I. Introduction

Top Previous Next References

Atherosclerosis remains, despite a recent decline, the most common cause of death in the Western world. The disease course of atherosclerosis is characterized by its chronicity, and progression in its initial stages is particularly insidious. Chronic inflammation is the pathological hallmark of atherosclerosis (Ross, 1986, 1993a, 1999), and inflammatory processes are instrumental in all stages of this disease. Even prior to the development of detectable intimal lesions, the expression pattern of the endothelium has been shown to be inflammatory in nature, conforming to the response-to-injury hypothesis as first postulated by the late Russell Ross (Ross and Glomset, 1973). Thus, in lesion-prone sites of the arterial tree, the endothelial expression of adhesion molecules is up-regulated, reflecting endothelial dysfunction secondary to unfavorable hemorheology (Nakashima et al., 1998) and/or hypercholesterolemia (Rosenfeld, 1991; Li et al., 1993; Sakai et al., 1997; Nakashima et al., 1998). In turn, this leads to the adhesion, extravasation, and intimal accumulation of circulating leukocytes (Nageh et al., 1997; Gerszten et al., 1998; Nakashima et al., 1998; Ramos et al., 1999; Dong et al., 2000), and thus to the development of the earliest detectable lesionthe fatty streakwhich consists solely of lipid-laden macrophages and T lymphocytes (Stary et al., 1994). These cell types are also present in more advanced plaques, in addition to smooth muscle cells and extracellular lipid and matrix deposits (Stary et al., 1994, 1995). The cellular constituents of the atherosclerotic lesion are thought to participate actively in the propagation of inflammation and, eventually, plaque destabilization (Ross, 1999; Sukhova et al., 1999). As well as contributing to the bulk of the lesion, plaque cells are involved in the production and degradation of extracellular matrix and contribute toward the formation of a necrotic lesion core by the elaboration of toxic mediators. These cellular functions are partly autonomous but to a large extent subject to autocrine and paracrine control mechanisms. A plethora of mediators has been shown to be involved in intercellular signaling in atheromatous tissue, including small molecules such as nitric oxide (Ignarro et al., 1999; Li and Forstermann, 2000), lipid mediators such as eicosanoids and sterols (Hajjar and Pomerantz, 1992; Edwards and Ericsson, 1999; Schnaper et al., 2000), and polypeptides such as cytokines (Frostegard et al., 1999; Meager, 1999).

Whereas fatty streaks are now known to develop even in utero under the influence of maternal hypercholesterolemia (Napoli et al., 1997), plaques rarely give rise to symptoms before the sixth or seventh decade of life. If primary prevention is to be the cardinal aim, the protracted nature of lesion development will necessitate a therapeutic strategy with a comparably prolonged duration of effectivity. In conjunction with the as yet perfunctory levels of prognostic accuracy for the identification of patients at risk of symptomatic atherosclerosis, this poses stringent demands with respect to the tolerability of any preventive intervention, including the use of immunomodulatory therapies.

The rate of atherogenesis largely depends on the level of exposure to major risk factors, including a positive family history, hypercholesterolemia, smoking, diabetes mellitus, and hypertension. Although the avoidance of risk factors undoubtedly constitutes the most rewarding approach to the prevention of atherosclerosis, it has thus far been frustrated by inadequate patient compliance and the influence of genetic factors in determining an individual's predisposition to atherosclerosis. This has led to the introduction of a variety of pharmacological interventions, including the widespread use of an extremely effective class of lipid-lowering drugs: the HMG-CoA reductase inhibitors, or so-called statins (Braunstein et al., 2001). Despite recent concerns regarding the induction of rhabdomyolysis, a rare and potentially lethal side effect of statin usage, these drugs continue to be the mainstay of most cholesterol-lowering regimens. In several clinical prevention trials (e.g., CARE; Ridker et al., 1998), statins have also been found to exert additional, lipid-independent, anti-inflammatory effects. These may contribute significantly to their anti-atherogenic properties, and this has indeed been corroborated in recent animal studies (Williams et al., 1998). Indeed, immunomodulation could be an attractive paradigm for the development of therapeutic alternatives to statins in atherosclerosis prevention. This may be of particular benefit to those whose lipid levels are (partially) unresponsive to statin therapy; as in a substantial number of patients in the U.S. National Cholesterol Education Program, LDL¹ cholesterol levels cannot be attained by statin monotherapy alone (Brown et al., 1998).

