JGV Papers in Press. Published May 16, 2014 as doi:10.1099/vir.0.064956-0

1

Development of an antemortem diagnostic test for enzootic nasal tumour

2

virus (ENTV-1) and detection of neutralizing antibodies in host serum

3

Scott R Walsh1, Kevin J. Stinson1, Paula I. Menzies2, Sarah K Wootton1*

4 5

1

6

Ontario, Canada

7

2

8

Guelph, Ontario, Canada

Department of Pathobiology, Ontario Veterinary College, University of Guelph, Guelph,

Department of Population Medicine, Ontario Veterinary College, University of Guelph,

9 10

Running title: Development of an antemortem diagnostic test for ENTV-1

11

SRW: [email protected]

12

KJS: [email protected]

13

PIM: [email protected]

14

SKW: [email protected]

15 16

6 Figures, 3 Tables

17

Summary: 250 words

18

Main text: 5494 words (including the figure legends)

19 20

For publication as a Standard ('full-length') paper

*

Corresponding author

ENA – enzootic nasal adenocarcinoma, ENTV – enzootic nasal tumour virus, PBMC – peripheral blood mononuclear cell, JSRV - Jaagsiekte sheep retrovirus, MLV – murine leukemia virus, ECa – ENTV-1 capsid protein 1

21

Summary

22

Enzootic nasal adenocarcinoma (ENA) is a contagious neoplasm of the nasal mucosa of sheep

23

and goats and is associated with enzootic nasal tumour virus (ENTV). Since ENA is a common

24

disease in North America and there are no vaccines against ENTV-1, diagnostic tests that can

25

identify infected animals and assist with eradication and disease surveillance efforts are greatly

26

needed. In this study, we endeavoured to develop a novel, non-invasive diagnostic tool that

27

could be used to not only validate clinical signs of ENA, but also to detect ENTV-1 infection

28

prior to the onset of disease signs (i.e. preclinical diagnosis). Cytology, serology and RT-PCR-

29

based diagnostic methods were investigated. Although the cytology-based assay was able to

30

detect ENTV-1 infection in some animals, it had poor sensitivity and specificity and thus was not

31

developed further as an antemortem diagnostic method. Three different assays, including

32

ELISA, western blot and virus neutralization were developed to detect the presence of ENTV-1

33

specific antibodies in sheep serum. While a surprisingly large number of sheep mounted an

34

antibody-mediated immune response against ENTV-1, and in some cases neutralizing,

35

correlation with disease status was poor. In contrast, RT-PCR on RNA extracted from nasal

36

swabs reliably detected exogenous ENTV-1 sequences, did not amplify endogenous ovine

37

betaretroviral sequences, demonstrated high concordance with immunohistochemical staining for

38

ENTV-1 envelope protein and had perfect sensitivity and specificity. This report describes a

39

practical and highly specific RT-PCR technique for the detection of clinical and preclinical ENA

40

that may prove beneficial in future control or eradication programs.

41

2

42

Introduction

43

Enzootic nasal adenocarcinoma (ENA) is an economically important contagious tumour of the

44

nasal mucosa of sheep and goats (De Las Heras et al., 2003). With the exception of Australia

45

and New Zealand, ENA has been recorded worldwide wherever sheep and goats are farmed with

46

a prevalence of up to 10% in some areas (De Las Heras et al., 2003). The true economic impact

47

of ENA is not known because affected animals are often culled before actual diagnosis, and

48

suspected disease incidence is rarely reported. Adenocarcinomas of the nasal cavity in sheep

49

have the highest reported prevalence in the United States, Canada, France, Germany, and Spain

50

(Caswell & Williams, 2007). Although the exact prevalence of ENA is unknown, it is not an

51

uncommon disease in North America (Walsh et al., 2010) and is very likely under-diagnosed.

52

Clinical signs of ENA include seromucosal nasal discharge leading to a “washed nose”

53

appearance, accompanied by snoring, coughing, wheezing and dyspnea. The duration of disease,

54

from the appearance of clinical signs to the time of death, varies from three weeks to one year or

55

more. ENA has been experimentally transmitted to goats using concentrated nasal fluid, proving

56

the infectious nature of this disease (De Las Heras et al., 1995). We recently conducted

57

transmission studies in newborn lambs and demonstrated the transmission of ENA using cell-free

58

tumour homogenate (Walsh et al., 2013).

59

The retrovirus, enzootic nasal tumour virus (ENTV), has been implicated in the etiology of this

60

lethal and contagious nasal tumour (Cousens et al., 1996, 1999). Using a mouse monoclonal

61

antibody against the envelope glycoprotein (Env) of ovine betaretroviruses (Wootton et al.,

62

2006a), of which ENTV is a member, we have been able to verify the presence of the virus in all

63

nasal tumour samples evaluated to date (Walsh et al., 2010). This monoclonal antibody is

64

exquisitely specific for ENTV-1 Env in immunohistochemical staining making it an ideal tool

3

65

for diagnostic purposes; however, acquisition of nasal biopsies is costly, therefore this tool has

66

been restricted to post-mortem use.

67

Currently, diagnosis of ENA depends on antemortem clinical signs, and histological findings

68

following post-mortem analysis. Serology has not been used for diagnostic purposes since there

69

is a reported lack of antibody production to the virus capsid in animals with ENA (Ortín et al.,

70

1998). This lack of immune response is postulated to be due to expression of endogenous ovine

71

betaretrovirus proteins in the thymus and lymphoid tissues during ontogeny (DeMartini et al.,

72

2003; Palmarini et al., 2000) and subsequent elimination of reactive B cells during negative

73

selection. Additionally, our inability to detect ENTV provirus in the peripheral blood

74

mononuclear cells (PBMCs) of either experimentally or naturally infected sheep at any point

75

during the course of infection rules out the use of whole blood for diagnostic purposes (Walsh et

76

al., 2013). Computed tomography scanning can detect tumours in the nose (Walsh et al., 2013)

77

of ENTV infected sheep, but this procedure is expensive, and is stressful for the animal as it

78

requires transport and anaesthesia. Also, detection of a tumour in these anatomical locations does

79

not conclusively implicate the involvement of ovine betaretroviruses. Therefore, the purpose of

80

this report was to evaluate the utility of RT-PCR and immunohistochemical analysis of materials

81

obtained from nasal swabs or nasal fluid for diagnosis of ENTV infection in sheep. With this

82

information in hand, it was envisaged that a combination RT-PCR/cytospin assay could be

83

implemented in a test-and-removal program aimed at eradicating ENTV from sheep flocks.

