Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
JOURNAL OF INTERFERON & CYTOKINE RESEARCH 23:601–612 (2003) © Mary Ann Liebert, Inc. Atlantic Salmon Interferon Genes: Cloning, Sequence Analysis, Expression, and Biological Activity BØRRE ROBERTSEN, VERONICA BERGAN, TORUNN RØKENES, RANNVEIG LARSEN, and ARTUR ALBUQUERQUE ABSTRACT In this work, we report cDNA cloning of two type I interferons (IFNs) from the head kidney of Atlantic salmon, called SasaIFN-a1 (829 bp) and SasaIFN-a2 (1290 bp). Both translate into 175 amino acid precursor molecules showing 95% amino acid sequence identity. The precursors have a putative 23 amino acid signal peptide, which suggests that the mature Atlantic salmon IFNs contain 152 amino acids (18.2 kDa). Salmon IFN appears to have five a-helices, similar to mammalian and avian type I IFNs, and showed 45% sequence identity with zebrafish IFN, up to 29% identity with mammalian IFN-a sequences, and 17%–18% sequence identity with mammalian IFN-b and chicken type I IFNs. Human embryonic kidney 293 cells transfected with the SasaIFN-a1 cDNA gene produced high titers of acid-stable antiviral activity, which protected salmonid cells against infectious pancreatic necrosis virus (IPNV) and also induced Mx protein in the cells. Poly(I)-poly(C) induced two IFN transcripts in head kidney of Atlantic salmon. Genomic IFN sequences contained four introns and five exons, which is different from the intronless type I IFN genes of birds and mammals. INTRODUCTION V of farmed Atlantic salmon in Norway and other countries. More knowledge about innate antiviral mechanisms of Atlantic salmon is important for a better understanding of its ability to resist viral infections. It is well established that type I interferons (IFN) mediate a major first line of defense against viral infection of higher vertebrates, but little is known about the IFN system of fish. IFNs are cytokines that induce cells into an antiviral state. Type I IFNs are induced by viruses in several cell types, preferentially by double-stranded (ds) RNA intermediates produced during viral replication.(1) Mammalian type I IFNs constitute a multigene family grouped into IFN-a, IFN-b, IFN-d, IFN-t, and IFN-v.(1) Type I IFNs are produced by most cell types, although IFN-a and IFN-v are primarily produced by leukocytes and IFN-b by fibroblasts.(1–3) IFN-d and IFN-t are produced by trophoblasts of pig and ruminants, respectively. In mammals, IFN-a is a family of 14–20 closely related genes, whereas IFN-b is encoded by one gene except in ungulates, which have 5 IFN-b genes.(1) Chicken possesses two distinct type I IFNs, one with a gene family of at least 10 members and one encoded by a single gene.(4–6) Mammalian type I IFNs constitute a clusIRAL DISEASES CAUSE HIGH LOSSES ter of genes lacking introns, in contrast to type II IFN (IFN-g), which is encoded by a single intron-containing gene.(7) IFN-g is produced by T lymphocytes and natural killer (NK) cells in response to mitogens, antigens, or interleukin-12 (IL-12) and is nonhomologous to type I IFNs.(7) Type I IFNs can also be distinguished from IFN-g by their stability at pH 2.(2) Although mammalian type I IFN vary in size (143–172 amino acids) and number of disulfide bonds (0–2), crystal structure determinations show that they share a common three-dimensional structure composed of 5 a-helices.(8–11) The antiviral action of type I IFNs is exerted through binding to the IFN-a/b receptor, which is composed of subunits IFNAR-1 and IFNAR-2.(12) Binding triggers signal transduction through the Jak-Stat pathway and results in induction of transcription of IFN-stimulated gene products, several of which encode proteins that inhibit viral replication.(3) Examples of IFN-induced antiviral proteins are Mx protein, protein kinase PKR and 2-5 oligoadenylate synthetase. One IFN gene has been cloned recently from zebrafish,(13) but IFN genes from salmonids have not yet been published. Atlantic salmon macrophages stimulated with poly(I)poly(C) produce type I IFN-like activity, which inhibits replication of infectious pancreatic necrosis virus (IPNV) and in- Department of Marine Biotechnology, Norwegian College of Fishery Science, University of Tromsø, N-9037 Tromsø, Norway. 601 602 ROBERTSEN ET AL. duces Mx protein in salmonid cell lines.(14,15) IFN-like activity also has been demonstrated in rainbow trout cells after infection with viruses,(16–18) and the molecular mass of the IFN was estimated to be 18 kDa and 94–100 kDa.(16,19) Sequence information about IFN from salmonids is of interest not only for studies of virus-host interactions in these economically important fish species but also with respect to evolutionary studies of IFN. In the present paper, we describe cDNA cloning, phylogenetic sequence analysis, and expression of two IFN genes from the head kidney of Atlantic salmon. Antiviral activity of salmon IFN is demonstrated against IPNV, which is one of the economically most important salmon pathogens. We also show that Atlantic salmon IFN I genes possess introns. MATERIALS AND METHODS Fish Atlantic salmon (Salmo salar) were obtained from the Aquaculture Research Station (Tromsø, Norway) where they were kept in 600 l tanks. For Northern blotting, fish weighing approximately 500 g kept in salt water at 8–9°C were injected intraperitoneally (i. p.) with 2.5 ml saline containing 2 mg/ml poly(I)-poly(C) (Amersham Pharmacia Biotec, Uppsala, Sweden) or with saline alone. Head kidney was harvested 14 h post-injection. For RACE cloning and RT-PCR of IFN, presmolts weighing 30–40 g kept in fresh water at 10°C were injected i.p. with either 0.2 ml saline containing 2 mg/ml poly(I)-poly(C) or 0.2 ml saline. Head kidneys from three control fish and three poly(I)-poly(C)-treated fish were harvested at 12, 24, and 48 h postinjection. In all experiments, organs were preserved in RNA Later™ (Ambion, Austin, TX) at 220°C before isolation of RNA. Prior to treatment, the fish were anesthetised with 0.005% ethyl p-aminobenzoate (Sigma, St Louis, MO). TABLE 1. Primer No. 1a 2 3b 4b 5 6 7 8 16 18 19 21 24 29 aR PRIMERS USED FOR CLONING Orientation Forward Reverse Forward Forward Reverse Forward Reverse Forward Forward Reverse Forward Reverse Forward Reverse 5 a/g, Y 5 c/t, and I 5 inosine. site is underlined b PstI AND Cell culture and virus TO cells originate from Atlantic salmon head kidney and were obtained from Dr. Heidrun Wergeland (University of Bergen, Norway).