`
`Contents lists available at SciVerse ScienceDirect
`
`Cytokine & Growth Factor Reviews
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c y t o g f r
`
`Survey
`
`Plasticity and cross-talk of Interleukin 6-type cytokines
`
`Christoph Garbers a, Heike M. Hermanns b, Fred Schaper c, Gerhard Mu¨ ller-Newen d, Joachim Gro¨ tzinger e,
`Stefan Rose-John e, Ju¨ rgen Scheller a,*
`
`a Institute of Biochemistry and Molecular Biology II, Medical Faculty, Heinrich-Heine University, Du¨sseldorf, Germany
`b Rudolf Virchow Center, DFG Research Center for Experimental Biomedicine, University of Wu¨rzburg, Wu¨rzburg, Germany
`c Institute of Biology, Otto-von-Guericke-University, Magdeburg, Germany
`d Institute of Biochemistry and Molecular Biology, RWTH Aachen University, Aachen, Germany
`e Institute of Biochemistry, Christian-Albrechts-University, Olshausenstrabe 40, Kiel, Germany
`
`A R T I C L E
`
`I N F O
`
`A B S T R A C T
`
`Article history:
`Available online 15 May 2012
`
`Keywords:
`Interleukin-6
`Cross-talk
`Plasticity
`Signal transduction
`
`Contents
`
`Interleukin (IL)-6-type cytokines are critically involved in health and disease. The duration and strength
`of IL-6-type cytokine-mediated signaling is tightly regulated to avoid overshooting activities. Here,
`molecular mechanisms of inter-familiar cytokine cross-talk are reviewed which regulate dynamics and
`strength of IL-6 signal transduction. Both plasticity and cytokine cross-talk are significantly involved in
`pro- and anti-inflammatory/regenerative properties of IL-6-type cytokines. Furthermore, we focus on IL-
`6-type cytokine/cytokine receptor plasticity and cross-talk exemplified by the recently identified
`composite cytokines IL-30/IL-6R and IL-35, the first inter-familiar IL-6/IL-12 family member. The
`complete understanding of the intra- and extracellular cytokine networks will aid to develop novel tailor-
`made therapeutic strategies with reduced side effects.
`
` 2012 Elsevier Ltd. All rights reserved.
`ß
`
`1.
`2.
`3.
`4.
`5.
`6.
`7.
`8.
`9.
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`IL-6-type cytokines, their receptors and main signaling pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Plasticity and intra-familiar cross-talk of IL-6-type cytokines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`IL-35: the first shared IL-6/IL-12 family cytokine
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Inter-familiar cross-talk of the major early pro-inflammatory cytokines IL-6, IL-1 and TNFa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Inter-familiar cross-talk between IL-6 and glucocorticoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Inter-familiar cross-talk of IL-6 and G-protein-coupled receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Inter-familiar cross-talk of IL-6 and receptor tyrosine kinase signaling: taming insulin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Conclusions and future directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`85
`86
`88
`88
`90
`91
`92
`93
`94
`94
`94
`
`1. Introduction
`
`Cytokines are released by a huge variety of different cell types
`and generally act locally as auto- and paracrine factors. Their
`functions are
`involved
`in multiple physiological processes,
`including differentiation, proliferation, migration and apoptosis
`of their target cells [1–4]. Therefore, their ability to signal and
`activate other cells must be tightly regulated, since uncoordinated
`
`* Corresponding author. Tel.: +49 211 8112724; fax: +49 211 8112726.
`E-mail address: jscheller@uni-duesseldorf.de (J. Scheller).
`
` 2012 Elsevier Ltd. All rights reserved.
`1359-6101/$ – see front matter
`http://dx.doi.org/10.1016/j.cytogfr.2012.04.001
`
`ß
`
`cytokine signaling correlates with several pathophysiological
`states,
`including acute and chronic
`inflammatory diseases,
`neoplastic disorders, cancer and autoimmune diseases [5].
