Regulation of Interferon-γ

Regulation of Interferon-γ During Innate and Adaptive Immune Responses- 1. INTRODUCTION

슬로싱킹 2024. 12. 31. 21:17

 

Regulation of Interferon-γ.pdf
0.67MB

 

 

1. INTRODUCTION

The canonical Th1 cytokine, interferon-γ (IFN-γ), is critical for innate and adaptive immunity against viral and intracellular bacterial infections. In humans, genetic deficiencies in the interleukin (IL)-12/IL-23/IFN-γ pathways that result in decreased IFN-γ induction or signaling are associated with strikingly increased susceptibility to mycobacterial infections (Filipe-Santos et al., 2006).

 

Susceptibility to what are normally weakly pathogenic mycobacterial strains is greatly increased in such patients, whereas susceptibility to the more pathogenic mycobacteria that cause leprosy and tuberculosis has been observed less frequently and primarily in individuals with incomplete loss-of-function mutations in these pathways (Casanova and Abel, 2002). Systemic infections with Salmonella are also more common (de Jong et al., 1998), but, unlike the risk for mycobacterial infection, are most often observed in those with defects in IL-12/IL-23 production or signaling rather than IFN-γ signaling; this difference suggests that risk for Salmonella infection may result both from a defect in IFN-γ production and the production of other cytokines, like IL-17 (MacLennan et al., 2004).

 

IFN- γ is also involved in tumor control (Ikeda et al., 2002; Rosenzweig and Holland, 2005). IFN-g directly enhances the immunogenicity of tumor cells and stimulates the immune response against transformed cells. Human tumors can evade this form of control by becoming unresponsive to IFN- γ (Kaplan et al., 1998). Mice with targeted genetic deficiencies resulting in the loss of IFN- γ induction, production, or responsiveness are highly susceptible to infections due to intracellular bacteria, including mycobacteria, Salmonella (John et al., 2002), Listeria (Harty and Bevan, 1995), intracellular protozoans (including Toxoplasma and Leishmania), and certain viruses (Dalton et al., 1993; Huang et al., 1993; Jouanguy et al., 1999). Such mice also display a greater range, number, and aggressiveness of naturally occurring and induced tumors (Kaplan et al., 1998).

 

The importance of IFN- γ in the immune system stems in part from its ability to inhibit viral replication directly, but most importantly derives from its immunostimulatory and immunomodulatory effects.

 

IFN- γ , either directly or indirectly, upregulates both major histocompatibility complex (MHC) class I and class II antigen presentation by increasing expression of subunits of MHC class I and II molecules, TAP1/2, invariant chain, and the expression and activity of the proteasome. IFN-γ also contributes to macrophage activation by increasing phagocytosis and priming the production of proinflammatory cytokines and potent antimicrobials, including superoxide radicals, nitric oxide, and hydrogen peroxide (Boehm et al., 1997).

 

As described below, IFN-γ also controls the differentiation of naive CD4 T cells into Th1 effectors, which mediate cellular immunity against viral and intracellular bacterial infections.

 

Although necessary for clearing many types of infections, excess IFN-γ has been associated with a pathogenic role in chronic autoimmune and autoinflammatory diseases, including inflammatory bowel disease, multiple sclerosis, and diabetes mellitus (Bouma and Strober, 2003; Neurath et al., 2002; Skurkovich and Skurkovich, 2003).

 

IFN-γ enhances lymphocyte recruitment and prolonged activation in tissues (Hill and Sarvetnick, 2002; Savinov et al., 2001). Thus, the induction, duration, and amount of IFN-γ produced must be closely controlled and delicately balanced for optimum host wellness.

 

The primary sources of IFN-γ are natural killer (NK) cells and natural killer T (NKT) cells, which are effectors of the innate immune response, and CD8 and CD4 Th1 effector T cells of the adaptive immune system. NK and NKT cells constitutively express IFN-γ mRNA, which allows for rapid induction and secretion of IFN-γ on infection.

 

In contrast to NK and NKT cells, naive CD4 and CD8 T cells produce little IFN-γ immediately following their initial activation.

