Serological and molecular evidence of Q fever among small ruminant flocks in Algeria

Accepted Manuscript Title: Serological and molecular evidence of Q fever among small ruminant flocks in Algeria Author: H. Khaled K. Sidi-Boumedine S. Merdja P. Dufour A. Dahmani R. Thi´ery E. Rousset A. Bouyoucef PII: DOI: Reference:

S0147-9571(16)30038-8 http://dx.doi.org/doi:10.1016/j.cimid.2016.05.002 CIMID 1068

To appear in: Received date: Revised date: Accepted date:

29-10-2015 2-5-2016 12-5-2016

Please cite this article as: Khaled H, Sidi-Boumedine K, Merdja S, Dufour P, Dahmani A, Thi´ery R, Rousset E, Bouyoucef A.Serological and molecular evidence of Q fever among small ruminant flocks in Algeria.Comparative Immunology, Microbiology and Infectious Diseases http://dx.doi.org/10.1016/j.cimid.2016.05.002 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Title: Serological and molecular evidence of Q fever among small ruminant flocks in Algeria

2

Authors: H. Khaleda*, K. Sidi-Boumedineb, S. Merdjaa, P. Dufourb, A. Dahmania, R. Thiéryb,

3

E. Roussetb, A. Bouyoucefa

4

a

5

b

LBRA, Institute of Veterinary Sciences, University Blida 1, Algeria ANSES, Laboratory of Sophia-Antipolis, Animal Q Fever Unit, Sophia Antipolis, France

6 7

Highlights

8

Q fever is a zoonosis widely reported in the world. The causative agent is Coxiella burnetii, an

9

obligate intracellular bacterium. The infection is often asymptomatic in ruminants, but it can

10

lead to reproductive disorders with bacterial shedding in the environment. Between 2011 and

11

2013, a study was undertaken in small ruminants flocks in different regions of Algeria. A total

12

of 35 flocks were visited and 227 sera and 267 genital swabs were collected from females after

13

abortions or lambing period in order to investigate Q fever infection. Indirect ELISA was used

14

to detect specific antibodies against C. burnetii and real time PCR for detecting bacterial DNA.

15

Our survey indicated that 58% (95% IC= 40-76%) of flocks had at least one positive animal

16

(17 seropositive flocks) and individual seroprevalence was estimated at 14.1% (95% IC= 11.8-

17

16.4%) (32 seropositive animals). The bacterial excretion has been observed in 21 flocks (60%),

18

and 57 females have proved C. burnetii excretion (21.3%). These results suggest that C. burnetii

19

distribution is high at flock’s level. Therefore seropositive and shedder animals can be found

20

all over the country. Further studies are needed in other regions and different animal species to

21

better understand the distribution and incidence of this disease, as well as human exposure, and

22

establish an adequate prophylaxis program.

23 24 25 26

Abstract: 1

27

Q fever, a commonly reported zoonosis worldwide, is caused by infection with Coxiella

28

burnetii, an obligate intracellular bacterium. The infection is often asymptomatic in ruminants,

29

but it can lead to reproductive disorders with bacterial shedding into the environment. Between

30

2011 and 2013, a study was undertaken in small ruminant flocks in different regions of Algeria.

31

A total of 35 flocks were visited and 227 sera and 267 genital swabs were collected from

32

females after abortions or the lambing period to investigate Q fever infection. Indirect ELISA

33

was used to detect specific antibodies against C. burnetii and real-time PCR for detecting

34

bacterial DNA. Our survey indicated that 58% (95% CI = 40-76%) of flocks had at least one

35

positive animal (17 seropositive flocks) and individual seroprevalence was estimated at 14.1%

36

(95% CI= 11.8-16.4%) (32 seropositive animals). Bacterial excretion was observed in 21 flocks

37

(60%), and 57 females showed evidence of C. burnetii shedding (21.3%). These results suggest

38

that C. burnetii distribution is high at the flock level and that seropositive and infected (shedder)

39

animals can be found all over the country. Further studies are needed in other regions and on

40

different animal species to better understand the distribution and incidence of Q fever, as well

41

as human exposure, and to develop an adequate prophylaxis program.

42

Keywords: Coxiella burnetii, Q fever, sheep, goat, ELISA, Real-time qPCR, Algeria

43

2

44

1. Introduction

45

Coxiella burnetii is an obligate intracellular bacterium that causes the zoonotic disease Q fever.

