Mi

Ra

b

a

ARRAA

KANIAId

1

ghrcTcbans(cDcdp

Pf

0d

Vaccine 29 (2011) 6976– 6985

Contents lists available at ScienceDirect

Vaccine

j ourna l ho me pag e: www.elsev ier .com/ locate /vacc ine

DCK cell line with inducible allele B NS1 expression propagates delNS1nfluenza virus to high titres

. van Wielinka,b,∗, M.M. Harmsena, D.E. Martensb, B.P.H. Peetersa, R.H. Wijffelsb, R.J.M. Moormanna

Central Veterinary Institute of Wageningen UR (CVI), P.O. Box 65, 8200 AB Lelystad, the NetherlandsWageningen University, Bioprocess Engineering, P.O. Box 8129, 6700 EV Wageningen, the Netherlands

r t i c l e i n f o

rticle history:eceived 25 March 2011eceived in revised form 4 July 2011ccepted 11 July 2011vailable online 23 July 2011

eywords:vian influenza

a b s t r a c t

Influenza A viruses lacking the gene encoding the non-structural NS1 protein (delNS1) have potentialuse as live attenuated vaccines. However, due to the lack of NS1, virus replication in cell culture isconsiderably reduced, prohibiting commercial vaccine production. We therefore established two stableMDCK cell lines that show inducible expression of the allele B NS1 protein. Upon induction, both celllines expressed NS1 to about 1000-fold lower levels than influenza virus-infected cells. Nevertheless,expression of NS1 increased delNS1 virus titres to levels comparable to those obtained with an isogenicvirus strain containing an intact NS1 gene. Recombinant NS1 expression increased the infectious virus

S1 proteinnducible expressionpoptosis

nterferonelNS1

titres 244 to 544-fold and inhibited virus induced apoptosis. However, NS1 expression resulted in onlyslightly, statistically not significant, reduced levels of interferon-� production. Thus, the low amount ofrecombinant NS1 is sufficient to restore delNS1 virus replication in MDCK cells, but it remains unclearwhether this occurs in an interferon dependent manner. In contrast to previous findings, recombinantNS1 expression did not induce apoptosis, nor did it affect cell growth. These cell lines thus show potentialto improve the yield of delNS1 virus for vaccine production.

© 2011 Elsevier Ltd. All rights reserved.

. Introduction

Influenza A virus is the causative agent of a highly conta-ious disease that affects both humans and animals. Outbreaks ofighly pathogenic avian influenza (AI) strains among poultry haveesulted in transmissions to humans, often with fatal outcome. Vac-ination of poultry is an important tool to control such outbreaks.raditional vaccines consist of inactivated viral particles whichlosely resemble the circulating strain. Currently vaccine strains areeing developed with improved characteristics, which include livettenuation, and the ability to serologically differentiate betweenaturally infected and vaccinated-only animals (DIVA). Influenzatrains lacking the gene encoding the non-structural NS1 proteindelNS1) are able to replicate in a host, albeit with reduced effi-iency, indicating that NS1 is dispensable for virus replication.elNS1 viruses are therefore considered as live-attenuated vac-

ines [1] for both animals [2,3] and humans [4]. Furthermore,elNS1 virus could be used as DIVA vaccine in poultry, as theresence of antibodies against NS1 in infected animals, but not

∗ Corresponding author at: Central Veterinary Institute of Wageningen UR (CVI),.O. Box 65, 8200 AB Lelystad, the Netherlands. Tel.: +31 320 238787;ax: +31 320 238668.

E-mail address: rutger.vanwielink@wur.nl (R. van Wielink).

264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2011.07.037

in vaccinated animals, allows for better surveillance of AI out-breaks and thus may improve worldwide acceptance of vaccination[5].

NS1 facilitates the infection of a cell in many different waysand the reduced replication efficiency of delNS1 virus is a directeffect of the absence of these functions. In cells infected with intactinfluenza virus, NS1 is expressed at very high levels shortly afterinfection [6–8] and it inhibits the cellular antiviral response byreducing the type I interferon (IFN) expression (such as IFN-�or IFN-�) at various levels. NS1 binds to retinoic-acid induciblegene I (RIG-I) that recognizes both cytoplasmic dsRNA and 5′-triphosphate-containing RNA, products that are generated duringinfluenza infection [9]. NS1 also binds directly to dsRNA [1]. Bybinding to these molecules, NS1 inhibits the activation of severalpathways that lead to IFN induction. Furthermore, NS1 also limitsIFN expression by inhibiting cellular pre-mRNA processing (includ-ing IFN pre-mRNA) and mRNA nuclear export [6]. The efficiencyand mechanism by which different influenza strains block the IFNresponse is strain dependent [10,11]. NS1 can also block the func-tion of two antiviral proteins, dsRNA-dependent protein kinaseR (PKR) and 2′-5′-oligoadenylatesynthetase (OAS) [1], and regu-

late both viral genome replication and translation [12] and proteinsynthesis [6].

The NS1 gene is localized on the eighth vRNA segment andsplicing of the primary NS1 mRNA results in the generation of

accine

apRtidb(watebeis

IlaktiIiwd

stym[ospriaii

Mdiiri(pifGadcAesMdwroer

R. van Wielink et al. / V

n mRNA that encodes the nuclear export protein (NEP) [6], arotein that plays a role in the nucleocytoplasmic export of viralNPs and regulation of viral genome transcription and replica-ion [13]. During development of delNS1 strains it is thereforemportant to retain the essential NEP protein. The NS1 protein isivided into two distinct functional domains: an N-terminal RNA-inding domain (residues 1–73) and a C-terminal effector domainresidues 74–230), which predominantly mediates interactionsith host-cell proteins [6]. Phylogenetic analysis of NS1 amino

cid sequences indicated that NS1 proteins can be divided intowo groups, termed alleles A and B [14,15]. The homology withinach allele is 93–100%, whereas between the two alleles it cane as little as 62%. Allele B NS1 is found, with a few exceptions,xclusively in avian influenza virus strains, whereas allele A NS1s found in all human, swine and equine strains and some aviantrains.

Infection of cells with delNS1virus induces higher levels ofFN than infection with a wild-type strain and results in aower virus yield [16,17]. In cells and animals with a low orbsence of IFN response, such as Vero cells, STAT1 or PKRnock-out mice, and embryonated eggs younger than 8 days,he virus replicates to high titres [16–19], which indicates thatnhibition of the IFN response is the main function of NS1 [1].nfluenza strains with C-terminally truncated NS1 genes showntermediate effects on IFN inhibition and virus replication [16],

hich correlates with the presence of the dsRNA binding proteinomain.

In large-scale influenza vaccine production, a serum-free anduspension cell-culture based process is preferred above the tradi-ional production in embryonated hen’s eggs [20]. However, theield of delNS1 virus, as reported by others [17] is below theinimal yield that is required for commercial vaccine production

21], even with IFN deficient Vero cells. Higher virus titres can bebtained by using a high multiplicity of infection (m.o.i.) [22], butuch large amounts of seeding virus are not feasible in large scaleroduction. Another approach to increase virus yield is transientecombinant expression of NS1 in the production cell line beforenfection [7]. However, host-cell encoded NS1 expression inducespoptosis [23–26]. This makes development of stable NS1 express-ng cell lines difficult, whereas transient expression is not feasiblen large scale production.