84

Results

85

Pilot study to evaluate the efficacy of a RT-PCR based assay to detect ENTV-1 in nasal

86

exudates

4

87

Efforts to develop an antemortem diagnostic test were initiated when we received 15 nasal

88

exudate samples from a research flock of Suffolk rams in Quebec, Canada. All of these rams

89

had, at some point within the last two years, displayed clinical signs of ENA, including nasal

90

discharge and noisy breathing. Samples were collected in Trizol LS (2:3 ratio of sample to

91

Trizol LS) to preserve RNA. Based on sequence information that we recently obtained from ten

92

North American ENTV-1 isolates (Walsh et al., 2010), primers were designed against the U5

93

and gag regions to specifically detect exogenous ENTV-1 (Fig. 1(a)). RNA extracted from

94

healthy sheep lung (a negative control) and naturally acquired ENA tissue (a positive control)

95

were screened for the presence of ENTV-1 genomic RNA by RT-PCR. In this RT-PCR assay,

96

an ENTV-1 positive sample produced a PCR product of approximately 592 bp (see positive

97

control Fig. 1(b), lane 15). No amplification was detected in the negative control (Fig. 1(b), lane

98

17) confirming the specificity of these primers for exogenous ENTV-1. After confirming the

99

ability of the RT-PCR assay to selectively amplify exogenous ENTV-1 by sequencing of the

100

product from QC-12 (data not shown), RNA was extracted from all 15 nasal exudate samples

101

and subjected to cDNA synthesis and RT-PCR for both GAPDH, as a control for RNA quality,

102

and ENTV-1 (Table 1). The RT-PCR results for a subset of animals are shown in Fig. 1(b). Of

103

the 15 samples submitted for RT-PCR analysis, four were positive for ENTV-1 (Table 1).

104

Three RT-PCR-positive and two RT-PCR-negative rams from the putative ENA cluster in

105

Quebec were euthanized and submitted for post mortem analysis. Upon gross inspection, the

106

three RT-PCR positive rams had visible tumours while the nasal cavity of the two RT-PCR

107

negative rams looked unremarkable. Tissues harvested from the nasal cavities were subjected to

108

histology and immunohistochemical staining with an ENTV-1 envelope specific monoclonal

109

antibody. ENTV-1 envelope expression and histopathological features consistent with ENA

5

110

were detected in all three RT-PCR/tumour-positive cases (QC-1, QC-12 and QC-15) (Table 1).

111

For one of the RT-PCR-negative cases (QC-6), a small ENTV-1 envelope positive lesion was

112

identified in the nose by immunohistochemistry (IHC). The other RT-PCR-negative sample

113

appeared normal upon histological and immunohistochemical examination (QC-7) (Table 1).

114

Results from the pilot study revealed a reasonable correlation between RT-PCR positivity and a

115

positive result by IHC for envelope protein expression, which is the current gold standard test.

116

Lack of sample availability limited our ability to test the remaining cases by IHC and therefore

117

limited our ability to analyse the specificity and sensitivity of the RT-PCR test on this cohort of

118

animals.

119

Study to evaluate serology, cytology and RT-PCR based assays for antemortem detection of

120

ENTV-1

121

Given the promising results of the pilot study we decided to launch a full-scale study involving a

122

flock of 80 Horned Dorset sheep from Ontario with a history of ENA. Additional control

123

samples were obtained from ten sheep chosen at random from a research flock at the University

124

of Guelph with no history of ENA and which has been closed for 25 years to all new animal

125

additions. In this study, we expanded our sample repertoire to include nasal swabs, cytology

126

samples and serum. Two nasal swabs were collected per animal. One swab was placed in RLT

127

buffer for subsequent RNA extraction and the other was applied to a charged microscope slide

128

for cytological examination.

129

Gross and histopathological detection of ENA showed good correlation with RT-PCR

130

results

131

Total RNA was extracted from nasal swabs and subjected to RT-PCR as described above. Based

132

on the RT-PCR results (Table 2), seven ewes that were RT-PCR-negative for ENTV-1 infection,

6

133

14 ewes that were RT-PCR-positive for ENTV-1 infection, and one lamb (2013-1) of

134

undetermined status but born to a mother who was RT-PCR-positive, were submitted for post-

135

mortem analysis. As shown in Fig. 2, tumours of various sizes, ranging from 0.5 cm to 8 cm,

136

were detected upon dissection of the nasal cavity. All tumours stained positive for ENTV-1

137

envelope expression by IHC (Table 2 and Fig. 3), whereas none of the sections from RT-PCR-

138

negative animals did (Table 2). Importantly, all of the animals that were RT-PCR-positive for

139

ENTV-1 had tumours whereas none of the animals that were RT-PCR-negative for ENTV-1 had

140

tumours. No evidence of tumours or envelope-positive cells was detected in the lamb born to a

141

mother who was RT-PCR-positive for ENTV-1 (Table 2). The specificity and sensitivity of the

142

RT-PCR test was perfect [100% for both (Table 3)] when the immunohistochemical staining for

143

envelope protein was used as the gold standard test for calculation purposes. Taken together,

144

these findings suggest that RT-PCR analysis of RNA extracted from nasal swabs represents a

145

sensitive and specific method for detection of ENA.