(20) TO cells were cultured in Eagle’s minimal essential medium (MEM) supplemented with 100 mg/ml streptomycin, 100 U/ml penicillin, 2 mM L -glutamine, 1% nonessential amino acids, and 5% fetal bovine serum (FBS) in 5% CO2 at 20°C. Chinook salmon embryo cells (CHSE-214) were cultured as TO cells except with 7.5% FBS. Human embryonic kidney 293 (HEK293) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with antibiotics and 10% FBS in 5% CO2 at 37°C. IPNV serotype Sp was propagated in CHSE-214 cells. The cell-free supernatants contained 3.1 3 107 50% tissue culture infective doses of virus per milliliter calculated according to the method of Reed and Muench.(21) CHSE-214 and IPNV were obtained from Dr. Ann Inger Sommer (Norwegian Institute of Fisheries and Aquaculture Ltd, Tromsø, Norway). Primers Table 1 shows a list of the primers used for cloning and RTPCR analysis of Atlantic salmon IFN. RNA isolation Total RNA was isolated using Trizol® Reagent (Invitrogen, Groningen, The Netherlands) following the manufacturer’s instructions. Poly(A)1 mRNA was purified from total RNA using the OligoTex™ mRNA isolation kit (Qiagen, Hilden, Germany). Suppressive subtractive hybridization (SSH) cDNA library TO cells were grown to subconfluency in 150-cm3 culture flasks containing 40 ml medium. The IFN system of the cells was stimulated by transfection of poly(I)-poly(C) as follows. EXPRESSION ANALYSES OF ATLANTIC SALMON IFN Sequence (59 to 39) aatacagtgcgcYgaRgcItggga ttgcagatgacgttttgtctctt actgcaggaagactacggaacaataattcggact actgcagggaaaactaacagcgaaacaaacagcta actgacagggtcccacatgat atagatgatggcgaggttgaggac tcaacttcttgaagtagcgtttcag cagagcgtgtgtcattgctgtga actaacagcgaaacaaacagctatttac aagactggggcatcctgctttg aacctgagaagagactgaaacgct actttataaactggtaagggcgtagc cattgctgtgactggatccgacaccaactac agtagcgtttcagtctcttctcaggtttca ATLANTIC SALMON IFN Fugene (60 ml) (Roche, Indianapolis, IN) alone or complexed with 4 mg poly(I)-poly(C) was diluted in 2 ml serum-free MEM. The transfection mixture was added to the cells, which were incubated for 8, 12, 16, or 22 h. Poly(A)1 mRNA was then purified from the cells as described. SSH of cDNA was carried out using the PCR-Select cDNA Subtraction kit from Clontech (Palo Alto, CA) using 2 mg mRNA for the cDNA synthesis. Driver cDNA was generated by reverse transcription of pooled mRNA from cells treated with Fugene alone, and tester cDNA was generated of pooled mRNA of cells transfected with poly(I)-poly(C) in Fugene. The resulting subtracted cDNA was used as a template in the initial PCR amplification of IFN. Initial PCR cloning The degenerate forward primer No. 1 (Table 1) was designed based on conserved regions in vertebrate IFN sequences, including a putative zebrafish IFN EST sequence (accession No. wz32679). Amplification of cDNA by a standard PCR was performed using the subtracted cDNA library as a template and primer No. 1 and the PCR primer 1 supplied with the cDNA subtraction kit from Clontech. PCR products were subcloned into the pCR4 TOPO vector (Invitrogen), sequenced, and analyzed by BLASTX for similarity with sequences deposited in the GenBank. Several cDNA fragments were obtained, which showed similarity with mammalian IFN-a sequences. RACE cloning of full-length IFN cDNA A cDNA library was generated from pooled samples of mRNA from poly(I)-poly(C)-stimulated fish using the Marathon cDNA amplification kit from Clontech. The reverse primer No. 2 (Table 1) was designed based on the salmon IFN cDNA sequences obtained from the SSH library. The 59-end of putative IFN cDNA was amplified using the RACE cDNA library as a template with primer No. 2 and the universal primer AP1 supplied with the RACE kit. Touchdown PCR was performed following the manufacturer’s recommendations. Products were subcloned into the pGEM T-Easy vector (Promega, Mannheim, Germany), sequenced, and analyzed for similarity to vertebrate IFN sequences. Based on differences in the 59-untranslated region (59-UTR) of these clones, two forward primers, No. 3 and No. 4 (Table 1), were designed to amplify full-length cDNA of the putative Atlantic salmon IFNs by PCR. Both primers were supplied with a PstI cut site for easier subcloning and analysis of the direction of the PCR product. Using the RACE cDNA library as a template, primer No. 3 or primer No. 4 and the universal primer AP1 supplied with the RACE kit were included in separate touchdown PCR reactions. Amplified products from each reaction were subcloned into the pGEM T-Easy vector. Of 20 clones picked, 6 different clones were sequenced completely. Two clones, named SasaIFN-a1 and SasaIFN-a2, were of full length and were investigated further in this work. Northern blot analysis mRNA (3 mg) from head kidney of saline-treated and poly(I)-poly(C)-treated fish was loaded into each well of a 1% glyoxal-based agarose gel (Ambion) and subjected to electrophoresis. Northern blotting was performed by the downward transfer method, using NorthernMax Transfer Buffer (Ambion), 603 onto a positively charged nylon membrane. The mRNA was immobilized by UV cross-linking. Prehybridization and hybridization were performed at 52°C in ULTRAHyb hybridization buffer (Ambion). A DIG-labeled 444-bp IFN probe corresponding to nucleotides 553–996 was amplified from the SasaIFN-a2 vector using primers No. 2 and No. 8 (Table 1) as described in the PCR DIG Probe Synthesis Kit (Roche). Hybridization was carried out overnight with the DIG-labeled IFN probe at 10 ng/ml. The membrane was washed once in 2 3 SSC containing 0.1% SDS for 10 min at ambient temperature and twice in 0.1 3 SSC, 0.1% SDS, at 52°C for 15 min. Hybridizing bands were detected with the DIG Luminescent Detection Kit (Roche). Semiquantitative RT-PCR assay of IFN TO cells were grown to subconfluency in 6–well culture plates containing 3 ml medium. Total RNA was extracted from untreated cells and from cells incubated with 1 mg/ml poly(I)poly(C) for 12, 24, or 48 h. Two wells were harvested for each time point, and the experiment was repeated three times. Total RNA was extracted from the head kidneys of control and