 

However, naive CD4 and CD8 T cells can gain the ability to efficiently transcribe the gene encoding IFN-γ (IFNG in humans and Ifng in mice) over several days in a process that is dependent on their proliferation, differentiation, upregulation of IFN-γ-promoting transcription factors, and remodeling of chromatin within the Ifn γ locus.

 

Naive CD8 T cells are programmed to differentiate into IFN-γ-producing cytotoxic effectors by default, whereas CD4 T cells can differentiate into a number of effector lineages, of which only Th1 CD4 effector T cells produce substantial amounts of IFN-γ.

 

The process of effector differentiation in CD4 T cells, and to a lesser extent in CD8 T cells, is influenced by the nature of the infecting pathogen and the cytokine milieu emanating from the innate immune system in response to the pathogen. These differences in priming conditions in turn can result in stable changes to the chromatin structure of the gene encoding IFN-γ, either facilitating high-level expression in Th1 CD4 and CD8 effector T cells or silencing expression in other effector lineages.

 

2. IFN-γ-PRODUCING CELLS

2.1. NK cells

NK cells are a key component of the innate immune system providing early cellular defense against viruses and other intracellular pathogens, and contributing to the early detection and destruction of transformed host cells. NK cells develop in the bone marrow from a common lymphoid progenitor that also gives rise to B and T cells.

 

NK cells express inhibitory receptors that recognize MHC class I molecules, which are expressed by all nucleated cells, serve as a marker for ‘‘self,’’ and inhibit NK cell activation.

 

Because antigen presentation via MHC class I is key to the CD8 T cell response, pathogens frequently downregulate MHC class I in an attempt to elude elimination by the CD8 T cell response; transformed cells often have reduced or absent MHC class I as well. Loss of MHC class I, in combination with binding of activating receptors to ligands on infected or transformed cells, leads to NK cell activation. Such activation causes NK cells to release cytotoxic granules containing perforin and granzymes, which induce programmed cell death in the target cell.

 

Activated NK cells are also the primary rapid and potent producers of IFN-γ during the early innate immune response, which contributes directly to the innate immune response and also shapes the type and quality of the adaptive immune response that is subsequently elicited.

NK cells transcribe Ifn-γ  at the earliest stages of NK cell development in the bone marrow. Ifn- γ transcription increases during NK maturation and is rapidly and potently induced by NK activation, which also triggers translation and secretion of IFN-γ (Stetson et al., 2003).

 

 

2.2. NKT cells

NKT cells share characteristics of both NK and T cell lineages in that they express several activating and inhibitory receptors of the NK repertoire, as well as a rearranged T cell receptor (TCR) in association with the CD3 signaling complex.

 

Many NKT cells express the CD4 coreceptor and surface markers associated with an activated or memory T cell phenotype: CD44hi, CD62Llo, CD69þ, and IL-25hi. The largest and best-characterized population of NKT cells includes those expressing an invariant T cell receptor (TCR) consisting of a rearranged Va14/Ja18 in the mouse and Va24/Ja18 in humans.

 

Invariant NKT cells recognize lipid antigens in the context of the nonclassical MHC class I molecule CD1d. These cells are derived from CD4þ CD8þ double-positive thymocytes (Egawa et al., 2005) that have been positively selected in the thymus by endogenous CD1d molecules presenting host-derived lipids (Gapin et al., 2001; Zhou et al., 2004a).

 

Mature invariant NKT cells are strongly reactive to the marine sponge-derived glycolipid a-galactosyl ceramide (a-GalCer) presented by CD1d, and as such a-GalCer has been used as an immunostimulatory molecule in mouse tumor and infection models (Gansert et al., 2003; Kawano et al., 1997).

 

NKT cells can be positively or negatively regulated by an array of cellular and microbial lipids presented by CD1d (Brutkiewicz, 2006), including phospholipids from Leishmania (Amprey et al., 2004), mycobacterial lipids (Fischer et al., 2004), and lysosomal glycosphingolipids (Zhou et al., 2004a).

