Independent Duplications of Bf and C3 Complement Genes in the Zebrafish

Independent Duplications ofBf andC3 Complement Genes in theZebrafish

R. GONGORA*, F. FIGUEROA & J. KLEIN

Max-Planck-Institut fu¨r Biologie, Abteilung Immungenetik, Corrensstrasse 42, D-72076 Tu¨bingen, Germany

(Received 26 April 1998; Accepted in revised form 13 August 1998)

Gongora R, Figueroa F, Klein J. Independent Duplications ofBf andC3Complement Genes in the Zebrafish.Scand J Immunol 1998;48:651–658

As part of an ongoing project aimed at the characterization of the MHC system in bony fishes, we haveattempted to identify the class III region in the zebrafish (a teleost), a region that in higher vertebrates containsgenes coding for complement proteins C2, Bf and C4. We obtained several genomic PAC clones byhybridization with a zebrafishBf probe, previously identified in our laboratory, and searched these for thepresence of other class III genes. We were able to obtain a secondBf-like gene, however, we were unable todetect anyC2- or C4-like genes. By using highly degenerated primers, we extended our search to ahepatopancreas cDNA library and amplified from it clones corresponding to three differentC3-like genes,and also theBf genes, but not anyC2- or C4-like genes. The zebrafish therefore contains twoBf and threeC3loci but apparently noC2andC4 loci. Independent duplications of theBf andC3genes in bony fishes suggestthat complement plays a prominent part in the immune response of this class of vertebrates.

Felipe Figueroa, Max-Planck-Institut fu¨r Biologie, Abteilung Immungenetik, Corrensstrasse 42, D-72076Tubingen, Germany

INTRODUCTION

The complement system plays a pivotal role in the host defenceagainst infection by micro-organisms [1]. It consists of plasmaproteins which, when activated by antibodies or cellular surfaces,interact in a cascade-like fashion to produce opsonizing sub-stances and the membrane attack complex capable of directcytolysis [2]. Two activation pathways, classical and alternative,lead to the cleavage of the central component C3 [3]. Theresulting C3b fragment then covalently tags micro-organismsfor clearing from the circulation or tissues [4]. The classicalpathway is initiated by the interaction of immunoglobulins withantigen and the attachment of the C1 components to the resultingimmune complexes, or by a direct interaction of the mannose-binding protein with polysaccharides on bacterial or viral sur-faces [5]. It leads to the formation of the C3 convertase from theC2 and C4 components, which cleaves the C3 molecule. Theinitiation of the alternative pathway involves the spontaneousactivation of the C3 component by the incorporation of watermolecules, the reaction being accelerated by activating substances.

The alternative pathway also leads to the formation of a C3convertase, composed of a fragment of factor B (Bf) and C3b.Both convertases ultimately cleave the C5 component and thusinitiate the formation of the membrane attack complex in thelytic pathway [1]. The alternative pathway is phylogeneticallyolder and is present even in jawless vertebrates (agnathans),which seem to lack the ability to produce antibodies [6].

Some of the components of the two pathways have been pro-duced by gene duplications, as indicated, for example, by theobservation that C2 and Bf, as well as C4 and C3, have homo-logous roles in the two convertases and that they are phylogen-etically related in sequence and in genome organization [7].Agnathans appear to possess only the alternative pathway actingas an opsonizing system [8]. Functional analysis suggests thepresence of both classical and alternative (and lytic) pathways incartilaginous and bony fishes [9, 10]. Molecular evidence forboth components of the classical C3 convertase (C2 and C4)exists only in mammals, although aC4-like sequence has beenalso found in Xenopus[11]. Components of the alternativepathway C3 convertase (Bf and C3) are known to exist inagnathans [12, 13], teleosts [14–17] and tetrapods [18–24].

C2, Bf and C4 are linked in mammals, forming part of theMHC class III region, whereasC3 resides on another chromo-some [25, 26];Xenopus Bf[19] andC4 [11] genes are also linked

Scand. J. Immunol.48, 651–658, 1998

q 1998 Blackwell Science Ltd

* Current address: Division of Developmental and Clinical Immunology,University of Alabama at Birmingham, 378 Wallace Tumor Institute, Birming-ham, AL 35294-3300, USA

to MHC, whereas linkage data forC3 are not available [23] andC2has not been yet detected in this species.Bf andC3sequenceshave been described in bony fishes [14–16]. In the zebrafish(Danio rerio) from the order Cypriniformes [27] and in otherteleosts (Satoet al., unpublished observations), MHC class I,class II andBf (class III) genes, are on different chromosomes.The aim of the present study was to characterize the genomicregion containing theBf gene in the zebrafish and to determinewhether it also contains genes coding for other complementcomponents as it does in mammals.

