The crystal structure of zebrafish IL-22 reveals an

Genes and Immunity (2014), 1–10 & 2014 Macmillan Publishers Limited All rights reserved 1466-4879/14 www.nature.com/gene

ORIGINAL ARTICLE

The crystal structure of zebrafish IL-22 reveals an evolutionary, conserved structure highly similar to that of human IL-22 P Siupka1, OJ Hamming1, M Fre´taud2,3, G Luftalla4, J-P Levraud2,3 and R Hartmann1 The class II cytokine family consists of small a-helical signaling proteins including the interleukin-10 (IL-10)/IL-22 family, as well as interferons (IFNs). They regulate the innate immune response and in addition have an important role in protecting epithelial tissues. Teleost fish possess a class II cytokine system surprisingly similar to that of humans, and thus zebrafish offers an attractive model organism for investigating the role of class II cytokines in inflammation. However, the evolution of class II cytokines is critical to understand if we are to take full advantage of zebrafish as a model system. The small size and fast evolution of these cytokines obscure phylogenetic analyses based purely on sequences, but one can overcome this obstacle by using information contained within the structure of those molecules. Here we present the crystal structure of IL-22 from zebrafish (zIL-22) solved at 2.1 Å, which displays a typical class II cytokine architecture. We generated a structure-guided alignment of vertebrate class II cytokines and used it for phylogenetic analysis. Our analysis suggests that IL-22 and IL-26 arose early during the evolution of the IL-10-like cytokines. Thus, we propose an evolutionary scenario of class II cytokines in vertebrates, based on genomic and structural data. Genes and Immunity advance online publication, 15 May 2014; doi:10.1038/gene.2014.18

INTRODUCTION The class II cytokine family includes all interferons (IFNs) (types I, II and III) together with interleukin-10 (IL-10) and its related cytokines IL-19, IL-20, IL-22, IL-24 and IL-26. They are encoded by genes that share a similar structure with five exons and four phase 0 introns, with the notable exceptions of type II (gamma) IFN genes, which have lost the last intron, and type I IFN genes in amniotes that contain no introns, and they were most likely created by a retrotransposition event followed by the loss of the original type I IFN gene.1–3 Despite displaying low amino-acid sequence identity within and between species, class II helical cytokines share a similar architecture. Structures have been reported for several human members of this family, namely IL-22, IL-10, IL-19, IL-20, several type I IFNs, type II IFNs and one type III IFN (IFNl3).4–10 They are all a-helical bundle-like proteins with six structural elements named A to F. Elements A, C, D, E and F are a-helices, whereas element B is more variable and contains a b-sheet harpin7,10 in some members. The protein core is formed by four helices (A, C, D and F) in up–up–down–down orientation.10 Type I IFNs in both mammals and fish are structurally distinct from the remainder of the class II cytokines, with the most notable difference being the F helix, which is straight in type I IFNs, while it bends almost 901 in all other class II cytokines.4,7,10,11 An expansion of the class II cytokine system seems to have taken place largely during the Cambrian explosion when vertebrates diverged from basal chordates. In the protochordate Ciona intestinalis, two proteins showing similarities to class II cytokines receptors have been found,2 although the ligands are still unknown. By contrast, the publication of completely or partially sequenced genomes from several fish species resulted in the identification of many class II cytokines in teleosts.2,12–16 Numerous candidate receptor chains have been identified in teleosts, with 16 members in zebrafish alone.17,18 Homologs for most ligands have also been reported, including IL-10, IL-20, IL-22,

IL-26 and IFNg (duplicated in zebrafish and other teleosts), and two groups of virally induced IFNs.2,12,13,15,19–21 Yet, owing to the low level of sequence conservation, the assignment of fish sequences to their mammalian counterparts is not an easy task.17 Most of the evolutionary analyses of class II cytokines were based on amino-acid sequence alignments, but the sequence similarity values between proteins within this family are too low for traditional phylogenetic analysis to give a clear answer regarding their evolutionary relationships.2,17 Thus, although it had been proposed that virally induced fish IFNs may be type III IFNs based on their gene structure and receptor domain organization,14,22 the crystal structures of zebrafish IFNs (zIFNs) proved this wrong: both groups of virally induced fish IFNs are type I IFNs.11 In many cases, the synteny also provided important clues, confirming the orthology of mammalian and fish IL-10-like cytokines (with fish IL-20-like corresponding to a common ancestor of IL-19, IL-20 and IL-24),15 as well as IFNg. Currently, only type III IFNs appear to be missing in fish, and they constitute a missing link in our understanding of class II cytokine evolution. However, as the crystal structure of IFNl3 most closely resembles that of IL-22, among all human class II cytokines, it has been speculated that type III IFNs may have originated within the tetrapod lineage from a duplication of IL-22.8 IL-22 was first described in mice and humans in 2000, and a zebrafish ortholog was described in 2006.12,23,24 In humans, IL-22 is secreted by activated T helper type 22 (Th22), Th1, Th17 and nautral killer (NK) cells and acts on epithelial cells from cutaneous, renal, digestive and respiratory systems.25–27 It binds to a receptor complex composed of the IL-22R1 chain (specific, high-affinity receptor chain) and IL-10R2 chain (shared with other IL-10-like cytokines and type III IFNs, low–affinity receptor chain).28,29 Binding to the receptor complex causes transphosphorylation of Janus kinase 1 and tyrosine kinase 2 kinases associated with the intracellular parts of IL-22R1 and IL-10R2, respectively.30 This leads

1 Department of Molecular Biology and Genetics, Aarhus University, Aarhus, Denmark; 2Macrophages et De´veloppement de l’Immunite´, Institut Pasteur, Paris, France; 3CNRS URA2578, Paris, France; 4UM2, Dynamique des Interactions Membranaires Normales et Pathologiques, Montpellier, France and 5CNRS, DIMNP UMR5235, Montpellier, France. Correspondence: Dr R Hartmann, Department of Molecular Biology and Genetics, Aarhus University, Gustav Wieds vej 10, Aarhus 8000, Denmark Received 3 February 2014; revised 2 April 2014; accepted 7 April 2014

