. 5C, 5D, 5G, 5H, 5K, 5L, 5O, 5P, 5S, 5T, 5W, 5X) were blocked by αT. Thus, βGlcCers increased activity of DC PKCα/β and PKCδ, which was modifiable with tocopherol isoforms in vitro.
Inhibition of PKC isoforms decreases development of DC subsets
Given the increased activity of PKCα/β and PKCδ in DCs in (Figs. 3 and 4, we determined whether pharmacologic inhibition of these PKC isoforms decreased βGlcCer-enhanced development of CD11c+CD11b+ DC subsets. The PKCα/β selective inhibitor Rö 32-0432 (IC50 of 9 nM for PKCα, 28 nM for PKCβ, and 108 nM for PKCε) (50, 51) was added at 30 min before the βGlcCer mixture and GM-CSF at each media change (days 3 and 6) during DC development. Rö 32-0432 mediated a dose-dependent inhibition of the βGlcCer-stimulated increase in mDCs and rDCs, but not alvDCs (Fig. 6A, 6F, 6K). There was no change to the basal level of DC subsets in cultures with GM-CSF (Fig. 6A, 6F, 6K) or cell viability (data not shown), indicating that the inhibition of PKCα/β blocked the βGlcCer-induced increase in DCs without inhibiting GM-CSF stimulation of DC differentiation. As expected, Rö 32-0432 treatment decreased both the proportion of DC subsets, which are p-PKCα/β+ (Fig. 6B, 6G, 6L), and the gMFI for p-PKCα/β within those subsets (Fig. 6C, 6H, 6M). Interestingly, Rö 32-0432 treatment of the cells also decreased the numbers of DCs with active p-PKCδ (Fig. 6D, 6E, 6I, 6J, 6N, 6O). Thus, Rö 32-0432 inhibited the βGlcCer-augmented generation of DCs, suggesting that PKCα/β was involved in the βGlcCer-augmented generation of numbers of mDCs and rDCs. Because there was also Rö 32-0432 inhibition of p-PKCδ, PKCδ was involved.
Interestingly, PKCα/β can increase PKCδ in leukocytes (52). Thus, it was assessed whether inhibition of PKCδ blocks the βGlcCer-augmented generation of DCs. To assess PKCδ function in DC differentiation, we treated the cells with delcasertib. Delcasertib is a small peptide that specifically competitively inhibits the binding of active PKCδ to RACK protein, thereby preventing PKCδ relocation to the membrane for interaction with its substrates in the membrane (53). There are no published IC50 values for delcasertib, but doses ranging from 10 nM to 1 µM delcasertib have been used for in vitro studies (53–55). Furthermore, delcasertib is specific to PKCδ isoform and notably does not affect PKCα (53). Delcasertib induced a dose-dependent reduction in the βGlcCer-stimulated increase in mDCs, rDCs, and alvDCs without altering the numbers of DCs in the GM-CSF–treated cells without βGlcCers (Fig. 7A, 7D, 7G) and without affecting cell viability (data not shown). There was no effect of delcasertib on the number of cells with active p-PKCδ (Fig. 7B, 7E, 7H) or the gMFI for p-PKCδ (Fig. 7C, 7F, 7I) because this peptide inhibitor does not prevent phosphorylation of PKCδ but specifically prevents the signaling function of active p-PKCδ. Thus, these data indicate that inhibition of PKCδ function blocked the βGlcCer enhancement of DC development, and that PKC isoforms were necessary for the βGlcCer-enhanced bone marrow development of CD11c+CD11b+ DC subsets.