poly(I)-poly(C)-treated fish as described. RNA (1 mg) from cells or head kidneys was reverse transcribed into cDNA in a reaction mixture containing 0.5 mg random hexamers, 0.5 mM dNTPs, 20 U RNasin, and 200 U M-MLV reverse transcriptase (all Promega) in 20 ml reaction buffer (50 mM Tris-HCl, 75 mM KCl, 3 mM MgCl2, 10 mM DTT, pH 8.3). The synthesis was performed at 42°C for 60 min. PCR was performed using 10% of the resulting cDNA as a template. A 220-bp IFN fragment was amplified by PCR using primer No. 6 and primer No. 7 (Table 1) at the following conditions: 96°C for 2 min, 30 cycles of 96°C for 20 sec, 58°C for 20 sec, 72°C for 20 sec. To confirm that all the samples contained an equal amount of cDNA, a 344-bp actin fragment was amplified by PCR using 10% of the cDNA as template and forward primer 59-cactcaaccccaaagccaacagg-39 and reverse primer 59-aaagtccagcgccacgtagcacag-39 at the following conditions: 96°C for 2 min, 20 cycles of 96°C for 20 sec, 68°C for 20 sec, 72°C for 30 sec. Identification of introns in IFN sequences from genomic DNA Total DNA was purified from full blood of three individual fish using the QIAamp DNA blood Midi kit (Qiagen). The DNA was used as template in standard PCR reactions used to identify introns in genomic IFN sequences from Atlantic salmon. The primer pairs used were Nos. 16 and 18, Nos. 19 and 21, and Nos. 24 and 29, respectively. Antiviral assay The SasaIFN-a1 gene was amplified by a standard PCR using primers No. 3 and No. 4 (Table 1). The product was subcloned into the eukaryotic TA expression vector pCR3.1 (Invitrogen), sequenced, and selected for insert in forward or reverse orientation with respect to the cytomegalovirus (CMV) promoter. Subconfluent HEK293 cells grown in 10 cm culture plates were transfected with 20 mg plasmid for 4 h using the calcium-phosphate coprecipitation method. The cells were given fresh medium and incubated for 48 h. Cell supernatants 604 ROBERTSEN ET AL. A FIG. 1. Nucleotide and amino acid sequences of the SasaIFN-a1 (A) and SasaIFN-a2 (B) precursors. The putative ORF is shown in capital letters along with the translated sequence. The predicted signal protein is underlined. Cysteines in the mature protein are circeled. The potential N-glycosylation site of the mature protein is boxed. In the 39-UTR of the nucleotide sequences, putative motifs associated with mRNA instability are shown in bold letters, and the polyadenylation site is underlined. were harvested, adjusted to pH 2 with HCl, and incubated for 24 h at 20°C. The supernatants were neutralized with NaOH and assayed for antiviral activity by the cytopathic effect (CPE) reduction assay.(22) The HEK293 cell supernatants were 2-fold serially diluted in cell medium, and 100 ml of each dilution was added to quadruplicate wells of subconfluent TO or CHSE-214 cells grown in 96-well culture plates. After 24 h, the cell supernatants were removed, and the cells were infected with IPNV (moi 5 0.1) in 100 ml MEM. Virus was absorbed for 1.5 h before 100 ml medium with serum was added. After 72 h, when complete destruction of unstimulated infected cells occurred, all cells were washed with phosphate-buffered saline (PBS), fixed, and stained by incubation with 1% (w/v) crystal violet in 20% ethanol for 10 min. The cells were then washed three times with distilled water and air dried. The stain was dissolved by addition of 100 ml 50% ethanol containing 0.05 M sodium citrate and 0.05 M citric acid, and the absorbance was read at 550 nm. The mean dye absorbance of lysed control cells was subtracted from the mean absorbance of noninfected control cells and infected stimulated cells. One unit of IFN activity was calculated as the reciprocal dilution of the HEK293 supernatant causing 50% reduction in absorbance at 550 nm. Immunoblot analysis CHSE-214 cells were seeded in 24-well culture plates and incubated for 24 h with 2-fold dilutions of supernatants from HEK293 cells transfected with the SasaIFN-a1 gene. The cells were washed once in PBS and lysed in 50 ml SDS-gel sample buffer. The extracts were boiled for 5 min, and 15 ml extract per lane was subjected to SDS-10% PAGE. (23) Blotting, blocking, and antibody incubation were performed as described.(14) A polyclonal rabbit antiserum generated against the rainbow trout Mx3 protein diluted 1:2000 was used to detect Mx protein. The antiserum was generated against a fragment of the rainbow trout Mx3 protein as described.(24) The plasmid used to express the rainbow trout Mx3 fragment was a gift from JoAnn C. Leong (Department of Microbiology, Oregon State University, Corvallis, OR). Horseradish peroxidase (HRP)-conjugated goat antirabbit antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) diluted 1:10,000 was used as secondary antibody. Detection was performed using the SuperSignal West Pico Chemiluminiscent Substrate (Pierce, Rockford, IL). After Mx protein detection, membranes were stripped and immunostained for actin using a polyclonal antiactin antibody (Sigma) diluted 1:500 as the primary antibody.(14) GenBank accession numbers The cloned salmon IFN cDNAs have GenBank accession numbers AY216594 (SasaIFN-a1) and AY216595 (SasaIFNa2). Accession numbers for sequences used in the creation of alignment and phylogenetic analysis are X92476 (chicken IFNa1), X92479 (chicken IFN-b), X84764 (duck IFN), P05003 ATLANTIC SALMON IFN 605 B FIG. 1. (horse IFN-a1), AAA30949 (horse IFN-v2), P01563 (human IFN-a2b), M25460 (human IFN-b), P37290 (human IFN-d), X58822 (human IFN-v1), NP_034634 (mouse IFN-a4), X14455 (mouse IFN-b), Z22706 (pig IFN-d), I46398 (sheep IFN-a), P56828 (sheep IFN-t), AAL76919 (woodchuck IFNa7), AAM95448 (zebrafish IFN cDNA), AJ544820 (zebrafish IFN gene). RESULTS Cloning and sequencing of Atlantic salmon IFN genes The cDNA cloning of Atlantic salmon IFN was based on induction of IFN mRNA by poly(I)-poly(C), which is a dsRNA with well-known IFN-inducing properties. SSH was used to generate cDNA from Atlantic salmon TO cells highly enriched in genes induced by poly(I)-poly(C). TO cells are derived from the head kidney of Atlantic salmon.