Functional Variation of IL-1R−Associated Kinases in the Conserved MyD88−TRAF6 Pathway during Evolution

This information is current as Xinyu Yan, Shenghui Chen, Huiqing Huang, Ting Peng, of September 27, 2021. Mengjiao Lan, Xia Yang, Meiling Dong, Shangwu Chen, Anlong Xu and Shengfeng Huang J Immunol published online 8 January 2020 http://www.jimmunol.org/content/early/2020/01/07/jimmun

ol.1900222 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2020/01/07/jimmunol.190022 Material 2.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 27, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2020 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published January 8, 2020, doi:10.4049/jimmunol.1900222 The Journal of Immunology

Functional Variation of IL-1R–Associated Kinases in the Conserved MyD88–TRAF6 Pathway during Evolution

Xinyu Yan,*,1 Shenghui Chen,*,1 Huiqing Huang,† Ting Peng,* Mengjiao Lan,* Xia Yang,* Meiling Dong,* Shangwu Chen,* Anlong Xu,*,‡ and Shengfeng Huang*,x

IL-1R–associated kinases (IRAK) are important regulators in the TLR/IL-1R pathways, but their function appears inconsistent between Drosophila, bony fishes, and vertebrates. This causes a difficulty to understand the IRAK functions. As a step to reveal the evolution of IRAKs, in this study, we performed comparative and functional analysis of IRAKs by exploiting the amphioxus, a pivotal taxon connecting invertebrates and vertebrates. Sequence and phylogenetic analysis indicated three major IRAK lineages: IRAK1/2/3 is a vertebrate-specific lineage, IRAK4 is an ancient lineage conserved between invertebrate and vertebrates, and Pelle is another ancient lineage that is preserved in protostomes and invertebrate deuterostomes but lost in vertebrate deuterostomes.

Pelle is closer neither to IRAK4 nor to IRAK1/2/3, hence suggesting no clear functional analogs to IRAK1/2/3 in nonvertebrates. Downloaded from Functional analysis showed that both amphioxus IRAK4 and Pelle could suppress NF-kB activation induced by MyD88 and TRAF6, which are unlike mammalian and Drosophila IRAKs, but, surprisingly, similar to bony fish IRAK4. Also unlike Dro- sophila IRAKs, no interaction was detected between amphioxus IRAK4 and Pelle, although both of them were shown capable of binding MyD88. These findings, together with previous reports, show that unlike other signal transducers in the TLR/IL-1R pathways, such as MyD88 and TRAF6, the functions of IRAKs are highly variable during evolution and very specialized in

different major animal taxa. Indeed, we suggest that the functional variability of IRAKs might confer plasticity to the signal http://www.jimmunol.org/ transduction of the TLR/IL-1R pathways, which in return helps the species to evolve against the pathogens. The Journal of Immunology, 2020, 204: 000–000.

oll-like receptors are pattern recognition receptors, This pathway is regulated by an important class of kinases, whereas IL-1R are cytokine receptors. They both share a the IL-1R–associated kinases (IRAKs) (6). In mammals, there T TLR/IL-1R (TIR) domain at their cytoplasmic tail, and are four IRAK genes, IRAK1, IRAK2, IRAK3 (or IRAKM), and both have important roles in the host-against-pathogen immune IRAK4. All four IRAKs have similar protein architecture: an responses (1, 2). Both types of receptors receive extracellular sig- N-terminal death domain, a ProST domain, a kinase domain, by guest on September 27, 2021 nals and induce the classic intracellular MyD88–TRAF6–NF-kB and a C-terminal domain (except IRAK4) (6). The death do- pathway (2, 3). Dysregulation of this pathway may cause chronic main is used for interaction with other signal transducers such inflammation and autoimmune diseases, etc (4, 5). as MyD88 and other IRAK members (7). The ProST domain is rich in serines, prolines, and threonines. Phosphorylation in this domain plays a role in regulating IRAKs’ own availability *State Key Laboratory of Biocontrol, Guangdong Key Laboratory of Pharmaceutical Functional Genes, School of Life Sciences, Sun Yat-sen University, 510275 Guangzhou, (8). IRAK1 is reported to undergo hyperphosphorylation in China; †Guangdong Food and Drug Vocational College, 510520 Guangzhou, China; ProST domain, leading to its release from MyD88 complex ‡School of Life Science, Beijing University of Chinese Medicine, 100029 Beijing, China; and xLaboratory for Marine Biology and Biotechnology, Qingdao National Laboratory and bind to TRAF6 (9). The kinase domain apparently confers for Marine Science and Technology, 266003 Qingdao, China kinase activity, which, however, may not be required for some 1X.Y. and S.C. contributed equally to this study. function of IRAKs. For example, IRAK3 lacks kinase activity, Received for publication February 22, 2019. Accepted for publication December 12, and the kinase inactive mutants of IRAK1 and 4 are still able to 2019. activate NF-kB (10–12). The C-terminal domain of IRAK1/2/3 is This work was supported by the National Key R&D Program of China (2018YFD0900503), reportedly important for the interaction with TRAF6 (carrying National Natural Science Foundation Projects 31800729, 31872595, and 1722052, the TRAF6 binding sites) (6, 13). Marine S&T Fund of Shandong Province (2018SDKJ0302-2), projects from Guangdong (201804010434, 201804020039, 2018A030310157, and 2017B030314021), and by In a typical signal transduction in mammals, TLRs/IL-1Rs use the National Supercomputer Center in Guangzhou and the Special Program for intracellular TIR domain to recruit MyD88 (2, 14, 15). Then, Applied Research on Super Computation of the NSFC-Guangdong Joint Fund. MyD88 recruits IRAKs via the death domains to form a signaling Address correspondence and reprint requests to Prof. Anlong Xu and Prof. Shengfeng complex (Fig. 1A) (7, 16). IRAK4 is suggested to be the first one Huang, State Key Laboratory of Biocontrol, Guangdong Key Laboratory of Pharma- ceutical Functional Genes, School of Life Science, Sun Yat-sen University, 135 to be recruited by MyD88, and its autophosphorylation after re- Xingang West Street, Haizhu District, Guangzhou 510275, Guangdong Province, cruitment leads to the subsequent recruitment and phosphorylation People’s Republic of China. E-mail addresses: [email protected] (A.X.) and of IRAK1 (10, 16–19). Hyperphosphorylated IRAK1 is then re- [email protected] (S.H.) leased from the MyD88 complex and interacts with TRAF6, which The online version of this article contains supplemental material. in turn activates NF-kB (20, 21). IRAK2 might have a redundant Abbreviations used in this article: BbIRAK4, B. belcheri IRAK4; BbMyD88, B. belcheri MyD88; BbPelle, B. belcheri Pelle; BbTRAF6, B. belcheri TRAF6; Co-IP, coimmu- function with IRAK1, but new evidence suggests a different role for noprecipitation; DrIRAK, Danio rerio IRAK; DrMyD88, D. rerio MyD88; IRAK, IRAK2 in the TLR/IL-1R signaling (6, 10, 22). In contrast, IRAK3 IL-1R–associated kinase; PBST, 0.05% Tween–20 in PBS; pI, isoelectric point; lacks kinase activity but is capable of suppressing the pathway by qRT-PCR, quantitative real-time PCR. preventing IRAK1/2 from the dissociation from MyD88 and from Copyright Ó 2020 by The American Association of Immunologists, Inc. 0022-1767/20/$37.50 the interaction with TRAF6 (23, 24).