Evolution of class II cytokines in vertebrates P Siupka et al

2 to activation and homodimerization of signal transducer and activator of transcription 3, which translocates to the nucleus and activates the expression of response genes.30–32 Although the main function of IL-22 is induction of anti-microbial defense proteins,26,32–36 it is also involved in wound regeneration and development, and has a role in tumorigenesis25,26,37–43 and antiviral defenses.44–46 Finally, IL-22 is an important factor in the etiology of some autoimmune diseases, such as psoriasis and Crohn’s disease.26,27,47,48 IL-22 has a soluble receptor, named IL22-binding protein, that seems to function as a decoy and thus modulates IL-22 activity.26,41 IL-22 has been identified in zebrafish,12 and despite its low sequence similarity, it was considered to be the ortholog of human IL-22 (hIL-22) owing to a highly conserved synteny with this cytokine (with the MDM1 gene on one side, and IL-26 then IFNg on the other side). Here, we present the three-dimensional structure of zIL-22 resolved by X-ray crystallography. The structure was compared with other known class II cytokine structures from humans and zebrafish. Structurally guided alignment allowed a more meaningful phylogenetic analysis than standard alignment procedures. Structurally speaking, IL-22 appears to belong to the subclade of ‘IL-10-related cytokines’, also including IL-19 and IL-20, but not IFNl, which seems to be slightly closer to type I IFNs despite its bent F helix. Based on these results, we propose a model of class II cytokines evolution in vertebrates. RESULTS Production of biologically active recombinant zIL-22 Recombinant zIL-22 protein was expressed in Escherichia coli, purified by Ni2 þ affinity chromatography under denaturing conditions, refolded in vitro and further purified using hydrophobic interaction chromatography (HIC) and size-exclusion chromatography (SEC). A highly pure protein preparation was obtained by this procedure (Figure 1a). The biological activity was tested by injection of purified IL22 into mpx:GFPi14 transgenic zebrafish larvae. The mpx:GFPi14 transgene is situated on a bacterial artificial chromosome containing 130 kb of the mxp upstream sequences controlling green fluorescent protein (GFP) expression.49 This line has been described as specifically labeling neutrophils.49 The GFP expression in enterocytes had previously been noted during a bacterial infection that also induced il22 expression.50 When injected to mpx:GFP/lysC:DsRED2 fish, the recombinant zIL-22 induced GFP expression specifically in enterocytes (Figure 1b). Enterocytes were clearly recognizable by their shape and location, and because the lysC:DsRED2 transgene also labels neutrophils, we could rule out an attraction of neutrophils to the gut. Thus, recombinant zIL-22 is biologically active and signals to gut epithelial cells, such as mammalian IL-22. Structure of zIL-22 Our highly purified zIL-22 preparation was used for crystallization as described in Materials and methods. The diffraction data were collected at the MAX Laboratory. The cryoprotected zIL-22 crystals diffracted to 2.1 Å and belonged to the primitive (P1) space group with four molecules in the asymmetric unit. The phases were obtained by molecular replacement using modified hIL-22 as a search model. The modifications were based on the zIL-22 protein sequence and include the introduction of a disulfide bridge between helix C and the C terminus (see Materials and methods for details). Detailed statistics for data collection and structure determination are presented in Table 1. The zIL-22 structure is presented in Figure 2a. The electron density map was easily interpretable, and an example of 2FoFc maps with model fitting in the electron density is shown in Figure 2b. All of the four chains present in the asymmetric unit display a highly similar structure, with minor differences in the orientation of solvent-exposed side Genes and Immunity (2014) 1 – 10

chains. Chain B represents the most complete model, and therefore this chain was used for further evolutionary analysis. In chain B, 136 out of 144 amino acids of the zIL-22 structure were built. No interpretable densities were observed for the missing residues: K144 at the C terminus, D27-E30 in the AB loop and S65D66-E67 in the C–D loop. The structure of zIL-22 is typical for class II cytokines with six a-helices (A–F) (Figure 2a) including the four up–up–down–down core helices (A, C, D, F). Helix A is divided into two parts by S10–S11: a short (five amino acids T4–D8) N-terminal part, termed A0 , and a long part of 15 amino acids (A11–N25). Like all IL-10-related cytokines, helix F is bent in the middle. Two disulfide bridges between C48–C141 and C49–C94 are present in the zIL-22 structure. They are connecting helix C with helices F and D, respectively. The structure was deposited in Protein Data Bank under the accession number 4O6K. Comparison of the zIL-22 structure with the structures of other class II cytokines The structure of zIL-22 was compared with those of other class II cytokines from either zebrafish (zIFNj1 and zIFNj2) or humans (hIL-22, hIL-19, hIL-20, IFNl3, hIL-10) using the SSM superimposition tool in the Coot. Examples of this comparison are shown in Figure 3, and the root mean square deviation (RMSD) values are presented in Table 2. The lowest pairwise RMSD value is observed for zIL-22 versus hIL-22 (1.860 Å). The most significant differences between these two sequences are a different position of the short helix B and a longer N terminus before helix A in hIL22. Following hIL-22, the structures of hIL-20 and hIL-19 are most similar to zIL-22 with RMSD values of 2.216 and 2.291 Å, respectively. In the case of zIL-22 versus hIL-20, the most significant differences are in the N terminus of both proteins. The short A0 helix is not present in hIL-20, which has an N-terminal b-hairpin instead. Helix B is either not present or unstructured in hIL-20, and helices D and E are longer than in zIL-22, but this might be conformational changes that result from binding to the receptor complex (Figure 3, top panel, right). The loop that separates the two parts of the hIL-19 helix A is longer (C10–D15) than that of zIL-22 (S10–S11). In both hIL-19 and hIL-22, helix B is located more equidistant from helices E and F. Helix C from hIL-19 is longer than that of zIL-22 (Figure 3, top panel, middle). The similarity of zIL-22 to other zebrafish cytokines is lower than to some of the human class II cytokines. In IFNj2, helix A is divided into two smaller helices at its C-terminal part and not its N-terminal part as in zIL-22. Also, the orientation of this helix is different in IFNj2. Helix F from IFNj2 is not bent, which seems to be a conserved feature among the type I interferon subfamily, and helix C is longer, leading to a slightly different shape of the molecule (Figure 3, bottom panel, right). The differences observed between the superimposed models of zIL-22 and IFNj2 are supported by the RSMD value of 2.663 Å, which is higher than the values with hIL-22, hIL-20 and hIL-19. This strongly indicates that the IL-10-related cytokines and type I IFNs diverged early in evolution before separation of mammals and teleost lineages. Our results indicate that hIL-22 and zIL-22 are true orthologs. Moreover, the close similarity of zIL-22 to hIL-20 and hIL-19 indicates that all these proteins originate from a common ancestor. This is further supported by the analysis of the disulfide bridge pattern in these molecules (Figure 4). zIL-22 has two disulfide bridges next to each other, involving helix C. hIL-20 and hIL-19 have three disulfide bridges, one connecting the N-terminal part of the protein with helix D and two adjacent involving helix C. A closer look at the position of disulfide bridges in superimposed pairs of zIL-22 and hIL-20 or hIL-19, respectively, show that two adjacent bridges are in a similar position in the structures of these cytokines (Figure 4, left and middle). However, they connect different parts of proteins. hIL-22 has two disulfide bridges, one connecting the N-terminal part of protein with the D–E loop and a & 2014 Macmillan Publishers Limited