`In a simplistic view, any cytokine binds to a unique cytokine
`receptor pair, leading to the activation of intracellular signaling
`pathways on target cells. Receptor proteins and major signaling
`pathways of IL-6-type cytokines have been investigated over the
`last two decades (reviewed
`in [6,7]). Most data have been
`generated under controlled experimental conditions using
`defined cell lines stimulated with single cytokines. This reduc-
`tionistic experimental approach was extremely successful to
`identify the basic signaling pathways, but when combined,
`
`Lassen - Exhibit 1019, p. 1
`
`
`
`86
`
`C. Garbers et al. / Cytokine & Growth Factor Reviews 23 (2012) 85–97
`
`cytokines, hormones and other stimuli might have additive,
`synergistic or antagonistic effects, a phenomenon referred to as
`cross-talk. In vivo, cross-talk is more likely the rule but the
`exception, since each cell has constantly to react to a variety of
`factors at almost
`the same
`time.
`In
`line with
`this,
`the
`reductionistic experimental approach was recently challenged
`by several studies dealing with cytokine cross-talk. Importantly, it
`was found that cytokine cross-talk does not only exist at the level
`of interfering signal transduction pathways but also at the level of
`cytokines and their receptors. Here, plasticity within cytokine/
`cytokine receptor binding sites allows, to a certain extent, that a
`cytokine can use different receptor combinations. Very little is
`known about how a cytokine discriminates between different
`options, but we feel that this decision will be driven by affinity and
`receptor abundance.
`In this review, we focus on intra- and interfamiliar IL-6-type
`cytokine cross-talk. Intra-familiar IL-6-type cytokine cross-talk
`was described for several cytokines of the IL-6-type family and is
`based on binding site/interface plasticity of the cytokine to its
`receptor(s).
`Inter-familiar
`IL-6-type cross-talk was
`recently
`highlighted by studies elucidating the connection between IL-6
`and IL-1 signaling pathways. Sections discussing the current
`knowledge of IL-6-type cytokine cross-talk are preceded by an
`introduction to IL-6-type cytokines, their basic receptor composi-
`tion and major signaling pathways. We hope that this review leads
`to a better perception of the emerging world of cytokine/cytokine
`receptor cross-talk and plasticity not only within the IL-6-type
`family of cytokines.
`
`2. IL-6-type cytokines, their receptors and main signaling
`pathways
`
`All cytokines of the IL-6 family belong to the four-helical bundle
`cytokine family. These proteins share little sequence homology
`and were only recognized to be related by prediction and analysis
`of their protein fold [8]. Interestingly, the four-helical bundle fold
`with an up–up–down–down topology has so far only been found in
`mediator proteins such as interleukins, cytokines and growth
`factors [9].
`Cytokines signal through a specific combination of b-receptors
`with supporting a-receptors to increase target cell specificity and
`to prevent unwanted cellular activation. Cytokine receptors for
`cytokines of the
`IL-6
`family belong to the
`immunoglobulin
`superfamily [8,9]. Two different types of cytokine receptors are
`distinguished. Short chain a-receptors consisting of 3 immuno-
`globulin-/fibronectin-like domains bind their cognate ligands and
`present
`these
`to
`long chain b-receptors consisting of 5–8
`immunoglobulin (Ig)-like/fibronectin-like domains. These
`long
`chain b-receptors
`form homo- or heterodimers and
`initiate
`intracellular signaling [10]. All cytokines of the IL-6 family signal
`via the long chain signal transducing b-receptor glycoprotein
`130 kDa (gp130). As shown in Fig. 1A, IL-6 and IL-11 use a
`homodimer of gp130 whereas all other members of the IL-6 family
`use heterodimers of gp130 and one of the other b-receptors
`leukemia inhibitory factor receptor (LIFR), oncostatin M receptor
`(OSMR) or WSX-1 (IL-27R). A somewhat special case
`is the
`cytokine IL-31, which binds to a heterodimer of OSMR and the IL-
`31 receptor A (IL-31RA) also known as gp130-like protein (GPL)
`[11]. Some cytokines need additional non-signaling a-receptors to
`bind to their b-receptors (Fig. 1A). These a-receptors are either
`membrane-bound such as IL-6R for IL-6, IL-11R for IL-11 and ciliary
`neurotrophic factor receptor (CNTFR) for CNTF or cardiotrophin-
`like cytokine (CLC), or are soluble such as Epstein-Barr virus
`induced gene 3 (EBI3) for IL-30 (p28) to form the composite
`cytokine IL-27 [12]. The cytokines OSM, LIF and IL-31 do not need
`
`an additional a-receptor. It is still not clear if cardiotrophin-1 (CT-
`1) needs an a-receptor [13] (Fig. 1A).
