Fish & Shellfish Immunology 23 (2007) 670e682www.elsevier.com/locate/fsi

Induction of antiviral genes, Mx and vig-1, by dsRNAand Chum salmon reovirus in rainbow trout monocyte/

macrophage and fibroblast cell lines

Stephanie J. DeWitte-Orr a, Jo-Ann C. Leong b, Niels C. Bols a,*

a Department of Biology, University of Waterloo, 200 University Avenue West, Waterloo, ON N2L 3G1, Canadab Hawaii Institute of Marine Biology, P.O. Box 1346, Kaneohe, HI 96744, USA

Received 29 August 2006; revised 16 January 2007; accepted 19 January 2007

Available online 1 February 2007

Abstract

The expression of potential antiviral genes, Mx1, Mx2, Mx3 and vig-1, was studied in two rainbow trout cell lines: monocyte/macrophage RTS11 and fibroblast-like RTG-2. Transcripts were monitored by RTePCR; Mx protein by Western blotting. Inunstimulated cultures Mx1 and vig-1 transcripts were seen occasionally in RTS11 but rarely in RTG-2. A low level of Mx proteinwas seen in unstimulated RTS11 but not in RTG-2. In both cell lines, Mx and vig-1 transcripts were induced by a dsRNA, polyinosinic: poly cytidylic acid (poly IC), and by Chum salmon reovirus (CSV). Medium conditioned by cells previously exposedto poly IC or CSV and assumed to contain interferon (IFN) induced the antiviral genes in RTS11. However, RTG-2 respondedonly to medium conditioned by RTG-2 exposed previously to CSV. In both cell lines, poly IC and CSV induced Mx transcriptsin the presence of cycloheximide, suggesting a direct induction mechanism, independent of IFN, was also possible. For CSV,ribavirin blocked induction in RTS11 but not in RTG-2, suggesting viral RNA synthesis was required for induction only inRTS11. In both RTS11 and RTG-2 cultures, Mx protein showed enhanced accumulation by 24 h after exposure to poly IC andCSV, but subsequently Mx protein levels declined back to control levels in RTS11 but not in RTG-2. These results suggest thatMx can be regulated differently in macrophages and fibroblasts.� 2007 Elsevier Ltd. All rights reserved.

Keywords: RTS11; RTG-2; Mx; vig-1; Antiviral; Reovirus; Poly IC; Chum salmon reovirus

1. Introduction

Manipulating the antiviral mechanisms of fish is of interest in aquaculture as a possible means of enhancingprotection against viral diseases but is limited by the relative lack of basic knowledge about such mechanisms andhow they are regulated. One potential target for manipulation is the Mx proteins [1e3]. Mx proteins were originallydiscovered in influenza-resistant laboratory mice [4], and have since been identified in many vertebrate species includ-ing fish [5]. They are dynamin-related members of the large GTPase super-family [5]. Most animals have multiple Mx

* Corresponding author. Tel.: þ1 519 888 4567x3993; fax: þ1 519 746 0614.

E-mail address: ncbols@sciborg.uwaterloo.ca (N.C. Bols).

1050-4648/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fsi.2007.01.017

671S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

isoforms, and rainbow trout express three Mx proteins [6]. Recently an antiviral role for Mx in fish has been described,with Atlantic salmon Mx1 inhibiting infectious pancreatic necrosis virus replication (IPNV) [7] and infectious salmonanemia virus (ISAV) [8]. Virus induced gene-1 (vig-1) was first identified in viral hemorrhagic septicemia virus(VHSV) exposed rainbow trout leukocytes [9]. Subsequently, viperin and cig5, which share sequence similaritywith vig-1, have been identified in humans and mice and possible antiviral activity has been demonstrated [10,11].

A variety of agents have been explored both in vivo and in vitro for their capacity to enhance Mx expression, onecommon stimulant is the synthetic double-stranded RNA (dsRNA), poly inosinic: poly cytidylic acid (poly IC) [1,12].Poly IC has also been shown to induce vig-1 expression [13]; however, concurrent induction of vig-1 and Mx has yet tobe demonstrated. Another potential inducer of antiviral genes that has yet to be examined is Chum salmon reovirus(CSV), which is an aquareovirus (AqVR) [14]. AqRV are non-enveloped with an icosohedral, double capsid and an11-segment dsRNA genome [15]. CSV was isolated from apparently healthy fish and caused little or no mortality inseveral salmon species [15]. This observation led to a study on the ability of CSV to stimulate host defenses in rainbowtrout and protect live fish against infections hematopoietic necrosis virus (IHNV), which is a rhabdovirus and a seriousviral pathogen for rainbow trout farms [16]. Although the mechanism was not investigated, induction of antiviralgene(s) is a possibility.

Central to the regulation of antiviral defenses in mammals, and presumably fish, is a network of interferons,coordinated by interacting signal transduction pathways. Interferons (IFNs) are cytokines induced by viruses thatact in an autocrine and paracrine fashion, through their cognate receptors, to induce expression of a battery of genes(interferon stimulated genes, ISGs) that collectively bestow an antiviral state. Interferons can affect the expression ofover 1000 different genes in mammalian cells [17]. One of the IFN-inducible proteins in mammals is the dsRNA-dependent protein kinase (PKR), which is a serine/threonine kinase [18]. PKR, which is synthesized inactive, isactivated by dsRNA; however, PKR activity is not limited to antiviral mechanisms and has increasingly been foundto be a regulator of diverse cellular responses to stress [19]. In fish only a handful of ISGs have been identified, in-cluding Mx and vig-1 [20e23]. A PKR-like gene has been recently identified in the crucian carp [24] and zebrafish[25], and PKR activity has been demonstrated in rainbow trout [26]. Poly IC has been shown under some conditionsto stimulate another type of antiviral defense, apoptosis, in some fish cell lines but not in others [27].

