JGV Papers in Press. Published September 10, 2014 as doi:10.1099/vir.0.067660-0

1

Title: Paramyxovirus-based production of Rift Valley fever virus replicon

2

particles

3 4

Running title: NDV based production of RVFV replicons

5 6

Paul J. Wichgers Schreura,1, Nadia Oreshkovaa,c, Frank Hardersb, Alex Bossersb, Rob J.

7

M. Moormanna,c, Jeroen Kortekaasa

8 9

a)

Department of Virology, Central Veterinary Institute of Wageningen University and

10

Research Centre, Lelystad, The Netherlands.

11

b)

12

University and Research Centre, Lelystad, The Netherlands.

13

c)

14

Veterinary Medicine, Utrecht University, Utrecht, The Netherlands

Department of Infection Biology, Central Veterinary Institute of Wageningen Department of Infectious Diseases and Immunology, Virology Division, Faculty of

15 16

1Correspondence:

17

[email protected]

18

Tel. +31 320 238423

19

Fax. +31 320 38153

20 21 22

Contents category: Animal viruses – Negative strand RNA

23

Word count summary: 188

24

Number of figures: 7

25

Number of tables: 1

Word count main text: 4559

26 27 28

Keywords:

Rift

Valley

fever

virus,

29

glycoproteins, pseudotyping, replicon

Newcastle

disease

virus,

paramyxovirus,

30

Summary

31

Replicon particle-based vaccines combine the efficacy of live-attenuated vaccines with

32

the safety of inactivated or subunit vaccines. Recently, we developed Rift Valley fever

33

virus (RVFV) replicon particles, also known as nonspreading RVFV (NSR) and

34

demonstrated that a single vaccination with these particles can confer sterile immunity

35

in target animals. NSR particles can be produced by transfection of replicon cells, which

36

stably maintain replicating RVFV S and L genome segments, with an expression

37

plasmid encoding the RVFV glycoproteins, Gn and Gc, normally encoded by the M-

38

genome segment. Here, we explored the possibility to produce NSR with the use of a

39

helper virus. We show that replicon cells infected with a Newcastle disease virus

40

expressing Gn and Gc (NDV-GnGc) are able to produce high levels of NSR particles. In

41

addition, using reverse-genetics and site-directed mutagenesis, we were able to create an

42

NDV-GnGc variant that lacks the NDV fusion protein and contains 2 amino acid

43

substitutions in respectively Gn and HN. The resulting virus uses a unique entry

44

pathway that facilitates the efficient production of NSR in a one-component system. The

45

novel system provides a promising alternative for transfection-based NSR production.

46

Introduction

47

RVFV causes recurrent outbreaks among ruminants and humans across the African

48

continent, several islands off the coast of southern Africa and the Arabian Peninsula.

49

The virus is mainly transmitted by Aedine and Culicine mosquito vectors. The tri-partite

50

viral genome consists of a small (S), medium (M) and large (L) RNA genome segment of

51

negative polarity (Elliott, 1990, 1997). The S segment encodes a nucleoprotein (N) and a

52

non-structural protein named NSs, which is the main virulence factor by acting as an

53

antagonist of host innate immune responses (Billecocq et al., 2004; Bouloy et al., 2001;

54

Ikegami et al., 2009a, b; Muller et al., 1995). The M segment encodes a glycoprotein

55

precursor (GPC) that is co-translationally cleaved into Gn and Gc (Kakach et al., 1988),

56

which are the two major structural glycoproteins involved in virus-cell attachment and

57

membrane fusion. The M segment additionally encodes a 78-kDa protein of unknown

58

function that is incorporated in virions in the mosquito vector (Weingartl et al., 2014)

59

and a non-structural protein (NSm) with anti-apoptotic properties (Terasaki et al., 2013;

60

Won et al., 2007). The L segment encodes an RNA-dependent RNA polymerase

61

responsible for transcription and replication.

62

Recently, we and others developed the first generation RVFV replicon particles,

63

also known as nonspreading RVFV (NSR) (Dodd et al., 2012; Kortekaas et al., 2011).

64

These particles are able to infect cells, followed by genome replication and protein

65

expression but, due to the absence of the genes encoding the structural glycoproteins Gn

66

and Gc, are incapable of forming progeny particles. Their inherent safety greatly

67

facilitates studies on RVFV entry and fusion [18, 19]. Remarkably, NSR induces a

68

protective immune response in mice and sheep after a single immunization (Kortekaas et

69

al., 2011; Kortekaas et al., 2012; Oreshkova et al., 2013). Very recently, we also showed

70

that NSR is a potent vaccine vector. A single intranasal or intramuscular immunization

71

of mice with NSR expressing the hemagglutinin protein of influenza, from the NSs

72

location of the S-segment, provided full protection against a lethal dose of influenza virus

73

(Oreshkova et al., 2014).

74

NSR is produced by transfection of baby hamster kidney (BHK) cells, that stably

75

maintain replicating L and S genome segments (replicon cells), with an expression

76

plasmid encoding RVFV Gn and Gc (Kortekaas et al., 2011). Although the transfection-

77

based production of NSR particles is relatively efficient, our laboratory is exploring

78

alternative systems for the cost-effective industrial-scale production of NSR vaccines

79

without the need for transfection or replicon cell lines. In this study, we evaluated

80

whether a paramyxovirus helper virus expressing the RVFV glycoproteins could be used

81

to produce NSR. The results show that a previously developed Newcastle disease virus

82

(NDV) expressing Gn and Gc (NDV-GnGc) is able to efficiently package RVFV L and S

83

genome segments into progeny NSR. Moreover, a NDV-GnGc mutant virus is described

84

that employs a unique entry mechanism, thereby enabling NSR production in a one-

85

component system.

86

Results

87

Production of NSR with NDV-GnGc

88

To evaluate whether RVFV L and S genome segments can be packaged by NDV

89

expressing Gn and Gc of RVFV, we infected BHK-Rep2 cells (Kortekaas et al., 2011) with

90

NDV-GnGc (Kortekaas et al., 2010). BHK-Rep2 cells are BHK-21 cells that constitutively

91

maintain replicating L and S genome segments of RVFV strain 35/74 (Kortekaas et al.,

92

2011). One day post infection supernatants were evaluated for the presence of progeny

93

NSR. As a positive control, BHK-Rep2 cells were transfected with pCAGGS-M, which is

94

known to facilitate NSR production (Kortekaas et al., 2011). Supernatants of both,

95

infected and transfected wells, contained up to 107 TCID50/ml of NSR progeny (Fig. 1).

