Alpha versus beta: are we on the way to resolve the mystery as to which is the endogenous ligand for natural killer T cells?

Abstract

Natural killer T (NKT) lymphocytes are a unique subset of cells that play a role in regulating the immune system. For the past decade, studies have focused upon attempts to define these cells and to determine the ligand(s) that are required for their development and peripheral activation. Many research groups have focused upon determining the mechanisms for activating or inhibiting NKT cells in an attempt to control immune-mediated disorders as well as infectious and malignant conditions by using different ligand structures. Alpha-anomeric glycolipids and phospholipids derived from mammalian, bacterial, protozoan and plant species have been suggested as potential ligands for these lymphocytes. Some of these ligands were structured in forms that can bind to CD1d molecules. The lack of alpha-anomeric glycosphingolipids in mammals and the modest effect of these ligands in human studies, along with recent data from animal models and humans on the NKT-dependent immunomodulatory effect of beta-glycosphingolipids, suggest that the beta-anomeric ligands have the potential to be the endogenous NKT ligand.

Natural killer T (NKT) cells

NKT cells are a unique T lymphocyte sublineage that have been implicated in the regulation of immune responses associated with a broad range of diseases, including autoimmunity, infectious diseases and cancer [1,2]. Given that low NKT cell numbers are associated with many different disease states, it is crucial to understand how they develop and become peripherally activated [2]. NKT cells differentiate from thymic precursors via signals emanating during T cell receptor (TCR) engagement by CD1d-expressing thymocytes [3]. These cells are reactive to the non-classical class I antigen-presenting molecule CD1d and recognize glycosphingolipid antigens, rather than peptides. The differentiation pathway of invariant natural killer T (iNKT) cells selected positively by CD1d molecules branches off from the pathway of mainstream thymocyte development at the double-positive (CD4⁺CD8⁺) stage [4]. It is, however, unclear whether a single ligand induces their peripheral activation. The exact definition of NKT cells is still subject to debate, and much of the misunderstanding arises from the different methods used for their isolation [5].

NKT cells play a role in the interface between innate and adaptive immunity, providing a powerful model to study the structural biology of glycolipid trafficking, processing and recognition [3]. They can either up- or down-regulate immune responses by promoting the secretion of T helper type 1 (Th1), Th2 or immune regulatory cytokines. They express a unique ability to suppress or enhance immunity, turning their manipulation into an attractive target for immunotherapy [1]. NKT cells also play a role in the differentiation of T regulatory (T_reg) cells and in the induction of peripheral tolerance. These lymphocytes induce maturation of dendritic cells (DCs), promoting them to induce an antigen-specific T and B cell response [6]. Their function in anterior chamber-associated immune deviation, oral tolerance, other tolerance systems and autoimmune diseases has been decribed [7–10].

CD1d proteins

NKT cells recognize glycosphingolipids presented by the non-classical major histocompatibility complex (MHC) molecule CD1d [11,12]. Antigens presented by CD1d have profound effects on the immune regulation of autoimmune, infectious and malignant disorders [13–15]. The CD1 family has a non-polar, hydrophobic antigen-binding groove that presents lipid antigens [16]. In humans, the CD1 family consists of the group I proteins, CDla, CDlb, CDlc and CDle, and the group II protein, CDld. Although rodents express only CDld, this single CD1 family member is able to acquire antigens in many subcellular compartments, as it is expressed broadly and traffics to all endosomal compartments [17]. While group I CD1 expression is limited to thymocytes and professional antigen-presenting cells (APCs), CD1d has a wider tissue distribution and can be found on many non-haematopoietic cells [11]. DCs and Vα14 iNKT cells recognize CD1d-associated glycosphingolipids (GSLs) via a semi-invariant TCR composed of an invariant Vα14-Jα18 chain that is paired preferentially with a restricted set of TCRβ chains [18].

The function of NKT cells is, therefore, dependent upon CD1d-restricted self and microbial antigen presentation and upon recognition of such antigens by an invariant T cell receptor on iNKT cells [19]. Analysis of the crystal structures and binding characteristics for an iNKT TCR with two CD1d-alpha-GalCer-specific Vbeta11⁺ TCRs that use different TCR Valpha chains showed the versatility of the TCR platform [20]. The structural and functional aspects of lipid presentation by CD1 molecules are therefore important in the context of the function of CD1-restricted T cells in anti-microbial responses, anti-tumour immunity and the regulation of tolerance and autoimmunity [21]. CD1d tetramers are used to monitor iNKT cell frequency, phenotype and lipid presentation by CD1d molecules [6]. This type of recognition is important for the cross-talk between NKT cells and DCs and for T_reg promotion [6].