MATERIALS AND METHODS

Fishes. Both adults and haploid embryos were derived from zebrafish(Danio rerio) laboratory stocks maintained at the Max-Planck-Institutfur Biologie, Tubingen. Haploid embryos were produced as describedpreviously [27].

Screening of the genomic PAC library. The zebrafish genomic librarywas obtained in the form of gridded filters from the Resource Centre ofthe German Human Genome Project (Berlin-Charlottenburg, Germany).The filters were prepared from anMboI partial digest of DNA isolatedfrom pooled red blood cells of 200 fish belonging to the inbred AB strain.The fragments were ligated into the pCYPAC 6 vector usingEscherichiacoli DH10B as the host, and spotted onto high-density nylon filters.Probe p03/2, corresponding to the first 700 bp of aD. rerio (Dare) BfcDNA clone [14], was labelled with (a32P)-dCTP (Amersham, Braunsch-weig, Germany) using the random priming method with the aid of theReady To Go DNA Labelling Kit (Pharmacia Biotech, Freiburg, Ger-many). The same buffer was used for both prehybridization and hybrid-ization at 658C, 7% SDS, 0.5M Na2PO4 (pH 7.2) and 1 mM EDTA. Thefilters were washed with 40 mM Na2PO4 (pH 7.2) and 0.1% SDS (rinsedtwice at room temperature and then twice for 30 min at 658C), and usedto expose an X-ray film with intensifying screen (XAR5, Kodak, Sttut-gart, Germany). Single positive clones were received from the ResourceCentre and their specificity was confirmed by polymerase chain reaction(PCR) amplification with primers specific for the sequence of interest.

Zebrafish cDNA library. The library was prepared from the spleen andhepatopancreas of a pool of 20 fishes from the laboratory noninbredKOC strain. The cDNA fragments were cloned in the lambda (l) gt10vector, as described previously [28].

PCR, cloning, and sequencing. Genomic and cDNA templates wereamplified as described previously [29] in the GeneAmp PCR system9600 (Perkin–Elmer, U¨ berlingen, Germany). Selected products wereisolated from 2% low-melting-point agarose gels (Gibco/BRL, Eggen-stein, Germany) and subcloned in pUC18 plasmid vector with the aid ofthe SureClone Ligation Kit (Pharmacia Biotech). Double-stranded DNA,prepared with the aid of a Qiagen Plasmid Kit (Qiagen, Hilden, Ger-many) was sequenced using an AutoRead Sequencing Kit (PharmaciaBiotech) in the Automated Laser Fluorescent (A.L.F.) DNA sequencer(Pharmacia Biotech). The primers Tu1279 (sense) and Tu1280 (anti-sense), corresponding to exons 3 and 4 of the zebrafishBf genes havebeen described previously [27]. To obtain longer sequences of the newBf gene (see below) from the cDNA library, two specific primers for thisgene were designed: BfN1 (50-CAGGACAGTGATCAGAGTCACGTT-30, antisense, codons 128–136) and BfN3 (50-GGTAACCACTTATTC-GTGTAATC-30, sense, codons 94–102), and used in combination withanchor primers in both arms of thel gt10 vector: gt10-F (50-TTGAG-CAAGTTCAGCCTGGTTAAG-30, sense) and gt10-R (50-CTTATGAG-TATTTCTTCCAGGGTA-30, antisense). The search forC3-andC4-like

sequences in zebrafish was carried out using several combinations ofdegenerated PCR primers, corresponding to the conserved thioesterregion in these genes: the primers C3F (50-GGSTGTGGGGARCAGA-AYATGATC-30) and C3IIF (50-CMSMKGGCTGTGGGGARCARAM-CA-30) in the forward orientation (around the amino acid motifGCGEQTM), as well as C3R (50-RTAKGCTGTSAGCCADGTGCT-30) and C3IIIR (50-ACYTTMACCACRWAKGCTGTSAGC-30), in thereverse orientation (around the motif TWLTAFV). The primer degene-racy is indicated in the IUPAC code. The sequence data reported in thispaper are available from the DDBJ/EMBL/GenBank, accession numbersAF047412 to AF047415, and AF047837.

