The IL-10 family's history provides insight into the use of, and need for, new topological maps and sequences for the identification of new molecule superfamilies.
As part of an effort to better understand their relationships and roles in biology, various classification schemes have been devised which create groups, classes, families and superfamilies. While the criteria that distinguish a family from a superfamily are not always clear, the underlying basis for most classifications is 3-dimensional structure, function, or gene organization. Functional designations (e.g. proinflammatory, chemotactic) often give rise to loose confederations of structurally unrelated molecules. Structural designations, by contrast, are based on motifs and topologies that transcend both function and taxonomy. The cystine knot superfamily, for example, has two disulfide bond-generated loops through which passes a third disulfide bond. In plant enzymes, the first, second, fourth and fifth cysteines create loops; in animal growth factors, the second, third, fifth and sixth cysteines generate loops.1, 2
Cytokines have generally been placed in one of seven protein fold-based superfamilies. b-trefoil (cloverleaf) family, while the TNF superfamily demonstrates a "beta-jellyroll" type fold. The cystine-knot superfamily includes PDGF, NGF, glycoprotein hormones and TGF-beta related molecules (activins, GDFs, BMPs). A number of molecules are also defined as being "alpha-helical", and within this grouping are at least three subtypes: those with short alpha-helices [8-10 amino acids (aa), i.e. IL-2, IL-3 and IL-4]; alpha-helices (10-20aa, i.e. Epo, GH and gp130 cytokines); and those with multiple repeat alpha-helices (i.e. IFN-gamma and IL-5).4 In general, a-helix cytokines contain four alpha-helices. IFN-gamma and IL-5, however, contain eight. IL-10 contains six alpha-helices (both long and short).5 Based on this pattern, sophisticated bioinformatic algorithms have been developed to facilitate EST searching for IL-10-like molecules in various databases. In 2000, a new cytokine was a-helical profile similar to IL-10.6 Although it had limited overall aa homology to IL-10 (21%), 80% of the 50 aa that generate IL-10's alpha-helical structure were conserved. This molecule has subsequently been isolated and named IL-19.6 In 2001, a second IL-10-like molecule (IL-20) was also identified through bioinformatics.7 Based on these topologies and primary sequence, at least three other heretofore unrecognized IL-10 family members have subsequently been identified. These include IL-22/TIL, IL-24/mda-7 and IL-26/AK155 (plus numerous viral homologs).5, 8 Within the context of function, these molecules share little similarity. Within the context of discovery, these molecules are at the cutting edge of technology.
The structure common to all mature IL-10 family members is the alpha-helix. Although no empirical evidence exists, it is suggested that all possess six alpha-helices.5 IL-10 has four cysteines, only one of which is conserved among family members. Since IL-10 demonstrates a V-shaped fold that contributes to its dimerization (Figure 1), it appears that disulfide bonds are not critical to this structure. Amino acid identity of family members to IL-10 ranges from 20% (IL-19) to 28% (IL-20).5, 8
IL-10 was first described as a Th2 cytokine in mice that inhibited IFN-alpha and GM-CSF cytokine production by Th1 cells.9, 10 In mice, IL-10 is a homodimeric, glycosylated polypeptide of 17-21kDa (native/monomeric). It is 178 aa in length with an 18 aa signal sequence (ss) and a 160 aa mature segment.11, 12 As noted, it has six alpha-helices. The first four are N-terminal and are central to the molecule while the last two are C-terminal helices and shared with an adjoining IL-10 molecule, forming a non-covalent homodimer.8, 13 Mouse IL-10 is inactive on human cells.14 Mouse IL-10 shares 85% aa identity with rat IL-10,15 85% aa identity with cotton rat IL-10,16 and 84% aa identity with guinea pig IL-10.17
Human IL-10 is also 178 aa in length with an 18 aa ss and a 160 aa mature segment. Its molecular weight is approximately 18 kDa (monomer). Unlike mouse IL-10, human IL-10 contains no potential N-linked glycosylation site and is not glycosylated.8, 18 It also contains one fewer cysteine residue than mouse IL-10 (4 vs. 5). Both IL-10s show two intrachain disulfide bonds, and form nondisulfide-linked homodimers.19 The length of alpha-helices A -->F in human IL-10 are 21, 8, 19, 20, 12 and 23 aa, respectively.5 As noted, helices A --> D of one monomer noncovalently interact with helices E and F of a second monomer, forming a noncovalent V-shaped homodimer. Functional areas have been mapped on the IL-10 molecule. In the N-terminus, pre-helix A residues #1-9 are involved in mast cell proliferation, while in the C-terminus, helix F residues #152-160 mediate leukocyte secretion and chemotaxis.20 In contrast to mouse, human IL-10 is active on various species including mouse18 and canine cells.21 Mature human IL-10 is 72%, 74%, 76%, 77%, 76%, 71% and 96% aa identical to mouse,11 rat,15 pig,22 guinea pig,17 cotton rat,16 dog21 and monkey IL-10,23 respectively. In human, IL-10 circulates at a concentration of approximately 0.5 pg/mL.24
Rat IL-10 has been cloned and found to be highly orthologous to mouse IL-10. It has a mature length of 160 aa and possesses five cysteines plus two potential N-linked glycosylation sites.15, 25 As with human IL-10, rat IL-10 is active on mouse cells.25 Mature rat IL-10 is 85%, 85% and 84% aa identical to mouse,11 cotton rat16 and guinea pig17 IL-10, respectively.