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900222 2 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY

However, there are few studies on IRAKs from lower vertebrates gene and nested PCR were performed to amplify the coding sequence of and invertebrates. But it appears that the composition and function BbIRAK4 and BbPelle. Amplified sequences were inserted into the pGEX-T of the IRAK family could be quite different from mammals (25– easy vector (Promega) and verified by sequencing. The primers were showed in Table I. 32). For example, in lower vertebrates, birds lack IRAK1 and IRAK3, reptiles and amphibians lack IRAK3, and bony fishes lack Bioinformatic analysis IRAK2 (33). Interestingly, several reports suggested that bony fish The domain structure was predicted on the main page of the SMART Web IRAK4 could suppress MyD88-dependent NF-kB activation when site (http://smart.embl-heidelberg.de). BLASTP was performed to analyze transfected into mammalian cells (28, 34, 35), which is in contrast the sequence identities. The isoelectric point (pI) and m.w. were speculated to the activating role of mammalian IRAK4 (Fig. 1A). on ExPASy Web site (http://www.expasy.org/tools/). Multiple sequence alignments were analyzed using ClustalX 1.83 and were manually corrected As about nonvertebrates, whereas arthropods such as Drosophila using GeneDoc. Neighbor-joining tree was built using MEGA5 with 1000 have two IRAK homologs (Tube/IRAK4 and Pelle), it appears that bootstrap tests and handling gaps by pairwise deletion. poriferans and hemichordates have only IRAK4 (29, 33, 36, 37). It is suggested that the IRAK family derives from a single IRAK4- Quantitative real-time PCR like kinase in the last common metazoan ancestor (33, 38). Quantitative real-time PCR (qRT-PCR) was performed to determine the Functions of invertebrate Tube/IRAK4 and Pelle have been stud- mRNA expression profiles of BbIRAK4 and BbPelle in amphioxus. Healthy ied in Drosophila melanogaster. When activated, D. melanogaster adult individuals were chosen for tissue distribution analysis. Tissues (muscle, skin, gill, ovary, hepatic cecum, intestine) were harvested, and total Tube and D. melanogaster Pelle are recruited to D. melanogaster RNA was extracted. qRT-PCR was conducted on a LightCycler 480 instru- MyD88 to form a trimeric complex via the death domains, which ment using SYBR PrimeScript qRT-PCR kit (Takara). Data were quantified 2DD is essential for subsequent signal transduction (39–41). Although using the 2 Ct method based on the cycle threshold values of BbIRAK4, Downloaded from D. melanogaster Pelle and D. melanogaster MyD88 do not di- BbPelle,andGAPDH. All of the samples were analyzed in three replicates. rectly contact with each other, they bind to distinct surface of Primers used for qRT-PCR were listed in Table I. the death domain of Tube (39, 42). Pelle is autophosphorylated Construction of the expression vector and catalyzes the phosphorylation of Cactus, leading to the deg- For the expression of BbIRAK4 and BbPelle in Hela and HEK293T cells, radation of Cactus and the activation of Dorsal or Dif (Fig. 1A) PCR fragments encoding full-length protein sequences of BbIRAK4 and

(43, 44). BbPelle were fused with the cutting site of EcoRIandXho I restriction http://www.jimmunol.org/ That being said, the function of IRAKs in lower vertebrates and enzyme. The PCR fragments and pCMV-hemagglutinin vector (Clontech) invertebrates is far from being revealed. To gain more under- were both digested with EcoR I and Xho I restriction enzyme at 37˚C standing of the functional evolution of IRAKs, in this study we for overnight. Then, DNA T4 ligase (Takara) was used to ligate the fragments and vector. The recombinant expression vectors were veri- exploited an evolutionary model organism called amphioxus and fied by sequencing and the expression of proteins were confirmed by performed corresponding comparative and functional analysis. Western blot. Primers used for expression vector construction are Cephalochordate amphioxus is the most basal living lineage showed in Table I. among three chordate subphyla (cephalochordates, urochordates, Immunofluorescence analysis and vertebrates), representing a pivotal transitional taxon con- necting invertebrates and vertebrates (45–50). There have been Hela cells were digested by trypsin and added to a 24-well plate covered by guest on September 27, 2021 with pretreated coverslips (10 3 10 mm). After 16–24 h, Hela cells were extensive studies on the molecular functions of the TLR/IL-1R transfected with 500 ng BbIRAK4 or BbPelle recombinant expression pathways in amphioxus (51–65), which provide a robust premise vectors by PEI (Polysciences) according to the manufacturer’s instructions. for the function study of amphioxus IRAKs. Our study shows After 24 h, Hela cells were harvested and fixed in 4% formaldehyde so- that the functions of IRAKs are highly variable between dif- lution for 15 min. Then cells were permeabilized by washing three times in ferent major animal taxa, suggesting that IRAKs might confer 0.05% Tween–20 in PBS (PBST) and blocked nonspecific sites with 5% BSA in PBST at room temperature for 1 h. Next, cells were incubated with necessary plasticity to the signal transduction of the TLR/IL-1R anti-hemagglutinin Ab for 1 h and washed three times in PBST at room pathways. temperature. Hela cells were further incubated with Alex Fluor 594 Goat Anti-Mouse IgG (H + L) (Proteintech) at 4˚C overnight. After triple Materials and Methods washing in PBS, cells were labeled with DAPI for 5 min and washed three Animals and cells times in PBS. The coverslips with treated Hela cells were taken out from the 24-well plate and cover on the slides with MOWIOL R4-88 Reagent Adult Branchiostoma belcheri were obtained from local fishermen of (Calbiochem). Finally, Hela cells were photographed by Carl Zeiss Axio Zhanjiang, China, and cultured in aquaria supplied with circulating filtered Imager Z1 microscope. seawater and fed once every 2 d. For immune stimulation, each amphioxus were injected 15 ml LPS (1 mg/ml) in PBS into the coelom, and PBS was Luciferase reporter assays injected as negative control. Then, intestine and gill from five fishes were HEK293T cells were digested by trypsin and seeded in 48-well plates. After collected and frozen using liquid nitrogen immediately at 0, 2, 8, 24, and 16–24 h, cells were transfected with equivalent mixed expression plasmids, 48 h after injection. which consist of the indicated amount of expression vectors, 50 ng per well HEK293T and Hela cells were grown in DMEM supplemented with of the NF-kB response promoter luciferase reporter plasmid pNF-kB-Luc 10% FCS at 37˚C. (StrataGene), and 5 ng per well Renilla luciferase reporter plasmid pRL- RNA isolation and cDNA synthesis TK (Promega) to normalize the data due to different transfection efficiency between wells, and empty vectors to complement the total plasmid quantity The total RNA was isolated using TRIzol Reagent (Invitrogen) and of each well to the same. After 24 h, HEK293T cells were harvested and isopropanol was used for precipitation. After quality inspection with measured by dual-luciferase reporter assay system (Promega). Each exper- agarose gel electrophoresis and spectrophotometrically, the extracted iment was performed at least in triplicate and was repeated at least twice in RNA was reverse transcribed to synthesize the first strand cDNA using all cases. The data are shown as the fold change relative to that measured PrimeScript first Strand cDNA Synthesis Kit (Takara). Then, the cDNA in cells transfected with empty vector. was stored at 280˚C. Coimmunoprecipitation assay Cloning of B. belcheri IRAK4 and B. belcheri Pelle HEK293T cells were digested by trypsin and seeded in six-well plates. Using human IRAKs as baits, we blasted the genome database of B. belcheri After 16–24 h, cells were transfected with 4–6 mg expression plasmids. (http://genome.bucm.edu.cn/lancelet/index.php) and obtained the consen- At 20–36 h posttransfection, whole-cell extracts were prepared by using sus sequences of B. belcheri IRAK4 (BbIRAK4) and B. belcheri Pelle cell lysis buffer (Cell Signaling Technology) and incubated with primary (BbPelle). Then two pair gene-specific primers were designed for each Abs (Sigma-Aldrich) at 4˚C overnight. The next day, Protein G Sepharose The Journal of Immunology 3 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. Current knowledge of IRAKs in mammal, bony fish, amphioxus, and Drosophila.(A) The functions of IRAKs in different species. We suggest that BbIRAK4 and BbPelle can repress the activity of BbMyD88 and BbTRAF6, but so far no evidence suggests the direct interaction between BbTRAF6 with BbIRAK4 or BbPelle. (B) The domain architecture of human, zebrafish, amphioxus, and Drosophila IRAKs. Important functional residues were indicated. The C-terminal of Drosophila Tube lack kinase domain, which was indicated as white. In RD kinases, an arginine (R) residue immediately precedes the critical aspartate (D) residue in the kinase activation loop, which is vital for their catalytic activity. And in non-RD kinases, this residue is not arginine (R). The kinase activity of DrIRAK3 is unclear.