Sequence analysis. Database searches were carried out using theBLAST algorithm [30] and dendograms of the amino acid sequenceswere obtained using Poisson correction for multiple substitutions and theneighbour-joining method [31] in the version specified by the MEGApackage [32].

RESULTS

Identification of genomic clones bearing the Bf gene

As part of an ongoing project aimed at the characterization of theMHC in bony fishes we obtained genomic PAC clones contain-ing the zebrafishBf gene which had been identified previously inour laboratory [14]. Several positive signals were obtained in thescreening of the genomic PAC library with theBf-specific p03/2probe; of these, we randomly chose 10 clones and tested them forthe presence of theBf gene by PCR amplification with theprimers Tu1279 and Tu1280, which amplify parts of exon 3and 4 and the intervening intron [27]. Two of these clones (PAC7 and 10) yielded two clear PCR bands of 573 and 387 bp, whileanother two clones (PAC 1 and 3) yielded the 573 bp band, andone other clone (PAC 9) yielded the 387 bp band only. Wesequenced both bands derived from the PAC 7 clone. Bothsequences wereBf-like and contained an intron in the expectedposition. The intron sequences obtained from the upper andlower bands were 421 and 235 bp long, respectively. The codingsequence (positions 127–231 in Fig. 1) of the upper band matchedthat of theBf gene exactly (hereafter referred asBf-1) describedpreviously by Seegeret al. [14]. The sequence obtained from thelower band (Bf-2) differed at five nucleotide sites fromBf-1,three nonsynonymous (sites 158, 187 and 221 in Fig. 1) and twosynonymous sites. The introns of the two genes were could not bealigned except for short exon-flanking stretches (not shown).Similar restriction maps of the PAC clones (average size 100 kb)indicated that the clones all came from the same region of thechromosome (not shown). Since both sequences were obtainedfrom the same PAC clone, the haploid zebrafish genome mustcontain two copies of theBf gene.

To determine whether the second copy isBf- or C2-like, weobtained an additional sequence of this new gene by PCR ampli-fication from a zebrafish cDNA library. On the basis of thesequence obtained from the genomic PAC 7 clone, we designed aspecific primer forBf-2 (BfN1) and used it in combination withthe anchor primer gt10R. We obtained a product of, 250 bp.Based on the sequence of this product, another primer BfN3,

652 R. Gongora et al.

q 1998 Blackwell Science Ltd,Scandinavian Journal of Immunology, 48, 651–658

upstream of BfN1, was designed and used in conjunction withanother anchor primer, gt10F, to produce a fragment of, 1 kb.The specificity of both amplifications was established unambigu-ously on the basis of specific substitutions in comparison with theBf-1 sequence. We sequenced 1085 nucleotides ofBf-2, encom-passing one short consensus repeat (SCR) and the von Wille-brand domains (Fig. 1). We were unable to amplify the entirelength because of the short length of the inserts in the library. Thealignment of the amino acid sequence deduced from the cDNAwith other representativeBf sequences, as well as human andmouseC2 sequences, is shown in Fig. 2. The twoBf formsshowed 88% identity at the nucleotide level and 83% identity atthe amino acid level, including one two-amino-acid deletion inBf-2. No structural defect precluding functionality of the secondgene in the sequenced part was found.

The Bf gene is also duplicated inXenopus[33]. The aminoacid alignment (Fig. 2) and the phylogenetic analysis (Fig. 3) ofthe Xenopusand zebrafish sequences indicate that in each ofthe two species the second copy was generated by independentduplications. The possibility, however, of a common ancestralduplication followed by intraspecies homogenization of theduplicatedBf genes cannot be ruled out completely.

Since theC2 andBf genes are closely linked in mammals (in

humans they are less than 1 kb apart) [25], we searched ourgenomic clones (and later also the cDNA library) for aC2-likesequence using PCR amplification with highly degeneratedprimers (designed to amplify bothC2 andBf). All we obtained,however, were the aforementionedBf sequences (not shown).