46

The bacterium has been found worldwide in a wide range of animal hosts, including mammals,

47

birds and ticks [1]. Moreover, a spore-like form of C. burnetii can survive extracellularly,

48

contributing to its persistence and widespread dissemination in the environment [2]. Ruminants

49

represent the primary reservoir of this organism [3]. In these animals, Q fever is mainly

50

asymptomatic, but can be responsible for reproductive disorders, including abortions that

51

generally occur at the end of gestation, as well as stillbirths and delivery of weak and unviable

52

newborns [4]. These reproductive failures are accompanied by high levels of bacterial shedding

53

through vaginal secretions, birth products, faeces, urine, and milk [5, 6, 7]. Nevertheless, active

54

or passive surveillance for Q fever in ruminants is rarely performed; thus, the prevalence and

55

the incidence of Q fever cannot be accurately estimated anywhere in the world [3]. Therefore,

56

animal infection is often revealed after the notification of human clinical cases. Outbreaks are

57

generally associated with proximity to sheep and goats, particularly during parturition or

58

abortion, during dry and windy weather [3,8].

59

In Algeria, Q fever was described for the first time in 1948 in French soldiers and then in 1956.

60

Cases appear to be linked to contact with small ruminants [9]. In a seroprevalence study, a rate

61

of 5.4% was reported in children under 16 years in the south of the country [10]. Another study

62

mentioned a rate of 15.5% in inhabitants of an agro-pastoral region in the east [11].

63

Nevertheless, studies on Q fever in animals are still rare in Algeria [12].

64

The purpose of this study conducted in Algeria between 2011 and 2013 was to estimate Q fever

65

seropositivity and shedding in small ruminant flocks. These data on the Q fever epidemiological

66

situation in animals aims to improve the visibility of this neglected or unknown disease;

67

enhance knowledge and facilitate future comparative studies; and participate in the

68

development of a surveillance plan and/or appropriate monitoring for the country.

3

69 70

2. Materials and methods

71

2.1. Study site

72

The study took place in eight departments of Algeria, covering the geographical and climatic

73

diversity of the country. The chosen regions were as follows: Constantine, Skikda, and Ain

74

Defla with a Mediterranean climate; Bordj Bouareridj, Medea, Djelfa and El Bayadh with a

75

continental climate and Biskra with a Saharian climate (Fig. 1). According to headcount

76

statistics for 2012 from the Algerian Ministry of Agriculture, sheep predominate the Algerian

77

ruminant population and account for 80% of the total estimated livestock population with more

78

than 25 million heads, including 12 million ewes. Goats are second-most common species

79

(13%) and 58% are females. Pastoral livestock production is concentrated in the steppe (in the

80

north-central part of the country), harbouring the largest small ruminant population in Algeria.

81

During the summer seasons, transhumance and nomadism to the north-east and north-west

82

become a necessity, especially from May to September when the pastures can no longer feed

83

the flocks.

84 85

2.2. Sampling

86

For every farm visited, a survey was completed to provide information regarding abortion

87

antecedents, size, composition and production system at the flock level. For each sampled

88

animal, the species, age, symptoms (i.e. abortion or normal delivery) were also recorded. From

89

the same animal, a blood sample and a genital swab were taken. Sampling was performed

90

approximately one week after lambing or abortion, according to farmers’ observations. In

91

regard to local customs, the flock composition changes over time and animals are not marked.

92

Moreover, no vaccination against Q fever has ever been administered in Algeria.

4

93

The blood sample (5 mL) was collected from the jugular vein of each animal using a vacutainer

94

tube. Sampling was performed by qualified veterinarians as part of (routine?) sample collection

95

for surveillance with the full consent of the farmers. Sera were separated from clotted blood by

96

centrifugation at 1500×g for 15 min, aliquoted into clean 1.5 mL tubes and stored at -20°C until

97

analysis.

98

The swab was rubbed on the inner vaginal wall to insure the collection of cells and intracellular

99

bacteria. Each specimen was marked with a code including an individual sampling number and

100

accompanied by an information sheet with the flock characteristics and tested animals. The

101

specimen samples were analysed with the help of the OIE and the French Reference Laboratory

102

for animal Q fever (ANSES Sophia Antipolis, France). Samples were sent to this laboratory

103

after sending the samples under cold storage, and the transport period did not exceed 24 h.

104 105

2.3. Laboratory testing

106

2.3.1. ELISA

107

The specific anti-Coxiella burnetii antibodies in the serum samples were measured using a

108

commercially available indirect enzyme-linked immunosorbent assay (ELISA) kit (LSIVET

109

Ruminant Serum/Milk Kit, batch #: ElisaCoxLS-001, France) according to the manufacturer’s

110

instructions. The technique uses microtiter plates pre-coated with a purified C. burnetii antigen

111

of ovine origin. Results were expressed as a percentage of the optical density (OD) reading of

112

the test calculated as %OD=100 × (OD sample – ODm Negative Control)/(ODm Positive

113

Control – ODm Negative Control) where ODm is the measured OD. An animal was considered

114

positive (ELISA+) when its %OD had a value between 40% and 100%, highly positive

115

(ELISA++) for values greater than 100%, and negative (ELISA-) for values lower than 40%.

116

We used an ELISA recorder and software provided by SAFAS® (Monaco).