146

Immunohistochemical staining of nasal smears for detection of ENTV-1

147

To evaluate the use of nasal swab cytology as a method to diagnose ENA, slides from 80

148

exposed sheep and 10 unexposed controls were immunocytochemically stained for ENTV-1

149

envelope protein and blindly graded by three individuals. Slides were graded as (~) for no

150

staining, (*) for low intensity, (**) for moderate intensity staining in >10% of cells or high

151

intensity staining in 10% of cells (Fig. 4

152

(a, b, c and d), respectively). Isotype control staining was performed on a slide from a ENA

153

tumor bearing animal and no staining was observed (data not shown). Although grading results

154

were consistent for some samples (e.g. 2008-1, 2012-12, 2006-37, 2010-44 and 2011-50), overall

155

there was poor inter-grader agreement (Table 2). In addition, the quantity and quality of cells on

7

156

the slide varied from sample to sample as well as between different regions of the same slide

157

and, in a subset of cases, poor sample representation made grading difficult. Taken together,

158

these results suggest that ENTV-1-infected cells can be identified by immunocytochemical

159

staining of nasal smears, but that sample quality would need to be improved to allow appropriate

160

and consistent interpretation. The specificity and sensitivity, which was calculated using IHC as

161

the gold standard, was relatively low compared to the RT-PCR test with 66.7% and 50%

162

respectively (Table 3). These values were calculated from absolute positive/negative values

163

derived from a consensus of the results from all three graders and any result with no clear

164

consensus value was not included (Table 2).

165

Detection of ENTV-1-reactive sheep serum by ELISA and western blot analysis

166

To investigate whether sheep mount an antibody response to ENTV-1, serum samples were

167

tested for reactivity against recombinant ENTV-1 capsid (ECa) and envelope surface subunit

168

(ESU-IgG) proteins using an indirect ELISA (Fig. 5(a) and (b), respectively). Serum samples

169

from the previously described closed research flock with no history of ENA (Fig. 5, group i)

170

were used to determine the background cut-off value, which was calculated as the mean of the

171

naïve samples plus three times the standard deviation of those samples. Of the eight sheep that

172

had no lesions of ENA by IHC (Fig.5, group iii), three (37.5 %) tested positive by the ECa

173

ELISA and five (62.5%) were positive by the ESU-IgG ELISA. In the group of animals

174

confirmed to have ENA by post-mortem analysis (Fig. 5, group ii), 3/13 (23.1%) samples were

175

positive in the ECa ELISA and 5/13 (38.5%) samples were positive in the ESU-IgG ELISA.

176

Only four animals (2006-37, 2011-50, 2001-16 and 2008-1) had serum that reacted positively in

177

both the ESU-IgG and ECa ELISAs. Of the remaining 55 flock mates for which no post-mortem

178

data were available (Fig. 5, group iv), 8/55 and 15/55 serum samples were positive by ECa and

8

179

ESU-IgG ELISA, respectively. Three of those serum samples (2011-36, 2007-28 and 2011-23)

180

from group iv were positive on both ELISAs. Both ELISA-based tests show poor agreement with

181

immunohistochemical staining for envelope protein as the specificity and sensitivity of the ECa

182

ELISA was 62.5% and 23.1%, respectively and the ESU-IgG ELISA was 33.3% and 50.0%,

183

respectively (Table 3). Thus, the ELISA tests would not be suitable for diagnostic purposes.

184

Sera from animals submitted for postmortem analysis were analyzed by strip-western blotting to

185

investigate reactivity against denatured ENTV-1 antigens and correlation with ELISA results.

186

Recombinant ECa and ESU-IgG proteins were used as antigens and polyclonal rabbit anti-

187

ENTV-1 capsid antibody (rαECa) and monoclonal anti-envelope antibody (Wootton et al.,

188

2006a) were used as positive controls, respectively (Fig. 6, lane C). The majority of serum

189

samples (15/21), including many of those from the control naïve animals (Fig. 6), detected a

190

band by western blot corresponding to ECa (Table 2). Given that there was very little difference

191

in ECa immunoreactivity between serum from exposed and naïve animals, it is likely that serum

192

antibodies are binding non-specifically to ENTV-1 ECa. With a Kappa value was -0.034,

193

agreement between the ECa ELISA and ECa western blot was poor and was no different than

194

what you would expect by chance alone. Conversely, none of the naïve sheep serum samples

195

detected ESU-IgG in western blot (Fig. 6). With 6/10 ESU-IgG ELISA positive samples

196

reacting positively in western blot and only 1/11 ESU-IgG ELISA negative samples reacting

197

positively on western blot, there was moderate agreement between the ESU-IgG ELISA and the

198

ESU-IgG western blot (Kappa value of 0.516). Taken together, it would appear that sheep are

199

able to mount an antibody-mediated immune response specifically against exogenous ENTV-1

200

envelope protein.

201

Serum neutralization of ENTV-1 envelope-pseudotyped virions

9

202

Heat-inactivated serum samples from animals submitted for post-mortem analysis, and thus with

203

known disease status, were evaluated for their ability to inhibit infection mediated by ENTV-1

204

envelope pseudotyped MLV particles. Neutralization was defined as a 50% or greater reduction

205

in infection (as determined by the number of alkaline phosphatase positive foci) relative to

206

untreated virus. Neutralization was observed in 7/13 (53.8%) samples from the IHC positive

207

group and 3/8 (37.5%) samples from the IHC negative group (Table 2). Serum from naïve sheep

208

was also tested and no neutralization was observed (data not shown). Despite the fact that the

209

ESU-IgG ELISA and the neutralization assays both detect antibodies specific for the ENTV-1

210

envelope protein, the calculated Cohen's kappa coefficient was 0.427 demonstrating only

211

moderate agreement between the two tests.

212

Discussion

213

ENA is a common disease among North American sheep flocks (Walsh et al., 2010) and as

214

producer and veterinarian awareness increases, so too does the need for an accurate antemortem

215

diagnostic test. Based on our previous findings, PBMCs are not an appropriate source material

216

for detecting ENTV-1 since at no point during the course of infection can ENTV-1 be detected in

217

PBMCs by PCR (Walsh et al., 2013). In this study, we endeavoured to develop a novel, non-

218

invasive diagnostic tool that could be used to not only confirm clinical signs of ENA, but also to

219

detect ENTV-1 infection prior to the onset of disease (i.e. preclinical diagnosis). The latter

220

would assist in surveillance programs and eradication efforts to quarantine and/or eliminate

221

ENTV-1 infected animals from a flock, which would ultimately enhance the health and

222

profitability of the sheep industry. Using clinical samples from a flock with a long history of

223

ENA, we developed and tested serology, cytology, and RT-PCR based assays for utility as

224

diagnostic tools for identification of ENTV-1 infected or ENA-affected sheep. A comparison of

10

225

RT-PCR, ELISA and cytology results with post-mortem and immunohistological findings

226

showed poor correlation between the presence and absence of envelope positive ENA tumours

227

and results of cytology, ELISA and western blot. However, based on small numbers of animals,

228

RT-PCR on RNA extracted from nasal swabs proved to be 100% sensitive and specific. As an

229

indication of the sensitivity of this assay, a sheep less than eight months of age (2013-7), which

230

was positive by RT-PCR, was found to have a small neoplastic lesion comprised of

231

approximately 50 ENTV-1 Env-positive cells (Fig. 3).