βGlcCer effects on Ag processing by DC subsets in vitro
Both the number of DCs and their ability to process and present Ag to T cells influence the impact of DCs on allergic inflammation. We investigated Ag uptake and processing using the model allergen for DC Ag processing, DQ-OVA, a self-quenched DQ fluorescence-conjugated model allergen that when proteolytically degraded has increased fluorescence. All of the mDCs, rDCs, and alvDCs acquired Ag regardless of whether they were exposed to optimal βGlcCer or tocopherol isoforms (data not shown) with processing of DQ-OVA initiated by 30 min (Fig. 8). By 90 min, the mixture of optimal doses of βGlcCers in (Fig. 2 increased DQ-OVA processing by mDCs, but there was no effect of this βGlcCer mixture on processing by rDCs or alvDCs (Fig. 8A, 8C, 8E). The increased Ag processing by βGlcCer-stimulated mDCs was reversed by αT (Fig. 8A). In cultures with the 1/10 dilution of the βGlcCer mixture or the γT treatment, there was a small increase in DQ-OVA processing by DCs (Fig. 8B, 8D, 8F). Thus, at the 90-min time point in (Fig. 8, there was a differential effect of doses of the βGlcCers on the ability of DC subsets to process Ag. Both doses of the βGlcCers induced a small increase in Ag processing by the mDC phenotype, whereas for rDC and alvDC phenotypes, the 1/10 dose of the βGlcCer mix induced a small increase in Ag processing (Fig. 8).
Ag processing per DC ex vivo by lung DCs from pups of allergic and nonallergic mothers
Although there are increased numbers of mDCs, rDCs, and alvDCs and increased levels of βGlcCer in allergen-challenged pups of allergic mothers, which are both modified by maternal dietary supplementation with tocopherol isoforms (19, 20), it is not known whether the DCs from these pups have altered function on a per-cell basis. Therefore, CD11c+ DCs from lungs of allergen-challenged pups from allergic and nonallergic mothers (Fig. 9A) were isolated, and Ag processing was evaluated ex vivo in cultures with equal numbers of CD11c+ DCs for each treatment group. As controls for analysis of pup responsiveness to allergen, pups of allergic mothers had increased bronchoalveolar lavage (BAL) eosinophils, a hallmark of allergy, and this eosinophilia was reduced by maternal dietary supplementation with αT and elevated by maternal dietary supplementation with γT (Fig. 9B, 9C) as we previously described (19, 20). Saline challenges of pups of allergic mothers that do not have allergic inflammation have BAL cells similar to allergen-challenged pups of nonallergic mothers (5). Also, we have reported that in offspring of allergic mothers, maternal diet supplementation with αT decreases and maternal diet supplementation with γT increases lung Th2 cytokines IL-4 and IL-5 and the chemokines CCL11 and CCL24 (19, 20). In the ex vivo cultures for analysis of DC function, there were more Ag-positive mDCs, rDCs, and alvDCs than CD103+ DCs and plasmacytoid DCs (Fig. 9D, 9E). γT treatment in vivo resulted in a small increase in percent of lung mDCs or rDCs with Ag acquisition ex vivo (Fig. 9E). There was little effect of in vivo treatments on Ag processing by the DCs ex vivo (Fig. 9F, 9G). Moreover, although mDCs from pups of allergic mothers had a similar percentage of cells processing Ag compared with rDCs and alvDCs (Fig. 9D, 9E), mDCs processed more Ag per cell as indicated by increased gMFI (Fig. 9F, 9G). Thus, in vitro βGlcCer and tocopherol isoforms had a greater effect on numbers of DCs than on the per-cell Ag processing function of DCs.