(20) PCR amplification of IFN-like cDNA fragments from TO cells was accomplished us- Continued. ing the subtracted cDNA as a template and a degenerate forward primer designed from the conserved vertebrate type I IFN sequence motif YSACAW at the C-terminal. The reverse primer was one supplied by the SSH kit. These sequences allowed design of a specific gene primer that could be used to amplify the 59-end of Atlantic salmon IFN by RACE cloning. Two 59-end IFN products were obtained using cDNA from head kidney of poly(I)-poly(C)-stimulated fish. Based on these sequences, new forward primers were designed that could amplify the whole open reading frames (ORFs) of Atlantic salmon IFN by 39RACE cloning. Two putative IFN cDNAs were amplified, 829 bp and 1290 bp in length. BLASTX analysis of the sequences showed lowest E-value for a putative zebrafish IFN precursor (2e-25), followed by woodchuck IFN-a7 precursor (2e-07), mouse IFN-a4 precursor (8e-7), and other mammalian type I IFNs. Figure 1 shows the nucleotide sequences and the predicted translation of the two cDNAs. The most likely ORF of both cDNAs contained 528 nucleotides starting at position 37 in the small cDNA clone and position 502 in the large cDNA. Both translated ORFs gave a 175-amino acid peptide with a 606 ROBERTSEN ET AL. FIG. 2. ClustalX alignment of the deduced signal protein (a) and mature protein (b) of SasaIFN-a1 and other vertebrate IFN sequences. Amino acids identical to the SasaIFN-a1 sequence are shown shadowed and in bold letters. Dashes indicate gaps. The helixes A, B, C, D, and E are underlined in the SasaIFN-a1 sequence and in sequences of IFNs for which the crystal structure has been solved. molecular mass of 20.8 kDa, which represents IFN precursor molecules of Atlantic salmon. The two clones showed 98% nucleotide sequence identity and 95% amino acid sequence identity. The hydrophobic 23 amino sequence at the amino-terminal end was identified as the signal peptide sequence using the SignalP program version 1.1 (www.cbs.dtu.dk/services/SignalP/).(25) This suggested that the mature Atlantic salmon IFNs contain 152 amino acids, representing a molecular mass of 18.2 kDa. The putative mature salmon IFNs have a cysteine as the first amino acid, similar to the mammalian IFN-a, and were named SasaIFN-a1 (small cDNA) and SasaIFN-a2 (large cDNA). The cysteine content (5 residues) of the signal sequences is notably higher than in other IFN signal sequences, whereas the mature proteins contain only 2 cysteine residues. The salmon IFNs have an isoelectric point of 9.2. An N-linked glycosylation motif (NYSA) is present at amino acid 124. A polyadenylation signal is present 10–12 nt upstream of the poly(A) tail in both cDNAs. Four mRNA destabilization elements ATTTA occur in the 39-UTR of the cDNAs (Fig. 1). Two of them are present in overlapping copies of the sequence UUAUUUAU reported in the 39-UTR of human IFN-b mRNA and other cytokines.(26,27) Alignment of salmon IFN with other vertebrate type I IFNs An alignment of the SasaIFN-a1 precursor protein with 24 different members of vertebrate type I IFNs was created using the CLUSTALX program.(28) Figure 2 shows the alignment with 10 selected members from different classes of IFN, with the signal sequences in Figure 2a and the mature proteins in Figure 2b. The signal sequence of the salmon IFN showed hardly any homology with IFN signal sequences of higher vertebrates (0–9% identity), pig IFN-d being the exception with 19% identity (Table 2). In contrast, salmon and zebrafish signal sequences showed marked homology (32% identity). Analysis of secondary structure using the PHD program(29,30) at the Protein Predict server (www.embl-heidelberg.de/predictprotein/) suggested that Atlantic salmon IFN possesses 5 alpha helixes (A, B, C, D, and E), which overlap with the helixes of other type I IFNs (Fig. 2b). The most conserved amino acids ATLANTIC SALMON IFN 607 TABLE 2. COMPARISON OF SASAIFN-a1 AMINO ACID SEQUENCE WITH OTHER TYPE I IFN SEQUENCES (% IDENTITY ) IFN type SasaIFN-a2 Zebrafish IFN Mouse IFN-a4 Human IFN-a2b Human IFN-v1 Sheep IFN-t Mouse IFN-b Pig IFN-d Chicken IFN-a1 Chicken IFN-b Signal peptide Mature protein 96 32 9 9 5 0 5 19 4 0 97 45 29 28 26 23 17 18 18 17 appeared to occur in and adjacent to the helixes. The salmon IFN showed a high degree of similarity with the zebrafish IFN, including the same pattern of cysteines in the mature protein. The alignment confirmed that the most conserved region of type I IFNs is in the C-terminal region from amino acid 125 to 136 of salmon IFN. The long loop region between helix A and B of the fish IFNs shows very little homology with AB loops from other IFNs. The CD loop of salmon IFN contains a larger gap than other IFNs, including the zebrafish IFN. The Atlantic salmon IFNs showed 45% identity with the zebrafish IFN with respect to both amino acid and nucleotide sequence. As shown in Table 2, salmon IFN showed highest sequence identity with the IFN-a sequences (#29%) and lowest identity with mouse and human IFN-b (17%) and the chicken IFN sequences (17%–18%). In the region from amino acid 125 to amino acid 136, the salmon IFN showed about 60% sequence identity with mammalian IFN-a. To analyze more rigorously the degree of evolutionary relatedness between type I IFN of fish and higher vertebrates, we performed a phylogenetic analysis of the sequences using the PAUP program(31) based on the CLUSTAL X alignment. The phylogenetic tree generated by the Neighbor-Joining method (Fig. 3), indicates that the fish IFNs form a cluster separate from the other vertebrate type I IFNs and are not more related to the IFN-a cluster than the IFN-b or the chicken IFNs. The tree was bootstrapped 1000 times and showed 100% support for the grouping of the fish sequences. The same conclusion was drawn using the Parsimony method of PAUP. Biological activity of SasaIFN-a1 To study whether the SasaIFN-a1 gene encoded a biologically active IFN, it was subcloned into a eukaryotic expression vector pCR3.1 in forward and reverse orientation from the CMV promoter. The constructs were transfected into human HEK293 cells. Media supernatants were harvested 48 h later, acid treated FIG. 3. Unrooted phylogenetic tree showing the relationships between the Atlantic salmon SasaIFN-a1 amino acid sequence and other known type I IFN sequences. The tree was constructed by the Neighbour-Joining method using the CLUSTALX and PAUP programs. 