Structure of IL-6 and its gene The protein
IL-6 is a glycoprotein with a molecular weight ranging from 21 to 28 kDa with an isoelectric point of 5-6 (Fuller et al. 1987). IL-6 can undergo post-translational modifications including N- and O- linked glycosylations (May et al. 1988) and also phosphorylations (Santhanam et al. 1989). The Human IL-6 consists of 212 amino acids including a hydrophobic signal sequence of 28 amino acids (Hirano et al. 1986). Human IL-6 shows homology with IL-6 from the mouse and rat by 65%
and 68% at the DNA level and 42 and 58% at the protein level respectively (Northemann et al.
1989). The mouse and rat protein sequences are identical by 93% (Northemann et al. 1989). Both the C- and the N-terminus play a critical role for its biological functions (Brakenhoff et al. 1989).
A computer-aided structural analysis predicts that IL-6 consists of four anti-parallel α helices with two long and one short loop connections (Paonessa et al. 1995). This tertiary structure is, in other respects, shared by other molecules that are: ciliary neurotrophic factor (CNTF), leukaemia inhibitory factor (LIF), oncostatin M (OSM), cardiotrophin-1 (CT-1), growth promoting activity (GPA) and the interleukin-11 (Kishimoto et al. 1995). In spite of a low amino acid sequence homologies (30%) Bazan et al proposed to group them into the “IL-6-related” or “IL-6-like”
subfamily (Bazan 1990). In addition the members of this family use a common receptor subunit for signal transduction (see “receptors” section) and by consequence they exert similar and overlapping physiological responses that is one major characteristic of the cytokines (Gadient & Otten 1997).
Genetic
The human and mouse IL-6 genes are respectively of approximately 5 and 7 kilobases in length and both consist of five exons and four introns (Yasukawa et al. 1987). The genomic genes for human and murine IL-6 are mapped respectively to chromosomes 7p21 and 5 (Sehgal et al. 1986). Several potential transcriptional control elements such as glucocorticoid-responsive elements (GRFs), an activating protein-1 (AP-1) binding site, a c-fos serum-responsive element (c-fos SRE) homology, a c-fos retinoblastoma control element (c-fos RCE) homology, a cAMP-responsive element (CRE) and a nuclear factor (NF)-κB binding site have been identified within the IL-6 promotes (Kamimura et al. 2003).
The IL-6 receptors
The system receptor consists of two molecules (Figure 20). They both belong to the type I membrane proteins (extracellular N-terminus, one transmembrane domain). One is an 80kDa IL-6 binding protein (IL-6Rα also known as gp80 or CD126) of low-affinity and the other a 130 kDa signal transducer, glycoprotein 130 (gp130 also known as IL-6Rβ or CD130). gp130 does not bind IL-6 but constitutes in association with the IL-6Rα a high affinity binding site (see also figure 23).
The gp130 subunit is shared by the other IL-6-related molecules (Heinrich et al. 1998, Hibi et al.
1990, Taga et al. 1989, Yamasaki et al. 1988). This purpose explains why IL-6 type cytokines have overlapping actions (notion of redundancy) despite the low sequence homologies that they display (figure 21).
Figure 20: IL-6 receptors and gp130 structures in mouse and human.
Amino acid residues are denoted with the single-letter code. JAK kinases are constitutively associated with gp130 (see text for description)
Figure 21: The gp 130 is a shared transducer among IL-6 family cytokines.
This explains why the members of the IL-6 family have redundant effects.
IL-6Rα and gp130 belong to the cytokine receptor class I family (Figure 20). Indeed both receptors contain an Ig-like domain and fibronectin (FN) type-III domains including a four cystein motif and tryptophan-serine-any-tryptophan-serine (W-S-X-W-S) motifs located in the extracellular region (Heinrich et al. 1998, Kamimura et al. 2003). The four cystein residues and the W-S-X-W-S are responsible for the ligand binding and thus is called the cytokine-binding module (CBM). Besides the CBM, gp130 has three additional FN type-III domains in its extracellular region. IL-6Rα is important for ligand binding but it only has 82 amino acids in its cytoplasmic domain indicating that it could play a minor role in signal transduction (Hibi et al. 1990, Taga et al. 1989). In fact this role has been shown to be played by the gp130 subunit. Indeed despite the fact that gp130 has no intrinsic kinase activity, it contains the Box -motifs (Box1 and Box2) in its cytoplasmic domain which have been shown to be crucial for the gp130 association with Janus Associated Kinases (JAK, named also Just Another Kinase which are tyrosine kinase) by which downstream signaling cascades are initiated (Kamimura et al. 2003).
The IL-6 soluble receptor
The expression of gp130 is ubiquitous (Saito et al. 1992) while that of IL-6Rα is more restricted.
Interestingly, IL-6Rα has been shown to exist not only in the membrane-bound for but also in a soluble form (sIL-6Rα) (Van Wagoner et al. 1999) that is generated either by proteolytic cleavage of the membrane-bound form of 80 kDa to produce the 45-55 kDa soluble form or by an alternative splicing of its mRNA leading to the formation of sIL-6R lacking the transmembrane domain (Jones et al. 1998, Rose-John & Heinrich 1994).
