Characterization of Amphioxus IFN Regulatory Factor Family Reveals an Archaic Signaling Framework for Innate Immune Response This information is current as of September 30, 2021. Shaochun Yuan, Tingting Zheng, Peiyi Li, Rirong Yang, Jie Ruan, Shengfeng Huang, Zhenxin Wu and Anlong Xu J Immunol 2015; 195:5657-5666; Prepublished online 16 November 2015; doi: 10.4049/jimmunol.1501927 Downloaded from http://www.jimmunol.org/content/195/12/5657

Supplementary http://www.jimmunol.org/content/suppl/2015/11/14/jimmunol.150192 http://www.jimmunol.org/ Material 7.DCSupplemental References This article cites 41 articles, 11 of which you can access for free at: http://www.jimmunol.org/content/195/12/5657.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 30, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*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 © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Characterization of Amphioxus IFN Regulatory Factor Family Reveals an Archaic Signaling Framework for Innate Immune Response

Shaochun Yuan,*,1 Tingting Zheng,*,1,2 Peiyi Li,* Rirong Yang,* Jie Ruan,* Shengfeng Huang,* Zhenxin Wu,* and Anlong Xu*,†

The IFN regulatory factor (IRF) family encodes transcription factors that play important roles in immune defense, stress response, reproduction, development, and carcinogenesis. Although the origin of the IRF family has been dated back to multicellular organ- isms, invertebrate IRFs differ from vertebrate IRFs in genomic structure and gene synteny, and little is known about their functions. Through comparison of multiple amphioxus genomes, in this study we suggested that amphioxus contains nine IRF members, whose

orthologs are supposed to be shared among three amphioxus species. As the orthologs to the vertebrate IRF1 and IRF4 subgroups, Downloaded from Branchiostoma belcheri tsingtauense (bbt)IRF1 and bbtIRF8 bind the IFN-stimulated response element (ISRE) and were upreg- ulated when amphioxus intestinal cells were stimulated with poly(I:C). As amphioxus-specific IRFs, both bbtIRF3 and bbtIRF7 bind ISRE. When activated, they can be phosphorylated by bbtTBK1 and then translocate into nucleus for target gene tran- scription. As transcriptional repressors, bbtIRF2 and bbtIRF4 can inhibit the transcriptional activities of bbtIRF1, 3, 7, and 8 by competing for the binding of ISRE. Interestingly, amphioxus IRF2, IRF8, and Rel were identified as target genes of bbtIRF1, bbtIRF7, and bbtIRF3, respectively, suggesting a dynamic feedback regulation among amphioxus IRF and NF-kB. Collectively, to http://www.jimmunol.org/ our knowledge we present for the first time an archaic IRF signaling framework in a basal chordate, shedding new insights into the origin and evolution of vertebrate IFN-based antiviral networks. The Journal of Immunology, 2015, 195: 5657–5666.

he IFN regulatory factors (IRFs) were initially identified as DNA sequences similar to the IFN-stimulated response element regulators of the type I IFNs system in the 1980s (1). To (ISRE) (5). Except IRF1 and 2, vertebrate IRFs possess an IRF- T date, nine members, IRF1–9, have been determined in hu- associated domain 1 (IAD1) in the C-terminal region that is re- mans and mice, and an additional IRF-10 is specific to chicken and sponsible for homo- and heteromeric interactions with other family some species of teleosts, including Danio rerio and Paralichthys members or other transcription factors (5). olivaceus (2–4). All vertebrate IRFs share a well-conserved DNA- Studies during the past two decades have revealed five mammalian by guest on September 30, 2021 binding domain (DBD) in the N-terminal region that recognizes IRFs, IRF1, 3, 5, 7, and 8, which serve as positive regulators of type I IFN and IFN-stimulated genes (2, 6). IRF3 and 7, both residing in the cytosol in a latent form in unstimulated cells, are essential for the *State Key Laboratory of Biocontrol, Department of Biochemistry, College of Life Sciences, Sun Yat-Sen (Zhongshan) University, Guangzhou 510275, People’s Repub- retinoic acid–inducible gene 1/melanoma differentiation-associated lic of China; and †Beijing University of Chinese Medicine, Beijing 100029, People’s gene 5–mediated type I IFN gene induction (7). Upon viral infection, Republic of China IRF3 is activated through phosphorylation by inhibitor of NF-kB 1S.Y. and T.Z. contributed equally to this work. kinase (IKK) ε and/or TANK-binding kinase 1 (TBK1), causing it to 2Current address: Shanghai Institute of Immunology, School of Medicine, Shanghai undergo nuclear translocation (8). The subsequent activation of IRF7 Jiao Tong University, Shanghai, China. drives the induction of the IFN-a/b cascade in a positive feedback ORCID: 0000-0002-8755-7543 (S.Y.). way. The formation of an IRF7 and IRF3 heterodimer, rather than an Received for publication September 2, 2015. Accepted for publication October 18, IRF3 homodimer, is presumed to be crucial for the production 2015. of IFN-a/b (9). In addition to the retinoic acid–inducible gene This work was supported by National Nature Science Foundation of China 1/melanoma differentiation-associated gene 5 signaling, mam- Grants 31270018 and 31470846, National Basic Research Program (973) Grant 2013CB835303, New Star of Pearl River on Science and Technology of Guangz- malian IRFs play critical roles in the TLR-mediated IFN responses hou Grant 2014J2200017, and by an open project from State Key Laboratory of (7). The IFN-b induction by the TLR4–Toll/IL-1R domain–con- Biocontrol Grant SKLBC13KF05. S.Y. also is supported by the Guangdong b Outstanding Youth Fund. taining adapter inducing IFN- pathwayismainlymediatedbyIRF3 rather than IRF7, whereas TLR9-MyD88–dependent type I IFN in- The sequences presented in this article have been submitted to National Center for Biotechnology Information Protein Database (http://www.ncbi.nlm.nih. duction is mainly mediated by IRF7 in primary dendritic cells (10). gov/) under accession numbers KM506887–KM506895. The activation of IRF7 requires the formation of a complex consisting Address correspondence and reprint requests to Dr. Anlong Xu, College of Life of MyD88, TNFR-associated factor (TRAF) 6, and IRF7 as well as Sciences, Sun Yat-Sen (Zhongshan) University, 135 West Xingang Road, Guangzhou TRAF6-dependent ubiquitination (10, 11). IRF5 and IRF1 have also 510275, People’s Republic of China. E-mail address: [email protected] been suggested to interact with MyD88 and to act as positive regu- The online version of this article contains supplemental material. lators of IFN gene induction (12). However, IRF4 can compete with Abbreviations used in this article: bbt, Branchiostoma belcheri tsingtauense;Co-IP, coimmunoprecipitation; DBD, DNA-binding domain; IAD1, IFN regulatory factor– IRF5, but not with IRF7, for MyD88 interaction, hence acting as a associated domain 1; IKK, inhibitor of NF-kB kinase; IRF, IFN regulatory factor; negative regulator of TLR signaling (13). In addition to the functions ISRE, IFN-stimulated response element; SG, subgroup; TBK1, TANK-binding kinase 1; in pattern recognition, multiple IRFs (IRF1, 2, 4, and 8) have TRAF, TNFR-associated factor. attracted attention, as they play central roles in the development Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 of immune cells such as dendritic, myeloid, NK, B, and T cells www.jimmunol.org/cgi/doi/10.4049/jimmunol.1501927 5658 CHARACTERIZATION OF AMPHIOXUS IRFs