117

5

118

2.3.2. Quantitative PCR

119

DNA was extracted from vaginal swab specimens using the QIAamp DNA Mini Kit (Qiagen,

120

Courtaboeuf, France) in a biosafety level-3 laboratory. DNA extracts were tested using a

121

quantitative real-time polymerase chain reaction (qPCR) in-house method with the IS1111 gene

122

target for C. burnetii [5]. The qPCR assays were carried out on 5 µL of DNA extract (with

123

uracil-N-glycosylase to hydrolyse the uracil-glycosidic bonds in DNA containing dUTP,

124

thereby preventing carry over from previous qPCR reactions). The thermal cycling conditions

125

included an initial step of 50°C for 2 min, one cycle at 95°C for 10 min to activate the Taq

126

polymerase and 40 cycles at 95°C for 15 s and at 60°C for 1 min. Amplifications were

127

performed in an automated DNA thermal cycler and data were analysed using the

128

accompanying software (Applied Biosystems 7500, version 2.0.5). A sample was considered

129

positive if the value of the threshold cycle (Ct) of the target gene was below 40. The bacterial

130

load present in each specimen was quantified by converting the Ct values of the target gene into

131

estimated quantities of C. burnetii using serial dilutions of known concentrations of the external

132

positive control (C. burnetii Nine Mile strain calibrated at 3x109 bacteria/mL). An internal

133

positive control (ruminant-specific Gapdh gene) was used to rule out false negatives caused by

134

PCR inhibition. In addition, two negative control samples (NCS) were used to monitor

135

contamination during manipulation of samples and of qPCR mixes respectively: (1) PBS was

136

included in every series of 10 samples for DNA extraction and (2) DNase-RNase-free water

137

was tested along with DNA samples for PCR runs. In the case of suspected PCR inhibition, the

138

qPCR was repeated after a 1:10 dilution of the DNA extract.

139 140

2.4. Statistical analysis

141

Data were analysed using STATISTICA software (version 11.0). A chi-squared test was used

142

to detect significant differences at the flock level (abortion antecedents, size, composition and

6

143

production system) and at the individual level (species, age, abortion or normal delivery). A

144

probability of less than 5% was considered statistically significant. For comparisons involving

145

small sample sizes, probabilities were calculated using the Fisher exact test. The Kruskal-Wallis

146

test was used to compare excretion means in different serological categories, and a Bonferroni

147

test was used to correct the significance level for multiple comparisons. Finally, the odds ratio

148

(OR) was calculated to quantify the association between positive qPCR with ELISA positive

149

results relative to ELISA negative results. The confidence interval (CI) was calculated using

150

the Miettinen method.

151 152

3. Results

153

3.1. Sampling obtained

154

In total, 35 small ruminant flocks were studied, from which 227 sera samples from sheep and

155

goat females and 267 genital swabs were collected. Blood samples from six flocks were

156

discarded due to haemolysis. Descriptive characteristics of flocks and results obtained for each

157

flock using ELISA (n = 29) and qPCR (n = 35) are given in Table 1. Statistical analyses at the

158

flock and individual levels are given in Table 2.

159 160

3.2. Determination of seroprevalence

161

A farm was considered positive for Coxiella burnetii when at least one animal showed a positive

162

ELISA results. Our results indicate that 17 of 29 tested flocks showed evidence of previous

163

infection with C. burnetii, i.e. 58% at the flock level (95% CI= 40-76%). A prevalence rate of

164

14.1% (95% CI= 11.8-16.4%) was determined at the individual level (32/227).

165

Overall seroprevalence between regions ranged from 11.8% in the Constantine and Djelfa

166

departments to 19.2% in the Biskra department. Flock seroprevalence varied from 0% to 50%

167

among departments but differences among flocks were not statistically significant (p=0.83).

7

168

Only two tested flocks were seronegative.. Interestingly, seropositive animals were observed in

169

80% of flocks with known abortion history, but in only 43% of flocks without any abortion

170

antecedents,. However, the statistical association between abortion history and seropositivity

171

was weak and non-significant (p=0.06). There was no association between prevalence rates and

172

flock size: comparisons of flocks comprising more than 100 animals (66.7%) and less than 100

173

animals (56.5%) were not significant (p=0.65). Further, no significant differences were

174

observed between the prevalence rate observed in sheep flocks and mixed flocks (71.4% and

175

46.7%, respectively, p=0.26). Curiously, similar seroprevalence rates were observed between

176

sedentary flocks (55.6%) and transhumant flocks (60%) (p=0.83).

177

There were no species-specific differences in seroprevalence: 12.2% of tested sheep and 15.0%

178

of tested goats included seropositive animals (p=0.62). Aborted females showed a rate of 16.2%

179

whereas 8.4% of females with normal lambing were seropositive; again, this difference was not

180

statistically significant (p=0.13). The highest rate of seroprevalence was observed in

181

primiparous females (20%) compared with only 12.9% in multiparous females. However, the

182

statistical relationship of this difference was weak and thus non-significant (p= 0.08).