NKT plasticity

NKT cells have been shown to be remarkably versatile in function during various immune responses [22]. This subset of lymphocytes has diverse influences in various disease models and a unique ability to suppress or enhance immunity in different microenvironments [1]. The term ‘plasticity’ is sometimes used to describe their function. NKT plasticity may require a duality in function, such as a capability for both Th1 and Th2 cytokine secretion in different immune backgrounds. Whether or not this plasticity is mediated by a ligand still remains unclear. NKT plasticity may evolve from the use of different ligands or from different signals in the immune microenvironment [23]. The same ligand can generate different types of immune responses by NKT cells in different microenvironments. In light of the differing effects of a given ligand in vivo and in vitro, the net effect of NKT activation may not result from the binding of a single ligand, but rather from the sum of the effects of a variety of mediators [3,17]. It may also depend on cell–cell interaction or on the influence of cytokines or co-stimulatory molecules. The DC–NKT and/or NKT–T_regs cross-talk may be of importance in determining NKT plasticity [7,24,25]. Organ-specific factors also play a role in NKT plasticity, with different responses generated in different organs by an identical stimulus. Antigen presentation and APCs may play an important role under these conditions [26]. Originally identical stimuli may reach NKT cells via different pathways through presentation by different APCs [1,27]. On the other hand, plasticity may result from the natural programmes of different subsets of cells [28]. NKT cells are a heterogeneous population of lymphocytes that can differ in their CD1d reactivity and CD expression. Apart from inherent heterogeneity between different NKT populations, alterations in cellular membranes with altered lipid raft properties can affect raft-bound receptors, such as CD1d, and may add to the variety of responses [29,30]. Distinct NKT cell subsets have been suggested to play positive and negative regulatory roles and to define a new immunoregulatory axis, with broad implications for tumour immunity and other immunological and disease settings [31]. For NKT cells, it is unclear as yet whether their extensive functional capacities can be attributed to a single population that is sensitive to environmental cues or if functionally distinct NKT cell subpopulations exist [22]. Thus, NKT plasticity may be considered the result of several of the above-mentioned factors, with CD1d-dependent ligands being the final link in a chain of factors that determines the end response [30,32].

NKT ligands

The ligands for iNKT cells have been the subject of much research over the last decade. A number of ligands that can be presented by CD1d to NKT or other CD1d-restricted T cells have been identified, including glycolipids from a marine sponge, bacterial glycolipids, normal endogenous glycolipids, tumour-derived phospholipids and glycolipids and non-lipidic molecules [13]. Many of these agonists are being used as therapeutic agents to activate NKT cells in animal models and in early studies in cancer patients [33]. For peripheral NKT activation multiple classes of self-derived lipids, in addition to pathogen-derived lipids, appear to play a role [34].

The identification of physiologically relevant candidate ligands for positive selection or activation has proved technically challenging, however, due largely to the fact that the ligands for iNKT cells are lipids [34]. Studies have suggested that NKT cells die within hours of stimulation [12]. These results imply that their direct impact on the immune system is associated with an initial cytokine burst released before their death [12].

Ligand structures relate to their propensity to bind CD1d molecules and, as a consequence, alter the function of NKT cells [35]. To optimize NKT-based immunotherapy strategies, structural and kinetic analyses lead to the design of optimal NKT cell agonists [24].

Alpha-galactosylceramide (alpha-GalCer)

The alpha-anomericity of the carbohydrate is considered to be an important requisite for the CD1d-specific activation of NKT cells [36]. Characterization of NKT cells over recent years has been based upon their reactivity to the synthetic glycolipid alpha-GalCer in a CD1d-dependent manner [36]. Administration of alpha-GalCer is associated with potent activation of NKT cells, rapid and robust cytokine production and activation of a variety of cells of the innate and adaptive immune systems [37]. Administration of alpha-GalCer induces the secretion of both interleukin (IL)-4 and interferon (IFN)-γ. Repeated administration favours the production of Th2 cytokines [38,39]. In some mouse models of autoimmunity, NKT cell deficiency exacerbates disease, suggesting that NKT cells play a role in suppressing autoimmunity. Conversely, specific activation of NKT cells with alpha-GalCer generally protects mice against the development of autoimmunity [40]. Alpha-GalCer is generally more beneficial for treatment of Th1-mediated autoimmune diseases. However, alpha-GalCer treatment in mice was associated with detrimental side effects, and treatment efficacy was influenced by a variety of parameters, resulting occasionally in disease exacerbation rather than protection. Alpha-GalCer has also been shown to be hepatotoxic in mice, limiting its use in human testing [41].

