www.elsevier.com/locate/vetmic

Veterinary Microbiology 118 (2006) 101–106

Diagnosis of Coxiella burnetii-related abortion in Italian

domestic ruminants using single-tube nested PCR

Antonio Parisi a,*, Rosa Fraccalvieri a, Mariassunta Cafiero a, Angela Miccolupo a,Iolanda Padalino a, Cosimo Montagna a, Federico Capuano b, Roldano Sottili a

a Istituto Zooprofilattico Sperimentale della Puglia e della Basilicata, V. Manfredonia 20, 71100 Foggia, Italyb Istituto Zooprofilattico del Mezzogiorno, V. della Salute 2, 80055 Portici (NA), Italy

Received 31 March 2006; received in revised form 19 June 2006; accepted 26 June 2006

Abstract

Coxiella burnetii, an obligate intracellular parasite with a worldwide distribution, is the causative agent of acute and chronic

Q fever in humans. Although infection is often unapparent in cattle, sheep and goats, there is increasing evidence that C. burnetii

infection in these species is associated with abortion and stillbirth. This paper describes the introduction of a single-tube nested

PCR protocol for the diagnosis of C. burnetii-related abortion in domestic ruminants in Italy. A total of 514 aborted foetuses

from cattle (n = 138) and sheep and goat (n = 376), collected from 301 farms, were analyzed from January 2001 to March 2005.

Ninety-seven of 514 (18.9%) animals tested PCR-positive, with 16/138 (11.6%) cattle and 81/376 (21.5%) sheep and goat.

Eleven of 102 (10.8%) farms with reproductive disorders in cattle and 37/199 (18.6%) farms with reproductive disorders in

sheep and goats were infected with C. burnetii. A greater incidence was observed in three of the seven investigated provinces

( p < 0.01), with rates of infected farms of up to 23.8%. Data showed that almost all the C. burnetii-related abortions were

recorded between October and April ( p < 0.01). These findings suggest that Q fever in humans is largely underestimated in

Italy, probably because its occurrence is obscured by flu-like symptoms in acute forms.

# 2006 Elsevier B.V. All rights reserved.

Keywords: Abortion; Single-tube nested PCR; Coxiella burnetii; Italy

1. Introduction

Coxiella burnetii, an obligate intracellular parasite

with a worldwide distribution, is the causative agent of

acute and chronic Q fever in humans and abortions and

* Corresponding author. Tel.: +39 0804057858;

fax: +39 0804057753.

E-mail address: parisi.izs@tin.it (A. Parisi).

0378-1135/$ – see front matter # 2006 Elsevier B.V. All rights reserved

doi:10.1016/j.vetmic.2006.06.023

infertility in animals (Marrie, 1990, 1995; Maurin and

Raoult, 1999).

Ticks are considered to be the natural primary

reservoirs of C. burnetii responsible for the spread of the

infection in wild animals and for transmission to

domestic animals (Norlander, 2000). Cattle, sheep and

goats are the main sources of human infection (Lang,

1990). Infected animals shed highly stable bacteria in

urine, faeces, milk and through placental and birth

.

A. Parisi et al. / Veterinary Microbiology 118 (2006) 101–106102

Fig. 1. Area of the research. Examined farms were located in two

Italian regions, (&) Puglia (1–5) and ( ) Basilicata (6–7), including

seven provinces.

fluids. Infectionvia inhalation of aerosolized organisms

or ingestion of raw milk or fresh dairy products has been

reported in humans and animals (Fishbein and Raoult,

1992; Tissot Dupont and Raoult, 1993). The highest risk

factor for acquisition of Q fever is exposure to parturient

infected animals (Raoult and Marrie, 1995). In the

summer and autumn of 1993, a large outbreak of human

Q fever occurred in north-eastern Italy involving 58

people during a 5-months period. The only risk factor

for acquisition of Q fever in this population was expo-

sure to migrating sheep flocks (Selvaggi et al., 1996).