DCs from lungs of pups of allergic mothers exhibited enhanced activation of T cells ex vivo, but this was modified in DCs from pups of allergic mothers with dietary tocopherol supplementation
To evaluate ex vivo T cell activation by DCs from pups of allergic and nonallergic mothers, we isolated lung CD11c+ cells from pups born to allergic or nonallergic moms fed either a basal or tocopherol-supplemented diet. The DCs were plated at equal numbers of CD11c+ DCs and then cocultured with CFSE-labeled CD4+ OTII spleen T cells in the presence of OVA peptide. T cell activation was quantified by CFSE dilution in T cells over 72 h. For pups of allergic mothers, there were increased numbers of mDCs and rDCs within the cultures plated with equal numbers of CD11c+ DCs, and the proportion of rDCs was reduced by maternal dietary supplementation with αT (Fig. 10A, 10B). DCs from pups from allergic mothers demonstrated enhanced ex vivo activation of OTII cells (Fig. 10C, 10D). This is consistent with the changes in proportion of mDCs and rDCs in the total DCs from these mice (Fig. 10A, 10B) and with our previous reports of increased total numbers of lung DCs in pups of allergic mothers (19, 20, 24). Furthermore, the increase in ex vivo activation of OTII cells by DCs from lungs of pups from allergic mothers was blocked by maternal dietary αT supplementation (Fig. 10C) and enhanced by maternal dietary γT supplementation (Fig. 10D). Because the proportions of mDCs, rDCs, and alvDC phenotypes were increased in the ex vivo cells plated at equal cell numbers (Fig. 10A, 10B), when data were expressed as number of proliferating OTII cells per total number of the mDCs, rDCs, and alvDC phenotypes in the culture, there was no increase in DC activation of OTII cells. These data suggest that the increased number of mDCs, rDCs, and alvDC phenotypes that most efficiently process Ag (Fig. 9) participate in the increase in OTII proliferation (Fig. 10).
In this study, we demonstrated that the lipids βGlcCers and tocopherols, which modulate allergic responsiveness in offspring of allergic mothers, directly regulated the hematopoietic development of DC subsets in vitro. The differentiation of CD11c+CD11b+ DC subsets, which are involved in initiation of responses to allergens (56), was enhanced in a dose-dependent manner by βGlcCers in vitro. Interestingly, the optimal effective dose for each βGlcCer isoform was different and was consistent with the proportion of these isoforms in the fetal livers of allergic mothers (24). A mixture of these isoforms at the optimal physiologic dose increased the number of mDCs, rDCs, and alvDC phenotypes in BMDC cultures. Moreover, the βGlcCer-induced increase in these subsets was modifiable by tocopherol isoforms. The βGlcCer-induced increase in DCs was blocked by αT and potentiated by γT. In an analysis of cell signals for DC differentiation, activation of PKCα and PKCδ was enhanced by βGlcCer or γT exposure, while αT blocked the βGlcCer-induced enhancement of PKC activity. Pharmacologic inhibitors of PKCα and PKCδ isoform activity also blocked the βGlcCer-induced increase in numbers of CD11c+CD11b+ DC subsets. Functionally, βGlcCers had minimal effect on Ag processing per DC in vitro. Moreover, there was no effect on Ag processing per cell ex vivo by DCs from pups of allergic mothers that were exposed to maternal βGlcCers. Thus, these novel studies indicate that βGlcCers directly increase numbers of CD11c+CD11b+ DC subsets that process Ag, and this effect can be modulated by tocopherol isoforms. Furthermore, we have identified PKC as a downstream mediator of βGlcCer exposure, and that this βGlcCer-induced increase in PKC activation is regulated by tocopherol isoforms. These results have important implications for understanding allergic disease and the development of allergic predisposition in offspring of allergic mothers.