 

Invariant NKT cells contribute to innate or preadaptive immunity to tumors and a range of infections, and help to maintain self-tolerance and to prevent autoimmunity. They do so by the rapid secretion of cytokines, including IFN-γ , tumor necrosis factor (TNF)-a, IL-4, IL-5, IL-13, granulocyte-macrophage colony-stimulating factor (GM-CSF), and IL-2. Furthermore, NKT cells are capable of Fas ligand- and perforin-dependent cytotoxic killing on TCR stimulation (Kronenberg, 2005; Kronenberg and Gapin, 2002).

 

 

2.3. CD8 T cells

Effector and memory CD8 T cells—often referred to as CTL—are important in the control of infection with viruses and certain other intracellular pathogens, and transformed host cells (Glimcher et al., 2004; Harty et al., 2000; Williams et al., 2006).

 

CD8 CTL are activated by signals from the TCR and costimulatory molecules in response to cognate MHC class I: peptide complexes presented by infected cells.

 

Similar to NK cells, CTL mediate the destruction of target cells via the directed release of cytotoxic granules containing perforin and granzymes onto an infected cell. Activated CD8 CTL also mediate target cell killing by the interaction of Fas ligand with Fas on target cells.

 

CD8 CTLs also produce copious amounts of IFN-γ in response to activation via the TCR or in response to IL-12 and IL-18 (Glimcher et al., 2004). IFN-γ in turn increases expression of MHC class I, thus making infected cells more readily recognizable by CD8 CTLs.

 

2.4. CD4 T cells play multiple roles in adaptive immunity

In contrast to CD8 T cells that are hardwired for cytolytic function and IFN-γ production, CD4 T cells can adopt a variety of functional effector responses. At present, four distinct CD4 effector lineages have been described: Th1, Th2, Th17, and regulatory T cells (Tregs), of which only Th1 CD4 effector T cells produce large amounts of IFN-γ (Dong, 2006; Harrington et al., 2006; Murphy and Reiner, 2002; Weaver et al., 2006).

 

The contributions of Th1 and Th2 CD4 T cells to cell-mediated or antiparasitic and humoral immunity, respectively, have been extensively studied (Murphy and Reiner, 2002). Th1 cells combat infection by intracellular pathogens by producing IFN-γ and IL-2 to stimulate and sustain an effective cellular immune response. Th2 cells, on the other hand, produce the cytokines IL-4, IL-5, and IL-13 that promote clearance of infe ction by multicellular helminths.

 

In mice, IL-4 and IL-5 secretion induces antibody (Ab) class switching in activated B cells from IgM and IgD to IgG1 and IgE and augments IgA production, respectively. IgG1 and IgA are functionally important for neutralization and targeting of extracellular pathogens for phagocytosis and killing by macrophages and neutrophils. IgE targets helminths for attack by eosinophils and triggers the activation of mast cells, thereby inducing mucus secretion, smooth muscle contraction, and vasodilatation to facilitate expulsion of helminths, while also playing a predominant role in asthma and allergy (Bischoff, 2007; Grimbaldeston et al., 2006).

 

CD4 Th17 cells are thought to protect against extracellular bacteria, particularly in the gut, by the secretion of the cytokines IL-17a, IL-17f, and IL-22 (Liang et al., 2006). Th17 cells may also be primary mediators of experimental autoimmune encephalomyelitis (EAE), a mouse model of multiple sclerosis, and of psoriasis and inflammatory bowel disease.

 

The fourth lineage of CD4 T cells, the Treg lineage, inhibits self-reactive adaptive responses that cause autoimmunity. Natural Tregs appear to arise as a separate lineage of CD4 T cells during thymic development (Sakaguchi, 2005). In addition to these natural Tregs, other Tregs can also be induced from naive CD4 T cell precursors as part of the adaptive immune response (Lohr et al., 2006).

 

Tregs dampen the immune response and prevent autoimmunity by direct suppression via cell–cell contact of effector T cells as well as by secretion of cytokines. Interestingly, Th17 cells share overlapping developmental factors with Treg cells (Weaver et al., 2006), suggesting that the generation of these two CD4 T cell lineages must be finely balanced to provide protective immunity without undue risk for autoimmunity.