Although infection is often unapparent in cattle,

sheep and goats, there is increasing evidence that C.

burnetii infection in these species is associated with

abortion and stillbirth (Aitken, 1998). C. burnetii may

cause abortion in sheep and goats; abortion in cows is

believed to be rare even if reproductive disorders and

mastitis can occur (To et al., 1998).

Routine diagnosis of Q fever is usually made by

serological tests including immunoflorescence, com-

plement fixation and enzyme-linked immunosorbent

assay (ELISA) (Peter et al., 1987), which have the

potential disadvantage of indicating exposure rather

than detecting the organism. The polymerase chain

reaction (PCR) technique has become a useful tool to

detect C. burnetii in biological samples (Berri et al.,

2000; Stein and Raoult, 1992). A PCR assay performed

with primers based on a repetitive, trasposon-like

element (Trans PCR) (Willems et al., 1994) has proved

to be highly specific and sensitive for the laboratory

diagnosis of C. burnetii infections, as it detects even

very few copies of a specific DNA sequence.

The aim of this study was to establish a single-tube

nested PCR protocol to enhance the sensitivity of a

diagnostic PCR test (Trans PCR) in detecting C.

burnetii. Two successive amplifications were per-

formed in a single tube using a thermal profile which

selectively extended first the external, then the internal

primer pair. This kind of approach has been used

successfully to improve sensitivity in the diagnosis of

infectious disease (Fournier and Raoult, 2003; Erlich

et al., 1991; Tilston and Corbitt, 1995).

2. Materials and methods

Total genomic DNA of C. burnetii Nine Mile strain

(Dade Behring Inc.) was serially diluted and used to

study the sensitivity of two PCR protocols: Trans PCR

and single-tube nested PCR. From January 2001 to

March 2005, 514 aborted foetuses from 301 farms

were collected and analyzed. The farms were located

in two southern Italian regions: Puglia and Basilicata

(Fig. 1). Different samples, mainly brain, liver and

placenta, were analyzed from each animal, based on

reports in the literature that these tissues were good

targets for the diagnosis of C. burnetii-related abortion

(Masala et al., 2004). Animals in which at least one

target tested PCR-positive were considered to be

infected. DNA was extracted from tissues using the

QIAmp DNATissue kit (Qiagen S.A.) according to the

instructions of the manufacturer. Five microlitres of

DNA solution were used in a total volume of 50 ml.

The Trans PCR assay was performed as described by

Berri et al. (2000). The internal primers, 261F and

463R, were designed on the basis of the sequence

available in GenBank (M80806) using DNAMan

software (Linnon Biosoft). Table 1 shows the primer

sequence and PCR conditions. The final reaction

mixture for Trans PCR consisted of 1 mM of each

primer, and for single-tube nested PCR of 0.5 mM of

each internal primer and 0.005 mM of each external

primer, 1 ml of ACGU + T dNTP Mix (SIGMA),

0.5 unit of UDG (SIGMA), 1U of Taq DNA

polymerase (Eppendorf) and 1X Taq buffer (Eppen-

dorf). The PCR protocols consisted of a preliminary

incubation at 37 8C for 10 min to remove possible

carryover contaminations followed by a denatura-

tion step at 94 8C for 10 min to inactivate UDG.

A. Parisi et al. / Veterinary Microbiology 118 (2006) 101–106 103

Table 1

Sequences of primers and PCR conditions

Protocol Primer Sequence Gene Amplimer

length (bp)

PCR conditions (8C/s) Number of

PCR cyclesDenaturation Annealing Extension

Trans Trans1 50TATGTATCCACCGTAGCCAGT C30 IS1111 687 94/30 66–62/30a 72/60 5

PCR Trans2 50CCCAACAACACCTCCTTATTC30 61/30 35

Single Trans1 50TATGTATCCACCGTAGCCAGTC30 IS1111 687 94/30 66–62/30a 72/60 5

Tube Trans2 50CCCAACAACACCTCCTTATTC30 61/30 10

Nested 261 F 50GAGCGAACCATTGGTATCG30 203 94/30 54/20 72/20 35

463 R 50CTTTAACAGCGCTTGAACGT30

a Annealing temperature was decreased �1 8C for each cycle of amplification.