The aim of this paper is to develop a stable NS1 expressingadin–Darby canine kidney (MDCK) cell line for high-yielding

elNS1 virus production. To overcome apoptosis induction, anducible NS1 expression system is used. A commonly usednducible expression system is based on the prokaryotic Tetepressor protein (TetR) which allows expression of a gene ofnterest (GOI) under control of a tetracycline-response elementTRE) in a cell line that is stably transfected with a transactivatorrotein [27]. Two variants of this transactivator exist that respond

n an opposite manner to the doxycycline (Dox) antibiotic usedor GOI induction. Using the Tet-off system Dox addition repressesOI expression, whereas in the Tet-on Advanced system Doxddition induces GOI expression. An adherently growing, serumependent MDCK Tet-off cell line is commercially available. Foronstruction of cell lines stably expressing NS1 with the Tet-ondvanced system, which allows more tight regulation of genexpression, the previously generated MDCK-SFS cell line growingerum-free and in suspension [20] was used. We obtained twoDCK Tet-on Advanced cell lines that showed inducible pro-

uction of allele B NS1 protein. We subsequently determinedhether upon NS1 induction these two cell lines supported

eplication to high titres of an influenza virus that either containedr lacked an intact NS1 gene. Furthermore, we determined theffect of NS1 induction on IFN-� production and the apoptoticesponse.

29 (2011) 6976– 6985 6977

2. Material and methods

2.1. Cell lines and virus strains

MDCK-SFS and Vero-SF cells [20] were grown adherently inserum-free UltraMDCK culture media (LonzaBiowhittaker, Basel,Switzerland) and Optipro (Invitrogen, Carlsbad, CA, USA) respec-tively. MDCK Tet-off cells, obtained from Clontech (Mountain View,CA, USA), were grown in DMEM with 5% Tet-system approved fetalbovine serum (FBS). All culture media were supplemented with4 mM glutamine and 100 units/ml penicillin and 100 �g/ml strep-tomycin (Gibco). Human embryo kidney (293T) and MDCK cellsused for the generation of reassortant virus were cultured in Gluta-max medium (Invitrogen) supplemented with 10% FBS. Cells weregrown at 37 ◦C and 5% CO2. Cell density and viability were deter-mined with a Countess automated cell counter (Invitrogen).

The influenza virus strains A/Puerto Rico/8/34 H1N1 (PR8; ATCCVR-95, Manassas, VA, USA), A/turkey/Wisconsin/68 H5N9 (kindlyprovided by Dr. G. Koch, CVI), A/turkey/Turkey/1/05 H5N1 (Vet-erinary Laboratories Agency, Pirbright, UK) were propagated inembryonated chicken eggs.

2.2. Generation of recombinant influenza viruses

Viral RNA of influenza virus strain A/turkey/Turkey/1/05 H5N1was isolated from allantoic fluid of infected eggs using a highpure viral RNA isolation kit (Roche Applied Science, Penzberg,Germany). Universal influenza genome primer uni12 [28] wasused for reverse transcriptase (RT) reactions with the Superscriptfirst strand synthesis system (Invitrogen). HA segment forward (5′-TATTGCTCTTCAGCCAGCAAAAGCAGGGGTWYAATCTGTC-3′) andreverse (5′-ATATGCTCTTCGATTAGTAGAAACAAGGGTGTTTTTAAY-TAC-3′) primers and NA segment forward (5′-TATTGCTCTTCAG-CCAGCAAAAGCAGGAGTTCAAAATGAATCC-3′) and reverse (5′-ATATGCTCTTCGATTAGTAGAAACAAGGAGTTTTTTGAACAAAC-3′)primers with Sap I restriction sites were used for PCR reactionswith the Expand high fidelity PCR system (Roche). The resultingcDNAs were inserted in vector pPOLISAP1RIB [29]. The sequenceof each insert was verified (NA, Genbank Acc. nr. EF619973;HA, Genbank Acc. nr. EF619980) by sequence analysis using theBigDye terminator v1.1 cycle sequencing kit (Applied Biosystems,Carlsbad, CA, USA).

The sequence of the HA gene encoding the multi basic cleavagesite of A/turkey/Turkey/1/05 H5N1 was replaced by the corre-sponding sequence encoding the consensus cleavage site of a lowpathogenic HA gene of subtype H6 by means of PCR mutagene-sis as described previously [30]. Using this method, the H5 genesequence CCTCAAGGAGAGAGAAGAAGAAAAAAGAGAGGACTATTTencoding the multi basic cleavage site PQGERRRRKKRGLF wasconverted into the sequence CCAGAGATTGAAACTAGAGGACTTTTTencoding the low pathogenic H6 subtype cleavage site PQIETRGLF.The resulting gene is referred to as H5(6).

In order to generate an NS segment that only encodes the NEPprotein but not the NS1 protein we used a synthetic gene seg-ment (GenScript Corporation, Piscataway, NJ, USA) correspondingto the NS gene segment of influenza virus strain PR8 (EMBL Acc. nr.AF389122) in which the nucleotide sequence spanning the exactintron sequence (nt 57–528) was deleted. The NS1-deletion genesegment was inserted in pHW2000 using BsmBI restriction sites asdescribed earlier [28].

The bi-directional 8-plasmid reverse genetics system of strainPR8 was a gift of Dr. Hoffman and Dr. Webster [31]. To rescue

recombinant virus, a mixture of 15 × 105 293T and 5 × 105 MDCKcells was transfected with equal amounts of the eight plasmidscontaining the different gene segments, using Lipofectamine2000 (Invitrogen). At 12 h post transfection (hpt) the transfection

6 accine

mwSMwiwvr

2e

eps[tn(tamiN

vstapcftw(aw

lpi2(cGoa

2

DiTRANc(m3ar

978 R. van Wielink et al. / V

ixture was replaced by Glutamax medium supplementedith 0.3% bovine serum albumin (ChemieBrunschwig AG, Basel,

witzerland) and 250 ng/ml TPCK-trypsin (Sigma–Aldrich, St. Louis,O, USA). Virus was harvested at 48 hpt and 96 hpt. Virus stocksere generated in 7- to 9-day-old embryonated eggs and virus

dentity was confirmed by sequence analysis. Using this system weere able to rescue the PR8::H5(6)N1 and PR8::H5(6)N1delNS1

iruses. These strains are referred to as H5(6)N1 and delNS1,espectively.

.3. Generation of stable MDCK cells with inducible NS1xpression

The pTet-on Advanced expression system for inducible genexpression was obtained from Clontech. It includes the improvedTET-on Advanced and pTRE-tight vectors that make the inducibleystem more sensitive to Dox while reducing leaky gene expression32]. Synthetic DNA fragments (GenScript Corporation) encodinghe allele A NS1 from A/Puerto Rico/8/34 H1N1 (Genbank Acc.o. V01104) and allele B NS1 from A/turkey/Wisconsin/68 H5N9Genbank Acc. no. U85378) were inserted in the pTRE-tight vec-or after mutating the 3′-splice site for generation of NEP mRNA,s described previously [33]. Furthermore, a synthetic DNA frag-ent encoding the N-terminal 99 amino acids of allele ANS1 was

nserted into pTRE-tight. The resulting plasmids are pTRE-tight-S1allA, pTRE-tight-NS1allB and pTRE-tight-NS1allA1–99.

After transfection of MDCK-SFS cells with the pTet-On Advancedector using Fugene HD (Roche) stably transformed cells wereelected with 400 �g/ml G418 (Promega, Fitchburg, WI, USA). Func-ional expression of the rtTA-Advanced transactivator protein wasssessed by transient cotransfection of the selected cell lines withTRE-tight-LUC, that encodes firefly luciferase, and a vector thatonstitutively expresses Renilla luciferase as a transfection control,ollowed by induction with different concentration of Dox (Clon-ech). Two days later both the Renilla and firefly luciferase activitiesere measured using the Dual-Luciferase reporter assay system

Promega). One MDCK Tet-on Advanced cell line that showed highnd inducible firefly luciferase activity was selected for furtherork.

Both the MDCK Tet-on Advanced and the MDCK Tet-off cellines (Clontech) were cotransfected with either one of the threeTRE-tight derived plasmids and a linear hygromycin marker

n a ratio of 20:1. Stably transformed cells were selected with00 �g/ml hygromycin B (Clontech) and either 300 �g/ml G418Tet-on Advanced cells) or 1 �g/ml puromycin (Clontech; Tet-offells). Cell propagation was performed in the presence of 200 �g/ml418, 1 �g/ml puromycin and/or 100 �g/ml hygromycin. Nonef these antibiotics were used during virus infections and otherssays.