183

Of the 32 positive cases, only 8 had a %OD value higher than 100% (ELISA++) and considered

184

as highly positive, 24 were positive (ELISA+) with a %OD value between 40 and 100% (Fig.

185

2).

186 187

3.3. Bacterial shedding via the vaginal route

188

The shedder status of a flock was confirmed when at least one animal presented a positive qPCR

189

result. Of the tested flocks, 21 had at least one shedding female (60%) and shedding was

190

demonstrated in 57 of the 267 tested females with variable quantities of C. burnetii in their

191

vaginal secretions (21.3%). The intra-flock percentages ranged from 5 to 50% for 14 flocks,

192

from 55 to 85% for 5 flocks, and 1 flock in the Ain Defla department showed 100% of positive

8

193

cases. In contrast, 14 flocks were non-shedders, and 8 flocks had only one shedding animal.

194

The maximum quantity of bacteria, i.e. 1.31x108 bacteria per vaginal swab, was observed in a

195

ewe. We observed 241 females with 0- 5x102 bacteria per vaginal swab, 8 females with 5x102-

196

1x103 bacteria per vaginal swab, 7 females with 1x103-1x104 bacteria per vaginal swab, 6

197

females with 1x104-1x106 bacteria per vaginal swab and 5 females with more than 1x106

198

bacteria per vaginal swab (Fig. 3).

199

At the flock level, there were no significant statistical differences for comparisons involving

200

flock characteristics: abortion history (61.1% for flocks with abortion antecedents vs. 58.8%

201

without antecedents, p=0.98); size (45.4% for large flocks and 70.1% for flocks with less than

202

100 animals, p=0.14) composition (45% for sheep flocks and 80% for mixed flocks, p=0.06) or

203

production system (66.7% in sedentary flocks against 56.5% in transhumant flocks, p=0.56).

204

At the individual level, no statistical relationship was detected between bacterial shedding and

205

species (19.7% and 30.2% for ovine and caprine, p=0.12), symptoms (22.3% for abortions and

206

22.9% for normal deliveries, p=0.71) or age (18.3% for primiparous females and 22.9% for

207

multiparous females, p=0.39).

208 209

3.4. Evaluation of concordance between ELISA and qPCR results

210

Because 6 flocks were not tested using ELISA, only 29 flocks were included in the analysis of

211

possible associations among results of ELISA on sera and qPCR on vaginal secretions obtained

212

from the same animal.

213

Our results showed that 7 of the 12 (58.3%) seronegative flocks had at least one qPCR positive

214

animal, and 10 of the 17 seropositive flocks (58.8%) were shedders: there were no significant

215

differences in shedder status in regard to serological status (p= 0.98). Vaginal excretion of C.

216

burnetii was detected in 19.5% (38/195) of seronegative animals and in 21.9% (7/32) of

217

seropositive animals, again without any statistical significance (p=0.75).

9

218

Concordance between ELISA and qPCR methods was detected in 9 flocks showing positive

219

ELISA/positive qPCR animals, and in 4 flocks with negative ELISA/negative qPCR animals.

220

At the individual level, only 7 animals were positive for both types of test and 163 were negative

221

for both types of test (Table 3). The mean bacterial load in vaginal secretions, expressed in

222

number of genome equivalents (GE) in log10/mL, for animals with ELISA negative results was

223

2.77 (standard deviation (σ)=2.7) compared with 3.03 (σ=1.5) for seropositive animals and 4.77

224

(σ=4.7) for high seropositive animals, without statistical significance (Kruskal-Wallis, p= 0.39).

225

However, Bonferroni correction showed that differences were robustly significant between

226

excreting animals with negative ELISA results (ELISA-) and positive ELISA (ELISA+) results

227

(p˂0.0001), and between excreting animals with ELISA- and ELISA++ (p˂0.0001). Excreting

228

animals with ELISA+ and ELISA++ did not differ statistically (p=0.94). Thus, the double

229

analysis of PCR and ELISA results at the individual level showed that strong shedders are more

230

likely to be seropositive.

231 232

4. Discussion

233

The epidemiology and evolution of animal Q fever have not been extensively studied, if at all,

234

in most countries, including Algeria. The disease is generally not suspected by practitioners

235

after observation of abortions cases. Therefore, at veterinary diagnostic laboratories, tests for Q

236

fever are not part of routine differential diagnosis for abortion cases. Here, we undertook the

237

first descriptive study on Q fever in the main areas of small ruminant production in Algeria to

238

estimate its prevalence. Due to differences in study design, sampling approaches and applied

239

methods, comparisons with other surveys are difficult., We nevertheless point out some trends

240

in regard to our study.

241

The overall seroprevalence rate at the flock level in our study was 58.6% (71.4% in sheep flocks

242

and 46.4% in mixed flocks) with or without abortion events. These rates are higher than those

10

243

reported by Masala et al. (2010) in flocks with reproductive disorders (47% for goat flocks and

244

38% for sheep flocks) [13]. At the individual level, the seroprevalence rate was estimated at

245

14.1%, a rate slightly higher than those described in Italy [13] and India [14], which are of the

246

order of 9%. The higher rate found in our study suggests that Q fever has already spread through

247

the country via environmental exposure to Coxiella burnetii.