To enable rational drug design aimed at immunomodulation in atherosclerosis, the pivotal inflammatory processes involved in this disease need to be delineated. In this regard, extensive efforts have been devoted to outlining the involvement of cytokines, because these cell-regulatory proteins are known to be key players in the initiation and control of inflammation in general. The term "cytokine" was first coined in the 1970s and encompasses a large number of (glyco)proteins involved in cell-to-cell signaling. Cytokines are conventionally classified by assignment to one of six families: interleukins, the tumor necrosis factor family, interferons, colony-stimulating factors, growth factors, and chemokines (Henderson and Higgs, 2000). Considerable overlap between these families exists, however, and alternative methods of subdivision have been suggested. Depending on the aim of classification it may be preferable to distinguish cytokines with an essentially proinflammatory mode of action [including tumor necrosis factor (TNF), interleukin-12 (IL-12), IL-18, and interferon  (IFN)] from those with largely anti-inflammatory properties (including IL-4, IL-10, IL-13, and the endogenous IL-1 receptor antagonist, IL-1ra) or T helper cell type I (Th1; including IL-2, IFN, and TNF) from T helper cell type II (Th2; including IL-3, IL-4, IL-5, IL-6, IL-10, and IL-13) cytokines. Alternatively, it may be desirable to identify cytokines according to their major function, such as those effecting chemoattraction [chemokines, including monocyte chemoattractant protein-1 (MCP-1), RANTES, macrophage inflammatory protein-1 (MIP-1), IL-8, and IL-16] or on the basis of receptor sequence homology (e.g., those employing the gp130 signal transduction protein, such as IL-6, IL-11, IL-12, oncostatin M, and cardiotrophin-1). Nonetheless, a substantial degree of pleiotropism in cytokine effector functions makes most of these subdivisions somewhat arbitrary.

Members of each conventional cytokine family have been found to be involved in atherogenesis, and all cell types present in the atherosclerotic plaque are capable of producing and responding to cytokine mediators. It is conceivable, therefore, that intervention in cytokine signaling could provide effective prevention and/or treatment of atherosclerosis, and proof-of-principle data to this effect have been obtained in a variety of in vitro and in vivo studies, although this has not yet yielded clinically applicable protocols. In this review, we shall focus mainly on interleukins in our aim to outline the results that have been achieved to date in delineating the pathophysiological role and the therapeutic potential of cytokines in atherosclerosis. In addition, we shall discuss the potential of the modulation of cytokine activity as a therapeutic approach to the primary and secondary prevention of atherosclerosis. Following an overview of the roles ascribed to a variety of interleukins in the pathogenesis of atherosclerosis, we shall describe recent progress in this field and perceived future opportunities.

By definition, interleukins are produced mainly by leukocytes and exert their effects mainly on leukocytes. Endothelial cells and smooth muscle cells, however, also express a variety of interleukins and/or their respective receptors, and their effects in atherogenesis are therefore by no means restricted to macrophages and T cells. Thus far, more than 30 major members of the interleukin family have been identified, and the majority of these have been shown to play a role in atherogenesis. As applies to cytokines in general, it is possible to subdivide the interleukins into families according to the homology of their amino acid sequences or the homology of the receptor complexes to which they bind (Fig. 1). Of these subgroups, the gp130 receptor family comprises principally pro-atherogenic interleukins, but most other families have both anti- and pro-atherogenic members (e.g., IL-1 family, IL-2 family, and c receptor family). It has not proved feasible to pinpoint an interleukin that acts as the cardinal culprit in the atherosclerotic process. On the contrary, it seems rather more likely that the delicate balance between pro- and anti-inflammatory signals that generally serves to keep inflammation in check, goes awry in atherosclerosis, leading to a self-perpetuating mechanism of lesion formation (Ross, 1993b; Tedgui and Mallat, 2001). Considering the extensive interplay of soluble mediators in the atherosclerotic plaque, however, it may prove possible to devise an anti-atherosclerotic therapy aimed at modifying the effect of a single interleukin, provided that due attention is paid to the mechanisms of redundancy, which have been shown to exist in cytokine signaling. In doing so, candidate interleukins cannot be identified solely by virtue of a demonstrated systemic or local modulation of their expression in the course of atherogenesis. On the contrary, it is of paramount importance to determine whether cytokine responses that have been observed in relation to the development of atherosclerosis are compensatory to, contributory to, or merely associated with this disease. Making this distinction will require well designed intervention studies in animal models, in which the effect of attenuation or administration of a particular interleukin can be evaluated. The currently favored approach entails the up- or down-regulation of interleukin expression in atherosclerosis-prone mouse strains by means of gene insertion ("transgenics") or gene deletion ("knockouts"), respectively. Administration of an interleukin or its ablation by specific antibodies/antagonists, however, can also provide valuable data regarding its role in atherogenesis. When pertinent, the results of such studies will be discussed in the next section.

