Advances in Research of Fish Immune-relevant Genes

8/20/2019 Advances in Research of Fish Immune-relevant Genes

1/24

Review

Advances in research of fish immune-relevant genes: A comparative overview

of innate and adaptive immunity in teleosts

Lv-yun Zhu a,b,c, Li Nie a,b,c, Guan Zhu a,b,c,d, Li-xin Xiang a,b,c,⇑, Jian-zhong Shao a,b,c,⇑

a College of Life Sciences, Zhejiang University, Hangzhou 310058, People’s Republic of Chinab Key Laboratory for Cell and Gene Engineering of Zhejiang Province, Hangzhou 310058, People’s Republic of Chinac Key Laboratory of Animal Virology of Ministry of Agriculture, Hangzhou 310058, People’s Republic of Chinad Department of Veterinary Pathobiology, College of Veterinary Medicine & Biomedical Sciences, Texas A&M University, 4467 TAMU, College Station, TX, USA

a r t i c l e i n f o

Article history:

Available online 12 April 2012

Keywords:

FishImmune-relevant genesInnate immunityAdaptive immunity

a b s t r a c t

Fish is considered to be an important model in comparative immunology studies because it is a represen-tative population of lower vertebrates serving as an essential link to early vertebrate evolution. Fishimmune-relevant genes have received considerable attention due to its role in improving understandingof both fish immunology and the evolution of immune systems. In this review, we discuss the currentunderstanding of teleost immune-relevant genes for both innate and adaptive immunity, including pat-tern recognition receptors, antimicrobial peptides, complement molecules, lectins, interferons and sig-naling factors, inflammatory cytokines, chemokines, adaptive immunity relevant cytokines andnegative regulators, major histocompatibility complexes, immunoglobulins, and costimulatory mole-cules. The implications of these factors on the evolutionary history of immune systems were discussedand a perspective outline of innate and adaptive immunity of teleost fish was described. This reviewmay provide clues on the evolution of the essential defense system in vertebrates.

2012 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402. Innate immunity relevant genes and signaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

2.1. Pattern recognition receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402.2. Antimicrobial peptides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422.3. Complement molecules. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432.4. Lectin family members . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452.5. Cytokines in innate immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

0145-305X/$ - see front matter 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.dci.2012.04.001

Abbreviations: PRRs, pattern-recognition receptors; PAMPs, pathogen-associated molecular patterns; TLRs, Toll-like receptors; RLRs, RIG-I like receptors; CLRs, C-typelectin receptors; LRR, leucine-rich repeats; TIR, Toll-IL-1 receptor; ssRNA, single-strand RNA; dsRNA, double strand RNA; NF- jB, nuclear factor-jB; cDC, conventionaldendritic cells; pDCs, plasmacytoid DCs; RIG-1, retinoic acid-inducible gene I; MDA5, melanoma differentiation-associated gene 5; LGP2, laboratory of genetics and

physiology 2; CARD, caspase activation and recruitment domains; RD, repressor domain; CTD, C-terminal domain; ISREs, IFN-stimulated response elements; PGRPs,peptidoglycan recognition proteins; AMPs, antimicrobial peptides; Nramp, natural resistance-associated macrophage protein; NO, nitric oxide; Bf, B factor; Df, D factor;C1INH, complement component 1 inhibitor; CRD, carbohydrate recognition domain; DC-SIGN, DC-specific ICAM-3 grabbing nonintegrin; TCR, T cell receptors; KLH, keyholelimpet hemocyanin; APC, ntigen presenting cells; LECT2, leukocyte cell-derived chemotaxin 2; IFNs, interferons; ILs, interleukins; TNFs, tumor necrosis factors; MBL,mannose-binding lectin; CRP, C-reactive protein; IPNV, infectious pancreatic necrosis virus; PHA, phytoagglutinin; DreIFN B, zebrafish IFN allele B; IFN-crel, IFN-c related;IRF, interferon-regulatory transcription factors; DBD, DNA-binding domain; CAB cells, crucian carp blastulae embryonic cells; MITA, The mediator of IRF3 activation; ISGs,interferon-stimulated genes; ICS, IFN-containing supernatant; ICE, interleukin-converting enzyme; THD, TNF homology domain; CRD, cysteine-rich domain; TNFR, TNFreceptor; TRAF, TNF receptor-associated factor; DD, death domain; AP-1, activator protein-1; TD, TRAF domain; BAC, bacterial artificial chromosome; MHC, majorhistocompatibility complex; b-2m, b-2 microglobulin; IgSF, immunoglobulin superfamily; Ii, invariant chain; Igs, immunoglobulins; BCR, B cell receptor; RAG1,recombination-activating gene 1; RAG2, recombination-activating gene 2; RSS, recombination signal sequence; TD, thymus dependent; TGF-b, transforming growth factor-b;MAPK, mitogen-activated protein kinase; STAT, signal transducer and activator of transcription protein; SOCS, suppressors of cytokine signaling; PIAS, protein inhibitor of activated STATs; KIR, kinase inhibitory region; RLD, RING-finger-like zinc-binding; AD, acidic domain; IRAK, including IL-1R-associated kinase; DIGIRR, double-Ig-IL-1R related molecule; SIGIRR, single-Ig-IL-1R related molecule.⇑ Corresponding authors. Address: College of Life Sciences, Zhejiang University, Hangzhou 310058, People’s Republic of China. Tel.: +86 571 88273287; fax: +86 571

88206582.

E-mail addresses: [email protected] (L.-x. Xiang), [email protected] (J.-z. Shao).

Developmental and Comparative Immunology 39 (2013) 39–62

Contents lists available at SciVerse ScienceDirect

Developmental and Comparative Immunology

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / d c i

http://dx.doi.org/10.1016/j.dci.2012.04.001mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.dci.2012.04.001http://www.sciencedirect.com/science/journal/0145305Xhttp://www.elsevier.com/locate/dcihttp://www.elsevier.com/locate/dcihttp://www.sciencedirect.com/science/journal/0145305Xhttp://dx.doi.org/10.1016/j.dci.2012.04.001mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.dci.2012.04.001


2/24

2.5.1. Interferons and signaling factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452.5.2. Interleukin-1 family members and receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482.5.3. Tumor necrosis factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492.5.4. Chemokines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3. Adaptive immunity relevant genes and their signaling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513.1. Major histocompatibility complex (MHC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513.2. Immunoglobulins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513.3. Costimulatory molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.4. Adaptive immunity relevant cytokines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524. Negative immune-regulators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

1. Introduction

Immunology has been studied for more than 200 years, startingfrom the discovery of cowpox in 1796. The majority of what weknowtoday about the composition, function,andmodulationof im-mune systems are mainly derived from works on mammals. Fromthe evolutionarypoint of view, researchon immunesystems in low-

er vertebrates will become indispensable for better understandingof the evolutionary history of immune systems throughout verte-brates as a whole. As an important link to vertebrate evolution, fishis believed to be an excellent model and an indispensable compo-nent of comparative immunology. This ancient population, fromwhom adaptive immunity originates, possesses complicated innateimmune networks appropriate for innate immunity investigations.Moreover, it serves as a bridge between innate and adaptiveimmunity, thus providing insights into the early events in thedevelopment of the adaptive immune system. However, immunol-ogy in fishes was traditionally less appreciated, largely due to thelack of model organisms for genetic manipulation. The lack of suffi-cient knowledge in fish immunity limits the investigation of immune system evolution, the development of vaccines, and the

selection of disease-resistant breeds. More recently, the emergenceof zebrafish (Danio rerio) as a newmodel organismand theadvance-ment in genome sequencing technology and bioinformatics havegreatly expedited the discovery and functional delineation of genesassociated with immunity in fish (Lieschke and Trede, 2009; Tredeet al., 2004; Van Muiswinkel, 2008). In fact, numerous immune-rel-evant genes for bothinnate and adaptiveimmunity, includingthoseencoding cytokines, complements, lectins, immunoglobulins, andcertain cellsurface molecules,havebeen characterizedfromvariousfishspecies,withChinesescientists contributinggreatly to thesead-vances. In this review, weattempt to give an overview of the recentadvances in fish immune-relevant gene studies by researchers inChina and worldwide.

2. Innate immunity relevant genes and signaling

2.1. Pattern recognition receptors

The recognition of microbial pathogens mediated by pattern-recognition receptors (PRRs) is critical to the initiation of innateimmune responses. PRRs sense the conserved molecular structureof a pathogen, known as pathogen-associated molecular patterns(PAMPs), and induces subsequent host immunity through multiplesignaling pathways that contribute to the eradication of the path-ogen ( Janeway and Medzhitov, 2002). To date, several classes of PRRs, such as Toll-like receptors (TLRs), RIG-I like receptors (RLRs),NOD-like receptors (NLRs), and C-type lectin receptors (CLRs) have

been characterized from many species, including humans, rodents,birds, and teleost fishes (Table 1).

