Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 May 12.
Published in final edited form as: Immunity. 2010 Aug 27;33(2):216–228. doi: 10.1016/j.immuni.2010.08.003

Lysosomal α-Galactosidase Controls the Generation of Self Lipid Antigens for Natural Killer T Cells

Alexandre Darmoise 1,5, Susann Teneberg 2, Lauriane Bouzonville 1, Roscoe O Brady 3, Michael Beck 4, Stefan HE Kaufmann 5, Florian Winau 1,5,*
PMCID: PMC4018304  NIHMSID: NIHMS575853  PMID: 20727792

SUMMARY

Natural Killer T (NKT) cells are lipid-reactive, CD1d-restricted T lymphocytes important in infection, cancer, and autoimmunity. In addition to foreign antigens, NKT cells react with endogenous self lipids. However, in the face of stimulating self antigen, it remains unclear how overstimulation of NKT cells is avoided. We hypothesized that constantly degraded endogenous antigen only accumulates upon inhibition of α-galactosidase A (α-Gal-A) in lysosomes. Here, we show that α-Gal-A deficiency caused vigorous activation of NKT cells. Moreover, microbes induced inhibition of α-Gal-A activity in antigen-presenting cells. This temporary enzyme block depended on Toll-like receptor (TLR) signaling and ultimately triggered lysosomal lipid accumulation. Thus, we present TLR-dependent negative regulation of α-Gal-A as a mechanistic link between pathogen recognition and self lipid antigen induction for NKT cells.

INTRODUCTION

Natural killer T (NKT) cells are a subpopulation of unconventional T lymphocytes that express a restricted T cell receptor (TCR) repertoire and several molecules characteristic for NK cells (Bendelac et al., 2007; Kronenberg, 2005). Following activation, NKT cells respond by a rapid burst of cytokines secreting mainly interferon-γ(IFN-γ) and interleukin-4 (IL-4), thus regulating the quality of downstream immune responses (Bendelac et al., 2007). Therefore, NKT cells play a role in various disease conditions including infections (Tupin et al., 2007), cancer (Cui et al., 1997; Dhodapkar, 2009), and autoimmunity (Shi and Van Kaer, 2006), such as diabetes (Hong et al., 2001; Sharif et al., 2001) and multiple sclerosis (Miyamoto et al., 2001). NKT cells recognize lipid antigens primarily belonging to the group of glycosphingolipids (GSLs) presented by nonclassical, major histocompatibility class I (MHC-I)-like CD1d molecules mainly expressed on dendritic cells (DCs) (Brigl and Brenner, 2004). The first described and most potent NKT cell antigen is α-galactosylceramide (αGalCer), initially isolated from the marine sponge Agelas mauritianus (Kawano et al., 1997). Invariant NKT cells (iNKT) are defined by their reactivity with αGalCer and, thus, are readily detectable by binding to αGalCer-loaded CD1d-tetramers (Matsuda et al., 2000; Benlagha et al., 2000). In the murine system, iNKT cells are predominantly located in peripheral tissues such as liver and spleen (Bendelac et al., 2007).

Upon infection, iNKT cells can be directly activated by pathogens that contain glycolipid antigens such as Sphingomonas spp (Mattner et al., 2005; Kinjo et al., 2005) and Borrelia burgdorferi (Kinjo et al., 2006) that cause a multisystem inflammatory disorder called Lyme disease. Moreover, Mycobacterium tuberculosis contains a glycolipid antigen entity, phosphatidylinositolmannoside (PIM), that is recognized by a subpopulation of iNKT cells (Fischer et al., 2004).

In addition to exogenous antigens, iNKT cells react with self lipids. Initial evidence for the existence of endogenous iNKT cell antigens was provided by experiments demonstrating that tail-truncated CD1d fails to select iNKT cells in the thymus (Chiu et al., 2002). Since truncated CD1d is unable to traffic to lysosomes, these results suggest that lysosomal lipids have to be loaded onto CD1d for proper iNKT cell selection and activation (Chiu et al., 2002). Further, these endogenous lipids most likely represent GSLs because iNKT cells are not able to recognize CD1d-expressing antigen presenting cells (APCs) lacking β-glucosylceramide (Stanic et al., 2003), which is the common precursor molecule for the majority of GSLs. Moreover, analysis of mice deficient for hexosaminidase B (Hex-B) reveals a lack of iNKT cells (Zhou et al., 2004). Since Hex-B is the lysosomal enzyme required for degradation of globotetraosylceramide (Gb4) and isoGb4 (iGb4) into globotriaosylceramide (Gb3) and isoGb3 (iGb3), respectively, it has been proposed that the GSLs downstream of Hex-B could represent the endogenous lipid ligands for iNKT cell selection and activation (Zhou et al., 2004). Accordingly, Hex-B-deficient DCs fail to activate iNKT cells in salmonella infection (Mattner et al., 2005). In contrast to Gb3, iGb3 proves to be a potent antigen to stimulate iNKT cells (Zhou et al., 2004; Mattner et al., 2005).

However, in the presence of stimulating self antigen, it is not known as to how uncontrolled activation of iNKT cells that potentially leads to autoimmunity is prevented. We hypothesized that under normal conditions, endogenous antigen such as iGb3 is constantly degraded to lactosylceramide, which prevents intra-lysosomal concentrations required for efficient CD1d loading and productive iNKT cell induction. Only if lysosomal α-galactosidase A (α-Gal-A), the rate-limiting enzyme of iGb3 turnover, is blocked would endogenous antigen accumulate and reach the threshold for subsequent iNKT cell activation.

The majority of pathogens potentially causing infection of the host lack glycolipid antigens to directly stimulate iNKT cells. However, facing this challenge, the host developed a pathway to ensure proper iNKT cell activation upon infection. Accordingly, pathogens lacking iNKT cell antigens induce the generation of self lipid antigens. This holds true for salmonella infection, which potently induces iNKT cell activation, strictly dependent on CD1d-presented self antigens and IL-12 secretion by microbe-exposed DCs (Brigl et al., 2003). Moreover, bacterial infections can promote the de novo synthesis of GSLs (De Libero et al., 2005). Further, recent evidence demonstrates that stimulation of DCs through Toll-like receptor (TLR)-9 using CpG oligonucleotides elicits the production of charged GSLs that mediate iNKT cell activation in the presence of type I interferons (Paget et al., 2007). However, how TLR engagement affects lipid antigen generation remains elusive.

Here, we show that DCs deficient in α-Gal-A activity induced vigorous iNKT cell activation in the absence of exogenous antigen. Reconstitution of α-Gal-A-deficient (Gla−/−) DCs with recombinant enzyme fully abrogated the iNKT cell response. Moreover, lack of α-Gal-A promoted iNKT cell stimulation in vivo. Endogenous iNKT cells in Gla−/− hosts showed signs of chronic exposure to self antigens. Beyond the model of α-Gal-A deficiency, we demonstrate that bacterial infection and microbial products induced inhibition of α-Gal-A activity in wild-type DCs. This temporary enzyme block strictly depended on TLR signaling through the adaptor protein MyD88 and ultimately triggered lysosomal lipid accumulation. Taken together, we propose that TLR-dependent inhibition of α-Gal-A in DCs mechanistically couples pathogen recognition with generation of lipid antigens for iNKT cell activation.

RESULTS

DCs Lacking α-Gal-A Activity Induce iNKT Cell Responses

We hypothesized that catabolic GSL intermediates such as iGb3 are normally degraded by the action of α-Gal-A in the lysosome, thus preventing antigen presentation to iNKT cells. Following this idea, endogenous antigen has to be rescued from degradation to be sufficiently abundant for iNKT cell activation. To this end, we generated bone marrow-derived DCs from Gla−/− and wild-type (WT) mice and cocultured them with iNKT cells isolated from liver and thymus of Vα14-transgenic animals. As positive control, we pulsed both antigen-presenting cells (APCs) with αGalCer and added a blocking antibody against CD1d in parallel in order to assess CD1d restriction of T cell activation. Two days after incubation, we measured 3H-thymidine incorporation into the DNA of expanding cells detected by counts per minute (cpm). In the absence of exogenous antigen, Gla−/− DCs induced vigorous proliferation of liver as well as thymic iNKT cells in contrast to WT DCs (Figure 1A and Figure S1A available online). The iNKT cell response was fully abrogated in the presence of blocking antibody, indicating that self antigen was presented through CD1d. DCs deficient for α-Gal-A normally activated iNKT cells by αGalCer, excluding any impairment of the lipid antigen presentation pathway (Figure 1A). Moreover, Gla−/− DCs efficiently presented ovalbumin to conventional T cells, showing normal peptide antigen presentation in APCs lacking α-Gal-A (Figure S1B).