232

The RT-PCR assay described in this study has many advantages. In addition to being highly

233

specific for exogenous ENTV-1, it is non-invasive and sample collection is simple to perform. In

234

fact, when we retested nasal swabs that were collected by the producer we found that the animals

235

had the same RT-PCR status irrespective of who collected the sample (data not shown). Since

236

the sampling materials are inexpensive, more than one sample per animal can be collected at one

237

time so as to have a backup should the RNA extraction fail. By generating cDNA using oligo dT

238

and random primers, RNA quality and presence of ENTV-1 could be assayed from the same

239

cDNA. However, the RT-PCR assay was not amenable to multiplexing so separate PCR

240

reactions were required to assess ENTV-1 and GAPDH. This assay could be useful in situations

241

where the veterinarian or producer would like confirmation that clinical signs are in fact due to

242

ENA and not some other condition such as nasal bots or chronic rhinitis. Indeed, one of the

243

sheep (2001-16) in our study had experienced chronic rhinitis and nasal discharge for many

244

months and appeared to be suffering from ENA. This animal was repeatedly negative for

245

ENTV-1 by RT-PCR and upon necropsy it was discovered that the animal had an extensive

246

bacterial infection but no evidence of ENA.

11

247

Sheep are thought to be immune tolerant to exogenous ENTV-1 infection. It has been

248

hypothesized that immune cells reactive to endogenous ovine betaretroviruses, which are

249

transcribed in the thymus and peripheral lymph nodes during ontogeny (Spencer et al., 2003),

250

undergo clonal deletion rendering sheep unable to mount an immune response against invading

251

exogenous ovine betaretroviruses. A study by Ortin et al. compared serum from healthy sheep

252

with serum from sheep affected with ENA for the presence of antibodies against ENTV-1 capsid

253

fused to glutathione S-transferase (GST). Although some serum samples reacted positively to

254

the antigen, the results were difficult to interpret due to cross-reactivity with the GST protein

255

alone (Ortín et al., 1998). For this reason, we chose to produce recombinant ENTV-1 capsid

256

protein using a hexa-His tag bacterial expression system and to remove the His tag prior to using

257

it as an antigen. The results presented here indicate that sheep can respond immunologically to

258

exogenous ENTV-1. Antibodies reactive against both the capsid and the envelope protein of

259

ENTV-1 were detected in the serum of sheep from a flock with a high incidence of ENA in both

260

ELISA and western blot analysis. There was, however, no consistency within ENA affected and

261

disease-free groups, such that animals in both groups had immunoreactive antibodies. Moreover,

262

since serum samples from naïve sheep reacted positively with ECa in immunoblot analysis, this

263

suggests that the immunoreactivity of serum samples from the flock of sheep with a history of

264

ENA is likely non-specific. Conversely, the ESU-IgG ELISA and immunoblot analysis detected

265

immunoreactive antibodies in serum samples from the flock with a history of ENA but not in the

266

naïve sheep serum samples, suggesting that these interactions were indeed specific and that

267

sheep are able develop antibodies against the ENTV-1 envelope protein.

268

Results from the virus neutralization assay revealed that the ability of a serum sample to

269

neutralize a virus pseudotyped with the ENTV-1 Env protein did not correlate well with the

12

270

detection of ENTV-1 envelope protein by IHC. The results of the ESU-IgG ELISA and the

271

neutralization assay showed only moderate correlation (with a Cohen’s kappa value of 0.427)

272

despite the fact that both assays detect antibodies directed against the ENTV-1 envelope protein.

273

The ESU-IgG ELISA contains only the ENTV-1 surface subunit whereas the neutralization assay

274

involves the full-length envelope protein, including the ectodomain region of the transmembrane

275

(TM) subunit. Epitopes in the ectodomain of HIV-1 gp41 (Hessell et al., 2010; Zwick et al.,

276

2001) have been shown to be important targets for broadly neutralizing antibodies so it is likely

277

that the ELISA negative neutralizing antibodies in our study represent antibodies targeting the

278

ENTV-1 envelope ectodomain. Therefore, the different composition of the antigens used in the

279

two assays could explain this lack of correlation.

280

Despite the fact that serology cannot reliably be used to diagnose ENTV-1 infection, it is clear

281

that some sheep are able to mount an immune response against the virus. It is likely that the

282

antibodies detected in these experiments represent low affinity antibodies because high affinity

283

antibody producing cells would be deleted in the thymus (Raimondi et al., 2007). Low affinity

284

antibody-producing cells on the other hand could escape clonal deletion and be induced to

285

expand when exposed to exogenous ENTV-1 antigens on antigen-presenting cells in the

286

periphery. We do not however, know whether disease-free sheep that mounted an immune

287

response against ENTV-1 were at one point infected with the virus and controlled the infection.

288

Nevertheless, it is possible that the immune system may be a factor in determining the outcome

289

of ENTV-1 infection, particularly in animals where tolerance to ENTV-1 is broken and

290

neutralizing antibodies are developed. Lastly, animals confirmed to have ENA by IHC and who

291

lack neutralizing antibodies could shed ENTV-1 chronically and potentially infect a significant

292

proportion of the flock.

13

293

Given that ENTV-1-infected sheep can harbor infectious tumours and exist in a flock unnoticed

294

for a considerable amount of time, or be sold as healthy animals, there is an urgent need for a

295

rapid and reliable antemortem diagnostic test for ENA for use in individual sheep. This report

296

describes a practical and highly specific RT-PCR technique for the detection of clinical and

297

preclinical ENA that may prove beneficial in future control or eradication programs.