TLRs were the first PRRs to be characterized (Akira et al., 2006).To date, at least 10 and 12 functional TLRs have been characterizedin human and mouse, respectively. All of them are type I trans-membrane proteins that consist of three parts: an N-terminal ecto-domain containing leucine-rich repeats (LRR) that mediate therecognition of PAMPs, a transmembrane region with one a-helix,and a C-terminal intracellular Toll-IL-1 receptor (TIR) domain that

activates downstream signaling pathways (Kawai and Akira, 2011).TLR1, TLR2, TLR4, TLR5, and TLR6 are localized on the cell surfaceand recognize PAMPs of bacteria, fungi, and protozoa, whereasTLR3, TLR7, TLR8, and TLR9 are expressed within intracellular com-partments and recognize nucleic acids. Bacterial PAMPs are mainlyrecognized by TLR1, TLR2, TLR4, TLR5, TLR6, TLR7, and TLR9 (Ku-mar et al., 2011). Viral nucleic acids, including single-strand RNA(ssRNA), double strand RNA (dsRNA), and DNA, are sensed byTLR7/8, TLR3, and TLR9, respectively (Gerlier and Lyles, 2011).After recognizing the respective PAMPs, TLRs will change their con-formation to allow homo- or heterophilic interactions with eachother and recruit TIR domain-containing adaptor molecules (TRIF,MyD88, TRAM) to their own TIR domains (Kawai and Akira,2010). Heterodimers of TLR1-TLR2 and TLR2-TLR6 are expressed

on the cell surface and recruit TIRAP and MyD88 to induce nuclearfactor-jB (NF-jB) activation, resulting in the production of inflam-matory cytokines in conventional dendritic cells (cDC) and macro-phages. TLR2 has been found to be expressed in the endosome of inflammatory monocytes. It can induce the production of type IIFN mediated by IRF3 and IRF7 in response to a virus by recruitingTIRAP and MyD88 (Barbalat et al., 2009). TLR3 is the only TLR thatcannot recruit MyD88; however, it can recruit TRIF to initiate TRIF-dependent signaling in cDCs and macrophages, leading to theinduction of type I IFN and inflammatory cytokines with the helpof IRF3 and NF-jB, respectively. TLR4 is the only TLR that activatesboth the MyD88- and TRIF-dependent pathways (Kawai and Akira,2010). By recruiting TIRAP and MyD88, TLR4 can activate NF-jBand lead to the production of inflammatory cytokines. Then, it will

be transported into phagosomes that contain bacteria and recruitTRAMand TRIF to activate IRF3 and NF-jB for the induction of typeI IFN and inflammatory cytokines. TLR5, which is mainly expressedon the cell surface of the lamina propria of DCs, recruits MyD88 toinduce NF-jB activation and production of inflammatory cyto-kines. In cDCs and macrophages, TLR7 and TLR9 heterodimers re-cruit MyD88 to induce NF-jB-dependent inflammatory cytokineproduction. While in plasmacytoid DCs (pDCs), TLR7 and TLR9 re-cruit MyD88 to activate IRF7 and lead to the production of type IIFN. After CpG DNA stimulation, TLR9 moves to the early endo-somes and triggers MyD88-dependent NF-jB activation. After-wards, it will be transported to lysosome-related organelles forIRF7 activation and stimulation of type I IFN production (Sasaiet al., 2010).

In recent years, a number of TLRs characterized in teleosts werefound to have distinct features andgreater diversity compared with

40 L.-y. Zhu et al. / Developmental and Comparative Immunology 39 (2013) 39–62


3/24

that of mammals. Atpresent, at least 17TLRshave beenfound intel-eosts, among which TLR14, TLR19, TLR20, TLR21, TLR22, and TLR23are non-mammalian TLRs and TLR5S is a soluble isoform of TLR5that appears to be unique in fish (Palti, 2011). So far, counterpartsof mammalian TLR6 and TLR10 have not been identified in fish.The first piscine member of the interleukin-1/TLR family wascharacterized from rainbow trout (Sangrador-Vegas et al., 2000).

Since then, numerous TLRs have been found in teleosts, includingTLR1 in pufferfish (Tetraodon nigroviridis) (Wu et al., 2008); TLR3in common carp (Cyprinus carpio) (Yang and Su, 2010); TLR1 andTLR2 in orange-spotted grouper (Epinephelus coioides) (Wei et al.,2011); TLR5S, TLR20, and TLR21 in cafish (Ictalurus punctatus)(Baoprasertkul et al., 2007); and TLR22 in large yellow croaker(Pseudosciaena crocea) (Xiao et al., 2011).

The genomic organizations of TLRs genes are not quite con-served. For instance, the zebrafish TLR1 gene is comprised of twoexons and one intron, whereas the mammalian TLR1 does not havean intron. The TLR2 gene has one and two introns in human andmurine, respectively, but has no intron in zebrafish and has tenin pufferfish (Wu et al., 2008). A novel member of the TLR familyfound in fish, TLR22, has displayed functions similar to mammalian

TLR3, which is responsible for recognizing dsRNA and initiating theimmune response against dsRNA viruses (Matsuo et al., 2008).

TLR4 senses lipopolysaccharide (LPS) of gram-negative bacteria inmammals, whereas TLR4 in zebrafish has no function in LPS recog-nition and is not present in most other fish species ( Jault et al.,2004). Studies on the molecules that regulate the TLR-initiated sig-naling pathways in teleost fish have also been performed. In zebra-fish, MyD88 was found to be more important than TIRAP inmediating NF-jB activation with insensitivity to LPS stimulation,

whereas both are critical for TLR4-mediated LPS recognition inmammals (Liu et al., 2010). In conclusion, although several TLRsand their signaling pathways are conserved from fish to mammals,a considerable number among them have distinct features and aremuch more diverse in fish.

The RIG-I-like receptor (RLR) family is comprised of three mem-bers: retinoic acid-inducible gene I (RIG-1), melanoma differentia-tion-associated gene 5 (MDA5), and laboratory of genetics andphysiology 2 (LGP2). All of them are DExD/H box RNA helicasesresponsible for the detection of cytoplasmic viral RNA. The RLR family members share a similar molecular structure: a tandem cas-pase activation and recruitment domains (CARD) in the N-terminal(with the exception of LGP2), a DExDH/H box RNA helicase domainin the center, and a repressor domain (RD) embedded within the C-

terminal domain (CTD) (with the exception of MDA5) (Yoneyamaet al., 2004). RIG-I and MDA5 have been well-studied and have

Table 1

Pattern recognition receptors.

PRR Location Recognized PAMPs Mediated signaling pathway Adaptors

TLR

TLR1⁄ Plasma membrane Triacyl lipoprotein Associate with TLR2 mediated signaling pathway TIRAP;MyD88

TLR2⁄ Plasma membrane Lipoproteins; LAM NF-jB signaling pathway TIRAP;MyD88

TLR3⁄

Endosome dsRNA; poly(I:C) NF-jB signaling pathway; IRF3 mediated type I IFNproduction

TRIF

TLR4⁄ Plasma membrane LPS NF-jB signaling pathway TIRAP;MyD88

Phagosome GIPIs; LPS NF-jB signaling pathway; IRF3 mediated type I IFNproduction

TRAM;TRIF

TLR5⁄ Plasma membrane Flagellin NF-jB signaling pathway MyD88TLR6⁄ Plasma membrane Diacyl lipoprotein Associate with TLR2 mediated signaling pathway TIRAP;

MyD88TLR7⁄TLR8⁄ Endosome ssRNA NF-jB signaling pathway; IRF7 mediated type I IFN

productionMyD88

TLR9⁄ Endosome CpG DNA NF-jB signaling pathway MyD88Lysosome-relatedorganelles (LRO)

IRF7 mediated type I IFN production MyD88

TLR10 Endosome Unknown Unknown UnknownTLR11⁄ Plasma membrane Profilin NF-jB signaling pathway MyD88

RLR

RIG-1⁄

Cytoplasm dsRNA NF-jB signaling pathway; IRF3 and IRF7 mediated type IIFN production

MAVS

MDA5⁄ Cytoplasm Poly(I:C); dsRNA NF-jB signaling pathway; IRF3 and IRF7 mediated type IIFN production

MAVS

NLR

NOD1⁄ Cytoplasm iE-DAP dipeptide; Virus NF-jB signaling pathway; IRF3 mediated IFNb production RIP2NOD2⁄ Cytoplasm MDP; ssRNA NF-jB signaling pathway; MAPK signaling pathway; IRF3

mediated IFNb productionRIP2;MAVS

NLRC4 Cytoplasm Flagellin Inflammasome mediated signaling pathway; IL-1bmediated signaling pathway

–

NLRP1 Cytoplasm Lethal toxins; MDP Inflammasome mediated signaling pathway; IL-1bmediated signaling pathway

–

NLRP3 Cytoplasm Toxins; Intracellular bacteria derived molecules;Viral ssRNA and dsRNA

Inflammasome mediated signaling pathway; IL-1bmediated signaling pathway

–

CLR

DC-SIGN⁄ Plasma membrane High mannose; Fucose RAF1-p65 acetylation pathway; NF-jB signaling pathway –

Dectin-1 Plasma membrane b-1,3-glucan SYK-CARD9 pathway; NF-jB signaling pathway; IL-1bmediated signaling pathway –

Dectin-2 Plasma membrane High mannose SYK-CARD9 pathway; NF-jB signaling pathway –Mincle Plasma membrane a-Mannose SYK-CARD9 pathway; NF-jB signaling pathway –

* indicate PPRs identified in fish. Abbreviations: GIPIs, Glycoinositolphospholipids; LAM, Lipoarabinomannan; iE-DAP, g-D-glutamyl-meso-diaminopimelic acid; MDP,muramyl dipeptide; MAPK, Mitogen-activated protein kinase; SYK, spleen tyrosine kinase; CARD9, caspase-recruitment domain family member 9.

L.-y. Zhu et al. / Developmental and Comparative Immunology 39 (2013) 39–62 41


4/24

been found to play different roles in the recognitionof RNA viruses.LGP2 is thought to have less function in virus recognition andsignal transduction, and possibly acts as a regulator of the RIG-Iand MDA5 signaling pathways (Yoneyama et al., 2005). RIG-I rec-ognizes RNA sequences marked with 50triphosphorylated (50ppp)ends and containing a secondary motif like polyuridine as well asa short dsRNA, while MDA5 prefers to recognize poly(I:C) and long,stable dsRNA (Hornung et al., 2006; Kato et al., 2008; Saito et al.,2008). Although RIG-I and MDA5 recognize different kinds of viruses, they are involved in the same signaling pathway that acti-vates NF-jB, IRF3, and IRF7, thus leading to the induction of proin-flammatory cytokines and type I IFN (Eisenacher and Krug, 2011).MAVS (also known as IPS-1, VISA, or Cardif) is the essential adaptorinvolved in this pathway through a CARD domain in the N-terminal(Seth et al., 2005). MAVS is located on the outer membrane of themitochondria and in the membrane of peroxisomes (Dixit et al.,2010; Scott, 2010). When stimulated by virus infection, the CARDdomain of MAVS will interact with the CARD of RIG-I or MAD5,then recruits TRAF3 and TRAF6 to mediate the activation of IRF3,IRF7, and NF-jB (Xu et al., 2005a). Upon phosphorylation, IRF3and IRF7 form homo-/or heterodimers, translocate into the nu-cleus, bind to IFN-stimulated response elements (ISREs), and in-duce type I IFN production and IFN-stimulated gene expression(Honda and Taniguchi, 2006; Honda et al., 2005). In addition totype I IFN production, signaling through RLRs also leads to IFN-kexpression through the same mechanisms (Donnelly and Kotenko,2010).