96

Uninfected and non-transfected negative control cells did not produce any NSR progeny.

97

These results show that BHK cells maintaining replicating RVFV L and S genome

98

segments can efficiently be infected by NDV-GnGc and subsequently produce progeny

99

NSR.

100 101

Use of NDV-GnGc-∆F replicons to produce NSR

102

Although NDV-GnGc is a potent helper virus for the production of NSR, the production

103

of NDV-GnGc, due to the absence of a polybasic cleavage site in the F-protein, depends

104

on embryonated chicken eggs (Dortmans et al., 2011; Kortekaas et al., 2010). To

105

circumvent the use of embryonated chicken eggs for the production of NDV-GnGc and to

106

preserve optimal safety, we constructed an NDV-GnGc replicon virus that can efficiently

107

grow in cell culture. Using NDV reverse genetics (Peeters et al., 1999; Romer-Oberdorfer

108

et al., 1999), we deleted the entire F gene from NDV-GnGc thereby generating NDV-

109

GnGc-∆F (Fig. 2(a),(b),(d)). In addition, a QM5 cell line expressing the F protein of NDV

110

strain Herts/33 (de Leeuw et al., 2005) (QM5-FHerts) was constructed (Fig. 2(c)). The FHerts

111

fusion protein contains a polybasic cleavage site facilitating infection in various cell

112

types. As expected, NDV-GnGc-∆F was unable to grow on QM5 cells but able to grow on

113

QM5-FHerts cells, resulting in the formation of NDV specific syncytia (Fig. 2(e)). Titers of

114

NDV-GnGc-∆F reached up to 107 TCID50/ml. Infection of BHK-Rep2 cells, at different

115

multiplicities of infection (MOIs), with NDV-GnGc or NDV-GnGc-∆F showed that NDV-

116

GnGc-∆F is equally efficient in the production of progeny NSR as NDV-GnGc (Fig. 2(f)).

117 118

Stability of NDV-mediated NSR production

119

To investigate whether NDV-assisted NSR production can be stable over several cell

120

passages, we passaged BHK-rep2 cells infected with NDV-GnGc-∆F several times and

121

monitored virus titers in the culture media. In two independent experiments, a rapid

122

decrease in NSR titers in the supernatants of sequentially passaged cells was observed

123

(data not shown), whereas in one experiment, NSR titers detected in the supernatant of

124

passaged infected cells remained relatively stable over 3 cell passages (Fig. 3).

125

Remarkably, starting from cell passage 2, mild cytopathic effects (CPE) were noted. CPE

126

increased in time and after cell passage 3, all cells died. This suggested that NDV-GnGc-

127

∆F had gained the ability to spread to a certain extent.

128

To evaluate whether the supernatant of cell passage 3 contained infectious NDV-

129

GnGc-∆F, besides infectious NSR, we inoculated fresh BHK-21 cells with the collected

130

supernatant. The induction of CPE and the presence of NDV antigens in these cells

131

confirmed the presence of small amounts of infectious NDV-GnGc-∆F. Interestingly,

132

evaluation of supernatants of passage 4-7 revealed a remarkably stable ratio of NSR and

133

infectious NDV-GnGc-∆F particles of about 1000:1. Likely, limited amounts of infectious

134

NDV-GnGc-∆F are sufficient for high level NSR production. From here on the

135

combination of high-level infectious NSR particles and low-level infectious NDV-GnGc-

136

∆F particles is referred to as NSRNDV (Fig. 3).

137 138

Next-generation sequence analysis of NSRNDV

139

To evaluate possible nucleotide changes in the genomes of NDV-GnGc-∆F and NSR in

140

NSRNDV, resulting from a putative adaptation to the BHK cells, we sequenced the RNA

141

isolated from NSRNDV passage 7 (Fig. 3) using Illumina-based next-generation

142

sequencing. The analysis, with a minimal coverage of 200, revealed that none of the

143

genome segments of either NDV-GnGc-∆F or NSR contained insertions or deletions.

144

Moreover, there was no evidence of recombination events. Nevertheless, as shown in

145

Table 1, several single nucleotide polymorphisms (SNPs) were detected in both the NSR

146

and NDV-GnGc-∆F genome. NSR contained 6 SNPs and NDV-GnGc-∆F contained 4

147

SNPs present in ≥ 50% of sequence reads. The non-synonymous SNPs in NSR did not

148

give clues about the mechanism responsible for the changed phenotype. In contrast, the

149

non-synonymous SNPs detected in NDV-GnGc-∆F in the Gn (Y303S) and HN

150

(I192M) proteins could have effects on virus entry, fusion and subsequent spread.

151 152

Y303S in Gn and I192M in HN are responsible for in vitro spreading of

153

NDV-GnGc-∆F

154

To investigate a possible role of the amino acid substitutions in HN and Gn in the NDV-

155

GnGc-∆F genome in the in vitro spreading phenotype of NSRNDV, we decided to construct

156

single and double site-directed NDV-GnGc-∆F mutants (Fig. 4(a)). All NDV-GnGc-∆F

157

site-directed mutants were viable and able to grow efficiently in QM5-FHerts cells up to

158

titers of 10E7 TCID50/ml (Fig. 4(b)). Remarkably, NDV-GnGc-∆F-HN* and NDV-Gn*Gc-

159

∆F-HN* containing the I192M substitution in HN showed increased syncytia

160

formation in the QM5-FHerts cells (Fig. 4(c)).

161

Infection of BSR cells with the site-directed mutants, grown on QM5-FHerts cells,

162

showed a significant variation in plaque formation (Fig. 4(c)). NDV-GnGc-∆F

163

predominantly infected single cells, whereas NDV-Gn*Gc-∆F and NDV-Gn*Gc-∆F-HN*

164

containing the Y303S substitution in Gn were able to spread to neighbouring cells.

165

Interestingly, BSR cells infected with the double site-directed mutant NDV-Gn*Gc-∆F-

166

HN* showed large virus plaques. Passage of supernatants to fresh cells revealed that

167

NDV-Gn*Gc-∆F and NDV-Gn*Gc-∆F-HN* are able to produce limited amounts of

168

infectious progeny in BSR cells.