(14). IRF1, 3, 5, and 8 can also modulate cellular responses in- Biotechnology Information protein database with accession numbers volved in tumorigenesis (15). KM506887–KM506895 (http://www.ncbi.nlm.nih.gov/). In addition to a large number of studies in mammals, some IRFs Plasmid constructions in fish, such as IRF3, 5, 7, 8, 9, and 10 in Japanese flounder (Paralichthys olivaceus) (3, 16, 17); IRF1, 2, 3, 4, 7, and 8 in For the expression of bbtIRFs in HEK 293T cells, PCR fragments encoding the full-length amino acids of bbtIRFs linked to 59-Flag tag and rainbow trout (Oncorhynchus mykiss) (18–20); IRF3, 7, and 9 in 59-hemagglutinin tag were inserted into the expression vector pcDNA 3.0 crucian carp (Carassius auratus) (21–23); and IRF1, 5, 7, and 10 (Invitrogen) and designated bbtIRFs-Flag and bbtIRFs-hemagglutinin, in zebrafish (Danio rerio) (4, 24, 25) have also been found to respectively. The full-length bbtIRF8 was inserted into pCMV- 9 regulate the induction of fish type I IFNs, indicating functional (Clontech) fused with 5 -Myc tag and designated bbtIRF8-Myc. To study the subcellular localization, the full-length bbtIRFs were inserted conservation of IRFs in fish and mammals. A search of genomic into pEGFP-N1 (Clontech). To study the transcriptional activity, the full- and expressed sequence tag databases also revealed the existence length bbtIRFs were inserted into pCMV-BD (Promega). Vectors of of IRF-like genes in invertebrate groups. For example, the IRF2 bbtMyD88, bbtTRAF6, bbtRel, and bbtp105 have been described in homolog identified in pearl oyster (Pinctada fucata) has been our previous studies (31–33). found to have similarity with vertebrate IRF1 and is involved in Preparation of nuclear extraction and EMSA NF-kB activation (26). However, given that type I IFN genes have been found only in vertebrates, IFN-based antiviral responses are Nuclear extract from bbtIRF overexpressed cells was obtained using a NE- PER nuclear and cytoplasmic extraction reagent kit (Thermo Scientific). apparently not present in invertebrates (27, 28), suggesting that the The protein concentration was measured by Pierce BCA protein assay kit IRF family has an ancient evolutionary origin and arose much (Thermo Scientific). Oligonucleotides for ISRE were synthesized and earlier than the IFN system (5, 29). Thus, it is of interest to in- biotinylated by Invitrogen (ISRE sense, 59-TGC AGG GAA ACT GAA Downloaded from 9 9 9 vestigate the function of invertebrate IRFs in the absence of the ACT AAT-3 ; ISRE antisense, 5 -ATTAGT TTC AGT TTC CCT GCA-3 ; ISRE mutant sense, 59-TGC AGG CAA ACT CAA ACT AAT-39; ISRE IFN system and attempt to understand how the IFN-based antiviral mutant antisense, 59-ATT AGT TTG AGT TTG CCT GCA-39). Comple- response coevolved with the IRF family after the divergence of mentary oligonucleotide pairs were annealed at 95˚C for 10 min in 10 mM vertebrates from invertebrate-like ancestors. The amphioxus, of Tris and 1 mM EDTA and cooled slowly at room temperature to create the most basal extant chordate lineage, is a key species model for double-stranded biotinylated and unbiotinylated probes. EMSA was per- formed using a chemiluminescent nucleic acid detection module (Thermo such investigation (30). Characterization of amphioxus IRFs will http://www.jimmunol.org/ Scientific) according to the manufacturer’s protocol. For competitive help to understand the origin of vertebrate IRFs and the coevo- coimmunoprecipitation (Co-IP) between the biotinylated ISRE motif and lution between IRF and IFN-based antiviral responses. bbtIRFs, nuclear extracts were incubated with double-stranded biotinylated ISRE and Streptavidin Sepharose (GE Healthcare). Materials and Methods Culture of amphioxus intestinal cells and poly(I:C) Cell culture and transient transfection transfection HEK 293T cells and HeLa cells were cultured in DMEM medium (Life Adult Chinese amphioxus B. belcheri were obtained from Zhangjiang, Technologies) supplemented with 10% FCS (Life Technologies) and an- China, and reared in aerated sea water with algae. Three days before tibiotics (streptomycin and penicillin; Life Technologies) in a 5% CO2 dissection, amphioxi were transferred to sea water without algae to evac- by guest on September 30, 2021 incubator at 37˚C. Transient transfection was conducted with Lipofect- uate the intestine. To reduce microbial abundance, on the day prior to amine 2000 (Invitrogen) for HEK 293T cells and jetPEI (Polyplus dissection, amphioxi were transferred to sea water that was filtered with a Transfection) for HeLa cells according to the manufacturers’ instructions. 0.45-mm filter and contained 10 mg/ml penicillin. After amphioxi were Gene cloning anesthetized, the amphioxus intestines were extracted, dissected into pieces, and digested for 2 h at 23˚C with 1% collagenase type II (Life Based on the gene models of Branchiostoma belcheri (bbe) IRFs (081950, Technologies). Then the cells were suspended and cultured in medium 246490, 108300, 011020, 108290, 108301, 138190, 108280, and 175880) (DMEM high glucose [HyClone], DMEM/F12 [HyClone], and Leiboviz’s in B. belcheri genome (http://mosas.sysu.edu.cn/genome), partial se- L15 [HyClone] at ratio of 1:1:1) supplemented with 10% FBS (Life quences of Branchiostoma belcheri tsingtauense (bbt) IRFs were amplified Technologies) and antibiotics (penicillin and streptomycin; Life Technol- from another Chinese amphioxus B. belcheri tsingtauense cDNA library ogies) at 23˚C. To test the expression pattern of amphioxus IRFs, the with specific primer pairs. Using the GeneRacer kit (Invitrogen), 59RACE primary amphioxus intestine cells were transfected with poly(I:C) (Sigma- and 39RACE were conducted with the primer pairs derived from the ob- Aldrich) at final concentration of 3 mg/ml by Lipofectamine 2000 and tained partial sequences of bbtIRFs. The full-length bbtIRF sequences collected at 0, 4, 6, 8, 12, and 16 h after poly(I:C) transfection. were then obtained from the same cDNA library. The deduced protein Sectional in situ hybridization was performed according to the sequences of nine bbtIRFs were deposited in the National Center for protocol described previously (34). Luciferase reporter assay, Co-IP,