3

Figure 1. Purification of zIL-22. (a) Sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels of samples from different steps of purification: Ni, supernatant after incubation with Ni-NTA beads; L, bacteria lysate; Wb, wash of Ni-NTA beads; E, protein eluate from Ni-NTA beads; D1, after first dialysis; D2, after final dialysis buffer; Tev þ , after additional incubation with TEV protease; P1, fractions corresponding to the first peak in the HIC chromatograph; GF, fractions corresponding to the peak in the SEC chromatograph; P2, fractions corresponding to second peak in the HIC chromatograph (left). Chromatographs from HIC (top, right) and SEC (bottom, right). (b) Biological activity of recombinant zIL-22. Fluorescence and transmitted light overlay of 4 days postinfection (dpf ) lysC:DsRed2/mpx:gfp double-transgenic larvae injected at 3 dpf with 8 ng of recombinant IL-22 (zIL-22) or 8 ng of bovine serum albumin (BSA) as a control. Gut cells expressing GFP are indicated by an arrow.

second connecting helix C with helix F. The first disulfide bridge from hIL-22 seems to be equivalent to the first disulfide bridge present in hIL-20 and hIL-19 (direct comparison not shown) as & 2014 Macmillan Publishers Limited

they connect similar parts of each protein, whereas the second is equivalent to the first disulfide bridge from zIL-22. This leads us to speculate that the ancestor of IL-22, IL-20 and IL-19 had three Genes and Immunity (2014) 1 – 10


4 disulfide bridges, and that during the evolution of zIL-22 and hIL-22, a different one was lost in each branch. This statement is supported by the sequence alignment of IL-22 proteins from

Table 1.

Statistics of crystallographic data for the zIL-22

Parameter Data collection statistics (MAX Lab) Energy (keV) Space group Unit cell dimensions a, b, c (Å) a, b, g (1) Mosaicity (1) Resolution (Å) No. of unique reflections Redundancy Completeness (%) oI4/os4 Rmerge (%) Refinement statistics Rcryst (%) Rfree (%) Est. coord. error (Å) Est. phase error (1) RMSD from ideal Bond length (Å) Bond angle (1) Wilson B-factor (Å2) No. of atoms No. of water molecules Ramachandran plot Favored (%) Allowed (%) Outliers (%)

different vertebrates. It shows that two adjacent disulfide bridges in helix C can be found only in teleost IL-22. In amphibians, birds and mammals, the hIL-22 disulfide bridge pattern seems to be the rule (Supplementary Figure 1a). In accordance with their putative common origin, in humans, IL-22, IL-20 and IL-19 along with IL-24 and IL-26 share components of their receptor complexes.7,51–53

Value for zIL-22 13.776 P1 46.94, 50.44, 65.8 72.03, 85.94, 62.2 0.157 27.671–2.1 29 007 3.86 (3.42) 98.31 16.67 (3.74) 10 (44.9) 0.222 0.246 0.30 28.74 0.007 1.017 22.98 4392 261 97.49 1.93 0.58

Abbreviations: RMSD, root mean square deviation; zIL-2, zebrafish interleukin-2. Values within parentheses are the values for the highest resolution shell.

Phylogeny of class II cytokines The phylogeny of class II cytokines has been the subject of substantial debate.2,11–17,22 Most of the conclusions were drawn from sequence alignments and subsequent calculation of phylogenetic trees.2,12–16 However, as cytokines are small proteins showing low sequence conservation, the construction of reliable alignments, and thus reliable trees is a challenging task that does not necessarily yield clear answers.2,17 On the other hand, this family is well conserved at the structural level. Inclusion of structural information can lead to better phylogenetic conclusions. Recently, Hamming et al.11 used a structural approach to analyze phylogenetic connections between zj1 and j2 and human class II cytokines, showing that zIFNs belong to the type I IFN subfamily. We used a multiple structure-guided alignment of class II cytokine proteins from humans and zebrafish, generated by the DALI server (Figure 5a). This alignment was used to calculate a phylogenetic tree that is shown in Figure 5b. In this tree, IFNs and IL-10-like cytokines form two separate clades, confirming results obtained previously for IFNs.11 Three algorithms were used to test the robustness of this consensus tree (parsimony, distance and maximum likelihood), and although they yield comparable results, not all algorithms gave significant bootstrap values for all nodes (Figure 5c).

DISCUSSION We produced a recombinant version of the protein considered to be the zebrafish orthologue of hIL-22. This protein was able to activate the mpx promoter in enterocytes of zebrafish larvae, and was thus biologically active and gut-tropic, as would be expected

Figure 2. Structure of zIL-22. (a) Cartoon representation of the structure of zIL-22. Structural elements are labeled A0 through F. Two disulfide bridges are shown in yellow. (b) Example of 2FoFc maps using refined phases along with the refined model. (c) Protein sequence of zIL-22. Helices are marked with red boxes, residues missing in the structure model are marked in gray and schematic representation of disulfide bridges is shown with black lines. Genes and Immunity (2014) 1 – 10

& 2014 Macmillan Publishers Limited


5

Figure 3. Comparison of zIL-22 to humans and zebrafish class II cytokines. Comparison of zIL-22 (red) to hIL22 (PDB; salmon), hIL-19 (PDB; blue), hIL-20 (PDB; green), hIL-10 (PDB; pink), hIL-28b (PDB; pale yellow) and zIFNj1 (PDB; cyan). The molecules have been superimposed in Wincoot (http://www.ysbl.york.ac.uk/~lohkamp/coot/wincoot.html) using the SSM superimpose tool.