169

We subsequently investigated the feasibility of using the NDV-Gn*Gc-∆F and

170

NDV-Gn*Gc-∆F-HN* viruses to produce NSR. BHK-Rep2 cells were infected with the

171

different site-directed mutants (grown on QM5-FHerts cells) and the levels of NSR progeny

172

in the supernatants were determined. The results show that none of the mutations had a

173

significant effect on the transient NDV-GnGc-∆F-mediated production of NSR (Fig. 4(e,

174

first bars). However, passage of the supernatants of infected BHK-Rep2 cells to fresh

175

wild-type BHK cells resulted in a rapid decrease of NSR production when using NDV-

176

GnGc-∆F, NDV-Gn*Gc-∆F or NDV-GnGc-∆F-HN* (Fig. 4(e)). In contrast, incubation of

177

cells with passaged supernatant of NDV-Gn*Gc-∆F-HN*-infected BHK-Rep2 cells again

178

resulted in NSR progeny. Altogether, these results indicate that the autonomous

179

production of NSRNDV, as observed previously (Fig. 3), is mediated by two amino acid

180

substitutions in NDV-GnGc-∆F in respectively HN and Gn.

181 182

NSRNDV production in a one-component system

183

To investigate whether NSRNDV could be produced repeatedly, we passaged supernatants

184

of NDV-Gn*Gc-∆F-HN* based NSRNDV, in BSR cells. The titers of NSR and NDV-Gn*Gc-

185

∆F-HN* were determined by monitoring eGFP expression and CPE, respectively. Of

186

note, immunoperoxidase monolayer assays (IPMA) demonstrated that CPE correlated

187

with Gn expression from the NDV genome. When no complete CPE was observed after

188

initial infection, due to lower levels of infectious NDV-Gn*Gc-∆F-HN*, cells instead of

189

supernatants were passaged. Ratios between NSR and NDV in each passage of NSRNDV

190

varied between 1:100 to 1:1000 (Fig. 4(f)). Altogether, these results indicate that NSRNDV

191

production is relatively stable upon passage and can be produced in a one-component

192

system (Fig. S1).

193 194

Spread NDV-Gn*Gc-∆F-HN* is mediated by RVFV Gn and Gc

195

To investigate whether RVFV Gn and Gc are involved in the cell entry of NDV-Gn*Gc-

196

∆F-HN* we incubated the virus with serial dilutions of a RVFV neutralizing serum. As a

197

control, we evaluated neutralization of NDV-GnGc and NDV-GnGc-∆F. In these

198

experiments, infection by NDV-GnGc or NDV-GnGc-∆F was not neutralized, whereas

199

infection by NDV-Gn*Gc-∆F-HN* was completely prevented by serum diluted up to 540

200

times (Fig. 5). Altogether, these results strongly suggest that infectivity of NDV-Gn*Gc-

201

∆F-HN* depends on the RVFV glycoproteins.

202 203

NSRNDV induces a protective immune response in mice

204

To investigate whether NSRNDV is equally effective as an NSR vaccine produced by

205

transfection, we performed a vaccination challenge experiment in mice. Mice were either

206

vaccinated with NSR (105.8 TCID50), NSRNDV (containing 105.8 TCID50 NSR and 102.8

207

TCID50 NDV-GnGc-∆F) or NDV-GnGc-∆F (102.8 TCID50) and challenged 3 weeks post

208

vaccination. As expected, the low amount of NDV-GnGc-∆F alone was unable to provide

209

protection and the challenge virus disseminated in high amounts to the brain and liver

210

(Fig. 6). In contrast, NSR and NSRNDV-vaccinated mice were completely protected from a

211

lethal RVFV challenge dose without any virus disseminating to the organs. Protection

212

could be correlated to high levels of neutralizing antibodies, which were detected before

213

challenge infection (Fig. 6(b)). In conclusion, NSR produced by the NDV-Gn*Gc-∆F-HN*

214

helper virus is able to protect mice against a lethal challenge dose.

215

Discussion

216

In previous work, we demonstrated that nonspreading RVFV (NSR) is highly effective in

217

preventing viremia and clinical disease in the natural target species. Our laboratory is

218

currently developing NSR production systems that are cost-effective and scalable. Here,

219

we evaluated the feasibility to produce NSR using a Newcastle disease virus (NDV) to

220

provide the RVFV structural glycoproteins Gn and Gc.

221

NDV-GnGc is able to induce a neutralizing immune response in mice and sheep

222

(Kortekaas et al., 2010; Kortekaas et al., 2012), strongly suggesting that Gn and Gc,

223

when expressed from the NDV genome, are correctly processed and assembled. The

224

ability of NDV-GnGc to complement the RVFV glycoproteins in the BHK-Rep2 cells was

225

therefore

226

complementation is explained by a combination of factors: i) high level expression and

227

correct processing of GnGc from the NDV genome, ii) none or low-level of competition

228

between NSR and NDV genome replication, and iii) the ability of NDV and NSR to infect

229

the same cell. Likely the highly similar nature of the NSR and NDV genomes, both

230

negative-sense, single-strand RNA, does not result in unbalanced use of host cell

231

machinery. We expect that similar methods may be successful for production of other

232

bunyavirus replicons, although we envisage that efficiencies will vary between certain

233

combinations of helper and replicon virus.

not

totally

unexpected.

Overall,

the

success

of

the

NDV-based

234

NDV-GnGc is based on lentogenic NDV strain LaSota and requires trypsin-like

235

proteases for cleavage and subsequent activation of the F protein. Growth of lentogenic

236

NDV is therefore most efficient in the allantoic cavity of embryonated eggs, where

237

trypsin-like proteases are abundant (Goldhaft, 1980; Kortekaas et al., 2010). To

238

circumvent the need for embryonated eggs, mesogenic or velogenic NDV strains can be

239

used that contain a polybasic-cleavage site in the F protein and thereby do not require

240

trypsin-like proteases for infectivity. However, these viruses are not avirulent and are

241

therefore not suitable for vaccine applications. As a safe alternative, we constructed an

242

NDV-GnGc-∆F replicon which can be amplified in a QM5 cell line constitutively

243

expressing a fusogenic F protein. The ability to construct a cell line stably expressing

244

high levels of a fusogenic F protein (Fig. 2(c)) may be partly related to the absence of the

245

HN protein. Indeed, several reports show that F mainly induces syncytia and cell death

246

in the presence of HN (de Leeuw et al., 2005; McGinnes et al., 2002; Sergel et al., 2000).