Cells known to express IL-10 include CD8+ T cells,26, 27 microglia,28 CD14+ (but not CD16+) monocytes,29, 30 Th2 CD4+ cells (mice),31 keratinocytes,32 hepatic stellate cells,33 Th1 and Th2 CD4+ T cells (human),34 melanoma cells,35 activated macrophages,36, 37 NK cells,38 dendritic cells,39 B cells (CD5+ and CD19+)40, 41 and eosinophils.42
IL-19 was originally isolated from an Epstein-Barr virus-transformed human B-cell library.6 It is 177 aa in length and contains an 18 aa ss plus a 159 aa mature segment. The mature region contains six cysteines and two potential N-linked glycosylation sites. When expressed in transfected cells, IL-19 is approximately 35-40 kDa in size (predicted MW = 21 kDa), suggesting extensive glycosylation. There is a potential alternate splice site that adds 38 aa to the N-terminus. The significance of this observation is unknown. If IL-19 is aligned to optimize its alpha-helical configurations, it shares less than 20% aa identity with IL-10. More than 80% of the 50 aa that generate IL-10's dimeric structure are conserved in IL-19, suggesting it likely forms noncovalent dimers. Cells known to secrete IL-19 include monocytes and B cells.6, 43
IL-20 was initially identified in a human keratinocyte library during a bioinformatics search for molecules expressing amphipathic helices. When cloned, it was found to consist of a 176 aa polypeptide that contains a 24 aa ss, a 152 aa mature segment, six cysteines, and no potential N-linked glycosylation sites.7 It is not known if this molecule forms a dimer; there are reports both for and against this configuration.7, 44 There is 26% aa identity between human IL-20 and IL-10, and 41% aa identity between IL-20 and IL-19 with conservation of all six cysteines.6, 7 The IL-20 gene maps to chromosome 1 with IL-10, IL-19 and IL-24/mda-7.7 Mouse and human IL-20 show 76% aa identity in the mature segment.7, 45 Cells known to produce IL-20 are monocytes43 and (perhaps) keratinocytes.7, 46
Human IL-22 is a 179 aa polypeptide that contains a 33 aa ss and a 146 aa mature region.47, 48 The mature molecule contains three potential N-linked glycosylation sites plus four cysteines, only two of which are conserved in IL-10.5 While it is suggested to form a homodimer,44 it can also apparently accommodate a monomeric configuration.5, 8 Thus, its native form is unclear. There is 24% aa identity between mature human IL-22 and IL-10. Mouse IL-22 is 179 aa in length with a 33 aa ss, 146 aa mature segment and four potential N-linked glycosylation sites.49, 50 Its native molecular weight is approximately 25 kDa.49 While there is only one human gene, there are two IL-22 genes (IL-22 alpha and IL-22 beta) in select mouse strains (C57/Bl), which are the result of gene duplication.50 Each gene is 179 aa in length, yet demonstrates differences in aa at positions #36 (Val to Ile), #103 (Val to Ile), and #112 (Gln to Arg). In C57 mice, it is not known if the IL-22 beta gene is functional as it lacks essential promoter elements. In the standard a-form of IL-22, there is also a polymorphism with an Asn substituted for a Ser at position #84.50 There is 81% aa identity in the mature regions of human IL-22 and mouse IL-22 alpha. Cells known to express IL-22 include NK cells and CD4+ Th1 cells.43