The samples were then subjected to different thermal

cycle conditions (Table 1). The PCR reactions were

performed in an automated DNA thermal cycler

(GeneAmp 9700, Applied Biosystem). Five micro-

litres of PCR products were examined by electro-

phoresis in 1.5% agarose gel, stained with ethidium

bromide, and photographed under UV transillumina-

tion to visualize the bands. Fragments size was

calculated by Totalab 2.0 (Nonlinear dynamics) with a

100 bp ladder as standard size marker (New England

Biolabs). Statistical analysis was performed with

StatView 5.0 software (SAS Institute Inc.). Chi-square

test was performed in order to analyze nominal

variables.

Fig. 2. Comparison of the sensitivity of the PCR protocols. (I) Trans

PCR. (II) Single-tube nested PCR. In the first lane 100 bp molecular

ladder. (A–F) dilutions of C. burnetii Nine Mile strain: (A) 1 pg; (B)

500 fg; (C) 250 fg; (D) 100 fg; (E) 50 fg; (F) 5 fg; (G) no DNA.

3. Results

Analysis of samples from serial dilutions of C.

burnetii genomic DNA established a detection limit of

50 fg for the Trans PCR, although a faint band was

detected for the last dilution, and of 5 fg for single-

tube nested PCR (Fig. 2). The single tube protocol

amplified a single band of the expected size (203 bp).

The results of PCR detection in the clinical samples

are illustrated in Table 2. Out of a total of 514

examined animals, 97 (18.9%) tested PCR-positive,

16/138 (11.6%) cattle and 81/376 (21.5%) sheep and

goats. A statistically significant difference ( p < 0.01)

was found between PCR-positive cattle as compared

to PCR-positive sheep and goats. A total of 301 farms

presenting abortion were investigated. Farms where

even a single animal tested PCR-positive were

considered to be infected. Eleven of 102 (10.8%)

farms presenting reproductive disorders in cattle and

37/199 (18.6%) farms presenting reproductive dis-

orders in sheep and goats were infected with C.

burnetii. On one farm C. burnetii-related abortions

occurred concomitantly in cattle, sheep, and goats.

There was no statistically significant difference

( p > 0.05) in the prevalence of C. burnetii infection

on cattle farms and on sheep–goat farms (Table 2). A

greater incidence was observed in Provinces 2, 3 and 4

with overall rates of PCR-positive farms of 23.8%,

17.4% and 21.3%, respectively ( p < 0.01) (Table 2).

Data showed that almost all the C. burnetii-related

abortions were recorded between October and April

( p < 0.01) (Table 3). Bacteriological analyses were

carried out to check for the main abortifacient agents

such as Brucella spp., Salmonella spp., Listeria

monocytogenes, and Campylobacter spp. All the

samples analyzed were negative. PCR analyses

were carried out to detect Chlamydophila abortus,

A. Parisi et al. / Veterinary Microbiology 118 (2006) 101–106104

Table 2

PCR results of examined animals and farms, broken down according to species and province of origin

Province Animals Farms

Sheep/goat Cattle Sheep/goat Cattle

Examined PCR+ (%) Examined PCR+ (%) Examined PCR+ (%) Examined PCR+ (%)

1 45 3 (6.7) 19 1 (5.3) 25 2 (8.0) 14 1 (7.1)

2a 132 43 (32.6) 76 13 (17.1) 57 18 (31.6) 52 8 (15.4)

3 25 6 (24.0) 11 1 (9.1) 16 3 (18.8) 7 1 (14.3)

4 69 24 (34.8) 15 1 (6.7) 33 9 (27.3) 14 1 (7.1)

5 4 1 (25.0) 0 0 (0.0) 3 1 (33.3) 0 0 (0.0)

6 83 3 (3.6) 9 0 (0.0) 51 3 (5.9) 8 0 (0.0)

7 18 1 (5.6) 8 0 (0.0) 14 1 (7.1) 7 0 (0.0)

TOTAL 376 81 (21.5) 138 16 (11.6) 199 37 (18.6) 102 11 (10.8)

a On one farm all the species (cattle, sheep, goat) had reproductive disorders.