In the past few years, several RLR family members have beencloned in teleost fish, including RIG-1 in grass carp (Ctenopharyng-odon idella) and channel catfish (Rajendran et al., 2012; Yang et al.,2011a), and LGP2 and MDA5 in rainbow trout (Oncorhynchus my-kiss) and grass carp (Chang et al., 2011a; Su et al., 2010). In contrastto investigations in mammals, the detailed signaling pathways of RLR family members remain elusive. For instance, although RIG-1and MDA5 appear to be well-conserved among vertebrates andcan activate fish antiviral responses, LGP2 appears to be an activa-

tor of the antiviral system in rainbow trout, which is different fromits role in mammals as the suppressor of RIG-1- and MAD-5-acti-vated IFN responses (Komuro and Horvath, 2006). Further studiesare needed to clarify the specific mechanisms involved in thosepathways in teleost fish.

NLRs are a newly identified cytoplasmic PRR family character-ized by the presence of an N-terminal effector binding domain, acentral nucleotide-binding domain, and a C-terminal leucine-richrepeat domain (Chen et al., 2009b). The three domains of NLRsare responsible for downstream signal generation through pro-tein–protein interaction, self-regulation and oligomerization, andpattern recognition, respectively (Martinon and Tschopp, 2005).Based on differences in N-terminal domains, the mammalian NLR family can be divided into three classes: the CARD-containing

NODs and IPAF, the PYD-containing NALPs, and the BIR-containingNAIPs (Carneiro et al., 2007). NLRs play its biological functionthrough two mechanisms: (1) the activation of NF-jB and MAPKsignaling pathway; and (2) the activation of caspase-1, which leadsto IL-1b secretion and programmed cell death fundamental to themolecular platform known as the inflammasome (Ogura et al.,2006; Wilmanski et al., 2008).

Recently, members of NLR family have also been discovered infish. Three distinct subfamilies have been identified in zebrafish,specifically NLR-A, NLR-B and NLR-C. The NLR-A and NLR-B resem-ble mammalian NODs and NALPs, respectively, while NLR-C seemsunique to teleost fish with portions resembling mammalian NOD3and NALPs (Laing et al., 2008). After the discovery of NLRs in zebra-fish, a bunch of NLRs have been identified in many other teleosts,

including channel catfish, rainbow trout, and Japanese flounder. Inchannel catfish, a total of 22 NLRs, including6 of theNLR-A subfam-

ily, 2 of the NLR-B subfamily, 11 of the NLR-C subfamily, and 3unclassified genes, have been characterized. All of these NLRs wereubiquitously expressed in channel catfish tissues and their functionfor initiation of immune responses against intracellular bacterialpathogen is tissue-specific (Rajendran et al., 2012). Several otherrelatively well-studied NLRs in teleost fish include NODs and twoNOD-2 alternatively spliced variants identified in rainbow trout.Theyshare 63%and 61%similaritieswith humanNOD2 respectively,and theirexpression wasmodulated by the inflammatory stimulantIFN-c and IL-1b (Chang et al., 2011b). Both NOD-1 and NOD-2 havebeen characterized in grass carp. They showed different expressionpatterns in responseto LPS, PGN, andpoly(I:C) stimulation,suggest-ing that these two family members might play different roles in thedefense against bacterial and varial infections. The grass carp NOD-2 exhibited significant sequence similarity and close phylogeneticrelationship to the rohu (Labeo rohita) NOD-2 (Chen et al., 2010e).RohuNOD-2 is believedto beimportantto theembryonicinnate de-fense against pathogen infection, indicating that the grass carpNOD-2 may also been involved in the embryonic immune protec-tion(Swainet al., 2012). In Japaneseflounder, a NLR-C-like moleculewas identified. In jfNLR-C over-expressed cells, the expression of IFN-c and IL-1b could be significantly induced by LPS-stimulation.Since the PRRs that recognize LPS have not been found in fish, theNLR-C in teleosts may be responsible for enhancing NF-jB activa-tionupon LPS-stimulation(Unajak etal., 2011). Despitethe progressin NLRs in teleosts, little is known about the precise mechanism of NLR activation and the initiation of subsequent signaling cascades.Further studies are necessary to clarify these issues.

Another group of intracellular PRR family members called pep-tidoglycan (PGN) recognition proteins (PGRPs) have also beenidentified in zebrafish. These PGRPs could bind to/hydrolyze PGNsof bacterial cell walls and participate in the regulation of TLR sig-naling pathways in teleost fish. For instance, TLR2 may work withPGRPs in PG recognition and may activate NF-jB and inflammatorycytokines through the MyD88-dependent pathway (Chang and Nie,2008; Chang et al., 2009).

2.2. Antimicrobial peptides

Naturally occurring antimicrobial peptides (AMPs) are widelypresent in plants, invertebrates such as insects, and all vertebrates.They are ancient components of the innate immune system. Thefirst AMP to be discovered was cecropins in pupae of the cecropiamoth (Hyalophora cecropia)in1981(Steiner et al., 1981). Since then,more than 1000 AMPs have been found, with newly discovered pep-tides continuously being added to the AMP database. There are twodifferent viewpoints about the anti-bacterial mechanisms of AMPs:(1) the amphiphilic structure of AMPs can selectively bind to thebacterial membrane and form transmembrane channels, which de-struct of their membrane integrity and kill incursive bacteria; and

(2)AMPs candirectly enter thebacterial cell to interact with specificintracellular targets to interfere withbacterial growth and metabo-lism, thus playing a role in bacterial death (Wimley, 2010).

Recent studies have cloned some important AMP genes in fish,and have provided more evidence that fish AMPs are importantcomponents of the non-specific immune system. The major groupsof AMPs studied in fish include defensins, natural resistance-asso-ciated macrophage protein (Nramp), NK-lysin, and hepcidin.Defensins could be classified into three subfamilies (e.g. a-defen-sins, b-defensins, and h-defensins) according to the distributionof the disulphide bonds formed by the six conserved cysteineresidues. They play important roles both in innate immunity asmicrobicidal agents and in adaptive immunity as enhancers (Zhuand Gao, 2012). Multiple b-defensins have been identified in bony

fish. For instance, two b-defensins were identified in common carp( C. carpio) (Marel et al., 2012), three were cloned in zebrafish



5/24

(D. rerio) (Zou et al., 2007a), four were detected in rainbow trout(O. mykiss) (Casadei et al., 2009; Falco et al., 2008), and many otherswere found in other fish species including olive flounder (Paralich-thys olivaceus) (Nam et al., 2010), medaka (Oryzias latipes) (Zhaoet al., 2009), and orange-spotted grouper (E. coioides) ( Jin et al.,2010). The orange-spotted grouperb-defensin is a newly identifiedfamily member with 63 amino acids whose sequence shares veryhigh identity with that of green puffer, medaka, rainbow trout,and zebrafish b-defensin-2 (57.14–87.30%) and very low identitywith other b-defensins in fish. In addition to the considerable anti-bacterial activity of grouper b-defensin, it also exhibited antiviralactivity against iridovirus (Rana grylio virus) infection and playsa pivotal role in reproduction/endocrine regulation based on itsdominant expression in conserved pituitary gland and testis. Theseantibacterial and antiviral activities, especially in mucosal systems,have also been explored in other fishb-defensins (Guo et al., 2012;Marel et al., 2012). The discovery of a chemotactic activity of fishb-defensin to head-kidney leucocytes suggests that fish b-defensinscan not only serve in innate immunity against the invasion of viralpathogens, but is also involved in host adaptive immune responses(Cuesta et al., 2011).

The NK-lysingene was initiallyidentified in human andhas beenfound to possessbroad lyticability against many kinds of pathogensincluding bacteria, fungi, protozoa, and parasites. Whereas onlyonecopy of the NK-lysin gene exists in human, three distinct NK-lysintranscripts have been cloned in channel catfish (Wang et al.,2006c,d). With the conservation of all the gene structures, proteinfeatures, and expression patterns of human NK-lysin, its anti-path-ogenic functions can be easily imagined.

Nramp is a macrophage-related AMP in vertebrates. It is an inte-gral membrane molecule with structural features similar to that of transporters, allowing Nramp to interact withperipheral ATP-bind-ing subunits and to limit pathogen invasion. Nramp also plays animportant role in IFN-c/LPS induced L -arginine transport acrossthe macrophage membrane, which induces nitric oxide (NO)production as both signaling molecule and antimicrobial agent

(Blackwell et al., 1994). Nramp genes have been discovered in fish,including common carp (C. carpio), turbot (Scophthalmus maximus),and Japanese flounder (P. olivaceus) (Chen et al., 2006a, 2007b; Saeijet al., 1999). Both turbot and Japanese flounder Nramps are trans-membrane proteins with 12 transmembrane domains. Their de-duced amino acid sequences share high identities of up to 60%with other vertebrate Nramps. Interestingly, Nramp mRNA couldalso be detected in fish embryos, with its expression level graduallyincreasing from 4 to 80 h post-fertilization. This finding suggeststhat Nramp is self-generated during embryogenesis and that it hasessential roles in protecting embryos against various pathogens inthe aquatic environment.

Hepcidin, a cysteine-rich amphipathic peptide, is the mostwidely studied AMP. Initially, it was considered to exert its antimi-

crobial properties on cellular pathogens by penetrating the patho-gen’s plasma membrane to increase permeability and lead to itsdeath(Krauseet al., 2000;Parket al., 2001). However,since hepcidinhas been found to act as an important iron regulator by binding toferroportin (a key iron exporter on macrophages) and inducingferroportin-mediated endocytosis and proteolysis, emerging stud-ies have focused on its regulation of iron metabolism and indirecthost defense (Ganz, 2011; Kroot et al., 2012; Nicolas et al., 2001;Shi and Camus, 2006). To date, hepcidin homologs have been foundin various teleost fish, including marine medaka (Oryzias melastig-ma) (Bo et al., 2011), orange-spotted grouper (E. coioides) (Zhouetal., 2011),ayu(Plecoglossus altivelis) (Chen etal., 2010b), large yel-low croaker (P. crocea) (Wang et al., 2009a), turbot (S. maximus)(Chen et al., 2007a), Japan sea bass (Lateolabrax japonicas) (Ren

et al., 2006), and black porgy ( Acanthopagrus schlegelii) (Yang et al.,2007). In general, most fish species contain more than one kind of

hepcidin; in particular, seven variants were found in black porgy(Yang et al., 2007).