Other synthetic glycosphingolipids

The biological activities of glycosphingolipids are highly variable, with minor changes in molecular structure. Synthesized glycosphingolipids provide insight into the various functional pathways associated with receptor binding of these ligands [42]. Sphingolipid analogues can be used in various systems: as anti-cancer agents, as probes of protein targets of bioactive lipids and of glycosphingolipid distribution in bilayers, as modulators of cholesterol-enriched microdomains that facilitate fusion of alphaviruses with membranes, as enhancers of membrane permeabilization induced by cholesterol-specific cytolysins and as probes for the selective internalization of glycosphingolipids in the caveolae of living mammalian cells [30,32,42].

Because NKT cells are attractive targets for the treatment of human autoimmune diseases, recent efforts have focused upon developing NKT cell agonists with superior treatment efficacy over alpha-GalCer [40]. OCH is the sphingosine-truncated derivative of alpha-GalCer and stimulates NKT cells to produce Th2 cytokines selectively. Administration of OCH suppressed autoimmune diseases, such as experimental autoimmune encephalomyelitis, diabetes in non-obese diabetic (NOD) mice and collagen-induced arthritis, by inducing Th2-biased T cells [38]. KRN 7000, an analogue with a di-unsaturated 20-carbon chain (C20:2), stimulates IL-4 production yet inhibits IFN-γ[43]. α-C-GalCer, a C-glycoside analogue, was reported to generate a longer-lasting reaction with a cleaner Th1 response [44,45]. A recent study evaluated 16 analogues of α-GalCer for the CD1-mediated TCR activation of naive human NKT cells and their anti-cancer efficacy [46].

Alterations in the degree of saturation, length of the lipid chain and sugar moieties are relevant for the GSL function. One example is the trans 4,5-unsaturation of the sphingosine backbone, which promotes closer packing and lower compressibility of ceramide analogues in the lipid–water interface relative to comparable saturated compounds[47]. Another is variation in the length of the long-chain base and in the structure of the carbohydrate-containing polar head group of (2S, 3R) (or D-erythro-)-β-lactosylceramide (LacCer), which may affect the mechanism of endocytic uptake from the plasma membrane [48].

Exogenous natural ligands

Glycosphingolipids and phospholipids with alpha-anomeric sugars attached to the lipid chain that are derived from mammalian, bacterial, protozoan and plant species have been identified as potential natural ligands for NKT cells [30,35,49]. NKT cells are activated during an infectious assault by the presentation of a neo-self glycolipid by the release of proinflammatory or anti-inflammatory cytokines by APCs, which can jump-start the immune system [30,50]. A dual recognition of self and microbial ligands was suggested to underline innate-like, anti-microbial functions that are mediated by the CD40L induction that is associated with massive cytokine release by NKT cells [47,51]. The Gram-negative, lipopolysaccharide-free bacterium Sphingomonas paucimobilis contains GSL [36]. Soluble CD1d-Ig dimers loaded with this lipid extract bind specifically to NKT cells, and the GSLs extracted from this bacterium are able to stimulate NKT cells in a CD1d-specific manner.

The semi-invariant αβTCRs can recognize isoglobotrihexosylceramide (iGb3), which is a mammalian glycosphingolipid, and a microbial α-glucuronylceramide, which is found in the cell walls of Gram-negative, lipopolysaccharide (LPS)-negative bacteria [52]. iGb3 is recognized by NKT cells under pathological conditions such as cancer and autoimmune disease [52,53]. Studies in mice that were deficient for iGb3 synthase revealed that these mice developed and reproduced normally. In addition, these mice had normal numbers of NKT cells in the thymus, spleen and liver [54]. Upon administration of α-galactosylceramide, activation of NKT cells and DCs was similar in iGb3S^–/– and iGb3S^+/– mice, which suggested that iGb3 is unlikely to be an endogenous CD1d lipid ligand that determines thymic NKT selection.