Table 3

Frequency distribution of examined samples during survey period

Month Animals

Sheep/goat Cattle

Examined PCR+ Examined PCR+

January 52 10 10 2

February 67 18 17 0

March 95 28 29 7

April 82 8 20 3

May 13 0 12 0

June 5 1 5 0

July 9 0 6 0

August 1 0 4 0

September 5 0 7 0

October 19 9 6 0

November 14 1 15 2

December 14 6 7 2

Toxoplasma gondii and Neospora caninum. Twenty-

eight animals (sheep and goats from 15 farms) were

PCR-positive for C. abortus, seven of them from three

farms were also PCR-positive for C. burnetii. Ten

bovine foetuses from nine farms were PCR-positive

for N. caninum and five animals from three sheep and

goat farms were PCR-positive for T. gondii. No

association was observed among C. burnetii, N.

caninum and T. gondii.

4. Discussion

Abortions constitute an important economic burden

for the livestock industry due to death loss and

decreased milk production. The incidence of infec-

tious abortions is very high throughout Italy, although

no reliable data are available (Capuano et al., 2004).

Our results indicate that C. burnetii plays an

important role in abortion of sheep and goats, whereas

these events are less frequent in cattle. Statistically

significant differences in the prevalence of C. burnetti-

related abortion ( p < 0.01) of cattle and sheep and

goats are ascribable to the different clinical manifes-

tations of the disease which has an enzootic trend in

sheep and goats but occurs only occasionally in cattle.

This finding is supported by the fact that there was no

statistical difference between the results from cattle

farms and from sheep and goat farms ( p > 0.05).

Compared to previous studies in Sardinia (Italy), we

detected a higher incidence of C. burnetii in the region

we examined (Masala et al., 2004).

C. burnetii is widespread in southern Italy since the

agent has been detected in 15.9% of the farms

investigated with at least one event of abortion

associated with C. burnetii. On one farm C. burnetti-

related abortion concomitantly affected cattle, sheep

and goats. A higher incidence was recorded in

Provinces 2, 3 and 4 (Fig. 2) comprising 83.3% of

the farms tested PCR-positive for C. burnetii.

Although animal coxiellosis is believed to be a tick-

borne disease, almost all the abortions were observed

during the colder months when ticks are not active. A

likely explanation for this is that even if ticks are an

important reservoir in the environment, C. burnetii is

widely present among domestic animals. Cattle

frequently shed C. burnetii in milk (Kim et al.,

2005), and small ruminants maintain carrier status for

a long period after abortion and are thus an important

A. Parisi et al. / Veterinary Microbiology 118 (2006) 101–106 105

source of infection for both humans and other animals

(Berri et al., 2002). Abortion is the clinical

manifestation of a widespread occurrence of C.

burnetii in the animal population and is mainly

concentrated during the reproductive season of small

ruminants.

As also reported by other authors (Sting et al.,

2000; Schopf et al., 1991), concomitant infection with

C. burnetii and C. abortus on three sheep and goats

farms was detected by PCR analysis in our investiga-

tion.

Isolation of C. burnetii is not performed routinely

for diagnostic purposes in veterinary medicine

because it is a tedious procedure. Cultural recovery

of C. burnetii in embryonated eggs or by cell culture is

time consuming, hazardous, expensive and requires

extensive laboratory support. Polymerase chain reac-

tion assays provide a valuable new approach which is

sensitive, easy to perform and safe for laboratory

personnel. The nested PCR approach has been used for

the diagnosis of human C. burnetii infection (Zhang

et al., 1998; Spyridaki et al., 1998), although this

method considerably increases the chances of PCR

contamination due to additional handling of ampli-

cons from the first amplification. In this study we

modified the Trans PCR method to enhance its

sensitivity in detecting C. burnetii DNA in animal

tissues. Single-tube nested PCR has been used

successfully in the diagnosis of infectious diseases

requiring a high sensitivity (Erlich et al., 1991; Tilston

and Corbitt, 1995). Recently, Fournier and Raoult

(2003) developed a single tube amplification protocol

to diagnose acute Q fever from human serum. This

method presents the advantages of the nested

procedure while minimizing the likelihood of con-

tamination. The nested primers were designed to have

a lower melting point than the external primers. The

optimal cycle profile was determined empirically

using a touchdown profile for the external amplifica-

tion, then the annealing temperature was lowered to an

optimal level for nested primer activity. Another focal

step of the amplification procedure was to optimize the

ratio of the external and internal primer concentration.