FIGURE 1. Domain topology of nine amphioxus IRF members and their correspondence among three amphioxus species. All nine bbtIRFs contain the DBD at their N-terminal position. Five conserved Trp in the DBD domains were shown. Six bbtIRFs (BbtIRF2, 3, 5, 6, 7, and 8) contain C-terminal IAD1, and bbtIRF4 has a pair of C2H2-type zinc fingers. Bbe7 indicated bbeIRF7, and Bbe_V2_94 indicated that bbtIRF7 is located on scaffold_94 of the B. belcheri genome (assembly of V2 version). BF7 indicated bfIRF7, and Bf_V2_7 indicated that bfIRF7 is located on scaffold_7 of B. floridae genome (assembly of V2 version). Others follow this instruction. The same color box indicated amphioxus IRF genes with close sequence similarity. The Journal of Immunology 5659 and immunofluorescence imaging were conducted according to the (Fig. 3A). As transcription factors, vertebrate IRFs bind to protocol described previously (31–33). A ISRE ( /GNGAAANNGAAACT), which is found in the pro- moters of type I IFN genes and many other genes that participate Results in immunity and oncogenesis. Further reporter assays showed that Identification and sequence analysis of nine amphioxus IRFs five transcription activators (bbtIRF1, 3, 5, 7, and 8) specifically Several studies have suggested that 10–13 IRF-like genes may activated ISRE reporter, but not NF-kB, AP-1, and PU box reporters exist in the Branchiostoma floridae genome (5, 29, 35). However, in a dose-dependent manner (Fig. 3B, Supplemental Fig. 2A). only nine (gene models 209310, 201596, 68560, 232921, 88707, Particularly, bbtIRF1 effectively recognizes the promoters of human 68559, 89979, 118813, and 178789) are confirmed to have cor- IFN-a1, IFN-a2, IFN-a6, and IFN-b (Fig. 3C). EMSAs confirmed responding orthologs in B. belcheri, whose genomic assembly has that bbtIRF1 and bbtIRF8 bind ISRE in a dose-dependent manner been recently completed (Fig. 1). Genomic comparison between (Fig. 3D). Unlike mammalian IRFs, which can form heterodimers B. floridae and B. belcheri showed that nine amphioxus IRFs have to exert their activity, no heterodimer of the transcription activators conserved genomic loci (Fig. 1). For further sequence comparison, bbtIRF1, 3, 7, and 8 could be identified (Fig. 3E, Supplemental the full lengths of nine IRFs were cloned from another Chinese Fig.2B,2C).However,bbtIRF3,5,6,7,and8withIAD1did amphioxus (B. belcheri tsingtauense) cDNA library and desig- form homodimers, whereas bbtIRF1, 4, and 9 without IAD1 did nated bbtIRF1–bbtIRF9. All nine bbtIRFs contain the character- istic DBD domain of ∼120 aa in the N-terminal position, and six (bbtIRF2, 3, 5, 6, 7, and 8) possess a C-terminal IAD1, suggesting that bbtIRFs are well conserved with respect to protein architec- Downloaded from ture (Fig. 1). Additionally, bbtIRF4 possesses a pair of C2H2-type zinc fingers in the C-terminal (Fig. 1). Similar to vertebrate IRFs, the DBDs of bbtIRF1, 2, 4, 8, and 9 contain the well-conserved Trp repeats spaced by 11–24 residues, whereas the first Trp un- derwent mutation to Phe in bbtIRF3, 5, 6, 7, and the fourth Trp was mutated to leucine in bbtIRF6 (Fig. 1, Supplemental Fig. 1A). http://www.jimmunol.org/ Genomic structure comparison further showed that the DBDs of nine bbeIRFs are well conserved in intron/exon organization, whereas the middle region and IAD1 show significant differences in the pattern of splice junction (Supplemental Fig. 1B). More- over, the ratio of the rate of nonsynonymous to synonymous nu- cleotide mutations of amphioxus IRF DBD is ,0.25, and that of amphioxus IRF IAD1 is ,0.62 (Supplemental Table I). Thus, the DBD domain is likely under purifying selection pressure, sug- gesting that it might have developed a more specialized function by guest on September 30, 2021 than that found in IAD1. Phylogenetic analysis based on deuterostome IRF DBDs sug- gested that predecessors of vertebrate IRF1 subgroup (SG) and IRF4SG might have already existed in the last ancestor of the deuterostome lineage, and it confirmed the orthology between amphioxus IRF1/IRF8 with vertebrate IRF1SG/IRF4SG (Fig. 2). The phylogenetic analysis also suggested that amphioxus has developed its exclusive IRF repertoire, such as IRF3, 5, 6, and 7 (Fig. 2). Many immune and stress gene repertoires have been shown to display rapid expansion and diversification both in the Florida and Chinese lancelet. A notable case is TLR, as 85% of lancelet TLRs became species specific (having no corresponding orthologs in the other lancelet species) within 130 million y ago (36). However, orthologs of nine amphioxus IRFs are supposed to be shared among three amphioxus species (Fig. 2), suggesting that nine amphioxus IRFs may have stabilized and specialized their functions within the last 130 million y. Transcriptional activity of bbtIRFs To reveal the functions of the amphioxus IRF family, full lengths of nine bbtIRFs were first inserted into pCMV-BD to make fusion of Gal4 DBD and bbtIRFs. Then, luciferase reporter assays were performed with a reporter construct pL8G5, which contains FIGURE 2. Maximum likelihood tree of the DBDs of deuterostome both Gal4- and LexA-binding sites. LexA-Vp16 (a fusion protein of IRFs. The standard maximum likelihood method was used to infer the molecular tree. Settings are the WAG+Freqs. [+F] model, the gamma the LexA DBD and the Vp16 transactivation domain) dramatically distribution with invariant sites [G+I] and 200 bootstrap repeats. Align- stimulates the transcription of this reporter gene. When coex- ment was created using the MUSCLE based on the DBDs of deuterostome pressed with bbtIRF1, 3, 5, 7, and 8, transcriptional activity of IRFs. Numbers at nodes indicate bootstrap values. Hemichordata: Sacco- LexA-Vp16 was significantly increased, suggesting that these five glossus kowalewskii (Sk). Echinodermata: Strongylocentrotus purpuratus bbtIRFs are transcription activators. Alternatively, four bbtIRFs (Sp). Cephalochordata: B. belcheri (Bbe), B. belcheri tsingtauense (Bbt), (bbtIRF2, 4, 6, and 9) function as transcription repressors and B. florida (Bf). Vertebrata: Danio rerio (Dr) and Homo sapiens (Hs). 5660 CHARACTERIZATION OF AMPHIOXUS IRFs Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 3. Transcriptional activities of bbtIRFs. (A) Analyses of the transcription activity of bbtIRFs in HEK 293T cells. bbtIRFs were inserted into the pCMV-BD vector to fuse with the GAL4-binding domain. The pL8G5 reporter gene contains both Gal4- and LexA-binding sites, whereas pLexA-Vp16 encodes a fusion protein of the LexA DBD and the VP16 transactivation domain. Means 6 SEM are shown. *p , 0.05 versus corresponding control (by Student t test). (B) All bbtIRFs were analyzed for their transcription abilities in activation of ISRE, NF-kB, AP-1, PU.box, IFN-as (IFN-a1, IFN-a2, and IFN-a6), and IFN-b. Results were obtained from luciferase reporter assays by cotransfection of individual bbtIRFs with the indicated luciferase reporters into HEK 293T cells. Histograms corresponding to (B) can be found in (C) and Supplemental Fig. 2A. (C) bbtIRF1 can activate the ISRE reporter in a dose- dependent manner and recognize the promoter region of human IFN-as (IFN-a1, IFN-a2, and IFN-a6) and IFN-b when overexpressed in HEK 293T cells. Means 6 SEM are shown. *p , 0.05 versus corresponding control (by Student t test). (D) EMSAs showed that bbtIRF1 and bbtIRF8 directly interacted with the ISRE sequence in a dose-dependent manner. bbtIRF, nucleus lysates of cells transfected with expression plasmids of certain bbtIRFs; competitor, ISRE without biotin; control 1, nucleus lysates of unstimulated cells; control 2, nucleus lysates of cells transfected with empty vector; mutant, mutant bio- ISRE replaced natural bio-ISRE; N, no protein. (E) Co-IP results showed that no heterodimer can be formed between bbtIRF3 and any one of bbtIRF1, 7, and 8. Homodimer of bbtIRF3 served as positive control. (F) Co-IP assays showed that bbtIRF3, 7, and 8 with IAD1 can form homodimers, whereas bbtIRF1 without IAD1 could not. (G) Co-IP assays showed that bbtIRF5 and 6 could form homodimers, whereas bbtIRF2, 4, and 9 could not. All Co-IP results are representative of at least two independent experiments. not (Fig. 3F, 3G), suggesting that IAD1 might be essential for Supplemental Fig. 2D). Further EMSAs confirmed that except homodimerization. bbtIRF6, bbtIRF2, 4, 5, and 9 can bind ISRE (Fig. 4B). Co-IP As transcriptional repressors, bbtIRF2 and 4, but not bbtIRF6 assays showed that neither bbtIRF2 nor bbtIRF4 can form hetero- and 9, could suppress the ISRE-dependent transcription activities dimers with any one of bbtIRF1, 3, 7, and 8 (Fig. 4C, Supplemental of bbtIRF1, 3, 7, and 8 in a dose-dependent manner (Fig. 4A, Fig. 2E). However, when coexpressed with bbtIRF1, 3, 7, and 8 in The Journal of Immunology 5661