`Except for IL-31, all IL-6-type cytokines bind to the b-receptor
`gp130, indicating a high degree of binding site plasticity within
`gp130 and the respective cytokines. Crystal structures of the
`cytokine/receptor complex have been solved for IL-6 in complex
`with IL-6R and gp130 and for the CBM of gp130 in complex to LIF
`[14,15] (Fig. 1B). Binding of IL-6-type cytokines is mainly mediated
`by ionic and hydrophobic interactions [16], however, inspection of
`the electrostatic surface potential of site II of IL-6, CNTF, LIF and
`OSM (constituted by helices A and C) and the cytokine binding
`module (CBM) of gp130 (domains 2 and 3) reveals almost no
`similarity (Fig. 1B). Per definition, site II and site III are binding sites
`within the cytokine which interact with the b-receptors, whereas
`site I is the binding site for the a-receptor (Fig. 1C and D).
`Therefore, the interaction between site II of the respective cytokine
`and the CBM of gp130 has been described as a chemical plasticity
`resulting in a slightly different orientation of the cytokine and the
`CBM of gp130 [17].
`Gp130 is expressed on all cells whereas the other receptor
`subunits show a more restricted expression profile [5]. Since no
`cytokine of the IL-6 family can signal via gp130 alone, it follows
`that the expression of the additional a- and b-receptor subunits
`determines whether a given cell will be responsive to the cytokine
`[5]. This is illustrated by the two cytokines IL-6 and IL-11, both
`signaling through a homodimer of gp130 plus additional specific
`a-receptors (Fig. 1A). In this case, expression of the a-receptors IL-
`6R and IL-11R determines responsiveness of a target cell. The IL-6R
`is mainly expressed on hepatocytes and some leukocyte popula-
`tions, including macrophages, monocytes, neutrophils, B- and T-
`cells [18], whereas the IL-11R has been detected on lymphocytes,
`B-cells, macrophages, endothelial cells, hematopoietic cells and
`osteoclasts [19]. These data, however, did not include quantitative
`analysis of a-receptor expression, and it is not possible to predict, if
`a cell, expressing both a-receptors, reacts more efficiently to IL-6
`or IL-11. Moreover, comprehensive data covering all receptors and
`cell types are still missing, and it is so far unclear whether cells can
`adapt their receptor expression profile,
`for example during
`pathophysiologic conditions. Examples of activation-dependent
`receptor presentation were shown for CD4 T cells, which lose
`membrane-bound IL-6R during activation via a process called
`ectodomain shedding mediated by ADAM proteases [20,21] and for
`hepatic cells or fibroblasts, which internalize gp130 rapidly in
`response to IL-1b [22].
`In case of IL-6, an additional mechanism for cellular activation
`has been recognized. Besides the membrane-bound
`IL-6R, a
`soluble form of the IL-6R (sIL-6R) is present in many body fluids.
`The sIL-6R is generated by limited proteolysis by ADAM10 or
`ADAM17 of the membrane-bound IL-6R and to a lesser extent by
`translation from an alternatively spliced mRNA in which the exon
`coding for the transmembrane region has been skipped [18]. The
`sIL-6R can still bind its ligand IL-6 and the complex of IL-6 and sIL-
`6R interacts with gp130 and thus activates cells which do not
`express IL-6R and are therefore unresponsive to IL-6 alone. This
`signaling paradigm has been called IL-6 trans-signaling [23],
`whereas signaling via the membrane-bound IL-6R is referred to as
`classic signaling [18] (Fig. 1A). The complex of IL-6 and sIL-6R can
`stimulate all cells of the body and thereby mimic signals from
`other cytokines of the IL-6 family [5]. Interestingly, IL-6 trans-
`signaling is mimicked by the viral ortholog of IL-6 (vIL-6) encoded
`by the human herpes virus 8 (HHV8), which directly binds to and
`stimulates gp130 and does not depend on IL-6R [24,25].