248

A slightly higher prevalence rate was observed for flocks with more than 100 animals compared

249

with that of small flocks (66.7% and 56.5%, respectively), perhaps due to animal overcrowding

250

in livestock buildings, where high density may influence animal welfare and the occurrence of

251

infectious diseases. Moreover, the observation of high C. burnetii seropositivity rates in flocks

252

with a history of abortion (80% in this study), in contrast to low rates for flocks without abortion

253

antecedents (43%), has also been reported in previous studies [15,16,17] in which abortions

254

were linked to Q fever. At the individual level, our results also suggest an association between

255

high seropositivity in aborted females compared with females with normal lambing. Although

256

the difference was not statistically significant, a slight increase was observed for aborted

257

females (16.2%) compared with 8.4% for females with normal deliveries, providing further

258

support that Q fever is involved in these abortions. Finally, another albeit non-significant trend

259

was observed: primiparous females (20.0%) were more often seropositive than multiparous

260

females (12.9%). A similar pattern was observed in (country) where seropositivity was 28% for

261

primiparous and 19% for biparous females [18]. This pattern likely indicates recent bacterial

262

circulation. Further studies need to be conducted on a larger sample sizes to confirm these trends

263

in Algeria.

264

We detected C. burnetii DNA in 57 vaginal samples (21.3%) and 21 flocks were thus

265

considered as shedder flocks. These excreting females may represent a source of contamination

266

for other animals and the environment, leading to the risk of human epidemics as already

267

reported in several countries [19]. We observed very similar proportions of shedders between

11

268

sheep and goat females. After abortions or normal lambings, there were no differences in

269

bacterial shedding, observed in 22.9 and 17.2% of females, respectively. These rates are similar

270

to those described in Rousset et al. (2009) with 23% for aborted and 11% for non-aborted

271

females in goat flocks where a Q fever abortion outbreak had occurred [20]. However, shedding

272

bacteria in flocks without clinical signs of abortion have been observed in other studies: in

273

Alsaleh et al. (2011) with a rate of 22.5% [21] and in Dubuc-Forfait et al. (2009) with a rate of

274

20% [22]. Nevertheless, the latter study had bacterial quantities clearly lower than in Q fever

275

aborted flocks.

276

Despite the good sensitivity and specificity of the ELISA technique used in this study [23], the

277

serological diagnosis of Q fever in small ruminants is not, in itself, appropriate, because

278

immunological responses do not prove bacterial shedding, but only provide evidence of

279

previous and/or present exposure to C. burnetii [24]. In our study, the very similar results

280

obtained between shedder animals with negative serology (19.5%) and positive serology

281

(21.9%) contrast with studies done on goat flocks. For example, Dubuc-Forfait et al. (2009)

282

report percentages of 26% and 72% respectively in asymptomatic flocks sampled less than one

283

month after parturition and 42% and 91% for flocks with confirmed Q fever cases [22]. In

284

another study on clinical Q fever outbreaks, percentages were 43.3% and 91.2% for animals

285

sampled on the day of normal delivery or abortion [7]. In our study, the two methods provided

286

similar diagnosis in 9 flocks and only for 7 animals (3%). Thus, the methods show no

287

association in terms of qualitative results. Interestingly, in terms of quantitative results, there

288

was a significant positive association, confirming the observation reported in Dubuc-Forfait et

289

al. (2009) for goats in south-eastern France.

290

The main purpose of this study was to assess the presence of Q fever in small ruminant flocks

291

from eight departments of Algeria using serological and molecular methods. This study has

292

helped to better determine the occurrence of Q fever in Algerian small ruminants flocks,

12

293

although the statistical analyses lack power due to small sample sizes. The ELISA results

294

clearly demonstrate the contact of tested females with the infectious agent. In addition, C.

295

burnetii shedding was demonstrated not only in aborted females, but also after normal

296

deliveries. Moreover, these animals may be chronically infected and shed bacteria in future

297

pregnancies, thus participating in environmental contamination and consequently the spread of

298

the infection. Attempts to isolate and genotype the circulating strains are currently underway.

299

The lack of knowledge on Q fever may increase infection risks for livestock and humans.

300

Therefore, raising the awareness of practitioners, farmers, and testing laboratories is an absolute

301

necessity.

302

Our study represents a preliminary step for future studies on animal and human Q fever that

303

can be extended to other species and other parts of the country using multidisciplinary scientific

304

approaches. The ultimate goal is to develop an epidemiological surveillance system adapted to

305

Algeria and its specificities.

306 307

Acknowledgments

308

The authors would like to thank the Francophone University Association (AUF) for providing

309

funds for the study and Ms. Carolyn Engel-Gautier for improving the English form of the

310

manuscript.