We found that the optimal primer ratio was 1:100, as

different ratios could be responsible for artefacts

during amplification. Moreover, to eliminate carry-

over contaminations, which can result in false positive

PCR reactions, we used ACGU + T dNTP Mix and

uracil DNA glycosylase. In addition to its extreme

sensitivity, the single-tube nested PCR technique

presents the advantage of amplifying a shorter region

than the Trans PCR method, thereby increasing the

chances of performing a diagnosis in tissues where

DNA may be degraded because of autolysis, as often is

the case in animal foetuses.

The Trans PCR assay proved to be a highly

sensitive method to detect C. burnetii DNA in animal

milk, faeces and tissues (Willems et al., 1994; Lorenz

et al., 1998; Berri et al., 2000). Comparison of the

sensitivities of the two PCR methods was made using

serially diluted C. burnetii pure genomic DNA. Our

protocol enhanced sensitivity over the Trans PCR in

detecting C. burnetii by 10- to 100-fold and it proved

to be an effective tool for the diagnosis of animal

coxiellosis. Finally, our study contributed further

insights into this important disease in Italy. The

findings of this investigation suggest that human Q

fever is largely underestimated in southern Italy,

probably because it is masked by flu-like symptoms in

acute forms (Norlander, 2000). More investigations

are necessary to elucidate the actual spread of C.

burnetii, the incidence of acute or chronic Q fever it

causes in humans and of abortion it causes in the

animal population.

Acknowledgement

This work was supported by a grant from Ministry

of Health (IZS/PB07/01).

References

Aitken, I.D., 1998. Clinical aspects and prevention of Q fever in

animals. Eur. J. Epidemiol. 5, 420–424.

Berri, M., Laroucau, K., Rodolakis, A., 2000. The detection of

Coxiella burnetii from ovine genital swabs, milk and fecal

samples by the use of a single touchdown polymerase chain

reaction. Vet. Microbiol. 72, 285–293.

Berri, M., Souriau, A., Crosby, M., Rodolakis, A., 2002. Shedding of

Coxiella burnetii in ewes in two pregnancies following an

episode of Coxiella abortion in a sheep flock. Vet. Microbiol.

85, 55–60.

Capuano, F., Parisi, A., Cafiero, M., Picaro, L., Fenizia, D., 2004.

Coxiella burnetii: what is the reality? Parassitol 46, 131–

134.

Erlich, H.A., Gelfand, D., Sninsky, J.J., 1991. Recent advances in the

polymerase chain reaction. Science 252, 1643–1651.

A. Parisi et al. / Veterinary Microbiology 118 (2006) 101–106106

Fishbein, D.B., Raoult, D., 1992. A cluster of Coxiella burnetii

infections associated with the exposure to vaccinated goats and

their unpasturized dairy products. Am. J. Trop. Med. Hyg. 47,

35–40.

Fournier, P.E., Raoult, D., 2003. Comparison of PCR and serology

assays for early diagnosis of acute Q fever. J. Clin. Microbiol. 41,

5094–5098.

Kim, S.G., Kim, E.H., Lafferty, C.J., Dubovi, E., 2005. Coxiella

burnetii in bulk tank milk samples. United States Emer. Infect.

Dis. 11, 619–621.

Lang, G.H., 1990. Coxiellosis (Q fever) in animals. In: Marrie, T.J.

(Ed.), Q fever: the disease, vol. 1. CRC Press, Boca Ranton, FL,

pp. 24–42.

Lorenz, H., Jager, C., Willems, H., Baljer, G., 1998. PCR detection

of Coxiella burnetii from different clinical specimens, especially

bovine milk, on the basis of DNA preparation with a silica

matrix. Appl. Environ. Microbiol. 64, 4234–4237.