.4. NS1 mRNA quantification by qRT-PCR

MDCK cells (4 × 105) were incubated in 0, 0.1 and 1 �g/mlox for 24, 48 and 72 h in 6-well plates, after which RNA was

solated with the RNeasy plus kit (Qiagen, Hilden, Germany).he NS1 mRNA concentration was then determined by real-timeT-PCR using the QuantiTect SYBR Green RT-PCR kit (Roche).

calibration curve was used for both allele A NS1 and allele BS1 genotypes, made by a serial dilution of RNA isolated fromells 24 h post infection (hpi) with H1N1 or H5N9 influenza virusm.o.i. 5). NS1 mRNA levels were corrected for the amount of actin

RNA. Allele A NS1 forward (5′-CAAAACATGGATCCAAACACTG-′) and reverse (5′-GAATCCGCTCCACTATCTGCT-3′) primers,llele B NS1 forward (5′-ACAGGGGGTTTGATGGTGA-3′) andeverse (5′-CTTTGGAGGGAGTGGAGGTC-3′) primers, and canine

29 (2011) 6976– 6985

actin forward (5′-GGCATCCTGACCCTGAAGTA-3′) and reverse(5′-GGGGTGTTGAAAGTCTCGAA-3′) primers were used.

2.5. SDS-PAGE and Western blot analysis

MDCK NS1allAoff, NS1allA1–99on, NS1allBon1, NS1allBon2 andTet-on Advanced cells were incubated for 24 h with or without1 �g/ml Dox and then subjected to reducing SDS-PAGE, usingNuPAGE® Novex® 12% Bis–Tris precast gels (Invitrogen). MDCK-SFS cells taken 24 h after infection with either H1N1 or H5N9 virus(m.o.i. 5) were used as a source of authentic NS1. As a positivecontrol for the truncated allele A NS1, MDCK Tet-on Advancedcells were incubated for 24 h with 1 �g/ml Dox, transiently trans-fected with pTRE-tight-NS1allA1–99, and subsequently incubatedfor another 24 h. Polypeptides were transferred to polyvinyli-denedifluoride membranes. NS1 was detected by immunoblottingusing a monoclonal mouse antibody against a peptide near theN-terminus of allele A NS1 (SC130568; Santa Cruz Biotechnology,Santa Cruz, CA, USA) or a custom-prepared rabbit antiserum (anti-NS1allB217–230) directed against a keyhole limpet hemocyaninconjugated peptide (CKQKRYMARRVESEV; N-terminal acetylation)encompassing the C-terminal 14 amino acids of allele BNS1 of H5N9virus with an additional N-terminal cysteine for conjugation pur-poses. The peptide-specific antibodies from the anti-NS1allB217–230antiserum were affinity purified using this same (unconjugated)peptide (GenScript Corporation). After subsequent incubation with,respectively, peroxidase-conjugated rabbit anti-mouse or goatanti-rabbit immunoglobulins (Dako, Glostrup, Denmark), proteinlevels were visualized with ECL plus (GE Healthcare, Bucking-hamshire, UK) and quantified with a Storm840 imaging system(MolecularDynamics, GE Healthcare).

2.6. Virus infection

For analysis of delNS1 virus replication by all four NS1 express-ing cell lines, the two control cell lines and the Vero cells, cellswere incubated in 96-well plates (3 × 104 cells/well), for 24 h withor without 1 �g/ml Dox. The following day cells were infected intriplicate with delNS1 virus (m.o.i. 0.001) and after 1 h, they werewashed twice and provided with 120 �l medium with 2 �g/mltrypsin-TPCK, without Dox or selection antibiotics and without FBSfor MDCK Tet-off cell lines. Three day post infection (dpi), virustitres were determined by M-gene specific qRT-PCR.

To study the infection kinetics of delNS1 and H5(6)N1 inNS1allBon1 and NS1allBon2, 1 × 106 cells per well were incu-bated in 6-well plates, with or without 1 �g/ml Dox. After 24 hincubation, one well per cell line was used to determine the celldensity. Next, cells were infected in triplicate with either delNS1 orH5(6)N1 virus at an m.o.i. of 0.01. One hour later cells were washedwith PBS and supplied with 5 ml UltraMDCK medium containing2 �g/ml trypsin-TPCK with or without Dox. Supernatant sampleswere taken 1, 24, 48 and 72 hpi and stored at −80 ◦C before deter-mining the virus yield.

The infectious influenza virus titre was measured by determin-ing the tissue culture infective dose required to infect 50% (TCID50)of MDCK cells. The influenza virus genome equivalent concentra-tion was derived from the amount of viral RNA copies determinedby M-gene specific qRT-PCR. These two assays were previouslydescribed [20].

2.7. Apoptosis measurement

MDCK NS1allBon1, NS1allBon2 and Tet-on Advanced controlcells were cultured for 24 h in the presence (1 �g/ml) or absenceof Dox, trypsinized and then allowed to adhere to the surfaceof a 96-well plate (1 × 104 cells/well) for 1.5 h. To determine

accine

iiDDettdsPpamcn

2

2

cwvtwuamITPaIR

2

c(vvGwtdcw1viEtwTbatso

2

it1

R. van Wielink et al. / V

nhibition of apoptosis induction by virus infection, cells werenfected with 5 m.o.i. of either delNS1, H5(6)N1 or mock-infected.ox was supplied to cells which were previously cultured withox. To determine apoptosis induction by host-cell-encoded NS1xpression, culture media with or without 1 �g/ml Dox was addedo cells that were previously cultured without Dox. 24 h after infec-ion or NS1 induction caspase-3 and caspase-7 activities wereetermined using the Caspase-Glo 3/7 Assay kit (Promega). Briefly,upernatant was removed and 25 �l Caspase-Glo reagent and 25 �lBS was added to each well. After 1 h incubation at room tem-erature the luminescence of each sample was measured using

GloMax-Multi luminometer (Promega). Results are given as theeans of triplicate experiments and represent the increase in

aspase activity as compared to control cells (mock infected andon-induced).

.8. IFN- ̌ quantification

.8.1. IFN- ̌ mRNA RT-PCRMDCK NS1allBon1, NS1allBon2 and Tet-on Advanced control

ells were incubated for 24 h in 6-well plates (6 × 105 cells/well),ith or without 1 �g/ml Dox. Cells were then infected with delNS1

irus (m.o.i. 3) or mock-infected. After 18 h, RNA was isolated usinghe RNeasy plus kit (Qiagen). The concentration of IFN-� mRNAas determined by real-time RT-PCR (Onestep RT-PCR kit, Qiagen),sing canine IFN-� forward (5′-CCAGTTCCAGAAGGAGGACA-3′)nd reverse (5′-CCTGTTGTCCCAGGTGAAGT-3′) primers and Taq-an probe (5′-6FAM-CCTGGAGGACGTCAAAGAGAAGGA-XT-PH).

FN-� mRNA levels were corrected for the amount of actin mRNA.he concentration of actin mRNA was determined by real-time RT-CR using QuantiTect SYBR Green RT-PCR kit (Roche), using caninectin primers (Section 2.4). A calibration curve was used for bothFN and actin mRNA by a ten-fold serial dilution (100 to 10−7) ofNA isolated from delNS1-infected MDCK-SFS cells.