Several potential natural glycosphingolipids have been suggested to activate NKT cells. Potential activators include glycosphingolipids, such as glycosylphosphatidylinositol [13,30], isoglobotirhexosylceramide [55] and α-glucuronsylceramide [56,57], and phospholipids such as phosphatidylcholine and phosphatidylinositol [58,59]. Not all are stimulatory, however, and some of these may even exert an inhibitory signal to NKT cells.

Beta-structured GSLs

There are few examples in nature in which the human body has receptors for antigens or ligands that do not exist in the body. Thus far, α-anomeric D-glycosylceramides have not been detected in mammals.

β-glucosylceramide (β-GC) is a naturally occurring glycolipid that is a metabolic intermediate in anabolic and catabolic glycosphingolipid pathways [33]. Its synthesis from ceramide is catalyzed by glucosylceramide synthase [30,32]. Beta-anomeric-structured glycosphingolipids are normal constituents of cell membranes [60,61]. In view of the lack of α-structured GSLs in mammals and the lack of a profound effect of alpha-anomeric compounds when tested in humans with cancers, recent studies have suggested that endogenous β-structured glycosphingolipids may be the potential endogenous ligands for NKT cells [32].

Gaucher's disease [62] is the most common glycolipid storage disorder, and is caused by the reduced activity of the lysosomal enzyme glucocerebrosidase. In this disease the substrate, β-glucosylceramide, accumulates in the cells of the reticuloendothelial system [63]. The decreased activity of glucosylceramide synthase also results in elevated serum β-GC levels [63]. Patients with Gaucher's disease have altered humoral and cellular immune profiles due, in part, to changes in cellular membrane properties [30]. There is circumstantial evidence indicating β-GC involvement in NKT cell regulation. Studies of patients with Gaucher's disease from Israel and Portugal showed an increase of NKT cells in the peripheral blood of these patients [64,65].

In vitro, treatment of NKT cells with β-GC decreases cell proliferation in the presence of DCs [66], and CD1d-bound β-GC inhibits NKT cell activation by α-GalCer [67]. Following its binding to CD1d, β-GC exerts a direct effect on NKT cells. It can also inhibit the α-GalCer-mediated activation [67]. Glucosylceramide synthase deficiency may also be associated with altered ligand presentation by CD1d [30].

β-galactosylceramide (β-GalCer)-deficient mice exhibit normal NKT development and function, and cells from these animals stimulate NKT hybridomas [68]. These hybridomas do not react to CD1d1 expressed by a β-GC-deficient cell line. Human β-GC synthase cDNA transfer restores recognition of the mutant cells expressing CD1d1 by Vα14Jα18 NKT hybridomas. Although suppression of β-GC synthesis inhibits antigen presentation to NKT cells, β-GC does not activate NKT hybridomas [68]. β-D-GalCer (C12) efficiently diminished the number of detectable NKT cells in vivo without inducing cytokine expression. Binding studies have demonstrated that both α-GalCer and β-D-GalCer are equally efficient in reducing the number of NKT cells [67].

β-glycosphingolipids exert a beneficial effect in several models of immune-mediated disorders and cancers, and their effect was reviewed recently in detail [69]. β-GC alleviates immunologically incongruous disorders and may be associated with the ‘fine tuning’ of the immune response by changes in the plasticity of NKT lymphocytes. β-GC alleviated concanavalin A (ConA)–NKT-mediated immune hepatitis in mice, which is an effect associated with a decreased serum level of IFN-γ and a reduced expression of the transcription factor signal transducer and activator of transcription-1 (STAT-1) [66]. The beneficial effect of β-GC was associated with a decrease in the number of intrahepatic NKT cells. β-GC generated a Th2 response that was associated with the alleviation of colitis in a murine Th1-mediated colitis model [70]. A significant amelioration of hepatic fibrosis was observed in β-GC-treated mice [71]. In contrast, in hepatocellular carcinoma-harbouring mice, β-GC resulted in a Th1 immune shift that was associated with the suppression of tumour growth and improved survival [70], which further supported the hypothesis that β-GC can alter the plasticity of NKT cells.