IL-24 was originally isolated from actively proliferating melanoma cells.51 It consists of a 207 aa precursor with an extended 47 aa signal sequence and a 160 aa mature segment.52 Contrary to initial reports, it is secreted and has a native molecular weight of approximately 35 kDa.53 Within the mature segment, there are three potential N-linked glycosylation sites and two cysteines. Thus, it differs markedly from all other known IL-10 family members in the potential for disulfide-bonding.51 In its mature segment, it shares 20% aa identity with IL-10. Human IL-24 is active on rat cells. IL-24 in mouse is a 27 kDa, glycosylated polypeptide also known as FISP.54 It is 220 aa in length with an extended 60 aa ss and a 160 aa mature segment. It shares 69% and 84% aa identity with human and rat IL-24, respectively, in its mature segment.52, 54 Rat IL-24 is also known as mob (mobster)-5 and c49a. It was discovered after Ha-ras-activation in cell lines.52 It is 183 aa in length and contains a 23 aa ss plus a 160 aa mature segment.52, 55 As with human IL-24, there are only two cysteines in the mature segment. It is 69% aa identical to human.55 Cells known to express IL-24 include Th2 cells (in mouse),53 melanocytes,51, 55, 56 breast epithelium,56 fibroblasts (in rat),55 monocytes,43, 54 vascular smooth muscle,57 NK cells,58 B cells58 and CD4+ CD45RA+ (naïve) T cells.43
IL-26/AK155 is a 36 kDa homodimer that was initially identified in the supernatant of cultured herpesvirus-transformed T cells.59 Its open reading frame encodes a 171 aa precursor that contains a 21 aa ss and a 150 aa mature segment. Its mature segment shares 25% aa identity with human IL-10 and retains the typical six alpha-helix pattern seen in IL-10. Remarkably, there is no mouse counterpart to IL-26.8 Cells known to express IL-26 include CD4+ CD45RO+ T cells and NK cells.43
The receptors for the IL-10 family are all members of the class II cytokine receptor family (CRF2).53 All functional receptors are heterodimers composed of an alpha or type 1 (R1) chain and alpha beta or type 2 (R2) chain. At a minimum, it would appear that there are at least two alpha and two beta chains involved in each signaling receptor complex.60 Traditional nomenclature has often defined the alpha subunit as being the ligand-binding molecule and the beta subunit as being the signal transducing subunit. For at least one member of the IL-10 family, both subunits can serve as ligand-binding moieties47, 61 and the alpha-subunit in this case generally serves as a STAT3 docking port (i.e. the first step in signal transduction).60 Class II cytokine receptors are characterized by the presence of two fibronectin type III (FNIII) domains in their extracellular region. These typically are 90-100 aa globular modules that consist of seven beta-strands that form two beta-sheets (reminiscent of Ig domains). These domains are typically associated with cell surface adhesion molecules.
View Larger Image
IL-10 R1: Human IL-10R1 is a 90-110 kDa, type I transmembrane glycoprotein that is expressed on a limited number of cell types.63 The open reading frame encodes a 578 aa protein that contains a 21 aa ss, a 215 aa extracellular region, a 25 aa transmembrane segment, and a 317 aa cytoplasmic domain. There are two FNIII motifs within the extracellular region and a STAT3 docking site plus a JAK1 association region within the cytoplasmic domain.60, 63, 64 IL-10R1 binds human IL-10 with a K
IL-20 R1/RA: Human IL-20R1, also known as zcytor7, is a 553 aa type I transmembrane protein of which little is known structurally.66 Originally identified by Zymogenetics, Inc., it has since been identified as one member of the IL-20 signaling complex.7 Overall, it shares approximately 21% aa identity with mature IL-10R1.
IL-22 R/CRF2-9: Human IL-22R, also known as zcytor11, is a 574 aa type I transmembrane protein that contains a 574 aa extracellular region, a 23 aa transmembrane segment, and a 346 aa cytoplasmic domain.47 It shares approximately 15% aa identity overall to IL-10R1, and 17% aa identity within the extracellular region.