247

Although to our knowledge this is the first report that describes the construction of an

248

NDV replicon, several replicons have been developed for other paramyxoviruses,

249

including the Respiratory syncytial virus (RSV) (Malykhina et al., 2011) and Sendai

250

virus (Yoshizaki et al., 2006). Likely, deleting and subsequently trans-complementing

251

glycoproteins can be generally applied for members of the Paramyxoviridae.

252

After observing that a mixture of NSR and NDV-GnGc-∆F virus (named NSRNDV),

253

contained low levels of autonomously replicating NDV, the genome sequences of the

254

respective viruses were determined by next-generation sequencing. This analysis

255

revealed the presence of several SNPs in both genomes. We decided to focus on the SNPs

256

of NDV-GnGc-∆F that were present in ≥ 95% of sequence reads and located in the

257

glycoprotein genes, since these proteins are involved in receptor binding and fusion. The

258

ag transition in HN, resulting in amino acid change I192M, was of particular

259

interest, because this mutation was previously described and characterized by Estevez et

260

al. (Estevez et al., 2011), who showed that the I192M substitution decreased

261

hemagglutinating activity and increased fusion activity. Our results are in full

262

agreement with these observations and emphasize that paramyxovirus fusion activity is

263

a dynamic process that can be influenced by subtle changes in the HN globular head

264

(Porotto et al., 2012). We hypothesize that the HN mutation increases the binding

265

capacity of the NDV virion to the cell surface due to prolonged receptor contact, although

266

this remains to be confirmed experimentally.

267

The SNP in Gn was not reported previously. The mutation does not significantly

268

affect the infectivity of NSR. Titers of NSR produced by NDV-Gn*Gc-∆F and NDV-

269

Gn*Gc-∆F-HN* were only slightly decreased compared to NSR titers produced by NDV-

270

GnGc-∆F. It is important to note that the Y303S substitution in Gn is required for the

271

F-independent infection of NDV-GnGc-∆F and that infectivity of NDV-Gn*Gc-∆F-HN*

272

can efficiently be blocked by RVFV-specific neutralizing antibodies. By combining these

273

two observations we hypothesise that Gn and Gc are incorporated in NDV-GnGc-∆F

274

virions as non-functional complexes that cannot mediate infectivity. However, our

275

findings suggest that a single substitution in the Gn protein (Y303S) results in a

276

functional GnGc complex that is capable to facilitate cell entry of NDV-GnGc-∆F virions

277

in the absence of F complementation. Whereas NDV normally uses a class-I fusion

278

protein (Chang & Dutch, 2012) to fuse at the plasma membrane, the ability of NDV-

279

Gn*Gc-∆F-HN* to autonomously replicate in tissue culture is apparently attributed to a

280

switch from a class I to a class II pH-dependent fusion protein-mediated entry (Fig. 7).

281

The mechanisms of entry of the NDV variants described in this study will be subject of

282

future research.

283

Wild-type RVFV induces strong CPE in various mammalian cell lines including

284

the BHK and BSR cell lines used in this study. Remarkably, BHK-Rep2 cells, which

285

stably maintain replicating RVFV L and S genome segments in the absence of a

286

replicating M-segment, do not show signs of CPE. Most likely, RVFV-mediated

287

cytotoxicity is caused by high-level expression of the M-segment encoded glycoproteins

288

Gn and Gc. In NSRNDV-infected cultures CPE was observed as well, although this CPE

289

was consistently less pronounced than CPE resulting from RVFV infection. Considering

290

that CPE resulting from incubation of cells with NSRNDV, is solely attributed to NDV-

291

Gn*Gc-∆F-HN*, CPE was used to determine titers of NDV-Gn*Gc-∆F-HN* virions in

292

NSRNDV preparations. Titers of NSR particles in NSRNDV preparations were determined

293

by detection of eGFP, expressed from the S-segment of the NSR genome. Remarkably,

294

multiple passage of NSRNDV revealed only minor variation in NSR progeny production

295

and ratio’s between NSR and NDV in NSRNDV preparations were relatively stable over

296

time.

297

To efficiently propagate NSRNDV in vitro, relatively high MOI infections are

298

required. Low MOI infections do not result in the production of progeny since this

299

depends on co-infection of a cell with NSR and NDV. In vivo administration of NSRNDV

300

will, due to the dilution effect, always result in low MOI infections and consequently

301

NSRNDV is unable to spread efficiently. This, together with the absence of the major

302

virulence factor NSs, highlights the safety of the NSRNDV vaccine.

303

In conclusion, we here show that NDV-GnGc and especially the variant referred

304

to as NDV-Gn*Gc-∆F-HN* can be used for efficient production of NSR. This novel system

305

could be further evaluated as a cost-effective alternative for transfection-based NSR

306

production.

307 308

Methods

309 310

Ethics statement

311

The animal experiment was conducted in accordance with the Dutch Law on Animal

312

Experiments (Wod, ID number BWBR0003081) and approved by the Animal Ethics

313

Committee of the Central Veterinary Institute (Permit Number: 2013050).

314 315

Cells and growth conditions

316

All cell lines were routinely maintained at 37°C and 5% CO2. BHK-21, BHK-Rep2 cells

317

and BSR cells were grown in Glasgow Minimum Essential Medium (GMEM; Invitrogen,

318

Bleiswijk, The Netherlands) supplemented with 4% tryptose phosphate broth (TPB;

319

Invitrogen), 1% non-essential amino acids (NEAA; Invitrogen), 5% fetal bovine serum

320

(FBS; Bodinco, Alkmaar, The Netherlands) and 1% penicillin-streptomycin (Invitrogen).

321

QM5 cells were routinely grown in QT35 medium (Invitrogen) supplemented with 5%

322

FBS and 1% penicillin-streptomycin. QM5-FHerts cells were grown in the same medium as

323

QM5 cells, supplemented with 0.5 mg/ml Geneticin (G-418; Promega, USA).

