Molecular and Cellular Probes (1998) 12, 407–416

Article No. ll980205

Development of a PCR-based method coupled with a

microplate colorimetric assay for the detection of Porcine
Parvovirus and application to diagnosis in piglet tissues and
human plasma

C. Arnauld,1 O. Legeay,1 Y. Laurian,2 R. Thiery,1 M. Denis,1 P. Blanchard1 and

A. Jestin1
1
Centre National d’Etudes Vétérinaires et Alimentaires (CNEVA), Unité de Biologie
Moléculaire, BP 53, 22 440 Ploufragan, France and 2Centre Hospitalier Universitaire
de Bicêtre, Laboratoire d’Hématologeie, 78, rue du Général Leclerc,
94 275 Le Kremlin Bicêtre cedex, France

(Received 16 June 1998, Accepted 24 August 1998)

A new method for Porcine Parvovirus (PPV) diagnosis was developed. The method is based on
polymerase chain reaction (PCR) amplification followed by hybridization and colorimetric detection
of PCR products in microwell plates.
A highly specific and sensitive amplification step was ensured by primers carefully selected in
the VP2 structural gene and optimized PCR conditions. Uracyl-DNA-Glycosylase (UDG) in
combination with dUTP was used to avoid false-positive results, and 100 copies of internal control
(IC) were added to each PCR reaction to reveal any false-negative samples. Biotinylated amplified
fragments were hybridized on specific capture probes covalently linked to microwell plates.
Finally, the detection of hybridized PCR products was performed by means of a colorimetric
reaction, which was automated.
The method permitted the detection of 103 copies (6 fg) of replicative form DNA (RF-DNA) in
20 mg of lung sample, and 500 copies (3 fg) in 100 ll of plasma. It was used to analyse 24 field
piglet tissue samples, and 35 human plasma or serum specimens collected from patients treated
with porcine Factor VIII concentrates.  1998 Academic Press

KEYWORDS: Porcine parvovirus, PCR, microplate, detection, diagnosis, viral biosecurity.

INTRODUCTION

Porcine Parvovirus (PPV), a small DNA virus identified return to oestrus, reduced litter size, mummified foet-
in 1967,1 is recognized throughout the world as the uses, stillbirths and occasional abortions.5,6 Infections
major causative agent in a syndrome of reproductive of adult swine causes a viremia with mild fever that
failure in swine.2–4 Porcine Parvovirus infection of generally remains unnoticed.
non-immune sows, depending on the stage of preg- After decades of epizooties,3,7 PPV has now become
nancy, produces early embryonic death, delayed ubiquitous in most herds, with periodic outbreaks

∗ Author to whom all correspondence should be addressed at: Centre National d’Etudes Vétérinaires et Alimentaires (CNEVA), Unité
de Biologie Moléculaire, BP 53, 22 440 Ploufragan, France. Tel: +33 02 967 0130; Fax: +33 02 9678 6861.