In this report the capacity of poly IC and CSV to induce Mx1, Mx2, Mx3, and vig-1 has been compared in RTS11and RTG-2, which are rainbow trout monocyte/macrophage and fibroblast cell lines respectively. Expression of an-tiviral gene transcripts has been found to differ subtly between these cells, varying with the inducing treatments.At the protein level, Mx appeared more stable in RTG-2 than in RTS11. Overall the results suggest that regulationof antiviral mechanisms in fish is complex, but unraveling the regulatory systems holds out the promise of beingable to optimize treatments for effectively inducing antiviral states.

2. Materials and methods

2.1. Cell culture

Two cell lines were used in this study, a fibroblast gonadal cell line, RTG-2, and a macrophage-like cell line,RTS11, both derived from rainbow trout. RTG-2 was obtained from the American Type Culture Collection(ATCC), whereas RTS11 [28] was developed in this laboratory. The routine growth of these cells lines has beendescribed in detail previously [27,29]. Briefly, cultures were maintained at 18e21 �C in Leibovitz’s L-15 supple-mented with fetal bovine serum (FBS) and penicillin G/streptomycin sulfate. These items were purchased from Sigma(St. Louis, MO). RTG-2 cultures were grown in 75-cm2 plastic tissue culture flasks while RTS11 cultures were grownin 25-cm2 plastic tissue culture flasks; both flasks were purchased from Nunc (Roskilde, Denmark).

2.2. Virus propagation and infection

Chum salmon reovirus (CSV) was obtained from ATCC and routinely propagated on monolayers of a chum salmonembryonic cell line, CHSE-214. CSV containing medium (CCM) was collected 7 days post infection, passed througha 0.2-mm filter, and kept frozen at�80 �C until used. Virus titration was performed on monolayers of CHSE-214 cellsgrown in a 96-well plate (Falcon/Becton-Dickinson, Franklin Lakes, NJ). Viral suspensions were diluted from 10�1 to10�6 and 6 wells were inoculated with 200 ml of each dilution. Cultures were incubated at 21 �C for 3 days. Following

672 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

this period, the cell monolayers were scored for the appearance of cytopathic effects and the final titer, expressed asTCID50/ml, was estimated using the Karber method [30].

2.3. Cell treatments

RTG-2 were seeded at 2.3 � 106 cells per 25-cm2 tissue culture flasks, while RTS11 were seeded at 1.0 � 107 cellsper 12.5-cm2 tissue culture flask. All cells were plated in L-15 medium without FBS supplementation. Cells wereallowed to attach overnight at normal growing temperatures before being treated.

2.3.1. Interferon-containing conditioned medium (ICCM)RTS11 and RTG-2 were treated with a stimulant, either 50 mg/ml poly IC (Sigma) or chum salmon reovirus (CSV;

104.3 TCID50/ml) for 24 h. After this, the medium was removed and the cells were washed two times with Dulbecco’sphosphate-buffered saline (D-PBS; Sigma), and fresh medium added. As RTS11 cultures are semi-adherent, stimulantexposures were terminated slightly differently. The cells were collected, centrifuged and washed twice with PBS ina 15-ml polypropylene centrifuge tube (Falcon/Becton-Dickinson), and plated in new tissue culture-treated flasks withfresh medium. The cells were then allowed to condition this medium for 24 h before it was collected, filtered (0.2 mmfilter), and immediately added, undiluted, to different cell cultures for measuring gene expression. RTS11 ICCM wasapplied to RTS11 cultures, and RTG-2 ICCM was applied to RTG-2 cultures. In this study poly IC-ICCM denotesinterferon-containing conditioned medium from poly IC stimulated cells, while CSV-ICCM denotes ICCM fromCSV stimulated cells. Control cultures for these experiments received conditioned medium from untreated cellsand are indicated as CM-control.

2.3.2. Stimulant studyRTS11 and RTG-2 cultures were treated with 50 mg/ml poly IC, CSV (104.3 TCID50/ml), undiluted poly IC-ICCM

or CSV-ICCM. Cells were incubated for 24 h at 18 �C. Poly IC-treated cultures, poly IC-ICCM and correspondingcontrol cultures were in serum-free medium, while CSV-treated cultures, CSV-ICCM and corresponding control cul-tures were in medium containing 5% FBS.

2.3.3. Inhibitor studyRTS11 were pretreated with inhibitors for 30 min prior to treatment with a stimulant. Inhibitor treatments included

two PKR inhibitors, 1.5 mM 2-aminopurine (2-AP), and a second commercially available oxindole functionalizedwith an imidazole, referred to as compound 16 [31] (0.5 mg/ml), a translation inhibitor, cycloheximide (0.1 mg/ml)and an antiviral compound, ribavirin (100 mg/ml). All inhibitors were purchased from Sigma with the exception ofcompound 16, which was purchased from Calbiochem (La Jolla, CA). Following the 30 min pretreatment, cellswere treated with poly IC, CSV or ICCM as described in section 2.3.2.

2.4. RTePCR

2.4.1. RNA extractionCells were collected at specific time points, pelleted and washed with D-PBS. RNA was extracted using GenElute

mammalian total RNA miniprep kit (Sigma). RNA was eluted using nuclease free water (Bioshop, Burlington, ON).

2.4.2. cDNA synthesisTwo micrograms of RNA was treated with 1 U RNnase-free, DNase I (Roche, Laval, Qc) for 30 min at 37 �C and

5 min at 75 �C to remove any contaminating genomic DNA from the samples. These samples were then used to obtaincDNA as follows. 1 ml of 0.5 mg/ml oligo-(dT)23 anchored primer (Sigma) was added to each of the RNA samples, thesamples were heated to 70 �C for 10 min and immediately put on ice. Afterward, 4 ml 5� buffer (250 mM TriseHCl(pH 8.3), 375 mM KCl, 15 mM MgCl2; Invitrogen, Burlington, ON), 1 ml 10 mM deoxynucleotide triphosphate(dNTP, Sigma) mix, 2 ml 0.1 M dithiothreitol (DTT, Invitrogen) and 80 U Superscript II reverse transcriptase (Invi-trogen) were added to each sample. The reaction was allowed to sit at room temperature for 10 min prior to 42 �Cfor 50 min and 95 �C for 5 min. The resulting cDNA was diluted 1:20 in nuclease-free water and stored at �80 �C.