`In
`principle, the soluble forms of the other two membrane-bound a-
`receptors CNTFR and IL-11R can also form biologically active
`soluble complexes, CNTF/sCNTFR [26] and IL-11/sIL-11R [27]. It is,
`however, unlikely that trans-signaling of CNTF and IL-11 via
`
`Lassen - Exhibit 1019, p. 2
`
`
`
`C. Garbers et al. / Cytokine & Growth Factor Reviews 23 (2012) 85–97
`
`87
`
`Fig. 1. Receptor complexes and major JAK/STAT signaling pathways of the IL-6-type family of cytokines. (A) All cytokines of the gp130-signaling family use gp130 (red) as at
`least one part of the signal-transducing b-receptor complex. Some cytokines need additional b-receptor molecules directly involved in signal transduction into the cell (LIF-R
`(dark orange), OSM-R (orange), WSX-1 (wheat), GPL (dark red) or non-signaling a-receptors (IL-6R (blue), IL-11-R (green), CNTF-R (dark grey), EBI3 (grey)). Cytokines of this
`family are IL-6 (dark blue), IL-11 (green), CNTF (black), CLC (black), CT-1 (grey), OSM (grey), LIF (grey), IL-27 (light grey) and IL-31 (lightest grey). CLC may require an
`additional a-receptor. The dominant STAT factor activations are marked in bold, however, this general assignment might not be true for all cells. (B) Comparison of the
`binding site II of IL-6-type cytokines IL-6, CNTF, LIF and OSM (yellow circle) and the complex of IL-6 (site II) with gp130 (cytokine binding module (CBM) domains 2–3, yellow
`circle). (C) Structure of IL-6. The binding sites of IL-6 to the IL-6R (site I) and to gp130 (site II and site III) are indicated. (D) Schematic assembly of the hexameric gp130-
`signaling complex. Sites I, II and III of IL-6 are indicated with I, II, III. Site I is in contact with IL-6R, site II with gp130-CBM and site III with gp130-Ig (D1).
`
`sCNTFR and sIL-11R is relevant in vivo, since no naturally occurring
`sCNTFR and sIL-11R were found so far. On the other hand, IL-30
`signals as composite cytokine IL-27 only via the trans-signaling
`mechanism, since EBI3 is only expressed as a soluble protein.
`It was shown that a soluble form of gp130 (sgp130) is the
`natural
`inhibitor of
`IL-6 trans-signaling
`[28,29]. Therefore,
`recombinant sgp130 is used as a molecular tool to distinguish
`between classic- and trans-signaling. IL-6 is known to exhibit pro-
`inflammatory and regenerative activities [30]. For example, IL-6 /
` mice are more susceptible in an inflammatory bowel disease
`model as compared to wild type mice [31] but protected in mouse
`models of rheumatoid arthritis [32,33]. It turned out that for pro-
`inflammatory signals,
`IL-6 acts via the sIL-6R whereas the
`regenerative activities of IL-6 and the induction of the hepatic
`acute-phase response are induced via the membrane-bound IL-6R
`[30]. This has led to the evaluation of an Fc-fusion protein of
`sgp130 (sgp130Fc) as a therapeutic principle to neutralize the
`
`pro-inflammatory activities of IL-6, which would not compromise
`other beneficial activities of IL-6 [34,35]. Importantly, sgp130Fc
`does not inhibit the signaling of the other IL-6-type cytokines [36].
`First clinical trials with the sgp130Fc protein are planned for 2012.
`Despite a molar excess of sIL-6R over IL-6, free IL-6 and IL-6 in IL-6/
`sIL-6R complexes are present which allows both classic and trans-
`signaling. Under these conditions, sgp130 was, however, able to
`trap all free IL-6 molecules in IL-6/sIL-6R/sgp130 complexes,
`resulting in inhibition of classic signaling by sgp130 [37].