.8.2. BioassayMDCK NS1allBon1, NS1allBon2 and Tet-on Advanced control

ells were incubated for 24 h in 6-well plates (6 × 105/well), with1 �g/ml) or without Dox. Cells were then infected with delNS1irus (m.o.i. 3) or mock-infected. After 25 h, supernatant was har-ested and ultra-filtrated (Vivaspin 500; Sartorius, Goettingen,ermany) to remove virus. The absence of virus in the filtrateas confirmed by incubation of adherent MDCK-SFS cells with

he filtrate in a immunoperoxidase monolayer assay, as previouslyescribed [20]. To determine the levels of secreted IFN, MDCK-SFSells were seeded in 96-well plates (5 × 104/well) and incubatedith a serial two-fold dilution series of virus-free supernatant for

8 h. Cells were then infected (m.o.i. 3) with Newcastle diseaseirus (NDV) strain LaSota that was generated by reverse genet-cs from a Full-Length NDV cDNA copy (NDFL) that expresses anGFP reporter gene (NDFL-tagEGFP) [34] and incubated for a fur-her 48 h. After removal of supernatant, the fluorescence of eachell was measured using a GloMax-Multi fluorimeter (Promega).

he amounts of IFN present in the supernatants were estimatedased on their ability to induce an antiviral state in MDCK-SFS cellsnd inhibit NDFL-tag EGFP replication. The supernatant dilutionhat resulted in the median EGFP signal was determined for eachample. The reciprocal dilution was used as a relative concentrationf IFN present in the supernatant.

.8.3. IFN- ̌ dependent luciferase reporter system

To prepare conditioned medium (CM), MDCK-SFS cells were

nfected with delNS1 virus (m.o.i. 0.6) without the addition ofrypsin. After 15 h, supernatant was ultra-filtrated (Amicon Ultra-5 100k; Millipore, Billerica, MA, USA) to remove virus. This

29 (2011) 6976– 6985 6979

virus-free supernatant was further used to stimulate IFN produc-tion in MDCK cells.

MDCK NS1allBon1, NS1allBon2 and Tet-on Advanced controlcells were incubated for 24 h in 96-well plates (1.5 × 104/well),with or without 1 �g/ml Dox. Next, cells were washed and tran-siently cotransfected with the reporter plasmid carrying the fireflyluciferase gene under the control of the IFN-� promoter (p125Luc,kindly provided by Takashi Fujita, Kyoto University, Japan [35]) andthe Renilla luciferase control plasmid pGL4.73 (Promega), usingFugene HD (Roche). Cells were then incubated for 7 h, followedby IFN stimulation by addition of either 10 �g/ml poly-IC or 1%CM, or mock-stimulated. The firefly luciferase activity was mea-sured after 40 h incubation using the Dual Luciferase Reporter AssaySystem (Promega) and normalized to Renilla luciferase activity. Alltransfection experiments were conducted in duplicate.

2.9. Statistical analysis

Data were analysed for statistical significance by the Studentst-test, using the Minitab software package (Minitab Inc., Coventry,UK). Differences in mean were assumed significant if p < 0.05.

3. Results

3.1. Generation of MDCK cells with stable NS1 expression

Transfection of MDCK-SFS cells with the pTet-on Advancedplasmid resulted in 98 stable cell lines. These were screened fortransactivator activity by transient transfection with a luciferasereporter plasmid. A cell line that showed strong luciferase induction(177-fold) upon Dox addition was selected. Subsequent transfec-tion of this cell line and a commercially available MDCK Tet-offcell line with either one of the pTRE-tight NS1allA, pTRE-tightNS1allA1–99 or pTRE-tight NS1allB plasmids yielded 30 stablytransformed cell lines. Fifteen of these cell lines originated fromtransfection with pTRE-tight NS1allB and ten cell lines from trans-fection with pTRE-tight NS1allA1–99. Only five (all Tet-off) cell lineswere obtained from transfection with pTRE-tight NS1allA. These30 cell lines were then screened for NS1 mRNA expression andincreased production of delNS1 virus upon induction of NS1 expres-sion. Four cell lines, NS1allAoff, NS1allA1–99on, NS1allBon1 andNS1allBon2, that unequivocally showed NS1 mRNA expression andappeared to support growth of delNS1 virus to higher titres uponNS1 induction (results not shown) were selected for further studies.

3.2. NS1 mRNA and protein expression

The optimal Dox concentration and induction period leading tohigh NS1 mRNA levels was next determined. Fig. 1 gives the ratio ofNS1 mRNA levels relative to the NS1 RNA (the sum of mRNA, vRNAand cRNA) levels in cells infected with the corresponding wild typevirus. Note that we used the total amount of NS1 RNA present ininfected cells as a reference, which comprises viral mRNA, vRNAand cRNA. However, in transfected cells only NS1 mRNA is present.The ratio between NS1 mRNA and total NS1 RNA levels in infectedcells is about 0.2 at 24 hpi [8]. Thus, the relative NS1 mRNA levels inour stable MDCK cell lines compared to NS1 mRNA levels after virusinfection are probably 5-fold higher than that reported in Fig. 1.

Each of the four cell lines showed NS1 mRNA expressionafter 24 h induction, which did not further increase upon pro-longed induction (Fig. 1A–D). The NS1allAoff cell line showedabout 10,000-fold lower NS1 mRNA level than cells infected with

H1N1 virus in the non-induced state (with Dox). Furthermore theNS1 mRNA level did not change upon induction by Dox omis-sion (Fig. 1A). Similarly, the NS1allA1–99on cell line showed nochange in NS1 mRNA level upon induction (with Dox) and an

6980 R. van Wielink et al. / Vaccine 29 (2011) 6976– 6985

F levelsl �g/m(

acs(iwo

uabdeBMrmi

o

FpiwNu

ig. 1. Inducible expression of NS1 genes in stable MDCK cell lines. The NS1 mRNA

ines were determined by qRT-PCR 24, 46 and 70 h after induction with 0, 0.1 or 1.0panels A and B) or H5N9 virus (panels C and D) are presented.

bout 100-fold lower NS1 mRNA level than H1N1 virus infectedells (Fig. 1B). However, both NS1allBon1 and NS1allBon2 cell lineshowed inducible expression of NS1 mRNA that, in the inducedwith Dox) state, was about 1000-fold lower than in H5N9 virusnfected cells (Fig. 1C and D). For further studies, NS1 expression

as induced by adding 1 �g/mlDox to Tet-on cell lines or by Doxmission from Tet-off cell lines, both for a period of 24 h.

The NS1 protein expression was analysed by Western blottingsing allele-specific antibodies for detection of allele A NS1 (Fig. 2A)nd allele B NS1 (Fig. 2B). Both Western blots show many faintands representing aspecific binding to proteins of MDCK cells thato not produce NS1 (Fig. 2A, lane 11; Fig. 2B, lanes 11 and 12). How-ver, a band representing allele A NS1 (Fig. 2A, lanes 1–5) and allele

NS1protein (Fig. 2B, lanes 1–6) is readily observed in samples fromDCK cells 24 hpi with H1N1 or H5N9 virus. Furthermore, a band

epresenting the truncated allele A NS1 is visible near the 6 kDa

arker (Fig. 2A, lane 6), following transient transfection of Dox

nduced MDCK Tet-on Advanced cells with pTRE-tight-NS1allA1–99.NS1 bands could not be detected in NS1allAoff and NS1allA1–99

n cell extracts (Fig. 2A, lanes 7–10). Both NS1Bon1 and NS1Bon2

ig. 2. Western blot analysis of NS1 protein expression. MDCK NS1allAoff, NS1allA1–99onresence of 1.0 �g/ml Dox or without Dox. Panels A and B represent Western blot analy

nfected with H1N1 (allele A NS1) or H5N9 (allele B NS1) influenza virus at 24 hpi were usith H1N1 (panel A, lanes 1–5) or H5N9 (panel B, lanes 1–6) virus was included to allowS1, Dox induced MDCK Tet-on Advanced cells were transiently transfected with pTRE-tsed as negative control (panel A lane 11and panel B lanes 11 and 12). Relevant molecula

in MDCK NS1allAoff (A), NS1allA1–99on (B), NS1allBon1 (C) and NS1allBon2 (D) celll Dox. The relative NS1 mRNA level as compared to cells infected with H1N1 virus

cell lines show a strong band at the expected position afterinduction of NS1 expression (Fig. 2B, lanes 9 and 11), which is pre-dominantly absent without induction (Fig. 2B, lanes 8 and 10). Thefaint band at this position in non-induced cells is not assumed torepresent NS1 since it is also visible in cells that do not expressNS1 and were incubated with and without Dox (Fig. 2B, lanes 11and 12). A serial dilution of MDCK cells 24 hpi with H5N9 virus(Fig. 2B, lanes 1–6) was used to quantify the NS1 protein content ofthe induced NS1allBon cell lines. Densitometric analysis revealedthat the NS1allBon1 and on2 cell lines produced, respectively,1200- and 500-fold lower NS1 protein levels than H5N9 infectedcells.