Figure 1. DCs Lacking α-Gal-A Activity Induce iNKT Cell Responses.

Figure 1

Open in a new tab

(A) DCs from wild-type (WT) or Gla−/− mice were pulsed with αGalCer (100 ng/ml) for 16 hr, or were left untreated, before coculture with liver iNKT cells from Vα14-Jα18 transgenic mice. Additionally, DCs were treated with CD1d-blocking antibody (20 μg/ml) or isotype control (iso, 20 μg/ml). Two days after stimulation, antigen-specific proliferation was determined by [3H]-thymidine incorporation as depicted by counts per minute (CPM).

(B) DCs from WT or Gla−/− mice were treated with DGJ (1 mM), or human recombinant α-Gal-A (Replagal, 20 μg/ml), or were left untreated. In parallel, DCs were pulsed with iGb3 (1 μg/ml) for 16 hr. After washing, DCs were cocultured with liver iNKT cells. Moreover, DCs additionally lacking CD1d expression (Cd1d1−/−) were used as APCs. Three days after stimulation, cytokines released in the culture supernatants were determined by ELISA.

(C) DCs from WT, Gla−/−, Gla−/−Cd1d1−/−, or β-Gal-deficient (Glb1−/−) mice were cocultured with the invariant NKT cell hybridoma DN32.D3 or the noninvariant TCB11. Additionally, WT APCs were preincubated with 1 mM DGJ, and Gla−/− DCs were reconstituted with 10 or 20 μg/ml recombinant α-Gal-A. One day after coculture, IL-2 secretion of hybridoma cells was measured by ELISA. (A–C) Graphs show mean ± SD of triplicate cultures.

(D) IB4-mediated blocking of endogenous antigens abrogates iNKT cell activation by α-Gal-A-deficient APCs. Gla−/− and WT DCs were pulsed for 6 hr with iGb3 (400 ng/ml), αGalCer (100 ng/ml), or were left unpulsed. In addition, DCs were preincubated with 0.5–5 μg/ml IB4 for 2 hr or were left untreated. Subsequently, APCs were cocultured with the invariant NKT cell hybridoma DN32.D3 for 1 day, prior to measurement of IL-2 by ELISA. Depicted is the percentage of the remaining IL-2 response of iNKT cells upon IB4 treatment compared to untreated DCs (Unpulsed, 200 pg/ml; iGb3, 277 pg/ml and 638 pg/ml; αGalCer, 2866 pg/ml and 3824 pg/ml, for WT or Gla−/− DCs, respectively). In the right panel, Gla−/− DCs were preincubated with 0.5–2 μg/ml IB4 or were left untreated. Subsequently, APCs were cocultured with primary liver iNKT cells, and secretion of IFN-γ and IL-4 was measured by ELISA 2 days after stimulation. Depicted is the percentage of the remaining cytokine response of liver iNKT cells upon IB4 treatment compared to untreated DCs (758 pg/ml for IFN-γ and 304 pg/ml for IL-4)

(E) WT and Gla−/− DCs were incubated with 100 μg/ml NB-DGJ for the indicated period of time or were left untreated. Subsequently, APCs were cocultured with the invariant NKT cell hybridoma DN32.D3, and IL-2 secretion was measured by ELISA 1 day after stimulation. Depicted is the percentage of the remaining IL-2 response of iNKT cells upon NB-DGJ treatment compared to untreated DCs (60 pg/ml for WT and 235 pg/ml for Gla−/− DCs). Graphs show mean ± SEM of triplicate cultures. (A–E) Results depicted are representative of at least three independent experiments with similar results. *p < 0.05; **p < 0.01; ***p < 0.001 as analyzed by Student’s t test. NS, not significant.

To demonstrate effector functions of iNKT cells, we cocultured Gla−/− and WT DCs with liver iNKT cells and measured cytokine secretion of IFN-γ and IL-4 by ELISA. Without addition of exogenous antigen, α-Gal-A-deficient DCs potently elicited IFN-γ and IL-4 secretion by iNKT cells in sharp contrast to untreated WT DCs (Figure 1B). Importantly, cytokine secretion by activated iNKT cells was fully abrogated when Gla−/− DCs additionally lacking CD1d (Gla−/−Cd1d1−/−) were used for T cell stimulation (Figure 1B). In parallel, we treated WT DCs with the specific α-Gal-A inhibitor deoxygalactonojirimycin (DGJ) (Fan et al., 1999). Blocking of α-Gal-A enzymatic activity in WT DCs induced CD1d-restricted cytokine secretion by activated liver iNKT cells (Figure 1B). Finally, we reconstituted α-Gal-A-deficient APCs with recombinant enzyme prior to functional T cell assays and showed that iNKT cell activation was completely abolished (Figure 1B). To substantiate our findings with primary iNKT cells, we performed analogous experiments using autoreactive NKT cell hybridoma. Both α-Gal-A-deficient DCs and WT DCs treated with DGJ strongly stimulated the invariant NKT cell hybridoma DN32.D3 in contrast to the nonin-variant NKT cell hybridoma TCB11 (Figure 1C). Consistently, enzyme reconstitution or coculture with Gla−/−Cd1d1−/− DCs abolished DN32.D3 stimulation (Figure 1C). Notably, β-galactosidase-deficient DCs failed to increase the autoreactive response of DN32.D3 cells, underlining the specific function of α-Gal-A in generation of endogenous antigens for iNKT cells (Figure 1C). As control, Gla−/− APCs normally stimulated invariant NKT cell hybridoma by exogenous lipid antigens (Figure S1C).

Next, we aimed at further defining the entity of the endogenous antigens that accumulate in α-Gal-A deficiency. For this purpose, we used the isolectin B4 (IB4), which specifically binds to digalactose residues in α1,3-glycosidic linkage as it is present in iGb3, and thus blocks lipid recognition by T cells. In this context, we pretreated lipid-incubated WT and α-Gal-A-deficient DCs as well as unpulsed APCs with IB4, prior to coculture with the invariant NKT cell hybridoma DN32.D3. IB4 incubation had no effect on the presentation of αGalCer by WT or Gla−/− DCs (Figure 1D). However, the T cell response to iGb3-pulsed DCs was inhibited in a dose-dependent manner and reduced by 70%–80% at high IB4 concentrations. Similarly, IB4-mediated blocking fully abrogated iNKT hybridoma responses induced by Gla−/− APCs (Figure 1D). In addition, IB4-mediated block of self antigen recognition drastically reduced the α-GalA-dependent activation of primary liver iNKT cells (Figure 1D). In order to define the class of lipid antigen involved in α-Gal-A-mediated iNKT cell induction, we incubated WT and Gla−/− DCs for different periods of time with N-butyl-DGJ (NB-DGJ) that inhibits β-glucosylceramide-synthase, which produces the main precursor for the majority of GSLs. Figure 1E shows the reduction of iNKT cell-derived IL-2 upon DC treatment with NB-DGJ compared to untreated DCs. Although short-term treatment of Gla−/− DCs with NB-DGJ had virtually no effect on iNKT cell stimulation, inhibitor incubation for up to 5 days reduced activation of iNKT cells to ~50% (Figure 1E). At the concentrations tested, NB-DGJ was not toxic and did not affect the phenotype of DCs (Figure S2A). As further controls, we excluded that the capacity of α-Gal-A-deficient APCs to strongly activate iNKT cells was due to increased CD1d expression, different APC phenotype (Figures S2B and S2C), or differential localization of complexes of self lipids and CD1d to lipid rafts (Figure S3).

Taken together, we demonstrate in vitro that lack of α-Gal-A enzymatic activity promotes the production of self antigens to activate iNKT cells.