Table 2. RMSD of the alignment of selected type II cytokine structures RMSD values for cytokine pair (A˚)

Cytokine

zIL-22 zIFNj1 zIFNj2 hIL-22 hIL-19 hIL-20 IFNl3

zIFNj1

zIFNj2

hIL-22

hIL-19

hIL-20

IFNl3

hIL-10

2.879

2.663 1.787

1.860 3.277 3.090

2.291 2.651 2.832 2.176

2.216 2.741 2.932 2.154 1.274

2.687 3.477 2.999 2.727 3.127 2.966

3.341 3.054 3.509 2.505 1.909 3.251 3.065

Abbreviations: hIL, human interleukin; IFN, interferon; RMSD, root mean square deviation; zIL, zebrafish interleukin.

for an IL-22 ortholog. We solved the crystal structure of zIL-22, which displayed the typical fold of a class II cytokine belonging to the IL-22/IL-10 family. We then used this structural information to gain further insight into the evolutionary origin of class II cytokines. We performed structure-guided alignment of available class II cytokines from humans and zebrafish using the DALI server. Based on this alignment, several evolutionary trees were calculated using different algorithms. The tree topologies were not strictly identical, but in all cases the tree contains two major clades: a type I IFN clade and an IL-10/IL-22-like family clade (Figure 5 and Supplementary Figure 2). This agrees with the observed structures of these proteins as type I IFNs and the IL-22/ IL-10 family are structurally distinct, most prominently displayed by the bend of helix F in the IL-22/IL-10 family. Both type II and type III IFNs are placed outside the two major clades, and they are difficult to assign to any of the two groups purely based upon the tree calculations. However, in all consensus trees, IFNl is closer to the type I IFN clade than to the IL-10 clade. Interestingly, zIL-22 was generally placed at the root of the IL-10-like cytokines clade, suggesting that IL-22 was one of the first proteins to diverge from & 2014 Macmillan Publishers Limited

the common ancestor of the IL-10-like cytokine subfamily. However, the data are not firm on this subject, in particular since we do not have a crystal structure of zIL-10 or IL-26. IL-22 has a key role in the immune response against pathogens as well as wound healing and organ development by acting on epithelial cells. The functions of fish IL-22 are not as well established but appear to be fairly well conserved with an ability to induce anti-microbial peptides,54 a constitutive expression in epithelia exposed to the outside world such as gills,55 and a specific activity on enterocytes (this study). In addition, zebrafish with IL-22 or IL-22R knockdown have gut-developmental problems (MF and GL, unpublished observations). Altogether, this suggests an immune and regulatory function of IL-22 appearing early in vertebrate evolution, which is in agreement with the phylogenetic analysis of IL-10-like subfamily presented by Wang et al.15 However, it is important to stress that the approach we are using is imperfect and that some manual judgment is involved in generating the structural alignment. Nevertheless, we hold the data presented here as a ‘best qualified guess’. We tried to reconstitute the events that led to the present-day class II cytokine family in vertebrates (Figure 6) by joining available data from sequence comparison studies and synteny with our structural analysis. Our model predicts the existence of a primordial class II cytokine, which probably arose during early chordate evolution, before the two rounds of whole-genome duplication (WGD) that shaped the genome of the ancestors of jawed vertebrates.56 Because of the current configurations of loci encoding class II cytokines, a minimal set of genes had to exist in the last common ancestor of teleosts and mammals: (1) one locus containing at least one type I IFN gene—linked to the ahrgap27 gene, which is linked to type I IFN genes in both zebrafish and African clawed frogs (ensemble release 73, scaffold GL172684.1). The link to the ahrgap27 loci is then assumed lost in the retrotransposition event that lead to the intron-less type I IFN genes known in tetrapods, (2) one locus with genes encoding IFNg, IL-26 and IL-22, found in tandem, and linked to the dyrk2 Genes and Immunity (2014) 1 – 10


6 gene, as it is still today in humans and zebrafish, and (3) one locus containing genes encoding IL-10 and an ‘IL-19/-20/-24-like’ cytokine in tandem, linked to the dyrk3 gene. The events that

led to this configuration can then be traced back in time using reasonable assumptions. The link with the dyrk genes suggests that the two last loci share an ancestor, and we propose that they

Figure 4. The ancestor of IL-22 had three disulfide bridges. Representation of disulfide bridges in zIL-22 (yellow) superimposed with hIL-19 (orange; left panel), hIL-20 (orange; middle panel) and hIL-22 (orange; right panel). Structural elements and disulfide bridges have been labeled, see text for details. The molecules were superimposed in WinCoot using the SSM superimpose tool.

Figure 5. Structure-guided alignment of class II cytokines. (a) Multiple alignment of zIL-22 with other class II cytokines with known structures. The alignment was created using the DALI server. Regions corresponding to the helices are indicated by letters A and B above the alignment, and b-strands are marked with green boxes. (b) Phylogenetic tree obtained using a distance method (neighbor-joining) based on the multiple alignment from (a). (c) Consensus cladograms calculated based on the alignment presented in (a), with the IFN-g sequence chosen as an outgroup. For each node, the bootstrap support values (out of 1000 replicates) are given for distance (blue), parsimony (red) and maximumlikelihood (green) methods by the top, middle and bottom numbers, respectively. Genes and Immunity (2014) 1 – 10



7

Figure 6. (a) A model showing the organizations of cytokine genes and their neighbors in the last common ancestor to teleost fish and tetrapods. The model is based upon synteny data. (b) Hypothetical model of class II cytokine evolution in vertebrates, the model is based on structural data, the synteny data as well as on information from available literature (see text for details).

arose during the second vertebrate-specific WGD. Thus, starting from a single primordial class II cytokine, the first WGD generated two ancestral proteins: a non-IFN class II cytokine and a primordial anti-viral IFN (pvi). The nICIIC then underwent tandem duplications, from which IFNg first originated along the ancestor of all IL10-like cytokines, and then this gene was split into an IL-26- and an IL-22-like cytokine. The second WGD led to the duplication and differentiation of the non-IFN class II cytokine into a primordial IL-22/IL-26-like protein and a primordial IL-10/IL-20-like protein. Most likely, IFNg existed before this duplication but was then lost at the third loci. After the second WGD, there would then have been a parallel evolution to create the IL-22/IL-26 family and the IL-10/IL-19/IL-20/ IL-24 family situated on loci 2 and 3, respectively. This is further supported by the analysis presented by Wang et al.,15 where the IL-22/IL-26 branch is separated from the common ancestor of the IL-10 and the IL-20-like subfamily of cytokines (IL-20 SFCs: IL-20, IL19 and IL-24) before each member of this family originated. The ancestral IL-20, IL-26, IL-10 and IL-20 SFCs proteins diverged from these two primordial proteins. Thus, we hypothesize that IL-22, IL26, IL-10 and IL-20 existed before the separation of the teleost and tetrapod lineages. This is in contradiction with a previous model16 that was proposed a decade ago when the available information on fish cytokines was limited. After the separation of piscine and tetrapods lineages, each of the primordial class II cytokines diverged into lineage homologs. In the case of IL-22, a different disulfide bridge was lost in each lineage during that event. & 2014 Macmillan Publishers Limited