Search for C4-like sequences in zebrafish

Similarly, since theBf andC4 genes are closely linked in mam-mals, we tried to detect the presence of aC4-like sequence inclones bearing theBf region by PCR amplification with highlydegenerate primers spanning the conserved thioester region oftheC3 andC4 genes. We used several combinations of primers,but could not amplify anyC4- or C3-related sequences from thegenomic PAC clones. Using the primers C3IIF and C3IIIR,however, we amplified an actinin-like sequence from the clonesPAC 7 and 9. Hence, in the zebrafish this gene is linked to theBflocus (Fig. 4). When compared with a mammalian sequence(thus far only the cDNA sequence is available), a correspondencewas found at both ends of our clone; the middle part, starting withGT and ending with AG, had no counterpart in the mammaliansequence and presumably represents an intron.

We then extended the search to the cDNA library using the

Bf and C3 Genes in Zebrafish653


Fig. 1. Alignment of the zebrafish,Dare, Bf-2 nucleotide (coding) sequence with the previously describedBf-1 sequence; dashes indicate identitywith the top sequence and indels are indicated by asterisks. The ends of the short consensus repeats (SCR) region and the von Willebrand factordomain are indicated by arrows, and the position of the intron 3 by a vertical bar.



Fig. 2. Alignment of human (Hosa,accession number X72875, 21), mouse (Mumu,M57890, 18),Xenopus AandB (Xela, D29796, 19 andD49373, 32), medaka fish (Orla, D84063, 15), zebrafish (Dare Bf-1,U34662, 14,andBf-2, this paper), and lamprey (Laja, D13568, 13), Bf, aswell as mouse (M57891, 18) and human C2 (X04481, 41) amino acid sequences. Only the length corresponding to the (incomplete) Bf-2 sequenceis shown; identity with the consensus sequence is indicated by dashes; indels are indicated by asterisks.

same primers. We could amplify three differentC3-like sequ-ences, and the sequence of thea-2-macroglobulin, a proteaseinhibitor phylogenetically related to the C3/C4/C5 family andcontaining the thioester bond [34]. TheC3-like sequences weredesignatedC3-1 (obtained with the primers C3F and CIIIR),C3-2 (obtained with primers C3F and C3R) andC3-3 (obtainedwith primers (C3F/C3IIIR). The nucleotide and amino acid sequ-ences of these three clones are shown in Fig. 5. The sequenceswere identified unambiguously asC3-like by BLAST analysis (atthe DNA and protein levels), and by phylogenetic analysis (Fig.6). An exhaustive search for aC4-like sequence was, however,unsuccessful.

The three zebrafishC3 sequences were more similar to eachother than any of them was to the troutC3 sequence [16]. At thesame time, however, the zebrafish sequences differed from oneanother by many substitutions.C3-2 andC3-3 formed the mostclosely related pair although even they differed at 20 of the 173sites at the DNA level and at 10 of the 58 positions at the aminoacid level. Since the three forms could be shown (by PCRamplification and sequencing) to be present in individual haploid

embryos, they must represent genes at three different loci ratherthan alleles. The sequence alignment in Fig. 5 and the phylo-genetic tree in Fig. 6 show that the duplication events that pro-duced the three zebrafishC3 loci took place after the separationof the lineages leading to the trout (Salmoniformes) and zebrafish(Cypriniformes). The possibility of a single earlier duplicationevent followed by the homogenization in the zebrafish of theduplicatedC3 genes remains.

DISCUSSION

The vertebrate complement system seems to have evolved bygene duplication from a simpler system. The pairsC1r/C1 s, C2/Bf, C6/C7 and the groupC3/C4/C5are good examples of thisorigin [4, 34]. The duplications took place throughout the periodof vertebrate evolution, since agnathans are the first group inwhich a differentiated system can be detected [8]. In the presentpaper we provide molecular evidence that the duplications mayhave been even more extensive in bony fishes than in othervertebrate classes. While all the tetrapods studied thus far seemto possess singleBf (with the exception ofXenopus, [33]) andC3loci, in the zebrafish there are at least twoBf loci (Bf-1 andBf-2)and threeC3 loci (C3-1, C3-2, and C3-3). The presence ofseveral structurally and functionally distinct C3 forms has alsobeen reported for other bony fishes. Lambriset al. [16] havefound three active forms of C3 (C3-1, C3-3 and C3-4) in therainbow trout (Salmo gairdneri), an old, quasi-tetraploid teleostspecies. A complete nucleotide sequence is available for theC3-1 form only, but the existence of at least two additionalC3genes has been proposed on the basis of partial N-terminal aminoacid sequences and different biochemical properties [17]. TheC3-2 form has no haemolytic activity and its tryptic peptide mapdiffers in several peptides (20% of the total) from that of the C3-1form [35]. Sunyeret al. [36] described the presence of five formsof C3 (C3-1 to C3-5) in the gilthead sea bream (Sparus aurata), amodern teleost fish. Tryptic peptide maps, N-terminal amino acidsequences, and certain biochemical particularities suggest theexistence of at least threeC3genes in this species [36]. The threeaforementioned species – zebrafish (Cypriniformes), rainbowtrout (Salmoniformes) and gilthead sea bream (Perciformes),