`Finally, binding of IL-6-type cytokines to signal-transducing
`transmembrane b-receptors induces the activation of receptor-
`bound Janus kinases (JAKs) and subsequent signaling pathways
`including signal transducers and activators of transcription (STAT)
`transcription factors, mitogen-activated protein kinases (MAPK),
`in particular the extracellular signal regulated kinases (Erk1/2),
`and phosphoinositide 3-kinase (PI3K). Several control mechanisms
`and negative
`feedback regulators are
`limiting
`IL-6 signaling
`
`Lassen - Exhibit 1019, p. 3
`
`
`
`88
`
`C. Garbers et al. / Cytokine & Growth Factor Reviews 23 (2012) 85–97
`
`pathways at different levels, including receptor internalization,
`expression of suppressors of cytokine signaling proteins (SOCS),
`the action of protein inhibitors of activated STATs (PIAS) and
`protein tyrosine phosphatases (PTP). For a detailed description of
`intracellular signaling of IL-6-type cytokines, we recommend the
`review article by Heinrich et al. [6].
`
`3. Plasticity and intra-familiar cross-talk of IL-6-type cytokines
`
`The well-documented plasticity of gp130 as common b-chain
`receptor of IL-6-type cytokines is not the only example of receptor
`cross-talk within the IL-6 family. OSM can signal via two different
`pairs of b-receptors, namely LIFR/gp130 (type I) and OSMR/gp130
`(type II) [6], which broadens the spectrum of cells that can be
`activated by OSM. This feature seems to be restricted to human OSM,
`since in mice, OSM exhibits only high affinity for the heterodimeric
`receptor pair OSMR/gp130 [38,39]. Interestingly, ongoing studies
`indicate that rat OSM resembles human OSM in signaling via both
`type I and type II receptor complexes (own unpublished results,
`H.M.H.). Plasticity is not restricted to cytokine/b-chain receptor
`interaction. The IL-6R can also interact with different ligands.
`Besides IL-6 [30], the IL-6R was shown to bind to CNTF, albeit at
`lower affinity compared to CNTF/CNTFR interaction. The CNTF/IL-6R
`complex was able to induce the formation of a heterodimer of
`gp130/LIFR for the activation of signal transduction [40] (Fig. 2).
`Another ligand for the IL-6R was recently shown to be IL-30, the IL-
`27 cytokine subunit p28 [41] (Fig. 2). The binding of IL-30 to IL-6R
`leads to
`plus assembly of the b-receptor signaling complex
`activation and phosphorylation of STAT1 and STAT3. Like in the
`case of CLC, the cytokine-like factor-1 (CLF) was needed for efficient
`secretion of IL-30, but not for signaling. During formation of the b-
`receptor signaling complex, CLF is replaced by one of the b-
`receptors. CLF interacts with site III of the cytokines CLC and IL-30. In
`the case of CLC, CLF is replaced by the LIFR upon formation of the
`signaling competent ternary receptor complex [42]. For IL-30, the
`common architecture of the cytokine/b-receptor complex is not
`possible in which site II interacts with CBM of gp130 and site III the
`Ig-like domain of LIFR or OSMR or a second gp130 in case of gp130
`homodimer
`formation. As a consequence,
`IL-27
`(IL-30/EBI3)
`interacts with gp130 via site III and with WSX-1 via site II.
`Therefore, in the case of IL-30/CLF, CLF must be replaced by gp130
`during b-receptor complex formation. The receptor composition of
`IL-30/IL-6R has not been identified [41]. The proposed gp130/WSX1
`heterodimer seemed to be likely, but was recently challenged [43].
`The authors demonstrate that the beneficial effects of IL-30 in the
`treatment of liver injury also occur in WSX-1-deficient mice, making
`the signaling via a gp130/WSX-1 heterodimer impossible. Therefore,
`the signal-transducing b-receptors for IL-30/IL-6R remains to be
`
`determined (Fig. 2). In IL-6 trans-signaling, IL-6 can induce signaling
`via the soluble IL-6R [23,44]. Interestingly, CNTF can also form a
`biologically active complex with sIL-6R but a naturally occurring
`sCNTFR was not described so far (Fig. 2) [40]. It remains to be shown,
`whether IL-30 is also able to form a biologically active composite
`cytokine with the sIL-6R (Fig. 2). This is of importance, since in
`humans about 30–50 ng/ml sIL-6R are present in the circulation. In
`pathophysiological conditions sIL-6R levels rise 2–3fold. Whether
`the biological activity of CNTF/sIL-6R and IL-30/sIL-6R complexes
`would be inhibited by sgp130 was not investigated.