HEK 293T cells, bbtIRF4 could compete for the binding of ISRE T|A|N)(T|C|G|A|N)TT were identified and collected by a Perl (Fig. 4D). Thus, these results suggested that bbtIRF2 and 4 function script. Finally, a total of 214 candidate target genes of the am- as transcriptional repressors by competing with other IRFs for the phioxus IRF family were identified. Gene functions assembled by binding of ISRE. KEGG showed that the putative target genes of amphioxus IRFs primarily participate in immune defense and cell growth and The presence of a TBK1–IRFs signal axis in amphioxus death, which are coincident with those in vertebrates (Fig. 6A). As Subcellular distribution of bbtIRFs was accessed using the GFP shown in Fig 6B, some predicted genes contain pathogen-associated fusion proteins, and results showed that bbtIRF1, 2, and 4 were molecular pattern recognition domains, such as chitin-binding do- restricted to the cell nucleus, bbtIRF6, 8, and 9 were distributed main, Ig, leucine-rich repeats, and CLECT. Genes related to apo- widely throughout the cytoplasm and nucleus, and bbtIRF3, 5, and ptosis, such as EDA-like (gene ID 154370), caspase 8–like (ID 7 were localized mainly in the cytoplasm (Fig. 5A). In vertebrates, 112340), Bcl2L1 (ID 197290), and genes involved in complement cytoplasmic IRF3 can be activated through phosphorylation by system, such as 042710 with fibrinogen-related domain and 060170 IKK kinase and/or TBK1 (8, 37), resulting in the IRF3 dimer- with C1Q domain were also identified. More interestingly, several ization and removal of an autoinhibitory structure to allow inter- transcriptional factors, including bbeRel (gene ID 274880), action with other coactivators and the translocation into nucleus. bbeIRF2 (ID 246490), and bbeIRF8 (ID 108280), were identified as Because bbtIRF3, 5, and 7 are mostly distributed in the cytosol putative IRF target genes in amphioxus. and have the nuclear localization sequences (Fig. 5B), to analyze k whether a similar activation mechanism is present in amphioxus, Dynamic feedback regulation of amphioxus IRFs and NF- B we first performed Co-IP assays and showed that both bbtIRF3 To confirm the genomic screening results, the upstream sequences Downloaded from and 7 interact with bbtTBK1, but not with bbtIKKa/b (Fig. 5C). of 17 candidate target genes, including bbeIRF2 (gene ID 246490), Then, reporter assays were performed and results showed that bbeIRF8 (ID 108280), bbeRel (gene ID 274880), bbeEDA-like (ID when coexpressed with bbtTBK1 in HEK 293T cells, the ISRE- 154370), and others were cloned from the B. belcheri genome and dependent transcription activities of bbtIRF3 and 7 were increased inserted into pGL3 luciferase vectors to test whether they could be in a dose-dependent manner (Fig. 5D). Moreover, in the presence recognized by bbtIRF1, bbtIRF3, bbtIRF7, or bbtIRF8 (Fig. 7A).