Fish hepcidin genes consist of three exons and two introns. Thefirst exon encodes the signal peptide and a small part of the prodo-main. The other two exons encode the remaining prodomain andthe mature peptide (Bo et al., 2011; Chen et al., 2007a, 2005; Shikeet al., 2004; Wang et al., 2009a; Yang et al., 2007). Most of theknown fish hepcidin propeptides are 85–90 amino acids in lengthand contain 20–26 amino acids in mature peptides (Rajanbabu andChen, 2011b). The propeptides of fish hepcidins consist of threeportions: signal peptide, prodomain, and mature peptide. Eightcysteine residues forming four disulfide bridges in mature peptideregion are well conserved at identical positions in most of the fishhepcidins. The cleavage site (Lys–Arg or Arg–Arg) for propeptideconvertase processing, which is essential for maturation of hepci-din, is also conserved from fish to mammals (Hentze et al., 2010;Nemeth et al., 2004; Piperno et al., 2009). However, there are stilldiscrepancies among fish hepcidin and between fish hepcidins andmammalian counterparts in terms of structure and expression pat-tern. For instance, analysis of all known fish hepcidin sequences orESTs revealed that 9 fish mature hepcidins posses only 4 cysteineresidues, 2 contain 6, and 1 contain 7 (Wang et al., 2009a). The dis-tributions of fish hepcidins are not consistent across different fishspecies. Some fish hepcidins are predominantly expressed in liversimilar to that in mammals (Douglas et al., 2003; Zhou et al.,2011), while some others could also be detected in head kidney,spleen, and gill (Chen et al., 2007a; Yang et al., 2007). Several fishhepcidins, such as black porgy ( A. schlegelii) hepc6 even were ex-pressed at higher levels in head kidney than in liver (Wang et al.,2009a; Yang et al., 2011b). Functionally, two hepcidin isoforms(hepc2 and hepc6) identified in black porgy have been shown tohave wide antimicrobial spectrum against bacteria, fungi, andother microbes (Yang et al., 2011b). The expression of Croceinecroaker (P. crocea) hepcidins was affected by iron status after inter-ference by cadmium exposure in liver, suggesting that fish hepc-edin might play a role in iron metabolism upon heavy metal

exposure (Chen et al., 2008a). Furthermore, fish hepcidins were re-cently found to have antiviral and antitumor activities (Cai et al.,2012; Chang et al., 2011c; Rajanbabu and Chen, 2011a). All theseobservations suggest the existence of functional versatility of hep-cidins in fish.

Aside from the major members of the AMP family in fish dis-cussed above, some other AMPs have also been documented. Theseinclude bactericidal permeability-increasing protein (Xu et al.,2005b), piscidin (Sun et al., 2007), pelteobagrin (Su, 2011), andcathelicidin (Lu et al., 2011). Evidence implies the existence of anAMP spectrum in teleost fish similar to that in mammals and thatthey function as the first line of antimicrobial response.

2.3. Complement molecules

The complement systems have been studied for nearly120 years. According to known results, the origin of the comple-ment systems can be traced back to the coelenterate (Dishawet al., 2005). The complement systems consist of a number of dif-ferent plasma proteins whose interactions play essential roles infighting pathogens. There are three different ways to activate thecomplement on the surface of invading pathogens: the classicalpathway, the mannan-binding lectin pathway, and the alternativepathway. Although initiation of each approach depends on differ-ent elements, they all produce the same anti-infectious effects:(1) opsonization or the promotion of pathogen phagocytosis byphagocytes with complement receptors due to the large numberof activated complement proteins generated and covalently bound

with the invading pathogen; (2) recruitment of phagocytic cells tosites of inflammation and promotion of their activation; and (3)



6/24

formation of membrane-attack complexes to punch holes in thecell membranes of pathogens, thus causing their death. In additionto the major functions of complements in innate immunity, somestudies reported that a series of complements could also playessential roles in adaptive immunity, immunologic memory, andeven in tissue regeneration and tumor growth (DeAngelis et al.,2006; Markiewski et al., 2008; Qu et al., 2009).

Recent studies have shown that bony fish have fully developedcomplement activation pathways. The classical pathway is themost well-studied pathway in fish; in fact, its first subcomponentnamed C1q was first identified in channel catfish ( I. punctatus)(Dodds and Petry, 1993). The full lengths of two C1q-like cDNAshave been cloned in mandarin fish, and both had three compo-nents: a short amino-terminal region; a collagen-like region withhigh repeats of Gly-X-Y triplets, which is conserved throughoutvertebrates; and a globular C1q domain essential for C1q bindingto the constant region of an immunoglobulin or directly to the sur-face of pathogens. However, the specific sequence of GEKGEP,which is present in mannose-binding lectin and other defense col-lagens of mammals and is implicated in opsonophagocytosis, couldbe found in several fish C1q genes only. This finding indicates thatC1q serves not only as a collagen lectin but also as an opsonin infish (Lao et al., 2008). More functional evidence that C1q is in-volved in the classical pathway has been documented in zebrafish(Hu et al., 2010b). In this study, they identified the composition of c1qA, c1qB, and c1qC genes. The synteny and organization of thethree genes were all shown to be well conserved fromfish to mam-mals. Global evolutionary analysis revealed that the production of c1qA, c1qB, and c1qC might be due to gene duplication from a singlecopy of C1qB and that the C1q family is closely related to the Emufamily. For functional analysis, they prepared globular heads of zebrafish C1qA (ghA), C1qB (ghB), and C1qC (ghC) proteins andtheir antibodies. Hemolytic inhibitory assays using anti-ghA, -ghB, and -ghC showed that the heads of C1qA, C1qB, and C1qCcould bind to zebrafish IgM and human IgG and IgM, respectively.Their findings suggest that zebrafish C1qs are involved in the clas-

sical pathway and that they, including their IgM or IgG bindingsites, were structurally and functionally conserved from fish to hu-man. The findings of this study improved current understanding of the evolutionary history of the C1q family and C1q-mediatedimmunity.

In addition to the well-known functions of C1q in the comple-ment systems, several other special and important functions havebeen elucidated in fish. For instance, a newly discovered zebrafishC1q-like molecule was found to be able to inhibit apoptosis in zeb-rafish embryos. Extensive morpholino-mediated knockdown andmRNA over-expression assays demonstrated that this C1q-likemolecule plays an anti-apoptotic role through inhibition of thep53 and caspase 3/9 pathways and was ultimately involved invascular and brain development during embryogenesis (Mei

et al., 2008b). Another C1q family member identified in Gibel carp(Carassius auratus gibelio) was specifically located in follicular epi-thelial cells and was released to the perivitelline space and eggenvelope upon fertilization. This factor could have a role in oocytematuration and ovulation by participating in follicular cell apopto-sis and deletion (Mei et al., 2008a). All other components of the dif-ferent stages of the complement classical pathway have beendiscovered in teleost fish. Chinese researchers have contributed alot through the identification of C2, C6, and C9 factors in large yel-low croaker (P. crocea) and grass carp (C. idella) (Wei et al., 2009).

In the alternative complement pathway, plasma B factor (Bf)can combine with the C3(H2O) generated by the spontaneoushydrolysis of the C3 factor, which is subsequently cleaved by theplasma cathepsin D factor (Df) to form C3 convertase. Both Bf

and Df have been identified in fish species. Bf factors have beenconfirmed in rainbow trout (O. mykiss) (Sunyer et al., 1998), carp

(C. carpio) (Nakao and Yano, 1998), zebrafish (D. rerio) (Gongoraet al., 1998), medaka (O. latipes) (Kuroda et al., 1996), and large yel-low croaker (P. crocea) (Wei et al., 2009). Likewise, the Df gene andprotein have been isolated from carp (C. carpio) and was found tobe dramatically up-regulated by pathogens such as Ichthyophthiri-us multifiliis (Gonzalez et al., 2007; Nakao and Yano, 1998). Sinceother key factors of the alternative pathway, including comple-ment H factor and its receptor, have also been identified in bonyfish (Sun et al., 2010b), it has been proposed that the completealternative complement pathway probably originated in ancientvertebrates.

Several complement components, such as C3 and Bf, have beenconfirmed to be maternal factors that are transferred from mater-nal source into eggs in different fish species, including spottedwolfish ( Anarhichas minor Olafsen) (Ellingsen et al., 2005), rainbowtrout ( O. mykiss) (Løvoll et al., 2006), common carp ( C. carpio)(Huttenhuis et al., 2006), grass carp (Ctenopharyngodon idellus)(Shen et al., 2011b), Atlantic salmon (Salmo salar ) (Løvoll et al.,2007), and zebrafish (D. rerio) (Wang et al., 2008). For instance,complement-mediated bacteriolysis detected in the extracts of zebrafish fertilized eggs could be inhibited by anti-C3 antibodyor inactivated by heating at 45 C (Wang et al., 2008). This bacteri-olytic activity was found to be associated with the integration of lysozyme in zebrafish eggs (Wang and Zhang, 2010). When femalezebrafish was immunized with inactivated pathogens, C3 and Bf were significantly induced and a corresponding rise of these factorswas detected in offsprings. As a result, the hemolytic activity of fer-tilized-egg extracts derived from immunized fish was significantlyincreased (Wang et al., 2009b). These data suggest that the com-plement components transferred from mother to offsprings maybe critical for the protection of early embryos against pathogenicinfections. Interestingly, only the alternative pathway-relatedcomponents appear to contribute to complement-mediated mater-nal immunity. Further investigations are needed to clarfy the exactroles of complement systems in early immune-defense in fish(Wang et al., 2008).

Although there is little evidence to confirm the existence of themannan-binding lectin pathway in teleost fish in the past, thehomologues of mannose-binding lectin (which is the initiatingcomponent capable of activating this pathway) have been charac-terized in three members of the carp (Cyprinidae) family, includingzebrafish (D. rerio), goldfish (C. auratus), and crucial carp (C. carpio)(Vitved et al., 2000), as well as in turbot (S. maximus) (Zhang et al.,2010a) and channel catfish (I. punctatus) (Zhang et al., 2011). How-ever, functional evidences to confirm the existence of mannan-binding lectin pathway in fish are still lacking.