3.3. Effect of NS1 on virus replication

The delNS1 total viral particle yield, as determined by the

virus genome equivalent titre, after infection of induced and non-induced cells was compared (Fig. 3). NS1 expression increaseddelNS1 virus yield 346-fold in NS1allBon1 cells (p = 0.015) and 82-fold in NS1allBon2 cells (p = 0.050). The virus yield obtained with

, NS1allBon1, NS1allBon2 and Tet-on Advanced cells were incubated for 24 h in thesis of allele A NS1 and allele B NS1, respectively. Cell extracts of MDCK-SFS cells

ed as source of authentic NS1. A serial 5-fold dilution series of cell samples infected densitometric NS1 quantification. As a positive control for the truncated allele A

ight-NS1allA1–99 (panel A, lane 6). Non-infected MDCK Tet-on Advanced cells werer weight markers (kDa) are indicated on the right of each panel.

R. van Wielink et al. / Vaccine 29 (2011) 6976– 6985 6981

Fig. 3. Effect of NS1 gene induction of four stable MDCK cell lines on production ofdelNS1 virus. Cells were induced with 1.0 �g/ml Dox (+) or not induced (−) and after24 h infected with 0.001 m.o.i. delNS1 virus. Three dpi the virus genome equivalentconcentration was measured by qRT-PCR. The two MDCK cell lines expressing onlythe transactivator and Vero cells were included as controls. Geometric mean titresad

bolHdyaconcow

vciiafeHtbo

N7st

3

cu(atciw

Fig. 4. Replication of delNS1 and H5(6)N1 viruses in NS1allBon1 (panel A) andNS1allBon2 (panel B) cell lines. Cells were either induced with 1 �g/ml Dox (closedsymbols, full line) for 24 h or not-induced (open symbols, dotted line), followed byinfection with 0.01 m.o.i. delNS1 or H5(6)N1 virus. The virus genome equivalentconcentration of delNS1 (squares) and H5(6)N1 (circles) viruses was determined byqRT-PCR. The titre of infectious delNS1 virus was determined by TCID (triangles).

nd 95% confidence interval of virus infections performed in triplicate (Vero cells inuplicate) are presented. Differences were significant (*) when p < 0.05.

oth induced NS1allBon cell lines was comparable to the virus yieldf Vero cells whereas non-induced NS1allBon cells yielded virusevels comparable to the Tet-on control cells, which lack a NS1 gene.owever, the virus yield of NS1allA1–99on cells and NS1allAoff cellsid not differ upon Dox treatment and was comparable to the virusield of the corresponding Tet-on and Tet-off control cells. Infectiont a higher m.o.i. (0.01) resulted in higher maximum titres for allell lines and a smaller difference between the delNS1 virus yieldbtained with induced and non-induced NS1allBon cell lines (dataot shown). Thus, only the two NS1allBon cell lines, in which weould demonstrate inducible NS1 expression, showed higher yieldsf delNS1 virus upon NS1 expression. Therefore only these cell linesere used in further studies.

We next determined the kinetics of delNS1 and H5(6)N1 totaliral particle production by induced and non-induced NS1allBonells (Fig. 4). The H5(6)N1 virus contains an intact NS1 gene buts further isogenic to delNS1 virus. Induction of NS1 expressionncreased delNS1 total viral particle yield by NS1allBon1 cellst 70 hpi 23-fold (p = 0.001; Fig. 4A) and by NS1allBon2 cells 5-old (p = 0.035; Fig. 4B). The titres at 70 hpi (about 108 TCID50quivalent/ml) are comparable to those found after infection with5(6)N1 virus (Fig. 4A and B), although the H5(6)N1virus appeared

o replicate somewhat faster. Induction of allele B NS1 expressiony both cell lines did not increase the replication speed nor the yieldf H5(6)N1 virus.

Strikingly, the induction of NS1 expression by NS1allBon1 andS1allBon2 cells increased the infectious delNS1 virus titre at0 hpi 244- and 544-fold, respectively (Fig. 4A and B), which is sub-tantially more than the, respectively, 23- and 5-fold increase inotal viral particle yield.

.4. Effect of NS1 on IFN- ̌ expression

We next determined the effect of NS1 expression by NS1allBonells on the induction of IFN-� expression by delNS1 virus infectionsing three different methods: quantification of IFN-� mRNA levelsFig. 5A), quantification of secreted IFN using a bioassay (Fig. 5B)nd the expression of a luciferase reporter gene under control of

he IFN-� promoter (Fig. 5C–E). Production of NS1 by NS1allBon2ells reduced the IFN-� mRNA expression after delNS1 virusnfection 3.5-fold (p = 0.045), whereas no significant difference

as observed for the NS1allBon1 and Tet-on cell lines (Fig. 5A).

50

Geometric mean titres and 95% confidence interval of the mean are presented ofvirus infections performed in triplicate.

Furthermore, IFN-� mRNA was not detectable in mock-infectedcells. The concentration of IFN secreted by delNS1 virus-infectedcells was determined by titration of MDCK-SFS cells with virus-freesupernatant, followed by infection of these cells with NDV thatexpresses EGFP for its quantification. NDV is known to be highlysusceptible to IFN. Thus cells stimulated with IFN produce lessNDV virus as assessed by the level of fluorescence. The productionof NDV viruses was translated to a relative IFN concentration asdescribed in Section 2. Although NS1 expression by NS1allBon1cells reduced IFN secretion almost 3-fold, this difference wasnot statistically significant. The IFN expression of mock-infectedNS1allBon1, NS1allBon2 and TET-on cells was below detec-tion limits, since the ultra-filtrated supernatant of these cellsdid not reduce EGFP expression after infection with NDFL-tagEGFP.

NS1 expression by NS1allBon2 cells reduced IFN secretion toan even lower extent (Fig. 5B). In the third assay, both NS1allBoncell lines showed slightly lower, although not statistically signif-icant, expression of a reporter gene under control of an IFN-�promoter when NS1 expression was induced, as compared to non-induced cells upon stimulation of IFN-� production by either CM(Fig. 5C) or poly-IC (Fig. 5D). These differences were smaller withoutstimulation of IFN-� production (Fig. 5E). Although the individualassays mostly do not yield statistically significant differences, thecombined results of these three assays suggest that induction of

NS1 expression by NS1allBon cells slightly reduces IFN production.However, this effect is minimal and therefore not likely to be criticalin the process of virus replication.

6982 R. van Wielink et al. / Vaccine 29 (2011) 6976– 6985

Fig. 5. Effect of NS1 expression on INF-� production determined with three dif-ferent methods. Before each assay, NS1allBon1, NS1allBon2 and Tet-on Advancedcontrol cells were induced for 24 h with 1.0 �g/ml Dox (+) or not induced (−). Theamount of IFN-� mRNA was determined by qRT-PCR, after infection with delNS1 ormock infection and normalized for the amount of actin mRNA (panel A). The relative

Fig. 6. Effect of NS1 expression on apoptosis and cell growth. NS1allBon1,NS1allBon2 and pTet-on Advanced control cells were induced with 1.0 �g/ml Dox(+), non-induced (−) or infected with delNS1 virus, followed by measurement of thecombined caspase-3 and -7 activity 24 h later (panel A). The results show the relativeincrease in caspase activity as compared to the corresponding non-induced, non-infected cells. The growth rate of these cells during a 48 h cultivation period in the

presence or absence of Dox is presented in panel B. Mean and their 95% confidenceinterval of triplicate experiments are given.