Fig. 3. Phylogenetic tree of Bf and C2 amino acid sequences fromFig. 2. The tree was produced by the neighbor-joining method withthe Poisson correction. Numbers on branches are bootstrap valuesobtained from 500 replications.

Fig. 4. Alignment of the zebrafish (Dare) actinin-like nucleotide sequence with the human cDNA sequence (Hosa, M86407), taken from ref. [43].Dashes indicate identity with the top sequence, dots unavailability of sequence information. The intron in theDare sequence, for which nocorresponding sequence is known in humans or any other vertebrate, is indicated by arrows.

belong to evolutionarily widely separated orders. Sunyeret al.(7th Congress of the ISDCI, abstract) also described twoforms of Bf in the rainbow trout, one (Bf-2) behavesfunctionally as a Bf-like protein, but its partial internal andN-terminal amino acid sequence is very similar, although notidentical, to the sequence of the other form (Bf-1) translatedfrom a full-length cDNA clone. Thus, in the rainbow trout, asin the zebrafish, there might exist twoBf-like loci, although it

is uncertain whether they are the result of a tandem duplication orof tetraploidization.

The presence of several recently duplicatedC3andBf genes inthe same species (e.g. zebrafish), might mean that in the fishesthe complement system may play a more important part in theimmune system than in the tetrapods. It may, perhaps, compen-sate for the relative unsophistication of the adaptive immunesystem as indicated, for example, by the restriction of immuno-globulin classes to IgM, and possibly IgD [37], and the generallydelayed antibody response [38].

An interesting possibility is that different C3 convertases maybe involved in specialized responses, each convertase arisingfrom the different combinations of the various Bf and C3 com-ponents in the same animal. The different duplicate copies ofthese components therefore may have been selected for andmaintained in response to different requirements. The descriptionof different functional properties and binding to complementactivators in trout and sea bream C3 forms is consistent with thisspeculation [36, 39]. Specialization to slightly different functionshas also been described for the two forms of humanC4. The C4Aand C4B components are nearly (99%) identical, and yet arefunctionally distinct as a consequence of different binding affin-ities to amino or hydroxyl groups [40].

The possibility that C2 and C4 components may not exist inbony fishes is intriguing. The C2 component has thus far been



Fig. 5. Nucleotide (A) and amino acid (B)sequence alignments of zebrafishC3 forms(Dare-1, – 2 and – 3), compared with otherrepresentativeC3 andC4 sequences; dotsindicate unavailability of sequenceinformation, and dashes, identity with theconsensus sequence. Source of sequences:Hosa(K02765, 22),Gaga(fowl, U16848,24), Saga(trout, L24433, 16),Dare (thispaper), andLaja C3 (D10087, 12),Hosa(K02403, 42) andXenopus(D78003, 11)C4; the human sequence is theC4A isoform.

Fig. 6 Phylogenetic tree of theC3 andC4 nucleotide sequences shownin Fig. 5(A). The tree was produced by the neighbor-joining andKimura’s two parameter methods. Numbers correspond to bootstrapvalues obtained from 500 replications.

identified only in mammals [18, 41], the C4 component in mam-mals and in amphibians [11, 42]. We and others have carried outextensive searches for these two components in bony fishes [10,14, 16, 35, 36, 39] using primers designed to detect bothC2 andBf or C3andC4genes, but without success. This, combined withthe fact that Bf and C3 components have been identified inagnathans [12, 13], that the depletion of a Bf-like component inthe domestic fowl abolishes both the classical and the alternativecomplement pathway activities [20], and that, in the sea bream,C3 may participate in both pathways [39], suggests that the C2and C4 components may be lacking in bony fishes. We postulatetherefore that the differentiation of C2 and C4 components tookplace in tetrapods after their divergence from fishes in responseto a more restricted functional specialization, and that in fishesthe C2 and C4 functions have been assumed by the various Bf-and C3-like forms.