When sIL-6Rα is associated with IL-6 they can act in an agonistic manner on cells expressing only gp130 (Hirano et al. 1997). Thus because of the existence of the sIL-6Rα, IL-6 can act on cells lacking the IL-6Rα subunit acquiring thus a “new” function that could not be played by the IL-6 alone. This mechanism called the “trans-signaling” accounts for the large functional diversity of IL-6 (Rose-John & Heinrich 1994, Hirano et al. 1994). It should be noted that a protein is obtained by the fusion of the coding sequence of IL-6R and IL-6R was produced (Chebath et al. 1997). This so called “chimeric” protein (IL6IL6R) display a much higher binding affinity to gp130 than non-fusioned IL-6 and sIL-6R.
The glycoprotein130 also has a soluble form but, in contrast to sIL-6Rα, this form has been revealed to act in an antagonistic manner (Rose-John & Heinrich 1994, Jostock et al. 2001).
The IL-6 signal transduction pathways The binding of IL-6 to its receptors
IL-6, IL-6Rα and gp130 have been demonstrated to form a hexameric complex with a stoichiometry of 2:2:2 (Paonessa et al. 1995, Ward et al. 1994). Mutagenesis combined to X-ray crystallography revealed 3 sites in the IL-6 protein that are crucial for the binding of the cytokine and its signal transduction (Figure 22). The actual paradigm concerning the IL-6 binding temporal sequence is:
IL-6 binds to IL-6Rα through IL-6 site 1 forming a heterodimer. This latter contacts the gp130 CBM trough the IL-6 site 2 forming a ternary complex that is not able to generate biological effect (Chow et al. 2001). Finally, after the homodimerisation of gp130, two trimers are assembled together via the site 2 and 3 of IL-6 and the Ig-like domain of gp130 (Chow et al. 2001) that is now able to generate IL-6 intracellular signaling pathways (Somers et al. 1997).
Figure 22: Model of the hexameric IL-6 receptor complex From (Paonessa et al. 1995)
The IL-6 intracellular signal transduction pathways
The principal informations concerning the precise gp130 cytoplasmic domains that are involved in the signaling pathways have been achieved in studies in which chimeric receptors were generated in knock-out gene targeting of the different tyrosine residues of the gp130 cytoplasmic domain. As previously described (see receptors section), the gp130 subunit contains Box-motifs in its cytoplasmic domain (see figure 20). These motifs are known to promote the association of gp130 with JAKs which are non-receptor tyrosine kinases (Heinrich et al. 1998). The gp130 dimerisation brings the JAKs into the proximity of gp130 resulting in JAKs activation (by phosphorylation). The activated JAKs then phosphorylate several tyrosine (Y) residues of cytoplasmic domain of gp130.
These phosphorylated tyrosine residues recruit (Darnell 1997): (1) the tyrosine phosphatase SHP-2 via its SH2 domain (by the gp130 tyrosine residue in position 759) and also (2) the STAT molecules (tyrosine residues at the 767th and 814th position for STAT3, tyrosine residues 905thand 915th for STAT1) leading to the generation of two major gp130 signaling pathways (figure 23): the so called “Y759 derived SHP-2/ERK MAPK” cascade and the “YXXQ-mediated STAT” pathway (Fukada et al. 1996, Kamimura et al. 2003, Heinrich et al. 1998).
Figure 23: Intracellular signaling pathways generated through gp130.
The formation hexameric initiates signal transduction of the JAK/STAT and ERK/MAPKinase pathways
More precisely, concerning the ERK/MAPK signal, after being recruited by Y759, SHP-2 is phosphorylated by JAKs and then interacts with the adaptor protein Grb2 (for Growth factor-receptor protein-2) that is constitutively associated with Sos (son-of-sevenless) that is a GDP/GTP exchanger for Ras. The GTP form of Ras transmits signals that lead to the activation of the ERK (Extracellular signal-related kinase)/ MAPK (Mitogen-Activated Protein Kinase) cascade (Fukada et al. 1996).
Regarding the STAT pathway after the recruitment of the STAT (STAT1 and STAT3) molecules by YXXQ residues these latter are phosphorylated by JAKs molecules. Finally the activated STATs form homo or heterodimers that enter the nucleus and activate several genes.
Here it is important to note that given the fact that gp130 generates multiple signal transduction pathways it could generate different but also contradictory signals in a given cell (Fukada et al.
1998). For example the Pheochromocytoma PC12 cells extend neurites when stimulated with IL-6 when pretreated with NGF. In these cells it has been shown that the ERK/MAPK positively controls
the neurite extension whereas the STAT3 pathway contradicts it (Ihara et al. 1997). We can conclude that the final physiological effect of IL-6 is the resultant of the balance between ERK/MAPK and JAK/STAT pathways. This so called “net effect” is determined itself by the expression pattern of the signaling molecules in the target cells (figure 24). This mode of action proposed by Hirano et al. the “Signal orchestration model” explains the multiplicity of action of the pleiotropic IL-6 (Kamimura et al. 2003).
Figure 24: “The Signal orchestration model”.
Given the fact that gp130 activates two different pathways the biological effect (or net effect) by the expression pattern of the signaling molecules in the target cells. This model explains why a same molecule can exert two different actions on the same cell.