324 325

Viruses

326

RVFV strain 35/74 (Barnard, 1979) was grown and titrated on BHK cells. NDV-GnGc

327

(Kortekaas et al., 2010) was grown in embryonated chicken eggs and titrated on QM5

328

cells. Viral titers were determined by serial dilution and expressed as 50% tissue culture

329

infective dose (TCID50) using the Spearman-Kärber algorithm. In the absence of CPE or

330

eGFP expression, viral infection was visualized by IPMA (see below).

331 332

Construction of a QM5 cell line stably expressing the F protein of the NDV

333

Herts strain

334

QM5 cells were transfected with pCIneo-FHerts (de Leeuw et al., 2005) using jetPEI

335

transfection reagents (Polyplus-transfection SA, Illkirch, France) according to the

336

manufacturers’ instructions. Two days post transfection, G-418 selection reagent was

337

added to the growth medium at a concentration of 0.5 mg/ml. One week post selection,

338

individual clones (>50) were sub-cultured and finally screened for FHerts expression by

339

IPMA. The clone with the most stable and homogenous F expression after numerous

340

passages was used for further studies and is from here on referred to as QM5-FHerts.

341

342

NDV reverse-genetics

343

The rescue of NDV virus by reverse genetics was performed as described previously

344

(Peeters et al., 1999; Romer-Oberdorfer et al., 1999) with some minor modifications.

345

Briefly, 6 × 106 QM5-FHerts cells (per well of 6 wells plate) were infected with a Fowlpox

346

virus expressing T7 polymerase (FP-T7) (Chaudhry et al., 2007) in Opti-MEM containing

347

0.2% FBS (Invitrogen). Two hours post FP-T7 infection, medium was replaced (Opti-

348

MEM, 0.2% FBS) and cells were transfected with pCIneo-NP, pCIneo-P, pCIneo-L and a

349

plasmid encoding the complete cDNA of NDV strain LaSota (pNDFL) using jetPEI

350

transfection reagents according to manufacturers’ instructions. Three hours post

351

transfection medium was replaced with QT-35 medium containing 5% FBS and three

352

days post transfection supernatants were filtered (0.2 µM to remove FP-T7) and used to

353

inoculate fresh QM5-FHerts cells.

354 355

Construction of pNDFL-GnGc-∆F

356

To delete the F gene in the infectious clone of NDV-GnGc, a fusion PCR strategy was

357

conducted. Briefly, two PCR fragments were generated from pNDFL-GnGc (Kortekaas et

358

al., 2010) that flank the F-gene and contain about 20 nucleotides overlap using primer

359

pairs 1-2 and 3-4 (Table S1). The two purified fragments were subsequently used as

360

template in a fusion PCR reaction based on primers 1 and 4. The resulting fragment

361

was digested with HpaI and SpeI and ligated to pNDFL-GnGc digested with the same

362

enzymes. Plasmids were confirmed to have the expected sequences using Sanger

363

sequencing. The construct was designed to comply with the rule-of-6 (Peeters et al.,

364

2000).

365 366

IPMA

367

Immunoperoxidase monolayer assays (IPMAs) were performed as previously described

368

with some minor modifications (de Boer et al., 2010). Briefly, infected cell monolayers

369

were fixed with 4% (w/v) paraformaldehyde for 15 min, washed with PBS and

370

permeabilized with 1% Triton in PBS (5 min). After washing with washing buffer (PBS,

371

0.05% v/v Tween-20), cells were incubated for 30 min with blocking buffer (PBS, 5%

372

horse serum). Subsequently, primary antibody incubations were performed for 1 h at

373

37°C in blocking buffer. Expression of F protein was detected with monoclonal antibody

374

8E12A8C3 (de Leeuw et al., 2005), of Gn protein with monoclonal antibody 4-39-cc

375

(Keegan & Collett, 1986) and of Gc protein with a polyclonal rabbit antiserum against

376

the Gc ectodomain (de Boer et al., 2012). Secondary, HRP-conjugated antibody

377

incubations (DAKO) were also performed in blocking buffer for 1 h at 37°C. Between

378

incubations cells were washed 3 times with washing buffer. Antibody binding was

379

visualized using AEC substrate.

380 381

Next-generation sequencing

382

A volume of 6 ml cell culture supernatant of NSRNDV passage 7 was concentrated to 125

383

µl using an Amicon Ultra 30K centrifugal filter (Merck, Darmstadt, Germany).

384

Subsequently, 400 µl Trizol LS (Sigma-Aldrich, Missouri, United States) was added to

385

the concentrate and RNA was isolated using the Direct-zolTM RNA Miniprep kit (Zymo

386

research, California, United States) according the manufacturers’ instructions. Libraries

387

were prepared using the ScriptSeqTMv2 RNA-Seq Library Preparation kit (Epicentre

388

Biotechnologies, Wisconsin, United States). Quality and quantity of the library were

389

determined using the Bioanalyzer with the High Sensitivity DNA kit (Agilent

390

Technologies, California, United States) and the Qubit dsDNA HS Assay kit (Life

391

Technologies, New York, United States).

392

sequencing of the libraries was subsequently performed with an Illumina MiSeq V2

393

instrument. Raw sequencing data were trimmed and subjected to quality control using

394

in-house scripts. Virus-specific reads were mapped against the reference sequences using

395

Bowtie2 (Langmead & Salzberg, 2012) and local-sensitive mapped reads were visualized

396

using Tablet (Milne et al., 2013).

Cluster generation and paired-end 250 bp

397 398

Site-directed mutagenesis of NDV-GnGc-∆F

399

Site-directed mutants of NDV-GnGc-∆F were constructed using the Quickchange II XL

400

Site-Directed Mutagenesis kit (Agilent Technologies, Amstelveen, Netherlands) in

401

combination with standard cloning techniques using pNDFL-GnGc-∆F as a template.

402

Used primers (7-10) are listed in Table S1. Site-directed mutant infectious clones were

403

confirmed to have the expected sequence using Sanger sequencing.

404 405

In vitro neutralization of NDV-GnGc variants

406

Virus (MOI 0.05) was incubated with serial dilutions of a convalescent sheep serum

407

containing RVFV antibodies or a negative control serum for 1.5 h in complete GMEM

408

medium and subsequently added to 200.000 BHK cells per well of a 24 wells plate. Two

409

days post inoculation, infected cells were visualized using IPMA.