of bbtTBK1, nuclear distribution of bbtIRF3 and 7 were increased Twelve of the 17 candidate genes were confirmed to be targets of http://www.jimmunol.org/ and phosphoserines were detected in HEK 293T cells (Fig. 5E, bbtIRF1, 3, 7, or 8, indicating 80% accuracy. Promoter regions of 5F). Thus, we assumed the presence of a TBK1–IRFs signal axis bbeIRF2, bbeEDA-like, bbecaspase 8–like, bbeBcl2L1, and 154440 in amphioxus. with CLECT domain can be specifically recognized by bbtIRF1; promoters of bbeIRF8, bbeTrim2a-like (299800), and 060170 with Target genes of amphioxus IRF family C1Q domain are recognized by bbtIRF7; promoters of bbeRel, gene Because most amphioxus IRFs can bind to ISRE, we conducted 142900 and 001450 are recognized by bbtIRF3; promoter of gene genomic screening of IRF target transcripts by collecting the 2-kb 170570 is recognized both by bbtIRF1 and bbtIRF8 (Fig. 7B, sequences upstream of ATG of all amphioxus B. belcheri tran- Supplemental Fig. 3A). scripts. Then, sequences with the ISRE mark sequence AA(G|C|A| To further reveal the dynamic transcriptional regulation among by guest on September 30, 2021 G|N)(T|C|A|G|N)GAAA or its reverse strand sequence TTTC(G|C| IRFs and NF-kB in amphioxus, we cultured primary amphioxus

FIGURE 4. Transcriptional regulation among bbtIRFs. (A) The effects of bbtIRF2, 4, 5, 6, and 9 on the ISRE activation of bbtIRF1, 3, 7, and 8 were analyzed. Results were obtained from luciferase reporter assays by cotransfection of indicated bbtIRFs with the ISRE luciferase reporter into HEK 293T cells. The line chart corresponding to (A) is presented in Supplemental Fig. 2D. (B) EMSAs showed that bbtIRF2, 4, 5, and 9, but not bbtIRF6, can bind directly with the ISRE motif. (C) Co-IP results showed that no heterodimer can be formed between bbtIRF4 and any one of bbtIRF1, 3, 7, and 8. The homodimer of bbtIRF8 served as positive control. (D) Biotin pull-down assays showed that bbtIRF4 competitively binds ISRE motif with bbtIRF1, 3, 7, and 8. All Co-IP results are representative of at least two independent experiments. 5662 CHARACTERIZATION OF AMPHIOXUS IRFs Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 5. Functional analyses of bbtIRF3 and 7. (A) Subcellular localizations of GFP-tagged bbtIRFs in HeLa cells. Nucleus stained by DAPI (original magnification 3400). (B) Nuclear localization sequences of bbtIRF3 and 7. (C) Co-IP results indicated that bbtIRF3 and 7 can interact with bbtTBK1, but not with bbtIKKa/b.(D) Both bbtIRF3 and 7 activated the ISRE reporter synergistically with bbtTBK1. The indicated amounts of expression plasmids were transfected into HEK 293T cells. (E) Phosphoserines detected in bbtIRF3 and bbtIRF7 when cotransfected with bbtTBK1 in HEK 293T cells. (F) The nuclear distributions of bbtIRF3 and 7 were increased when cotransfected with bbtTBK1 in HEK 293T cells. intestine cells and transfected them with FITC-labeled poly(I:C) bbtMyD88, both bbtIRF3 and 7 can mount the NF-kB activation (Fig. 7C), as the digestive system is thought to comprise the major mediated by bbtMyD88 in a dose-dependent manner, suggesting immune organs of amphioxus and contain many immune-related that bbtIRF3 and 7 may be involved in amphioxus MyD88/NF- cells, including lymphocyte-like, monocyte-like, and macrophage- kB–dependent signaling (Supplemental Fig. 3D). like cells (30). Then, RT-PCR assays were conducted and results showed that transcripts of bbeIRF1 and bbeIRF8 were upregulated in 2 h and sustained to 12 h, whereas transcripts of bbeIRF2 were Discussion upregulated in 16 h when amphioxus intestinal cells were trans- Characterization of amphioxus IRFs presents a separate fected with poly(I:C) (Supplemental Fig. 3B). Using anti-bbeIRF8 evolutionary event for nonvertebrate deuterostome IRFs mAb, which can recognize the Flag-tagged bbtIRF8 and the en- IRF-like genes have been found in sea sponges, tracing the origin of dogenous bbeIRF8 (Fig. 7D), we confirmed that the protein level the IRF family to multicellular organisms (5, 29). However, in- of endogenous bbeIRF8 was significantly upregulated 8 h after vertebrate IRFs differ from vertebrate IRFs in genomic structure poly(I:C) transfection (Fig. 7E). Because bbeRel was verified as and syntenic gene arrangement (5, 29). For example, no IRF-like the target gene of bbtIRF3, we further performed reporter assays sequence was found in Drosophila and nematodes, but up to nine and showed that bbtRel could mount the ISRE-dependent tran- IRFs were found in cephalochordates. Several studies that tried to scriptional activity of bbtIRF3 and 7, but not bbtIRF1 and 8 (Fig. verify the relationship between amphioxus and vertebrate IRFs 7F, Supplemental Fig. 3C). Moreover, when coexpressed with have linked bfIRF1 (gene ID 209310, as BF4 in their study) to The Journal of Immunology 5663 Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 6. Target genes of the bbtIRF family. (A) The KEGG metabolism pathway of bbtIRF candidate target genes. (B) Domain topology of several amphioxus IRF putative target genes, including chitin-binding domain type (ChtBD2), domain abundant in complement control proteins (CCP), trypsin-like serine protease (Tryp_SPc), Ig (IG), IgC2 type (IGc2), fibronectin type 3 domain (FN3), tyrosine kinase, catalytic domain (TyrKc), Src homology 3 domain (SH3), leucine-rich repeat (LRR), tetratricopeptide repeats (TPR), ankyrin repeats (ANK), plasmid replication region DNA-binding N term (KfrA-N), calcium-dependent carbohydrate binding modules (TNF, CLECT), NHL repeat (NHL), death effector domain (DED), caspase, IL-1b converting enzyme (ICE) homologs (CASc), B cell lymphoma (BCL), fibrinogen-related domains (FBG), Rel homology domain (RHD), Ig-like plexins, transcription factors (IPT), and complement component C1q domain (C1Q). vertebrate IRF1SG (including IRF1 and 2), and bfIRF8 (ID amphioxus IRFs, amphioxus IRF1 is the only one that could bind 118813, as BF3 in their study) to vertebrate IRF4SG (including the promoters of human IFN-a1, IFN-a2, and IFN-a6. Moreover, IRF4, 8, and 9) (5). Nehyba et al. (5) suggested that bfIRF1 and both amphioxus IRF1 and 8 can bind ISRE and the promoter of bfIRF8 locate in chordate linkage group 6 and chordate linkage human IFN-b and be upregulated when amphioxus intestinal cells group 3, which represent the origin loci of vertebrate IRF1SG and were stimulated with poly(I:C), suggesting that amphioxus IRF1 IRF4SG, respectively (5). Our phylogenetic analysis of nine am- and 8 have possessed characteristics key for the recruitment of the phioxus IRFs led to the similar observation and supported the vertebrate IFN system. orthology between nonvertebrate deuterostome IRF1 and verte- In addition to genes with a clear relationship to vertebrate IRF1SG brate IRF1SG. Besides, we provided additional information to and IRF4SG, amphioxus has lineage-specific IRFs, such as IRF3, 5, support the functional relationship between amphioxus IRF1/IRF8 and7.Ininfectedcells,mammalianIRF3and7areactivatedby and vertebrate IRF1SG/IRF4SG. We showed that among nine TBK1-mediated phosphorylation, leading to their dimerization, 5664 CHARACTERIZATION OF AMPHIOXUS IRFs Downloaded from http://www.jimmunol.org/