All these thriving works described above imply that all comple-ment systems play major roles in innate immunity and that theyform a bridge to adaptive immunity with a robust and rapidresponse against invadingpathogens. However, the triggered enzy-

matic proteolytic cascades of the complement systems are double-edged swords to the hosts themselves. As a result, there are severalregulatory mechanisms involved in the complement systems toprevent damage to the host cells (Qu et al., 2009). Various fluid-phase and membrane-bound complement regulatory moleculeshave been discovered in fish. For instance, complement component1 inhibitor (C1INH), an important element in controlling the clas-sical pathway, has been cloned from large yellow croaker (Pseudos-ciana crocea). Western blot analysis demonstrated the existence of C1INH proteins in various tissues. When infected with Vibrioanguillarum, the C1INH protein is significantly up-regulated inmost immune-related tissues (Wei et al., 2010). In addition, ananalogue of CD59, which could block the formation of membraneattack complex by binding to a-chain of C8 and b domain of C9,

has been cloned from rainbow trout (O. mykiss) and large yellowcroaker (P. crocea) (Liu et al., 2007a; Papanastasiou et al., 2007).



7/24

The recombinant protein of large yellow croaker CD59 could dra-matically inhibit erythrocyte lysis in tilapia (Oreochromis niloticus)when challenged with large yellow croaker and tilapia serum. Incontrast, no change in red cell lysis was observed after introductionof serum from chicken and mouse. These findings indicate thatCD59 has a species-selective inhibitory effect on complementsystems.

2.4. Lectin family members

Biomembranes contain an amount of carbohydrates mainly inthe form of glycoproteins and glycolipids. Lectin refers to a groupof sugar-binding proteins that can specifically bind to these carbo-hydrate structures independent of enzymes and immunoglobulins,and in turn agglutinates various cells to exert its biological func-tion. Lectins are an enormous superfamily that consists of a greatnumber of members throughout almost all living creatures, includ-ing virus, bacteria, fungi, protists, plants, and animals. In animals,the lectins are pivotal components of innate immune response byinducing phagocytosis, activating platelet, initiating complementsystem, and enhancing the natural killer cell activity (Osorio andReis e Sousa, 2011). With the emergence of adaptive immunity invertebrates, the lectins have corresponding functions as regulatorsof adaptive immune responses by recognizing bacterial or viralcomponents on dendritic cells (DCs) (van Vliet et al., 2008), pro-moting signals that initiate or modulate cytokine responses andinducing lymphocyte maturation, and polarization to the invadingpathogens (den Dunnen et al., 2010). A total of 15 major lectin fam-ilies have been identified in animals, among which C-type lectins,galectins, F-type lectins, rhamnose-binding lectins, and intelectinshave also been demonstrated to exist in fish species (Vasta et al.,2011).

C-type lectin is a large family that includes all known collectinsand selectins in animals. Its most basic feature is its ability to inter-act with carbohydrate residues in glycoprotein through the C-typecarbohydrate recognition domain (CRD) or C-type lectin domain.

Aside from the mannan-binding lectin in the complement systemdescribed above, several other members of the C-type lectin familyhave been discovered in rainbow trout ( O. mykiss) (Zhang et al.,2000), common carp (C. carpio) (Savanet al., 2004), Japanese eel ( An- guilla japonica) (Tasumi et al., 2002), bighead carp ( Aristichthys nobi-lis) (Pan et al., 2010), zebrafish (D. rerio) (Lin et al., 2009a), turbot (S.maximus) (Zhang et al., 2010a), and grass carp (C. idellus) (Liu et al.,2011). All these C-type lectins couldbe significantlyinduced by bac-terial invasion and have shown different functions of agglutinationand inhibition with various invading pathogens. For instance, thepurified bighead carp C-type lectin possesses anti-Vibrio harveyiactivity but no antifungal activity, whereas recombinant turbotC-type lectin can agglutinate the Gram-negative fish pathogenListonella anguillarum butcannot functionagainst the gram-positive

pathogen Streptococcus iniae.DC-specific ICAM-3 grabbing nonintegrin (DC-SIGN), also

known as CD209, is a type II transmembrane protein and is animportant member of the C-type lectin family. The nucleotide se-quence of DC-SIGN was first cloned from the placental gene poolin 1992 and is mainly expressed in monocyte-derived dendriticcells (Curtis et al., 1992). After years of silence on DC-SIGN re-search, it was reported to serve as an important adhesion receptormediating DC maturation and migration, T cell activation, andother activities related to the immune response in 2000 (Geijten-beek et al., 2000a,b). Afterwards, DC-SIGN was found to have theability to bind to gp120 protein of HIV, which is involved in virusimmune escape; implying that the novel pathogenesis of HIV is in-duced by mannose C-type lectin receptors. DC-SIGN mediates the

interaction between DC and T cells after interacting with ICAM-3and ICAM-1, thus providing favorable conditions for T cell recep-

tors (TCR) to recognize the ligands (Ploquin et al., 2004). The firstfish DC-SIGN was identified in zebrafish with a full length cDNAof 1421 bp, encoding a 343-amino acid peptide. Found to be highlyconserved compared with that of mammals, zebrafish DC-SIGNcontains a typical CRD in the extracellular domain, as well as adileucine (LL) motif and a triacidic cluster in the intracellular re-gion believed to be essential for internalization and signal trans-duction of DC-SIGN. Functional analysis showed that T cellactivation and antibody production in vivo induced by exogenouskeyhole limpet hemocyanin (KLH) antigen could be significantlyblocked by anti-DC-SIGN antibody treatment. Phagocytosis analy-sis substantiated that the DC-SIGN protein is not directly involvedin antigen uptake by antigen presenting cells (APC), but might playa role in APC migration and aggregation, followed by T cell activa-tion downstream (Lin et al., 2009a). Since fish dendritic cells havealso been discovered in the last 2 years (Aghaallaei et al., 2010;Lugo-Villarino et al., 2010), DC-SIGN is probably a functional mol-ecule on fish DC that participates in fish adaptive immuneresponse. In addition, the C-type lectin receptor detected in ayu(P. altivelis) could also interact with other innate immune-relatedfactors such as leukocyte cell-derived chemotaxin 2 (LECT2), whichpossesses neutrophil chemotactic activity (Chen et al., 2010a).Demonstration of these two reciprocal molecules could completelydetermine the function of LECT2 signaling and demonstrate themultiple functions of the C-type lectin system.

Intelectin is a newly discovered Ca2+-dependent enteric lectinthat serves as a receptor for pathogen recognition. Since its firstidentification in mouse, intelectins have been widely discoveredin fish and mammals (Komiya et al., 1998). To date, full lengthsequences of intelectins have been cloned in several fish species,including grass carp ( C. idella) (Chang and Nie, 2007), catfish( I. punctatus) (Takano et al., 2008), and zebrafish ( D. rerio) (Linet al., 2009b). Although sequence exploration, phylogenetic analy-ses, and expression pattern have been performed, functional inves-tigation of fish intelectins remains limited.

2.5. Cytokines in innate immunity

Cytokines are a family of lowmolecular weight proteins that areoften glycosylated and are secreted by activated immune-relatedcells upon induction by various pathogens such as parasitic, bacte-rial, or viral components (Salazar-Mather and Hokeness, 2006).They can modulate immune responses through an autocrine orparacrine manner upon binding to their corresponding receptors.Cytokines are derived from macrophages, lymphocytes, granulo-cytes, DCs, mast cells, and epithelial cells, and can be divided intointerferons (IFNs), interleukins (ILs), tumor necrosis factors (TNFs),colony stimulating factors, and chemokines (Savan and Sakai,2006). In innate immunity, macrophages can secret IL-1, IL-6,IL-12, TNF-a, and chemokines such as IL-8 and MCP-1, all of which

are indispensable for macrophage, neutrophil, and lymphocyterecruitment to the infected tissues and their activation as pathogeneliminators (Svanborg et al., 1999). Meanwhile, cytokines releasedby phagocytes in tissues can also induce acute phase proteins,including mannose-binding lectin (MBL) and C-reactive protein(CRP), and promote migration of DCs (DeVries et al., 1999). Allthese cytokines have been found in bony fish, and their functionsand signaling are being explored with great progress (Fig. 1).

2.5.1. Interferons and signaling factors

In human and other mammals, IFNs are the first line of defenseagainst virus infections and can be induced through different sig-naling pathways in response to pathogen infection or pathogen-associated molecular pattern stimulation. Since its first discovery

in chicken in 1957 (Decker et al., 2005), a large number of IFNshave been identified in various species of vertebrates. They are



8/24

classified into three groups with different structures and functions(e.g. type I IFNs, type II IFNs, and type III IFNs) and interact withdifferent cell-surface receptors. Type I IFNs consist of about 20members, including IFN-a, IFN-b, IFN-x, IFN-j, IFN-e, and limitin,while both type II and type III IFNs only have one member calledIFN-c and IFN-k, respectively (Altmann et al., 2003; Iversen andPaludan, 2010). In recent years, progress has been made in fishIFN research. Since IFN-like activity was first discovered in a per-manent cell line from fathead minnow (Pimephales promelas) stim-ulated with infectious pancreatic necrosis virus (IPNV) (Gravell andMalsberger, 1965), similar observations were also described in var-ious fish species induced by different viruses (de Kinkelin andDorson, 1973; Mathews and Vorndam, 1982; Tamai et al., 1993).

Careful property analysis in grass carp (Ctenocpharyngodon idellus)revealed that virus-induced fish IFN shares similar properties withmammalian type I IFN (Shao et al., 1998). Grass carp type I-like IFNcan be produced by leucocytes from head kidney, spleen, andperipheral blood (Xiang and Shao, 2000). In addition to antiviralactivity, it has regulatory effects on macrophage activation, and Tor B cell proliferation (Shao and Xiang, 2001). Notably, IFN-c-likeactivity was also identified from phytoagglutinin (PHA)-stimulatedgrass carp serum and leucocytes (Shao et al., 2001). The activity of this IFN-c-like factor was not neutralized by the antibody specificto grass carp type I IFN, but was highly sensitive to heat (56 C) andacid (pH 2.0). These findings demonstrate that both type I and typeII IFNs exist in fish and play important roles in immune responses.