311 312

References

313

[1] T.J. Marrie, D. Raoult, 2015. Coxiella burnetii (Q Fever). Mandell, Douglas, and Bennett's

314

Principles and Practice of Infectious Diseases (Eighth Edition), Volume 2, 2015, Pages 2208-

315

2216.e2

13

316

[2] P. Ratmanov, H. Bassene, F. Fenollar, A. Tall, C. Sokhna, D. Raoult, D. Mediannikov, The

317

correlation of Q fever and Coxiella burnetii DNA in household environments in rural Senegal,

318

Vector Borne Zoonotic Dis., 13, (2013) 70-72.

319

[3] EFSA (European Food Safety Authority), Panel on Animal Health and Welfare (AHAW).

320

S. More, J.A. Stegeman, A. Rodolakis, H.J. Roest, P. Vellema, R. Thiéry, H. Neubauer, W. van

321

der Hoek, K. Staerk, H. Needham, A. Afonso, M. Georgiev, J. Richardson, Scientific opinion

322

on Q Fever, EFSA Journal, 8 (2010) 1593-1709.

323

[4] E. Rousset, K. Sidi-Boumedine. 2015. Chapter 2.1.12. Q fever. In: Manual of diagnostic

324

tests and vaccines for terrestrial animals (mammals, birds and bees). 8th ed., O.I.E.

325

http://www.oie.int/en/international-standard-setting/terrestrial-manual/access-online/

326

[5] A. Joulié, K. Laroucau X., Bailly, M. Prigent P. Gasqui E. Lepetitcolin B. Blanchard E.

327

Rousset K. Sidi-Boumedine E. Jourdain. Circulation of Coxiella burnetii in a naturally infected

328

flock of dairy sheep: Shedding Dynamics, environmental contamination, and genotype

329

diversity, Appl. Environ. Microbiol., 15 (2015) 7253-7260.

330

[6] I. Astobiza, F.F. Barandika, F. Ruiz-Fons, A. Hurtado, I. Povedano, R.A. Juste, A.L. Garcia-

331

Perez, Four-year evaluation of the effect of vaccination against Coxiella burnetii on reduction

332

of animal infection and environmental contamination in a naturally infected dairy sheep flock,

333

Appl. Environ. Microbiol., 77 (2011) 87-99.

334

[7] R. de Crémoux, E. Rousset, A. Touratier, G. Audusseau, P. Nicollet, D. Ribaud, V. David,

335

M. Le Pape, Coxiella burnetii vaginal shedding and antibody responses in dairy goat flocks in

336

a context of clinical Q fever outbreaks. FEMS Immunology and Medical Microbiology, 64

337

(2012) 120-122.

338

[8] ECDC (European Centre for Disease Prevention and Control), Panel with representatives

339

from the Netherlands, France, Germany, UK, United States. Asher, A., Bernard, H., Coutino,

340

R., Durat, G., De Valk, H., Desenclos, J-C., Holmberg, J., Kirkbridge, H., More, S,

14

341

Scheenberger, P., van der Hoek, W., van der Poel, C., van Steenbergen, J., Villanueva, S.,

342

Coulombier, D., Forland, F., Giesecke, J., Jansen, A., Nilsson, M., Guichard, C., Mailles, A.,

343

Pouchol, E., Rousset, E. 2010. Risk assessment on Q fever. ECDC Technical report

344

doi:102900/28860. Available online: www.ecdc.europa.eu

345

[9] M. Pierrou, G. Mimoune, G. Vastel, Une importante épidémie de fièvre Q (175 cas)

346

observée à Batna (Algérie). Presse Med., 64, (1956) 471-473.

347

[10] N. Dumas, Rickettsioses et chlamydioses au Hoggar (République Algérienne): Sondage

348

épidémiologique. Bull. Soc. Path. Ex. 77 (1984) 278-283.

349

[11] A. Lacheheb, D. Raoult, Seroprevalence of Q-fever in Algeria. Cli. Mic. Inf., 15 (2009)

350

167-168.

351

[12] W.I. Yahiaoui, F. Afri-Bouzebda, Z. Bouzebda, A. Dahmani, Sondage sérologique de la

352

fièvre Q chez les ovins par la méthode ELISA et prévalence des avortements dans la région de

353

Ksar El Boukhari (Algérie), Tropicultura, 32 (2013) 22-27.

354

[13] G. Masala, R. Porcu, G. Sanna, G. Chessa, G. Cillara, V. Chisu, S. Tola, Occurrence,

355

distribution, and role in abortion of Coxiella burnetii in sheep and goats in Sardinia, Italy, Vet.

356

Microbiol. 99 (2004) 301-305.

357

[14] V.M. Vaidya, S.V.S. Malik, K.N. Bhilegaonkar, R.S. Rathore, S. Kaur, S.B. Barbuddhe,

358

Prevalence of Q fever in domestic animals with reproductive disorders. Comp. Immun.