View larger version (44K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the receptor specificity and mechanism of action of interleukin families thought to be involved in atherogenesis. Most receptors have been found to consist of heterodimeric complexes, frequently incorporating an interleukin-specific chain in addition to a common chain that is shared by the interleukin family members (including IL-2R, c, and gp130). Receptor activation initiates intracellular phosphorylation cascades that are mediated by kinases (including p38 MAPK, c-Jun N-terminal kinase, and JAKs), resulting in the activation and/or nuclear translocation of transcription factors (including AP-1, STATs, NF-B). Binding of these factors to DNA consensus sequences, in conjunction with the required cofactors, effects the expression of specific patterns of pro- and/or anti-inflammatory mediators.

Since the effects exerted by cytokines may differ significantly depending on their local environment, it will also be necessary to distinguish between the role of systemic and local variations in cytokine levels. This type of information could in the future be derived from cell- or organ-specific gene overexpression through the use of specific promoters and gene deletion by means of the cre-lox system (Perkins, 2002) or by comparison of the effects of local and systemic administration of cytokines.

The IL-1 family comprises four proteins that share considerable sequence homology and contain a -pleated sheet structure (Dinarello, 1997): IL-1, IL-1, IL-1 receptor antagonist (IL-1Ra), and IL-18 (also known as IFN-inducing factor). Release of mature IL-1 requires extracellular calpain-mediated cleavage of a pro-IL-1, whereas mature IL-1 is derived proteolytically from pro-IL-1 by intracellular IL-1-converting enzyme (ICE or caspase-1) activity. Upon binding of IL-1 or IL-1 to the IL-1 receptor type I (IL-1RI), IL-1R accessory protein (IL-1RIAcP) is recruited by the receptor complex, and intracellular signal transduction is triggered through a p38 mitogen-activated protein kinase (MAPK)-activated phosphorylation cascade. Due to extensive signal amplification, minute amounts of IL-1 can have considerable biological activity, and as little as 1 ng/kg intravenous IL-1 causes symptoms in humans. The signaling cascade culminates in the nuclear translocation of the transcription factors nuclear factor kappa B (NF-B) and activating protein-1 (AP-1) and the ensuing transcription of a variety of proinflammatory genes, including autocrine amplification of IL-1 production (Suzuki et al., 1989). In addition to the IL-1RI, IL-1 may also bind to the so-called type II interleukin-1 receptor, the expression of which appears to be regulated by IL-4 (Colotta et al., 1993b). Binding of IL-1 to this receptor does not result in cellular activation, and IL-1RII is therefore presumed to act as a decoy that negatively regulates IL-1 activity.

A further member of the IL-1 cytokine family, IFN-inducing factor, has been termed IL-18, on the basis of its pleiotropic Th1-inducing effects (Ushio et al., 1996). It has been assigned to the IL-1 family on the grounds of sequence homology (26% with IL-1) and similarity of the IL-18 receptor to IL-1R (Torigoe et al., 1997; Dinarello, 1999). Like IL-1, IL-18 is dependent on ICE for proteolytic processing, and on nuclear translocation of NF-B for transcriptional activation.

Owing to its proinflammatory effects on endothelial cells (Jirik et al., 1989; Loppnow and Libby, 1989a,b; Sironi et al., 1989; Suzuki et al., 1989; Sica et al., 1990b; Bochner et al., 1991; Clinton et al., 1992; Collins et al., 1995; Garcia et al., 2000), smooth muscle cells (Loppnow and Libby, 1989a, 1990; Wang et al., 1991; Clinton et al., 1992; Braun et al., 1995; Stanford et al., 2000), and macrophages (Sica et al., 1990b), and due to its production by all of these cell types in atherosclerotic lesions (Moyer et al., 1991; Tipping and Hancock, 1993; Galea et al., 1996), IL-1 was one of the first cytokines to be considered instrumental in the propagation of vessel wall inflammation in atherosclerosis. It is thought to facilitate early lesion formation by increasing leukocyte adhesion to endothelial cells (Bevilacqua et al., 1985; Wang et al., 1995) and mediating leukocyte transmigration (Moser et al., 1989; Furie anf McHugh, 1989). Subsequently, locally produced IL-1 may serve to maintain an inflammatory milieu by autocrine and paracrine stimulation of cytokine (Jirik et al., 1989; Loppnow and Libby, 1989a,b, 1990, 1992; Sironi et al., 1989; Sica et al., 1990a,b; Wang et al., 1991; Clinton et al., 1992; Li et al., 1995; Taki et al., 1999; Garcia et al., 2000; Stanford et al., 2000) and adhesion molecule expression (Osborn et al., 1989; Bochner et al., 1991; Braun et al., 1995; Collins et al., 1995). In the advanced plaque, IL-1-induced up-regulation of matrix metalloproteinases may destabilize the proteinaceous scaffold of the cap and thereby have a hand in plaque rupture (Galis et al., 1995; Libby et al., 1995); this hypothesis is corroborated clinically by the fact that a particular IL-1 gene polymorphism has been found to be associated with myocardial infarction in chlamydia pneumoniae seropositive patients (Momiyama et al., 2001), and that pericardial fluid levels of IL-1 are raised in patients with unstable angina pectoris (Oyama et al., 2001).