Fig. 1. Schematic representation of the known cytokine network regulating inflammatory cells functions in fish, including cell proliferation, differentiation, survival or

apoptosis, and numerous gene expression.



9/24

Since it was first cloned in zebrafish (D. rerio), IFN genes havebeen identified from various fish species, including Atlantic salmon(S. salar ), rainbow trout (O. mykiss), fugu (Takifugu rubripes), andcatfish ( I. punctatus) (Long et al., 2004; Robertsen et al., 2003;Zou et al., 2004, 2005a,b, 2007b). Studies suggest that teleost spe-cies possess several genes encoding virus-induced IFNs; specifi-cally, 4 genes in zebrafish, 4 in catfish, and 11 in Atlantic salmon(Aggad et al., 2009; Long et al., 2006; Sun et al., 2009). The classi-fication and nomenclature of fish virus-induced IFNs remain con-troversial. Several reports classified them as type I IFNs based onputative structural features, whereas others suggested that theyare type III IFNs since they are encoded by genes with four intronsand their receptor structures are more similar to that of type IIIIFNs (Robertsen, 2006). With the progress in the study of fish IFNsignaling pathway, researchers tend to agree with the former pointof view and further divided them into three subtypes (e.g. IFNa,IFNb, and IFNc) depending on sequence and expression patternsor named them as IFN1, 2, 3, 4, etc. according to the order in whichthey were discovered in a certain species (Sun et al., 2009). For in-stance, the zebrafish IFN allele B (DreIFN B) gene is considered atype I IFN (Wang et al., 2006b). Some researchers propose thatthese virus-induced IFNs should be referred to as IFNu indepen-dently from the current nomenclature in mammals, since theyshare combined features of mammalian type I and type III IFNs(Stein et al., 2007).

Based on cysteine patterns, virus-induced fish IFN can be di-vided into two groups: the 2 cysteine containing group (group I)

and the 4 cysteine containing group (group II) (Zou et al., 2007b).The two groups bind to two distinct receptor complexes. Group Ibinds to a receptor comprised of CRFB1 and CRFB5, while groupII binds to a receptor comprised of CRFB2 and CRFB5. This bindingpattern differs from that of mammalian type I IFNs, which all bindto a common ternary receptor complex containing two chains of IFNAR1 and IFNAR2 (Aggad et al., 2009) (Fig. 2).

TypeII IFNis encodedby a singlegenein mammals. In contrast, itconsists of two in fish: IFN-c with similar functions to itshomologues in mammals and a teleost specific IFN-c related(IFN-crel)molecule (ZouandSecombes, 2011). The IFN-cmoleculesprimarily promote cell-mediated immunity and have been identi-fiedin lotof species,includingrainbowtrout(O. mykiss), channelcat-fish (I. punctatus), and Atlantic salmon (S. salar ) (Martin et al., 2007;Robertsen et al., 2006; Zou et al., 2005a,b). Meanwhile, IFN-crel,whichis believed to beduplicatedfromthe IFN-c gene, has not beenfully elucidated in terms of function. A recent report on grass carp(C. idella) IFN-crel (gcIFN-crel) revealed that gcIFN-crel productionin vivo could be activated by bacterial PGN, LPS, and the interferoninducer poly(I:C) possibly through a NOD2-dependent mechanism(Chen et al., 2010f ). These findings suggest that both fish type Iand type II IFNs mightparticipatein anti-viral defense,a fundamen-talfunction conservedfromfishto mammals. In addition, type II fishIFN may also be involved in bacterial immune responses (Furneset al., 2009; Stolte et al., 2008; Strandskog et al., 2008).

Interferon-regulatory transcription factors (IRF) are a family of transcription factors essential for modulating the expression of

Fig. 2. Diagram of the deduced anti-virus signalingof typeI IFNsin teleost fish. Virus-infected cells recognizeviral pathogen-associated molecular patterns (PAMP) by patternrecognition receptors (PRR). After TRAF3 or TRAF6 recruitment by PRRs and their associated adaptors, downstream kinases catalyze the phosphorylation of IRF3 or IRF7respectively. These, in turn, formdimers, translocate into the nucleus, attach to the specific motif in the IFN1 (IFN-a1) or IFN3 (IFN-c2) promoter, and regulate the expressionof type I IFNs in collaboration with AP-1 and NF-jB. The newly produced type I IFNs could bind to two groups of receptors on most host cells in fish, both of which contain acommon low affinity CRFB5 chain and a distinct high affinity CRFB1 or CRFB2 chain. After phosphorylation of STAT1 and STAT2 by TYK2 and JAK1 kinases in the cytoplasm,activated transcription factors form a dimer and cross the nuclear membrane associated with IRF9. Following this, the transcription factor complexes bind to interferon-stimulated response elements (ISRE) motifs and the interferon-stimulated genes (ISGs) could be finally induced to exert antiviral functions. Fish IRF3 is an ISG distinct from

that of tetrapods due to its unique ability for a self positive feedback loop of the IFN regulatory system in order to trigger cascade amplification during antivirus response inteleost.



10/24

IFN genes and IFN-stimulated genes by binding to characteristicelements in their promoters. The IRF family consists of ten mem-bers (IRF-1–IRF-10) in vertebrates, although IRF-10 is absent in hu-man and mouse (Chen and Royer, 2010). Recently, a globalcharacterization of IRF family members was performed in fishand a number of other vertebrates and non-vertebrate deutero-stomes. Results showed that all IRF family members ubiquitouslyexist in all vertebrates. They share a relatively high identity ingenomic structure and syntenic gene arrangement, implying thatthey might originate from an analogous pattern under similarselective pressure during evolution. All IRF members have beenpredicted from zebrafish and stickleback. In addition, more thanone IRF-4 gene (three in zebrafish and two in other fish species)and a novel IRF-11 gene have been demonstrated to be specificto teleost fish (Huang et al., 2010; Stein et al., 2007).

Aside from the in silico analyses mentioned above, several IRFfamily members (e.g. IRF-1, IRF2, IRF3, and IRF7) have been clonedfrom fish tissues or cell lines. Phylogenetic analysis showed thatfish IRF family members can be divided into two clusters: one con-sisting of IRF-1 and IRF-2, and another cluster of all other remain-ing IRFs (Shi et al., 2010). IRF-1 was the first IRF family memberknown to activate the IFN-b gene, whereas IRF-2 was describedas a multifunctional transcription factor that exhibits both tran-scriptional activating and repressing activities by antagonizingthe effect of IRF-1 (Sato et al., 2001). IRF-1 and IRF-2 genes havebeen cloned fromC. auratus andE. coioides. They encode a 289-ami-no acid Carassius IRF-1 (CaIRF-1) protein and a 336-amino acid Epi-nephelus IRF-2 (Ec IRF-2) protein, respectively (Shi et al., 2008,2010). Both CaIRF-1 and Ec IRF-2 proteins have an N-terminalDNA-binding domain (DBD) that contains five or six tryptophanresidues, which can form a helix-turn-helix domain and can recog-nize a specific DNA element. CaIRF-1 and Ec IRF-2 could be widelydetected in various fish tissues. Expression of CaIRF-1 and Ec IRF-2 were significantly elevated when challenged with IFN, Grass carphemorrhage virus (GCHV), flounder birnavirus (FBV), poly(I:C), orLPS. This finding suggests that CaIRF-1 and Ec IRF-2 play essential

roles in IFN signaling. Subcellular localization analysis of CaIRF-1and Ec IRF-2 showed that they are nuclear-localized proteins in fishcells. CaIRF-1 has two nuclear localization signals (NLS), whereasmammalian IRF-1 proteins have a single NLS.

Among all IRF members, IRF-3 has the greatest structuralhomology to IRF-7. Both IRF-3 and IRF-7 are early IRFs activatedby TLRs and other signaling modes that play a pivotal role in theinitial induction of type I IFNs (Ozato et al., 2007) (Fig. 2). IRF-3is required for the activation of early-phase IFNs, including IFN-b,which in turn amplify the expression of late-phase IFN-a genesby IFN-induced IRF-7 through the STAT1 pathway. Although typeI fish IFN genes cannot be classified as IFN-a or IFN-b, zebrafishIFN1 (IFN-a1) and IFN3 (IFN-c2) were recently found to be regu-lated by IRF-3 and IRF7, respectively. This finding suggests that

zebrafish IFN1 (IFN-a1) resembles mammalian IFN-b and thatIFN3 (IFN-c2) resembles mammalian IFN-a (Sun et al., 2011a).Recently, IRF-3 and IRF-7 were also characterized in crucian carp( C. auratus) and crucian carp blastulae embryonic (CAB) cells(Sun et al., 2010a; Zhang et al., 2003). IRF-7 expression could besignificantly elevated when CAB cells were treated with GCHV,UV-inactivated GCHV, or CAB-produced IFN protein. Similarly,IRF-3 could be up-regulated by IFN, poly(I:C), and other IFN induc-ers, as was observed in mammals. Typical ISRE motifs were identi-fied in fish IRF-3 promoters (e.g. two in zebrafish and one in theother fish species compared with one in frog IRF-3 promoter),but not in all tetrapods. Either IFN or poly(I:C) was found to inducephosphorylation and cytoplasm-to-nuclear translocation of IRF-3.Recombinant fish IFN was able to upregulate IFN production,

which could be enhanced by overexpression of STAT1 and im-paired by knockdown of STAT1 (Yu et al., 2010a). This finding sug-

gests that the STAT1 pathway might be involved in the antiviralactivity of fish IFNs. In addition, a positive feedback loop of theIFN regulation system that can trigger cascade amplification inantiviral response through the STAT1 pathway may be present inlower vertebrates, including fish and amphibians.