Masala, G., Porcu, R., Sanna, G., Chessa, G., Cillara, G., Chisu, V.,

Tola, S., 2004. Occurrence, distribution, and role in abortion of

Coxiella burnetii in sheep and goats in Sardinia. Italy Vet.

Microbiol. 99, 301–305.

Marrie, T.J., 1990. Epidemiology of Q fever. In: Marrie, T.J. (Ed.), Q

fever: the disease. CRS Press, Boca Raton, FL, pp. 49–70.

Marrie, T.J., 1995. Coxiella burnetii (Q fever) pneumonia. Clin.

Infect. Dis. 21, 253–264.

Maurin, M., Raoult, D., 1999. Q Fever. Clin. Microbiol. Rev. 12,

518–553.

Norlander, L., 2000. Q fever epidemiology and pathogenesis.

Microbes Infect. 2, 417–424.

Peter, O., Dupuis, M., Peacock, M.G., Burgdorfer, W., 1987. Com-

parison of enzyme-linked immunosorbent assay and comple-

ment fixation and indirect fluorescent-antibody tests for

detection of Coxiella burnetii antibody. J. Clin. Microbiol.

25, 1063–1067.

Raoult, D., Marrie, T.S., 1995. Q Fever Clin. Infect. Dis. 20, 489–

495.

Selvaggi, T.M., Rezza, G., Scagnelli, M., Rigoli, R., Massu, M., De

Lalla, F., Pellizzer, G., Tramarin, A., Bettini, C., Zampieri, L.,

Belloni, M., Dalla Pozza, E., Marangon, S., Marchioretto, N.,

Togni, G., Giacobbo, M., Todessato, A., Binkin, N., 1996.

Investigation of a Q fever outbreak in Northern Italy. Eur. J.

Epidemiol. 12, 403–408.

Spyridaki, I., Gikas, A., Kofteridis, D., Psaroulaki, A., Tselentis, Y.,

1998. Q fever in the Greek island of Crete: detection, isolation,

and molecular identification of eight strains of Coxiella burnetii

from clinical samples. J Clin Microbiol. 36, 2063–2067.

Stein, A., Raoult, D., 1992. Detection of Coxiella burnetii by DNA

amplification using polymerase chain reaction. J. Clin. Micro-

biol. 30, 2462–2466.

Sting, R., Simmert, J., Mandl, J., Seemann, G., Bay, F., Muller, K.F.,

Schmitt, K., Mentrup, T., 2000. Coxiella burnetii infections and

infections with bacteria of the genus Chlamydia in dairy cattle.

Berl. Munch. Tierarztl. 113, 423–430.

Schopf, K., Khaschabi, D., Dackau, T., 1991. Enzootic abortion in a

goat herd, caused by mixed infection with Coxiella burnetii and

Chlamydia psittaci. Case report. Tierarztl. Prax. 19, 630–634.

Tilston, P., Corbitt, G., 1995. A single tube nested PCR for the

detection of hepatitis C virus RNA. J. Virol. Methods 53, 121–

129.

Tissot Dupont, H., Raoult, D., 1993. Epidemiologie de la fievre Q.

Bull. Epidem. Hebd. 5, 17–18.

To, H., Htwe, K., Kako, N., Kim, H.S., Yamagushi, T., Fukushi, H.,

Hirai, K., 1998. Prevalence of Coxiella burnetii infection in

dairy cattle with reproductive disorders. J. Vet. Med. Sci. 60,

859–861.

Willems, H., Thiele, D., Frohlich-Ritter, R., Krauss, H., 1994.

Detection of Coxiella burnetii in cow’s milk using the poly-

merase chain reaction (PCR). J. Vet. Med. Ser. 41, 580–587.

Zhang, G.Q., Hotta, A., Mizutani, M., Ho, T., Yamaguchi, T.,

Fukushi, H., Hirai, K., 1998. Direct identification of Coxiella

burnetii plasmids in human sera by nested PCR. J. Clin. Micro-

biol. 36, 2210–2213.