3.5. Effect of NS1 on apoptosis

Apoptosis induction was determined by measuring the com-bined activity of capase-3 and caspase-7, two proteins involved inthe activation of apoptosis [36]. Expression of NS1 in MDCK cellsis known to induce apoptosis [23,24,26]. Furthermore, infection ofcells with delNS1 virus is known to induce apoptosis more effi-ciently than infection with virus containing an NS1 gene [37–41],indicating that NS1 can inhibit apoptosis during an infection. Withrespect to the apoptosis induction by NS1 expression, no signifi-cant difference in caspase activity (Fig. 6A) or growth rate (Fig. 6B)was observed between Dox-induced and non-induced NS1allBon1and NS1allBon2 cell lines, which indicates that allele B NS1 expres-sion in these cells does not induce apoptosis. As a positive control,cells were infected with delNS1, which yielded a four to nine-foldincrease in caspase activity (Fig. 6A).

We next analysed the effect of NS1 expression by NS1allBoncells on apoptosis induction by both delNS1 and H5(6)N1 virusinfections. As expected, delNS1 virus induced about 60% higher cas-pase activity than H5(6)N1 virus in pTet-on Advanced control cellsand Dox addition did not affect caspase activity (Fig. 7A). However,expression of NS1 24 h prior to delNS1 virus infection reduced thecaspase activity of NS1allBon1 cells 2.1-fold (p = 0.017; Fig. 7B) andNS1allBon2 cells 2.0-fold (p = 0.018; Fig. 7C). The caspase activityin the induced cells was comparable to those in cells infected withH5(6)N1 virus. Expression of NS1 prior to infection with H5(6)N1virus did not affect caspase activity of NS1allBon2 cells (Fig. 7C) but

did reduce caspase activity of NS1allBon1 cells 1.4-fold (Fig. 7B)which was, however, not statistically significant (p = 0.074). Thus,NS1 expression by NS1allBon cells reduces apoptosis induction by

amount of excreted IFN was determined after infection of the cells with delNS1, fol-lowed by removal of virus from the supernatant 24 h later by ultra-filtration. The IFNin the supernatant was titrated by stimulating MDCK-SFS cells with a serial dilutionof the supernatant before infection with NDFLtag-EGFP virus. The relative amountof IFN present in the supernatant was determined by the dilution of the supernatantat 50% EGFP signal (panel B). Furthermore, the cells were transfected with an IFN-�dependent firefly luciferase reporter plasmid, followed by stimulation of the IFN-�production with 1% CM (panel C), 10 �g/ml poly-IC (panel D) or mock (panel E). Theluciferase activity was normalized by the Renilla luciferase activity. The mean and95% confidence interval of the mean of triplicate experiments (luciferase reporterassay in duplicate) are presented for each assay. Differences were significant (*)when p < 0.05.

R. van Wielink et al. / Vaccine

Fig. 7. Inhibition of apoptosis by NS1 expression. Tet-on Advanced control cells(panel A), NS1allBon1 (panel B) and NS1allBon2 (panel C) cells were induced with1.0 �g/ml Dox (+) or non-induced (−), followed by infection with delNS1 or H5(6)N1virus. The combined caspase-3 and -7 activity was measured 24 hpi. The relativeita

at

4

osteeetnMnti[NlcslswHik[oudil[

Nicw

ncrease in caspase activity as compared to the corresponding non-infected cells washen calculated. Mean and their 95% confidence interval of triplicate experimentsre given. Differences were significant (*) when p < 0.05.

virus lacking an NS1 gene to a level that is comparable with cellshat are infected with a virus that contains an intact NS1 gene.

. Discussion

Two stable MDCK cell lines that show inducible expressionf allele B NS1 from the A/turkey/Wisconsin/68 H5N9 influenzatrain were established: NS1allBon1 and NS1allBon2. Upon induc-ion, these cell lines showed 500 and 1200-fold lower allele B NS1xpression levels than virus-infected cells, as assessed by West-rn blot analysis. Attempts to generate stable cell lines similarlyxpressing either the full-length allele A NS1 or the allele A NS1 N-erminal domain from the A/PR/8/34 H1N1 influenza strain wereot successful. A previous attempt by others to establish a stableDCK cell line expressing allele A NS1 from A/Ty/Ont/66 was also

ot successful, which the authors attributed to apoptosis induc-ion due to leaky NS1 expression [26]. Transient NS1 expressions known to induce apoptosis in several cell lines, including MDCK23,24,26]. However, induction of allele B NS1 expression in the twoS1allBoncell lines did not induce apoptosis nor did it affect cel-

ular growth rate. The relatively low NS1 expression level by theseell lines is likely not the cause of the lack of apoptosis induction,ince others previously found apoptosis induction at comparablyow allele A NS1 expression levels [26]. Noticeably, most of thetudies into the effect of transient NS1 expression on apoptosisere done with allele A NS1. There is only one example of an5N2 strain containing allele BNS1 that does not induce apoptosis

n porcine cells [25]. The efficiency of apoptosis induction is alsonown to vary between allele A NS1 genes from different strains12,25,41,42]. Thus, it is plausible that the successful establishmentf the NS1allBon cell lines is related to the specific allele B NS1 genesed. One cell line with stable expression of an NS1-GFP protein waseveloped for the production of delNS1 virus, however no further

nformation on the origin of the NS1 protein, the NS1 expressionevels, virus replication efficiency and cell growth were described43].

Apart from apoptosis induction by recombinant NS1 expression,

S1 can also down-regulate the induction of apoptosis due to virus

nfection. This is suggested by the lower apoptotic response inells infected with wild-type influenza virus than in cells infectedith the corresponding isogenic strains lacking a functional

29 (2011) 6976– 6985 6983

NS1gene [37–40]. This inhibition of the apoptotic response wasattributed to activation by NS1 of the phosphatidylinositol 3-kinase(PI3 K)-Akt signalling pathway, which is known to result in ananti-apoptotic response [42,44]. We observed that recombinantallele B NS1expression suppresses the apoptotic response inducedby delNS1 virus to the same level as observed with NS1-expressingvirus, despite the 1000-fold lower NS1expression level as com-pared to virus-expressed NS1. Others similarly observed that lowNS1 expression levels are enough to saturate the intracellularAkt-signalling pathway, leading to apoptosis suppression [42].Other methods by which NS1 could have downregulated apo-ptosis are inhibition of PKR, OAS/RNAse L or the JNK/AP-1 stresspathway [6].

NS1 expression in both NS1allBon cell lines increased the delNS1virus yield comparable to those found for Vero cells infected withthe same virus. Restoration of virus production upon recombinantallele B NS1 expression could be related to its function as an IFNantagonist. Organisms and cells with low or absent IFN response,like Vero cells, STAT1 or PKR knock-out mice and young embry-onated eggs, are capable of efficient delNS1 virus replication due totheir IFN deficiency [18]. In addition, IFN expression in MDCK cellswas reduced 5-fold after transient expression of allele ANS1 [7]. Inour studies, however, the NS1 expression by both NS1allBon celllines had only a minor effect on IFN production, as was observedin three independent assays. Different assays were used to covermultiple levels in the IFN cascade, by looking at either activation ofthe IFN-� promoter, IFN-� transcription directly or secretion of IFN,and by stimulating IFN induction with either delNS1 virus, poly-ICor CM. Because allele B NS1 expression only has a minimal effecton IFN induction, it appears that the increase in virus productionby recombinant NS1during delNS1 infection is not IFN related. Thelack of IFN inhibition could be connected to the origin of the NS1gene, since inhibition of IFN by NS1 is strain dependent [10,12] andallele B NS1 was found to be less efficient in the suppression of IFN-� production than allele ANS1 [15,45]. Another possible cause couldbe the low expression level. During infection, NS1 is expressed inhigh quantities, suggesting that at least one of the many functionsof NS1 requires such a high expression level. This implicates thatnot all of the NS1 functions, e.g. inhibition of IFN induction, arenecessarily complemented by about 1000-fold lower NS1 expres-sion level in the NS1allBon cell lines. Another function of NS1 thatlikely does not occur in the NS1allBon cells upon Dox induction isinhibition of cellular mRNA preprocessing. Such a function wouldmost likely affect cellular processes and limit cell division. This washowever not observed, as cells grew with comparable high rates,both induced and non-induced.