β-GC ameliorated graft-versus-host disease (GVHD) in both Th1- and Th2-mediated murine models and affected the immune systems differentially in these incongruous models [27]. The beneficial effect of β-GC was associated with an increase in the intrahepatic : peripheral NKT lymphocyte ratio in the semi-allogeneic Th1-mediated acute model. This ratio, however, was decreased in the chronic Th2-mediated GVHD model. The administration of β-GC led to a decreased serum level of IFN-γ and an increased serum level of IL-4 in the Th1-mediated model, whereas this was reversed in the Th2-mediated model.

β-GC resulted in a significant amelioration in animal models for type 2 diabetes and non-alcoholic fatty liver disease [72–75]. A significant amelioration of the metabolic changes that are characteristic of leptin-deficient mice was observed: the liver size and hepatic fat content were decreased, glucose tolerance was nearly normalized and serum triglyceride levels were decreased. β-GC reduced pancreatic and liver steatosis and stimulated insulin secretion in the Cohen diabetes-sensitive rat, which is a lean model of non-insulin-resistant, nutritionally induced diabetes. β-GC increased the immune response against the hepatitis B virus in association with an altered distribution of NKT and CD8 cells, suggesting that β-GC could be used as a potent adjuvant for overcoming non-responsiveness to the hepatitis B virus (HBV) vaccine, which may be NKT-dependent [76].

Several potential mechanisms of the NKT-dependent, immunomodulatory effect of β-glycosphingolipids were reviewed recently [69]. β-GSLs seem to affect NKT cells differently than does α-GalCer. In vitro, β-GSLs inhibit NKT cell proliferation without stimulating cytokine expression [66], and in vivo, β-GSLs exert an opposing effect on NKT cells in different microenvironments. Studies have shown that the immunomodulatory effect of β-GC is dependent upon DCs [24,66,77] and that β-GSLs can promote T_regs. This effect can be mediated via DCs or by direct cross-talk between NKT cells and T_regs[30,78]. The immunomodulatory effect of glucosylceramide can be associated with its role as an active ligand for NKT cells in a CD1d-dependent manner or via the displacement of an activating ligand from the MHC class I-like molecule, CD1d. Recent studies in animal models and in humans have shown that GC can promote T_regs via activation of DCs (Y. Ilan, unpublished data). β-GSLs may also affect NKT cells by the alteration of lipid rafts on cell membranes [79]. CD1d is localized to lipid rafts, and disruption of these lipid rafts can inhibit NKT activation without impairing CD1d-ligand binding [80]. Raft disruption has been shown to inhibit IL-6/STAT3 and IFN-γ/STAT1 signalling [81]. The administration of naturally occurring β-GSLs can alter lipid raft composition and structure, thereby affecting the intracellular signalling machinery [50,67].

Supporting human studies

Enzyme replacement therapy (ERT) for patients with GD has proved effective in the management of the key disease manifestations, including a reduction in organomegaly, improvements in the blood counts and biochemical parameters, a decrease in bone-related pain and compensatory growth in children [82–87]. An increasing number of complications that are not associated directly with the known consequences of GD and that, paradoxically, may be related to the long-term use of ERT, have been reported.

The life expectancy of patients with GD type 1 (GD1) was evaluated by comparing the survival data from patients who were enrolled in the International Gaucher Registry (ICGG). The overall estimated life expectancy at birth for patients with GD1 was 9 years fewer than that estimated for the reference population [88]. The causes of death, however, could not be attributed to the glycolipid storage deficiency or to complications of GD that were seen typically in the pre-ERT era, such as bleeding or infections [89]. These observations have led to the notion that there is a reduction in the levels of β-GC that is caused by high dosages of ERT which, in turn, may have detrimental effects [90].

GD1 is associated with high resting energy expenditure, a low circulating adiponectin level and peripheral insulin resistance [91]. ERT may be associated with a decrease in resting energy expenditure. The prevalence of overweight individuals is lower in untreated GD patients than in the general population [91]. The prevalence of type 2 diabetes increases significantly with ERT, resulting in a comparable prevalence of type 2 diabetes among patients receiving ERT and the general population [91]. Long-term treatment with ERT results in a larger-than-average weight gain. The data suggest that a decrease in the serum level of GC has a deleterious effect on metabolic pathways that are associated with insulin resistance. Two recent reviews published in light of these data described the potent immunomodulatory, anti-inflammatory and anti-malignant effects of GC and suggested that the long-term suppression of GC by ERT may impact the outcome of some patients with GD [69,90].