(Roche) was added to the mixture and incubated at 4˚C for 4–6 h. Then the below. Then we cloned the full-length cDNA of BbIRAK4 and BbPelle mixture was washed three times with cell lysis buffer and analyzed by with nested RT-PCR (Table I). Western blot. Protein architectural diversity of the IRAK genes Results The coding sequence of BbIRAK4 is 1488 bp, encoding 495 aa Identification and cloning of two amphioxus IRAK genes with a theoretical pI of 5.19 and a calculated molecular mass We performed an extensive search in the genome and tran- of 54.6 kDa. It shares ∼84% protein sequence identity with scriptome databases and identified two amphioxus IRAK genes. its B. floridae ortholog and ∼40% identity with IRAK4 from So now it is clear that the genomes of amphioxus, including vertebrates and other invertebrates. BbIRAK4 contains an B. belcheri and B. floridae, only encode two IRAK genes. In N-terminal death domain (residues 1–127), a ProST domain this study, we designated them as IRAK4 and Pelle, respectively, (residues 128–222) and a C-terminal kinase domain (residues according to the results of sequence, phylogenetic and function analyses 223–495), which is more similar to vertebrate IRAK4 than 4 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY

Table I. Primers used for PCR amplification

Primers Primer sequence (5ˊ to 3ˊ) Gene-specific primers BbIRAK4-F 59-TATTTTTGTGTGGTCTGAGAAC-39 BbIRAK4-R 59-TCCCACTCTACTCTCTTAAAATC-39 BbIRAK4-NF 59-ATGGGAAGCCTGACAAATATGCCAG-39 BbIRAK4-NR 59-TCAAATACAGTCCAAGTTCTTTTTCA-39 BbPelle-F 59-TTGGACTACAGGAGACCGGACGGA-39 BbPelle-R 59-ATCAAACGCCTTCTGTCCAGTC-39 BbPelle-NF 59-ATGGCGGCTGCTGTTCTCGGTAGCA-39 BbPelle-NR 59-TCAAACACCACCACGCCAATCAAA-39 Primers for qRT-PCR BbIRAK4-rtF 59-ACTGGTGGAGGTCTTACTAC-39 BbIRAK4-rtR 59-TGGCTGGTCTGAGATGTT-39 BbPelle-rtF 59-CCGAAGTTCTCAAGCCTCTA-39 BbPelle-rtR 59-TTGCCGAAGTTCTCATCAG-39 Primers for construction of expression vector BbIRAK4-vF 59-CCGGAATTCCGatgggaagcctgacaaatatgccag-39 BbIRAK4-vR 59-CCGCTCGAGCGGtcaaatacagtccaagttctttttca-39 BbPelle-vF 59-CCGGAATTCCGatggcggctgctgttctcggtagca-39 BbPelle-vR 59-CCGCTCGAGCGGtcaaacaccaccacgccaatcaaa-39 Downloaded from F, outer forward; NF, inner forward; NR, inner reverse; R, outer reverse; rtF, qRT-PCR forward; rtR, qRT-PCR reverse; vF, vector forward; vR, vector reverse. to Drosophila Tube because Tube lacks the kinase domain performed phylogenetic analysis. Two phylogenetic trees were (Fig. 1B). constructed for the kinase domain and the death domain, respec-

In contrast, the coding sequence of BbPelle is 1353 bp, encoding tively (Fig. 2). The tree based on kinase domains suggests that http://www.jimmunol.org/ 450 aa with predicted molecular mass of 50.5 kDa and theoretical IRAK proteins are clearly distinct from other related kinase pro- pI of 7.64. It shares ∼86% protein sequence identity with its B. teins such as TAK1, TBK1, and IKKs. In contrast the tree based floridae counterpart but shares only ∼35% identity with vertebrate on death domains indicates that the death domain of IRAK pro- IRAK1/2/3 and other invertebrate Pelle. BbPelle contains a death teins is more similar to MyD88 (mainly mediating NF-kB acti- domain (residues 1–117), a ProST domain (residues 118–221), vation) than to FADD and TRADD (mainly mediating apoptosis). and a kinase domain (residues 222–450) but lacks the C-terminal Therefore, both trees suggest that the IRAK family is a distinct region presented in vertebrate IRAK1/2/3 (Fig. 1B). In this sense, kinase family. BbPelle is more similar to Drosophila Pelle. Both trees suggest that the IRAK family could be separated into Further analysis shows that the kinase domain of amphioxus three groups: the IRAK4 group, which is conserved between by guest on September 27, 2021 IRAK4 shares conserved residues with other IRAK4, such as the protostomes and deuterostomes; the IRAK1/2/3 group, which is invariant lysine residues for ATP binding, the threonine/serine specific to vertebrates; and the Pelle group, which is also conserved residues important for kinase activity, and the tyrosine gate- between protostomes and invertebrate deuterostomes but lost in keeper residue, which are all important for IRAK4 function vertebrate deuterostomes. The Pelle group tends to represent the (Supplemental Fig. 1). As for amphioxus Pelle, its kinase intermediate state between IRAK4 and IRAK1/2/3. According to domain is clearly shorter than that of other invertebrate Pelle the two phylogenetic trees, amphioxus IRAK4 belongs to the and vertebrate IRAK1/2/3, which might cause differences in IRAK4 group, whereas amphioxus Pelle is closer to invertebrate function. Amphioxus Pelle also preserves the invariant lysine Pelle than to vertebrate IRAK1/2/3 (Fig. 2). In line with this, we residues and the conserved threonine residues, which are found that amphioxus Pelle and other invertebrate Pelle have not critical for the phosphorylation for Drosophila Pelle and ver- evolved a distinct C-terminal region, which is presented in ver- tebrate IRAK1/2 (Supplemental Fig. 1). tebrate IRAK1/2/3 (Fig. 1B). It is worth noting that in the kinase- Protein kinases can be classified as RD or non-RD kinase based based tree, amphioxus Pelle is clustered with invertebrate Pelle on the residue that immediately precedes the critical aspartate (D) with high confidence, whereas in the death domain based tree, residue in the kinase activation loop region with arginine (R) amphioxus Pelle is slightly closer to vertebrate IRAK4s (although residue or without (66). Like vertebrate IRAK4, amphioxus not as close as amphioxus IRAK4 to vertebrate IRAK4s) (Fig. 2). IRAK4 belongs to the RD kinase groups, whereas amphioxus This suggests that the death domain of Pelle might evolve slower Pelle, together with other invertebrate Pelle and vertebrate than its kinase domain. Taken together, the IRAK family of am- IRAK1/2/3, are non-RD kinases (Fig. 1B, Supplemental Fig. 1). phioxus and other invertebrates does not have an exact functional This indicates that amphioxus IRAK4 could be ortholog of counterpart for mammalian IRAK1/2/3. vertebrate IRAK4. Whereas amphioxus Pelle, like Drosophila Taken together, IRAK4, Pelle, and IRAK1/2/3 are three major Pelle (33), could share a certain degree of functional similarity IRAK lineages in animals. IRAK4 is an ancient lineage conserved with vertebrate IRAK1/2/3. between invertebrate and vertebrates. Pelle is also an ancient Taken together, between human, zebrafish, amphioxus, and lineage but lost in vertebrates. IRAK1/2/3 is a vertebrate-specific Drosophila IRAKs, there are obvious differences in terms of domain lineage. Judging from the sequence and phylogenetic positions (and motif) architectures. We expect that this type of architectural (Fig. 2), we suspect there might be a chance that IRAK1/2/3 was diversity may reflect on their functional diversity. derived from Pelle and diverged in later evolution. The phylogenetic origins of the three IRAK lineages Expression profile of amphioxus IRAK4 and Pelle To further understand the sequence evolution of the IRAK family, qRT- PCR was used to investigate the expression patterns of we collected related sequences from a wide range of species and BbIRAK4 and BbPelle in different tissues and during immune The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 2. Phylogenetic trees of IRAKs based on the sequence of kinase domains (A) and death domains (B). The tree was constructed using the neighbor-joining method. Numbers on the lines indicate the percentage bootstrap values for 1000 replicate. Maximum likelihood trees were also constructed with the Posisson model, which were similar to neighbor-joining trees (data not shown). The accession numbers for the used protein sequences are as follows: Homo sapiens IRAK1, NP_001560.2; H. sapiens IRAK2, NP_001561.3; H. sapiens IRAK3, NP_009130.2; H. sapiens IRAK4, NP_001107654.1; H. sapiens TAK1, NP_663304.1; H. sapiens TBK1, NP_037386.1; H. sapiens IKKa, NP_001269.3; H. sapiens IKKb, NP_001547.1; Mus musculus IRAK1, NP_001171444.1; M. musculus IRAK2, NP_751893.3; M. musculus IRAK3, NP_082955.2; M. musculus IRAK4, NP_084202.2; M. musculus TAK1, NP_0033342.1; M. musculus TBK1, NP_062760.3; M. musculus IKKa, NP_031726.2; M. musculus IKKb, O88351.1; DrIRAK1, XP_005166760.1; DrIRAK3, XP_003198307.2; DrIRAK4, NP_956457.1; D. rerio TAK1, NP_001018586.1; D. rerio TBK1, NP_001038213.2; D. rerio IKKa, NP_956611.1; D. rerio IKKb, NP_001116737.1; D. melanogaster Pelle, NP_476971.1; D. melanogaster Tube, NP_001189164.1; D. melanogaster TAK1, NP_524080.1; D. melanogaster IKKb, NP_524751.3; Ciona intestinalis IRAK4, XP_002122012.1; Saccoglossus kowalevskii IRAK4, XP_002739502.1; Caenorhabditis elegans PIK1, NP_001255742.1; Gallus gallus IRAK2, NP_001025776.1; G. gallus IRAK4, NP_001025909.1; Anolis carolinensis IRAK1, XP_008102118.1; A. carolinensis IRAK2, XP_008115205.1; A. carolinensis IRAK4, XP_003221322.2; Xenopus tropicalis IRAK1, NP_001006713.1; X. tropicalis IRAK2, NP_001007501.1; Xenopus tropicalis IRAK4, NP_001116877.1; Eriocheir sinensis Pelle, AKJ26285.1; E. sinensis Tube, AGT21373.1; Scylla paramamosain (SpPelle), AGY49577.1; S. paramamosain (SpTube), AGY49576.1; Camponotus floridanu Pelle, XP_011265220.1; C. floridanu Tube, EFN72687.1. 6 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 3. Expression profile of BbIRAK4 and BbPelle.(A) Tissue expression of BbIRAK4.(B) Tissue expression of BbPelle.(C and D) Expression profile of BbIRAK4 (C)orBbPelle (D) in amphioxus gill after LPS challenge. (E and F) Expression profile of BbIRAK4 (E) and BbPelle (F) in amphioxus intestine after LPS challenge. Expression level changes were calculated by normalization to the expression of GAPDH. The samples at time 0 h were untreated with either PBS or LPS. Tissues from five fishes were collected for each sample. Two parallel experiments were performed in triplicate, and data are plotted as mean 6 SD (n = 3), *p , 0.05, **p , 0.01, ***p , 0.001. challenge (Fig. 3). Both BbIRAK4 and BbPelle show relatively half-smooth tongue sole, zebrafish, and abalone (26, 27, 67). high expression in the gill and skin and medium expression in the Moreover, the Drosophila Tube, the homolog of IRAK4, has intestine (Fig. 3A, 3B). These tissues, especially the gill and the been shown to participate in the formation of the dorsoventral skin, are considered the first defense line against microbial path- axis during development (66). ogens. Indeed, both BbIRAK4 and BbPelle were upregulated in the gill and the intestine after immune challenge with LPS, the Amphioxus IRAK4 inhibited MyD88 and TRAF6 by common cell wall component of bacteria (Fig. 3C–F). This ex- different mechanisms pression pattern suggests the participation of BbIRAK4 and In agreement with the findings in grouper (35, 68), BbIRAK4 was BbPelle in acute immune reaction. In addition, BbIRAK4 dis- dispersedly expressed in the cytoplasm (Supplemental Fig. 2). played very high expression in the ovary, whereas BbPelle was Thus, BbIRAK4 may exert its function in the cytoplasm but not in nearly absent (Fig. 3A, 3B). Because the harvested ovary mainly the nucleus, like mammalian IRAK4. Using luciferase reporter consisted of unspawned eggs, this observation suggests that assays, we found that transfecting human HEK293T cell line with BbIRAK4, but not BbPelle, is abundant as maternal transcripts high doses (100–300 ng) of BbIRAK4 slightly activated NF-kB and may have an important role in embryogenesis. In line with signal in an unstable way (Fig. 4A), which we suspected could be this, abundant maternal transcripts of IRAK4 were found in due to the minor perturbation conferred by the kinase activity of The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/