`Since gp130 shows a remarkable plasticity by integrating the
`binding of all IL-6-type cytokines to a single b-receptor, specific
`differences between the signal transduction of IL-6-type cytokines
`must be determined not only by receptor expression pattern but
`also by the accompanying second b-receptor chain. However, since
`some cytokines use the same homo- or heterodimeric receptor
`complexes such as IL-6 and IL-11 or LIF, CNTF, OSM, CLC and CT-1,
`it is still not clear whether the intracellular signaling of these
`cytokines is identical or if the signal transduction of the same b-
`receptor complex is different with respect to the cytokine bound.
`Moreover, it has to be stated that, although the described cross-talk
`of cytokines and cytokine receptors leads to multiple possible
`combination of active b-receptor complexes, cytokine binding
`mainly results in the activation of only a few major signaling
`pathways, which include the aforementioned STAT, MAPK and
`PI3K proteins. It is not clear how this is regulated in space and time
`for dynamic signal transduction. Even though the plasticity of
`cytokine/cytokine receptor interactions is the basis for the broad
`IL-6-type cytokine redundancy, differences in their signal trans-
`duction should be expected and were indeed demonstrated [45].
`
`4. IL-35: the first shared IL-6/IL-12 family cytokine
`
`is a
`family
`IL-6
`the
`latest discovered cytokine of
`The
`heterodimer composed of p35 (cytokine subunit of IL-12) and
`IL-27).
`IL-35 was discovered by
`EBI3 (a-receptor subunit of
`Collison et al. [46], although the ability of the two proteins to form
`a complex and to be secreted together from cells was already
`demonstrated ten years before [47]. IL-35 bridges the IL-6 and the
`IL-12 cytokine families. The IL-12 family consists of the two
`heterodimers IL-12 (p35/p40; cytokine/a-receptor) [48] and IL-23
`(p19/p40) [49]. IL-35 it is the first shared member of the two
`cytokine families and defines an inter-familiar cross-talk (Fig. 3).
`The recently
`identified b-receptors and the specific STAT
`proteins activated by IL-35 are unconventional [50]. IL-35 signals
`via three different pairs of b-receptors, a gp130 homodimer, an
`IL-12Rb2 homodimer or
`a
`gp130/IL-12Rb2 heterodimer.
`Whereas IL-35-induced gp130 homodimer formation leads to
`
`CNTF
`
`
`
`IL-3
`
`0 (p28)
`
`?
`
`CNTFRCNTFR
`
`gp130/LIFR
`
`
`High
`affinity
`
`IL 6RIL-6R
`
`gp130/LIFR
`
`
`sIL 6RsIL-6R
`
`gp130/LIFR
`
`Low
`affinity
`
`gp130/WSX-1
`
`
`EBI3EBI3
`
`
`
`High
`affinity
`
`IL 6RIL-6R
`
`gp130/?
`
`sIL 6RsIL-6R
`
`gp130/?
`
`Low
`affinity
`
`Fig. 2. Plasticity and intra-familiar cross-talk of CNTF and IL-30. CNTF binds to CNTFR (high affinity) and to membrane-bound and soluble IL-6R (low affinity). IL-30 binds to
`EBI3 (high affinity) and to membrane-bound IL-6R and very likely also to the soluble IL-6R. The receptor combination of IL-30/IL-6R might be the same as for IL-27 (IL-30/
`EBI3).
`
`Lassen - Exhibit 1019, p. 4
`
`
`
`C. Garbers et al. / Cytokine & Growth Factor Reviews 23 (2012) 85–97
`
`89
`
`IL-27
`
`IL-35
`
`EBI3
`
`IL-30
`
`(p28(p ))
`
`
`
`
`
`(IL-12( A))
`
`p35
`
`
`
`
`EBI3
`
`IL-12
`
`p35
`
`(IL-12A)(IL 12A)
`
`p40
`
`X-1
`
`gp130/WS
`
`STAT1/
`STAT3/
`STAT5
`
`gp130/gp130
`
`
`
`STAT1
`
`gp130/IL-12Rβ2
`
`
`
`STAT1/
`STAT4
`
`IL-12Rβ2/IL-1
`2Rβ2
`
`STAT4
`
`IL-12Rβ2/IL-12Rβ1
`
`
`
`STAT1/
`STAT3/
`STAT4/
`STAT5
`
`Fig. 3. Receptor composition of the novel IL-6/IL-12 cytokine IL-35. IL-35 binds to and activates three different combinations of receptor complexes, a gp130 homodimer
`(STAT1), a gp130/IL-12Rb2 heterodimer (STAT1/STAT4) and an IL-12Rb2 homodimer (STAT4). For comparison the receptor composition of IL-27 and IL-12 plus main STAT
`factors is given.