608 ROBERTSEN ET AL. TABLE 3. IFN ACTIVITY IN SUPERNATANTS OF HEK293 CELLS TRANSFECTED WITH THE SASAIFN-a1 GENE IFN activity (U/ml) Test cells Gene orientation CHSE-214 CHSE-214 TO TO Forward Reverse Forward Reverse a ND, Experiment 1 Experiment 2 5560 NDa 120400 ND 3060 ND 36900 ND no detectable activity. over night, and then assayed for antiviral activity against IPNV in Atlantic salmon TO cells or Chinook salmon CHSE-214 cells using the CPE reduction method. As shown in Table 3, supernatants from HEK293 cells transfected with the SasaIFN-a1 gene in the correct orientation showed relatively high levels of antiviral activity (36,000 and 120,000 U/ml) when tested on TO cells and 10-fold lower activity when tested on CHSE-214 cells. In contrast, HEK293 cells that were transfected with the SasaIFN-a1 gene in the reverse orientation gave no IFN activity in either cell system. This demonstrates that the SasaIFN-a1 cDNA translates into a protein, which is cleaved and secreted as an active IFN by human cells. CHSE-214 cells transfected with SasaIFN-a1 also produced IFN activity, but at much lower titers due to lower transfection rates (data not shown). As shown in Figure 4, supernatants of HEK293 cells transfected with the SasaIFN-a1 gene in the forward orientation of the CMV promoter induced Mx protein in the CHSE-214 cells after 24 h of stimulation. In contrast, supernatants from HEK293 cells transfected with the SasaIFN-a1 gene in the reverse orientation did not induce Mx protein in the cells. Similar results were obtained with TO cells (data not shown). FIG. 5. Northern blot analysis of salmon IFN mRNA in head kidney of Atlantic salmon treated with poly(I)-poly(C). Organs were harvested and extracted for RNA 14 h after injection of poly(I)-poly(C). IFN-specific mRNA was detected by hybridization with a DIG-labeled cDNA probe. Lanes 1 and 2, mRNA from two nontreated fish; lanes 3, 4, and 5, mRNA from three different poly(I)-poly(C)-treated fish. were, however, not able to detect IFN transcripts by Northern blotting when total RNA from stimulated fish or TO cells was electrophoresed. This indicated that IFN mRNA was present at relatively low levels even in the stimulated fish cells. Semiquantitative RT-PCR analyses, therefore, had to be used for time course studies of induction. These measurements showed highest level IFN transcripts in the head kidney of the fish 12 h after injection of poly(I)-poly(C) and diminishing levels 24 and 48 h later (Fig. 6 top). TO cells stimulated with poly(I)- Induction of type IFN salmon genes Northern blot analysis of mRNA showed that two IFN transcripts of about 850 nt and 1200 nt were expressed in head kidney of Atlantic salmon 14 h after injection of poly(I)-poly(C) (Fig. 5). Transcripts were not detected in untreated fish. We FIG. 4. Induction of Mx protein by SasaIFN-a1 in CHSE214 cells. Supernatants from HEK293 cells transfected with SasaIFN-a1 in forward and reverse (control) orientation of the CMV promoter were 2-fold serially diluted in cell medium and incubated with CHSE-214 cells for 24 h. Extracts of the cells were subjected to immunoblot analysis with polyclonal antibodies against Mx protein and actin. FIG. 6. RT-PCR analysis of IFN transcripts in head kidney (top) and TO cells (bottom) of Atlantic salmon. TO cells were incubated with poly(I)-poly(C) for 0, 12, 24, and 48 h and extracted for RNA. Atlantic salmon presmolts were injected with poly(I)-poly(C), and head kidneys were harvested and extracted for RNA 0, 12, 24, and 48 h later. RNA samples were reverse transcribed, and the cDNA was subjected to PCR with IFN-specific primers or actin primers as described. The intensity of the 220-bp IFN product and the 344-bp actin product are illustrated. ATLANTIC SALMON IFN poly(C) showed detectable IFN transcripts after 12 h and highest levels of IFN transcripts after 24 h of stimulation (Fig. 6 bottom). IFN transcripts were present at low or undetectable levels in control samples of cells or head kidney. Atlantic salmon IFN genes contain introns PCR of genomic DNA with primers designed from the salmon IFN cDNA sequences revealed that the salmon IFN genes contained introns. Primers No. 16 and No. 18 (Table 1) amplified a 493-bp genomic region containing a 294-bp intron flanked by salmon IFN cDNA sequences (GenBank accession number AY327544). The first sequence contained the start codon of salmon IFN and was assumed to be part of exon I. As shown in Figure 7, primers 24 and 29 amplified a 1048-bp product that contained 3 introns, 271, 126, and 311 bp in length, flanked by exon 1 and exon 4. This sequence must originate from another gene than that amplified by primers No. 16 and No. 18 because of the difference in intron 1 size. The combined exon sequence of the 1048-bp product was 16 and 19 nt different from the corresponding cDNA sequences of SasaIFN-a1 and SasaIFN-a2. This indicates the presence of a third salmon IFN gene or pseudogene. Primers No. 19 and No. 21 amplified a 525-bp genomic region containing a 329-bp intron flanked by two salmon IFN sequences assumed to be exon 4 and exon 5 (GenBank accession number AY327545). PCR of genomic DNA using various primers from the 39-UTR indicated that no more introns were present in the terminal part of the IFN genes. These data suggest that the salmon IFN genes contain 5 exons and 4 introns (Fig. 8). All introns possess the classic 59GT and 39AT intron splice motifs (Fig. 7). The first exon contained the 59-UTR, the 23 amino acid signal sequence, and the 22 first amino acids of the mature protein. Exons 2, 3, and 4 encoded 25, 59, and 26 amino acids, whereas exon 5 encoded 29 amino acids, the translation stop codon, and the 39-UTR. 609 DISCUSSION We describe cloning of two IFN genes from Atlantic salmon that both encode precursor proteins containing 175 amino acids, which are putatively cleaved into 152 amino acid mature proteins called SasaIFN-a1 and SasaIFN-a2. Several lines of evidence suggest that the cloned cDNAs encode type I IFNs. First, the translated amino acid sequences showed significant homology with mammalian IFN-a but no homology with IFN-g or non-IFN proteins. Second, human HEK293 cells transfected with the SasaIFN-a1 cDNA produced and secreted high titers of acid-stable antiviral activity, which protected salmonid cells against IPNV. In addition, the recombinant SasaIFN-a1 induced Mx protein in the fish cells. Mx protein is a major antiviral protein induced by type I IFN in mammals(3,32) and has recently been cloned from Atlantic salmon.