673S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

2.4.3. PCR reactionsAll PCR reactions contained: 0.5 ml 10 mM dNTP mix (Sigma), 1.25 U Taq polymerase (Sigma), 1.5 mM MgCl2

(Sigma), 2.5 ml 10� reaction buffer (500 mM KCl, 100 mM TriseHCl, pH 8.3, Sigma), 1.25 ml 10 mM forward andreverse primers, 2.5 ml diluted cDNA and nuclease-free water to a 25-ml total volume. All PCR reactions were set upon ice. The primer sequence, cycle number, and annealing temperature for each primer set are listed in Table 1. ThePCR reactions were run using a MasterCycler personal thermocycler (Eppendorf), all samples were run at least twicewith two independent sets of cDNA to verify results. Cycle conditions were as follows: 95 �C for 2 min, a set numberof cycles with 95 �C 45 s, primer-specific annealing temperature for 45 s, 72 �C for 45 s followed by 72 �C for 10 min.PCR products (10 ml) were visualized on a 1.5% agarose gel with 15 min post-stain in 0.5 mg/ml ethidium bromide(EtBr) and 15 min destain in MilliQ water followed by visualization under UV transillumination.

Amplified primer sequences were cloned into pGEM-Teasy (Promega, Madison, WI) and subsequently sequencedto confirm primer specificity, especially between Mx1, Mx2 and Mx3. All primers were found to amplify the gene ofinterest specifically.

2.5. Western blot analysis

Cells were seeded at treated as described above (Section 2.3) and collected at the completion of the exposures.Cells were washed twice in ice-cold PBS and lysed on ice in RIPA lysis buffer (150 mM NaCl, 1% NP-40, 0.5%sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 50 mM Tris, pH 8.0) with protease inhibitor cocktail(Sigma). The cells were lysed for 30 min and centrifuged at 10,000 � g for 15 min. The protein concentration ofthe cell extract was quantified using the Bradford method. Equal amounts of protein (20 mg) were boiled for 5 minin Laemmli buffer (135 mM Tris, 4% (w/v) SDS, 0.06% (w/v) bromophenol blue, 20% glycerol, 2% (w/v) 2-mer-captoethanol), put immediately on ice and added to each lane of a SDSe15% polyacrylamide gel. Electrophoresiswas performed using a Bio-Rad mini-PROTEAN II electrophoresis cell. SeeBlue pre-stained protein standard(Invitrogen) was included with each gel for protein size estimation. Proteins were electro-transferred to nitrocel-lulose membrane (Bio-Rad) overnight using the Bio-Rad mini-Trans Blot cell (60 mA). Equal loading of proteinsamples was confirmed by staining of total protein with 0.1% Ponceau S (in 5% w/v acetic acid). The stainedblots were scanned and de-stained using dH2O. Membranes were then blocked in 5% (w/v) nonfat dried milkin T-TBS (10 mM TriseHCl, 100 mM NaCl, 0.1% (v/v) Tween 20) for 1 h at room temperature. The primaryantibody was diluted in blocking solution: polyclonal rabbit anti-rainbow trout Mx [6] was diluted 1:2000.The polyclonal antiserum was raised against an Escherichia coli-expressed fragment of Mx3 and reacted withall three rainbow trout Mx proteins [6]. The membranes were incubated with the primary antibody for 1 h atroom temperature. Membranes were then washed three times in T-TBS and incubated with the secondary anti-body; horseradish-peroxidase-conjugated goat anti-rabbit (HþL) (BioRad), and visualized using the enhancedchemiluminescence (ECL) detection kit (Amersham Biosciences, Little Chalfont, UK). Each blot was performedat least twice with independent samples.

Table 1

Summary of primers used in study, including PCR product size and number of cycles used

Gene Primers PCR product

size (bp)

Annealing

temperature (�C)

Cycle # Primer

reference

b-actin (AJ438158) F 50 ATCGTGGGGCGCCCCAGGCACC 30 514 53 30 [49]

R 50 CTCCTTAATGTCACGCACGATTTC 30

Mx1 (U30253) F 50 ATGCCACCCTACAGGAGATGAT 30 452 53 30 [30]

R 50 AAAAAGGATAACAAAGGACT 30

Mx2 (U47945) F 50 CTTGGTAGACAAAGGCACAGAGGA 30 500 65 27 Unpublished

R 50 AAGTTCTTTCCAGAGCGATCCA 30

Mx3 (U47946) F 50 ATGCCACCCTACAGGAGATGAT 30 380 53 30 [34]

R 50 CCACAGTGTACATTTAGTTG 30

Vig-1 (AF076620) F 50 CAGTTCAGTGGCTTTGACGA 30 232 65 27 [9]

R 50 ACAAACTCCTCAAGGTATGG 30

Gene accession numbers are listed in parentheses.

674 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

3. Results

3.1. Expression in unstimulated cultures

RTePCR was used to study the expression of vig-1 and Mx1, Mx2 and Mx3 transcripts in two rainbow trout celllines, the macrophage-like RTS11, and the fibroblast-like RTG-2, whether from stimulated or unstimulated cultures.Mx and vig-1 transcripts was seen occasionally in unstimulated RTS11 cultures, but rarely or not at all in unstimulatedRTG-2 cultures (Fig. 1). Western blotting with antibodies against rainbow trout Mx was used to detect the Mx protein.A broad band at 70 kDa was repeatedly seen in unstimulated RTS11 cultures but was rare in RTG-2 cultures (Fig. 7).

3.2. Mx and vig-1 transcript induction by dsRNA and CSV

Exposures of 24 h to poly IC or Chum salmon reovirus (CSV) induced Mx1, Mx2, Mx3 and vig-1 in both RTS11 andRTG-2 (Fig. 1). For poly IC, gene expression seemed faster in RTS11, where faint bands of Mx2, Mx3 and vig-1 couldbe detected after 4 h exposures while only Mx2 transcript could be detected in RTG-2. Although CSV induced Mx1,Mx2, Mx3 and vig-1 transcript expression in both RTS11 and RTG-2, vig-1 transcript levels were consistently low inRTG-2. No transcripts were induced by a 4 h exposure to CSV. The production of CSV by RTS11 and RTG-2 cultureswas not seen until 12 days post infection, and relative to the cell line used to maintain the virus, CHSE-214, productionwas relatively low [32].