Interestingly, IL-19 has been found in birds and mammals and IL-24 in mammals only, and therefore they probably originated after/from the separation of tetrapods and piscine lineages. During the second WGD, the pviIFN gave rise to two types I IFN loci, which are still seen in fish. Primordial type I IFN then underwent a number of duplications and/or retrotransposition events, leading to the number of members observed nowadays and abolishing any synteny. These events clearly occurred independently in both teleost and tetrapod lineages. The origin of type III IFN is, unfortunately, rather unclear. It is situated at the base of our tree based on structure-guided alignment, more or less midway between the type I IFN- and the IL-10-like cytokine clades, and therefore the tree does not provide definitive information as to the origin of this cytokine. Moreover, type III IFN shares a significant structural homology with IL-22 from both humans and zebrafish.8 Furthermore, the genes encoding the IFNlR1 and IL-22R1 receptor chains lie in tandem57 in mammalian genomes, suggesting a recent duplication. Both genes are primarily expressed in epithelial tissues, and together with the IL-10R2 receptor chain, they form the functional type III IFN and IL22 receptor, respectively. Currently, we can trace the type III IFN system as far back as amphibians. All of this leave the possibility open that type III IFN might be a duplication of the IL-22 system and originated in tetrapods. The newly duplicated IFNlR1 chain would have changed its signal transducer and activator of transcription docking site to a signal transducer and activator of transcription 2 docking site to signal via signal transducer and Genes and Immunity (2014) 1 – 10


8 activator of transcription 2. The second option is that type III IFN represents an ancient form of pviIFN, which would also imply that pviIFN had a fold closer to the IL-22/IL-10 family than to type I IFN. However, given the present data, we cannot conclusively differentiate between the two scenarios, and thus more research is needed. Resolving structures of hIL-24 or hIL-26, as well as cytokines, from other species would be very informative. Furthermore, characterization of cytokines from early vertebrates such as sharks or lampreys would be very informative. Based on the comparison of zIL-22 to hIL-22 and literature data on the binding of the latter to its receptor complex,58–61 residues that might be involved in binding to the fish receptor have been proposed (Supplementary Figure 1b). Some of these residues are also conserved at the sequence level of all or almost all IL-22 proteins from different species (Supplementary Figure 1, top panel). It would be interesting to see if they, in fact ,have a role in the binding to the receptor. Unfortunately, so far the receptor complex for zIL-22 protein is not known. There are some candidates among known CRFBs proteins in fish, but this work is still in progress. MATERIALS AND METHODS Protein expression and purification The cDNA of the mature zIL-22 transcript was cloned into the pET-15b 6 His vector, using BamHI and NcoI restriction enzymes (Invitrogen, Carlsbad, CA, USA). N-terminally 6 His-tagged zIL-22 was expressed in Escherichia coli BL21(DE3) bacteria, as described previously.11 Cells were harvested by centrifugation at 4 1C, 4000 r.p.m for 30 min. The bacteria pellets were resuspended in lysis buffer (300 mM NaCl, 50 mM NaH2PO4, 8 M urea, pH 8.0) and homogenized using a high pressure homogenizer. The bacteria lysate was spun down at 4 1C, 12 000 r.p.m., and the supernatant was then incubated with shaking for 1 h at room temperature with 1 ml NiNTA beads per 10 ml supernatant. Afterwards, the Ni-NTA beads were transferred onto a microcellulose column and washed with washing buffer (500 mM NaCl, 50 mM NaH2PO4, 10 mM imidazole, 8 M urea, pH 8.0), and the protein was then eluted from the beads with elution buffer (20 mM Tris-HCl, 100 mM NaCl, 250 mM imidazole, 8 M urea, pH 7.5). Dithiothreitol was added to the sample to a final concentration of 25 mM, and the sample was incubated for 1 h at room temperature. Next, the protein sample was transferred to a dialysis bag and incubated overnight in the first dialysis buffer (50 mM Tris-HCl, 100 mM NaCl, 6 M urea, pH 6.5). The bag was then transferred to the refolding buffer (50 mM HEPES, 240 mM NaCl, 10 mM KCl, 0.5 M L-arginine, 0.75 M guanidine HCl, 0.5% Triton X-100, 0.5 mM glutathione (oxidized), 2 mM glutathione (reduced), pH 7.0) and incubated overnight at 4 1C, which was followed by incubation in the last dialysis buffer (20 mM HEPES, 200 mM NaCl, 5% glycerol, pH 7.0) for 8 h. After that time, 1.5 mg of TEV protease was added to the bag to cleave off the 6 His-tag overnight. After in vitro refolding, the sample was spun down at 4 1C, 8000 r.p.m. for 30 min, and (NH4)2SO4 was added to the supernatant to a final concentration of 1 M. This was followed by incubation on ice for 15 min and centrifugation at 12 000 r.p.m. for 15 min. The sample was loaded onto a HiTrap Butyl FF column (GE Healthcare, Brondby, Denmark) and eluted in a gradient of (NH4)2SO4 using HIC buffer A (20 mM HEPES, 50 mM NaCl, 0.1 mM EDTA, 1 M (NH4)2SO4, pH 7.0) and HIC buffer B (20 mM HEPES, 50 mM NaCl, 0.1 mM EDTA, pH 7.0). The peak fractions were collected, and the sample was concentrated by centrifugation using a 10 000 MWCO filter device (Vivaspin, GE Healthcare). The upconcentrated sample was further purified on a HiLoad 16/60 Superdex 75 column (GE Healthcare) using SEC buffer (20 mM HEPES, 225 mM NaCl, 5% glycerol, pH 7.0). The peak fractions were collected and concentrated as described above. The final protein sample (4.86 mg ml 1) was used for crystallization and biological experiments.

used were Alexa488-labeled goat anti-chicken (1/1000; Invitrogen) and Cy3-labeled goat anti-rabbit (1/500; Jackson Immunoresearch, West Grove, PA, USA). Confocal images of fixed larvae were taken using a Leica SPE inverted confocal microscope (Leica Microsystems, Copenhagen, Denmark) with a 16 (NA 0.5) objective. Images were processed with ImageJ (http://imagej.nih.gov/ij/) and Adobe Photoshop.