 

As described more fully below, the lineage choice between Th1, Th2, and Th17 is strongly influenced by the cytokine milieu produced by dendritic cells (DCs) and other antigen-presenting cells (APCs) at the time of T cell priming.

 

Along with signals via the TCR and costimulatory receptors, cytokine receptor signaling pathways then induce or upregulate expression of the lineage-specific transcription factors, T-box expressed in T cells (T-bet), GATA-binding protein-3 (GATA-3), and retinoid orphan receptor- γ  T (ROR-γ T), in developing Th1, Th2, and Th17 cells, respectively; these transcription factors appear to be the ‘‘master regulators’’ of these respective cell fates.

 

By contrast, Treg development, commitment, and/or survival require the forkhead family transcription factor, FoxP3 (Fontenot et al., 2005), though the mechanisms by which this transcription factor acts are incompletely understood.

 

3. SIGNALING PATHWAYS CONTROLLING IFN-γ PRODUCTION BY NK CELLS

3.1. NK receptors provide a dynamic rheostat to control NK cell responses

The cytoplasmic storage of lytic granules and continual transcription of effector cytokines, including Ifng(ifn-γ ) , position NK cells to respond within minutes to hours on activation, thereby contributing to early stages of immunity to infection. A fine balance of activating and inhibitory receptors are involved in the control of NK cell responses, and several of the receptors and downstream signaling events are shared with receptors found in other cells such as T cells (reviewed by Lanier, 2005 and Vivier et al., 2004).

 

Activating receptors recognize stress-induced ligands expressed on infected, damaged, or transformed cells and are balanced by the function of inhibitory receptors, which bind classical and nonclassical MHC class I molecules to prevent NK cell destruction of healthy host cells.

 

Damage, transformation, or infection can downregulate the expression of MHC class I molecules, reducing the strength of inhibitory signals and contributing to NK cell activation.

 

Inhibitory receptors contain immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that signal via phosphorylated Src homology 2 domain tyrosine phosphatase-1 (SHP-1), SHP-2, or SHIP to prevent sustained calcium signaling and decrease the phosphorylation of a number of intracellular signaling molecules including Syk, phospholipase C-γ (PLC-γ), Shc, Vav, zeta chain-associated protein kinase of 70 kDa (ZAP70), SH2-domain-containing leukocyte protein of 76 kDa (SLP-76), and linker for activation of T cells (LAT).

 

By contrast, activating receptors include the Fcγ receptor CD16, CD94/natural-killer group 2, member C (NKG2C), CD94/NKG2E, and NKG2D, in addition to Ly49H and Ly49D in the mouse, and in humans the short form killer immunoglobulin receptors (KIRs). These activating receptors couple to signaling adaptor molecules FcRγ, CD3 ξ , or DAP12 to transmit activating signals into the cytoplasm via phosphorylation of immunoreceptor tyrosine-based activating motifs (ITAMs), resulting in the activation of protein tyrosine kinases (PTKs) of the Src family.

 

Src PTKs then phosphorylate tyrosines on secondary PTKs of the Syk family, resulting in the recruitment of a number of cytosolic adaptor molecules including SLP-76, 3BP2, and LAT, which in turn lead to the activation of downstream signaling cascades, including mitogen-activated protein kinase (MAPK), PLC-γ, and the Son of sevenless (Sos)/Ras pathways. A fourth signaling adaptor, DAP10 functions as a costimulatory molecule similar to CD28 on the surface of T cells to activate Akt.

 

Together, these signaling pathways downstream of activating receptors lead to the induction of NK cell effector mechanisms, including the rapid release of cytotoxic molecules and secretion of cytokines including IFN-γ.

 

The signaling pathways leading to IFN-γ secretion have been defined in vitro using a number of pharmacological inhibitors and by Ab cross-linking of activating receptors. NK cells appear to use distinct signaling pathways to induce cytotoxicity and cytokine secretion effector mechanisms, with the latter being more heavily dependent on MAPK signaling.