α-Gal-A Deficiency Promotes Activation of iNKT Cells In Vivo

Next, we analyzed the impact of lacking α-Gal-A on iNKT cell activation in vivo. WT recipient mice received 1 × 106 WT, Gla−/−, or Gla−/−Cd1d1−/− DCs intravenously. As a positive control, we administered WT DCs pulsed with iGb3. We determined activation of liver iNKT cells measuring TCR downmodulation by flow cytometry 18 hr after DC transfer. Moreover, we measured upregulation of the activation marker molecules CD69 and CD25 on the surface of iNKT cells. Following stimulation with Gla−/− DCs, the frequency of αGalCer-loaded CD1d-tetramer+ cells among liver lymphocytes was reduced by half (~12%) when compared to unpulsed WT DCs (~23%) and Gla−/−Cd1d1−/− DCs (~24%) (Figure 2A). Since reduced frequencies of CD1d-tetramer+ iNKT cells indicate TCR down-modulation upon activation, α-Gal-A-deficient DCs induced CD1d-dependent stimulation of iNKT cells in vivo. The bar chart in Figure 2A shows the significant differences in liver iNKT cell percentages in recipient mice stimulated with the different types of DCs when compared to PBS-treated control animals. These findings were corroborated by increased expression of CD69 and CD25 molecules by liver iNKT cells subsequent to stimulation with Gla−/− in contrast to Gla−/−Cd1d1−/− DCs (Figure 2B).

Figure 2. Absence of α-Gal-A Mediates iNKT Cell Activation In Vivo.

Figure 2

Open in a new tab

(A) α-Gal-A-deficient DCs activate iNKT cells in vivo. Wild-type (WT) mice received intravenously (i.v.) 1 × 106 WT DCs, pulsed with iGb3 (1 μg/ml) for 16 hr as indicated. Alternatively, mice received 1 × 106 DCs generated from Gla−/− or Gla−/−Cd1d1−/− mice. As controls, recipient animals received PBS or untreated WT DCs. After 18 hr, liver mononuclear cells were analyzed by flow cytometry. The percentages of iNKT cells defined as αGalCer-loaded CD1d-tetramer+ TCRβ+ cells among B220− lymphocytes are depicted in the dot plots. The right panel shows the differences in liver iNKT cell percentages in recipient mice stimulated with the different types of DCs when compared to PBS-treated control animals. For the bar chart, the data from five independent experiments were pooled, including three mice per group. *p < 0.05; **p < 0.01; ***p < 0.001 as analyzed by one-way ANOVA with variance comparison between all groups followed by a Bonferroni post-test.

(B) Surface expression of CD25 and CD69 by iNKT cells was analyzed upon transfer of Gla−/−, Gla−/− Cd1d1−/−, or iGb3-pulsed WT DCs, as described in (A).

(C) Deficiency of α-Gal-A promotes iNKT cell expansion in vivo. Thymocytes from Vα14-Jα18 transgenic mice were subjected to CD8+ T cell depletion, labeled with CFSE, and injected intravenously (i.v.) into sublethally irradiated WT, Cd1d1−/−, Gla−/−, or Gla−/−Cd1d1−/− mice. Nine days after adoptive transfer, iNKT cells prepared from livers of recipient mice were identified by flow cytometric staining with anti-TCRβ and CD1d-tetramers as shown in the dot plots. In parallel, cotransferred conventional T cells were detected as TCRβ+, CD1d-tetramer− liver lymphocytes. Histograms of the left panel demonstrate the proliferation profiles of CFSE-labeled transferred iNKT cells, including their mean fluorescence intensity (MFI). The right panel depicts CFSE dilution of conventional T cells as controls. (B and C) Shown is one representative experiment of three performed with similar results.

Further, we examined the expansion of transferred iNKT cells in an α-Gal-A-deficient host environment. For this purpose, CFSE-labeled thymocytes isolated from Vα14-transgenic mice were transferred intravenously into sublethally irradiated Gla−/−, Gla−/− Cd1d1−/−, WT, or Cd1d1−/− recipient mice. Nine days after adoptive T cell transfer, lymphocytes from livers of recipient mice were stained with CD1d-tetramers to identify iNKT cells, and CFSE dilution profiles reflecting iNKT cell proliferation were determined by flow cytometry. Transferred iNKT cells in WT, Cd1d1−/−, and Gla−/− Cd1d1−/− recipients showed a classical homeostatic proliferation pattern with the majority of cells undergoing two to three rounds of cell division (Matsuda et al., 2002) (Figure 2C). In sharp contrast, the proliferation profile in Gla−/− mice showed that the majority of iNKT cells underwent four to six rounds of division. This was also reflected by the mean fluorescence intensity (MFI) of CFSE+ iNKT cells in Gla−/− mice, which was reduced by half compared to Gla−/−Cd1d1−/− recipients, indicating an increased expansion of iNKT cells in α-Gal-A deficiency (Figure 2C). In parallel, we determined the proliferation of cotransferred conventional T cells in Gla−/−, Gla−/−Cd1d1−/−, WT, or Cd1d1−/− recipient mice. Accordingly, CFSE dilution of TCRβ+ CD1d-tetramer− T cells was not enhanced in Gla−/− recipients when compared to control mice (Figure 2C). In addition, we tested the expansion of transferred iNKT cells in α-Gal-A-deficient and WT hosts upon exogenous antigenic stimulation with αGalCer. In contrast to the abundance of conventional T cells, the frequencies of transferred iNKT cells in liver and spleen were drastically increased in Gla−/− mice when compared to WT recipients (Figure S4).

Taken together, our findings demonstrate that lack of α-Gal-A facilitates iNKT cell activation in vivo.

Peripheral iNKT Cells Show Signs of Chronic Exposure to Self Antigens in Gla−/− Mice

Subsequently, we investigated the endogenous iNKT cell population of α-Gal-A-deficient mice with regard to signs of overstimulation possibly because of chronic antigen exposure. After staining thymus, spleen, and liver of Gla−/− mice using CD1d-tetramers, flow cytometry revealed that iNKT cell numbers were reduced mainly in peripheral tissues like liver and spleen (Figure 3A). Loss of peripheral iNKT cells could be either because of impaired thymic selection and subsequent decreased output to the periphery, or might be the consequence of increased activation-induced cell death due to overstimulation in peripheral tissues. In order to address the first option, we provided Gla−/− and WT mice with bromodeoxyuridine (BrdU) in the drinking water for 5 days. Subsequently, CD1d-tetramer+ thymocytes were stained for BrdU incorporation, reflecting their endogenous expansion. In addition, thymic iNKT cells were stained for CD44 and NK1.1 to determine their maturation stages. Flow-cytometric analysis revealed that the expansion of immature, semimature, and mature thymic iNKT cells in Gla−/− mice was comparable to WT animals as indicated by the percentage of BrdU+ cells (Figure 3B). Additionally, we analyzed the proliferation of conventional lymphocytes in the thymus of Gla−/− and WT mice following BrdU administration. After 18 hr, thymic lymphocytes were stained for heat stable antigen (HSA, CD24) that is highly expressed on immature thymocytes and subsequently decreases upon T cell maturation. Figure 3B shows the similar generation of CD1d-tetramer−, NK1.1−, HSA+ conventional lymphocytes in α-Gal-A-deficient and WT thymi. In striking contrast, the frequency of BrdU+ cells and, thus, the expansion of liver and splenic iNKT cells compared to conventional T lymphocytes and NK cells was markedly reduced in Gla−/− mice (Figure 3C). Thus, homeostasis of iNKT cells is specifically affected in the periphery of α-Gal-A-deficient mice.

Figure 3. Homeostasis of iNKT Cells Is Impaired in the Periphery of α-Gal-A-Deficient Hosts.

Figure 3

Open in a new tab

(A) Specific loss of peripheral iNKT cells is a hallmark of Gla−/− mice. Dot plots show percentages of αGalCer-loaded CD1d-tetramer+ CD44+ iNKT cells among liver mononuclear cells, splenocytes, and thymocytes from WT or Gla−/− mice measured by FACS. The bar charts demonstrate the percentages or absolute numbers of iNKT cells in liver, spleen, or thymus of α-Gal-A-deficient compared to WT mice.

(B) WT and Gla−/− mice received BrdU (0.8 mg/ml) in the drinking water. Five days later, mice were sacrificed and BrdU incorporation by immature (CD44− NK1.1−), semimature (CD44+ NK1.1−), and mature (CD44+ NK1.1+) thymic iNKT cells was analyzed after intracellular staining with a BrdU antibody. iNKT cells were identified as αGalCer-loaded CD1d-tetramer+ TCRβ+ cells. Additionally, the generation of conventional T cells was analyzed in α-Gal-A-deficient and WT thymi by detection of CD1d-tetramer−, NK1.1−, HSA+ lymphocytes. Depicted are the percentages of BrdU-incorporating cells.