FIGURE 7. Target genes of bbtIRFs and the dynamic regulation between amphioxus IRFs and NF-kBs. (A) The promoters of putative target genes were inserted in the pGL3 basic vector. (B) bbtIRF1 can recognize the promoter regions of bbeIRF2 and bbeEDA-like; bbtIRF7 recognized the promoter region of bbeIRF8; and bbtIRF3 recognized the promoter region of bbeRel in a dose-dependent manner. (C) Primary amphioxus intestinal cells were cultured and transfected with FITC-labeled poly(I:C) as described in Materials and Methods. Cell stained by FITC (right). Original magnification 3200 (left), 3400

(right). (D) Western blotting was used to test the specificity and efficiency of bbeIRF8 mAb. C1, C2, and C3 indicated three distinct mAb lines of bbeIRF8 by guest on September 30, 2021 based on distinct epitope. One and 3 indicated whole-cell lysis of amphioxus B. belcheri primary intestinal cells, whereas 2 and 4 indicated whole-cell lysis of Flag-tagged bbtIRF8 overexpressed HEK 293T cells. (E) Western blotting showed that the protein level of endogenous bbeIRF8 was upregulated when amphioxus intestinal cells were transfected with poly(I:C). (F) The ISRE-based transcription activities of bbtIRF3 and bbtIRF7 were synergistically mounted when coexpressed with bbtRel in HEK 293T cells. Means 6 SEM and shown. *p , 0.05, **p , 0.01 by Student t test. nuclear translocation, and the transcription of type I IFN genes (6). In in the amphioxus genome (28, 38). By seeking the target genes of the present study, we showed that bbtIRF3 and 7 bind to the ISRE amphioxus IRFs, we found that the putative target genes of am- motif, reside in the cytoplasm, and are phosphorylated by bbtTBK1, phioxus IRFs are mainly involved in innate immunity, cell death, suggesting the presence of a TBK1–IRFs axis in amphioxus. How- and growth. For example, bbtIRF1 activated the promoter of ever, amphioxus IRF3 and 7 did not form heterodimers, showing amphioxus EDA-like, caspase 8–like, and Bcl2L1, whose coun- differences between the vertebrate and amphioxus TBK1–IRFs axis. terparts in vertebrates are involved in the development of the ec- Besides serving as transcriptional activators, some amphioxus- toderm and related to immune regulation (39). bbtIRF1 also specific IRFs, including IRF2, 4, 6, and 9, function as transcrip- recognized the promoters of genes with pathogen-associated tion repressors. Similar to mammalian IRF2, which can repress the molecular pattern recognition domains, such as CLECT, sug- IRF1 transactivation of certain promoters through competition for gesting the dual roles of amphioxus IRF1 in both embryonic de- the same DNA-binding sites (1), bbtIRF2 and 4 mainly reside in the velopment and immune defense. An interesting finding is that nucleus and function as transcriptional suppressors by competing amphioxus IRF2 was found as the target gene of bbtIRF1, whereas with other IRFs for the binding of ISRE. An interesting observation IRF8 was the target gene of bbtIRF7. The upregulation of bbeIRF2 is that amphioxus IRF4 and 8 emerged by lineage-specific duplica- and bbeIRF8 when amphioxi were stimulated with poly(I:C) suggests tion have contrary transcriptional activities, suggesting the functional the feedback regulation among amphioxus IRFs. specialization of amphioxus-specific IRFs. Collectively, the amphi- The relationship between amphioxus IRFs and NF-kBs is also of oxus IRF family not only has genes linked to the predecessors of interest. In vertebrates, both transcription factors play an essential vertebrate IRF1SG and IRF4SG, but it also has lineage-specific IRFs role in immune cell development and function, cooperatively with specialized functions, suggesting a separate evolutionary event regulating the expression of many cytokine genes. For example, for nonvertebrate deuterostome IRFs. IRF3 binds to the p65 subunit of NF-kB to transactivate a set of NF-kB–dependent genes without binding to an ISRE. IRF-1 in- Amphioxus IRFs shed light on the evolution of IFN-based teracts with NF-kB to induce the production of inducible NO antiviral responses synthase, which is relevant for the production of NO as a defense Although the IFN system has been found to be conserved in all against bacterial infection and for elimination of tumor cells (40). tetrapods and fishes, gene models of IFNs have not been identified We have suggested that, unlike IRFs, amphioxus NF-kB has not The Journal of Immunology 5665