 

Inhibition of early signaling events by blocking activation of spleen tyrosine kinase (SYK) and Src kinases, or downstream events including phosphoinositol 3-kinase (PI3K) and MAPK pathways, greatly decrease the ability of NK cells to produce IFN-γ. More specifically, activation of MAPK pathways involving extracellular signal-regulated kinase (ERK) and p38 kinases leads to cytokine secretion, in part through the activation of Fos and Jun transcription factors. Furthermore, Syk/ Zap70/ NK cells fail to produce IFN-γ when stimulated in vitro through a number of different activating receptors. IFN-γ production in response to NKG2D activation specifically requires DAP12/Syk and stimulates Janus kinase 2 (JAK2), signal transducers and activators of transcription 5 (STAT5), PI3K, and MAPK pathways to induce secretion of IFN-γ and other cytokines. Finally, calcium flux is necessary for high-level IFN-g production in human NK cells in response to CD16 stimulation (Tassi and Colonna, 2005; Tassi et al., 2005).\

 

3.2. IL-12 is a potent activator of IFN-γ production in NK cells

In addition to NK cell activating receptors, a number of soluble and contact-dependent signals contribute to the activation of NK cells (reviewed by Newman and Riley, 2007), of which secretion of cytokines by infected cells and APCs are most well characterized. Among the cytokines that activate NK cells are IL-12, IL-18, IL-2, IL-15, and type I IFNs (IFN- α and IFN- β ) (reviewed by Lieberman and Hunter, 2002).

 

IL-12, originally identified based on its ability to enhance NK cell cytotoxic killing, has long been recognized for its ability to stimulate IFN-γ secretion by NK cells. Binding of IL-12 to its receptor activates the transcription factor STAT4, which can facilitate IFN-γ expression both directly and indirectly by upregulating the expression of CD44 and CD28 (Kobayashi et al., 1989).

 

CD80/CD86 ligation of the CD28 molecule on the surface of NK cells can activate the transcription factor NFkB and affect several NK cell functions, including enhancing IFN-γ production in vitro as well as in response to infection in vivo (Fitzgerald et al., 2000; Ghosh et al., 1993).

 

The ability of IL-12 to stimulate NK cell production of IFN-γ can be further increased by the presence of other soluble and cell-bound ligands, including type I IFNs, TNF- α , and IL-18.

 

Many of the synergistic stimuli that enhance IL-12 activation of IFN-γ production by NK cells share the ability to activate the transcription factor NFkB or STAT4. Binding of IL-18 to its receptor activates NFkB and MAPK (Strengell et al., 2003), and markedly augments IL-12-induced IFN-γ production by NK cells (Hoshino et al., 1999).

 

IFN-γ secreted by NK cells can bind to its receptor, IFN-γR, thereby activating STAT1 and upregulating T-bet expression (Afkarian et al., 2002). While this positive feedback loop is important for Th1 effector differentiation, it only modestly influences subsequent IFN-γ secretion by NK cells (Lee et al., 2000).

 

In addition to the IFN-γR, the receptor for type I IFNs (IFNAR), signals through STAT1. STAT1/ mice show impaired resistance to viral infection and tumors and decreased NK cytolytic function, but the ability of NK cells to secrete IFN-γ ex vivo in response to IL-12 stimulationis onlymarginallydiminished (Leeet al., 2000).

 

On viral infection, most nucleated cells produce type I IFNs (IFN- α and IFN- β), which are distinct from the type II interferon, IFN-γ. Type I IFNs bind IFNAR on infected and neighboring cells, activating STAT1, STAT2, and STAT4, which results in the inhibition of host and viral protein synthesis, resistance to viral replication, and increased presentation of antigens by MHC class I (Decker et al., 2005; Lieberman and Hunter, 2002).

 

NK cell cytotoxicity is largely dependent on type I IFNs (Nguyen et al., 2002a). IFN- β strongly enhances the cytotoxicity of human (Gerosa et al., 2005; Hanabuchi et al., 2006; Marshall et al., 2006) and murine NK cells (Hunter et al., 1997), but only modestly enhances IFN-γ production when used in combination with other stimuli and does not enhance on its own. IFN- α treatment of primary NK cells or the human NK-92 cell line induces low level, transient transcription of IFNG, which peaks around 1 h after stimulation, whereas IL-12 stimulation results in substantially longer and more robust IFNG transcription (Matikainen et al., 2001).