Because the IL-18 signal transduction cascade is similar to that activated by IL-1, it is perhaps unsurprising that IL-18 has also been found to up-regulate the expression of intercellular adhesion molecule 1 (ICAM-1) and cytokines by monocytes, including IL-1, IL-6, and IL-8 (Dinarello, 1999), and the production of vascular cell adhesion molecule-1 (VCAM-1) by endothelial cells (Vidal-Vanaclocha et al., 2000). It is, therefore, entirely conceivable that IL-18 may have pro-atherogenic properties, and Mallat et al. (2001a) have indeed demonstrated IL-18 in atherosclerotic plaques in human carotids, which is primarily localized to macrophages. They found the corresponding receptor, IL-18R, to be expressed on endothelial cells and macrophages and barely present on SMCs. These findings have subsequently been confirmed histologically and in vitro by Gerdes et al. (2002), who also demonstrated the functionality of the IL-18 receptor on these cells through IL-18-mediated induction of pro-atherogenic factors, including IL-6, IL-8, ICAM-1, and matrix metalloproteinases. In addition, the serum level of IL-18 has recently been identified as a strong predictor of cardiovascular death in stable and unstable angina (Blankenberg et al., 2002). The pro-atherogenic effects of IL-18 are thought to be mediated by IFN, since the induction of atherosclerosis by exogenous IL-18 is abrogated by IFN deficiency in apolipoprotein E knockout (apoE/) mice (Whitman et al., 2002). A role for IL-18 in plaque destabilization was suggested by the up-regulation of IL-18 mRNA levels in symptomatic and ulcerative atherosclerotic plaques (Mallat et al., 2001a).

In comparison with the proinflammatory reprobates of the IL-1 family, IL-1ra appears positively angelic. IL-1ra displays affinity for the IL-1R, but it does not induce a cellular response; it is therefore believed to be an endogenous inhibitor of IL-1 signaling (Dinarello, 1997). IL-1ra is produced by monocytes (Arend et al., 1990), macrophages (Janson et al., 1991), and smooth muscle cells (Beasley et al., 1995). Recombinant intracellular IL-1ra has been shown to counteract the IL-1-induced production of IL-6, IL-8, and monocyte chemotactic protein by human endothelial cells (Bertini et al., 1992), and to inhibit smooth muscle cell proliferation (Porreca et al., 1993). Moreover, vascular inflammation is the major phenotypic characteristic of IL-1ra-deficient mice (Nicklin et al., 2000), whereas atherogenesis is reduced in IL-1ra transgenic mice on a high fat diet (Devlin et al., 2002), and fatty streak formation is reduced in apoE/ mice by IL-1ra administration (Elhage et al., 1998). Il-1ra has been found to be present in carotid atherosclerotic plaques (Gottsater et al., 2002), and the relevance of IL-1ra to human atherosclerosis is underscored by the fact that certain IL-1ra alleles are associated with coronary artery disease (Francis et al., 1999) and restenosis (Kastrati et al., 2000; Francis et al., 2001).

This family of cytokines encompasses a group of interleukins which share a common receptor subunit, the "common  chain" (c chain), which acts in unison with a subtype specific  chain to initiate the signaling cascade. As the common receptor subunit was initially discovered in relation to IL-2, it has also been termed the "IL-2 receptor  chain" (Takeshita et al., 1990), and the group of cytokines that interact with this receptor has consequently been termed the "IL-2 family" (Leonard and Lin, 2000). The members of this interleukin family are primarily involved in T cell development and activation, and mutations of the c chain cause X-linked severe combined immunodeficiency in humans (Noguchi et al., 1993b) and lead to thymic hypoplasia in mice (Cao et al., 1995).