Maternal allergy and its effects on the in utero environment (14, 15) influence the developing immune system in offspring leading to allergic predisposition early in life (12). In contrast, after birth, breast milk allergen–IgG complexes may limit offspring development of responses to the allergens in these complexes in breast milk (57, 58). Interestingly, DCs from offspring of allergic mothers are essential for development of offspring allergy because adoptive transfer of donor cells from neonates from allergic mothers (without donor neonate allergen exposures) indicate that CD11c+ DCs, but not macrophages or Th cells, transfer allergic predisposition to neonatal mice from nonallergic mothers (17). In studies of maternal factors that modulate offspring DCs in vivo, we have previously shown that maternal allergy results in increased maternal βGlcCers that cross the placenta and lead to increased fetal and neonate CD11b+CD11c+ DC subsets (24). These offspring with increased numbers of DC subsets have an increased responsiveness to allergen and allergic phenotype (24). We now have demonstrated a mechanism of action of βGlcCers whereby βGlcCers directly increase the number of CD11c+CD11b+ DCs during hematopoiesis in vitro in a dose-dependent manner. Importantly, the proportions of βGlcCer isoforms found in the fetal liver of offspring of allergic mothers were optimal in vitro (24). Thus, βGlcCers directly enhanced the generation of an increase in number of DCs for Ag presentation. Mechanisms for βGlcCer regulation of hematopoiesis of DCs are in ongoing studies that are beyond the scope of this report.
IL-12 is reported to act as a proinflammatory signal during Th2-mediated lung responses to allergen in sensitized mice (59, 60). Interestingly, we demonstrated that when BMDCs were stimulated with HDMs in vitro, βGlcCer-treated mDCs and rDCs produced slightly more IL-12 than the DMSO solvent control-treated DCs, suggesting that this may contribute somewhat to increased allergic inflammation. Low numbers of alvDCs (<5%) produced any IL-12 regardless of HDM stimulation or βGlcCer treatment, highlighting a functional difference between DC subsets. Whether the βGlcCer-induced increase in HDM-stimulated IL-12 production by DCs contributes to the βGlcCer-mediated increase in allergic lung inflammation in offspring of allergic mothers is in ongoing studies. We also demonstrated that in the absence of allergen stimulation, rDCs and alvDCs produced IL-12 regardless of exposure to βGlcCers; this is consistent with studies in humans showing similar IL-12 production in DCs irrespective of atopic status of subjects (61, 62).
As efficient APCs, DCs profoundly influence the adaptive immune system and skewing of the Th1 and Th2 responses. DCs present Ag on MHC II, and DC costimulatory molecules are important for effective Ag stimulation of T cells. The expressions of MHC II and the costimulatory molecules CD80 and CD86 are not altered in vivo in offspring of allergic mothers or humans with elevated βGlcCer (19, 20, 63). Consistent with this, MHC II, CD80, CD86, CD70, 4-1BBL, GITRL, and OX40L expression had no or minimal change on DCs in vitro during βGlcCer regulation of DC development. Nevertheless, DCs isolated from neonates of allergic mothers exhibited increased stimulation of OTII cells when the total number of DCs were plated at equal numbers. However, when the total number of CD11c+ DCs were plated at equal numbers, there was still an increased proportion of mDCs and rDCs within the total DC population, suggesting that absolute numbers of APCs in offspring of allergic mothers contributes to increased activation of T cells. Another potential contributing factor to increased DC activation of T cells is Ag processing for presentation. However, for DCs from lungs of pups of allergic mothers, there was no change in the ex vivo Ag processing per DC, except a small increase by mDCs ex vivo. Among the DCs, there were more ex vivo CD11c+CD11b+ subsets that acquired Ag compared with CD103+ DCs or plasmacytoid DCs. For the BMDCs that were developed in vitro in the presence of βGlcCer, the greatest effect of βGlcCer was on increasing numbers of DCs because there was a 10-fold higher number of cells for the CD11b+CD11c+ DC subsets, but only a small increase in Ag processing by DCs. Thus, ex vivo and in vitro DCs are significantly impacted by βGlcCer-induced increases in numbers of DCs. This suggests a role for βGlcCer in hematopoiesis of DCs that ultimately increases allergen responsiveness in pups of allergic mothers because DCs are exposed to elevated βGlcCer levels during development in utero and during nursing.