FIGURE 8. Putative IRF-based signaling network in amphioxus B. belcheri. When challenged by specific pathogens, bbeSTATs are activated to induce the transcription activator bbeIRF1, which further activates transcription of the transcription repressor bbeIRF2. When bbeIRF3 and bbeIRF7 were activated by bbeTBK1, bbeIRF7 binds to the promoter of bbeIRF8 whereas bbeIRF3 binds to the promoter of bbeRel, leading to the subsequent expression of bbeIRF8 and bbeRel for the application of bbtIRF3/7- based responses to immune demands. Downloaded from http://www.jimmunol.org/ experienced gene duplication. Only two NF-kB genes were pre- 5. Nehyba, J., R. Hrdlickova´, and H. R. Bose. 2009. Dynamic evolution of immune B. floridae B. belcheri system regulators: the history of the interferon regulatory factor family. Mol. sent in both and genomes (33). Thus, it is Biol. Evol. 26: 2539–2550. possible that amphioxus IRFs can interact with NF-kB to control 6. Tamura, T., H. Yanai, D. Savitsky, and T. Taniguchi. 2008. The IRF family the specificity and magnitude of their transcription events. transcription factors in immunity and oncogenesis. Annu. Rev. Immunol. 26: 535–584. Our previous studies not only have identified gene models 7. Seth, R. B., L. Sun, and Z. J. Chen. 2006. Antiviral innate immunity pathways. similar to vertebrate RIG-I–like receptors, but also indicated that Cell Res. 16: 141–147. amphioxus TLRs and Nod-like receptors expanded to 39 and 73 8. Panne, D., S. M. McWhirter, T. Maniatis, and S. C. Harrison. 2007. Interferon regulatory factor 3 is regulated by a dual phosphorylation-dependent switch. J. members, respectively (41). In the present study, we assumed the Biol. Chem. 282: 22816–22822. by guest on September 30, 2021 archaic IRF signaling network in amphioxus B. belcheri as fol- 9. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, lows: when amphioxus cells are stimulated with virus or other Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master regulator of type-I interferon-dependent immune responses. Nature 434: 772– pathogens, bbeIRF3 and 7 can be phosphorylated by bbeTBK1 777. and then translocated into the nucleus for target genes transcrip- 10. Kawai, T., S. Sato, K. J. Ishii, C. Coban, H. Hemmi, M. Yamamoto, K. Terai, M. Matsuda, J. Inoue, S. Uematsu, et al. 2004. Interferon-a induction through tion. The subsequent expression of bbeIRF8 and bbeRel leads to Toll-like receptors involves a direct interaction of IRF7 with MyD88 and the magnification of IRF3/7-based responses to immune demands TRAF6. Nat. Immunol. 5: 1061–1068. (Fig. 8). Because several STAT-binding sites were found in the 11. Honda, K., H. Yanai, T. Mizutani, H. Negishi, N. Shimada, N. Suzuki, Y. Ohba, A. Takaoka, W. C. Yeh, and T. Taniguchi. 2004. Role of a transductional- promoter region of bbeIRF1, another path may be presented as transcriptional processor complex involving MyD88 and IRF-7 in Toll-like re- follows: when amphioxus cells were stimulated by pathogens, ceptor signaling. Proc. Natl. Acad. Sci. USA 101: 15416–15421. transcription of bbeIRF1 was induced by bbeSTATs. Then, the 12. Takaoka, A., H. Yanai, S. Kondo, G. Duncan, H. Negishi, T. Mizutani, S. Kano, K. Honda, Y. Ohba, T. W. Mak, and T. Taniguchi. 2005. Integral role of IRF-5 in genes related to apoptosis as well as the genes related to immune the gene induction programme activated by Toll-like receptors. Nature 434: 243– regulation, such as transcription repressor IRF2, were induced to 249. balance the transcriptional activation of amphioxus IRFs, possibly 13. Negishi, H., Y. Ohba, H. Yanai, A. Takaoka, K. Honma, K. Yui, T. Matsuyama, T. Taniguchi, and K. Honda. 2005. Negative regulation of Toll-like- to avoid excessive immune responses (Fig. 8). signaling by IRF-4. Proc. Natl. Acad. Sci. USA 102: 15989–15994. 14. Yanai, H., H. Negishi, and T. Taniguchi. 2012. The IRF family of transcription factors: inception, impact and implications in oncogenesis. OncoImmunology 1: Disclosures 1376–1386. The authors have no financial conflicts of interest. 15. Savitsky, D., T. Tamura, H. Yanai, and T. Taniguchi. 2010. Regulation of im- munity and oncogenesis by the IRF family. Cancer Immunol. Immunother. 59: 489–510. 16. Hu, G., X. Chen, Q. Gong, Q. Liu, S. Zhang, and X. Dong. 2013. Structural and References expression studies of interferon regulatory factor 8 in Japanese flounder, Para- 1. Harada, H., T. Fujita, M. Miyamoto, Y. Kimura, M. Maruyama, A. Furia, lichthys olivaceus. Fish Shellfish Immunol. 35: 1016–1024. T. Miyata, and T. Taniguchi. 1989. Structurally similar but functionally distinct 17. Ohtani, M., J. Hikima, S. D. Hwang, T. Morita, Y. Suzuki, G. Kato, H. Kondo, factors, IRF-1 and IRF-2, bind to the same regulatory elements of IFN and IFN- I. Hirono, T. S. Jung, and T. Aoki. 2012. Transcriptional regulation of type I inducible genes. Cell 58: 729–739. interferon gene expression by interferon regulatory factor-3 in Japanese flounder, 2. Ikushima, H., H. Negishi, and T. Taniguchi. 2013. The IRF family transcription Paralichthys olivaceus. Dev. Comp. Immunol. 36: 697–706. factors at the interface of innate and adaptive immune responses. Cold Spring 18. Holland, J. W., A. Karim, T. Wang, A. Alnabulsi, J. Scott, B. Collet, Harb. Symp. Quant. Biol. 78: 105–116. M. S. Mughal, C. J. Secombes, and S. Bird. 2010. Molecular cloning and 3. Suzuki, Y., M. Yasuike, H. Kondo, T. Aoki, and I. Hirono. 2011. Molecular characterization of interferon regulatory factors 4 and 8 (IRF-4 and IRF-8) in cloning and expression analysis of interferon regulatory factor 10 (IRF10) in rainbow trout, Oncorhynchus mykiss. Fish Shellfish Immunol. 29: 157–166. Japanese flounder, Paralichthys olivaceus. Fish Shellfish Immunol. 30: 67–76. 19. Holland, J. W., S. Bird, B. Williamson, C. Woudstra, A. Mustafa, T. Wang, 4. Feng, H., Y. B. Zhang, Q. M. Zhang, Z. Li, Q. Y. Zhang, and J. F. Gui. 2015. J. Zou, S. C. Blaney, B. Collet, and C. J. Secombes. 2008. Molecular Zebrafish IRF1 regulates IFN antiviral response through binding to IFNf1 and characterization of IRF3 and IRF7 in rainbow trout, Oncorhynchus mykiss: IFNf3 promoters downstream of MyD88 signaling. J. Immunol. 194: 1225– functional analysis and transcriptional modulation. Mol. Immunol. 46: 269– 1238. 285. 5666 CHARACTERIZATION OF AMPHIOXUS IRFs