FIGURE 4. Regulatory role of BbIRAK4 and BbPelle in NF-kB activity. HEK293T cells were cotransfected with NF-kB transcriptional luciferase by guest on September 27, 2021 reporter plasmid and Renilla luciferase reporter plasmid, together with BbIRAK4/Pelle/MyD88 vectors. (A) BbIRAK4 slightly activated NF-kB signal. (B and C) BbIRAK4 impaired BbMyD88-induced (B) and BbTRAF6-induced (C) NF-kB activation. (D) BbPelle could not activate NF-kB signal. (E and F) BbPelle suppressed BbMyD88-induced (E) and BbTRAF6-induced (F) NF-kB activation. The results are expressed as means 6 SD of three samples per treatment, and were confirmed by at least three separate experiments. *p , 0.05, **p , 0.01, ***p , 0.001.

BbIRAK4. In any case, this warrants further study. In contrast, BbIRAK4 might not affect NF-kB activation through the direct when B. belcheri MyD88 (BbMyD88) was cotransfected with interaction with BbTRAF6. BbIRAK4 even only 5 ng, the BbMyD88-induced NF-kB acti- vation could be significantly impaired. Morever, a nearly complete Amphioxus Pelle inhibited MyD88 and TRAF6 by blockage could be achieved with a slighted larger dose (50 ng) of different mechanisms BbIRAK4 (Fig. 4B). This observation coincides with the pre- BbPelle was also expressed in the cytoplasm like BbIRAK4 vious reports that bony fish IRAK4 could suppress NF-kBac- (Supplemental Fig. 2). But, BbPelle could not activate NF-kB tivation in human cells (34, 35). Meanwhile, B. belcheri TRAF6 even with large dosage (Fig. 4D), which is different to Drosophila (BbTRAF6)–mediated NF-kB activation was also suppressed by Pelle and human IRAK1 (43, 69). BbPelle also impaired the BbIRAK4, but the effect is limited and not correlated with the NF-kB activation when cotransfected with BbMyD88, although dosage of BbIRAK4 (Fig. 4C), hence implying that BbIRAK4 this required a larger dosage (Fig. 4E). BbPelle also negatively might not act on BbTRAF6 directly. Then we detected the ex- regulated BbTRAF6-mediated NF-kB activation, but different pression levels of TNF-a and IL-8, which are downstream genes from BbIRAK4, the suppressive effect of BbPelle on BbTRAF6 was of NF-kB after transfection. In line with the reporter assays, correlated with the transfection dosage (Fig. 4F). Further qRT-PCR BbIRAK4 could intensely downregulate the expression of TNF-a assays showed that BbPelle impaired the expression of TNF-a and IL-8 induced by BbMyD88 and also impair their expression induced by BbMyD88 or BbTRAF6 and the expression of IL-8 induced by BbTRAF6 (Fig. 5A–D). induced by BbTRAF6 (Fig. 5A–D). Further coimmunoprecipitation (Co-IP) assays showed that like Co-IP assays showed the direct binding between BbPelle and its mammalian counterparts, BbIRAK4 could use its death domain BbMyD88 via the death domains (Fig. 6B) but failed to show the to directly associate with BbMyD88 (Fig. 6A), suggesting that stable interaction between BbPelle and BbTRAF6 (Fig. 6C). The BbIRAK4 might suppress MyD88 by direct interaction. Immuno- latter observation might be due to the fact that BbPelle lacks fluorescence analysis in Hela cells confirmed that BbIRAK4 could the C-terminal region and TRAF6-binding motifs, which are re- colocalize with BbMyD88 in the cytoplasm (Fig. 6E, Supplemental quired for the interaction between vertebrate IRAK1/2 and TRAF6 Fig. 3). In contrast, the direct interaction between BbIRAK4 (Fig. 1B). BbPelle also colocalized with BbMyD88 in the cyto- and BbTRAF6 was not observed (Fig. 6C, 6E), suggesting that plasm of Hela cells (Fig. 6E, Supplemental Fig. 3), in agreement 8 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY Downloaded from http://www.jimmunol.org/