In addition to βGlcCer, maternal dietary lipids, including tocopherol isoforms, alter allergic predisposition in offspring of allergic mothers (19, 20). We have reported that maternal dietary supplementation with αT or γT decreases and increases, respectively, offspring responsiveness to allergen and numbers of DCs, but it was unknown whether the regulation of numbers of DCs was a direct or indirect effect. The present studies demonstrate that the tocopherols had a direct effect on the βGlcCer-induced increase in development of mDCs, rDCs, and alvDCs. αT blocked the βGlcCer-induced increase in DC development but did not alter the development of DCs in cultures without βGlcCer. This suggests that αT did not alter GM-CSF stimulation of the development of DCs but did block the effects of βGlcCer. Moreover, γT enhanced the βGlcCer-stimulated increase in DCs. Thus, the increase in DCs and allergen responsiveness in offspring of allergic mothers is modifiable by a direct effect of tocopherols on development of DCs.
Tocopherols have antioxidant and non-antioxidant functions in regulating cell signaling, including regulation of activation of PKC isoforms (12, 33–35). αT and γT bind the C1A regulatory domain of PKCα with αT antagonizing and γT potentiating PKCα activation. Both PKCα/β and PKCδ contain the C1A regulatory binding domain (64). Moreover, PKC isoforms are critical to the development of DCs (43, 47, 48, 65, 66). In novel data in this study, βGlcCer increased PKCα/β and PKCδ activation in DCs, which was blocked by αT. In contrast, γT in the absence of βGlcCer increased PKCα/β and PKCδ activation similar to the increased level of PKC activation by βGlcCer. Although almost all of the DCs had PKC activation, γT enhanced the βGlcCer-stimulated increase of active PKCα/β per cell in mDCs, but not in other DC subsets. Furthermore, our studies with PKCα/β and PKCδ inhibitors suggest that the βGlcCer-induced increase in numbers of mDCs and rDCs required both PKCα/β and PKCδ activation, while the increase in numbers of alvDCs required PKCδ, but not PKCα/β, activation. This may reveal a preferential role of PKCδ as compared with PKCα/β for downstream signaling in alvDCs compared with other DC subsets. Future and ongoing studies will address the function of βGlcCer activation of PKCs during hematopoiesis of DCs.
In summary, this study provides novel insight into βGlcCer regulation of the development of CD11c+CD11b+ DC subsets. The in vitro and ex vivo studies on DC subsets demonstrate that βGlcCer lipids, which have previously been shown to regulate allergic predisposition in offspring of allergic mothers, directly influence DC subset development and function. βGlcCers increased the numbers of CD11c+CD11b+ DCs, enhanced PKC activation within cells in these DC subsets, and induced a small increase in Ag processing per DC. The increase in proportion of mDCs and rDCs enhanced activation of T cells. Furthermore, these βGlcCer-induced increases in DCs were blocked by PKC inhibitors and αT, an antagonist of PKC, but were enhanced by γT, an agonist of PKC. This study advances the understanding of development of CD11c+CD11b+ DC subsets and their regulation by both endogenously synthesized (βGlcCer) and dietary (tocopherol) lipids. It also provides evidence for an important role for lipid regulation of PKC activation during development of DCs. These studies impact design of potential novel approaches for prevention or intervention in asthma and allergic disease.
This work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH) Grants U01 AI131337 and R01 AI127695 (to J.M.C.-M.), Marshall Klaus Perinatal Research Award (to J.D.L.), and the Pediatric Scientist Development Program (J.D.L.). Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the NIH (Award K12HD000850).
J.D.L., N.A., A.T., and K.T. performed experiments of dendritic cell differentiation and function, participated in figure and manuscript preparation, and did statistical analyses. J.M.C.-M. conceived of the study design and participated in performing experiments, statistical analyses, interpretations, and manuscript preparation.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Abbreviations used in this article:
alveolar-like dendritic cell
bone marrow–derived dendritic cell
geometric mean fluorescence intensity
house dust mite
monocyte-derived dendritic cell
MHC class II
protein kinase C
resident-phenotype dendritic cell
The authors have no financial conflicts of interest.