In addition to IL-2, the family includes IL-4 (Russell et al., 1993), IL-7 (Noguchi et al., 1993a), IL-9 (Russell et al., 1994), IL-15 (Giri et al., 1994a), and IL-21 (Vosshenrich and Di Santo, 2001). All members interact with receptor complexes consisting of an interleukin-specific  chain and the common c chain (Fig. 1). Moreover, the IL-4  chain is also a component of the IL-13 receptor complex (Zurawski et al., 1993), and for purposes of classification, we shall include IL-13 in this interleukin family. A substantial degree of functional redundancy is extolled by the IL-2 family members, which is comprehensible in view of considerable overlap in their signaling pathways. Thus, Janus kinase 1 (Jak1) and Jak3 have been found to be activated by the subtype-specific chains and the constant c chain, respectively (Miyazaki et al., 1994; Russell et al., 1994; Leonard and Lin, 2000), which ultimately cascades into the activation of transcription by the common downstream effector molecules "signal transducer and activator of transcription" 5a (Stat5a), Stat5b, and Stat3 (Lin et al., 1995; Lin and Leonard, 2000). IL-4 and IL-13 are somewhat distinct in activating Jak2 and Stat-6 via a c chain-independent pathway (Palmer Crocker et al., 1996).

IL-2 (Arbustini et al., 1991; Frostegard et al., 1999) and the IL-2R receptor (Kishikawa et al., 1993) are expressed in atheromatous tissue, but a direct causal role for IL-2 in atherogenesis remains to be proven. Nonetheless, serum IL-2 levels have been found to be elevated in ischemic heart disease (Mazzone et al., 1999) and especially unstable angina pectoris (Mizia-Stec et al., 2002), and the risk of acute myocardial infarction is increased following IL-2 treatment for cancer (Kragel et al., 1990). A possible explanation for the presumed pro-atherogenic effect of IL-2 may lie in its ability to induce a T helper cell shift toward a Th1 phenotype. T cells have been shown to be present in atherosclerotic lesions (Hansson et al., 1988), and Th1 cells, in particular, are believed to actively promote atherogenesis (de Boer et al., 1999; Frostegard et al., 1999; Huber et al., 2001; Laurat et al., 2001; Song et al., 2001). In its capacity as an autocrine stimulator of Th1 cell differentiation and proliferation (Kurt-Jones et al., 1987; Harel-Bellan et al., 1988), IL-2 may promote the expansion and activation of this T cell subset, and, consequently, plaque development.

Conversely, IL-4 is known to promote Th2-type responses (partly by autocrine activation) and to exert immunosuppressive effects on macrophages, including the suppression of proinflammatory cytokine production and the stimulation of IL-1ra elaboration (Paul, 1991). This cytokine is therefore considered to be potentially anti-atherogenic. The highly pleiotropic effects of IL-4, however, reserve a rather more complicated role for IL-4 in atherosclerosis. Thus, whereas mice deficient in Stat6, which is one of the mediators activated by IL-4, develop larger atherosclerotic lesions than their wild-type counterparts (Huber et al., 2001), IL-4 deficient mice do not display increased susceptibility to diet-induced atherosclerosis (George et al., 2000a). They have even been found to be relatively resistant to the acceleration of fatty streak formation by heat shock protein 65 or mycobacterium tuberculosis (George et al., 2000b). Similarly, reconstitution with IL-4-deficient bone marrow in LDLr/ mice reduces atherosclerotic lesion formation in the aortic arch and the thoracic aorta compared with reconstitution with wild-type bone marrow (King et al., 2002). Although IL-4 expression in atherosclerotic plaques appears to be limited (Uyemura et al., 1996), among the pro-atherogenic effects of IL-4 we may count the up-regulation of P-selectin (Khew-Goodall et al., 1999) and 15-lipoxygenase (Lee et al., 2001b) expression by endothelial cells, VCAM-1 (Barks et al., 1997) and matrix metalloproteinase 1 (MMP-1) (Sasaguri et al., 1998) expression by vascular smooth muscle cells, and the augmentation of CD36 receptor expression (Feng et al., 2000) and cholesterol esterification (Cornicelli et al., 2000) in macrophages. On the other hand, IL-4 has also been shown to inhibit smooth muscle cell proliferation (Vadiveloo et al., 1994; Sasaguri et al., 1998) and macrophage adhesiveness (Elliott et al., 1991). The net effect of IL-4 in atherosclerosis thus still hangs in the balance, and it may vary with the stage of the disease.