410 411

Vaccination and challenge of mice

412

Six-week-old female BALB/cAnCrl mice (Charles River Laboratories) were divided in 4

413

groups of 10 mice, kept in type III filter top cages under BSL-3 conditions, and allowed to

414

acclimatize for 6 days. At day 0, mice were vaccinated intramuscularly (thigh muscle)

415

with either medium (Mock), NSR-Gn (Oreshkova et al., 2013) (105.8 TCID50), NSRNDV

416

(NSR: 105.8 TCID50; NDV: 102.8 TCID50) or NDV-GnGc-∆F (102.8 TCID50) in 50 µl. Mice

417

were observed daily and three weeks post vaccination mice were challenged

418

intraperitoneally with 102.8 TCID50 of RVFV strain 35/74 in 0.1 ml medium. RVFV-

419

specific neutralization titers in sera were determined one day prior challenge, using a

420

previously developed RVFV neutralization assay (Kortekaas et al., 2011). Virus

421

dissemination to the livers was evaluated by qRT-PCR as described (Kortekaas et al.,

422

2012). Briefly, RNA was isolated of 200 µl 10% w/v organ suspension using the RNeasy

423

96 Kit (QIAGEN) according the manufacturers’ instructions. A volume of 5 µl purified

424

RNA was used as PCR template and RVFV specific RNA copy numbers in the 5 µl RNA

425

sample were determined using an in vitro synthesised RNA standard.

426

Acknowledgements

427

We like to thank Olav de Leeuw (CVI) for providing plasmids pCIneo-NP, pCIneo-P,

428

pCIneo-L and pCIneo-FHerts, Geoff Oldham (Institute for Animal Health, Compton,

429

United Kingdom) for providing FP-T7, Connie Schmaljohn (USAMRIID, Fort Detrick,

430

MD) for providing the Gn monoclonal antibody (4-39-cc) and Jet Kant (CVI) for technical

431

assistance. The project was performed within the Castellum program, which is

432

financially supported by the Dutch Ministry of Economic Affairs.

433

References

434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484

Barnard, B. J. (1979). Rift Valley fever vaccine--antibody and immune response in cattle to a live and an inactivated vaccine. Journal of the South African Veterinary Association 50, 155-157. Billecocq, A., Spiegel, M., Vialat, P., Kohl, A., Weber, F., Bouloy, M. & Haller, O. (2004). NSs protein of Rift Valley fever virus blocks interferon production by inhibiting host gene transcription. Journal of virology 78, 9798-9806. Bouloy, M., Janzen, C., Vialat, P., Khun, H., Pavlovic, J., Huerre, M. & Haller, O. (2001). Genetic evidence for an interferon-antagonistic function of rift valley fever virus nonstructural protein NSs. Journal of virology 75, 1371-1377. Chang, A. & Dutch, R. E. (2012). Paramyxovirus fusion and entry: multiple paths to a common end. Viruses 4, 613-636. Chaudhry, Y., Skinner, M. A. & Goodfellow, I. G. (2007). Recovery of genetically defined murine norovirus in tissue culture by using a fowlpox virus expressing T7 RNA polymerase. The Journal of general virology 88, 2091-2100. de Boer, S. M., Kortekaas, J., Spel, L., Rottier, P. J., Moormann, R. J. & Bosch, B. J. (2012). Acid-activated structural reorganization of the rift valley Fever virus gc fusion protein. Journal of virology 86, 13642-13652. de Boer, S. M., Kortekaas, J., Antonis, A. F., Kant, J., van Oploo, J. L., Rottier, P. J., Moormann, R. J. & Bosch, B. J. (2010). Rift Valley fever virus subunit vaccines confer complete protection against a lethal virus challenge. Vaccine 28, 2330-2339. de Leeuw, O. S., Koch, G., Hartog, L., Ravenshorst, N. & Peeters, B. P. (2005). Virulence of Newcastle disease virus is determined by the cleavage site of the fusion protein and by both the stem region and globular head of the haemagglutinin-neuraminidase protein. The Journal of general virology 86, 17591769. Dodd, K. A., Bird, B. H., Metcalfe, M. G., Nichol, S. T. & Albarino, C. G. (2012). Single-dose immunization with virus replicon particles confers rapid robust protection against Rift Valley fever virus challenge. Journal of virology 86, 42044212. Dortmans, J. C., Koch, G., Rottier, P. J. & Peeters, B. P. (2011). Virulence of Newcastle disease virus: what is known so far? Veterinary research 42, 122. Elliott, R. M. (1990). Molecular biology of the Bunyaviridae. The Journal of general virology 71 ( Pt 3), 501-522. Elliott, R. M. (1997). Emerging viruses: the Bunyaviridae. Molecular medicine 3, 572577. Estevez, C., King, D. J., Luo, M. & Yu, Q. (2011). A single amino acid substitution in the haemagglutinin-neuraminidase protein of Newcastle disease virus results in increased fusion promotion and decreased neuraminidase activities without changes in virus pathotype. The Journal of general virology 92, 544-551. Goldhaft, T. M. (1980). Historical note on the origin of the LaSota strain of Newcastle disease virus. Avian diseases 24, 297-301. Ikegami, T., Narayanan, K., Won, S., Kamitani, W., Peters, C. J. & Makino, S. (2009a). Rift Valley fever virus NSs protein promotes post-transcriptional downregulation of protein kinase PKR and inhibits eIF2alpha phosphorylation. PLoS pathogens 5, e1000287. Ikegami, T., Narayanan, K., Won, S., Kamitani, W., Peters, C. J. & Makino, S. (2009b). Dual functions of Rift Valley fever virus NSs protein: inhibition of host mRNA transcription and post-transcriptional downregulation of protein kinase PKR. Annals of the New York Academy of Sciences 1171 Suppl 1, E75-85.