There are two antiviral signaling pathways inducing IRF-3/7activation and type I IFN production in mammals: a TRIF-depen-dent TLR pathway and a TRIF-independent RLR pathway (Meylanand Tschopp, 2006). The full length cDNA sequence of zebrafishTRIF was cloned and was localized to the Golgi apparatus. Trans-fection and luciferase report assays revealed that zebrafish TRIFis responsible for the activation of the IFN promoter and the NF-jB response promoter. This function could be abolished by muta-tion of Ala359 of zebrafish TRIF protein to Pro or His. Althoughthe key elements of the two antiviral signalingpathways were evo-lutionarily conserved in zebrafish, the RIG-I/MAVS pathway wasmore sensitive and specific to poly(I:C) induction than TLR3/TRIF.A novel antiviral mechanism might exist in zebrafish, since LPScannot up-regulate type I IFN expression and no TRAM (which isan adaptor of the TLR4 signal pathway of IFN induction throughinteracting with TRIF in mammals) could be found in the zebrafishDNA database (Fan et al., 2008). On the other hand, a conservedRLR-triggered IFN response mediated by the MITA-TBK1-IRF3pathway was identified in crucian carp (C. auratus). The mediatorof IRF3 activation (MITA), which is essential for triggering RIG-I-mediated IFN-b induction, is an adaptor linking signal transductionbetween MAVS and downstream cytosolic kinase TBK1. Similar toits function in mammals, it activates IRF3/7-dependent IFN re-sponse in fish (Sun et al., 2011a). These findings greatly improvedunderstanding of the molecular and functional evolution of antivi-ral signaling pathways through type I IFN activation.

In the classical IFN pathways, the type I IFN receptor can triggerrapid phosphorylation and activation of receptor-associated JAKsupon binding to type I IFNs, which in turn activate STAT complexesand allow them to form into homodimers or heterodimers. Acti-vated STATs can translocate into the nucleus and associate with

IRFs, followed by attachment of the complexes to ISREs in the pro-moters of interferon-stimulated genes (ISGs) to induce ISG expres-sion, which ultimately exert host antiviral effects (Platanias, 2005)(Fig. 2). In addition to the classical pathway, there is emerging evi-dence showing that non-STAT pathways also play important rolesin the signaling of IFN-responses ( Joshi et al., 2010). Using CABcells, several ISGs have been discovered in fish, including CaIFI58,CaIFI56 , Mx1, PKR-like gene, Gig1, and Gig2 (Grass carp hemorrhagicvirus-induced gene 1 and 2, respectively) (Hu et al., 2004; Jianget al., 2009; Zhang and Gui, 2004a,b,c). All of the newly discoveredISGs could be induced by GCHV, UV-inactivated GCHV, CAB IFN-containing supernatant (ICS), or poly(I:C). Gig1 and Gig2, and thenewly produced type I IFN and exogenous recombinant IFN, exerttheir activity on ISGs through the JAK-STAT pathway in fish species

(Yu et al., 2010a; Zhang and Gui, 2004a).

2.5.2. Interleukin-1 family members and receptors

Interleukin-1 (IL-1) is an important early response pro-inflam-matory cytokine that mediates immune regulation in both innateand adaptive immunity. IL-1 could be secreted by monocytes, acti-vated macrophages, granulocytes, endothelial cells, activated Tlymphocytes, and many other cell types. There are 10 ligand pro-teins in the IL-1 gene family, the main members of which includeIL-1a, IL-1b, and IL-18. IL-1a and IL-1b share the same receptoron target cells and exert similar biological functions, although IL-1b shows more potent function in humoral immune response. Overthe years, IL-1b genes have been identified in various teleost fishspecies, including rainbow trout (O. mykiss), carp (C. carpio), sea

bass (Dicentrarchus labrax), channel catfish (I. punctatus), and yel-lowfin sea bream ( Acanthopagrus latus) (Fujiki et al., 2000; Jiang



11/24

et al., 2008; Scapigliati et al., 2001; Wang et al., 2006e; Zou et al.,1999). In general, only one IL-1b gene seems to exist in fish. How-ever, two IL-1-b-like genes encoding 280-amino acid peptides withhigh identity (94.3%) with each other have been cloned fromcatfish. Multiple alignments showed that catfish IL-1b genesshared low identity with that of mammals and higher vertebrates(approximately 20%) and low identity with other identified fish IL-1b genes (31–38%). There are differences in the expression levels of these two IL-1b genes in various tissues. IL-1b gene 1 is signifi-cantly expressed in liver, head kidney, spleen, intestine, and mus-cle, but minimally in stomach, brain, ovary, skin, and trunk kidney.In contrast, IL-1b gene 2 is highly expressed in all tested tissues ex-cept the brain. In addition, IL-1b gene 1 could be more significantlyinduced than IL-1b gene 2 after bacterial infection. These observa-tions suggest that these two catfish IL-1b genes may play differentroles in anti-bacterial response (Wang et al., 2006e).As inmost fishspecies, only one IL-1b gene could be identified in yellowfin seabream. This gene consists of a 121-bp 50-UTR, a 762-bp ORF, anda 342-bp 30-UTR and encodes a 253-amino acid IL-1b protein.The yellowfin sea bream IL-1b gene showed low identities withthat of higher vertebrates (25.3–26.6%) and had greatly variedidentities with different fish species (26.6–88.5%). This finding sug-gests that a rapid sequence divergence might have occurred in tel-eosts, indicating that there was rapid evolution of functions of fishIL-1b ( Jiang et al., 2008). All fish IL-1b genes identified displayed nointerleukin-converting enzyme (ICE) cut site, which is essential forIL-1b maturation from precursor to functional state in mammals.Therefore, the mechanisms underlying the processing of IL-1b pre-cursors in fish remain to be elucidated. The yellowfin sea bream IL-1b precursor was predicted to cut between Tyr87 and Thr88, andgives a mature peptide with 166-amino acid residues.

IL-1 activates target cells by binding to IL-1 receptors on the cellsurface and ultimately triggering inflammation to cope with path-ogen infection. Two IL-1 receptors (IL-1RI and IL-RII) that displayopposite functions in IL-1 signaling exist in mammals. IL-1RI isthe positive receptor participating in all known IL-1 functions,

whereas IL-1RII antagonizes IL-1 function as a decoy receptor. Fishalso possess both contrary IL-1Rs, indicating that the compositionof fish IL-1Rs was conserved in mammals. Both IL-1RI and IL-1RIIhave been found in various fish species, including salmon (S. salar ),rainbow trout (O. mykiss), pufferfish (Fugu rubripes), and zebrafish(D. rerio) for IL-1RI (Huising et al., 2004); and rainbow trout (O. my-kiss), gilthead seabream (Sparus aurata), and Japanese flounder (P.olivaceus) for IL-1RII (Fan et al., 2010; Lopez-Castejon et al.,2007; Sangrador-Vegas et al., 2000). The Japanese flounder IL-1RII gene contains a 1263-bp ORF, a 100-bp 50-UTR, and a 430-bp 30-UTR containing three repeated AU (ATTTA) motifs. It encodesa 420-amino acid IL-1RII protein that contains two Ig-like domainsin the extracellular region for recognizing IL-1 but no TIR domainin the intracellular region for signal transmission. In general, this

structure is conserved in mammals, although three Ig-like domainsexist in human and mouse IL-1RIIs. Fish IL-1RII also seems to func-tion as a negative regulator of IL-1 signaling in inflammatory re-sponse, indicating that this function was conserved in highervertebrates. However, some reports suggest that fish IL-1RII mighthave some function in signal transduction, since Japanese flounderIL-1RII possesses 4 proline residues important for signal transduc-tion in IL-1RI (Fan et al., 2010; Luheshi et al., 1993; Mirtella et al.,1995). These findings suggest that a new mechanism for IL-1RIImight exist in addition to its negative regulatory function, whichimplies that IL-1 signaling could be more complicated than previ-ously known.

2.5.3. Tumor necrosis factors

The TNF superfamily plays a key role in inflammation, host de-fense, autoimmunity, organogenesis, cellular apoptosis, and differ-

entiation. At present, at least 19 members of this family have beenidentified in mammals (Ware, 2003), including the well-knownTNF, Fas ligand, CD27 ligand, CD40 ligand, and TNF related apopto-sis inducing ligand (TRAIL) (Gruss, 1996; MacEwan, 2002). Theyshare similar structures and biological functions, with a conservedextracellular C-terminal domain called the TNF homology domain(THD). Most of them are type II transmembrane proteins. TheTHD binds to the cysteine-rich domain (CRD) of their correspond-ing receptors to initiate appropriate biological responses. Untilrecently, only a limited number of TNF superfamily members havebeen identified in teleosts, and a great number of them are similarto mammalian TNF-a. This factor has been identified from severalfish species, including mandarin fish (Siniperca chuatsi) (Xiao et al.,2007), zebrafish ( D. rerio) (Savan et al., 2005c), common carp(C. carpio) (Saeij et al., 2003), and turbot (Psetta maxima) (Ordaset al., 2007). Several other family members have also been charac-terized in teleosts, such as TRAIL-like (Chang et al., 2006; Gao et al.,2008), CsTL (Zhang et al., 2008), and BAFF (Ai et al., 2011). TNF-bdoes not seem to exist in fish. Similar to their counterparts inmammals, teleost TNF-a genes consist of four exons and three in-trons. However, the identity of fish TNF-as are generally lowerthan that in mammals, and the average size of TNF-as in teleostis a little larger than that of mammals (Goetz et al., 2004). Zebra-fish TNF-a can mediate cell death and regulate the expression of some essential molecules in relative pathways, suggesting thatthe function of this molecule was conserved in mammals (Wanget al., 2011). The first TRAIL gene cloned and characterized in tele-ost is in grass carp (C. idella) (Chang et al., 2006). This gene consistsof five extrons and four introns, similar to TRAIL genes in mam-mals. However, in mandarin fish (S. chuatsi), the last intron is sep-arated into two parts, so that this molecule contains a fifth intron(Gao et al., 2008). Consistent with its expression pattern in mam-mals, fish TRAIL transcripts can be detected in a wide range of tis-sues, such as spleen, kidney, and intestine. In grass carp, expressioncan also be detected in liver, which is distinct from the observationin adult human. Recently, fish TRAIL has been shown to function in

inducing apoptosis like its mammalian counterparts (Gao et al.,2008), indicating the conservation of the structure and biologicalfunctions of this molecule.