`
`the phosphorylation/activation of STAT1, the IL-12Rb2 homodimer
`phosphorylates only STAT4 and the gp130/IL-12Rb2 heterodimer
`phosphorylates both STAT1 and STAT4 leading to the formation of
`phosphorylated STAT1/STAT4 heterodimers (Fig. 3). Since IL-35 is
`composed of subunits from IL-12 and IL-27, it is not surprising that
`also the b-receptors are shared between the families, but the use of
`three different receptor pairings, resulting in the recruitment and
`activation of three different and distinct STAT proteins or STAT
`protein combinations, makes IL-35 a special family member. These
`findings furthermore underline the complex network of cytokine
`subunits and membrane-bound a- and b-receptors that are engaged
`in different combinations to induce biological actions. Since gp130
`homodimer formation after IL-6 or IL-11 binding leads to the strong
`tyrosine phosphorylation of STAT3 and a weaker STAT1 activation, it
`is not trivial to imagine how the same gp130 homodimer after
`binding of IL-35 leads to STAT1 but not to STAT3 phosphorylation.
`Keeping this in mind, a comprehensive comparison of IL-35-, IL-6-
`and IL-11-induced gp130 homodimer formation and signal trans-
`duction will help to unravel the basis of differential signal
`transduction from the same b-receptor complex depending on
`the bound cytokine. It can be speculated that binding of the different
`cytokines to the same receptor complex results
`in different
`orientations which might give access only to a specific set of
`tyrosine motifs in the intracellular region [51].
`Extensive cross-talk and sharing of subunits and receptors
`makes it difficult to address the biological functions of the different
`cytokines in vivo. In the case of IL-35, neither EBI3-deficient nor
`p35-deficient mice represent solely IL-35-deficiency, because EBI3
`is also involved in IL-27 signaling and p35 in IL-12 signaling.
`Therefore the reassignment of data achieved from these mice to
`the function of a single cytokine is impossible. One alternative
`strategy might be to compare p35- with p40-deficient mice. Both
`knock-out mice lack IL-12 signaling, but only p35-deficient mice
`additionally lack IL-35. Unfortunately, also interpretation of data
`from these animal experiments is complicated by cytokine subunit
`cross-talk, since p40 is shared between IL-12 [48] and IL-23 [49],
`making p40-deficient mice unresponsive to both cytokines. Finally,
`also the comparison of EBI3-deficient mice with IL-30-deficient
`mice is not helpful. Several reports have shown that IL-30 has
`functions independently of IL-27 [41,52,53], although the available
`data are conflicting. Shimozato et al. [53] reported that IL-30 alone
`can inhibit IL-27-mediated signaling, whereas Stumhofer et al. [52]
`identified
`IL-30 as a general antagonist of gp130-mediated
`signaling. In contrast, Crabe et al. [41] found that IL-30 is able
`to signal via the IL-6R. The latest report by Dibra et al. [43]
`underlines the agonistic signaling capacities of
`IL-30. They
`
`characterize IL-30 as an anti-inflammatory cytokine capable of
`inhibiting
`inflammation-induced
`liver
`injury, an ability not
`executed by IL-27 (IL-30/EBI3). To date it is unclear how these
`different findings can be included into one model of IL-30 action,
`but it is clear that the listed IL-30 functions, which are independent
`of EBI3, are also absent in IL-30-deficient mice. This makes it
`impossible to dissect the role of IL-35 in vivo by comparing EBI3-
`deficient with IL-30-deficient mice.
`In summary, it is currently not possible to generate conclusive
`IL-35-deficient mice and even data from the aforementioned
`knock-out animals of IL-30, EBI3, p40 have to be interpreted with
`caution due to intensive cytokine/cytokine receptor cross-talk.