Fig. 1. Induction of Mx and vig-1 transcripts by poly IC and chum salmon reovirus (CSV). RTS11 (monocyte/macrophage-like) and RTG-2

(fibroblast-like) cultures were exposed to 50 mg/ml poly IC or 104.3 TCID50/ml CSV for 4 or 24 h prior to RNA being collected and RTePCR

performed for transcripts of Mx1, Mx2, Mx3, vig-1 and b-actin (b-a). Control cultures for poly IC-treated cells had serum-free conditions, com-

pared to CSV control cultures which had a 5% FBS supplement.

675S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

3.3. Effect of cycloheximide on Mx and vig-1 transcript induction by dsRNA and CSV

Evidence for a direct pathway to Mx and vig-1 induction that does not require de novo protein synthesis was pur-sued using the inducers together with cycloheximide (CHX), a translation inhibitor. RTS11 and RTG-2 culturestreated with poly IC and CHX together for 24 h expressed Mx and vig-1 transcripts similar to cultures treated withpoly IC alone (Fig. 2). Cycloheximide only slightly affected the induction of Mx and vig-1 transcripts by CSV inRTS11 but almost completely blocked induction in RTG-2 (Fig. 2). These results suggest that poly IC can directlyinduce antiviral gene expression in RTS11 and RTG-2, without first inducing type I interferon production. This isalso true for CSV in RTS11. On the other hand, induction in RTG-2 by CSV appears to require new protein synthesiseither for the production of interferon-like proteins or to support CSV replication.

3.4. Effect of ribavirin on Mx and vig-1 transcript induction by dsRNA and CSV

The two cell lines responded differently to ribavirin. Mx and vig-1 transcript induction by CSV was almost com-pletely blocked by ribavirin in RTS11, whereas in RTG-2 induction continued (Fig. 3). Poly IC-induced Mx and vig-1

Fig. 2. Effect of translation inhibitor, cycloheximide, on induction of Mx and vig-1 transcripts by poly IC and CSV. RTS11 (monocyte/macro-

phage-like) and RTG-2 (fibroblast-like) cultures were exposed to 50 mg/ml poly IC or 104.3 TCID50/ml CSV with or without 0.1 mg/ml cyclohex-

imide (CHX). After 24 h of treatment, RNA was collected and RTePCR performed for transcripts of Mx1, Mx2, Mx3, vig-1 and b-actin (b-a).

Control cultures for poly IC-treated cells had serum-free conditions, compared to CSV control cultures which had a 5% FBS supplement.

676 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

transcript expression was not affected by ribavirin in either cell line (data not shown), which rules out a general effectof ribavirin on cell metabolism or a preferential toxicity to one cell line. When either cell line was exposed to CSV,ribavirin completely blocked the development of cytopathic effects and viral production [32]. Together these resultssuggest that viral gene expression is required for Mx and vig-1 induction only in RTS11.

3.5. Effect of PKR inhibitors on Mx and vig-1 transcript induction by dsRNA and CSV

In RTS11, induction of Mx and vig-1 expression by 24 h treatments with poly IC was completely blocked by thePKR inhibitors, 2-aminopurine (2-AP) and compound 16, while the inhibitors partially blocked CSV induction(Fig. 4). Following 4 h exposures to poly IC, compound 16 also blocked Mx and vig-1 expression in RTS11 (datanot shown). By contrast, Mx and vig-1 transcript induction by poly IC in RTG-2 was not inhibited by either PKRinhibitor after 24 h (Fig. 4). For CSV in RTG-2 cultures, compound 16 did not affect Mx induction, but might haveimpaired vig-1 induction. These results imply that PKR or PKR-like kinases are required for the direct inductionof Mx and vig-1 transcripts in RTS11 but are less important to RTG-2.

3.6. Mx and vig-1 transcript induction by interferon-containing conditioned medium (ICCM)

RTS11 and RTG-2 were treated with ICCM from respectively RTS11 and RTG-2 cultures. ICCM was mediumconditioned by RTS11 and RTG-2 previously exposed to either poly IC or CSV. ICCM induced strong Mx and vig-1 expression by 24 h in all conditions except one. Induction was seen as early as 4 h in RTS11 cultures treatedwith ICCM derived from poly IC cultures and could still be detected at 24 h as well (Fig. 5). By contrast, inductionwas not evident in RTG-2 cultures treated with poly IC-ICCM. RTG-2 control cultures receiving control CM demon-strated low levels of Mx2 and Mx3 transcripts, and these levels were not enhanced in cultures treated with poly IC-ICCM (Fig. 5). On the other hand, CSV-ICCM, from RTS11 and RTG-2 cultures, previously exposed to CSV, inducedMx and vig-1 expression in both RTS11 and RTG-2 (Fig. 6).

3.7. Mx expression at the protein level

As detected by Western blotting, Mx protein was induced in RTS11 and RTG-2 cultures after some of the treat-ments that induced Mx transcripts, but not after one treatment, ICCM. Mx protein levels were induced in bothRTS11 and RTG-2 cultures 24 h after exposure to either poly IC or CSV (Fig. 7). Interestingly, in RTS11 cultures

Fig. 3. Effect of the viral replication inhibitor, ribavirin, on induction of Mx and vig-1 transcripts by CSV. RTS11 (monocyte/macrophage-like)

and RTG-2 (fibroblast-like) cultures were exposed to CSV (104.3 TCID50/ml) with or without 100 mg/ml ribavirin. After 24 h of treatment, RNA

was collected and RTePCR performed for transcripts of Mx1, Mx2, Mx3, vig-1 and b-actin (b-a). Control cultures had a 5% FBS supplement.