FIGURE 5. Downstream genes were inhibited by BbIRAK4 and BbPelle. HEK293T cells were cotransfected with BbIRAK4/Pelle/MyD88/TRAF6 vectors. The expression level of TNF-a and IL-8 were detected through qRT-PCR assay. (A and B) The expression change of downstream genes induced by BbMyD88, when cotransfected with BbIRAK4 or BbPelle. (C and D) The expression change of downstream genes induced by BbTRAF6, when cotransfected with BbIRAK4 or BbPelle. The results are expressed as means 6 SD of three samples per treatment and were confirmed by at least three separate experiments. *p , 0.05, **p , 0.01, ***p , 0.001. DD, death domain. with the Co-IP assay. Taken together, we suggested that BbPelle when cotransfected BbMyD88 with BbIRAK4-ODD (Fig. 6G). by guest on September 27, 2021 acts on BbMyD88 to repress its activity. We also suggested that These results indicated that the death domain of BbIRAK4 was BbPelle might act on both upstream and downstream of BbTRAF6, mainly responsible for its suppressive function on BbMyD88. because BbPelle negatively regulates BbTRAF6 in a dosage- However, BbIRAK4-OKD (or other truncated mutants) had dependent manner without direct contact. As reference, there was no (or weak) influence on BbTRAF6-mediated NF-kB acti- a negative feedback regulation role of Drosophila Pelle in the Toll vation when cotransfected with BbTRAF6, suggesting that pathway (70). Drosophila Pelle possibly direct phosphorylation of the kinase domain of BbIRAK4 might be responsible for this Tube, leading to the disaggregation of membrane-associated Tube function (Fig. 6H). clusters (70). As for BbPelle, both the death domain and the ProST region contributed to the inhibitory effect on BbMyD88, whereas the No interaction detected between amphioxus IRAK4 and Pelle kinase domain was not required (Fig. 6I). Interestingly, BbPelle- In both Drosophila and vertebrates, Tube (IRAK4 in vertebrates) OProST sharply enhanced both the BbMyD88-induced and the can interact with both Pelle (IRAK1/2 in vertebrates) and MyD88 BbTRAF6-induced NF-kB activation (Fig. 6I, 6J), suggesting that to form the so-called Myddosome complex during signal the ProST domain was important for its suppressive function. transduction (Fig. 1A). Here, we performed Co-IP assays to test Based on these results, we suggested that BbPelle could bind to if BbPelle and BbIRAK4 could interact with each other for the MyD88 and then use the ProST region to inhibit the downstream regulation of the MyD88-TRAF6 pathway. Surprisingly, although NF-kB activation. both BbPelle and BbIRAK4 could interact with BbMyD88 Ex vivo functional analysis of zebrafish IRAKs (Fig. 6A, 6B), we did not observe the direct interaction between BbPelle and BbIRAK4 (Fig. 6D). However, part of BbPelle and We also analyzed the function of zebrafish Danio rerio IRAK BbIRAK4 could colocalize with each other (Fig. 6E, Supplemental (DrIRAK)1/3/4 in HEK293T cells. DrIRAK1 could significantly Fig. 3), indicating that they may form complexes through the in- activate NF-kB signal, whereas DrIRAK3 and DrIRAK4 could not teraction with MyD88. This suggested that a new functional model (Fig. 7A). Both DrIRAK3 and DrIRAK4 impaired D. rerio for the IRAK proteins in the basal chordate. MyD88 (DrMyD88)–induced NF-kB activation with dosage de- pendent manners, which coincided with the reports of other Functions of different domains in BbIRAK4 and BbPelle teleost IRAK4 (Fig. 7B, 7C) (34, 35). When cotransfected with To investigate which domain was essential for their suppressive DrIRAK1, DrIRAK3 prominently downregulated DrIRKA1- activity, we constructed several truncated mutants of BbIRAK4 and mediated NF-kB activation (Fig. 7D), whereas DrIRAK4 had BbPelle (Fig. 6F). Two truncated mutants of BbIRAK4 with death no effect on DrIRKA1-mediated NF-kB activation (Fig. 7E). domain still strongly suppressed BbMyD88-induced NF-kB As mammalian IRAK3 prevented IRAK1/2 from the dissoci- activation, whereas the suppressive activity was very weak ation from MyD88 and negatively regulated the pathway (23, The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 6. Functional analysis of BbIRAK4 and BbPelle. For colocalization assay, Hela cells were transfected with BbIRAK4/Pelle/MyD88/TRAF6 vectors. For Co-IP and reporter assays, HEK293T cells were used. (A and B) Co-IP assays showed the interaction between BbMyD88 with BbIRAK4 (A)or BbPelle (B). (C) Co-IP assays showed no interaction between BbTRAF6 with BbIRAK4 or BbPelle. (D) Co-IP assays showed no interaction between BbIRAK4 with BbPelle. (E) Colocalization of BbIRAK4, BbPelle, BbMyD88, and BbTRAF6 in Hela cells. Scale bar, 5 mm. (F) Truncated mutants were constructed. (G and H) The function of BbIRAK4 truncated mutants were analyzed through reporter assays. (I and J) The function of BbPelle truncated mutants were analyzed through reporter assays. Results were confirmed by at least three separate experiments The results of reporter assays are expressed as means 6 SD of three samples per treatment. *p , 0.05, **p , 0.01, ***p , 0.001. DD, death domain.

24), we presumed that DrIRAK3 may perform a similar unique evolutionary position between vertebrates and inverte- function as mammalian IRAK3. However, the role of brates, the study of amphioxus IRAKs may provide insight into DrIRAK4 was more close to amphioxus IRAK4. We also tried to the functional evolution of IRAK family (71). In this study, we use ZFL cells, derived from zebrafish liver, to conduct the luciferase showed that amphioxus has only two IRAK genes; one is the reporter assays, but the low transfection efficiency made the result ortholog to IRAK4, whereas the other one is the ortholog to in- not sufficiently reliable and repeatable (data not shown). vertebrate Pelle, rather than to vertebrate IRAK1/2/3. Phylogenetic trees show that the kinase domain of amphioxus Pelle might Discussion evolve faster than its death domain. As we known, the death do- The IRAK family plays a crucial role in the TLR/IL-1R–MyD88– main is served as a bridge to link MyD88 and IRAKs for forming TRAF6–NF-kB signaling pathway, but their function appears not Myddosome complex, whereas the kinase domain of IRAKs consistent between Drosophila and vertebrates, and even between phosphorylates downstream proteins or itself during signaling bony fishes and mammals (25–30). As amphioxus occupied a transduction (6, 7). Thus, these findings suggested that although 10 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY Downloaded from