Crystallization and structure determination The initial screening was carried out using the Index Screen (Hampton Research, Aliso Viejo, CA, USA) both at 4 1C and at room temperature. Crystals appeared in several conditions at 4 1C, which were subjected to optimization. Crystals were obtained in sitting drop trays, using a seeding technique in 25% PEG 1500, pH 5.6, 3% hexanediol and 25% PEG 1500, pH 5.2, 5% PEG 400 at 4 1C by mixing the protein sample with the crystallization reservoir in a 1:1 ratio. Crystals appeared in 7–8 weeks. These initial crystals were used for seeding, leading to crystal formation after 3 days. After 1 week, crystals were frozen in liquid nitrogen with 30% glycerol as cryoprotectant. The diffraction data were collected at the I911-3 beam line at the MAX Laboratory (Lund, Sweden) at 0.9 Å. Seven hundred and twenty images were collected per data set with a 0.5 degree oscillation. The data were indexed and scaled to 2.1 Å using the XDS software package.62 The crystals belonged to the P1 space group with unit cell lengths a ¼ 46.94 Å, b ¼ 50.44 Å, c ¼ 65.8 Å and angles a ¼ 72.031, b ¼ 85.941, g ¼ 62.21. The phase problem was resolved by molecular replacement using Phenix software.63 hIL-22 (PDB: 1M4R), in which all residues were in silico mutated to alanine, was used as a model. Based on the zIL-22 sequence, a disulfide bridge in helix C was introduced to the model. Aninitial model of the zIL-22 protein was built using AutoBuild from the Phenix software package. The missing residues were added manually using Coot,64 and the model was refined using Refine software from Phenix package.63 The structure was deposited in the Protein Data Bank under accession number 4O6K.

Structure comparison The three-dimensional structures were superposed, using SSM superimpose in Coot,65 and—where available—the amino acid matching was carried out using PyMol.66 All the structures were found in the Protein Data Bank (accession numbers hIL-22: 1M4R; hIL-19: 1N1F; hIL-28B: 3HHC; IFNj1: 3PIV; IFNj2: 3PIW; hIL-10: 2H24). As hIL-10 is an intercalated dimer, where the E and F helices of one subunit occupy the position of the other subunit, an artificial monomer model was generated by keeping regions A–D of one subunit and E and F of the other subunit.

Phylogenetic analysis Automated sequence-based alignment of IL-22 proteins was generated using ClustalX67 and visualized using JalView software.68 The IL-22 protein sequences from different organisms used in the alignment were found in the National Center for Biology Information Protein Database (NCBI accession numbers: zebrafish NP_001018628; African clawed frog NP_001131083; chicken XP_416079; cattle NP_001091849; rat ABF82262; mouse AAI16236; dog XP_53824; human AAQ89249; sheep CCF12129; rainbow trout CAR6817; naked mole-rat EHB08327; crab-eating macaque EHH66472; rhesus macaque NP_001181653; haddock CAN84587; goat ADL28382; Atlantic cod CAR63747). Structure-guided alignment was performed using DALI Lite server.69 As a query for database search, the zIL-22 chain B structure was used. The results were investigated, and protein structures of class II cytokines were chosen, based on Z-score, to generate alignment. One chain of each member was used, PDB accession numbers (chain index is given after slash): hIL-22 1YKB/C; hIL-19 1N1F/A; hIL-20 4DOH/C; hIL-10 1LK3/A; hIL-28B 3HHC/A; zIFNj1 3PIV/B; zIFNj2 3PIW/A; hIFNg 1EKU/B; hIFNa 1ITF/A; hIFNb 1AU1/A. After generating a multiple alignment guided by structure, phylogenetic analysis was performed as described before.11

CONFLICT OF INTEREST Biological activity Eight nanograms of recombinant zIL-22 protein (or bovine serum albumin as a control) were microinjected into the caudal vein of a 3-day-old Tg(lysC:DsRed2)nz50 Tg(mpx:EGFP)i114 double-transgenic zebrafish larvae.49 One day later, anesthesized larvae were fixed and processed for immunochemistry as described.59 Primary antibodies used were chicken polyclonal anti-GFP (1/500; Abcam, Cambridge, UK) and rabbit polyclonal anti-DsRed (1/300; Clontech, Montain View, CA, USA). Secondary antibodies Genes and Immunity (2014) 1 – 10

The authors declare no conflict of interest.

ACKNOWLEDGEMENTS The work was funded by the Danish Cancer Society (Grant: R20-A927 to RH), the Danish Council for Independent Research, Medical Research (Grant 11-107588 to RH) and the French Agence Nationale de la Recherche (Grant ANR-10-MIDI-009 Zebraflam).