 

While neither IFN- α nor IL-18 alone can stimulate substantial IFN-γ secretion by human or mouse NK cells, the combination of these two cytokines stimulates IFN-γ production in amounts similar to IL-12 alone (Matikainen et al., 2001). These data suggest that IFN- α can activate STAT4 in an IL-12-independent and transient manner that requires additional stimuli, either IL-12 or IL-18 to fully induce IFN-γ secretion from NK cells (Hunter et al., 1997). Human plasmacytoid dendritic cells (pDCs) and myeloid DCs (mDCs) can respond to microbial stimuli, resulting in the release of cytokines that influence NK cell viability and function.

 

Furthermore, pDCs and mDCs can provide accessory ligands that mediate NK cell activation. Activation of pDCs via Toll-like receptor 9 (TLR9) by various CpGs oligonucleotides stimulates the secretion of IFN- α and TNF- α . In human NK cells, IFN- α and TNF-α can synergize to induce IFN-γ secretion (Marshall et al., 2006). In response to poly (I:C), mDCs induce IFN-γ production by NK cells through cell–cell contact and IL-12-dependent mechanisms (Gerosa et al., 2005). Resting human NK cells, as well as T cells and macrophages, express the glucocortoid-induced TNF receptor (GITR), stimulation of which promotes NK cell lytic function and IFN-γ secretion in synergy with IL-2, IFN- α , and NKG2D signaling (Hanabuchi et al., 2006). The expression of the GITR ligand (GITRL) on the surface of activated pDCs (Hanabuchi et al., 2006) provides additional evidence that NK cell activation relies on accessory signals from DCs.

 

3.3. IL-15 and IL-2 regulate NK cell development and contribute to IFN-γ production

The development of mature NK cells requires IL-15 production from the bone marrow stromal cells and monocytes (Carson et al., 1997; Puzanov et al., 1996; Vosshenrich et al., 2005). In the secondary lymphoid tissues, IL-15 is thought to be produced and trans-presented to NK cells by APCs, thereby providing survival and/or proliferative signals necessary for homeostatic maintenance of mature, peripheral NK cells, and priming these cells for cytolytic activity and the production of IFN-γ (Lucas et al., 2007; Ma et al., 2006; Williams et al., 1998).

 

Transfer of mature NK cells into IL-15/ or IL-15R α / mice results in a dramatic loss of the transferred NK cells (Cooper et al., 2002; Koka et al., 2003), further demonstrating the need of NK cells for IL-15-induced survival signals. Like IL-12 and IFN- α , IL-15 acts in concert with IL-18 to stimulate IFN-γ production (Strengell et al., 2003). IL-2, which shares the common b and g chains of its receptor with IL-15, can also promote NK cell growth, differentiation, cytolytic activity, and IFN-γ production in vitro (Fujii et al., 1998). However, unlike IL-15, IL-2 is dispensable for NK cell development and function in vivo. IL-2 does not directly affect IL-12Rβ2 expression (Wang et al., 1999) and only minimally affects IFN-γ secretion by mouse NK cells (Chang and Aune, 2005), but exposure of NK cells to IL-2 prior to stimulation with IL-12 increases expression of the IL-12Rβ2, suggesting that IL-2 may enhance NK response to IL-12 in vivo (Wang et al., 2000).

 

3.4. TGF- β is a negative regulator of IFN- γ production and NK cell development

While IL-2, IL-15, IL-12, IL-18, and type I IFNs promote IFN-γ production by NK cells, excess or prolonged production of IFN-γ can lead to protracted inflammation and immune activation, resulting in inflammatory and autoimmune pathologies. The immunoregulatory cytokine transforming growth factor- β (TGF- β ) inhibits the expression of IFN-γ by NK cells (Li et al., 2006b). TGF-β induces the phosphorylation of SMAD2 and/or SMAD3 signaling proteins, which then bind with a common SMAD4 partner, translocate to the nucleus and bind to the promoters of target genes and recruit activating or repressive complexes (Massague, 1998). As shown by chromatin immunoprecipitation (ChIP), SMAD3/4 heterodimers can bind to the Ifng promoter and repress transcription (Yu et al., 2006).