(C) Following the same approach as described in (B), liver and splenic iNKT cells in Gla−/− and WT mice were analyzed for BrdU incorporation, including the detection of conventional T cells and NK cells as controls. (A–C) Results are representative of at least three independent experiments with similar results. Error bars depicted in graphs represent mean values ± SD. *p < 0.05; **p < 0.01; ***p < 0.001 as analyzed by Student’s t test. NS, not significant.

To address the question of peripheral loss of iNKT cells due to increased apoptosis, we intravenously injected a fluorogenic substrate (Flivo) for poly-caspases, a group of enzymes activated upon activation-induced cell death, into Gla−/− and WT mice. After 45 min, iNKT cells from liver and thymus were detected by CD1d-tetramers, and Flivo fluorescence was measured by flow cytometry. Liver iNKT cells in Gla−/− mice exhibited an increased frequency (up to 20%) of apoptotic cells compared to WT animals (Figure 4A). Moreover, conventional T cells and NK cells did not exhibit Flivo fluorescence in liver and spleen of Gla−/− mice (data not shown), indicating that increased apoptosis was specific for iNKT cells in α-Gal-A deficiency.

Figure 4. Peripheral iNKT Cells Show Signs of Chronic Exposure to Self Antigens in Gla−/− Mice.

Figure 4

Open in a new tab

(A) WT and Gla−/− mice received i.v. a fluorescent poly-caspase substrate (Flivo). After 45 min, mice were sacrificed, and substrate fluorescence was measured in αGalCer-loaded CD1d-tetramer+ TCRβ+ iNKT cells.

(B) Dot plots show the percentages of liver, spleen, and thymus iNKT cells expressing surface Ly49C/I in WT and Gla−/− mice. iNKT cells were identified as in (A).

(C) Peripheral iNKT cells in Gla−/− mice are tolerant. WT and Gla−/− mice received i.v. αGalCer in PBS containing 0.5% polysorbate 20 or PBS/0.5% polysorbate 20 alone (vehicle). After 3 hr, mice were sacrificed and liver mononuclear cells were isolated and stained with αGalCer-loaded CD1d-tetramers and TCRβ mAb. Cells were fixed and permeabilized prior to intracellular IFN-γ staining analyzed by flow cytometry. (B and C) Numbers in dot plots represent percentages of Ly49C/I+ or IFN-γ+ iNKT cells in the respective quadrants.

(D) Serum concentrations of IFN-γ in WT and Gla−/− mice were determined by ELISA 16 hr following i.p. injection of αGalCer in vehicle or vehicle alone. Error bars depicted in graphs represent mean values ± SEM (A–D). Results are representative of at least three independent experiments with similar results. At least three mice were used per group.

Of note, persistent antigenic stimulation of iNKT cells leads to upregulation of inhibitory receptors of the Ly49 family (Hayakawa et al., 2004). Therefore, we investigated the expression of Ly49 molecules on the surface of iNKT cells in α-Gal-A-deficient mice and found iNKT cells expressing Ly49C/I were markedly enriched in liver and spleen when compared to WT mice (Figure 4B). Finally, we examined whether iNKT cells in α-Gal-A deficiency display a state of functional tolerance to antigenic stimulation. For this purpose, we challenged Gla−/− mice compared to WT animals with 2 μg αGalCer and, 3 hr later, analyzed intracellular IFN-γ production of liver iNKT cells using flow cytometry. The intracellular IFN-γ response of iNKT cells to αGalCer in Gla−/− mice was impaired compared to that in WT animals (Figure 4C). The serum concentrations of IFN-γ, measured 16 hr after αGalCer challenge, were equally diminished in Gla−/− mice (Figure 4D). To conclude, our findings on increased apoptosis, augmented L49C/I expression, and functional unresponsiveness suggest that iNKT cells are prone to overstimulation and tolerance induction in α-Gal-A deficiency indicative for chronic antigen exposure.

DCs Deficient for α-Gal-A Amplify Antimicrobial Responses by iNKT Cells

In infection, DCs recognize microbes by binding of pathogen-associated molecular patterns (PAMPs) through TLRs and consequently produce self lipid antigens for iNKT cell activation (Brigl et al., 2003; Paget et al., 2007; Salio et al., 2007). Therefore, we investigated the impact of α-Gal-A deficiency on the antimicrobial iNKT cell response. For this purpose, we incubated Gla−/− compared to WT DCs with Listeria monocytogenes, or the TLR ligands LPS and CpG oligodeoxynucleotides (ODN), and subsequently measured cytokine secretion by iNKT cells. Infection and treatment with PAMPs mainly triggered IFN-γ production by iNKT cells when compared to the amount of secreted IL-4. Importantly, the magnitude of IFN-γ secretion induced by WT DCs incubated with bacteria or PAMPs was similar to the amount triggered by untreated Gla−/− DCs (Figure 5). Moreover, the IFN-γ response by activated iNKT cells was drastically amplified when α-Gal-A-deficient DCs were used as APCs upon bacteria or PAMP treatment (Figure 5). This increase in cytokine response was fully abrogated when Gla−/− Cd1d1−/− DCs were used for iNKT cell stimulation, reemphasizing that induced self antigens were presented through CD1d in α-Gal-A-lacking APCs. Of note, iNKT cell activation by pathogens or their microbial products strictly depended on TLR signaling because DCs deficient for the TLR adaptor molecule MyD88 failed to stimulate iNKT cells (Figure 5). The augmented IFN-γ secretion by iNKT cells upon stimulation with Gla−/− DCs treated with PAMPs was mediated by DC-derived IL-12 because addition of IL-12 neutralizing antibody to cocultures reduced the cytokine concentration to the amount of untreated Gla−/− DCs (Figure S5A). However, α-Gal-A-deficient DCs did not produce higher baseline concentrations of IL-12p70 when compared to WT DCs (Figure S5B). Finally, preincubation of α-Gal-A-deficient DCs with the inhibitor of β-glucosylceramide-synthase, NB-DGJ, did not alter the enhancement of the antimicrobial iNKT cell response (Figure S5C). This suggests that catabolic rather than newly synthesized GSLs are responsible for iNKT cell activation by Gla−/− APCs. Taken together, we show that DCs lacking α-Gal-A amplify the antimicrobial response of iNKT cells.

Figure 5. DCs Lacking α-Gal-A Amplify Antimicrobial Responses by iNKT Cells.

Figure 5

Open in a new tab

DCs from WT, Myd88−/−, Gla−/−, and Gla−/− Cd1d1−/− mice were incubated for 16 hr with L. monocytogenes, LPS (1 μg/ml), CpG ODN (2 μg/ml), or were left untreated, prior to coculture with liver iNKT cells. After 72 hr, IFN-γ release into the culture supernatants was quantified by ELISA. Results are representative of at least three independent experiments with similar results. Error bars depicted in column graphs represent mean values ± SEM of triplicate cultures. *p < 0.05; **p < 0.01; ***p < 0.001 as analyzed by Student’s t test. NS, not significant.

Microbes Inhibit α-Gal-A Activity in APCs Through MyD88-Dependent TLR Signaling

Since (1) TLR engagement induces iNKT cell activation by self antigens and (2) lack of α-Gal-A induces iNKT cell activation by self antigens, we hypothesized that TLR engagement is equivalent to an α-Gal-A block. Beyond the use of α-Gal-A-deficient mice as a model, we considered TLR-mediated inhibition of α-Gal-A activity as a physiological mechanism. To test this hypothesis, we incubated WT and MyD88-deficient DCs with L. monocytogenes, the TLR-4-ligand LPS, or the TLR-3-ligand polyI:C, a synthetic double-stranded RNA. After 2–20 hr, DCs were disrupted for subsequent measurement of α-Gal-A activity using the specific fluorogenic substrate 4-methylumbelliferyl-α-D-galactopyranoside. Exposure to listeriae or LPS resulted in a ~50% reduction of α-Gal-A enzyme activity compared to untreated WT DCs (Figure 6). In contrast, stimulation with polyI:C had no effect on α-Gal-A activity. Of note, enzymatic activity upon stimulation with LPS recovered to baseline levels after 20 hr incubation, indicating temporary inhibition of α-Gal-A by PAMPs. Importantly, enzyme block by bacteria and PAMPs was abolished in the absence of the TLR-adaptor molecule MyD88 (Figure 6). The absolute values of α-Gal-A activity were similar between untreated WT and Myd88−/− DCs or decreased upon treatment with DGJ, whereas no activity was detected in α-Gal-A-deficient DCs (Table S1). Further, we determined the transcriptional regulation of α-Gal-A and other GSL-degrading or biosynthetic enzymes by quantitative real-time RT-PCR. Expression of α-Gal-A was strongly upregulated in DCs treated with LPS (Figure S6). Thus, reduction of α-Gal-A enzymatic activity upon PAMP treatment of DCs was not because of transcriptional downregulation.