677S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

Mx protein levels had decreased back to control levels by 5 to 7 days, whereas the levels stayed at high in RTG-2(Fig. 7). The poly IC treatment was not cytotoxic to RTS11 or RTG-2 [27]. Likewise, cytotoxic effects were not ob-served in RTS11 cultures after CSV exposure and only became apparent in RTG-2 cultures after 5 days [32].

4. Discussion

4.1. Mx and vig-1 expression in unstimulated cells

In the absence of an experimental stimulus, expression of vig-1 and Mx was low or absent in the rainbow troutfibroblast (RTG-2) and monocyte/macrophage (RTS11) cell lines. As observed previously for RTG-2 [9], vig-1

Fig. 4. The effect of PKR inhibitors on the induction of Mx and vig-1 transcripts by poly IC and CSV. RTS11 (monocyte/macrophage-like) and

RTG-2 (fibroblast-like) cultures were exposed to 50 mg/ml poly IC or CSV (104.3 TCID50/ml) in the presence or absence of the PKR inhibitors, 2-

AP (1.5 mM) and compound 16 (0.5 mg/ml). After 24 h of treatment, RNA was collected and RTePCR performed for transcripts of Mx1, Mx2,

Mx3, vig-1 and b-actin (b-a). Control cultures for poly IC-treated cells had serum-free conditions, compared to CSV control cultures which had

a 5% FBS supplement.

678 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

transcripts were not expressed constitutively, and now this has been found to be the case for RTS11 as well. Mx tran-scripts were rarely found in RTG-2 but are occasionally detected in RTS11. Possibly a more sensitive method, such asquantitative PCR, might detect transcripts even in RTG-2. Mx protein could be detected in RTS11 but not commonlyin RTG-2 extracts, using Mx antiserum, which likely cross reacts with all three rainbow trout Mx proteins and detectsa 70e72 kDa polypeptide, which is the expected size for rainbow trout Mx [6]. Together, these data suggest that a lowlevel of Mx expression is possible in the absence of a stimulus in RTS11 and to a lesser extent in RTG-2. This couldindicate a unique expression pattern for Mx in macrophages. Interestingly, Atlantic salmon macrophages from somefish that had not been specifically stimulated showed expression of Mx protein [33]. In vivo, constitutive expression ofMx has been observed in several fish tissues [1,34,35].

4.2. Concurrent induction of Mx and vig-1 transcripts by poly IC and poly IC ICCM

Poly IC induced the three Mx genes and vig-1 concurrently in both RTS11 and RTG-2. This is the first demonstra-tion of the simultaneous induction of these four transcripts by poly IC. Previously, poly IC has been shown to induce

Fig. 5. Induction of Mx and vig-1 transcripts by interferon-containing conditioned medium (ICCM) from poly IC-treated cultures. ICCM was

prepared by treating RTS11 (monocyte/macrophage-like) and RTG-2 (fibroblast-like) cultures for 24 h with poly IC (50 mg/ml) and then allowing

them to condition medium without poly IC for 24 h. After 4 or 24 h, RNA was isolated and RTePCR performed for transcripts of Mx1, Mx2,

Mx3, vig-1 and b-actin (b-a).

Fig. 6. Induction of Mx and vig-1 transcripts by interferon-containing conditioned medium (ICCM) from CSV-treated cultures. ICCM was pre-

pared by treating RTS11 (monocyte/macrophage-like) and RTG-2 (fibroblast-like) cultures for 24 h with CSV (104.3 TCID50/ml) and then allow-

ing them to condition medium without CSV for 24 h. After 24 h, RNA was isolated and RTePCR performed for transcripts of Mx1, Mx2, Mx3,

vig-1 and b-actin (b-a).

679S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

Mx1 transcripts by 24 h in RTG-2 [36]. Vig-1 induction by dsRNA has not been previously demonstrated in any fishcell line. RTS11 might be slightly more sensitive to dsRNA, as induction of some genes could be seen as early as 4 hafter poly IC exposure, but this potential difference between the cell lines must wait confirmation by quantitative PCRapproaches.

Medium from cultures previously exposed to poly IC, which is referred to as poly IC induced interferon-containingculture medium (poly IC-ICCM), enhanced transcription of the three Mx genes and vig-1 concurrently in RTS11 butcaused little induction in RTG-2. This suggests that in RTS11 poly IC more rapidly and strongly induced the produc-tion of interferon activity, as indicated by the enhanced expression of Mx and vig-1. Experiments with RTG-2 indi-cated the production of some inducing activity without a stimulus because applying medium conditioned by controluntreated cells to new RTG-2 cultures caused expression of Mx2 and Mx3 transcripts. When the medium was condi-tioned by RTG-2 that previously had been exposed to poly IC, no enhancement of Mx2 and Mx3 transcript were seenand Mx1 and vig-1 transcripts remained absent. The simplest interpretation for this result is that in response to a 24 hpoly IC treatment, RTG-2 failed to produce and secrete additional IFN activity over the next 24 h. This was alsosupported by an experiment where Mx protein was monitored by Western blotting in RTG-2 cultures 24 h after theapplication of ICCM from RTG-2 previously exposed to poly IC for 24 h. No Mx induction was seen (data not shown).In earlier work RTG-2 had been shown to produce ICCM after poly IC but the poly IC had been complexed with dex-trans [9,26,36]. These could have improved dsRNA uptake and stability, leading to better, sustained IFN productionafter their removal from RTG-2 cultures.

4.3. Indirect and direct routes to Mx and vig-1 transcript induction by poly IC and poly ICCM

The commonly accepted sequence of events by which dsRNA induces Mx in fish cell cultures is a two-step process,which is referred to here as an indirect mechanism and likely is responsible for the induction of Mx and vig-1 tran-scripts by poly IC-ICCM in RTS11. The first step is the induction of interferon (IFN) synthesis and release of IFN intothe medium [36]. The second step likely involves IFN triggering the Jak-STAT pathway, which consists of Januskinases (Jak) and signal transducers and activators of transcriptions (STAT). These mediate the induction of manyIFN-stimulated genes (ISGs) initially via ISGF3 formation [37] and later through IRF-1 [38,39]. Several componentsof the Jak-STAT pathway have been identified in fish, including STAT1 and IRF-1 [40,41]. IRF-1 stimulates transcrip-tion by binding an IFN-stimulating responsive element (ISRE). The promoter of rainbow trout Mx1 contains an ISRE[36], and as well, the promoter for mandarin carp viperin (vig-1) has an IRF-1 binding site [42]. When only the second

Fig. 7. Induction of Mx protein by poly IC and CSV. RTS11 (monocyte/macrophage-like) and RTG-2 (fibroblast-like) cultures were treated with

50 mg/ml poly IC or CSV (104.3 TCID50/ml). After 24 or 48 h or 5 days, cell extracts were prepared and subjected to SDS gel electrophoresis

followed by Western blotting for Mx. Ponceau staining of blots is shown below the Westerns to illustrate protein loading between samples.