485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535

Kakach, L. T., Wasmoen, T. L. & Collett, M. S. (1988). Rift Valley fever virus M segment: use of recombinant vaccinia viruses to study Phlebovirus gene expression. Journal of virology 62, 826-833. Keegan, K. & Collett, M. S. (1986). Use of bacterial expression cloning to define the amino acid sequences of antigenic determinants on the G2 glycoprotein of Rift Valley fever virus. Journal of virology 58, 263-270. Kortekaas, J., de Boer, S. M., Kant, J., Vloet, R. P., Antonis, A. F. & Moormann, R. J. (2010). Rift Valley fever virus immunity provided by a paramyxovirus vaccine vector. Vaccine 28, 4394-4401. Kortekaas, J., Oreshkova, N., Cobos-Jimenez, V., Vloet, R. P., Potgieter, C. A. & Moormann, R. J. (2011). Creation of a nonspreading Rift Valley fever virus. Journal of virology 85, 12622-12630. Kortekaas, J., Antonis, A. F., Kant, J., Vloet, R. P., Vogel, A., Oreshkova, N., de Boer, S. M., Bosch, B. J. & Moormann, R. J. (2012). Efficacy of three candidate Rift Valley fever vaccines in sheep. Vaccine 30, 3423-3429. Langmead, B. & Salzberg, S. L. (2012). Fast gapped-read alignment with Bowtie 2. Nat Methods 9, 357-U354. Malykhina, O., Yednak, M. A., Collins, P. L., Olivo, P. D. & Peeples, M. E. (2011). A respiratory syncytial virus replicon that is noncytotoxic and capable of longterm foreign gene expression. Journal of virology 85, 4792-4801. McGinnes, L. W., Gravel, K. & Morrison, T. G. (2002). Newcastle disease virus HN protein alters the conformation of the F protein at cell surfaces. Journal of virology 76, 12622-12633. Milne, I., Stephen, G., Bayer, M., Cock, P. J. A., Pritchard, L., Cardle, L., Shaw, P. D. & Marshall, D. (2013). Using Tablet for visual exploration of secondgeneration sequencing data. Briefings in bioinformatics 14, 193-202. Muller, R., Saluzzo, J. F., Lopez, N., Dreier, T., Turell, M., Smith, J. & Bouloy, M. (1995). Characterization of clone 13, a naturally attenuated avirulent isolate of Rift Valley fever virus, which is altered in the small segment. The American journal of tropical medicine and hygiene 53, 405-411. Oreshkova, N., van Keulen, L., Kant, J., Moormann, R. J. & Kortekaas, J. (2013). A single vaccination with an improved nonspreading rift valley Fever virus vaccine provides sterile immunity in lambs. PloS one 8, e77461. Oreshkova, N., Cornelissen, L. A., de Haan, C. A., Moormann, R. J. & Kortekaas, J. (2014). Evaluation of nonspreading Rift Valley fever virus as a vaccine vector using influenza virus hemagglutinin as a model antigen. Vaccine. Peeters, B. P., de Leeuw, O. S., Koch, G. & Gielkens, A. L. (1999). Rescue of Newcastle disease virus from cloned cDNA: evidence that cleavability of the fusion protein is a major determinant for virulence. Journal of virology 73, 50015009. Peeters, B. P., Gruijthuijsen, Y. K., de Leeuw, O. S. & Gielkens, A. L. (2000). Genome replication of Newcastle disease virus: involvement of the rule-of-six. Archives of virology 145, 1829-1845. Porotto, M., Salah, Z. W., Gui, L., DeVito, I., Jurgens, E. M., Lu, H., Yokoyama, C. C., Palermo, L. M., Lee, K. K. & other authors (2012). Regulation of paramyxovirus fusion activation: the hemagglutinin-neuraminidase protein stabilizes the fusion protein in a pretriggered state. Journal of virology 86, 1283812848. Romer-Oberdorfer, A., Mundt, E., Mebatsion, T., Buchholz, U. J. & Mettenleiter, T. C. (1999). Generation of recombinant lentogenic Newcastle disease virus from cDNA. The Journal of general virology 80 ( Pt 11), 2987-2995.

536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553

Sergel, T. A., McGinnes, L. W. & Morrison, T. G. (2000). A single amino acid change in the Newcastle disease virus fusion protein alters the requirement for HN protein in fusion. Journal of virology 74, 5101-5107. Terasaki, K., Won, S. & Makino, S. (2013). The C-terminal region of Rift Valley fever virus NSm protein targets the protein to the mitochondrial outer membrane and exerts antiapoptotic function. Journal of virology 87, 676-682. Weingartl, H. M., Zhang, S., Marszal, P., McGreevy, A., Burton, L. & Wilson, W. C. (2014). Rift Valley Fever Virus Incorporates the 78 kDa Glycoprotein into Virions Matured in Mosquito C6/36 Cells. PloS one 9, e87385. Won, S., Ikegami, T., Peters, C. J. & Makino, S. (2007). NSm protein of Rift Valley fever virus suppresses virus-induced apoptosis. Journal of virology 81, 1333513345. Yoshizaki, M., Hironaka, T., Iwasaki, H., Ban, H., Tokusumi, Y., Iida, A., Nagai, Y., Hasegawa, M. & Inoue, M. (2006). Naked Sendai virus vector lacking all of the envelope-related genes: reduced cytopathogenicity and immunogenicity. The journal of gene medicine 8, 1151-1159.

554 Table 1. Genetic changes in NSRNDV detected by next‐generation sequencing              Segment  SNP  Frequency  Region  RVFV‐SeGFP  T10C  50%     C11A  100%     A1604T  100%  RVFV‐L  G1181A  100%     A1602C  100%     T6325C  100%  NDV‐GnGc‐∆F  A4085C  100%     A6951C  90%     A8367G  95%     T14476C  90%           Passage 7 culture supernatant is used as RNA template (Fig. 2).   SNPs present in ≥ 50% of sequence reads are shown   * Counted from the 4th methionine of the M‐segment  UTR = untranslated region, Pol. = polymerase 

3'UTR  3'UTR  5'UTR  Pol. gene  Pol. gene  5'UTR  Gn gene  M gene  HN gene  L gene    

   AA change   ‐   ‐   ‐  G388E   ‐   ‐  Y303S*  K155T  I192M   ‐    

  

555 556 557

Figure legends

558 559

Figure 1. Transfection and infection-mediated NSR production. BHK-Rep2 cells

560

stably maintaining RVFV L and SeGFP genome segments were transfected with pCAGGS-

561

M or infected with NDV-GnGc at an MOI of 0.5. One day post transfection or infection,

562

supernatants were collected and titrated on BHK-21 cells to determine the level of NSR

563

progeny. TCID50 values were determined two days post infection by evaluation of eGFP

564

expression. Bars represent means and standard errors of three experiments.