0890–8508/98/060407+10 $30.00/0  1998 Academic Press

408 C. Arnauld et al.

which can be linked to herd management or vac- quantitation in a 1% agarose gel (Eurobio).20 From
cination failures.8,9 The virus may be recovered from the total length of PPV genome (4973 b),21 it was
the herd environment and also as a contaminant of considered that 1 pg of RF-DNA corresponds to ap-
cell cultures or commercial trypsin.10,11 Furthermore, proximately 1·7×105 copies, and the molecular con-
PPV is suspected to be present in commercial products centration of the solution could therefore be deduced.
of porcine origin, as for human Parvovirus B19 that The following virus strains were used to test speci-
is frequently identified in blood products.12–14 Hence, ficity: a plasmid (pEMBL) containing the entire Human
laboratory diagnosis is essential for the control of this Parvovirus B19 genome except for the two palin-
pathogen for both animal health and viral biosafety dromic ends, kindly provided by F. Morinet (Hôpital
concerns. St Louis, Paris, France); the original strain of MVM
Serological tests are routinely used to establish the (Minute Virus of Mice) isolated by Crawford22 was a
status pattern of susceptible herds, but are insufficient gift of A. Op De Beeck (Université Libre de Bruxelles,
to clearly signal an outbreak. Diagnosis has to be Rhode St Genèse, Belgium); a CPV strain (Canine
confirmed by the detection of PPV antigens in foetal Parvovirus) provided by M. Remond (CNEVA, Alfort,
lung either by immunofluorescence microscopy (IF), France); the 89 384 strain of DPV (Duck Parvovirus)
or by haemagglutination tests (HA).4,15 Two molecular given by V. Jestin (CNEVA, Ploufragan, France); the
biological methods based on the Slot-blot technique16 Nia3 strain of Pseudorabies Virus (PRV) provided by
and polymerase chain reaction (PCR)17 have pre- J.B. Mac Ferran (Belfast, Ireland).
viously been described. Our aim was to develop an
alternative method for PPV detection with improved
specificity and sensitivity.
Clinical samples
The method presented here combines the PCR with
specific measures to secure maximum reliability.
A total of 24 piglet tissue samples (lung and brain)
The two major drawbacks of PCR, carry-over con-
were supplied from the LVD 29 (Laboratoire Vét-
taminations and inhibitions of Taq Polymerase were
érinaire Départemental, Quimper, France). These
overcome respectively by using an enzymatic de-
samples had been collected after post mortem ex-
contamination system18 and adding an internal control
amination of mummified foetuses or stillborn piglets
(IC). Integration of a simple detection protocol and
from breedings units where PPV infection was sus-
automation were also taken into account in order to
pected. They were subjected to immunofluorescence
allow the easy routine use of the test. The assay
analysis as described by Mengeling et al.5 by the
is based on hybridization of the biotinylated PCR
LVD 29.
products on microtiter plates and on a final col-
Thirty-five human serum and plasma samples were
orimetric detection step using a conventional en-
obtained from five patients admitted to Bicêtre Hos-
zymatic kit. Two probes, specific to either viral or IC
pital (Kremlin-Bicêtre, France). The patients (one
targets, were coated in separate wells and allowed
severe A haemophiliac and four who developed auto-
the specific capture of PCR products from each sample
antibodies to factor VIII) received infusions of com-
spiked with IC.
mercial factor VIII of porcine origin. Subsequent
This new diagnostic tool was used for the analysis of
identification of PPV was done by the firm in the
piglet tissue samples and blood specimens collected
commercialized batches by cell culture and PCR
from human patients treated with porcine coagulation
(data not available). Blood specimens were collected
factor VIII concentrates.
before, during, and after infusions and were subjected
by us to PPV-PCR analysis.
MATERIALS AND METHODS

Virus strains DNA extraction

The NADL-8 strain of PPV, provided by the National Tissue samples from piglets were cut into pieces with
Animal Disease Center (Ames, Iowa, USA), was used sterile blades. The nucleic acids were extracted from
for test optimization. The virus was propagated in 20 mg fractions with the Rapid Genomic DNA Isol-
porcine kidney primary cell culture and a modified ation Kit for Cells and Tissues (Pharmacia Biotech),
Hirt procedure was performed to purify the double- according to the supplier’s instructions. After ethanol
stranded replicative form DNA (RF-DNA) as described precipitation, the DNA pellets were suspended in
by Molitor et al.19 The amount of double stranded 50 ll of TE buffer (10 m Tris-HCl, 1 m EDTA, pH
DNA was estimated by ethidium bromide fluorescent 8·0), and stored at −20°C prior to PCR analysis.
 PCR-based method for the detection of Porcine parvovirus 409