IL-9 was initially identified as a mast cell and T cell growth factor (Renauld et al., 1990) and has subsequently been shown to lead to exaggerated Th2-type inflammatory responses (Godfraind et al., 1998; McLane et al., 1998) and thymic lymphomas (Renauld et al., 1994) in IL-9 transgenic mice. IL-9 is not entirely independent in its actions, however, since IL-9 production by T lymphocytes requires IL-2-mediated stimulation (Houssiau et al., 1992), and the mitogenic effect of IL-9 on T lymphocytes requires their preactivation (Uyttenhove et al., 1988). In a murine model of Gram-negative bacterial shock, IL-9 led to suppression of TNF, IL-12, and IFN, possibly mediated by an induction of IL-10 expression (Grohmann et al., 2000). In agreement with this study, IL-9 has been found to induce the expression of the intracellular cytokine signal inhibitors cytokine-inducible SH2-containing protein, suppressor of cytokine signaling (SOCS)-2 and SOCS-3 (Lejeune et al., 2001). SOCS-3, in particular, may impair signaling by pro-atherogenic cytokines that act through the gp130 receptor, including IL-6 and IL-12. Some of the activities of IL-9 may also be mediated by its induction of IL-22 (IL-TIF), which shares 22% sequence homology with IL-10 (Dumoutier et al., 2000). Although its role in atherosclerosis has thus far not been elucidated, it appears that IL-9 may be potentially anti-atherogenic through a deflection of the immune response from a Th1 to a Th2 type. Albeit that a caveat needs to be added, as overzealous stimulation of Th2 responses may well prove to be detrimental in the later stages of atherosclerosis. Thus, mast cells have been identified in advanced plaques (Kaartinen et al., 1994a; Jeziorska et al., 1997) and are presumed to promote plaque instability by the secretion of chymase (Kaartinen et al., 1994b; Kovanen, 1997) and the stimulation of calcification (Jeziorska et al., 1998). Their stimulation may promote, rather than impede, the development of atherosclerotic complications.

IL-15 is produced by a variety of cells, including monocytes (Musso et al., 1999) and endothelial cells (Oppenheimer-Marks et al., 1998; Krishnaswamy et al., 1999), and has an activity profile similar to IL-2, without sharing sequence homology (Waldmann and Tagaya, 1999). IL-15 mediates extravasation of lymphocytes through its stimulatory and chemotactic effects on natural killer cells (Carson et al., 1994; Allavena et al., 1997) and T lymphocytes (Giri et al., 1995; Sancho et al., 1999) and by the up-regulation of hyaluronan on the endothelium (Estess et al., 1999). Recently, atherosclerotic lesions in humans and apoE/ mice were found to contain IL-15-responsive T cells as well as IL-15 itself, which colocalizes with oxidized LDL-positive macrophages (Houtkamp et al., 2001, Wuttge et al., 2001). IL-15 may therefore accelerate atherogenesis by promoting the recruitment and antigen-independent induction of T lymphocytes.

Despite sharing only 20 to 25% sequence homology and differing from IL-4 in lacking an effect on T cell function (Zurawski and de Vries, 1994), IL-13 is highly akin to IL-4 with respect to its immunomodulatory properties (Opal and DePalo, 2000), which is likely to be attributable to IL-4R-mediated Stat6 activation by both cytokines (Hart et al., 1999). In monocytes, IL-13 attenuates the expression of a wide range of inflammatory cytokines, including IL-1, IL-6, IL-8, IL-10, IL-12, MIP-1, granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), IFN, and TNF, while up-regulating the expression of IL-1ra (de Waal Malefyt et al., 1993; Mijatovic et al., 1997). Nitric oxide production is inhibited by IL-13 in macrophages (Doherty et al., 1993; Bogdan et al., 1997) and smooth muscle cells (Ruetten and Thiemermann, 1997). The properties of IL-13 are not exclusively anti-inflammatory, however, as exemplified by the IL-13-mediated potentiation of IL-8 receptor expression, 15-lipoxygenase expression, and LDL oxidation by monocytes (Nassar et al., 1994; Folcik et al., 1997; Bonecchi et al., 2000), and of IL-8 and MCP-1 release in response to IL-1 or TNF in SMCs (Jordan et al., 1997). Moreover, IL-13 is known to enhance the transmigration of leukocytes by stimulating the endothelial expression of adhesion molecules (Bochner et al., 1995; Ying et al., 1997) and chemotactic factors (Goebeler et al., 1997). In analogy with IL-4, the overall effect of IL-13 in atherosclerosis is still controvertible.