9 REFERENCES 1 Pestka S, Krause CD, Walter MR. Interferons, interferon-like cytokines, and their receptors. Immunol Rev 2004; 202: 8–32. 2 Krause CD, Pestka S. Evolution of the class 2 cytokines and receptors, and discovery of new friends and relatives. Pharmacol Ther 2005; 106: 299–346. 3 Qi Z, Nie P, Secombes CJ, Zou J. Intron-containing type I and type III IFN coexist in amphibians: refuting the concept that a retroposition event gave rise to type I IFNs. J Immunol 2010; 184: 5038–5046. 4 Nagem RAP, Colau D, Dumoutier L, Renauld J-C, Ogata C, Polikarpov I. Crystal structure of recombinant human interleukin-22. Structure 2002; 10: 1051–1062. 5 Zdanov A, Schalk-Hihi C, Gustchina A, Tsang M, Weatherbee J, Wlodawer A. Crystal structure of interleukin-10 reveals the functional dimer with an unexpected topological similarity to interferon gamma. Structure 1995; 3: 591–601. 6 Chang C, Magracheva E, Kozlov S, Fong S, Tobin G, Kotenko S et al. Crystal structure of interleukin-19 defines a new subfamily of helical cytokines. J Biol Chem 2003; 278: 3308–3313. 7 Logsdon NJ, Deshpande A, Harris BD, Rajashankar KR, Walter MR. Structural basis for receptor sharing and activation by interleukin-20 receptor-2 (IL-20R2) binding cytokines. Proc Natl Acad Sci USA 2012; 109: 12704–12709. 8 Gad HH, Dellgren C, Hamming OJ, Vends S, Paludan SR, Hartmann R. Interferon-lambda is functionally an interferon but structurally related to the interleukin-10 family. J Biol Chem 2009; 284: 20869–20875. 9 Ealick S, Cook W, Vijay-Kumar S, Carson M, Nagabhushan T, Trotta P et al. Three-dimensional structure of recombinant human interferon-gamma. Science (80-) 1991; 252: 698–702. 10 Walter MR. Structural analysis of IL-10 and type I interferon family members and their complexes with receptor. Adv Protein Chem 2004; 68: 171–222. 11 Hamming OJ, Lutfalla G, Levraud J-P, Hartmann R. Crystal structure of zebrafish interferons I and II reveals conservation of type I interferon structure in vertebrates. J Virol 2011; 85: 8181–8187. 12 Igawa D, Sakai M, Savan R. An unexpected discovery of two interferon gamma-like genes along with interleukin (IL)-22 and -26 from teleost: IL-22 and -26 genes have been described for the first time outside mammals. Mol Immunol 2006; 43: 999–1009. 13 Zou J, Tafalla C, Truckle J, Secombes CJ. Identification of a second group of type I IFNs in fish sheds light on IFN evolution in vertebrates. J Immunol 2007; 179: 3859–3871. 14 Levraud J-P, Boudinot P, Colin I, Benmansour A, Peyrieras N, Herbomel P et al. Identification of the zebrafish IFN receptor: implications for the origin of the vertebrate IFN system. J Immunol 2007; 178: 4385–4394. 15 Wang T, Dıáz-Rosales P, Martin SaM, Secombes CJ. Cloning of a novel interleukin (IL)-20-like gene in rainbow trout Oncorhynchus mykiss gives an insight into the evolution of the IL-10 family. Dev Comp Immunol 2010; 34: 158–167. 16 Lutfalla G, Roest Crollius H, Stange-Thomann N, Jaillon O, Mogensen K, Monneron D. Comparative genomic analysis reveals independent expansion of a lineage-specific gene family in vertebrates: the class II cytokine receptors and their ligands in mammals and fish. BMC Genom 2003; 4: 29. 17 Stein C, Caccamo M, Laird G, Leptin M. Conservation and divergence of gene families encoding components of innate immune response systems in zebrafish. Genome Biol 2007; 8: R251. 18 Aggad D, Stein C, Sieger D, Mazel M, Boudinot P, Herbomel P et al. In vivo analysis of Ifn-g1 and Ifn-g2 signaling in zebrafish. J Immunol 2010; 185: 6774–6782. 19 Altmann SM, Mellon MT, Distel DL, Kim CH. Molecular and functional analysis of an interferon gene from the zebrafish, Danio rerio. J Virol 2003; 77: 1992–2002. 20 Stolte EH, Savelkoul HFJ, Wiegertjes G, Flik G, Lidy Verburg-van Kemenade BM. Differential expression of two interferon-gamma genes in common carp (Cyprinus carpio L.). Dev Comp Immunol 2008; 32: 1467–1481. 21 Robertsen B, Bergan V, Røkenes T, Larsen R, Albuquerque A. Atlantic salmon interferon genes: cloning, sequence analysis, expression, and biological activity. J Interferon Cytokine Res 2003; 23: 601–612. 22 Aggad D, Mazel M, Boudinot P, Mogensen KE, Hamming OJ, Hartmann R et al. The two groups of zebrafish virus-induced interferons signal via distinct receptors with specific and shared chains. J Immunol 2009; 183: 3924–3931. 23 Dumoutier L, Van Roost E, Ameye G, Michaux L, Renauld JC. IL-TIF/IL-22: genomic organization and mapping of the human and mouse genes. Genes Immun 2000; 1: 488–494. 24 Dumoutier L, Louahed J, Renauld J-C. Cloning and characterization of IL-10-related T cell-derived inducible factor (IL-TIF), a novel cytokine structurally related to IL-10 and inducible by IL-9. J Immunol 2000; 164: 1814–1819. 25 Wolk K, Sabat R. Interleukin-22: a novel T- and NK-cell derived cytokine that regulates the biology of tissue cells. Cytokine Growth Factor Rev 2006; 17: 367–380. 26 Wolk K, Witte E, Witte K, Warszawska K, Sabat R. Biology of interleukin-22. Semin Immunopathol 2010; 32: 17–31.


27 Wolk K, Witte E, Warszawska K, Schulze-Tanzil G, Witte K, Philipp S et al. The Th17 cytokine IL-22 induces IL-20 production in keratinocytes: a novel immunological cascade with potential relevance in psoriasis. Eur J Immunol 2009; 39: 3570–3581. 28 Xie MH, Aggarwal S, Ho WH, Foster J, Zhang Z, Stinson J et al. Interleukin (IL)-22, a novel human cytokine that signals through the interferon receptor-related proteins CRF2-4 and IL-22R. J Biol Chem 2000; 275: 31335–31339. 29 Kotenko SV, Izotova LS, Mirochnitchenko OV, Esterova E, Dickensheets H, Donnelly RP et al. Identification of the functional interleukin-22 (IL-22) receptor complex: the IL-10R2 chain (IL-10Rbeta ) is a common chain of both the IL-10 and IL-22 (IL-10-related T cell-derived inducible factor, IL-TIF) receptor complexes. J Biol Chem 2001; 276: 2725–2732. 30 Lejeune D, Dumoutier L, Constantinescu S, Kruijer W, Schuringa JJ, Renauld J-C. Interleukin-22 (IL-22) activates the JAK/STAT, ERK, JNK, and p38 MAP kinase pathways in a rat hepatoma cell line. Pathways that are shared with and distinct from IL-10. J Biol Chem 2002; 277: 33676–33682. 31 Wolk K, Kunz S, Witte E, Friedrich M, Asadullah K, Sabat R. IL-22 increases the innate immunity of tissues. Immunity 2004; 21: 241–254. 32 Wolk K, Witte E, Wallace E, Do¨cke W-D, Kunz S, Asadullah K et al. IL-22 regulates the expression of genes responsible for antimicrobial defense, cellular differentiation, and mobility in keratinocytes: a potential role in psoriasis. Eur J Immunol 2006; 36: 1309–1323. 33 Wilson MS, Feng CG, Barber DL, Yarovinsky F, Cheever AW, Sher A et al. Redundant and pathogenic roles for IL-22 in mycobacterial, protozoan, and helminth infections. J Immunol 2010; 184: 4378–4390. 34 Bao W, Jin L, Fu H-J, Shen Y-N, Lu G-X, Mei H et al. Interleukin-22 mediates early host defense against Rhizomucor pusilluscan pathogens. PLoS ONE 2013; 8: e65065. 35 Ota N, Wong K, Valdez Pa, Zheng Y, Crellin NK, Diehl L et al. IL-22 bridges the lymphotoxin pathway with the maintenance of colonic lymphoid structures during infection with Citrobacter rodentium. Nat Immunol 2011; 12: 941–948. 36 Zheng Y, Valdez Pa, Danilenko DM, Hu Y, Sa SM, Gong Q et al. Interleukin-22 mediates early host defense against attaching and effacing bacterial pathogens. Nat Med 2008; 14: 282–289. 37 Jiang R, Tan Z, Deng L, Chen Y, Xia Y, Gao Y et al. Interleukin-22 promotes human hepatocellular carcinoma by activation of STAT3. Hepatology 2011; 54: 900–909. 38 Zhang F, Shang D, Zhang Y, Tian Y. Interleukin-22 suppresses the growth of A498 renal cell carcinoma cells via regulation of STAT1 pathway. PLoS ONE 2011; 6: e20382. 39 Curd LM, Favors SE, Gregg RK. Pro-tumour activity of interleukin-22 in HPAFII human pancreatic cancer cells. Clin Exp Immunol 2012; 168: 192–199. 40 Thompson CL, Plummer SJ, Tucker TC, Casey G, Li L. Interleukin-22 genetic polymorphisms and risk of colon cancer. Cancer Causes Control 2010; 21: 1165–1170. 41 Huber S, Gagliani N, Zenewicz La, Huber FJ, Bosurgi L, Hu B et al. IL-22BP is regulated by the inflammasome and modulates tumorigenesis in the intestine. Nature 2012; 491: 259–263. 42 Paget C, Ivanov S, Fontaine J, Renneson J, Blanc F, Pichavant M et al. Interleukin-22 is produced by invariant natural killer T lymphocytes during influenza A virus infection: potential role in protection against lung epithelial damage. J Biol Chem 2012; 287: 8816–8829. 43 Ing WX, Ou MZ, Iu SL, Tao XU, Ang JW, Dong-gang XU. Interleukin-22 protects against acute alcohol-induced hepatotoxicity in mice. Biosci Biotechnol Biochem 2011; 75: 1290–1294. 44 Misse´ D, Yssel H, Trabattoni D, Oblet C, Lo Caputo S, Mazzotta F et al. IL-22 participates in an innate anti-HIV-1 host-resistance network through acute-phase protein induction. J Immunol 2007; 178: 407–415. 45 Zhang Y, Cobleigh Ma, Lian J-Q, Huang C-X, Booth CJ, Bai X-F et al. A proinflammatory role for interleukin-22 in the immune response to hepatitis B virus. Gastroenterology 2011; 141: 1897–1906. 46 Ivanov S, Renneson J, Fontaine J, Barthelemy A, Paget C, Macho Fernandez E et al. Interleukin-22 reduces lung inflammation during influenza A virus infection and protects against secondary bacterial infection. J Virol 2013; 87: 6911–6924. 47 Belle AB, Van, Heusch M, De, Lemaire MM, Hendrickx E, Warnier G, Fouser LA et al. IL-22 is required for imiquimod-induced psoriasiform skin inflammation in mice. J Immunol 2013; 188: 462–469. 48 Pan H-F, Li X-P, Zheng SG, Ye D-Q. Emerging role of interleukin-22 in autoimmune diseases. Cytokine Growth Factor Rev 2013; 24: 51–57. 49 Renshaw Sa, Loynes Ca, Trushell DMI, Elworthy S, Ingham PW. Whyte MKB. A transgenic zebrafish model of neutrophilic inflammation. Blood 2006; 108: 3976–3978. 50 Rendueles O, Ferrie`res L, Fre´taud M, Be´gaud E, Herbomel P, Levraud J-P et al. A new zebrafish model of oro-intestinal pathogen colonization reveals a key role for adhesion in protection by probiotic bacteria. PLoS Pathogen 2012; 8: e1002815.