 

In addition to directly inhibiting Ifng transcription, TGF-β inhibits the expression of T-bet, STAT4, and IL-12Rβ2 (Gorelik and Flavell, 2000; Lin et al., 2005), all of which are important for IFN-γ expression. In addition to blocking expression of IFN-γ, TGF-β also has selective effects on NK cell development and homeostasis. Mice with deficiencies in TGF- β signaling in NK cells develop increased numbers of NK cells as early as 3 weeks after birth (Laouar et al., 2005). dnTGF- β RII transgenic mice also have higher frequencies of NK cells that are sustained much longer after L. major infection than do control mice (Laouar et al., 2005).

 

IL-12, IL-15, and IL-18 decrease the ability of NK cells to respond to TGF -β by decreasing transcription and surface expression of the receptor for TGF -β and the SMAD2 and SMAD3 signaling molecules, and by partially rescuing expression of the transcription factor T-bet. Furthermore, enforced expression of T-bet in TGF- β -treated NK cells maintains IFN-γ production (Gorelik and Flavell, 2000; Yu et al., 2006).

 

4. CONTROL OF IFN- γ PRODUCTION BY NKT CELLS

NKT cells are activated on TCR recognition of lipid antigens presented by CD1d. The ability of NKT cells to produce large amounts of both Th1 and Th2 cytokines is well documented and contrasts sharply with conventional T cells, which produce one or the other. Nearly all unstimulated NKT cells transcribe Ifng and Il4 (the gene encoding IL-4) and well over half transcribe message for both cytokines (Matsuda et al., 2003; Stetson et al., 2003), indicating that the transcription factors and epigenetic status necessary for high-level expression of these cytokines are already in place prior to stimulation of NKT cells. Nonetheless, exposure to IL-12 plus IL-18 selectively induces the secretion of IFN-γ in NKT cells (Nagarajan and Kronenberg, 2007) as it does in NK cells and memory/effector Th1 and CD8 T cells.

 

However, NKT cells from mice deficient for either the IL-4R or IL-12R produce IFN-γ and IL-4 in amounts similar to controls in response to a-GalCer stimulation, which is not the case for conventional T cells.

 

As IL-12 and IL-4 play little role in influencing the types of cytokines produced by NKT cells, it is not surprising that attempts to ‘‘polarize’’ NKT cells by varying the dose, route, or timing of antigen has failed to alter the immediate cytokine profile (Matsuda et al., 2003). Thus, it has been proposed that NKT cells are relatively fixed in their ability to simultaneously express both Ifng and Il4. The mechanisms that allow both classes of cytokines to be produced by NKT cells remain poorly understood (Kronenberg, 2005).

 

One possibility is that only a fraction of NKT cells secretes IL-4 and another fraction secretes IFN-γ . In this model, the choice of cytokines secreted by an individual NKT cell could either be random, be determined by the differential activation of NK receptors on the surface of an individual NKT cell, or be determined by differences in the structure of the lipopeptides recognized by NKT cells, which in turn would result in qualitatively or quantitatively different signals via the TCR. For example, reduced aliphatic chain length on derivatives of a-GalCer led to increased Th2 cytokine secretion but impaired IFN-γ secretion as a result of reduced activation of the NFkB transcription factor c-Rel (Oki et al., 2004), whereas an analog of a-GalCer containing a carbon replacement resulted in increased IFN-γ secretion (Schmieg et al., 2003). Alternatively, there could be subsets of NKT cells more poised toward the production of IFN-γ or IL-4.

 

Finally, NKT cells may secrete both IFN-γ and IL-4, as the findings of Stetson et al. (2003) and Matsuda et al. (2003) suggest, and the resulting modulation of the immune response may be in part due to the differential induction of surface molecules, including those that are involved in cell–cell interactions between NKT cells and other responding cells such as NK cells or DCs.