The complex actions of IL-2 family members in the vascular wall are depicted in Fig. 2.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Overview of the complex actions of IL-2 family members in the vascular wall. All members are produced by cellular constituents of the atherosclerotic plaque (dotted line: EC, endothelial cell; SMC, smooth muscle cell; M, macrophage; TC, T cell) and exert effects that are considered to be either pro-atherogenic (continuous line) or anti-atherogenic (dashed line).

The common receptor subunit shared by the members of this family of cytokines, gp130, was first discovered as a signal transducer for IL-6 (Hibi et al., 1990). The other factors to employ this receptor subunit in combination with their own specific subunit, are IL-11 (Yin et al., 1993), IL-12 (Chua et al., 1994), leukemia inhibitory factor (Gearing et al., 1991), oncostatin M (Gearing et al., 1992), cardiotrophin-1 (Ip et al., 1992), ciliary neurotrophic factor (Pennica et al., 1995), and neurotrophin-1/B cell-stimulating factor-3 (Senaldi et al., 1999) (Fig. 1). Following gp130 binding, the Janus kinases Jak1, Jak3, and Tyk2 and the transcription factors Stat1 and Stat3 are phosphorylated (Heinrich et al., 1998). In this review, we shall restrict the discussion to the interleukin members of the gp130 family.

In addition, two novel heterodimeric interleukins with an activity profile similar to IL-12 have recently been identified. IL-23 is composed of a p19 subunit and the p40 subunit of IL-12 (Oppmann et al., 2000), and this cytokine acts through a receptor composed of IL-12R1 and a novel cytokine receptor subunit, IL-23R (Parham et al., 2002). IL-27 is made up of an IL-12 p40-related and an IL-12 p35-related protein and binds to the gp130-related receptor WSX-1/TCCR (Pflanz et al., 2002).

Endothelial cells, smooth muscle cells, and macrophages are capable of elaborating IL-6, and its expression has been observed in atherosclerotic lesions in humans, hypercholesterolemic rabbits, and apoE-deficient mice (Ikeda et al., 1992; Kishikawa et al., 1993; Seino et al., 1994; Rus et al., 1996; Sukovich et al., 1998; Schieffer et al., 2000). Although the endothelium is largely unresponsive to IL-6 (Podor et al., 1989), addition of the soluble IL-6R subunit (sIL-6R) enables endothelial cells to mount an inflammatory response to IL-6, by interacting with membrane-bound gp130 (Jones et al., 2001). This process has been termed "trans-signaling", and it may lead to increased endothelial cell adhesiveness by the up-regulation of E-selectin, ICAM-1, and VCAM-1, and the release of inflammatory mediators, including MCP-1, IL-8, and IL-6 itself (Modur et al., 1997; Romano et al., 1997). Thus, sIL-6R present in serum and/or elaborated locally by cells in the intima may serve to augment endothelial adhesion and extravasation of leukocytes into the atherosclerotic plaque. Monocytes and macrophages, on the other hand, produce IL-6R autonomously and therefore do not depend on ambient sIL-6R levels for IL-6-mediated modulation of gene expression (Akira and Kishimoto, 1996). The effector functions of IL-6 in cells of the monocyte/macrophage lineage include the differentiation of monocytes to macrophages (Chomarat et al., 2000), the up-regulation of acute phase response gene expression in hepatocytes and macrophages (Perlmutter, 1989), and the priming of macrophages for enhanced TNF production in response to lipopolysaccharide (LPS) administration (Cochran and Finch-Arietta, 1992). In smooth muscle cells, IL-6 induces proliferation directly (Nabata et al., 1990; Ikeda et al., 1991) and indirectly through the initiation of an autocrine loop mediated by the up-regulation of gp130 (Klouche et al., 1999). In addition, smooth muscle cells are stimulated by IL-6 to express ICAM-1 (Ikeda et al., 1993) and to evolve into foam cells (Klouche et al., 2000).