298

Materials and methods

299

Sample collection

300

All work with animals was conducted in strict accordance with the Canadian Council of Animal

301

Care (CCAC) guidelines. The animal use protocol was approved by the Animal Care Committee

302

(ACC) of the University of Guelph. All efforts were made to minimize suffering. Two groups of

303

sheep were sampled. The first group consisted of 15 Suffolk rams from a research flock in

304

Quebec, Canada, that had displayed clinical signs of ENA, including nasal discharge and noisy

305

breathing, on and off for more than two years. Approximately 0.5 ml of nasal exudate was

306

combined with 0.75 ml of the RNA preservative, Trizol LS (Life Technologies) (2:3 ratio),

307

transported on ice and stored at -80°C. RNA was extracted according to the manufacturer’s

308

instructions and stored at -80°C until cDNA synthesis was performed. The second group of

309

animals enrolled in the study consisted of eighty horned Dorset sheep aged 6 months to 16 years

310

belonging to a commercial flock with a long history of ENA. In this group, both nostrils were

311

sampled by gently rubbing the nasal turbinates with an 8-inch sterile cytobrush (Fisher

312

Scientific). Two nasal swabs were obtained per animal. The first swab was transferred to a

313

charged microscope slide (Superfrost Plus; Fisher Scientific) using standard procedures and

314

transported to the laboratory immersed in PBS. Once in the lab, the slides were fixed in 4%

315

paraformaldehyde for 10 min, and stored in PBS at 4˚C until immunocytochemical staining was

14

316

performed. For the second swab, the tip of the cytobrush was cut off and placed into a 1.5 ml

317

microfuge tube containing 0.5 ml of RLT buffer (Qiagen). Specimens were transported to the

318

laboratory on ice and frozen at -80 °C within 4 h of collection. Blood samples were obtained by

319

venipucture of the jugular vein using serum-separating vacutainer tubes (Becton Dickinson). A

320

subset of animals was submitted for post-mortem analysis and nasal tissue harvested for

321

formalin-fixation.

322

Immunohistochemical and immunocytochemical staining

323

Fixed tissues were embedded in paraffin, sectioned at 5 μm and stained with haematoxylin-eosin.

324

The avidin-biotin-peroxidase complex (ABC) method was used on paraffin-embedded tissue

325

sections for immunohistochemical demonstration of ENTV envelope protein expression as

326

described previously (Walsh et al., 2013). Immunocytochemical staining of nasal smears was

327

conducted as above except deparaffinization and hydration steps were excluded. For the nasal

328

smears, three randomly selected fields per slide were evaluated by three independent graders.

329

Production of ENTV-1 capsid antibody

330

The ENTV-1 capsid gene was amplified from ENA tumor tissue using forward ECaF 5'-

331

GCTAGCCCTGTTTTTGAAAATAATAACCAG-3' and reverse ECaR 5'-GAATTCTTAAGC

332

AATACCTTGCATGTAGTA-3' primers, which contain a NdeI and a EcoRI restriction site,

333

respectively. The capsid amplicon was cloned into the pET28a plasmid (Novagen) using

334

restriction sites included in the primers to produce, pET28aECa, which contains an amino 6xHis

335

tag fused to ENTV-1 capsid protein by a thrombin cleavage site. Capsid protein was expressed

336

and purified in BL21star bacteria cells using IPTG induction and the Ni-NTA Fast Start kit

337

according to the manufacturer’s instructions. Eluted protein was processed with the Thrombin

338

CleanCleave Kit according to the manufacturer’s instructions (Sigma) and a nickel column was

15

339

used to remove the histidine tag. Purified capsid protein lacking the histidine tag was sent to

340

Pacific Immunology for generation of rabbit hyper immune serum (rαECa). The resultant

341

polyclonal antibody was able to efficiently detect both ENTV and JSRV capsid protein at a

342

dilution of 1:10,000.

343

Enzyme-linked immunosorbent assay (ELISA)

344

Serum antibodies reactive with ENTV-1 capsid and envelope proteins were detected by indirect

345

ELISA. Recombinant antigens used in the ELISA were prepared as follows: Recombinant

346

capsid protein from ENTV-1 (ECa) was expressed in E. coli using the pET28a System

347

(Novagen) and purified using a nickel column as described above. ESU-IgG, a fusion protein

348

comprised of the surface (SU) subunit of the ENTV-1 envelope protein fused to the Fc domain of

349

human IgG1 (Van Hoeven & Miller, 2005), was purified from the supernatant of 293T cells

350

transfected with pCMV-ESU-IgG using a HiTrap protein-G sepharose column (GE Healthcare)

351

according to manufacturer instructions. Briefly, flat-bottomed 96-well plates (VWR

352

International) were coated with 200 ng of purified ECa (2 μg/ml) or ESU-IgG (2 μg/ml) per well

353

and incubated at 4°C overnight. Plates were then washed with PBS-0.5% Tween 20 and blocked

354

for 1 h at room temperature with PBS containing 5% non-fat dry milk. Serum samples were

355

diluted in blocking buffer and added to the 96-well plates. Plates were then incubated at room

356

temperature for 2 h. Following three washes in PBS-0.5% Tween 20, horseradish peroxidase

357

(HRP)-conjugated rabbit anti-sheep IgG (Life Technologies) (diluted 1:10,000 in blocking

358

buffer) was added and plates were incubated for 1 h at room temperature. Plates were washed

359

and the reaction was visualized by the addition of 100 μl of ABTS substrate (Mandel Scientific)

360

for 10 min. Absorbance was measured at 405 nm using a microplate reader (Bio-tek

361

Instruments). Rabbit polyclonal antibodies raised against bacterially-expressed ENTV-1 capsid

16

362

protein (ECa) (described above) and a mouse monoclonal antibody that cross-reacts with the

363

ENTV envelope protein (Wootton et al., 2006b) were used as positive controls with the

364

appropriate HRP-conjugated secondary antibodies.

365

Immunoblot analysis

366

ECa and ESU-IgG proteins resolved on a 15% SDS-PAGE gel were transferred to polyvinyl

367

difluoride (PVDF) membrane and blocked overnight at 4˚C in PBS-0.5% Tween 20 containing

368

5% non-fat dry milk. The membrane was then cut into evenly sized strips and probed with

369

individual sheep serum samples (diluted 1:50 in blocking buffer) at 4˚C for 16 h. The strips were

370

washed with PBS-0.5%Tween 20 and incubated with a 1:5,000 dilution of HRP conjugated

371

rabbit anti-sheep IgG (Life Technologies) in blocking buffer for 1 h at room temperature. The

372

membrane strips were developed using Western Lightning Plus ECL (PerkinElmer) and imaged

373

with a ChemiDoc XRS (Bio Rad).