TNF fulfills its functions by interacting with its specific recep-tors. The TNF receptor (TNFR) family members are classified intothree groups: TNF receptor-associated factor (TRAF) binding recep-tors, death domain (DD)-containing receptors, and decoy receptors(Aggarwal, 2003; Dempsey et al., 2003; Locksley et al., 2001). In-stead of having DDs, TRAF binding receptors (including TNFR2)have motifs that recruit TRAF proteins to exert functions (Chunget al., 2002). DD-containing TNFRs, including TNFR1 and FAS, acti-vate caspase cascades through DD-containing signaling adaptors,resulting in caspase activation and cell apoptosis. These adaptorsbind to TNFRs and to each other through homo- and hetero-typic

interactions to induce apoptosis. To date, six TRAF molecules havebeen identified in mammals (Xu et al., 2008). They share commonstructures, such as a single RING finger at the N-terminal (with theexception of TRAF1, which does not have an N-terminal RING fin-ger domain), and several zinc fingers and a TRAF domain at the C-terminal. The TRAF domain is responsible for binding to associatereceptors. Among the six TRAF molecules, TRAF2, TRAF5, andTRAF6 are adaptors that connect receptors to downstream kinasecascades. This binding results in activation of NF-jB and activatorprotein-1(AP-1) (Inoue et al., 2000), which in turn leads to apopto-sis, inflammation, or cell survival. TRAF1 can indirectly associatewith TNFR2 with the help of TAFR2 to form a heterodimeric com-plex. TRAF1 can also interact with TNFR1 and act as a substrate forcaspases activated by receptors with DDs. TRAF3 antagonizes the

effects of TRAF2 in NF-jB activation (Watts, 2005). TRAF4 is a un-ique member of this family in terms of both structure and function.



12/24

The most conserved residues R, Y, and S in TRAF1, 2, 3, and 5 arereplaced by S, F, and F in TRAF4, so that this molecule has reducedinteraction with the TNFR family. Few TNFRs interact with TRAF4molecules and only under given conditions (Kedinger and Rio,2007). Although the TRAF gene has been characterized in manyvertebrates, little is known about it in fish. The first detaileddescription of the structure and expression of TRAF in fish is TRAF1in grass carp (

C. idella). Compared with TRAF1 in other vertebrates,

its identity varies from 44% to 72% in terms of the most importantdomain and varies from 52% to 58% in reference to the TRAF do-main (TD) (Xu et al., 2008). In mammals, TRAF1 is selectively ex-pressed in spleen, lung, tonsils, and testis (Rothe et al., 1994), butnot in kidney, liver, heart, and brain. However, in grass carp, TRAF1mRNA is widely distributed even in heart, head kidney, thymus,brain, gill, liver, and spleen. TRAF1 may play a negative role inthe regulation of TNF signaling in teleost, although the main func-tion of this molecule and its homologs in bony fish remains to beinvestigated.

2.5.4. Chemokines

Chemokines refer to a group of chemotactic cytokine familymembers that are released by most infected tissues in the earlystages of infection. They are small heparin-binding secreted cyto-kines recruiting monocytes, neutrophils, and other effector cellsfrom vessels towards the focus of infection. Chemokines could bedivided into several families (e.g. CXC, CC, C, and CX3C) accordingto the organization of cysteines next to the amino terminal. Todate, approximately 50 members and 20 receptors in the chemo-kine family have been discovered in humans (Ransohoff, 2009; Sal-azar-Mather and Hokeness, 2006). The complex interplays amongdifferent chemokines and between chemokines and other cyto-kines remain the highlights of immune response studies.

The two most important and deeply studied families of chemo-kines are the CC chemokines (characterized by two adjacent cys-teine residues next to the N-terminal) and the CXC chemokines(characterized by two cysteine residues separated by one amino

acid next to the N-terminal). These two main chemokine familiesand their receptors have also been found in bony fish. cDNAhybridizations and bacterial artificial chromosome (BAC) finger-printing and clustering have revealed 26 CC chemokines in catfish,among which the full length cDNA of 23 CC chemokines were suc-cessfully sequenced (Bao et al., 2006; He et al., 2004; Peatmanet al., 2006). Compared with human chromosome 17, which con-tains the largest group of CC chemokine genes, several catfish CCchemokines share high identities on the same chromosomalstretch in humans, such as SCYA108, SCYA123, SCYA120, SCYA124,and SCYA104. In addition, SCYA101 and SCYA107 have been shownto be specific to fish. SCYA112 resembled human CCL20 located onchromosome 2. EST evidence suggests that other additional CCchemokines exist in catfish in addition to these 26 identified CC

chemokines. Whether all these chemokines are translated intofunctional proteins or some exist only as redundant molecules stillneed to be elucidated. After challenging with Edwardsiella ictaluri,only seven of the 26 CC chemokines were up-regulated in spleenand kidney of catfish; specifically, SCYA105, SCYA109, SCYA112,SCYA113, SCYA115, SCYA117, and SCYA125. These seven inducibleCC chemokines may be involved in the recruitment of inflamma-tory cells and in the induction of inflammatory response in fish.Moreover, the expression of SCYA116 and SCYA121 were down-regulated, while 16 of the 26 CC chemokines did not exhibit anychange in expression upon bacterial infection. Interestingly, CCchemokines are not only involved in bacterial defense, but alsoother biological functions in mammals. For instance, CC chemo-kines play important roles in the regulation of the neuroendocrine

system and antitumor immunity (Balkwill, 2012; Rostene et al.,2011). Thus, the different expression patterns of the 26 CC chemo-

kines in catfish after challenging with bacteria may imply thatthese CC chemokines may also exert various functions in additionto bacterial defense. The full-length cDNA of a CC chemokine(LycCC2) has been characterized in large yellow croaker (P. crocea).It consists of a 50-UTR of 44 bp, a 30-UTR of 242 bp, and 282-bp ORFencoding a peptide of 93 amino acids. Multiple alignments showedthat LycCC2 shares low identity with chemokines of mammals andother fish species, and a slightly higher identity with a small induc-ible cytokine from cichlid fish (Paralabidochromis chilotes). Similarresults were also seen in phylogenetic tree analysis. LycCC2 geneis widely expressed in various tissues except for the liver and isquickly up-regulated by either poly(I:C) or inactivated bacterialvaccine in spleen and kidney. In addition, in vitro determinationof LycCC2 chemotactic activity revealed that this chemokine pos-sesses the ability to attract peripheral blood leucocytes, especiallyat the concentration of 700lg/ml. These findings demonstrate thatLycCC2 is a novel CC chemokine with strong chemotactic activityand possible roles in inflammatory responses.

Investigations of fish CC chemokine receptors have accompa-nied fish CC chemokine discovery. A total of 17 CC chemokinereceptor members, together with their gene structure and organi-zation, have been determined in zebrafish (Liu et al., 2009). Theyare homogenous to CCR6, CCR7, CCR8, CCR9, and CCRL1 of mam-mals. No CCR1, CCR2, CCR3, CCR4, CCR5, and CCR10 homologscould be identified in zebrafish. Expression analyses revealed thatmajority of these receptors are expressed in fertilized eggs or earlyembryos, indicating they are of maternal origin. However, differen-tial expression patterns were found in adult fish, suggesting thatthey have functional diversity.

IL-8 is one of the first CXC chemokines to be discovered in fish(Lee et al., 2001) and is classified into two groups (Abdelkhaleket al., 2009; van der Aa et al., 2010). An IL-8 ortholog has beenidentified in fugu (T. rubripes) (Saha et al., 2007), black seabream( A. schlegelii) (Zhonghua et al., 2008), and Half-smooth tongue sole(Cynoglossus semilaevis) (Sun et al., 2011b), which are all economi-cally valued fish in China. C. semilaevis IL-8 (CsIL-8) possesses the

conserved CXC tripeptide motif (35CRC37) but lacks the typical ELR motif associated with IL-8 neutrophil-attracting ability, similar toother fish species but different from mammalian and bird IL-8.Functional analysis showed that CsIL-8 could be significantlystimulated upon bacterial challenge. It has chemotactic activity onkidney lymphocytes and peripheral blood leukocytes in a dose-dependent manner.

CXC chemokines can be subdivided into ELR + and ELR groupsaccording to the presence or absence of the ELR domain, respec-tively. Functionally, the ELR + group chemokines attract neutro-phils, while the ELR group chemokines attract lymphocytes andmonocytes. IL-8 is an ELR + CXC chemokine in mammals, but issuggested to be an ELR CXC chemokine in fish. Aside from IL-8,several other ELR + and ELR CXC chemokines have been identified

in fish. For example, an ELR

CXC chemokine with close evolution-ary relationship with CXCL14 has been identified in turbot (S. max-imus). A CXCL12-like homologue with no ELR motif was cloned inlarge yellow croaker (P. crocea) (Liu et al., 2007b; Wan and Chen,2009). This large yellow croaker CXCL12 (LycCXCL12) shared highidentity with other fish CXCL12 molecules (57–69%) but loweridentity with that of mammals (32–36%). Functional analysis todemonstrate the role of the ELR motif in ELR + CXC chemokine of black sea bream has been performed (Cai et al., 2009). The blacksea bream CXC chemokine (BS CXC) possesses an ELH motif insteadof ELR. The ELH motif could specifically recruit blood neutrophilsbut showed weak ability to induce macrophage migration, whichis quite different from mammalian ELR + CXC chemokines. Interest-ingly, when wild-type His of ELH motif was mutated into Arg to

form a complete ELR motif, the mutant BS CXC could induce bothchemotaxis of blood neutrophils and macrophage migration.



13/24

Meanwhile, shortening of the N-terminal of the ELR motif of BSCXC showed no effect on its chemotactic properties. However, allrecombinant BS CXC or CXC mutants cannot induce migration of mammalian blood neutrophils and macrophages, implying thatother motifs in addition to the ELR motif may be necessary forthe recruitment of higher animal blood leucocytes.

There are seven CXC receptors in mammals referred to asCXCR1–7 (Oliveira and Lukacs, 2003). Most CXCRs have been iden-tified in fish except for CXCR6, including CXCR1 in mandarin fish(S. chuatsi) (Chen et al., 2009a), CXCR1 and CXCR2

Advances in Research of Fish Immune-relevant Genes

Documents

Transcript of Advances in Research of Fish Immune-relevant Genes