`Presenting EBI3-deficient mice as IL-35-deficient mice [55] is
`misleading, since the in vivo roles of IL-35 cannot be dissected from
`IL-27. Time will
`tell whether sophisticated mouse-genetic
`approaches or
`the use of
`inhibitory antibodies, specifically
`targeting the EBI3:p35 heterodimer, will give rise to meaningful
`IL-35-deficient mice.
`Already in 1993, the shared IL-12/IL-23 subunit p40 has been
`shown to be able to inhibit signaling of IL-12 [56]. Later reports
`have proven the existence and antagonistic properties of an IL-
`12p40 homodimer, named IL-12p80 [57,58]. This homodimer is
`able to bind to the IL-12Rb1, thereby antagonizing IL-12-mediated
`signaling by inhibiting binding of IL-12 to its receptor, without
`mediating biologic activities on its own [59]. In vivo, up to one-
`third of the total amount of p40 has been shown to exist as dimeric
`p80 [60]. IL-12p40 can also inhibit IL-23 mediated signaling,
`because IL-23 signals via a heterodimer of IL-12Rb1 and IL-23R.
`[61]. The inhibitory actions of p80 might also contribute to the
`phenotypes seen in p35 and p40-deficient mice, again complicat-
`ing the generation of IL-35-deficient mice.
`Here, we have summarized cross-talk of the extracellular
`signaling network of cytokines, cytokine subunits and soluble
`receptors of the IL-6/IL-12 family. It is tempting to speculate
`that future work will unravel additional roles of single subunits,
`novel heterodimeric cytokines and agonistic or antagonistic
`functions of soluble cytokine receptors. Moreover, less attention
`was paid to the species specificity, since it is not clear whether
`the plasticity of cytokine receptor assembly within the IL-6/IL-
`12 family is the same in men, rats or mice. This is exemplified by
`differences for human and murine OSM [38,39] and by rat CNTF
`which is able to engage signaling via a heterodimer of gp130/
`LIFR in the absence of CNTFR [62], a feature that has not been
`observed for murine or human CNTF. Therefore, one has to be
`cautious with the transfer of findings from mice or rats to the
`human situation.
`
`Lassen - Exhibit 1019, p. 5
`
`
`
`90
`
`C. Garbers et al. / Cytokine & Growth Factor Reviews 23 (2012) 85–97
`
`5. Inter-familiar cross-talk of the major early pro-inflammatory
`cytokines IL-6, IL-1 and TNFa
`
`The duration and strength of IL-6-type cytokine-mediated
`signaling is tightly regulated to avoid overshooting pro-inflamma-
`tory activities. However, also anti-inflammatory activities have to
`be restricted since premature termination of the inflammatory
`response might prevent efficient anti-bacterial or regenerative
`activities.
`In the past decade much attention has been paid on the cross-
`talk elicited by the pro-inflammatory cytokines, IL-1b and TNFa
`with IL-6. Two important proteins were identified as critical
`regulators
`to
`influence
`IL-6-mediated signaling:
`the serine/
`threonine protein kinase p38, a stress-activated MAPK, and the
`transcription factor NF-kB (Fig. 4).
`The liver is one of the most important organs responding to IL-6
`[63]. IL-6 is crucial for the induction of hepatic class I and class II
`acute-phase gene expression and the regeneration of the liver after
`partial hepatectomy. These mechanisms are fundamental to cope
`with infections, toxic stress and traumata [64–66]. Already in the
`late 1980s it was recognized that IL-1b exerts a strong suppressive
`activity on the class II acute-phase protein release initiated by IL-6
`[67]. This complex species- as well as tissue-specific suppressive
`activity depends on at least two activities of IL-1b. An immediately
`acting reduction/prevention of the IL-6-induced STAT3 tyrosine
`phosphorylation [22,68,69] and delayed effects influencing either
`the binding of STAT3 to promoters of target genes or the prolonged
`STAT3 activation [70–72]. The rapid inhibition of early STAT3
`tyrosine phosphorylation was shown to be independent of protein
`de novo synthesis, the activity of tyrosine phosphatases and NF-kB
`activation; however, it required the activity of p38 MAPK [22,69].
`
`Pro-inflammatory cytokines can directly impact on the cell
`surface expression of gp130. It has been shown that gp130
`expression on human primary smooth muscle cells [73] and on
`mast cells [74] can be re