ACKNOWLEDGMENTS

We thank Ms Sandra Mu¨ller for technical help, Ms Niamh Nı´Bhleithın for editorial assistance, and Dr W. E. Mayer for kindlydonating the probe. We thank also the Resource Centre of theGerman Human Genome Project (Berlin) for the PAC clones. R.Gongora was supported by a TMR program grant of the Euro-pean Commission.

REFERENCES

1 Volanakis JE, Frank MM, eds. The Human Complement System inHealth and Disease. New York: Marcel Dekker Inc., 1998.

2 Esser AF. Big MAC attack: complement proteins cause leakypatches. Immunol Today 1991;12:316–18.

3 Lambris JD. The chemistry, biology, and phylogeny of C3. In: CruseJM, Lewis REJ, eds. Complement Today. Complement Profiles.Basel: Karger, 1993.

4 Campbell RD, Law ASK, Reid KBM, Sim RB. Structure, organiz-ation, and regulation of the complement genes. Annu Rev Immunol1988;6:161–95.

5 Liszewski MK, Farries TC, Lublin DM, Rooney IA, Atkinson JP.Control of the complement system. Adv Immunol 1995;61:201–83.

6 Dodds AW, Day AJ. The phylogeny and evolution of the comple-ment system. In: Whaley, K, Loos, M, Weiler, JM, eds. Complementin Health and Disease, 2nd, edn. Amsterdam: Kluwer AcademicPublishers, 1993:39–88.

7 Farries TC, Atkinson JP. Evolution of the complement system.Immunol Today 1991;12:295–300.

8 Nonaka M, Fuji S, Kaidoh Set al. Purification of a lamprey com-plement protein homologous to the third component of the mammal-ian complement system. J Immunol 1984;133:3242–9.

9 Jensen JA, Festa E, Smith DS, Cayer M. The complement system ofthe nurse shark: hemolytic and comparative characteristics. Science1981;214:566–9.

10 Nonaka M, Yamaguchi N, Natsuume-Sakai S, Takahashi M. Thecomplement system of rainbow trout (Salmo gairdneri). J Immunol1981;126:1489–94.

11 Mo R, Kato Y, Nonaka M, Nakayama K, Takashashi M. Fourthcomponent ofXenopus laeviscomplement: cDNA cloning and link-age analysis of the frog MHC. Immunogenetics 1996;43:360–9.

12 Nonaka M, Takahashi M. Complete complementary DNA sequenceof the third component of complement of lamprey. J Immunol1992;148:3290–5.

13 Nonaka M, Takahashi M, Sasaki M. Molecular cloning of a lampreyhomologue of the mammalian MHC class III gene, complementfactor B. J Immunol 1994;152:2263–9.

14 Seeger A, Mayer WE, Klein J. A complement factor B-like cDNAclone from the zebrafish (Brachydanio rerio). Mol Immunol 1996;33:511–20.

15 Kuroada N, Wada H, Naruse Ket al. Molecular cloning and linkageanalysis of the Japanese medaka fish complementBf/C2 gene.Immunogenetics 1996;44:459–67.

16 Lambris JD, Lao Z, Pang J, Alsenz J. Third component of troutcomplement. J Immmunol 1993;151:6123–34.

17 Sunyer JO, Zarkadis IK, Sahu A, Lambris JD. Multiple forms ofcomplement C3 in trout that differ in binding to complementactivators. Proc Natl Acad Sci USA 1996;93:8546–51.

18 Ishikawa N, Nonaka M, Wetsel RA, Colten HR. Murine complementC2 and factor B genomic and cDNA cloning reveals differentmechanisms for multiple transcripts of C2 and B. J Biol Chem1990;265:19040–6.

19 Kato Y, Salter-Cid L, Flajnik MFet al. Isolation of theXenopuscomplement factor-B complementary-DNA and linkage of the geneto the frog MHC. J Immunol 1994;153:4546–54.

20 Kjalke M, Welinder KG, Koch C. Structural analysis of chickenfactor B-like protease and comparison with mammalian complementprotein factor B and C2. J Immunol 1994;151:4147–52.

21 Mole JE, Anderson JK, Davison EA, Woods DEJ. Complete primarystructure for the zymogen of human complement factor B. J BiolChem 1984;259:3407–12.