(33) Third, transcripts of the IFN genes were not expressed constitutively but were induced in both live fish and TO cells by the dsRNA poly(I)poly(C), which is a well-known inducer of type I IFN. Moreover, the presence of several putative mRNA destabilization elements (ATTTA) in the 39-UTR of the IFN cDNAs is a typical feature of cytokines, including IFN. (26,27) Surprisingly, the Atlantic salmon IFN genes were found to contain 5 exons and 4 introns (Fig. 8). The gene structure has to be confirmed by cloning of full-length genomic salmon IFN sequences but is similar to the gene structure of a zebrafish IFN gene deposited in the gene bank (AJ544820), which also contains 5 exons. The sizes of IFN exons are similar in the two species, but the introns are generally smaller in salmon than in zebrafish (Fig. 8). The presence of introns in the fish IFN genes is in contrast to the typical intron lacking feature of type I IFN of mammals and birds. Loss of introns has, however, not been uncommon during evolution of vertebrates.(34) For example, introns in the coding region of the recombination activating gene rag 1 have been identified in teleosts but not in higher verte- FIG. 7. Partial genomic IFN sequence from Atlantic salmon. The nucleotide sequence was amplified from genomic DNA by PCR using primers No. 24 and No. 29 (Table 1) and corresponds to position 98–437 in the cDNA sequence of SasaIFN-a1. Exon sequences are shown in capital letters and introns in lowercase italic letters. 610 FIG. 8. Schematic diagram depicting the gene structure of putative Atlantic salmon IFN gene compared with zebrafish IFN gene (AJ544820). Values represent nucleotide numbers. Unshaded boxes represents exons, and shaded boxes represent UTRs. Lines represent introns and are drawn one fifth. brates.(35) It is thus possible that type I IFNs have lost introns during the evolution of higher vertebrates. Analysis of the zebrafish and pufferfish genomes may answer whether fish also possess intronless type I IFN genes. Genes of a novel human type I IFN designated IFN-l have recently been discovered on chromosome 19, which also contain 5 exons and 4 introns.(36) However, its relationship to the fish IFNs is uncertain because IFN-l did not appear in either BLASTX or BLASTP analysis of salmon IFN sequences, and CLUSTAL X alignment of SasaIFN-a1 and IFN-l1 showed only 16% sequence identity. The gene structure of fish IFN is clearly different from that of IFN-g from higher vertebrates, which have 3 introns and 4 exons with sizes different from the fish IFN exons.(37) The functional properties of SasaIFN-a1 are in agreement with those of type I IFN-like activity produced by Atlantic salmon macrophages stimulated with poly(I)-poly(C).(14,15) However, with the cloned salmon IFN, more IFN activity with a defined content is available for further studies of virus-fish interactions. Activity in supernatants from HEK293 cells transfected with the salmon IFN gene was at least 200-fold higher than in supernatants from stimulated macrophages. The observation that IFN transcripts were induced more rapidly in live fish than in TO cells in response to poly(I)-poly(C) suggests that some cell populations in the fish, perhaps leukocytes, are more reactive to poly(I)-poly(C) than are TO cells. The low abundance of IFN transcripts in the head kidney of the poly(I)poly(C) fish support that salmon IFN mRNA is short-lived, as suggested by the presence of the ATTTA elements in the 39UTR of the mRNA. The difference in cDNA size and sequence of SasaIFN-a1 and SasaIFN-a2 suggests that they represent distinct type I IFN genes. The expression of the two IFN genes in the head kidney was supported by Northern blot analysis of mRNA from poly(I)-poly(C)-treated Atlantic salmon, which showed two IFN transcripts with sizes similar to the cloned IFN cDNAs. The head kidney is the major hematopoietic organ of Atlantic salmon. The two IFNs may thus be produced by leukocytes. An important question is whether nonhematopoietic cells produce type I IFNs different from SasaIFN-a1 and SasaIFN-a2. In mammals, IFN-a are primarily produced by hematopoietic cells, whereas IFN-b is produced by fibroblasts.(1) The genome of salmonids is believed to be in part tetraploid.(38) This adds complexity to the organization of ROBERTSEN ET AL. salmon genes and may explain why some genes are present in duplicate. At present, we do not know if SasaIFN-a1 and SasaIFN-a2 are clustered on the same chromosome or if they are present on different chromosomal locations. Development of multigene families appears to be a common feature of type I IFNs of higher vertebrates.(39) In contrast, IFN-g of birds and mammals is the product of only one gene. The type I IFN genes of human are clustered, IFN-a have multiple copies, whereas IFN-b has one gene copy.(40) The diversification of IFN-a genes is suggested to have occurred after the separation of the main mammalian orders by duplication followed by gene conversion.(39) The genome of Atlantic salmon most probably encodes additional transcribed genes or pseudogenes, as indicated by sequence in Figure 7. Preliminary results indicate that truncated versions of salmon IFN are expressed (unpublished observations). The number of salmon IFN genes has yet to be determined, however. The putative salmon IFNs have a molecular mass in agreement with the 18 kDa determined for the low molecular mass IFN activity produced by rainbow trout gonad cells.