FIGURE 7. Regulatory role of DrIRAK1/3/4 in NF-kB activity. HEK293T cells were cotransfected with NF-kB transcriptional luciferase reporter plasmid and Rellina luciferase reporter plasmid, together with DrIRAK1/IRAK3/IRAK4/MyD88 vectors. (A) DrIRAK1 could significantly active NF-kB signal. (B) DrIRAK3 impaired DrMyD88-induced NF-kB activation. (C) DrIRAK4 impaired DrMyD88-induced NF-kB activation. (D) DrIRAK3 impaired http://www.jimmunol.org/ DrIRAK1-induced NF-kB activation. (E) DrIRAK4 had no influence on DrIRAK1-induced NF-kB activation. Results were confirmed by at least three separate experiments. The results are expressed as means 6 SD of three samples per treatment. **p , 0.01, ***p , 0.001. the IRAKs have a conserved position in TLR pathway, the specific Tube, causing the dissociation of Pelle–Tube–MyD88 complex, functions significantly vary between major animal taxa. and downregulate the pathway (70). However, we could not detect We cloned these two IRAK genes from the amphioxus B. belcheri. the interaction between BbIRAK4/BbPelle and BbTRAF6, which Both BbIRAK4 and BbPelle may be related to amphioxus immune may be due to the lacking of the C-terminal domain and the reaction according to their high expression in gill and skin. The TRAF6 binding sites. Interestingly, when eliminated the ProST by guest on September 27, 2021 gill of amphioxus belongs to digestive system and is the first domain of BbPelle, BbPelle-OProST enhanced the NF-kB acti- defense line of amphioxus immunity, and the skin of amphioxus is vation induced by BbMyD88 or BbTRAF6. The ProST domain is also a potential immune organ of amphioxus (71). We observed rich in serines, prolines, and threonines, and mammalian IRAK1 abundant maternal transcripts of BbIRAK4 in eggs, which has also was reported to undergo hyperphosphorylation in ProST domain, been reported in half-smooth tongue sole, zebrafish, and abalone which was important for its signal transduction (8, 72). Thus, we (26, 27, 67), suggesting a potentially important role in embryogen- suspect that hyperphosphorylation in BbPelle ProST domain was esis. For example, it is reported that Drosophila Tube, the homolog essential for its suppressive function. However, BbPelle without of IRAK4, participates in the formation of the dorsoventral axis kinase domain still downregulated BbMyD88 and BbTRAF6- during development (66). Although BbPelle transcripts were nearly mediated NF-kB activation, indicating that the ProST domain of absent in eggs, they indicated an obviously functional difference of BbPelle may be phosphorylated by some other kinases but not amphioxus IRAK4 and Pelle. autophosphorylation. In HEK293T cells, phosphorylated BbPelle BbIRAK4 slightly induced activation of NF-kB, consistent with may induce the degradation of BbTRAF6 or directly inhibit some previous findings of bony fish IRAK4 (34, 35). Amphioxus or endogenous downstream signal molecules, such as NFKB1 and bony fish IRAK4 may be not fully compatible with the mamma- NFKB2 which also contained death domain, and performed as lian system, because zebrafish IRAK4 could significantly stimu- inhibition function when cotransfected with BbTRAF6. And these late NF-kB activation in ZFL cells (26). BbIRAK4 showed a require further investigation. Considering that the kinase domain negative role in the MyD88–TRAF6–NF-kB pathway, in agree- of BbIRAK4 is essential for inhibiting BbTRAF6 induced NF-kB ment with the previous studies of bony fish IRAK4 (34, 35). Both activation, one possible explanation in amphioxus is that BbPelle BbIRAK4 and BbPelle could interact with BbMyD88 through maybe phosphorylated by BbIRAK4 when they interacted with their death domain, suggesting a possibility to form a complex BbMyD88 at the same time. similar to mammal Myddosome complex. Further truncated mu- We also analyzed the function of zebrafish IRAK1/3/4, indi- tants analysis confirmed that the death domain of BbIRAK4 and cating that the function of zebrafish IRAK1 and IRAK3 was similar BbPelle is essential for suppressing BbMyD88-induced NF-kB to their mammalian counterparts, whereas zebrafish IRAK4 was activation. However, we could not detect the direct interaction more similar to amphioxus IRAK4. Thus, there was no functional between BbIRAK4 and BbPelle. Thus, the Myddosome-like counterpart of amphioxus Pelle in vertebrate, and the functional complex in amphioxus might be IRAK4–MyD88–Pelle but not difference of amphioxus and mammalian IRAK4 was evolved after Pelle–IRAK4–MyD88. This may explain the functional difference the divergence of bony fishes and amphibian. of amphioxus IRAKs with mammals. We also observed a sup- In summary, our findings show that the composition and se- pressive role of BbPelle in MyD88–TRAF6–NF-kB pathway. quence of amphioxus IRAKs are more similar to invertebrates than Also, Drosophila Pelle was suggested to directly phosphorylate vertebrates. However, molecular functions of amphioxus IRAKs The Journal of Immunology 11 are quite different from Drosophila and mammals. Together with 24. Zhou, H., M. Yu, K. Fukuda, J. Im, P. Yao, W. Cui, K. Bulek, J. Zepp, Y. Wan, T. W. Kim, et al. 2013. IRAK-M mediates toll-like receptor/IL-1R-induced previous reports, we concluded that although IRAKs are con- NFkB activation and cytokine production. EMBO J. 32: 583–596. served members of the TLR/IL-1R pathways, their functions 25. Huang, R., J. Lv, D. Luo, L. Liao, Z. Zhu, and Y. Wang. 2012. Identification, (unlike those of MyD88 and TRAF6) are highly variable during characterization and the interaction of Tollip and IRAK-1 in grass carp (Cte- nopharyngodon idellus). Fish Shellfish Immunol. 33: 459–467. evolution and clearly specialized in different major animal taxa. 26. Phelan, P. E., M. T. Mellon, and C. H. Kim. 2005. Functional characterization of Considering their kinase activity, we further suggest that the full-length TLR3, IRAK-4, and TRAF6 in zebrafish (Danio rerio). Mol. functional variability of IRAKs might confer important plasticity Immunol. 42: 1057–1071. 27. Yu, Y., Q. Zhong, C. Li, L. Jiang, Y. Wang, Y. Sun, X. Wang, Z. Wang, and to the intracellular signal transduction of the TLR/IL-1R path- Q. Zhang. 2012. Identification and characterization of IL-1 receptor-associated ways, which in return could facilitate the species to evolve and to kinase-4 (IRAK-4) in half-smooth tongue sole Cynoglossus semilaevis. Fish adapt during the completion against the pathogens. Shellfish Immunol. 32: 609–615. 28. Brietzke, A., T. Goldammer, H. Rebl, T. Koryta´r, B. Ko¨llner, W. Yang, A. Rebl, and H. M. Seyfert. 2014. Characterization of the interleukin 1 receptor- Disclosures associated kinase 4 (IRAK4)-encoding gene in salmonid fish: the functional copy is rearranged in Oncorhynchus mykiss and that factor can impair TLR The authors have no financial conflicts of interest. signaling in mammalian cells. Fish Shellfish Immunol. 36: 206–214. 29. Li, X. C., X. W. Zhang, J. F. Zhou, H. Y. Ma, Z. D. Liu, L. Zhu, X. J. Yao, L. G. Li, and W. H. Fang. 2013. Identification, characterization, and functional References analysis of Tube and Pelle homologs in the mud crab Scylla paramamosain. PLoS One 8: e76728. 1. Takeuchi, O., and S. Akira. 2010. Pattern recognition receptors and inflamma- tion. Cell 140: 805–820. 30. Watthanasurorot, A., K. So¨derha¨ll, and P. Jiravanichpaisal. 2012. A mammalian 2. Akira, S., and K. Takeda. 2004. Toll-like receptor signalling. Nat. Rev. Immunol. like interleukin-1 receptor-associated kinase 4 (IRAK-4), a TIR signaling me- diator in intestinal innate immunity of black tiger shrimp (Penaeus monodon). 4: 499–511. Downloaded from 3. Martin, M. U., and H. Wesche. 2002. Summary and comparison of the signaling Biochem. Biophys. Res. Commun. 417: 623–629. mechanisms of the toll/interleukin-1 receptor family. Biochim. Biophys. Acta 31. Suzuki, K., S. Izumi, H. Tanaka, and T. Katagiri. 