20. Collet, B., G. C. Hovens, D. Mazzoni, I. Hirono, T. Aoki, and C. J. Secombes. 31. Yuan, S., S. Huang, W. Zhang, T. Wu, M. Dong, Y. Yu, T. Liu, K. Wu, H. Liu, 2003. Cloning and expression analysis of rainbow trout Oncorhynchus mykiss M. Yang, et al. 2009. An amphioxus TLR with dynamic embryonic expression interferon regulatory factor 1 and 2 (IRF-1 and IRF-2). Dev. Comp. Immunol. 27: pattern responses to pathogens and activates NF-kB pathway via MyD88. Mol. 111–126. Immunol. 46: 2348–2356. 21. Shi, J., Y. B. Zhang, T. K. Liu, F. Sun, and J. F. Gui. 2012. Subcellular locali- 32. Yuan, S., T. Liu, S. Huang, T. Wu, L. Huang, H. Liu, X. Tao, M. Yang, K. Wu, zation and functional characterization of a fish IRF9 from crucian carp Carassius Y. Yu, et al. 2009. Genomic and functional uniqueness of the TNF receptor- auratus. Fish Shellfish Immunol. 33: 258–266. associated factor gene family in amphioxus, the basal chordate. J. Immunol. 183: 22. Sun, F., Y. B. Zhang, T. K. Liu, L. Gan, F. F. Yu, Y. Liu, and J. F. Gui. 2010. 4560–4568. Characterization of fish IRF3 as an IFN-inducible protein reveals evolving 33. Yuan, S., J. Zhang, L. Zhang, L. Huang, J. Peng, S. Huang, S. Chen, and A. Xu. regulation of IFN response in vertebrates. J. Immunol. 185: 7573–7582. 2013. The archaic roles of the amphioxus NF-kB/IkB complex in innate immune 23. Zhang, Y. B., C. Y. Hu, J. Zhang, G. P. Huang, L. H. Wei, Q. Y. Zhang, and responses. J. Immunol. 191: 1220–1230. J. F. Gui. 2003. Molecular cloning and characterization of crucian carp (Car- 34. Yuan, S., K. Wu, M. Yang, L. Xu, L. Huang, H. Liu, X. Tao, S. Huang, and assius auratus L.) interferon regulatory factor 7. Fish Shellfish Immunol. 15: A. Xu. 2010. Amphioxus SARM involved in neural development may function 453–466. as a suppressor of TLR signaling. J. Immunol. 184: 6874–6881. 24. Zhang, Y. B., and J. F. Gui. 2012. Molecular regulation of interferon antiviral 35. Paps, J., P. W. Holland, and S. M. Shimeld. 2012. A genome-wide view of response in fish. Dev. Comp. Immunol. 38: 193–202. transcription factor gene diversity in chordate evolution: less gene loss in am- 25. Sun, F., Y. B. Zhang, T. K. Liu, J. Shi, B. Wang, and J. F. Gui. 2011. Fish MITA phioxus? Brief. Funct. Genomics 11: 177–186. serves as a mediator for distinct fish IFN gene activation dependent on IRF3 or 36. Huang, S., Z. Chen, X. Yan, T. Yu, G. Huang, Q. Yan, P. A. Pontarotti, H. Zhao, IRF7. J. Immunol. 187: 2531–2539. J. Li, P. Yang, et al. 2014. Decelerated genome evolution in modern vertebrates 26. Huang, X. D., W. G. Liu, Q. Wang, M. Zhao, S. Z. Wu, Y. Y. Guan, Y. Shi, and revealed by analysis of multiple lancelet genomes. Nat. Commun. 5: 5896. M. X. He. 2013. Molecular characterization of interferon regulatory factor 2 37. Balkhi, M. Y., K. A. Fitzgerald, and P. M. Pitha. 2010. IKKa negatively regulates (IRF-2) homolog in pearl oyster Pinctada fucata. Fish Shellfish Immunol. 34: IRF-5 function in a MyD88-TRAF6 pathway. Cell. Signal. 22: 117–127. 1279–1286. 38. Loker, E. S., C. M. Adema, S. M. Zhang, and T. B. Kepler. 2004. Invertebrate 27. Wang, P. H., L. S. Yang, Z. H. Gu, S. P. Weng, X. Q. Yu, and J. G. He. 2013. immune systems—not homogeneous, not simple, not well understood. Immunol. Nucleic acid-induced antiviral immunity in shrimp. Antiviral Res. 99: 270–280. Rev. 198: 10–24. 28. Langevin, C., E. Aleksejeva, G. Passoni, N. Palha, J. P. Levraud, and 39. Mikkola, M. L. 2008. TNF superfamily in skin appendage development. Cyto- Downloaded from P. Boudinot. 2013. The antiviral innate immune response in fish: evolution and kine Growth Factor Rev. 19: 219–230. conservation of the IFN system. J. Mol. Biol. 425: 4904–4920. 40. Hiscott, J., N. Grandvaux, S. Sharma, B. R. Tenoever, M. J. Servant, and R. Lin. 29. Huang, B., Z. T. Qi, Z. Xu, and P. Nie. 2010. Global characterization of inter- 2003. Convergence of the NF-kB and interferon signaling pathways in the feron regulatory factor (IRF) genes in vertebrates: glimpse of the diversification regulation of antiviral defense and apoptosis. Ann. N. Y. Acad. Sci. 1010: 237– in evolution. BMC Immunol. 11: 22. 248. 30. Yuan, S., J. Ruan, S. Huang, S. Chen, and A. Xu. 2015. Amphioxus as a model 41. Huang, S., S. Yuan, L. Guo, Y. Yu, J. Li, T. Wu, T. Liu, M. Yang, K. Wu, H. Liu, for investigating evolution of the vertebrate immune system. Dev. Comp. et al. 2008. Genomic analysis of the immune gene repertoire of amphioxus reveals

Immunol. 48: 297–305. extraordinary innate complexity and diversity. Genome Res. 18: 1112–1126. http://www.jimmunol.org/ by guest on September 30, 2021