Figure 6. TLR Stimulation Mediated through MyD88 Inhibits α-Gal-A Activity in DCs.

Figure 6

Open in a new tab

WT and Myd88−/− DCs were incubated with L. monocytogenes (MOI 5), exposed to LPS (1 μg/ml), or polyI:C (10 μg/ml) for the indicated period of time. After extensive washing, DC pellets were disrupted, and α-Gal-A activity was measured in lysates using a fluorogenic substrate specific for α-Gal-A. Activity of α-Gal-A is indicated as percentage of remaining enzyme activity as compared to WT, untreated DCs. Results are representative of at least five independent experiments with similar results.

Taken together, we demonstrate that TLR signaling leads to temporary inhibition of α-Gal-A activity in DCs.

α-Gal-A Is a Central Regulator of TLR-Induced Lysosomal GSL Accumulation

Inhibition or lack of α-Gal-A enzyme activity leads to intralysosomal accumulation of lipids upstream of α-Gal-A in the GSL degradation pathway. One of these GSLs is Gb3, which also functions as a receptor for Shiga toxin subunit B (STxB) in B cells and DCs (Pina et al., 2007). Since iGb3 detection requires complex techniques such as ion trap mass spectrometry (Li et al., 2008; Li et al., 2009), we used detection of the α-GalA substrate Gb3 as a reporter system for lipid accumulation upon α-Gal-A inhibition. To this end, we employed fluorochrome-coupled STxB for staining Gb3 by flow cytometry. Indeed, due to accumulation of α-Gal-A-dependent lipid substrates, STxB staining for Gb3 was drastically increased in α-Gal-A-deficient compared to WT DCs (Figure 7A).

Figure 7. α-Gal-A Is a Central Regulator of TLR-Induced Lysosomal GSL Accumulation.

Figure 7

Open in a new tab

(A) DCs from WT and Gla−/− mice were incubated with Alexa 488-coupled STxB (5 μg/ml) for 30 min at 37°C. After extensive washing, DCs were stained with CD11c mAb, washed, and analyzed by FACS. Numbers in dot plots represent percentages of STxB+ DCs in the respective quadrants.

(B) WT and MyD88-deficient DCs were incubated with Alexa 488-coupled STxB (5 μg/ml) for 30 min, before LPS (1 μg/ml), polyI:C (10 μg/ml), CpG ODN (2 μg/ml), or beads were added for 12 hr. In parallel, DCs were treated with bafilomycin A1 (50 nM) for 1 hr prior to incubation with STxB. After extensive washing, DCs were analyzed by flow cytometry. Histograms represent STxB fluorescence measured in untreated compared to treated CD11c+ DCs. (A and B) Shown is one representative experiment out of three performed with similar results.

Next, we incubated WT and MyD88-deficient DCs with fluorochrome-conjugated STxB and subsequently stimulated the cells with PAMPs such as LPS, CpG oligonucleotides, and polyI:C, or left DCs untreated. After 12 hr of incubation, cells were analyzed by flow cytometry. LPS and CpG oligonucleotides increased STxB staining of DCs compared to untreated cells, indicating GSL accumulation (Figure 7B). Most importantly, this increase in STxB staining was completely abrogated in MyD88-deficient DCs, which is consistent with the TLR-dependent inhibition of α-Gal-A activity to induce lipid accumulation. On the other hand, polyI:C did not induce lipid accumulation in DCs as indicated by lack of increased STxB staining (Figure 7B), which is in line with absent α-Gal-A inhibition upon stimulation with polyI:C. Moreover, we excluded nonspecific effects on GSL induction mediated by triggering of the endocytotic machinery because latex beads applied to DCs did not affect STxB staining (Figure 7B). Finally, we showed that lipid accumulation critically depended on lysosome function because blocking lysosomal acidification using bafilomycin fully abrogated STxB staining. In conclusion, we demonstrate that α-Gal-A is a central regulator of TLR-induced GSL accumulation.

DISCUSSION

Here we have demonstrated an essential function of the lysosomal enzyme α-Gal-A in the generation of self lipid antigens for iNKT cell activation. Using Gla−/− mice as a model for constitutive α-Gal-A deficiency, we have shown that DCs lacking the lysosomal enzyme strongly activate iNKT cells a priori, without additional sensitization or exogenously added antigen. Reconstituting α-Gal-A-deficient DCs with recombinant enzyme fully abolished iNKT cell stimulation, demonstrating the pivotal role of α-Gal-A in activation of iNKT cells by self antigen. Additionally, we induced α-Gal-A deficiency in wild-type DCs using the enzyme inhibitor DGJ, which resulted in productive antigen presentation of CD1d-restricted endogenous lipids. These experiments provide the proof of principle that α-Gal-A function is central to controlling the threshold of endogenous antigen availability. Further, the activation of an invariant in contrast to noninvariant NKT cell hybridoma by Gla−/− APCs corroborates the notion that endogenous lysosomal lipids mediate iNKT cell stimulation in α-Gal-A deficiency. Notably, β-Gal-deficient DCs did not amplify the induction of iNKT cells, indicating that nonspecific lysosomal lipid storage was not responsible for the α-Gal-A-mediated mechanism of iNKT cell activation. Moreover, the IB4-mediated abrogation of iNKT cell stimulation by α-Gal-A-deficient DCs strongly suggests that endogenous GSLs that bear a terminal Gal(α1,3)Gal disaccharide residue such as iGb3 are the respective antigens accumulating in α-Gal-A deficiency. Lack of α-Gal-A also facilitated iNKT cell stimulation in vivo, as shown by immunization of wild-type mice with α-Gal-A-deficient DCs, as well as adoptive transfer of thymic iNKT cells upon Gla−/− recipients. Of note, genetic combination of α-Gal-A deficiency and lack of CD1d totally abrogated iNKT cell activation, indicating α-Gal-A-dependent presentation of CD1d-restricted self antigen.

In 1967, the enzyme mediating the degradation of Gb3 to lactosylceramide was described as ceramidetrihexosidase, later called α-galactosidase A, and its enzymatic defect has been shown to cause Fabry disease (Brady et al., 1967). In patients, this lysosomal storage disorder is characterized by the accumulation of Gb3 (Desnick and Schuchman, 2002). Since endothelial cells are mainly affected by the stored lipids, thickened vessels are the main reasons for Fabry symptoms, including skin lesions, pain, and ischemia of heart, kidney, and brain (Beck, 2007). Although the corresponding mouse model of Gla−/− mice accumulates α-Gal-A-dependent lipids, animals appear healthy and do not show apparent clinical signs (Ohshima et al., 1997). An early study investigating the processing of exogenous glycolipids indicated a reduced frequency of splenic iNKT cells in Gla−/− mice (Prigozy et al., 2001). Consistently, we found decreased numbers of iNKT cells in liver and spleen of α-GalA-deficient animals. Reduced iNKT cell frequencies could be because of either impaired thymic development or loss of peripheral cells based on uncontrolled activation. Previous reports suggested that lysosomal storage disorders affect iNKT cell generation in the thymus (Gadola et al., 2006; Schumann et al., 2007). Although this has been studied in mice lacking β-hexosaminidase B or β-galactosidase, effects on thymic selection were not demonstrated in Gla−/− mice (Gadola et al., 2006). Moreover, a recent study showed that single lysosomal enzyme deficiencies do not interfere with thymic iNKT cell selection (Plati et al., 2009). Consistent with this, we observed normal kinetics of iNKT cell development in the thymus of Gla−/− mice. Therefore, reduced iNKT cell numbers in α-Gal-A deficiency are caused by peripheral loss putatively due to overstimulation by accumulated self antigen. Accordingly, we found increased activation-induced apoptosis of liver and splenic iNKT cells in α-Gal-A-lacking mice. Further, iNKT cells in Gla−/− hosts displayed functional tolerance to antigenic challenge, as reflected by markedly reduced cytokine responses to αGalCer stimulation. Finally, peripheral iNKT cells in an α-Gal-A-deficient environment express Ly49C/I receptors at higher frequencies. The family of inhibitory Ly49 molecules binds to MHC-I proteins in order to silence autoreactive T cell responses (Dimasi et al., 2004). Of note, chronic stimulation with glycolipid antigen upregulates inhibitory Ly49 receptors on the surface of iNKT cells (Hayakawa et al., 2004). Taken together, iNKT cells in peripheral tissues of α-Gal-A-deficient hosts exhibit signs of chronic antigen exposure.