565 566

Figure 2. Construction of NDV-GnGc-∆F and an F complementing cell line. A)

567

Schematic presentation of the NDV-GnGc and NDV-GnGc-∆F genome. B) QM5 cells

568

infected with NDV-GnGc and NDV-GnGc-∆F were stained for F, Gn and Gc expression

569

using IPMA one day post infection. C) IPMA of QM5 and QM5-FHerts cells using an α F-

570

monoclonal antibody. D) Confirmation of F deletion in NDV-GnGc-∆F genomic RNA

571

isolated from virus grown on QM5-FHerts cells and determined by conventional RT-PCR

572

using primers 5 and 6 (Table S1). E) DAPI staining of QM5 and QM5-FHerts cells infected

573

with NDV-GnGc-∆F one day post infection. Foci of syncytia forming cells are magnified.

574

F) Production of NSR progeny after infection of BHK-Rep2 cells at different MOIs with

575

NDV-GnGc or NDV-GnGc-∆F. One day post infection, supernatants were collected and

576

titrated on BHK-21 cells. TCID50 values were determined two days post infection by

577

evaluation of eGFP expression.

578 579

Figure 3. Adaptation of NDV-GnGc-∆F to BHK-Rep2 cells. A) BHK-Rep2 cells

580

stably maintaining RVFV L and SeGFP genome segments were infected with NDV-GnGc-

581

∆F at an MOI of 0.5. After reaching confluence, the cells were passaged to new plates

582

until, due to the presence of CPE, no cell passage was possible anymore. Starting from

583

passage 3, supernatants were passaged to fresh BHK-21 cells instead of sub-culturing

584

the BHK-Rep2 cells. NSR progeny and infectious NDV-GnGc-∆F present in the

585

supernatants was determined for each passage. The three pictures (eGFP expression,

586

phase contrast and Gn IPMA) represent the observed phenotype of BHK cells infected

587

with NSRNDV.

588 589

Figure 4. Construction and evaluation of NDV-GnGc-∆F site-directed mutant

590

mediated NSR production. A) Schematic presentation of the genomes of NDV-GnGc-

591

∆F, NDV-Gn*Gc-∆F, NDV-GnGc-∆F-HN* and NDV-Gn*Gc-∆F-HN*. B) Growth curve of

592

the site-directed mutants in QM5-FHerts cells. C) DAPI staining of QM5 and QM5-FHerts

593

cells and IPMA of BSR cells infected with the site-directed mutants. The mutants grown

594

on BSR cells were also evaluated for the presence of infectious progeny by testing

595

supernatants of infected BSR cells on fresh BSR cells. D) Confirmation of F deletion in

596

the NDV-GnGc-∆F site directed mutants genomic RNA isolated from virus grown on

597

QM5-FHerts cells and determined by conventional RT-PCR using primers 5 and 6 (Table

598

S1). E) Site-directed mutant-mediated NSR production. BHK-Rep2 cells were infected

599

with the site-directed mutants and 2 days post infection the level of NSR progeny was

600

determined in the supernatant (P0, first bar). The supernatant of P0 was subsequently

601

passaged to fresh BHK-21 cells. These cells were washed twice at 6 h post infection, and

602

at 4 days post infection the level of NSR progeny was determined (P1, second bar). Bars

603

represent means and standard errors of three experiments. F) Stability of NDV-Gn*Gc-

604

∆F-HN*-mediated NSR production. BSR cells were infected with NDV-Gn*Gc-∆F-HN*

605

and NSR and each 2-3 days post infection supernatants were passaged to fresh cells.

606

Levels of progeny NSR and NDV-Gn*Gc-∆F-HN* were determined after each passage by

607

titration using eGFP as a readout for NSR and CPE as a readout for NDV-Gn*Gc-∆F-

608

HN*. At the passages indicated with an asterisk (*) an additional cell passage was

609

performed.

610 611

Figure 5. Neutralization of NDV-GnGc, NDV-GnGc-∆F and NDV-Gn*Gc-∆F-HN*.

612

NDV-GnGc, NDV-GnGc-∆F and NDV-Gn*Gc-∆F-HN* were incubated with serial

613

dilutions of normal sheep serum or a sheep serum containing α-RVFV antibodies. After

614

1.5 h the virus-serum mixture was added to BHK-21 cells. Two days later cells were

615

fixed and evaluated for Gn expression using IPMA. Pictures (A) were taken from wells

616

containing the 1:40 serum dilutions and (B) neutralization titers are expressed as the

617

serum dilution able to block over 50% of infectious particles. The detection limit was a

618

serum dilution of 1:20.

619 620

Figure 6. Vaccination challenge experiment NSRNDV in mice. A) Survival curve of

621

mice vaccinated with either NSR, NSRNDV or NDV-GnGc-∆F (NDV). Three weeks post

622

vaccination mice were challenged with a lethal RVFV challenge dose. B) RVFV-specific

623

virus neutralization titers present in sera the day before challenge. C,D) RVFV

624

dissemination into the livers (C) and brains (D) of challenged mice determined by qRT-

625

PCR.

626 627

Figure 7. Model showing replication and processing of wild-type NDV (LaSota),

628

wild-type RVFV, NDV-GnGc-∆F and NDV-Gn*Gc-∆F-HN* virions. NDV enters

629

cells via fusion with the plasma membrane and progeny is formed by budding at the

630

plasma membrane. Due to the absence of trypsin like proteases the F protein is not

631

cleaved and progeny is not infectious. RVFV enters cells via the endosomal route and

632

buds from the trans-Golgi compartments. Resulting progeny is infectious. NDV-Gn*Gc-

633

∆F, previously grown on F complementing cells, enters cells via fusion with the plasma

634

membrane. Progeny virus is not infectious due to the absence of the F protein. Unknown

635

amounts of GnGc heterodimers are expected to be present in the viral envelope. NDV-

636

Gn*Gc-∆F-HN* virions follow the RVFV specific entry route, via endosomes, where after

637

infectious progeny is formed at the plasma membrane due to the presence of mutations

638

in HN and Gn.

639