2016. Identification and ex- 1592: 265–280. pression analysis of IRAK-4 cDNA and its gene from the ayu Plecoglossus 4. Verstak, B., P. Hertzog, and A. Mansell. 2007. Toll-like receptor signalling and altivelis altivelis. Fish. Sci. 82: 47–57. the clinical benefits that lie within. Inflamm. Res. 56: 1–10. 32. Valenzuela-Mun˜oz, V., and C. Gallardo-Esca´rate. 2014. Molecular cloning and k 5. O’Neill, L. A. 2008. Primer: toll-like receptor signaling pathways--what do expression of IRAK-4, IL-17 and I- B genes in Haliotis rufescens challenged rheumatologists need to know? Nat. Clin. Pract. Rheumatol. 4: 319–327. with Vibrio anguillarum. Fish Shellfish Immunol. 36: 503–509. 6. Flannery, S., and A. G. Bowie. 2010. The interleukin-1 receptor-associated ki- 33. Gosu, V., S. Basith, P. Durai, and S. Choi. 2012. Molecular evolution and http://www.jimmunol.org/ nases: critical regulators of innate immune signalling. Biochem. Pharmacol. 80: structural features of IRAK family members. PLoS One 7: e49771. 1981–1991. 34. Zou, P. F., X. N. Huang, C. L. Yao, Q. X. Sun, Y. Li, Q. Zhu, Z. X. Yu, and 7. Lin, S. C., Y. C. Lo, and H. Wu. 2010. Helical assembly in the MyD88-IRAK4- Z. J. Fan. 2017. Cloning and functional characterization of IRAK4 in large IRAK2 complex in TLR/IL-1R signalling. Nature 465: 885–890. yellow croaker (Larimichthys crocea) that associates with MyD88 but impairs 8. Kollewe, C., A. C. Mackensen, D. Neumann, J. Knop, P. Cao, S. Li, H. Wesche, NF-kB activation. Fish Shellfish Immunol. 63: 452–464. and M. U. Martin. 2004. Sequential autophosphorylation steps in the interleukin- 35. Li, Y. W., X. B. Mo, L. Zhou, X. Li, X. M. Dan, X. C. Luo, and A. X. Li. 2014. 1 receptor-associated kinase-1 regulate its availability as an adapter in Identification of IRAK-4 in grouper (Epinephelus coioides) that impairs MyD88- interleukin-1 signaling. J. Biol. Chem. 279: 5227–5236. dependent NF-kB activation. Dev. Comp. Immunol. 45: 190–197. 9. Martin, M. U., and C. Kollewe. 2001. Interleukin-1 receptor-associated kinase-1 36. Huang, Y., Y. H. Chen, Y. Z. Zhang, J. L. Feng, L. L. Zhao, H. X. Zhu, W. Wang, (IRAK-1): a self-regulatory adapter molecule in the signaling cascade of the toll/ and Q. Ren. 2015. Identification, characterization, and functional studies of a IL-1 receptor family. Signal Transduct. 1: 37–50. Pelle gene in the Chinese mitten crab, Eriocheir sinensis. Fish Shellfish Immunol. 10. Kawagoe, T., S. Sato, K. Matsushita, H. Kato, K. Matsui, Y. Kumagai, T. Saitoh, 45: 704–716. by guest on September 27, 2021 T. Kawai, O. Takeuchi, and S. Akira. 2008. Sequential control of toll-like 37. Yu, A. Q., X. K. Jin, M. H. Wu, X. N. Guo, S. Li, L. He, W. W. Li, and Q. Wang. receptor-dependent responses by IRAK1 and IRAK2. Nat. Immunol. 9: 684–691. 2014. Identification and characterization of Tube in the Chinese mitten crab 11. Meylan, E., and J. Tschopp. 2008. IRAK2 takes its place in TLR signaling. Nat. Eriocheir sinensis. Gene 541: 41–50. Immunol. 9: 581–582. 38. Gauthier, M. E., L. Du Pasquier, and B. M. Degnan. 2010. The genome of the 12. Lye, E., C. Mirtsos, N. Suzuki, S. Suzuki, and W. C. Yeh. 2004. The role of sponge Amphimedon queenslandica provides new perspectives into the origin of interleukin 1 receptor-associated kinase-4 (IRAK-4) kinase activity in IRAK-4- toll-like and interleukin 1 receptor pathways. Evol. Dev. 12: 519–533. mediated signaling. J. Biol. Chem. 279: 40653–40658. 39. Sun, H., B. N. Bristow, G. Qu, and S. A. Wasserman. 2002. A heterotrimeric 13. Ye, H., J. R. Arron, B. Lamothe, M. Cirilli, T. Kobayashi, N. K. Shevde, death domain complex in toll signaling. Proc. Natl. Acad. Sci. USA 99: 12871– D. Segal, O. K. Dzivenu, M. Vologodskaia, M. Yim, et al. 2002. Distinct mo- 12876. lecular mechanism for initiating TRAF6 signalling. Nature 418: 443–447. 40. Xiao, T., P. Towb, S. A. Wasserman, and S. R. Sprang. 1999. Three-dimensional 14. Takeda, K., and S. Akira. 2004. TLR signaling pathways. Semin. Immunol. 16: structure of a complex between the death domains of Pelle and Tube. Cell 99: 3–9. 545–555. 15. Janssens, S., and R. Beyaert. 2002. A universal role for MyD88 in TLR/IL-1R- 41. Schiffmann, D. A., J. H. White, A. Cooper, M. A. Nutley, S. E. Harding, mediated signaling. Trends Biochem. Sci. 27: 474–482. K. Jumel, R. Solari, K. P. Ray, and N. J. Gay. 1999. Formation and biochemical 16. Motshwene, P. G., M. C. Moncrieffe, J. G. Grossmann, C. Kao, M. Ayaluru, characterization of tube/pelle death domain complexes: critical regulators of A. M. Sandercock, C. V. Robinson, E. Latz, and N. J. Gay. 2009. An oligomeric postreceptor signaling by the Drosophila toll receptor. Biochemistry 38: 11722– signaling platform formed by the toll-like receptor signal transducers MyD88 11733. and IRAK-4. J. Biol. Chem. 284: 25404–25411. 42. Valanne, S., J. H. Wang, and M. Ra¨met. 2011. The Drosophila toll signaling 17. Burns, K., S. Janssens, B. Brissoni, N. Olivos, R. Beyaert, and J. Tschopp. 2003. pathway. J. Immunol. 186: 649–656. Inhibition of interleukin 1 receptor/toll-like receptor signaling through the al- 43. Shen, B., and J. L. Manley. 2002. Pelle kinase is activated by autophosphor- ternatively spliced, short form of MyD88 is due to its failure to recruit IRAK-4. ylation during toll signaling in Drosophila. Development 129: 1925–1933. J. Exp. Med. 197: 263–268. 44. Daigneault, J., L. Klemetsaune, and S. A. Wasserman. 2013. The IRAK homolog 18. Brikos, C., R. Wait, S. Begum, L. A. O’Neill, and J. Saklatvala. 2007. Mass Pelle is the functional counterpart of IkB kinase in the Drosophila toll pathway. spectrometric analysis of the endogenous type I interleukin-1 (IL-1) receptor PLoS One 8: e75150. signaling complex formed after IL-1 binding identifies IL-1RAcP, MyD88, and 45. Delsuc, F., H. Brinkmann, D. Chourrout, and H. Philippe. 2006. Tunicates and IRAK-4 as the stable components. Mol. Cell. Proteomics 6: 1551–1559. not cephalochordates are the closest living relatives of vertebrates. Nature 439: 19. Cheng, H., T. Addona, H. Keshishian, E. Dahlstrand, C. Lu, M. Dorsch, Z. Li, 965–968. A. Wang, T. D. Ocain, P. Li, et al. 2007. Regulation of IRAK-4 kinase activity 46. Huang, S., Z. Chen, X. Yan, T. Yu, G. Huang, Q. Yan, P. A. Pontarotti, via autophosphorylation within its activation loop. Biochem. Biophys. Res. H. Zhao, J. Li, P. Yang, et al. 2014. Decelerated genome evolution in Commun. 352: 609–616. modern vertebrates revealed by analysis of multiple lancelet genomes. Nat. 20. Yamin, T. T., and D. K. Miller. 1997. The interleukin-1 receptor-associated ki- Commun. 5: 5896. nase is degraded by proteasomes following its phosphorylation. J. Biol. Chem. 47. Putnam, N. H., T. Butts, D. E. Ferrier, R. F. Furlong, U. Hellsten, T. Kawashima, 272: 21540–21547. M. Robinson-Rechavi, E. Shoguchi, A. Terry, J. K. Yu, et al. 2008. The am- 21. Wesche, H., W. J. Henzel, W. Shillinglaw, S. Li, and Z. Cao. 1997. MyD88: phioxus genome and the evolution of the chordate karyotype. Nature 453: 1064– an adapter that recruits IRAK to the IL-1 receptor complex. Immunity 7: 1071. 837–847. 48. Holland, L. Z., R. Albalat, K. Azumi, E. Benito-Gutie´rrez, M. J. Blow, 22. Keating, S. E., G. M. Maloney, E. M. Moran, and A. G. Bowie. 2007. IRAK-2 M. Bronner-Fraser, F. Brunet, T. Butts, S. Candiani, L. J. Dishaw, et al. 2008. participates in multiple toll-like receptor signaling pathways to NFkappaB via The amphioxus genome illuminates vertebrate origins and cephalochordate bi- activation of TRAF6 ubiquitination. J. Biol. Chem. 282: 33435–33443. ology. Genome Res. 18: 1100–1111. 23. Kobayashi, K., L. D. Hernandez, J. E. Gala´n, C. A. Janeway, Jr., R. Medzhitov, 49. Schubert, M., H. Escriva, J. Xavier-Neto, and V. Laudet. 2006. Amphioxus and and R. A. Flavell. 2002. IRAK-M is a negative regulator of toll-like receptor tunicates as evolutionary model systems. Trends Ecol. Evol. (Amst.) 21: 269– signaling. Cell 110: 191–202. 277. 12 FUNCTIONAL VARIATION OF IRAKS IN MYD88–TRAF6 PATHWAY