Soluble IL-22 R/CRF2-10: A genetically distinct soluble receptor for IL-22 has recently been identified that is 40 kDa in size and 210 aa in length.67, 68, 69, 70 As with the other IL-10R family members, it contains two FNIII motifs. Two alternate splice forms have been noted in addition to the standard 210 aa form. One splice form is truncated and only 131 aa in length.69 The second splice form is extended by 32 aa, creating a 242 aa mature molecule.67, 69, 70 This long form is only found in placenta and may regulate both IL-22 and IL-20 activity. The standard form is strictly an antagonist for IL-22 activity (STAT activation).69 Cells known to make the standard form include plasma cells, monocytes, B cells, CD4+ T cells and type II greater alveolar cells.68
IL-10R2/CRF2-4: Human IL-10R2 is a 60 kDa, 325 aa glycoprotein that contains a 201 aa extracellular region, a 29 aa transmembrane segment, and a 76 aa cytoplasmic domain.71 Its cytoplasmic domain is associated with Tyk2 kinase.64 Relative to human IL-10R1, it shows 16% aa identity overall and 21% aa identity in the extracellular domain. Its designation as CRF2-4 follows that for CRF2-1 (IFN-alpha beta R), CRF2-2 (IFN-gamma receptor) and CRF2-3 (tissue factor receptor). The mouse IL-10R2 molecule is 329 aa in mature length, with a 100 aa cytoplasmic region, a 29aa transmembrane segment, and a 200 aa extracellular domain.72,73 It is 69% aa identical to human IL-10R2 overall, with 75% aa identity in the extracellular region.72
IL-20 R2/RB: IL-20R2 is also known as DIRS1 in the patent literature [Parnam, C.L. (1999) WO 9946379A2]. It is 311 aa in length and contains the typical FNIII domain. Over the entire ORF, there is 28% aa identity, human IL-10R2 to human IL-20R2.74 In combination with IL-20R1, IL-20R2 is a functional part of the IL-20 receptor.7
IL-10: Binds to IL-10R1 / IL-10R2
IL-19: Binds to IL-20R1 / IL-20R2
IL-20: Binds to IL-20R1 / IL-20R2 and IL-22R / IL-20R2
IL-22: Binds to IL-22R / IL-10R2
IL-24: Binds to IL-20R1 / IL-20R2 and IL-22R / IL-20R2
IL-26: No known receptor
It is clear that IL-10 and its family members share common receptors. Often, however, cytokines have distinct, if not antagonistic, functions. As an example, both IL-10 and IL-22 signal through IL-10R2 and each activate JAK1 and TYK2, thus resulting in STAT3 activation.53, 76 IL-22, however, also induces serine phosphorylation of STAT3 (IL-10 does not), an event that is associated with MAP kinase pathway activation.76, 77 This notable difference in signaling is reflected in at least one function of IL-10 and IL-22. IL-10 suppresses both IL-1 and TNF-alpha production by LPS-stimulated monocytes while IL-22 does not.47
IL-10: The volume of information published related to IL-10 function is considerable. In brief, the following information outlines general functions on select cells.
On T cells, the initial observations of IL-10 inhibition of IFN-gamma production is now suggested to be an indirect effect mediated by accessory cells.9, 10 Additional effects on T cells, however, include: IL-10 induced CD8+ T cell chemotaxis,20, 78 an inhibition of CD4+ T cell chemotaxis towards IL-8,20 suppression of IL-2 production following activation,79, 80 an inhibition of T cell apoptosis via Bcl-2 up-regulation,81 and an interruption of T cell proliferation following low antigen exposure accompanied by B7/CD28 costimulation.82
On B cells, IL-10 has a number of related, yet distinct functions. In conjunction with TNF-beta and CD40L, IL-10 induces IgA production in naïve (IgD+) B cells. It is believed that TGF-beta/CD40L promotes class switching while IL-10 initiates differentiation and growth.83, 84 When TGF-beta is not present, IL-10 cooperates with CD40L in inducing IgG1 and IgG3 (human), and thus may be a direct switch factor for IgG subtypes.85 Interestingly, IL-10 has divergent effects on IL-4 induced IgE secretion. If IL-10 is present at the time of IL-4 induced class switching, it reverses the effect; if it is present after IgE commitment, it augments IgE secretion.86 Finally, CD27/CD70 interaction in the presence of IL-10 promotes plasma cell formation from memory B cells.87