The delNS1 virus titres from induced NS1allBon cells were com-parable to those found for non-induced cells infected with H5(6)N1virus. This indicates that the NS1 produced by the NS1allBon cellscan fully facilitate the viral infection in a similar manner as the alleleA NS1 that originates from the wild-type influenza virus. This is inconcordance with the observation that virus replication efficiencyis not dependent on the NS1 allele type [12]. Replication of H5(6)N1strain, which yielded high titres already in the absence of Dox, wasnot improved by Dox-induced NS1 expression. Similar observationswere made after transient transfection with NS1, although repli-cation speed of the wild-type strain was somewhat increased [7].Hence there is no significant advantage for NS1 being present atthe start of the infection. The efficiency in infectious delNS1 virusparticle production, as determined by the ratio of infectious virusparticles to the number of virus genome copy equivalent (relativeinfectious virus titre), is 0.1% for both NS1allBon cell lines in the

absence of Dox. Induction of NS1 expression increased the effi-ciency 11 to 95-fold, due to the strong increase in infectious virustitre (244 to 544-fold). This is especially relevant for live attenuatedvaccine production, since these rely on infectious virus particles.

6 accine

mcvswNorivtccifl

tFticTalNFdt

A

ISHo

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

984 R. van Wielink et al. / V

The increased relative infectious titre due to NS1 inductionay give a clue to the mechanism by which recombinant NS1

an complement the replication of delNS1 virus. NS1 regulatesiral genome replication and transcription [46], and NS1 expres-ion reduces viral RNA accumulation in cells transiently transfectedith the influenza polymerase complex genes PB1, PB2, PA andP, in comparison to cells that do not express NS1 [12]. The lackf regulation of the influenza polymerase complex by NS1 couldesult in an altered ratio among vRNA, cRNA and mRNA, resultingn lower infectious viral titres while not affecting the number ofirus genome copy equivalent titre, which measures all three RNAypes. Thus, the low yield of delNS1 virus in normal MDCK cellsould be due to the loss of NS1 regulation of viral genome repli-ation and transcription rather than the inability of the virus tonterfere with the host cell’s antiviral response. This suggestion isurther supported by the observation that IFN expression is not aimiting factor for influenza replication in MDCK cells [7].

The inducible expression of NS1 allows for further studies intohe timing and level of NS1 expression in relation to its function.urthermore, these cells are able to replicate delNS1 virus to highitres required in industrial production. The requirement for Doxn the culture medium may however limit their commercial appli-ation due to restriction in use of antibiotics in vaccine production.hus, stable production cell lines with constitutive expression ofllele B NS1are more suited for this purpose. The feasibility of estab-ishing such cell lines is suggested by our observation that allele BS1 expression does not affect cell growth (nor induce apoptosis).inally, the use of allele B NS1 may be advantageous for the pro-uction of DIVA vaccines, because of the antigenic difference withhe more common allele A protein.

cknowledgements

This research was funded by the Impulse Veterinary Aviannfluenza Research in the Netherlands program of the Economictructure Enhancement Fund. The authors thank M. Tacken, R.eutink and G. Tjeerdsma (CVI) for their help with the developmentf various assays and technical assistance.

eferences

[1] Richt JA, Garcia-Sastre A. Attenuated influenza virus vaccines with modifiedNS1 proteins. Curr Top Microbiol Immunol 2009;333:177–95.

[2] Richt JA, Lekcharoensuk P, Lager KM, Vincent AL, Loiacono CM, Janke BH,et al. Vaccination of pigs against swine influenza viruses by using anNS1-truncated modified live-virus vaccine. J Virol 2006;80(November (22)):11009–18.

[3] Wang L, Suarez DL, Pantin-Jackwood M, Mibayashi M, Garcia-Sastre A, Saif YM,et al. Characterization of influenza virus variants with different sizes of the non-structural (NS) genes and their potential as a live influenza vaccine in poultry.Vaccine 2008;26(July (29–30)):3580–6.

[4] Wressnigg N, Voss D, Wolff T, Romanova J, Ruthsatz T, Mayerhofer I, et al. Devel-opment of a live-attenuated influenza B DeltaNS1 intranasal vaccine candidate.Vaccine 2009;27(May (21)):2851–7.

[5] Capua I, Cattoli G. Diagnosing avian influenza infection in vaccinated popula-tions by systems for differentiating infected from vaccinated animals (DIVA).Develop Biol 2007;130:137–43.

[6] Hale BG, Randall RE, Ortin J, Jackson D. The multifunctional NS1 protein ofinfluenza A viruses. J Gen Virol 2008;89(October (10)):2359–76.

[7] Seitz C, Frensing T, Hoper D, Kochs G, Reichl U. High yields of influenza A virusin Madin–Darby canine kidney cells are promoted by an insufficient interferon-induced antiviral state. J Gen Virol 2010;91(July (7)):1754–63.

[8] Vester D, Lagoda A, Hoffmann D, Seitz C, Heldt S, Bettenbrock K, et al. Real-time RT-qPCR assay for the analysis of human influenza A virus transcriptionand replication dynamics. J Virol Methods 2010;168(September (1–2)):63–71.

[9] Mibayashi M, Martinez-Sobrido L, Loo YM, Cardenas WB, Gale Jr M, Garcia-Sastre A. Inhibition of retinoic acid-inducible gene I-mediated induction of

beta interferon by the NS1 protein of influenza A virus. J Virol 2007;81(January(2)):514–24.

10] Hayman A, Comely S, Lackenby A, Murphy S, McCauley J, Goodbourn S, et al.Variation in the ability of human influenza A viruses to induce and inhibit theIFN-beta pathway. Virology 2006;347(March (1)):52–64.

[

29 (2011) 6976– 6985

11] Kochs G, Garcia-Sastre A, Martinez-Sobrido L. Multiple anti-interferonactions of the influenza A virus NS1 protein. J Virol 2007;81(July (13)):7011–21.

12] Wang Z, Robb NC, Lenz E, Wolff T, Fodor E, Pleschka S. NS reassortment ofan H7-type highly pathogenic avian influenza virus affects its propagation byaltering the regulation of viral RNA production and antiviral host response. JVirol 2010;84(November (21)):11323–35.

13] Robb NC, Smith M, Vreede FT, Fodor E. NS2/NEP protein regulates transcriptionand replication of the influenza virus RNA genome. J Gen Virol 2009;90(June(Pt 6)):1398–407.

14] Ludwig S, Schultz U, Mandler J, Fitch WM, Scholtissek C. Phylogenetic rela-tionship of the nonstructural (NS) genes of influenza A viruses. Virology1991;183(August (2)):566–77.

15] Munir M, Zohari S, Metreveli G, Baule C, Belak S, Berg M. Allele A and Bnon-structural protein 1 of avian influenza A viruses differentially inhibitIFN-� production in human and mink lung cells. J Gen Virol 2011;(June),doi:10.1099/vir.0. 031716-0.

16] Kochs G, Koerner I, Thiel L, Kothlow S, Kaspers B, Ruggli N, et al. Properties ofH7N7 influenza A virus strain SC35M lacking interferon antagonist NS1 in miceand chickens. J Gen Virol 2007;88(May (5)):1403–9.