In infection, DCs sense the presence of microbes by expression of a diverse set of TLRs binding to pathogen-associated molecular patterns (PAMPs) (Janeway and Medzhitov, 2002; Medzhitov, 2007; Takeda et al., 2003). TLR engagement by DCs triggers the subsequent production of self lipid antigens for induction of a predominant IFN-γ response by iNKT cells (Brigl et al., 2003; Paget et al., 2007; Salio et al., 2007). As described for salmonella infection, activation of iNKT cells by self antigens depends on IL-12 secretion by DCs (Brigl et al., 2003). Upon infection or treatment with PAMPs, we demonstrate that α-Gal-A-deficient DCs profoundly amplify antigen-specific IFN-γ production by iNKT cells. Whereas previous reports demonstrated CD1d-independent, LPS-mediated iNKT cell activation (Nagarajan and Kronenberg, 2007), we observed CD1d-restriction of iNKT cell responses induced by infected or PAMP-treated DCs. Moreover, the increase in cytokine secretion strictly depended on DC-derived IL-12. It is conceivable that the strong stimulatory capacity of Gla−/− DCs induces enhanced IFN-γ secretion by iNKT cells, which in turn promotes IL-12 production by DCs in terms of a positive feed-back loop. Recent studies suggest that TLR-activated DCs stimulate iNKT cells through presentation of newly synthesized lipids (Paget et al., 2007; Salio et al., 2007). However, inhibition of GSL neosynthesis using the β-glucosylceramide synthase blocker NB-DGJ had a minor effect on iNKT cell activation. Moreover, GSL synthesis has been blocked during DC differentiation (Salio et al., 2007), thus potentially reducing the overall GSL pool subsequently available in lysosomes. In our experiments, stimulation of iNKT cells by α-Gal-A-deficient DCs was not affected by short-term treatment with NB-DGJ. Therefore, antigen presentation of lipids relied rather on the accumulation of α-Gal-A-dependent catabolic intermediates in lysosomes than on GSL neosynthesis.

The physiological mechanism we propose for the generation of endogenous iNKT cell antigens involves the TLR-regulated inhibition of α-Gal-A. Incubation of DCs with L. monocytogenes or LPS rapidly induced α-Gal-A block. The reduction of enzymatic activity was not because of α-Gal-A downregulation, because infection or PAMP treatment of DCs rather increased α-Gal-A gene expression. Upon exposure to LPS, the enzyme activity recovered 20 hr postincubation. Thus, TLR-regulated enzyme inhibition turned out to represent a temporary α-Gal-A block. Importantly, α-Gal-A inhibition critically depended on the TLR adaptor molecule MyD88. The TLR-MyD88 axis has been described to regulate phagosome maturation (Blander and Medzhitov, 2006). It is thus tempting to speculate that MyD88 could also influence late endosomal and lysosomal function.

Furthermore, we investigated consequences of TLR signaling and α-Gal-A inhibition on the lipid content of DCs. We used fluorochrome-coupled STxB that specifically binds the α-Gal-A substrate Gb3 as lipid reporter system for GSL accumulation in DCs. Upon sensitization of DCs with PAMPs, we found that the TLR-4 and TLR-9 ligands, LPS and CpG ODN, respectively, induced GSL accumulation in a MyD88-restricted fashion. In contrast, the TLR-3 ligand polyI:C did not influence lipid abundance, which was consistent with its failure to mediate α-Gal-A inhibition. In the light of well-established evidence that TLR ligation leads to lipid antigen generation in DCs (Brigl et al., 2003; Paget et al., 2007; Salio et al., 2007), we provide an intertwining link between lysosomal enzyme inhibition and GSL abundance.

The self lipid iGb3 is a potent antigen for iNKT cells (Zhou et al., 2004; Mattner et al., 2005). However, its detection in tissues proved to be delicate. Although previous studies were not able to detect iGb3 by HPLC methods (Speak et al., 2007), more recent work demonstrated its presence in thymus and DCs using sensitive ion trap mass spectrometry (Li et al., 2008; Li et al., 2009). However, generation of iGb3 synthase-deficient mice revealed normal frequencies of iNKT cells in thymus and peripheral tissues (Porubsky et al., 2007). Although an alternative pathway of iGb3 synthesis can not be formally excluded, it is likely that other endogenous ligands exist that mediate iNKT cell selection in the thymus. Surprisingly, the studies investigating iGb3-synthase-lacking mice did not include TLR stimulation of DCs to measure self antigen-restricted iNKT cell activation (Porubsky et al., 2007). Although it has been reported that myeloid-derived suppressor cells (MDSCs) from iGb3-synthase-deficient mice stimulate iNKT cells after TLR ligation (De Santo et al., 2008), iGb3 could still represent the endogenous antigen that is induced upon TLR engagement in DCs to activate iNKT cells in peripheral tissues. Of note, it has been shown that iGb3 accumulates in Gla−/− mice (Li et al., 2009).

So far, iGb3 remains the placeholder of the veritable endogenous antigen for iNKT cells. Taking into account our data derived from the model of α-Gal-A deficiency, the iNKT cell-stimulating antigen has to be positioned upstream of α-Gal-A and downstream of Hex-B. Thus, we consider it worthwhile to analyze this biochemical layer for identification of the endogenous iNKT cell antigen. Taken together, we propose TLR-dependent block of α-Gal-A in DCs as mechanistic link between pathogen recognition and self lipid antigen induction for iNKT cells.

EXPERIMENTAL PROCEDURES

Mice

Ten to 18-week-old wild-type C57BL/6, Vα14-Jα18 transgenic, Jα18-deficient, Gla−/−, Gla−/−Cd1d1−/−, Cd1d1−/−, OT-1 TCR-transgenic, CD45.1 congenic, and Myd88−/− mice were bred under specific pathogen-free conditions. Vα14-Jα18 transgenic, Jα18-deficient, and Cd1d1−/− breeders were kindly provided by Dr. A. Bendelac, Chicago. β-Gal-deficient mice were provided by Dr. A. d’Azzo, Memphis. All mice were on C57BL/6 background and backcrossed for at least eight times. All procedures were approved by the Institutional Animal Care and Use Committee.

Antibodies and Chemicals

Monoclonal antibodies against murine CD4, CD5, CD8, NK1.1, CD25, CD44, Ly49C/I, TCR β chain, CD69, IFN-γ, IL-4, CD1d, MHC-I, MHC-II, CD80, CD86, and CD11c were purchased from BD Biosciences. The mCD1d-specific mAb K253 was a kind gift from Dr. S. Porcelli, New York. Neutralizing antibody against murine IL-12 was purchased from R&D Systems. Chemically synthesized αGalCer and isoglobotriaosylceramide (iGb3) were purchased from Enzo Life Sciences. Ultrapure LPS from Escherichia coli serotype 0111:B4, polyI:C, and type B CpG ODN 1826 were from Cayla (Toulouse, France). LPS from Salmonella enterica serotype abortus equi was from Sigma. 1-deoxygalactonojirimycin and N-butyl-deoxygalactonojirimycin were purchased from Calbiochem. N-acetyl-N-galactosamine, 4-methylumbelliferyl-α-D-galactopyranoside, MβCD, BrdU, and filipin III from Streptomyces filipinensis were purchased from Sigma Aldrich. The isolectin B4 was purchased from Vector Labs. Alexa Fluor 488-coupled recombinant Shiga Toxin B subunit was kindly provided by Dr. L. Johannes, Paris, France.