374

Virus neutralization assay

375

The ability of serum samples to neutralize ENTV-1 Env mediated infection was assessed by

376

neutralization assay. Briefly, NIH 3T3/LL2SN cells (Rai et al., 2001), which overexpress the

377

receptor for ENTV-1, Hyal2, were seeded in 12 well plates at 1x105 cells per well. After 16 h,

378

sheep serum samples were heat-inactivated at 56°C for 30 min, mixed at a 1:50 dilution with ~

379

400 infectious ENTV-1 Env-pseudotyped murine leukemia virus (MLV) particles expressing

380

human placental alkaline phosphatase (AP) in a total volume of 500 μL, and incubated for 30

381

min at 37°C. Note that the MLV particles were produced in LGPS/LAPSN cells [NIH 3T3 cells

382

that express Moloney murine leukemia virus (MoMLV) Gag and Pol (Miller et al., 1991) and

383

contain the LAPSN retroviral vector, which expresses a human placental alkaline phosphatase

384

(AP) gene (Miller et al., 1994)] transfected with pSX2EenvI4, a plasmid expressing a modified

17

385

version of the ENTV-1 envelope protein with enhanced pseudotyping efficiency (Walsh et al.,

386

personal communication), as described previously (Wolgamot et al., 1998). The virus-serum

387

mixtures were then added to NIH 3T3/ LL2SN cells in the presence of 8 μg/ml polybrene and

388

incubated for 4 h before replacing with 1 mL of fresh medium. Cells were fixed 48 hours later

389

with 3.7% formaldehyde and stained for AP activity as described (Van Hoeven & Miller, 2005).

390

The assay was performed in quadruplicate and neutralization was defined as a 50% or more

391

reduction in AP positive foci compared to PBS treated samples.

392

RNA extraction and RT-PCR

393

Total RNA was extracted from nasal exudate stored in Trizol LS according to the manufacturer’s

394

instructions. Total RNA was extracted from the nasal swabs stored in RLT buffer using the

395

RNeasy kit (Qiagen). cDNA libraries were synthesized using qScriptTM Flex cDNA SuperMix

396

(Quanta Biosciences) and a combination of random and oligo dT primers. A fragment of the

397

5'end of the ENTV genome was amplified (Fig. 1) using Platinum® PCR SuperMix (Life

398

Technologies) and the ENTV-U5-F (5'-GATGCTCCGTTCTCTCCTTATA-3') and GAG-R (5′-

399

GGGACGCGACGAATGTAGG-3′) (Walsh et al., 2010) primer pair. The efficiency of RNA

400

extraction and cDNA generation was tested using forward (5'-TGT TCC AGT ATG ATT CCA

401

CCC-3') and reverse (5'-ATA AGT CCC TCC ACG ATG CC-3') primers specific for exon and 3

402

and 6, respectively, of ovine glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Philbey et

403

al., 2006). Select RT-PCR products were sequenced to verify the identity of the amplicons.

404

Statistical Analysis

405

Sensitivity, specificity and kappa coefficient analysis was performed using GraphPad Prism v6

406

software. Specificity and sensitivity was calculated using the results of the anti-envelope

407

immunohistochemical staining of nasal tissue as the gold standard. Calculations included only

18

408

the results from the 22 animals from the 80+ horned Dorset flock for which post-mortem tissue

409

samples were available.

410

Acknowledgements

411

This study was supported by the Gartshore Memorial Sheep Research Fund and the National

412

Sciences and Engineering Research Council of Canada (NSERC). We acknowledge Dr. Gaston

413

Rioux and the Horned dorset producer for providing samples and Lisa Santry for assistance with

414

sample collection.

415

19

416 417

Figure Legends Fig. 1. Detection of ENTV-1 genomic RNA in nasal swabs. (A) A schematic of the ENTV-1

418

genome showing the location of primers used in the RT-PCR assay (arrows). VR stands for

419

variable region. (B) A representative image showing the results of RT-PCR amplification of

420

exogenous ENTV-1 from RNA extracted from nasal swabs. QC-1 to QC-6 represent samples 1-

421

6. Lanes 1, 3, 5, 7, 9, 11, 13, 15 and 17 contain RT-PCR products generated using ENTV-1

422

specific primers (592 bp product) and lanes 2, 4, 6, 8, 10, 12, 14, 16, and 18 contain RT-PCR

423

products (388 bp product) generated using primers specific for GAPDH. RNA isolated from

424

naturally acquired ENA (Pos) and healthy sheep lung (Neg) were used as a positive and negative

425

controls, respectively. H20; water only. M indicates 1Kb Plus DNA ladder (Life Technologies).

426 427

Fig. 2. Gross pathology of sheep nasal tumours. Representative images of tumours observed

428

in the nasal cavity at necropsy (A-D). Examples of a large, soft unilateral tumour obstructing the

429

entire nasal passage (A), a medium-sized, hard unilateral tumour (B) and a small soft tumour (C

430

and D) are shown. The arrow in D is pointing to the tumor. A piece of normal nasal turbinate is

431

shown for comparison. Images in A, B and C represent serial sections of the nose.

432 433

Fig. 3. Histopathology and immunohistochemical staining for ENTV-1 envelope protein

434

expression in nasal tumours. Representative images of hematoxylin and eosin stained nasal

435

tissues from ENA positive (A and C; animal number 2005-36) and ENA negative (E; animal

436

number 2012-76) sheep from a flock with endemic ENA. Immunohistochemical staining for

437

ENTV-1 envelope protein in ENA tissue obtained from macroscopically evident tumours (B and

438

D; animal number 2005-36) and macroscopically normal nasal tissue with an early envelope

439

positive lesion (F; animal number 2013-7). 20

440 441

Fig. 4. Nasal swab cytology detection of ENTV-1 envelope protein expression.

442

Representative images showing nasal swab cytology slides stained with an anti-envelope

443

monoclonal antibody and graded as ~ for no staining (A), * for low intensity (B), ** for

444

moderate intensity staining in >10% of cells or high intensity staining in 10% of cells (D).