680 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

step is required for antiviral gene induction, which is the case with ICCM, the response would be expected to bequicker. Indeed strong induction of Mx and vig-1 was seen with poly-ICCM by 4 h.

In addition to Mx and vig-1 induction by the above indirect mechanism, poly IC also appeared to directly induceexpression of the antiviral genes in both cell lines. This is because poly IC continued to induce Mx and vig-1 tran-scripts in the presence of cycloheximide (CHX) at a concentration that inhibited most protein synthesis in these cells[27]. Inhibiting protein synthesis would block the indirect process by blocking the first step, the synthesis and accu-mulation of IFN. A direct action by poly IC in the induction of ISG has been demonstrated in some studies with mam-malian cells [43,44], but this is a first for a piscine system. However, another stimulus does directly induce vig-1 inrainbow trout. In head kidney cells, vig-1 was induced by the VHSV G protein in the presence of CHX [9].

For primate cell lines, PKR appeared to have a role in this direct action [43]. Two PKR inhibitors blocked the in-duction by poly IC of Mx and vig-1 in RTS11 and of vig-1 in RTG-2. One inhibitor was 2-aminopurine (2-AP), whichis an adenine analog, and has been the classic inhibitor of PKR [45], although possibly other kinases could be inhibitedas well. The other, compound 16, has just recently become commercially available and is an oxindole functionalizedwith an imidazole and promises to be more specific as it is effective at lower concentrations [31]. Therefore, the directenhancement of Mx and vig-1 gene expression in RTS11 and of vig-1 in RTG-2 could be due to the activation by thedsRNA of a constitutively expressed PKR or PKR-like kinases.

4.4. Concurrent induction of Mx and vig-1 transcripts by CSV

This is the first demonstration in fish cell lines of Mx and vig-1 induction by an aquareovirus (AqRV) and of con-currently induced expression of the three Mx genes and a vig-1 gene by a virus. CSV belongs to subgroup A among theAqRV, and the only other AqRV to be studied for the capacity to induce Mx has been channel catfish virus (CCV),which belongs to the AqRV-D subgroup [14]. Injection of CSV induced an Mx transcript in the liver of channel catfish[35]. Most studies on Mx induction by viruses in vitro have involved rhabdoviruses, which unlike AqRV have a mem-brane. Several rhabdoviruses have been examined, including viral hemorrhagic septicemia virus (VHSV), Hiramerhabdovirus (HIRRV), and grass carp hemorrhagic virus (GCHV) [9,46,47]. In RTG-2, VHSV was shown to inducea transcript for Mx but not for vig-1 [9]. GCHV induced transcripts for two Mx genes in the carp cell line, CAB, butCHX blocked induction of one gene and not the other, implying different mechanisms of regulation [46]. By contrast,in the rainbow trout cell lines, the three Mx genes responded the same way, although differently in RTS11 than inRTG-2.

4.5. Indirect and direct routes to Mx and vig-1 transcript induction by CSV

In RTS11, CSVappeared to induce Mx and vig-1 transcripts directly. Supporting the direct action is the observationthat Mx but less so vig-1 transcripts were induced in the presence of CHX, which would have prevented the formationof ICCM and an indirect action. However, any direct pathway with CSV would have required transcription of viralgenes as ribavirin blocked induction. For mammalian cells, ribavirin has been shown to inhibit the transcription ofviral genes for reoviruses with little effect on host cell RNA synthesis [48]. Ribavirin is presumed to act similarlywith CSV and the rainbow trout cell lines, and indeed, ribavirin did not prevent Mx and vig-1 induction in RTS11by poly IC (data not shown). As PKR inhibitors blocked most induction, PKR and viral dsRNA appear to mediatethe induction of Mx and vig-1 transcripts in RTS11.

Whether CSV also induces Mx and vig-1 transcripts indirectly via the induction of interferon activity is less clear.Although no evidence of CSV production was seen in fish cell lines 24 h post infection [32], the ICCM could have hadsome CSV. Thus the antiviral gene expression induced by CSV ICCM could have been due to a few CSV acting bya direct mechanism. However, as many viruses do induce interferon in cells and CHX did not completely block Mxand vig-1 induction, the expectation is that in addition to the direct mechanism an indirect mechanism operates as wellin RTS11.

In RTG-2, CSV appeared to act primarily via an indirect mechanism, requiring the induction of IFN first to induceMx and vig-1 transcripts. This is supported by the observations that induction was almost completely blocked byCHX, which would block IFN synthesis, and that the medium conditioned by RTG-2 previously exposed to CSVwas able to induce Mx and vig-1 in RTG-2. CSV RNA replication/synthesis was not required because Mx inductionoccurred in the presence of ribavirin, although vig-1 induction did not. PKR inhibitors did not block the induction of

681S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

Mx transcripts by CSV or by CSV ICCM in RTG-2 (data not shown). In the mammalian literature, some examplesexist of viruses or viral components inducing IFN independent of PKR [49]. Possibly a CSV protein(s) rather thanRNA induces IFN and in turn Mx in RTG-2.

For reovirus infection of mammalian cells, the conversion of the capsid to intermediate subviral particles (ISVPs) isan essential step and the lysosomal proteases responsible for this can differ between macrophages and fibroblasts [50].As well as entering by endocytosis, ISVPs can under some circumstances penetrate cell membranes [51,52]. Little isknown about these steps in aquareoviruses, but possibly, CSV virions are processed and/or enter RTS11 and RTG-2differently, leading to different pathways for Mx and vig-1 induction.