NKT Cell Hybridoma Assays

Bone marrow-derived DCs (5 × 104/well) were incubated with the murine Vα14+ iNKT cell hybridoma DN32.D3 (5 × 104/well) or with the murine Vα8+ NKT cell hybridoma TCB11 (5 × 104/well) in 96-well round-bottom plates and a final volume of 200 μl complete RPMI medium, prior to incubation at 37°C and 5% CO2 for 20 hr. Where indicated, DCs were pulsed with different concentrations of αGalCer or iGb3 for 6 or 16 hr before addition of NKT cell hybridoma. Mouse IL-2 from coculture supernatants was measured by ELISA. For iGb3 blocking experiments, DCs were incubated with IB4 at 37°C for 2 hr at the indicated concentrations and washed twice in medium before stimulation of NKT cell hybridoma. To inhibit GSL synthesis, DCs were treated with 100 μg/ml NB-DGJ for the indicated period of time prior to iNKT cell assays. NB-DGJ was also present during the stimulation experiments at 100 μg/ml. To achieve cholesterol depletion, DCs were incubated with 10 mM MβCD in complete medium for 15 min at 37°C. After extensive washing, DCs were cocultured with iNKT cell hybridoma.

Supplementary Material

Supplemental Information

Acknowledgments

We thank S. Weber for expert technical assistance. We are grateful to Dr. S. Porcelli for sharing mAb K253 with us. We thank Dr. A. d’Azzo for providing us with β-Gal-deficient mice. We thank the NIH Tetramer Facility for CD1d-tetramers. This work was supported in part by the European Union Marie Curie Actions (MEST-CT-2005-020311 to A.D.) and the German Science Foundation (SFB 421 to S.H.E.K. and F.W.).

Footnotes

SUPPLEMENTAL INFORMATION

Supplemental Information includes six figures, one table, and Supplemental Experimental Procedures and can be found with this article online at doi:10.1016/j.immuni.2010.08.003.