17] Garcia-Sastre A, Egorov A, Matassov D, Brandt S, Levy DE, Durbin JE, et al.Influenza A virus lacking the NS1 gene replicates in interferon-deficient sys-tems. Virology 1998;252(December (2)):324–30.

18] Egorov A, Brandt S, Sereinig S, Romanova J, Ferko B, Katinger D, et al. Transfec-tant influenza A viruses with long deletions in the NS1 protein grow efficientlyin Vero cells. J Virol 1998;72(August (8)):6437–41.

19] Talon J, Salvatore M, O’Neill RE, Nakaya Y, Zheng H, Muster T, et al. InfluenzaA and B viruses expressing altered NS1 proteins: a vaccine approach. Proc NatlAcad Sci USA 2000;97(April (8)):4309–14.

20] van Wielink R, Kant-Eenbergen HC, Harmsen MM, Martens DE, Wijffels RH,Coco-Martin JM. Adaptation of a Madin–Darby canine kidney cell line tosuspension growth in serum-free media and comparison of its ability to pro-duce avian influenza virus to Vero and BHK21 cell lines. J Virol Methods2011;171(January (1)):53–60.

21] Genzel Y, Reichl U. Continuous cell lines as a production system for influenzavaccines. Expert Rev Vaccines 2009;8(December (12)):1681–92.

22] Solorzano A, Webby RJ, Lager KM, Janke BH, Garcia-Sastre A, Richt JA. Mutationsin the NS1 protein of swine influenza virus impair anti-interferon activity andconfer attenuation in pigs. J Virol 2005;79(June (12)):7535–43.

23] Zhang C, Yang Y, Zhou X, Liu X, Song H, He Y, et al. Highly pathogenic avianinfluenza A virus H5N1 NS1 protein induces caspase-dependent apoptosis inhuman alveolar basal epithelial cells. Virol J 2010;7:51.

24] Lam WY, Tang JW, Yeung AC, Chiu LC, Sung JJ, Chan PK. Avian influenza virusA/HK/483/97(H5N1) NS1 protein induces apoptosis in human airway epithelialcells. J Virol 2008;82(March (6)):2741–51.

25] Daidoji T, Koma T, Du A, Yang CS, Ueda M, Ikuta K, et al. H5N1 avian influenzavirus induces apoptotic cell death in mammalian airway epithelial cells. J Virol2008;82(November (22)):11294–307.

26] Schultz-Cherry S, Dybdahl-Sissoko N, Neumann G, Kawaoka Y, HinshawVS. Influenza virus ns1 protein induces apoptosis in cultured cells. J Virol2001;75(September (17)):7875–81.

27] Gossen M, Bujard H. Anhydrotetracycline, a novel effector for tetracyclinecontrolled gene expression systems in eukaryotic cells. Nucleic Acids Res1993;21(September (18)):4411–2.

28] Hoffmann E, Krauss S, Perez D, Webby R, Webster RG. Eight-plasmid sys-tem for rapid generation of influenza virus vaccines. Vaccine 2002;20(August(25–26)):3165–70.

29] Fodor E, Devenish L, Engelhardt OG, Palese P, Brownlee GG, Garcia-Sastre A.Rescue of influenza A virus from recombinant DNA. J Virol 1999;73(November(11)):9679–82.

30] Peeters BP, de Leeuw OS, Koch G, Gielkens AL. Rescue of Newcastle disease virusfrom cloned cDNA: evidence that cleavability of the fusion protein is a majordeterminant for virulence. J Virol 1999;73(June (6)):5001–9.

31] Hoffmann E, Neumann G, Kawaoka Y, Hobom G, Webster RG. A DNA transfec-tion system for generation of influenza A virus from eight plasmids. Proc NatlAcad Sci USA 2000;97(May (11)):6108–13.

32] Urlinger S, Baron U, Thellmann M, Hasan MT, Bujard H, Hillen W. Explor-ing the sequence space for tetracycline-dependent transcriptional activators:novel mutations yield expanded range and sensitivity. Proc Natl Acad Sci USA2000;97(July (14)):7963–8.

33] Alonso-Caplen FV, Nemeroff ME, Qiu Y, Krug RM. Nucleocytoplasmic transport:the influenza virus NS1 protein regulates the transport of spliced NS2 mRNAand its precursor NS1 mRNA. Genes Dev 1992;6(February (2)):255–67.

34] Al-Garib SO, Gielkens AL, Gruys E, Peeters BP, Koch G. Tissue tropism inthe chicken embryo of non-virulent and virulent Newcastle diseases strainsthat express green fluorescence protein. Avian Pathol 2003;32(December(6)):591–6.

35] Yoneyama M, Suhara W, Fukuhara Y, Sato M, Ozato K, Fujita T. Autocrine ampli-fication of type I interferon gene expression mediated by interferon stimulatedgene factor 3 (ISGF3). J Biochem 1996;120(July (1)):160–9.

36] Ludwig S, Pleschka S, Planz O, Wolff T. Ringing the alarm bells: signalling

and apoptosis in influenza virus infected cells. Cell Microbiol 2006;8(March(3)):375–86.

37] Stasakova J, Ferko B, Kittel C, Sereinig S, Romanova J, Katinger H, et al. InfluenzaA mutant viruses with altered NS1 protein function provoke caspase-1 activa-tion in primary human macrophages, resulting in fast apoptosis and release

accine

[

[

[

[

[

[

[

[

R. van Wielink et al. / V

of high levels of interleukins 1beta and 18. J Gen Virol 2005;86(January(1)):185–95.

38] Xing Z, Cardona CJ, Adams S, Yang Z, Li J, Perez D, et al. Differential regulationof antiviral and proinflammatory cytokines and suppression of Fas-mediatedapoptosis by NS1 of H9N2 avian influenza virus in chicken macrophages. J GenVirol 2009;90(May (5)):1109–18.

39] Zhirnov OP, Konakova TE, Wolff T, Klenk HD. NS1 protein of influenza A virusdown-regulates apoptosis. J Virol 2002;76(February (4)):1617–25.

40] Jackson D, Killip MJ, Galloway CS, Russell RJ, Randall RE. Loss of function of theinfluenza A virus NS1 protein promotes apoptosis but this is not due to a failureto activate phosphatidylinositol 3-kinase (PI3K). Virology 2010;396(January

(1)):94–105.

41] Wang X, Shen Y, Qiu Y, Shi Z, Shao D, Chen P, et al. The non-structural(NS1) protein of influenza A virus associates with p53 and inhibits p53-mediated transcriptional activity and apoptosis. Biochem Biophys Res Commun2010;395(April (1)):141–5.

[

29 (2011) 6976– 6985 6985

42] Zhirnov OP, Klenk HD. Control of apoptosis in influenza virus-infectedcells by up-regulation of Akt and p53 signaling. Apoptosis 2007;12(August(8)):1419–32.

43] Kochs G, Martinez-Sobrido L, Lienenklaus S, Weiss S, Garcia-Sastre A, Staeheli P.Strong interferon-inducing capacity of a highly virulent variant of influenza Avirus strain PR8 with deletions in the NS1 gene. J Gen Virol 2009;90(December(Pt 12)):2990–4.

44] Ehrhardt C, Wolff T, Pleschka S, Planz O, Beermann W, Bode JG, et al. InfluenzaA virus NS1 protein activates the PI3K/Akt pathway to mediate antiapoptoticsignaling responses. J Virol 2007;81(April (7)):3058–67.

45] Zohari S, Munir M, Metreveli G, Belak S, Berg M. Differences in the ability to

suppress interferon Beta production between allele A and allele B NS1 proteinsfrom H10 influenza A viruses. Virol J 2010;7(December (1)):376.

46] Falcon AM, Marion RM, Zurcher T, Gomez P, Portela A, Nieto A, et al. DefectiveRNA replication and late gene expression in temperature-sensitive influenzaviruses expressing deleted forms of the NS1 protein. J Virol 2004;78(8):3880–8.