(19) The salmon IFNs were notably shorter than the zebrafish IFN, which has a putative length of 163 amino acids. The salmon and zebrafish IFNs contain the same number and positions of cysteines and show 45% sequence identity. IFN-a from different mammals show similar degree of homology. In contrast to zebrafish IFN, the salmon IFNs possess a glycosylation motif. The degree of glycosylation varies among vertebrate type I IFNs, from none in most IFN-a to 3 in mouse I IFN-b and 4 in the chicken IFNs. Thus, glycosylation of IFN does not appear to be conserved during evolution. Like the zebrafish IFN, the salmon IFNs show no resemblance with a previously published sequence for flatfish IFN,(41) but the latter is apparently more similar to bacteriophage Mu genes than IFN genes(6) and should, therefore, not be classified as an IFN. The sequence similarity suggested that the salmon IFNs are more related to mammalian IFN-a than IFN-b or type I IFNs from birds. Phylogenetic analysis demonstrated, however, that the salmon and zebrafish IFNs form a group separated from mammalian IFN-a and IFN-b and the chicken type I IFNs. This supports phylogenetic analyses, which suggests that IFN-a and IFN-b appeared at later stage in evolution. In fact, the mammalian IFN-a and IFN-b genes are proposed to have originated by duplication of a progenitor after the divergence of birds about 250 million years ago.(39) Although chicken IFN-a and IFN-b apparently are not orthologs of mammalian IFN-a and IFN-b, they appear to be functional homologs.(4,42) IFN-a and IFN-b show only 30%–45% sequence identity in both mammals and chicken. Whether Atlantic salmon or other fish species possess such diverse type I IFN molecules remains to be determined. The signal sequence of the salmon IFNs showed hardly any homology with signal sequences of mammalian IFNs, which demonstrates that the degree of conservation in IFN structure is much larger than the conservation of signal sequences. Clearly, however, even human cells possess the ability to cleave and export the Atlantic salmon IFN precursor. This indicates that the protease(s) cleaving the preprotein and the export mechanisms have broad specificity. Still, there appears to be some conservation of cleavage and export mechanisms for type I IFN within the teleosts because the signal sequences of Atlantic ATLANTIC SALMON IFN salmon and zebrafish IFNs showed a marked degree of similarity. The presence of a 5 a-helix structure typical of type I IFNs suggests that SasaIFN-a1 has a similar conformation to other type I IFNs despite a relatively low percent sequence identity. The IFNs are members of the a-helical cytokine family, which adopts a common up-up-down-down 4-helix-bundle topology.(9) The mature Atlantic salmon IFN is one of the shortest IFNs yet detected. The shortness of the molecule is due to an 8 amino acid gap that occurs between helices C and D. Whether this is due to deletion(s) in the ancestral IFN gene or an insertion(s) in the higher vertebrate IFN genes is uncertain at present. The alignment revealed that some amino acid positions appear to be conserved in all type I IFNs and may be important for the stabilization and activity of the IFNs. These are Leu-19, Trp-75, Leu-94, Leu-120, Trp-130, and Arg-134 in the salmon IFN. Several other amino acids in the salmon IFN sequence are present in most, but not all, of the other IFN sequences and may, therefore, also have their origin in an ancestral IFN. In particular, both salmon and zebrafish IFN possess Cys-1 and Cys-97, which are present and thought to form a disulfide bond in most type I IFNs except mammalian IFN-b.(9–11) On the other hand, both salmon and zebrafish IFN lack the two other disulfide-forming cysteines (positions 29 and 138 of human IFNa2b), which are present in most other type I IFNs except mammalian IFN-b. This indicates that the ancestral vertebrate IFN possessed only one pair of cysteines and that the other pair appeared at a later stage but before divergence of the birds. The type I IFNs in vertebrates serve several important functions besides antiviral activity. Examples of accessory functions are activation of NK cell cytotoxic activity, antiproliferatory effects on cells, in vivo induction of memory CD8 T cell proliferation, upregulation of class I MHC expression, and inhibition of IL-12 expression.(43) Moreover, different IFNs exert regulatory effects on each other’s expression.(44) Accessory functions of type I IFNs in fish are likely to be present but have not yet been demonstrated. ACKNOWLEDGMENTS We thank Aud Øvervatn and Geir Bjørkøy for their assistance with transfection of HEK293 cells. This work was supported by the Research Council of Norway (grants 134168/120 and 151938/150). REFERENCES 1. SEN, G.C. (2001). Viruses and interferons. Annu. Rev. Microbiol. 55, 255–281. 2. PFEFFER, L.M., DINARELLO, C.A., HERBERMAN, R.B., WILLIAMS, B.R., BORDEN, E.C., BORDENS, R., WALTER, M.R., NAGABHUSHAN, T.L., TROTTA, P.P., and PESTKA, S. (1998). Biological properties of recombinant alpha-interferons: 40th anniversary of the discovery of interferons. Cancer Res. 58, 2489–2499. 3. SAMUEL, C.E. (2001). Antiviral actions of interferons. Clin. Microbiol. Rev. 14, 778–809. 4. LOWENTHAL, J.W., STAEHELI, P., SCHULTZ, U., SEKEL- 611 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. LICK, M.J., and MARCUS, P.I. (2001). Nomenclature of avian interferon proteins. J. Interferon Cytokine Res. 21, 547–549. SICK, C., SCHULTZ, U., and STAEHELI, P. (1996). A family of genes coding for two serologically distinct chicken interferons. J. Biol. Chem. 271, 7635–7639. SEKELLICK, M.J., FERRANDINO, A.F., HOPKINS, D.A., and MARCUS, P.I. (1994). Chicken interferon gene: cloning, expression, and analysis. J. Interferon Res. 14, 71–79. VILCEK, J., and SEN, G.C. (1996). Interferons and other cytokines. In: Fields Virology, 3rd ed. B.N. Fields, D.M. Knipe, and P.M. Howley (eds.) Philadelphia: Lippincott-Raven Publishers, pp. 375–399. KARPUSAS, M., NOLTE, M., BENTON, C.B., MEIER, W., LIPSCOMB, W.N., and GOELZ, S. (1997). The crystal structure of human interferon beta at 2.2-A resolution. Proc. Natl. Acad. Sci. USA 94, 11813–11818. RADHAKRISHNAN, R., WALTER, L.J., HRUZA, A., REICHERT, P., TROTTA, P.P., NAGABHUSHAN, T.L., and WALTER, M.R. (1996). Zinc-mediated dimer of human interferon-alpha 2b revealed by X-ray crystallography. Structure 4,, 1453–1463. RADHAKRISHNAN, R., WALTER, L.J., SUBRAMANIAM, P.S., JOHNSON, H.M., and WALTER, M.R. (1999). Crystal structure of ovine interferon-tau at 2.1 A resolution. J. Mol. Biol. 286, 151–162. SENDA, T., SHIMAZU, T., MATSUDA, S., KAWANO, G., SHIMIZU, H., NAKAMURA, K.T., and MITSUI, Y. (1992). Three-dimensional crystal structure of recombinant murine interferon- beta. EMBO J. 11, 3193–3201. MOGENSEN, K.E., LEWERENZ, M., REBOUL, J., LUTFALLA, G., and UZE, G. (1999). The type I interferon receptor: structure, function, and evolution of a family business. J. Interferon Cytokine Res. 19, 1069–1098. ALTMANN, S.M., MELLON, M.T., DISTEL, D.L., and KIM, C.H. (2003). Molecular and functional analysis of an interferon gene from the zebrafish, Danio rerio. J. Virol. 