4.6. Induction of Mx protein

Both poly IC and CSV caused Mx protein to accumulate in RTS11 and RTG-2 cultures by 24 h, but subsequentlyMx levels declined in RTS11 but not in RTG-2. This suggests that either posttranslational regulation of Mx differs inRTS11 and RTG-2 and/or the cell lines differ in their uptake and degradation of dsRNA. Previously, a 48 h poly ICtreatment had been shown to induce Mx in RTG-2 [53], and when the inducer was a complex of poly IC and dextran,Mx levels had begun to decline by 72 h [53]. For Mx protein induction by viruses in salmonid cell lines, ISAV but notIPNV induced Mx in SHK-1 and TO [54] and IHNV failed to induce in RTG-2 [53]. This is the first report of inductionwith CSV.

References

[1] Lockhart K, Bowden TJ, Ellis AE. Poly I:C-induced Mx responses in Atlantic salmon parr, post-smolts and growers. Fish Shellfish Immunol

2004;17:245e54.

[2] Salinas I, Lockhart K, Bowden TJ, Collet B, Secombes CJ, Ellis AE. An assessment of immunostimulants as Mx inducers in Atlantic salmon

(Salmo salar L.) parr and the effect of temperature on the kinetics of Mx responses. Fish Shellfish Immunol 2004;17:159e70.

[3] Leong JC, Trobridge GD, Kim CHY, Johnson M, Simon B. Interferon-inducible Mx proteins in fish. Immunol Rev 1998;166:349e63.

[4] Lindenmann J. Resistance of mice to mouse-adapted influenza A virus. Virol 1962;16:203e4.

[5] Lee SH, Vidal SM. Functional diversity of Mx proteins: variations on a theme of host resistance to infection. Genome Res

2002;12:527e30.

[6] Trobridge GD, Chiou PP, Leong JC. Cloning of the rainbow trout (Oncorhynchus mykiss) Mx2 and Mx3 cDNAs and characterization of trout

Mx protein expression in Salmon cells. J Virol 1997;71:5304e11.

[7] Larsen R, Rokenes T, Robertsen B. Inhibition of infectious pancreatic necrosis virus replication by Atlantic salmon Mx1 protein. J Virol

2004;78:7938e44.

[8] Kibenge MT, Munir K, Kibenge FS. Constitutive expression of Atlantic salmon Mx1 protein in CHSE-214 cells confers resistance to

infectious salmon anemia virus. Virol J 2005;2:75.

[9] Boudinot P, Massin P, Blanco M, Riffault S, Benmansour A. vig-1, a new fish genes induced by the rhabdovirus glycoprotein, has a virus-

induced homologue in humans and shares conserved motifs with the MoaA family. J Virol 1999;73:1846e52.

[10] Boudinot P, Riffault S, Salhi S, Carrat C, Sedlik C, Mahmoudi N, et al. Vesicular stomatitis virus and pseudorabies virus induce a vig1/cig5

homologue in mouse dendritic cells via different pathways. J Gen Virol 2000;81:2675e82.

[11] Chin KC, Cresswell P. Viperin (cig5), an IFN-inducible antiviral protein directly induced by human cytomegalovirus. Proc Natl Acad Sci

USA 2001;98:15125e30.

[12] Eaton WD. Anti-viral activity in four species of salmonids following exposure to poly inosinic:cytidylic acid. Dis Aquat Org 1990;9:

193e8.

[13] Alonso M, Leong JC. Suppressive subtraction libraries to identify interferon-inducible genes in fish. Marine Biotechnol 2002;4:74e80.

[14] Essbauer S, Ahne W. Viruses of lower vertebrates. J Vet Med 2001;48:403e75.

[15] Winton JR, Lannan CN, Fryer JL, Kimura T. Isolation of a new reovirus from Chum salmon in Japan. Fish Pathol 1981;15:155e62.

[16] La Patra S, Lauda KA, Jones GR. Aquareovirus interference mediated resistance to infectious hematopoietic necrosis virus. Vet Res

1995;26:455e9.

[17] Der SD, Zhou A, Williams BRG, Silverman RH. Identification of genes differentially regulated by interferon a, b, or gamma using

oligonucleotide arrays. Proc Natl Acad Sci USA 1998;95:15623e8.

[18] Sen GC. Novel functions of interferon-induced proteins. Cancer Biol 2000;10:93e101.

[19] Peel AL. PKR activation in neurodegenerative disease. J Neuropathol Exp Neurol 2004;63:97e105.

[20] Hansen JD, La Patra S. Induction of the rainbow trout MHC class I pathway during acute IHNV infection. Immunogenetics 2002;54:

654e61.

[21] Nam BH, Yamamoto E, Hirono I, Aoki T. A survey of expressed genes in leukocytes of Japanese flounder, Paralichthys olivaceus infected

with Hirame rhabdovirus. Dev Comp Immunol 2000;24:13e24.

[22] O’Farrell C, Vaghefi N, Cantonnet M, Buteau B, Boudinot P, Benmansour A. Survey of transcript expression in rainbow trout leukocytes

reveals a major contribution of interferon-responsive genes in the early response to a rhabdovirus infection. J Virol 2002;76(16):8040e9.

682 S.J. DeWitte-Orr et al. / Fish & Shellfish Immunology 23 (2007) 670e682

[23] Purcell MK, Kurath G, Garver KA, Herwig RP, Winton JR. Quantitative expression profiling of immune response genes in rainbow trout

following infectious haematopoietic necrosis virus (IHNV) infection or DNA vaccination. Fish Shellfish Immunol 2004;17:447e62.

[24] Hu CY, Zhang YB, Huang GP, Zhang QY, Gui JF. Molecular cloning and characterization of a fish PKR-like gene from cultured CAB cells

induced by UV-inactivated virus. Fish Shellfish Immunol 2004;17:353e66.