Genes and Immunity (2014) 1 – 10


10 51 Wang M, Liang P. Interleukin-24 and its receptors. Immunology 2005; 114: 166–170. 52 Trivella DBB, Ferreira-Juńior JR, Dumoutier L, Renauld J-C, Polikarpov I. Structure and function of interleukin-22 and other members of the interleukin-10 family. Cell Mol Life Sci 2010; 67: 2909–2935. 53 Donnelly RP, Sheikh F, Dickensheets H, Savan R, Young Ha, Walter MR. Interleukin-26: an IL-10-related cytokine produced by Th17 cells. Cytokine Growth Factor Rev 2010; 21: 393–401. 54 Monte MM, Zou J, Wang T, Carrington A, Secombes CJ. Cloning, expression analysis and bioactivity studies of rainbow trout (Oncorhynchus mykiss) interleukin-22. Cytokine 2011; 55: 62–73. 55 Corripio-Miyar Y, Zou J, Richmond H, Secombes CJ. Identification of interleukin-22 in gadoids and examination of its expression level in vaccinated fish. Mol Immunol 2009; 46: 2098–2106. 56 Kasahara M. The 2R hypothesis: an update. Curr Opin Immunol 2007; 19: 547–552. 57 Dumoutier L, Lejeune D, Hor S, Fickenscher H, Renauld J-C. Cloning of a new type II cytokine receptor activating signal transducer and activator of transcription (STAT)1, STAT2 and STAT3. Biochem J 2003; 370: 391–396. 58 Jones BC, Logsdon NJ, Walter MR. Structure of IL-22 bound to its high affinity IL-22R1 chain. Structure 2009; 16: 1333–1344. 59 Logsdon NJ, Jones BC, Allman JC, Izotova L, Schwartz B, Pestka S et al. The IL-10R2 binding hot spot on IL-22 is located on the N-terminal helix and is dependent on N-linked glycosylation. J Mol Biol 2004; 342: 503–514. 60 Wu PW, Li J, Kodangattil SR, Luxenberg DP, Bennett F, Martino M et al. IL-22R, IL-10R2, and IL-22BP binding sites are topologically juxtaposed

61

62 63

64 65

66 67 68

69

on adjacent and overlapping surfaces of IL-22. J Mol Biol 2008; 382: 1168–1183. Bleicher L, de Moura PR, Watanabe L, Colau D, Dumoutier L, Renauld J-C et al. Crystal structure of the IL-22/IL-22R1 complex and its implications for the IL-22 signaling mechanism. FEBS Lett 2008; 582: 2985–2992. Kabsch W. Xds. Acta Crystallogr D 2010; 66: 125–132. Adams PD, Afonine PV, Bunkoćzi G, Chen VB, Davis IW, Echols N et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D 2010; 66: 213–221. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D 2004; 60: 2126–2132. Krissinel E, Henrick K. Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallogr D 2004; 60: 2256–2268. Schrodinger LLC. The PyMOL Molecular Graphics System, Version 1.3r1, Schrodinger LLC: Portland, OR, USA, 2010. Larkin Ma, Blackshields G, Brown NP, Chenna R, McGettigan Pa, McWilliam H et al. Clustal W and Clustal X version 2.0. Bioinformatics 2007; 23: 2947–2948. Waterhouse AM, Procter JB, Martin DMa, Clamp M, Barton GJ. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 2009; 25: 1189–1191. Holm L, Rosenstro¨m P. Dali server: conservation mapping in 3D. Nucleic Acids Res 2010; 38: W545–W549.

Supplementary Information accompanies this paper on Genes and Immunity website (http://www.nature.com/gene)

Genes and Immunity (2014) 1 – 10