359

Microbiol. Infect. Dis. 33, (2010) 307-321.

360

[15] S.E. Sanford, G.K. Josephson, A. MacDonald, Coxiella burnetii (Q fever) abortion storms

361

in goat flocks after attendance at an annual fair. Can.Vet. J. 35 (1994) 376-378.

362

[16] E. Rousset, B. Durand, M. Berri, P. Dufour, M. Prigent, P. Russo, T. Delcroix, A.

363

Touratier, A. Rodolakis, M. Aubert, Comparative diagnostic potential of three serological tests

364

for abortive Q fever in goat flocks. Vet. Micobiol. 124 (2007) 286-294.

15

365

[17] A.L. Garcia-Perez, I. Astobiza, J.F. Barandika, R. Atxaerandio, A. Hurtado, R.A. Juste,

366

Investigation of Coxiella burnetii occurrence in dairy sheep flocks by bulk-tank milk analysis

367

and antibody level determination, J. Dai. Sci. 92 (2009) 1581-1584.

368

[18] E. Kennerman, E. Rousset, E. Gölcü, P. Dufour, Seroprevalence of Q fever (coxiellosis)

369

in sheep from the Southern Marmara Region, Turkey. Comp. Immun. Microbiol. Infect. Dis.

370

33 (2010) 37-45.

371

[19] M. Georgiev, A. Afonso, H. Neubauer, H. Needham, R. Thiery, A. Rodolakis, H. Roest,

372

K. Stark, J. Stegeman, P. Vellema, W. van der Hoek, S. More. Q fever in humans and farm

373

animals in four European countries, 1982 to 2010, Eurosurveillance Editorial Team. 8 (2013)

374

1-13.

375

[20] E. Rousset, M. Berri, B. Durand, P. Dufour, M. Prigent, T. Delcroix, A. Touratier, A.

376

Rodolakis, Coxiella burnetii shedding routes and antibody response after outbreaks of Q fever-

377

induced abortion in dairy goat flocks. App. Env. Mic. 75 (2009) 428-433.

378

[21] A. Alsaleha, J.L. Pellerin, A. Rodolakis, M. Larrata, D. Cochonneau, J.F. Bruyas, F. Fieni,

379

Detection of Coxiella burnetii, the agent of Q fever, in oviducts and uterine flushing media and

380

in genital tract tissues of the non pregnant goat. Comp. Immun. Microbiol. Infect. Dis. 34 (2011)

381

355-360.

382

[22] C. Dubuc-Forfait, E. Rousset, J.L. Champion, M. Marois, P. Dufour, E. Zerhaoui, R.

383

Thiéry, P. Sabatier. Démarche d’appréciation du risque d’excrétion de Coxiella burnetii dans

384

les troupeaux caprins laitiers dans le sud-est de la France, Epi. San. Ani. 55 (2009) 117-136.

385

[23] R. Kittelberger, J. Mars, G. Wibberley, R. Sting, K. Henning, G.W. Horner, K.M. Garnett,

386

M.J. Hannah, J.A. Jenner, C.J. Piggott, J.S. O'Keefe. Comparison of the Q-fever complement

387

fixation test and two commercial enzyme-linked immunosorbent assays for the detection of

388

serum antibodies against Coxiella burnetii (Q-fever) in ruminants: recommendations for use of

389

serological tests on imported animals in New Zealand. N. Z. Vet. J. 57 (2009) 262-268.

16

390

[24] J. Muskens, E. van Engelen, C. van Maanen, C. Bartels, T.J. Lam. Prevalence of Coxiella

391

burnetii infection in Dutch dairy flocks based on testing bulk tank milk and individual samples

392

by PCR and ELISA, Vet. Rec., 168 (2011) 79.

393 394

Table 1. Descriptive characteristics and results obtained by herd using ELISA and qPCR Herd

Herd characteristics abortion size antecede nt

compositi on

A1

Yes

˃100

mixed

A2

No

≤100

sheep

A3

Yes

≤100

sheep

A4

yes

≤100

mixed

A5

yes

≤100

mixed

A6

yes

˃100

sheep

A7

yes

˃100

sheep

A8

yes

≤100

mixed

A9

no

˃100

mixed

A10

no

≤100

sheep

A11

no

≤100

mixed

A12

yes

≤100

sheep

A13

yes

≤100

mixed

A14

no

≤100

mixed

A15

yes

≤100

sheep

A16

no

≤100

mixed

A17

no

≤100

sheep

Producti on system

Nb. of tested anima ls transhuma 12 nt transhuma 5 nt transhuma 7 nt transhuma 8 nt transhuma 9 nt transhuma 10 nt transhuma 16 nt transhuma 5 nt transhuma 9 nt transhuma 5 nt transhuma 7 nt transhuma 6 nt transhuma 8 nt transhuma 5 nt transhuma 6 nt transhuma 7 nt transhuma 8 nt