[25] Rothenburg S, Deigendesch N, Dittmar K, Koch-Nolte F, Haag F, Lowenhaupt K, et al. A PKR-like eukaryotic initiation factor 2a kinase

from zebrafish contains Z-DNA binding domains instead of dsRNA binding domains. Proc Natl Acad Sci USA 2005;102:1602e7.

[26] Garner JN, Joshi B, Jagus R. Characterization of rainbow trout and zebrafish eukaryotic initiation factor 2a and its response to endoplasmic

reticulum stress and IPNV infection. Dev Comp Immunol 2003;27:217e31.

[27] DeWitte-Orr SJ, Zorzitto JR, Sutton LP, Bols NC. Preferential induction of apoptosis in the rainbow trout macrophage cell line, RTS11, by

actinomycin D, cycloheximide and double stranded RNA. Fish Shellfish Immunol 2005;18:279e95.

[28] Ganassin RC, Bols NC. Development of a monocyte/macrophage-like cell line, RTS11, from rainbow trout spleen. Fish Shellfish Immunol

1998;8:457e76.

[29] Bols NC, Lee LEJ. Cell lines: availability, propagation and isolation. In: Hochachka PW, Mommsen TP, editors. Biochemistry and Molecular

Biology of Fishes. Amsterdam: Elsevier Science; 1994. p. 145e59.

[30] Karber J. Beitrag zur kollektiven behandlung pharmakologischer reihenversuch. Arch Exp Pathol Pharmakol 1931;162:480e3.

[31] Jammi NV, Whitby LR, Beal PA. Small molecule inhibitors of the RNA-dependent protein kinase. Biochem Biophys Res Commun

2003;308:50e7.

[32] DeWitte-Orr SJ, Bols NC. Cytopathic effects of chum salmon reovirus to salmonid epithelial, fibroblast and macrophage cell lines. Virus Res

2007;126:159e71.

[33] Nygaard R, Husgard S, Sommer AI, Leong JC, Robertsen B. Induction of Mx protein by interferon and double-stranded RNA in salmonid

cells. Fish Shellfish Immunol 2000;10:435e50.

[34] Tafalla C, Aranguren R, Secombes CJ, Figueras A, Novoa B. Cloning and analysis of expression of a gilthead sea bream (Sparus aurata) Mx

cDNA. Fish Shellfish Immunol 2004;16:11e24.

[35] Plant KP, Thune RL. Cloning and characterization of a channel catfish (Ictalurus punctatus) Mx gene. Fish Shellfish Immunol

2004;16:391e405.

[36] Collet B, Secombes CJ. The rainbow trout (Oncorhynchus mykiss) Mx1 promoter: structural and functional characterization. Eur J Biochem

2001;268:1577e84.

[37] Sen GC. Viruses and interferons. Annu Rev Microbiol 2001;55:255e81.

[38] Platanias LC. Mechanisms of type-1- and type-II-interferon-mediated signalling. Nat Rev Immunol 2005;5:375e86.

[39] Shuai K, Liu B. Regulation of Jak-STAT signaling in the immune system. Nat Rev Immunol 2003;3:900e11.

[40] Collet B, Secombes CJ. Type I-interferon signalling in fish. Fish Shellfish Immunol 2002;12:389e97.

[41] Zhang Y, Gui J. Molecular characterization and IFN signal pathway analysis of Carassius auratus CaSTAT1 identified from the cultured cells

in response to virus infection. Dev Comp Immunol 2004;28:211e27.

[42] Sun BJ, Nie P. Molecular cloning of the viperin gene and its promoter region from the mandarin fish Siniperca chuatsi. Vet Immunol

Immunopathol 2004;101:161e70.

[43] Memet S, Besancon F, Bourgeade M-F, Thang MN. Direct induction of interferon gamma and interferon-a/b-inducible genes by double-

stranded RNA. J Interferon Res 1991;11:131e41.

[44] Decker T. Double-stranded RNA and interferon-a induce transcription through different molecular mechanisms. J Interferon Res

1992;12:445e8.

[45] Hu Y, Conway TW. 2-aminopurine inhibits the double-stranded RNA-dependent protein kinase both in vitro and in vivo. J Interferon Res

1993;13:323e8.

[46] Zhang YB, Li Q, Gui JF. Differential expression of two Carassius auratus Mx genes in cultured CAB cells induced by grass carp hemor-

rhage virus and interferon. Immunogenetics 2004;56:68e75.

[47] Lin OE, Ohira T, Hirono I, Saito-Taki T, Aoki T. Immunoanalysis of antiviral Mx protein expression in Japanese flounder (Paralichthysolivaceus) cells. Dev Comp Immunol 2005;29:443e55.

[48] Rankin JT, Eppes SB, Antczak JB, Joklik WK. Studies on the mechanism of the antiviral activity of ribavirin against reovirus. Virology

1989;168:147e58.

[49] Stewart MJ, Blum MA, Sherry B. PKR’s protective role in viral myocarditis. Virology 2003;314:92e100.

[50] Golden JW, Bahe JA, Lucas WT, Nibert ML, Schiff LA. Cathespin S supports acid-independent infection by some reoviruses. J Biol Chem

2004;279:8547e57.

[51] Borsa J, Morash BD, Sargent MD, Copps TP, Lievaart PA, Szekely JG. Two modes of entry of reovirus particles into L cells. J Gen Virol

1979;45:161e70.

[52] Tosteson MT, Nibert ML, Field BN. Ion channels induced in lipid bilayers by subvirion particles of the nonenveloped mammalian reoviruses.

Proc Natl Acad Sci USA 1993;90:10549e52.

[53] Trobridge GD, Chiou PP, Kim CH, Leong JC. Induction of the Mx protein of rainbow trout Oncorhynchus mykiss in vitro and in vivo with

poly I:C dsRNA and infectious hematopoietic necrosis virus. Dis Aquat Org 1997;30:91e8.

[54] Jensen I, Robertsen B. Effect of double-stranded RNA and interferon on the antiviral activity of Atlantic salmon cells against infectious

salmon anemia virus and infectious pancreatic necrosis virus. Fish Shellfish Immunol 2002;13:221e41.