22 De Bruijn MHL, Fey GH. Human complement component C3:cDNA coding sequence and derived primary structure. Proc NatlAcad Sci USA 1985;82:708–12.

23 Lambris JD, Pappas J, Mavroidis Met al. The third component ofXenopuscomplement: cDNA cloning, structural and functionalanalysis, and evidence for an alternate C3 transcript. Eur J Immunol1995;25:572–8.

24 Mavroidis M, Sunyer OJ, Lambris JD. Isolation, primary structure,and evolution of the third component of chicken complement andevidence for a new member of thea2-macroglobulin family. JImmunol 1995;154:2164–74.

25 Campbell RD, Trowsdale J. Map of the human MHC. ImmunolToday 1993;14:349–52.

26 Figueroa F. The primate class III major histocompatibility complexregion encoding complement components and other genes. In:Blancher A, Socha WW, Klein J, eds. Molecular Biology andEvolution of Blood Group and MHC Antigens in Primates. Berlin:Springer-Verlag, 1997:433–48.

27 Bingulac-Popovic J, Figueroa F, Talbot WSet al. Mapping ofMhcclass I and class II regions to different linkage groups in the zebra-fish, Danio rerio. Immunogenetics 1997;46:129–34.

28 Ono H, Klein D, Vincek Vet al. Major histocompatibility complexclass II genes of zebrafish. Proc Natl Acad Sci USA 1992;89:11886–90.

29 Gongora R, Figueroa F, Klein J. The.HLA-DRB9gene and the originof HLA-DRhaplotypes. Human Immunol 1996;51:23–31.

30 Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic localalignment search tool. J Mol Biol 1990;215:403–10.

31 Saitou N, Nei M. The neighbor-joining method: A new method forreconstructing phylogenetic trees. Mol Biol Evol 1987;4:406–25.



32 Kumar S, Tamura K, Nei M. MEGA Molecular EvolutionaryGenetic Analysis Version 1.0. University Park, PA: The Pennsylva-nia State University, 1993.

33 Kato Y, Salter-Cid L, Flajnik MF, Amikawa C, Sasaki M, NonakaM. Duplication of the MHC-linkedXenopuscomplement factorBgene. Immunogenetics 1995;42:196–203.

34 Sottrup-Jensen L, Stepanik TM, Kristensen Tet al. Common evolu-tionary origin of alpha-2-macroglobulin and complement compo-nents C3 and C4. Proc Natl Acad Sci USA 1985;82:9–13.

35 Nonaka M, Irie M, Tanabe K, Kaidoh T, Natsuume-Sakai S,Takahashi M. Identification and characterization of a variant of thethird component of complement (C3) in rainbow trout (Salmogairdneri) serum. J Biol Chem 1985;260:809–15.

36 Sunyer JO, Tort L, Lambris JD. Structural C3 diversity in fish. JImmunol 1997;158:2813–21.

37 Wilson M, Bengton E, Miller NW, Clem LW, Du Pasquier L, WarrG. A novel chimeric Ig heavy chain from a teleost fish sharessimilarities to IgD. Proc Natl Acad Sci USA 1997;94:4593–7.

38 Wilson MR, Warr GW. Fish immunoglobulins and the genes thatencode them. Annu Rev Fish Dis 1992;2:201–21.

39 Sunyer JO, Tort L, Lambris JD. Diversity of the third form ofcomplement, C3, in fish: functional characterization of five formsof C3 in the diploid fishSparus aurata. Biochem J 1997;326:877–81.

40 Carroll MC, Fathallah DM, Bergamaschini L, Alicot LM, IsenmanDE. Substitutions of a single amino acid (aspartic acid for hystidine)converts the functional activity of human complement C4A to C4B.Proc Natl Acad Sci USA 1990;87:6868–72.

41 Bentley DR. Primary structure of human complement componentC2. Biochem J 1986;239:339–45.

42 Belt KT, Carroll MC, Porter RR. The structural basis of the multipleforms of human complement component C4. Cell 1984;36:907–14.

43 Beggs AH, Byers TJ, Knoll JHM, Boyce FM, Bruns GAP, KunkelLM. Cloning and characterization of two human skeletal musclealpha-actinin genes located on chromosome 1 and 11. J Biol Chem1992;267:9281.



Independent Duplications of Bf and C3 Complement Genes in the Zebrafish

Documents

Transcript of Independent Duplications of Bf and C3 Complement Genes in the Zebrafish