Whereas homozygous deletion of gp130 in mice leads to intrauterine death due to myocardial hypoplasia (Yoshida et al., 1996), IL-6-deficient mice develop normally despite an attenuated acute phase response and impaired cellular immunity to virus infection (Kopf et al., 1994). This is a reflection of the functional redundancy in gp130-mediated signaling and thus of the extent to which the other members of the gp130 family can take over IL-6-mediated functions. IL-6 was initially described as a lymphocyte stimulatory factor but has since been found to exert a plethora of inflammatory effects (Hirano et al., 1990). With the possible exception of IL-1, IL-6 is the cytokine with the most extensively studied pro-atherogenic profile. Causality has been established through the exacerbation of early atherosclerosis by recombinant IL-6 in various atherosclerosis-prone murine models (Huber et al., 1999). Interestingly, the progression of atherosclerotic lesions to an advanced phenotype appears to be inhibited by IL-6 in apoE-deficient mice, uncovering a potentially biphasic mode of action in atherogenesis (Elhage et al., 2001), which is perhaps partly explained by its observed anti-inflammatory properties (Barton et al., 1996; Xing et al., 1998) and its inhibition of macrophage class A scavenger receptor expression (Liao et al., 1999). Nonetheless, inhibition of IL-6 signaling may be considered to constitute an attractive therapeutic strategy for the prevention of coronary heart disease (Stein and Kung Sutherland, 1998; Yudkin et al., 2000).

Clinically, elevated levels of IL-6 and its hepatic by-product C-reactive protein (Verma et al., 2002) are associated with increased risks of coronary and peripheral atherosclerosis (Erren et al., 1999; Mazzone et al., 1999; Flex et al., 2002; Bermudez et al., 2002; Kato et al., 2002; Stenvinkel et al., 2002; van der Meer et al., 2002), myocardial infarction (Ridker et al., 2000b; Ikeda et al., 2001), and the risk of death of patients with cardiovascular disease (Volpato et al., 2001), and IL-6 has been suggested to mediate the pro-atherogenic properties of cytomegalovirus (Blankenberg et al., 2001). In a large multicenter study, IL-6 gene polymorphisms were found to correlate with the severity of coronary artery disease and the risk of myocardial infarction (Georges et al., 2001), and carotid atherosclerosis has been shown to be independently linked with an IL-6 promoter polymorphism (Rauramaa et al., 2000; Rundek et al., 2002), as has the risk of coronary artery disease (Humphries et al., 2001). In addition, lower levels of soluble IL-6 receptor, a naturally occurring IL-6 antagonist, are linked with the risk of myocardial infarction (Ueda et al., 1999). Although these clinical findings do not establish causality, they have identified a strong association between IL-6 levels and atherosclerosis.

Despite sharing considerable redundancy with IL-6 with respect to its signaling and effector functions, IL-11 has been judged to be a more anti-inflammatory member of the gp130 family of cytokines based on the net effect of its pleiotropic actions (Schwertschlag et al., 1999; Taki et al., 1999). In macrophages, recombinant IL-11 has been found to attenuate macrophage expression of TNF, IL-1, IL-12, and nitric oxide following an LPS challenge (Trepicchio et al., 1996; Leng and Elias et al., 1997). These effects are direct and mediated by NF-B down-regulation (Trepicchio et al., 1997), as is IL-11-mediated attenuation of smooth muscle cell proliferation and cytokine production (Zimmerman et al., 2002). In endothelial cells, IL-11 provides protection against immune-mediated injury (Mahboubi et al., 2000), and inhibits apoptosis through up-regulation of survivin (Mahboubi et al., 2001). In CD4+ lymphocytes, IL-11 has been found to induce a shift from a Th1 to a Th2 phenotype (Bozza et al., 2001). This effect has been put to use in immunomodulatory treatment employing IL-11 in psoriasis (Trepicchio et al., 1999) and Crohn's disease (Sands et al., 1999), and it may also offer therapeutic possibilities in the setting of atherosclerosis.

Activated monocytes are the primary source of IL-12 (D'Andrea et al., 1992), a cytokine that induces proliferation (Gately et al., 1991) and a shift toward a Th1 expression pattern in lymphocytes (Hsieh et al., 1993). IL-12 was originally implicated in atherosclerosis by Uyemura et al. (1996), who observed an abundance of p40 mRNA and IL-12 p70 protein in atherosclerotic lesions, and up-regulation of IL-12 production by monocytes following the addition of highly oxidized LDL. Subsequently, atherosclerotic lesions in apoE-deficient mice were found to contain IL-12, and their progression to be accelerated by daily injections of recombinant IL-12 (Lee et al., 1999). Conversely, a selective defect of macrophage IL-12 synthesis due to 12/15-lipoxygenase deficiency reduces lesion formation in atherosclerosis-prone Apobec-1//ApoE/ mice (Zhao et al., 2002). In clinical practice, raised serum levels of IL-12 have been found to be associated with acute myocardial infarction (Zhou et al., 2001a).

The genes encoding the members of this familyIL-3, IL-5, and GM-CSFare clustered on the human chromosome 5 (van Leeuwen et al., 1989) (Fig. 1). Their products bind to receptor complexes consisting of a common  chain (c) and a cytokine-specific