77, 1992–2002. JENSEN, I., and ROBERTSEN, B. (2002). Effect of doublestranded RNA and interferon on the antiviral activity of Atlantic salmon cells against infectious salmon anemia virus and infectious pancreatic necrosis virus. Fish Shellfish Immunol. 13, 221–241. NYGAARD, R., HUSGÅRD, S., SOMMER, A.-I., LEONG, J.A.C., and ROBERTSEN, B. (2000). Induction of Mx-protein by interferon and double-stranded RNA in salmonid cells. Fish Shellfish Immunol. 10, 435–450. DE SENA, J., and RIO, G.J. (1975). Partial purification and characterization of RTG-2 fish cell interferon. Infect. Immun. 11, 815–822. DORSON, M., BARDE, A., and DE KINKELIN, P. (1975). Egtved virus induced rainbow trout serum interferon: some physicochemical properties. Ann. Microbiol. (Paris) 126, 485–489. ROGEL-GAILLARD, C., CHILMONCZYK, S., and DE KINKELIN, P. (1993). In vitro induction of interferon-like activity from rainbow trout leucocytes stimulated by Egtvedt virus. Fish Shellfish Immunol. 3, 383–394. LIN, S., LIN, C., and HSU, Y. (2001). Establishment of an antiviral assay and the identification and partial purification of interferon-like protein from rainbow trout gonad cells (RTG-2). Zool. Stud. 40, 240–246. WERGELAND, H.I., and AAKRE JOHANSEN, R. (2001). A salmonid cell line (TO) for production of infectious salmon anaemia virus (ISAV). Dis. Aquatic Organisms 44, 183–190. REED, L., and MUENCH, H. (1938). A simple method of estimating fifty percent end points. Am. J. Hyg. 27, 493–497. MEAGER, A. (1987). Quantification of interferons by anti-viral assays and their standardization. In: Lymphokines and Interferons: A Practical Approach. M.J. Clemens, A.G. Morris, and A.J.H. Gearing (eds.) Oxford: IRL Press, pp. 129–147. 612 23. LAEMMLI, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 24. TROBRIDGE, G.D., CHIOU, P.P., and LEONG, J.A.C. (1997). Cloning of the rainbow trout (Oncorhynchus mykiss) Mx2 and Mx3 cDNAs and characterization of trout Mx protein expression in salmon cells. J. Virol. 71, 5304–5311. 25. NIELSEN, H., ENGELBRECHT, J., BRUNAK, S., and VON HEIJNE, G. (1997). Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng. 10, 1–6. 26. PEPPEL, K., VINCI, J.M., and BAGLIONI, C. (1991). The AUrich sequences in the 39 untranslated region mediate the increased turnover of interferon mRNA induced by glucocorticoids. J. Exp. Med. 173, 349–355. 27. CHEN, C.Y., and SHYU, A.B. (1995). AU-rich elements: characterization and importance in mRNA degradation. Trends Biochem. Sci. 20, 465–470. 28. THOMPSON, J.D., GIBSON, T.J., PLEWNIAK, F., JEANMOUGIN, F., and HIGGINS, D.G. (1997). The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882. 29. ROST, B., and SANDER, C. (1994). Combining evolutionary information and neural networks to predict protein secondary structure. Proteins 19, 55–72. 30. ROST, B., and SANDER, C. (1993). Prediction of protein secondary structure at better than 70% accuracy. J. Mol. Biol. 232, 584–599. 31. SWOFFORD, D.L. (2000) PAUP: Phylogenetic Analysis Using Parsimony and Other Methods. (Software). Sunderland, MA: Sinauer Associates. 32. HALLER, O., FRESE, M., and KOCHS, G. (1998). Mx proteins: mediators of innate resistance to RNA viruses. Rev. Sci. Technol. 17, 220–230. 33. ROBERTSEN, B., TROBRIDGE, G., and LEONG, J.C. (1997). Molecular cloning of double-stranded RNA inducible Mx genes from Atlantic salmon (Salmo salar L.). Dev. Comp. Immunol. 21, 397–412. 34. ROY, S.W., FEDOROV, A., and GILBERT, W. (2003). Largescale comparison of intron positions in mammalian genes shows intron loss but no gain. Proc. Natl. Acad. Sci. USA 100, 7158–7162. 35. WILLETT, C.E., CHERRY, J.J., and STEINER, L.A. (1997). Characterization and expression of the recombination activating genes (rag1 and rag2) of zebrafish. Immunogenetics 45, 394–404. 36. KOTENKO, S.V., GALLAGHER, G., BAURIN, V.V., LEWISANTES, A., SHEN, M., SHAH, N.K., LANGER, J.A., SHEIKH, F., DICKENSHEETS, H., and DONNELLY, R.P. (2003). IFN- ROBERTSEN ET AL. 37. 38. 39. 40. 41. 42. 43. 44. lambdas mediate antiviral protection through a distinct class II cytokine receptor complex. Nat. Immunol. 4, 69–77. KAISER, P., WAIN, H.M., and ROTHWELL, L. (1998). Structure of the chicken interferon-gamma gene, and comparison to mammalian homologues. Gene 207, 25–32. ALLENDORF, F.W., and THORGAARD, G.H. (1984). Tetraploidy and the evolution of salmonid fishes. In: Evolutionary Genetics of Fishes. B.J. Turner (ed.) New York: Plenum, pp. 1– 53. ROBERTS, R.M., LIU, L., GUO, Q., LEAMAN, D., and BIXBY, J. (1998). The evolution of the type I interferons. J. Interferon Cytokine Res. 18, 805–816. DIAZ, M.O., POMYKALA, H.M., BOHLANDER, S.K., MALTEPE, E., MALIK, K., BROWNSTEIN, B., and OLOPADE, O.I. (1994). Structure of the human type-I interferon gene cluster determined from a YAC clone contig. Genomics 22, 540–552. TAMAI, T., SHIRAHATA, S., NOGUCHI, T., SATO, N., KIMURA, S., and MURAKAMI, H. (1993). Cloning and expression of flatfish (Paralichthys olivaceus) interferon cDNA. Biochim. Biophys. Acta 1174, 182–186. STAEHELI, P., PUEHLER, F., SCHNEIDER, K., GOBEL, T.W., and KASPERS, B. (2001). Cytokines of birds: conserved functions—a largely different look. J. Interferon Cytokine Res. 21, 993–1010. BIRON, C.A., and SEN, G.C. (2001). Interferons and other cytokines. In: Fields Virology, 4th ed. D.M. Knipe, P.M. Howley, D.E. Griffin, M. Martin, B. Roizman, and S.E. Straus (eds.) Philadelphia: Lippincott-Raven, pp. 321–351. MARIE, I., DURBIN, J.E., and LEVY, D.E. (1998). Differential viral induction of distinct interferon-alpha genes by positive feedback through interferon regulatory factor-7. EMBO J. 17, 6660–6669. Address reprint requests or correspondence to: Dr. Børre Robertsen Norwegian College of Fishery Science University of Tromsø N-9037 Tromsø Norway Tel: 147 77 64 61 30 Fax: 147 77 64 60 20 E-mail: [email protected] Received 14 March 2003/Accepted 9 July 2003 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.