Based on the preclinical data suggesting that GC improves insulin resistance, hepatic steatosis and metabolic syndrome in animal models of diabetes and non-alcoholic steatohepatitis (NASH) [73,74], the safety and efficacy of oral GC were studied recently in a clinical trial [92]. In an open-label phase I/II study, patients with liver biopsy-proven NASH were treated with oral GC for 24 weeks and followed for 16 additional weeks. No treatment-related adverse events were noted. Oral administration of GC led to reduced aspartate transaminase (AST), alanine aminotransferase (ALT) and gamma glutamyl transferase levels and to improved histological findings in liver biopsies of some of the patients. Serum haemoglobin A1C, the glucose tolerance test, triglycerides and low-density lipoprotein (LDL) cholesterol also improved in some of the treated patients. The liver fat content, as assessed by magnetic resonance imaging (MRI), was reduced [69]. An increase in the percentage of peripheral blood NKT cells was noted in some of the treated patients. The results of this preliminary study indicate that β-GC may be a promising treatment modality for patients with NKT-associated disorders [69].

Optimization of β-glycosphingolipids as NKT ligands: are there several endogenous ligands?

The above preclinical and clinical data, although implying that β-GC may be a natural endogenous ligand for NKT cells, does not rule out the option of other β-GSLs as additional ligands. Each of the different natural β-GSLs may be associated with different signals to NKT cells, and each may exert a different effect in different immune microenvironments. In fact, part of the plasticity of NKT cells described above may be related to the potential role of different natural ligands, β-GSLs, to alter NKT function in different directions. Alternatively, NKT plasticity may involve different subsets of cells being activated by different ligands from this family. Each of these cells may play a role in different diseases.

Recent studies determined methods to optimize the use of β-GSLs as ligands for NKT cells by using different natural β-GSLs (e.g. β-lactosylceramide), using different ratios of combinations of two natural β-GSLs (e.g. β-GC with β-lactosylceramide), and by altering the structure of the glucose moiety, the length of the lipid chain and the number of double bonds in the β-GC molecule [93]. The β-anomeric structure of the glycosphingolipid molecule, along with alteration of the carbon number in the acyl chain and the use of several combinations of β-analogues in different doses, was associated with an altered effect on lipid raft structure. This alteration was then followed by downstream effects on STAT signalling, which resulted in beneficial effects in models of NKT cell-dependent, immune-mediated, metabolic or malignant disorders [93].

A reduction of glycosphingolipids by the pharmacological inhibition of glucosylceramide synthase with atrial natriuretic peptide–dominant negative mutant (AMP–DNM) [94–97] or GENZ-123346 [98] was shown to improve insulin signalling and insulin resistance. Although this observation seems to be in contrast with part of the results presented above, it supports further the concept that β-GSLs are a family of immunomodulatory molecules. Blockage of an enzyme in the β-GSL pathway will decrease one compound and increase the levels of the other compounds in the chain, each of which may have a different effect on NKT cells. The glycosphingolipid composition of the cell membrane has an effect on the insulin receptor and the downstream signalling pathway. Alteration in insulin receptors by GSLs occurs via natural up-regulation of their biosynthesis. Glycosphingolipids modulate the activity of the insulin receptor negatively [99]. In obesity-driven insulin resistance, several factors contribute to the presence of excess GSLs in plasma membranes. Inhibition of the synthesis pathway for glycosphingolipids under these conditions restores the signalling capacity of the insulin receptor [29,100]. This effect may or may not be associated with a direct or indirect immunomodulatory effect of β-GC on NKT cells. Overall, the data support the hypothesis that alteration of β-GSLs exerts an effect on metabolic pathways, some of which may be mediated at least partially by NKT cells.

Summary

NKT cells remain enigmatic as to their identity, as do the ligands required for their development and peripheral activation. Much of the data relating to NKT cells over the last decade have been derived from the use of alpha-anomeric GSLs, which have been either natural or synthesized. The lack of these compounds in mammals, the fact that these compounds have only a modest effect in clinical trials, and recent data on β-GSLs generated in both animal models and humans suggest that beta-anomeric ligands may be the endogenous natural ligands for NKT cells. Approaches that use these compounds solely or in combination, or that introduce rationally designed β-glycosphingolipids to interfere with glycolipid metabolism, are attractive means for the development of NKT cell-based immune therapies.

Disclosure

YI receives consultancy fees from ENZO Therapeutics; Chiasma Pharmaceuticals; Member of the Scientific Advisory Board of Orammune.