446 447

Fig. 5. ELISA detection of antibodies reactive against ENTV-1 capsid and envelope

448

proteins. (A) Sheep serum samples tested in an indirect ELISA using purified ENTV-1 capsid

449

(ECa) and (B) envelope (ESU-IgG) proteins as antigens. Serum sample were assayed three times

450

and the absorbance is presented as the mean of three replicates ± standard deviation. The dotted

451

line represents the background cut off value. Group i, serum from control sheep with no history

452

of ENA; group ii, serum from sheep confirmed to have ENA, group iii, serum from sheep

453

confirmed to be free of ENA; and group iv, sheep of unknown ENA status. Chequered bars

454

represent RT-PCR-positive sheep with unknown disease status. Note that all animals in group i

455

were RT-PCR negative for ENTV-1.

456 457

Fig. 6. Representative Western blot analysis of sheep serum against ENTV-1 envelope and

458

capsid proteins. Serum samples from sheep with ENA (samples 1-3) and from disease-free

459

sheep that were housed with diseased animals (samples 4-6) were tested for reactivity against

460

ENTV-1 envelope (ESU-IgG) and capsid (ECa) proteins in a strip western blot. Naïve serum

461

was obtained from a high health status research flock with no history of ENA. ELISA results for

462

the various serum samples are shown above (for ESU-IgG) and below (ECa). Monoclonal anti21

463

JSRV envelope antibody (Wootton et al., 2006a), which cross reacts with ENTV envelope

464

(Wootton et al., 2006b), and rabbit anti-capsid antibody (rαECa) were used as positive controls

465

(C) for the ESU-IgG and ECa western blots, respectively. Lanes 1 (2011-5), 2 (2012-49), 3

466

(2010-4), 4 (2012-12), 5 (2006-38), and 6 (2007-1).

22

467

References

468 469

Caswell, J. & Williams, K. (2007). Enzootic nasal tumour of sheep. In Palmer’s Pathol Domest Anim, 5th edn., pp. 640–643. Edited by M. Maxie. Philadelphia, PA,: Elsevier Saunders.

470 471 472 473

Cousens, C., Minguijon, E., Garcia, M., Ferrer, L. M., Dalziel, R. G., Palmarini, M., De Las Heras, M. & Sharp, J. M. (1996). PCR-based detection and partial characterization of a retrovirus associated with contagious intranasal tumours of sheep and goats. J Virol 70, 7580–7583.

474 475 476 477

Cousens, C., Minguijon, E., Dalziel, R. G., Ortin, A., Garcia, M., Park, J., Gonzalez, L., Sharp, J. M. & de las Heras, M. (1999). Complete sequence of enzootic nasal tumour virus, a retrovirus associated with transmissible intranasal tumours of sheep. J Virol 73, 3986–3993.

478 479 480

DeMartini, J. C., Carlson, J. O., Leroux, C., Spencer, T. & Palmarini, M. (2003). Endogenous retroviruses related to jaagsiekte sheep retrovirus. Curr Top Microbiol Immunol 275, 117–137.

481 482 483 484 485

Hessell, A. J., Rakasz, E. G., Tehrani, D. M., Huber, M., Weisgrau, K. L., Landucci, G., Forthal, D. N., Koff, W. C., Poignard, P. & other authors. (2010). Broadly neutralizing monoclonal antibodies 2F5 and 4E10 directed against the human immunodeficiency virus type 1 gp41 membrane-proximal external region protect against mucosal challenge by simian-human immunodeficiency virus SHIVBa-L. J Virol 84, 1302–1313.

486 487 488

De Las Heras, M., García De Jalón, J. A., Minguijón, E., Gray, E. W., Dewar, P. & Sharp, J. M. (1995). Experimental transmission of enzootic intranasal tumours of goats. Vet Pathol 32, 19–23.

489 490

De Las Heras, M., Ortín, A., Cousens, C., Minguijón, E. & Sharp, J. M. (2003). Enzootic nasal adenocarcinoma of sheep and goats. Curr Top Microbiol Immunol 275, 201–223.

491 492 493

Miller, A. D., Garcia, J. V., von Suhr, N., Lynch, C. M., Wilson, C., & Eiden, M. V. (1991). Construction and properties of retrovirus packaging cells based on gibbon ape leukemia virus. J Virol 65, 2220–2224.

494 495 496

Miller, D. G., Edwards, R. H, & Miller, A. D. (1994). Cloning of the cellular receptor for amphotropic murine retroviruses reveals homology to that for gibbon ape leukemia virus. Proc Natl Acad Sci USA 91, 78–82.

497 498 499 500 501

Ortín, A., Minguijón, E., Dewar, P., García, M., Ferrer, L., Palmarini, M., Gonzalez, L., Sharp, J. & De Las Heras, M. (1998). Lack of a specific immune response against a recombinant capsid protein of Jaagsiekte sheep retrovirus in sheep and goats naturally affected by enzootic nasal tumour or sheep pulmonary adenomatosis. Vet Immunol Immunopathol 61, 229–37. 23

502 503 504 505

Palmarini, M., Hallwirth, C., York, D., Murgia, C., de Oliveira, T., Spencer, T. & Fan, H. (2000). Molecular Cloning and Functional Analysis of Three Type D Endogenous Retroviruses of Sheep Reveal a Different Cell Tropism from That of the Highly Related Exogenous Jaagsiekte Sheep Retrovirus. J Virol 74, 8065–8076.

506 507 508

Philbey, A. W., Cousens, C., Bishop, J. V, Gill, C. A., DeMartini, J. C. & Sharp, J. M. (2006). Multiclonal pattern of Jaagsiekte sheep retrovirus integration sites in ovine pulmonary adenocarcinoma. Virus Res 117, 254–263.

509 510 511 512 513

Rai, S. K., Duh, F.-M., Vigdorovich, V., Danilkovitch-Miagkova, A., Lerman, M. I. & Miller, A. D. (2001). Candidate tumour suppressor HYAL2 is a glycosylphosphatidylinositol (GPI)-anchored cell-surface receptor for jaagsiekte sheep retrovirus, the envelope protein of which mediates oncogenic transformation. PNAS 98, 4443–4448.