ELISA

qPCR

Nb. of positive cases (%) 0

Nb. of positive cases (%) 0

0

2 (40)

3 (42.9)

4 (57.1)

4 (50)

0

0

0

3 (30)

0

2 (12.5)

0

0

1 (20)

0

2 (22.2)

1 (20)

4 (80)

0

3 (42.9)

0

1 (16.7)

1 (12.5)

1 (12.5)

1 (20)

1(20)

2 (33.3)

0

0

0

1 (12.5)

0 17

B1 B2 B3 B4 C1 C2 C3 D1

yes yes yes no yes yes no no

≤100 ≤100 ≤100 ≤100 ≤100 ≤100 ≤100 ≤100

mixed mixed mixed sheep mixed mixed mixed sheep

D2

no

≤100

sheep

D3

no

˃100

sheep

E1

yes

˃100

sheep

E2 F1 F2 G1 G2 H1

no no no no yes no

≤100 ≤100 ≤100 ≤100 ≤100 ≤100

sheep sheep sheep sheep sheep sheep

H2

yes

≤100

sheep

Total

/

/

/

Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary Sedentary transhuma nt transhuma nt transhuma nt transhuma nt Sedentary Sedentary Sedentary Sedentary Sedentary transhuma nt transhuma nt /

8 7 6 4 7 8 6 5

nt. 1 (14.3) 1 (16.7) 0 1 (14.3) 2 (25) 0 nt.

3 (37.5) 0 3 (50) 0 7 (100) 3 (37.5) 0 0 (0)

8

1(12.5)

0

9

1 (11.1)

1 (11.1)

20

5 (25)

1 (5)

6 7 6 8 6 7

0 2 (28.2) 0 nt. nt. nt.

1 (16.7) 6 (85.7) 4 (66.7) 5 (62.5) 3 (50) 0

6

nt.

1 (20)

32 (14.1)

57 (21.3)

395 396 397 398

nt. not tested mixed: sheep and goats

399 400 401 402 403 404 405 406

18

407 408 409 410

Table 2. Statistical analysis in herds (region, abortion history, size, composition and

411

production system) and individuals (species, symptoms and age of females) Variables

ELISA Negative P cases value

Region

Positive cases (%) 18 (13.5) 2 (11.8) 3 (14.3)

Djelfa Biskra Bordj Bouareridj Skikda El Bayadh yes

2 (11.8) 5 (19.2) 2 (15.4)

15 21 11

12 (80)

3

no

6 (43)

8

≤100

13 (56.5) 4 (66.7)

10

4

0.26

mixed sedentary

10 (71.4) 7 (46.7) 5 (55.6)

8 4

0.82

transhumant

12 (60)

8

sheep

26 (12.2) 6 (15)

187

27 (16.2) 5 (8.4)

140

Medea Constantine Ain Defla

Abortion history

Size

˃100 Composition

Productive system

Species

sheep

goat Symptoms

abortion normal delivery

qPCR

115

0.83

15 18

nt. 0.06

0.65

2

0.62

34

55

0.13

Positive cases (%) 19 (14.3) 6 (24) 10 (47.6) 1 (4.5) 2 (7.7) 10 (76.9) 8 (57.1) 1 (21.3) 11 (61.1) 10 (58.8) 17 (70.1) 5 (45.4) 9 (45)

Negative P cases value

11

0.06

12 (80) 8 (66.7) 13 (56.5) 44 (19.7) 13 (30.2) 41 (22.3) 16 (22.9)

3 4

0.56

114

1.29

19 11 21 24 3 6 12 7

0.98

7 7

0.14

6

10 179

0.12

30 156

0.71

54

19

Age of females

primiparous

14 (20)

56

multiparous

18 (12.9)

139

0.08

15 (18.3) 42 (22.9)

67

0.39

141

412 413

Table 3. Concordance between bacterial excretion in the 3 serological class results

414

Serology Excretion PCR + PCR Total

ELISA -

ELISA +

ELISA ++

38 156 195

5 19 24

2 6 8

415

416 417

Fig. 1. Location of Algerian departments from which samples were collected (each black dot

418

represents a tested flock).

419 420

20

421 422 423 424 425 426 427 428 429 430 Number of animals 220

431 432

200 180 160

433

140 120

434 435

100 80 60

436

40 20

437 438 439 440

0

ELISA- (OD<40)

ELISA+ (40≤OD<100)

ELISA++ (OD≥100)

Optical Density categories

Fig. 2. Distribution of various ELISA classes

441 442 443 444 445 446 21

447 448 449 450 451 452 453 454 455 456 Number of animals 240

457 458

220 200 180

459 460

160 140 120

461 462

100 80 60

463 464

40 20 0

465 466 467

0-500

500-10E3

10E3-10E4

10E4-10E6

>10E6

Number of bacteria per swab

Fig. 3. Distribution of various qPCR classes

468 469 470 471

22