References

  1. Beck M. New therapeutic options for lysosomal storage disorders: enzyme replacement, small molecules and gene therapy. Hum Genet. 2007;121:1–22. doi: 10.1007/s00439-006-0280-4. [DOI] [PubMed] [Google Scholar]
  2. Bendelac A, Savage PB, Teyton L. The biology of NKT cells. Annu Rev Immunol. 2007;25:297–336. doi: 10.1146/annurev.immunol.25.022106.141711. [DOI] [PubMed] [Google Scholar]
  3. Benlagha K, Weiss A, Beavis A, Teyton L, Bendelac A. In vivo identification of glycolipid antigen-specific T cells using fluorescent CD1d tetramers. J Exp Med. 2000;191:1895–1903. doi: 10.1084/jem.191.11.1895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Blander JM, Medzhitov R. On regulation of phagosome maturation and antigen presentation. Nat Immunol. 2006;7:1029–1035. doi: 10.1038/ni1006-1029. [DOI] [PubMed] [Google Scholar]
  5. Brady RO, Gal AE, Bradley RM, Martensson E, Warshaw AL, Laster L. Enzymatic defect in Fabry’s disease. Ceramidetrihexosidase deficiency. N Engl J Med. 1967;276:1163–1167. doi: 10.1056/NEJM196705252762101. [DOI] [PubMed] [Google Scholar]
  6. Brigl M, Brenner MB. CD1: antigen presentation and T cell function. Annu Rev Immunol. 2004;22:817–890. doi: 10.1146/annurev.immunol.22.012703.104608. [DOI] [PubMed] [Google Scholar]
  7. Brigl M, Bry L, Kent SC, Gumperz JE, Brenner MB. Mechanism of CD1d-restricted natural killer T cell activation during microbial infection. Nat Immunol. 2003;4:1230–1237. doi: 10.1038/ni1002. [DOI] [PubMed] [Google Scholar]
  8. Chiu YH, Park SH, Benlagha K, Forestier C, Jayawardena-Wolf J, Savage PB, Teyton L, Bendelac A. Multiple defects in antigen presentation and T cell development by mice expressing cytoplasmic tail-truncated CD1d. Nat Immunol. 2002;3:55–60. doi: 10.1038/ni740. [DOI] [PubMed] [Google Scholar]
  9. Cui J, Shin T, Kawano T, Sato H, Kondo E, Toura I, Kaneko Y, Koseki H, Kanno M, Taniguchi M. Requirement for Valpha14 NKT cells in IL-12-mediated rejection of tumors. Science. 1997;278:1623–1626. doi: 10.1126/science.278.5343.1623. [DOI] [PubMed] [Google Scholar]
  10. De Libero G, Moran AP, Gober HJ, Rossy E, Shamshiev A, Chelnokova O, Mazorra Z, Vendetti S, Sacchi A, Prendergast MM, et al. Bacterial infections promote T cell recognition of self-glycolipids. Immunity. 2005;22:763–772. doi: 10.1016/j.immuni.2005.04.013. [DOI] [PubMed] [Google Scholar]
  11. De Santo C, Salio M, Masri SH, Lee LY, Dong T, Speak AO, Porubsky S, Booth S, Veerapen N, Besra GS, et al. Invariant NKT cells reduce the immunosuppressive activity of influenza A virus-induced myeloid-derived suppressor cells in mice and humans. J Clin Invest. 2008;118:4036–4048. doi: 10.1172/JCI36264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Desnick RJ, Schuchman EH. Enzyme replacement and enhancement therapies: lessons from lysosomal disorders. Nat Rev Genet. 2002;3:954–966. doi: 10.1038/nrg963. [DOI] [PubMed] [Google Scholar]
  13. Dhodapkar MV. Harnessing human CD1d restricted T cells for tumor immunity: progress and challenges. Front Biosci. 2009;14:796–807. doi: 10.2741/3278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dimasi N, Moretta L, Biassoni R. Structure of the Ly49 family of natural killer (NK) cell receptors and their interaction with MHC class I molecules. Immunol Res. 2004;30:95–104. doi: 10.1385/IR:30:1:095. [DOI] [PubMed] [Google Scholar]
  15. Fan JQ, Ishii S, Asano N, Suzuki Y. Accelerated transport and maturation of lysosomal alpha-galactosidase A in Fabry lymphoblasts by an enzyme inhibitor. Nat Med. 1999;5:112–115. doi: 10.1038/4801. [DOI] [PubMed] [Google Scholar]
  16. Fischer K, Scotet E, Niemeyer M, Koebernick H, Zerrahn J, Maillet S, Hurwitz R, Kursar M, Bonneville M, Kaufmann SH, Schaible UE. Mycobacterial phosphatidylinositol mannoside is a natural antigen for CD1d-restricted T cells. Proc Natl Acad Sci USA. 2004;101:10685–10690. doi: 10.1073/pnas.0403787101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gadola SD, Silk JD, Jeans A, Illarionov PA, Salio M, Besra GS, Dwek R, Butters TD, Platt FM, Cerundolo V. Impaired selection of invariant natural killer T cells in diverse mouse models of glycosphingo-lipid lysosomal storage diseases. J Exp Med. 2006;203:2293–2303. doi: 10.1084/jem.20060921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hayakawa Y, Berzins SP, Crowe NY, Godfrey DI, Smyth MJ. Antigen-induced tolerance by intrathymic modulation of self-recognizing inhibitory receptors. Nat Immunol. 2004;5:590–596. doi: 10.1038/ni1069. [DOI] [PubMed] [Google Scholar]
  19. Hong S, Wilson MT, Serizawa I, Wu L, Singh N, Naidenko OV, Miura T, Haba T, Scherer DC, Wei J, et al. The natural killer T-cell ligand alpha-galactosylceramide prevents autoimmune diabetes in non-obese diabetic mice. Nat Med. 2001;7:1052–1056. doi: 10.1038/nm0901-1052. [DOI] [PubMed] [Google Scholar]
  20. Janeway CA, Jr, Medzhitov R. Innate immune recognition. Annu Rev Immunol. 2002;20:197–216. doi: 10.1146/annurev.immunol.20.083001.084359. [DOI] [PubMed] [Google Scholar]
  21. Kawano T, Cui J, Koezuka Y, Toura I, Kaneko Y, Motoki K, Ueno H, Nakagawa R, Sato H, Kondo E, et al. CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides. Science. 1997;278:1626–1629. doi: 10.1126/science.278.5343.1626. [DOI] [PubMed] [Google Scholar]
  22. Kinjo Y, Wu D, Kim G, Xing GW, Poles MA, Ho DD, Tsuji M, Kawahara K, Wong CH, Kronenberg M. Recognition of bacterial glycosphingolipids by natural killer T cells. Nature. 2005;434:520–525. doi: 10.1038/nature03407. [DOI] [PubMed] [Google Scholar]
  23. Kinjo Y, Tupin E, Wu D, Fujio M, Garcia-Navarro R, Benhnia MR, Zajonc DM, Ben-Menachem G, Ainge GD, Painter GF, et al. Natural killer T cells recognize diacylglycerol antigens from pathogenic bacteria. Nat Immunol. 2006;7:978–986. doi: 10.1038/ni1380. [DOI] [PubMed] [Google Scholar]
  24. Kronenberg M. Toward an understanding of NKT cell biology: progress and paradoxes. Annu Rev Immunol. 2005;23:877–900. doi: 10.1146/annurev.immunol.23.021704.115742. [DOI] [PubMed] [Google Scholar]
  25. Li Y, Teneberg S, Thapa P, Bendelac A, Levery SB, Zhou D. Sensitive detection of isoglobo and globo series tetraglycosylceramides in human thymus by ion trap mass spectrometry. Glycobiology. 2008;18:158–165. doi: 10.1093/glycob/cwm129. [DOI] [PubMed] [Google Scholar]
  26. Li Y, Thapa P, Hawke D, Kondo Y, Furukawa K, Furukawa K, Hsu FF, Adlercreutz D, Weadge J, Palcic MM, Wang PG, Levery SB, Zhou D. Immunologic glycosphingolipidomics and NKT cell development in mouse thymus. J Proteome Res. 2009;8:2740–2751. doi: 10.1021/pr801040h. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Matsuda JL, Naidenko OV, Gapin L, Nakayama T, Taniguchi M, Wang CR, Koezuka Y, Kronenberg M. Tracking the response of natural killer T cells to a glycolipid antigen using CD1d tetramers. J Exp Med. 2000;192:741–754. doi: 10.1084/jem.192.5.741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Matsuda JL, Gapin L, Sidobre S, Kieper WC, Tan JT, Ceredig R, Surh CD, Kronenberg M. Homeostasis of V alpha 14i NKT cells. Nat Immunol. 2002;3:966–974. doi: 10.1038/ni837. [DOI] [PubMed] [Google Scholar]
  29. Mattner J, Debord KL, Ismail N, Goff RD, Cantu C, III, Zhou D, Saint-Mezard P, Wang V, Gao Y, Yin N, et al. Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections. Nature. 2005;434:525–529. doi: 10.1038/nature03408. [DOI] [PubMed] [Google Scholar]
  30. Medzhitov R. Recognition of microorganisms and activation of the immune response. Nature. 2007;449:819–826. doi: 10.1038/nature06246. [DOI] [PubMed] [Google Scholar]
  31. Miyamoto K, Miyake S, Yamamura T. A synthetic glycolipid prevents autoimmune encephalomyelitis by inducing TH2 bias of natural killer T cells. Nature. 2001;413:531–534. doi: 10.1038/35097097. [DOI] [PubMed] [Google Scholar]
  32. Nagarajan NA, Kronenberg M. Invariant NKT cells amplify the innate immune response to lipopolysaccharide. J Immunol. 2007;178:2706–2713. doi: 10.4049/jimmunol.178.5.2706. [DOI] [PubMed] [Google Scholar]
  33. Ohshima T, Murray GJ, Swaim WD, Longenecker G, Quirk JM, Cardarelli CO, Sugimoto Y, Pastan I, Gottesman MM, Brady RO, Kulkarni AB. alpha-Galactosidase A deficient mice: a model of Fabry disease. Proc Natl Acad Sci USA. 1997;94:2540–2544. doi: 10.1073/pnas.94.6.2540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Paget C, Mallevaey T, Speak AO, Torres D, Fontaine J, Sheehan KC, Capron M, Ryffel B, Faveeuw C, Leite de Moraes M, et al. Activation of invariant NKT cells by toll-like receptor 9-stimulated dendritic cells requires type I interferon and charged glycosphingolipids. Immunity. 2007;27:597–609. doi: 10.1016/j.immuni.2007.08.017. [DOI] [PubMed] [Google Scholar]
  35. Pina DG, Johannes L, Castanho MA. Shiga toxin B-subunit sequential binding to its natural receptor in lipid membranes. Biochim Biophys Acta. 2007;1768:628–636. doi: 10.1016/j.bbamem.2006.12.011. [DOI] [PubMed] [Google Scholar]
  36. Plati T, Visigalli I, Capotondo A, Buono M, Naldini L, Pia CM, Biffi A. Development and maturation of invariant natural killer T cells in the presence of lysosomal engulfment. Eur J Immunol. 2009;39:2748–2754. doi: 10.1002/eji.200939639. [DOI] [PubMed] [Google Scholar]
  37. Porubsky S, Speak AO, Luckow B, Cerundolo V, Platt FM, Grone HJ. Normal development and function of invariant natural killer T cells in mice with isoglobotrihexosylceramide (iGb3) deficiency. Proc Natl Acad Sci USA. 2007;104:5977–5982. doi: 10.1073/pnas.0611139104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Prigozy TI, Naidenko O, Qasba P, Elewaut D, Brossay L, Khurana A, Natori T, Koezuka Y, Kulkarni A, Kronenberg M. Glycolipid antigen processing for presentation by CD1d molecules. Science. 2001;291:664–667. doi: 10.1126/science.291.5504.664. [DOI] [PubMed] [Google Scholar]
  39. Salio M, Speak AO, Shepherd D, Polzella P, Illarionov PA, Veerapen N, Besra GS, Platt FM, Cerundolo V. Modulation of human natural killer T cell ligands on TLR-mediated antigen-presenting cell activation. Proc Natl Acad Sci USA. 2007;104:20490–20495. doi: 10.1073/pnas.0710145104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Schumann J, Facciotti F, Panza L, Michieletti M, Compostella F, Collmann A, Mori L, De Libero G. Differential alteration of lipid antigen presentation to NKT cells due to imbalances in lipid metabolism. Eur J Immunol. 2007;37:1431–1441. doi: 10.1002/eji.200737160. [DOI] [PubMed] [Google Scholar]
  41. Sharif S, Arreaza GA, Zucker P, Mi QS, Sondhi J, Naidenko OV, Kronenberg M, Koezuka Y, Delovitch TL, Gombert JM, et al. Activation of natural killer T cells by alpha-galactosylceramide treatment prevents the onset and recurrence of autoimmune Type 1 diabetes. Nat Med. 2001;7:1057–1062. doi: 10.1038/nm0901-1057. [DOI] [PubMed] [Google Scholar]
  42. Shi FD, Van Kaer L. Reciprocal regulation between natural killer cells and autoreactive T cells. Nat Rev Immunol. 2006;6:751–760. doi: 10.1038/nri1935. [DOI] [PubMed] [Google Scholar]
  43. Speak AO, Salio M, Neville DC, Fontaine J, Priestman DA, Platt N, Heare T, Butters TD, Dwek RA, Trottein F, et al. Implications for invariant natural killer T cell ligands due to the restricted presence of isoglobo-trihexosylceramide in mammals. Proc Natl Acad Sci USA. 2007;104:5971–5976. doi: 10.1073/pnas.0607285104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stanic AK, De Silva AD, Park JJ, Sriram V, Ichikawa S, Hirabyashi Y, Hayakawa K, Van KL, Brutkiewicz RR, Joyce S. Defective presentation of the CD1d1-restricted natural Va14Ja18 NKT lymphocyte antigen caused by beta-D-glucosylceramide synthase deficiency. Proc Natl Acad Sci USA. 2003;100:1849–1854. doi: 10.1073/pnas.0430327100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Takeda K, Kaisho T, Akira S. Toll-like receptors. Annu Rev Immunol. 2003;21:335–376. doi: 10.1146/annurev.immunol.21.120601.141126. [DOI] [PubMed] [Google Scholar]
  46. Tupin E, Kinjo Y, Kronenberg M. The unique role of natural killer T cells in the response to microorganisms. Nat Rev Microbiol. 2007;5:405–417. doi: 10.1038/nrmicro1657. [DOI] [PubMed] [Google Scholar]
  47. Zhou D, Mattner J, Cantu C, III, Schrantz N, Yin N, Gao Y, Sagiv Y, Hudspeth K, Wu YP, Yamashita T, et al. Lysosomal glycosphingo-lipid recognition by NKT cells. Science. 2004;306:1786–1789. doi: 10.1126/science.1103440. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Information
NIHMS575853-supplement-Supplemental_Information.pdf (395.9KB, pdf)

RESOURCES