514 515

Raimondi, G., Turnquist, H. R. & Thomson, A. W. (2007). Frontiers of immunological tolerance. Methods Mol Biol 380, 1–24.

516 517 518

Spencer, T. E., Mura, M., Gray, C. A., Griebel, P. J. & Palmarini, M. (2003). Receptor usage and fetal expression of ovine endogenous betaretroviruses: implications for coevolution of endogenous and exogenous retroviruses. J Virol 77, 749–753.

519 520 521

Van Hoeven, N. S. & Miller, A. D. (2005). Improved enzootic nasal tumour virus pseudotype packaging cell lines reveal virus entry requirements in addition to the primary receptor Hyal2. J Virol 79, 87–94.

522 523 524 525

Walsh, S. R., Linnerth-Petrik, N. M., Laporte, A. N., Menzies, P. I., Foster, R. A. & Wootton, S. K. (2010). Full-length genome sequence analysis of enzootic nasal tumour virus reveals an unusually high degree of genetic stability. Virus Res 151, 74–87. Elsevier B.V.

526 527 528

Walsh, S. R., Linnerth-Petrik, N. M., Yu, D. L., Foster, R. a, Menzies, P. I., Diaz-Méndez, A., Chalmers, H. J. & Wootton, S. K. (2013). Experimental transmission of enzootic nasal adenocarcinoma in sheep. Vet Res 44, 66.

529 530 531

Wolgamot, G., Rasko Miller, J. E. J., & Miller, A. D. (1998) Retrovirus Packaging Cells Expressing the Mus dunni Endogenous Virus Envelope Facilitate Transduction of CHO and Primary Hematopoietic Cells. J Virol 72, 10242-10245.

532 533 534 535

Wootton, S. K., Metzger, M. J., Hudkins, K. L., Alpers, C. E., York, D., DeMartini, J. C. & Miller, A. D. (2006a). Lung cancer induced in mice by the envelope protein of jaagsiekte sheep retrovirus (JSRV) closely resembles lung cancer in sheep infected with JSRV. Retrovirology 3, 94.

24

536 537 538

Wootton, S. K., Halbert, C. L. & Miller, A. D. (2006b). Envelope proteins of Jaagsiekte sheep retrovirus and enzootic nasal tumor virus induce similar bronchioalveolar tumors in lungs of mice. J Virol 80, 9322-9325.

539 540 541 542

Zwick, M. B., Labrijn, A. F., Wang, M., Spenlehauer, C., Saphire, E. O., Binley, J. M., Moore, J. P., Stiegler, G., Katinger, H. & other authors. (2001). Broadly neutralizing antibodies targeted to the membrane-proximal external region of human immunodeficiency virus type 1 glycoprotein gp41. J Virol 75, 10892–10905.

543 544

25

Tables Table 1. RT-PCR and histology results from pilot study Animal ID

RT-PCR ENTV-1

RT-PCR GAPDH

IHC for Env

QC-1 QC-2 QC-3 QC-4 QC-5 QC-6 QC-7 QC-8 QC-9 QC-10 QC-11 QC-12 QC-13 QC-14 QC-15

+ + + +

+ + + + + + + + + + + + + + +

+ ND ND ND ND +§ ND ND ND ND + ND ND +

§

small hyperplastic lesion that stained positive for ENTV-1 Env, (+) positive, (-) negative, ND (not determined)

26

Table 2: Summary of sample analysis from sheep submitted for post-mortem examination

Tissue Animal ID

Necropsy Findings

Nasal Swab IHC for Env

RTPCR GAPDH

RT-PCR ENTV-1

Cytology #1

Cytology #2

Serum Cytology #3

ECa ELISA

ESUIgG ELISA

ECa WB

Env WB

Virus Neutralization‡

2001-16 Normal + P/S P/S ~ + + + - (11%) 2007-1 Normal + * ~ ~ + - (22%) 2008-1 Normal + ~ ~ ~ + + + + + (95%) 2012-12 Normal + *** ** ** + + + + (96%) 2012-18 Normal + ~ P/S ~ + + + - (27%) 2012-76 Normal + * ** ~ + - (26%) 2006-38 Normal + ~ ~ * + + - (32%) 2013-1 Normal + ND ND ND + + + (88%) 2005-26 Tumor + + + P/S ~ ~ + - (9%) 2006-37 Tumor + + + ** ** ** + + - (30%) 2008-31 Tumor + + + * * ** + - (6%) 2010-43 Tumor + + + * ~ ~ + + + + (76%) 2010-44 Tumor + + + ~ ~ ~ + + (89%) 2011-50 Tumor + + + ** ** *** + + + - (43%) 2012-49 Tumor + + + ~ ~ ~ + - (27%) 2012-72 Tumor + + + * *** ** + (57%) 2012-81 Tumor + + + ~ * ~ + +/+ + (87%) 2009-34 Tumor + + + P/S P/S P/S + + (58%) 2005-36 Tumor + + + P/S P/S P/S - (15%) 2010-4 Tumor + + + * ** * + + + + (70%) 2012-7 Tumor + + + P/S ~ * + (51%) 2013-7§ Tumor¶ + + + ND ND ND ND ND ND ND ND (+) positive, (-) negative, (¶) IHC positive early neoplastic lesion, (§) no serum sample obtained, (P/S) poor sample, (~) - no staining, (*) low intensity staining, (**) moderate intensity staining, (***) high intensity staining, (WB) western blot, (ND) not determined, (‡) value in brackets represents the % reduction in infectivity, grey background highlights samples positive by IHC for envelope expression (gold standard), (ESU-IgG) SU domain of ENTV Env fused to the Fc domain of human IgG, (ECa) ENTV-1 capsid protein.

27

Table 3. Specificity and sensitivity values for the diagnostic tests evaluated

Test RT-PCR Cytology ECa ELISA ESU-IgG ELISA Virus Neutralization CI – confidence interval

Specificity Result 100% 66.70% 62.50% 33.30% 62.50%

Sensitivity

95% CI 63.1% - 100% 22.3% - 95.7% 24.5% - 91.5% 11.8% - 61.6% 24.5% - 91.5%

28

Result 100% 50% 23.10% 50.00% 53.80%

95% CI 76.8% - 100% 18.7% - 81.3% 5.0% - 53.8% 11.8% - 88.2% 25.1% - 80.8%