Mast cells and NK cells are also impacted by IL-10. On mast cells, IL-10 induces histamine release while blocking GM-CSF and TNF-alpha release. This effect may be autocrine as IL-10 is known to be released by mast cells in rat.88,89 As evidence of its pleiotrophic nature, IL-10 has the opposite effects on NK cells. Rather than blocking TNF-alpha and GM-CSF production, IL-10 actually promotes this function on NK cells.90 In addition, it potentiates IL-2 induced NK cell proliferation and facilitates IFN-?gamma secretion in NK cells primed by IL-18.90, 91, 92 In concert with both IL-12 and/or IL-18, IL-10 potentiates NK cell cytotoxicity.92
IL-10 has a pronounced anti-inflammatory impact on neutrophils. It inhibits the secretion of the chemokines MIP-1 alpha, MIP-1 beta and IL-8,93, 94 and blocks production of the proinflammatory mediators IL-1 beta and TNF-alpha.94 In addition, it decreases the ability of neutrophils to produce superoxide, and as a result interferes with PMN-mediated antibody-dependent cellular cytotoxicity.95 It also blocks IL-8 and fMLP-induced chemotaxis, possibly via CXCR1.96
On dendritic cells (DCs), IL-10 generally exhibits immunosuppressive effects. It would appear to promote CD14+ macrophage differentiation at the expense of DCs.97 Macrophages, while phagocytic, are poor antigen-presenting cells. IL-10 seems to decrease the ability of DCs to stimulate Tcells, particularly for Th1 type cells.98, 99 How IL-10 accomplishes this is unclear, as the data within the literature is conflicting. Relative to MHC-II expression, it can be down-regulated,100, 101 unchanged,99, 102 or up-regulated.98, 101 With respect to B7-1/CD80, IL-10 will either up-regulate98, 101 or down-regulate100, 101, 103 its expression. B7-2/CD86 plays a key role in T cell activation. For this molecule, IL-10 is involved in both up-regulation98, 100 and down-regulation.100, 101, 102 Perhaps the most significant modulation, however, occurs with CD40 (IL-10 seems to reduce its expression).101, 103 At the regional level, IL-10 may block immunostimulation by inhibiting Langerhans cell migration in response to proinflammatory cytokines.104 Alternatively, IL-10 blocks an inflammation-induced DC maturation step that normally involves CCR1, CCR2 and CCR5 down-regulation and CCR7 up-regulation. This blockage, with retention of CCR1, CCR2 and CCR5, results in a failure of DCs to migrate to regional nodes. The result is an immobile DC that will not stimulate T cells but will bind (and clear) proinflammatory chemokines without responding to them.105
On monocytes, IL-10 has a number of documented effects. For example, IL-10 seems to clearly reduce cell surface MHC-II expression.20, 106, 107, 108 It also inhibits IL-12 production following stimulation.109 While it promotes a monocyte to macrophage transition in conjunction with M-CSF,110 the phenotype of the macrophage is not clear (i.e. CD16+/cytotoxic vs. CD16
IL-19: To date, there is no known function for IL-19.8
IL-20: The only suggested function for IL-20 involves keratinocytes. Partly based on overexpression studies, it is proposed that IL-20 induces keratinocyte differentiation and proliferation.7, 46 IL-20R is expressed on keratinocytes and endothelial cells,7 and keratinocytes likely express the IL-20 cytokine.114
IL-22: The functions associated with IL-22 are limited. IL-22 induces acute phase proteins in hepatocytes (i.e. haptoglobin, alpha 1-antichymotrypsin)8, 48 and may reduce IL-4 production by certain Th2 cells.47 In addition, pancreatic acinar cells, which synthesize inactive precursor forms of digestive enzymes, are now known to express PAP1 (pancreatitis-associated protein-1) and osteopontin in response to IL-22.115 These are two molecules that may have protective or trophic effects during inflammation.
IL-24: IL-24/mda-7 was initially believed to induce melanocyte/melanoma cell differentiation. Melanoma cells lack IL-24 expression, while melanocytes express IL-24. In the presence of IL-24, melanoma cell growth was inhibited (G2/M arrest).51, 57 Thus, it appeared that mda-7/IL-24 may be a tumor suppressor molecule.56,116 Subsequent studies have shown that IL-24 can inhibit tumor growth in cells other than melanoma (i.e. non-small cell lung carcinoma, breast cancer, prostate cancer),117, 118, 119 and that apoptosis can be induced in transformed cells without negatively impacting any normal cell types.57, 113, 119 Aside from its ability suppress tumors, IL-24 has also been shown to induce IL-6 and TNF-alpha secretion by monocytes, demonstrating an activity antagonistic to that of IL-10.58
IL-26: To date, almost nothing is known about IL-26 activity.