50. Louis, A., H. Roest Crollius, and M. Robinson-Rechavi. 2012. How much does 61. Yin, D., W. Li, M. Fu, L. Chen, F. Ma, and P. Jin. 2016. Identification and the amphioxus genome represent the ancestor of chordates? Brief. Funct. characterization of a TAB1 gene involved in innate immunity of amphioxus Genomics 11: 89–95. (Branchiostoma belcheri). Gene 575: 294–302. 51. Yuan, S., J. Zhang, L. Zhang, L. Huang, J. Peng, S. Huang, S. Chen, and A. Xu. 62. Wang, R., S. Huang, X. Fu, G. Huang, X. Yan, Z. Yue, S. Chen, Y. Li, and A. Xu. 2013. The archaic roles of the amphioxus NF-kB/IkB complex in innate immune 2017. The conserved ancient role of chordate PIAS as a multilevel repressor of responses. J. Immunol. 191: 1220–1230. the NF-kB pathway. Sci. Rep. 7: 17063. 52. Yuan, S., T. Liu, S. Huang, T. Wu, L. Huang, H. Liu, X. Tao, M. Yang, K. Wu, 63. Song, X., P. Jin, J. Hu, S. Qin, L. Chen, J. Li-Ling, and F. Ma. 2012. Involvement Y. Yu, et al. 2009. Genomic and functional uniqueness of the TNF receptor- of AmphiREL, a Rel-like gene identified in Brachiastoma belcheri, in LPS- associated factor gene family in amphioxus, the basal chordate. J. Immunol. 183: induced response: implication for evolution of Rel subfamily genes. Genomics 4560–4568. 99: 361–369. 53. Yuan, S., S. Huang, W. Zhang, T. Wu, M. Dong, Y. Yu, T. Liu, K. Wu, H. Liu, 64. Yuan, S., H. Liu, M. Gu, L. Xu, S. Huang, Z. Ren, and A. Xu. 2010. Charac- M. Yang, et al. 2009. An amphioxus TLR with dynamic embryonic expression terization of the extrinsic apoptotic pathway in the basal chordate amphioxus. pattern responses to pathogens and activates NF-kappaB pathway via MyD88. Sci. Signal. 3: ra66. Mol. Immunol. 46: 2348–2356. 65. Fan, S., S. Chen, Y. Liu, Y. Lin, H. Liu, L. Guo, B. Lin, S. Huang, and A. Xu. 54. Huang, S., X. Wang, Q. Yan, L. Guo, S. Yuan, G. Huang, H. Huang, J. Li, 2008. Zebrafish TRIF, a Golgi-localized protein, participates in IFN induction M. Dong, S. Chen, and A. Xu. 2011. The evolution and regulation of the mucosal and NF-kappaB activation. J. Immunol. 180: 5373–5383. immune complexity in the basal chordate amphioxus. J. Immunol. 186: 2042– 66. Towb, P., H. Sun, and S. A. Wasserman. 2009. Tube Is an IRAK-4 homolog in a 2055. toll pathway adapted for development and immunity. J. Innate Immun. 1: 309– 55. Huang, S., S. Yuan, L. Guo, Y. Yu, J. Li, T. Wu, T. Liu, M. Yang, K. Wu, H. Liu, 321. et al. 2008. Genomic analysis of the immune gene repertoire of amphioxus reveals 67. Ge, H., G. Wang, L. Zhang, Z. Zhang, S. Wang, Z. Zou, S. Yan, and Y. Wang. extraordinary innate complexity and diversity. Genome Res. 18: 1112–1126. 2011. Molecular cloning and expression of interleukin-1 receptor-associated 56. Zhang, Q. L., Q. H. Zhu, M. Z. Liang, F. Wang, J. Guo, X. Y. Deng, J. Y. Chen, kinase 4, an important mediator of toll-like receptor signal pathway, from Y. J. Wang, and L. B. Lin. 2018. Comparative transcriptomic analysis provides small abalone Haliotis diversicolor. Fish Shellfish Immunol. 30: 1138–1146. insights into antibacterial mechanisms of Branchiostoma belcheri under Vibrio 68. Li, Y. W., F. Zhao, Z. Q. Mo, X. C. Luo, A. X. Li, and X. M. Dan. 2016. parahaemolyticus infection. Fish Shellfish Immunol. 76: 196–205. Characterization, expression, and functional study of IRAK-1 from grouper, 57. Yuan, S., K. Wu, M. Yang, L. Xu, L. Huang, H. Liu, X. Tao, S. Huang, and Epinephelus coioides. Fish Shellfish Immunol. 56: 374–381. Downloaded from A. Xu. 2010. Amphioxus SARM involved in neural development may function 69. Li, S., A. Strelow, E. J. Fontana, and H. Wesche. 2002. IRAK-4: a novel member as a suppressor of TLR signaling. J. Immunol. 184: 6874–6881. of the IRAK family with the properties of an IRAK-kinase. Proc. Natl. Acad. 58. Yang, M., S. Yuan, S. Huang, J. Li, L. Xu, H. Huang, X. Tao, J. Peng, and A. Xu. Sci. USA 99: 5567–5572. 2011. Characterization of bbtTICAM from amphioxus suggests the emergence 70. Towb, P., A. Bergmann, and S. A. Wasserman. 2001. The protein kinase Pelle of a MyD88-independent pathway in basal chordates. Cell Res. 21: 1410–1423. mediates feedback regulation in the Drosophila toll signaling pathway. Devel- 59. Yuan, S., X. Dong, X. Tao, L. Xu, J. Ruan, J. Peng, and A. Xu. 2014. Emergence of opment 128: 4729–4736. the A20/ABIN-mediated inhibition of NF-kB signaling via modifying the ubiq- 71. Yuan, S., J. Ruan, S. Huang, S. Chen, and A. Xu. 2015. Amphioxus as a model

uitinated proteins in a basal chordate. Proc. Natl. Acad. Sci. USA 111: 6720–6725. for investigating evolution of the vertebrate immune system. Dev. Comp. http://www.jimmunol.org/ 60. Peng, J., X. Tao, R. Li, J. Hu, J. Ruan, R. Wang, M. Yang, R. Yang, X. Dong, Immunol. 48: 297–305. S. Chen, et al. 2015. Novel toll/IL-1 receptor homologous region adaptors act as 72. Gottipati, S., N. L. Rao, and W. P. Fung-Leung. 2008. IRAK1: a critical signaling negative regulators in amphioxus TLR signaling. J. Immunol. 195: 3110–3118. mediator of innate immunity. Cell. Signal. 20: 269–276. by guest on September 27, 2021