Nucleic acids from human blood samples were specificity, PCR amplification proceeded with 5 cycles
extracted according to the method published by of 94°C for 20 s, 57°C for 20 s, 72°C for 20 s, followed
Hornsleth et al.24 One hundred microlitres of serum by 35 cycles of 94°C for 15 s, 56°C for 20 s, 72°C for
or plasma were incubated for 1 hour at 50°C with 20 s and a final extension step at 72°C for 10 min.
400 ll of lysis buffer (50 m Tris-HCl pH 8, 100 m The absence of PCR carry-over contaminations was
NaCl, 10 m EDTA, 0·5% lauroyl sarcosine), and assessed by introduction, in each PCR run, of two
10 ll of Proteinase K buffer (50 m Tris-HCl pH 8, PCR negative controls in which DNA sample and IC
1 m CaCl2, 20 mg ml−1 Proteinase K). Lysates were were replaced by water.
extracted by phenol-chloroform-isoamylalcohol treat-
ment, followed by precipitation with cold 98% eth-
anol containing 0·2  LiCl. After centrifugation
Construction of the internal control
(15 min, 15 000 g), the supernatant was discarded.
The vacuum dried pellet was suspended in 50 ll of
The internal control (IC) molecule (220 bp) was gen-
TE buffer and stored at −20°C prior to PCR analysis.
erated by PCR using PPV strain NADL-8 DNA as
Two extraction controls were processed for each
template and the 3708-30 and SYIC primers. The
set of 10 samples. They were composed of lung or
sequence of the latter is composed of the 3906-28
plasma from specific pathogen free (SPF) swine from
primer at its 5′ end, a specifically rearranged 11 b
our experimental unit.24
sequence and 7 b complementary to the viral
sequence at its 3′ end (Fig. 1). The use of this primer
induces a deletion of 6 b. Its sequence is indicated
PCR primers
in Fig. 1. The PCR was set up in a total volume of
100 ll with 104 copies of PPV RF-DNA, 5 U of Taq
The PCR primers were designed from the sequence
Polymerase Eurobio (added at the end of the initial
published by Ranz et al.,21 with the two programs
denaturation step), (1X) EZ buffer, 1·5 m MgCl2
Oligo (Medprobe, Norway) and PC Rare
100 l of each dATP, dCTP, dGTP and dTTP, 0·4 l
(Griffais, Institut Pasteur Paris, France). The oligo-
of each primers, and 2 U of UDG. The following
nucleotides were synthesized using an automated 391
thermal programme was used: 10 min at 37°C, 10 min
DNA synthesizer (Applied Biosystems). The primers
at 95°C, 5 cycles of 94°C for 30 s, 56°C for 30 s,
coded 3708-30 (upper primer) and 3906-28 (lower
72°C for 30 s, followed by 35 cycles of 94°C for 20 s,
primer), both located within the VP2 structural gene,
55°C for 20 s, 72°C for 20 s and a final extension step
permitted amplification of a 226 bp PPV DNA frag-
at 72°C for 10 min. The product was purified using
ment. Their sequences are 3708-30: 5′-CCA GCA
Micron 100 columns (Amicon) and quantified by
GCT AAC ACA AGA AAA GGT TAT CAC-3′, and
optical density (OD) measurement at 260 nm. Con-
3906-28: 5′-GTC CAT GTT GGT AAT CCA TTG TAA
sidering the IC length, the number of IC copies in the
ATC T-3′. For PCR products microplate detection, the
solution was calculated as described above for PPV
3708-30 primer was 5′-labelled using Biotin Amidite
viral DNA. Serial dilutions in TE buffer were aliquoted
(Perkin Elmer) and coded 3708-30B.
and stored at −20°C prior to use.
In order to establish the validity of the IC, PCR
products resulting from co-amplification of viral and
PCR amplification
IC targets were analysed as previously described in
the laboratory.25,26
The PCR test was set up in a total volume of 20 ll
with 2·5 ll of DNA sample, 100 IC copies (see below),
(1X) EZ buffer (Eurobio), 1·5 m MgCl2, 200 l of
each dATP, dCTP and dGTP, 400 l dUTP (Boehringer Analysis of PCR products
Mannheim), 0·4 l of 3708-30B and 3906-28
primers, 0·4 U of UDG (Gibco BRL-Life technologies), Throughout optimization steps, PCR products from
1 U of Taq Polymerase Eurobio, 0·22 lg of TaqStart amplification without IC addition were analysed by
Antibody (a neutralizing monoclonal antibody to Taq electrophoresis on a 2·5% agarose gel (Eurobio)
DNA polymerase used as Hot Start technique) (Clon- stained with ethidium bromide. Polymerase chain
tech) and 0·8 ll of its specific dilution buffer. Using reaction products from samples spiked with IC were
the GeneAmp PCR system 9600 (Perkin Elmer), the analysed by hybridization and colorimetric detection
reaction mixture was held for 10 min at 37°C for in microtiter plates. Most of the parameters involved
UDG activity, then for 10 min at 95°C for UDG in the hybridization were subjected to optimization.
and TaqStart Antibody inactivation. For enhanced Optimized conditions are described here.
410 C. Arnauld et al.

*****
CAPV probe 3' CTTGGTTTACCGCACT 5'
deletion
5'...AATATAATGATGATGAACCAAATGGTGCTATAAGATTTACAATGGATTACCAACATGGACACT...3'

PPV"+"

3' ACTACTATTCTCGCGTTATTCTAAATGTTACCTGGTTGTACCTG 5' SYIC

Primer
rearranged sequence 3906-28 primer

3'...TTATATTACTACTACTTGGTTTACCACGATATTCTAAATGTTACCTAATGGTTGTACCTGTGA...5'
PPV"–"
PCR with 3708-30
and SYIC primers
*****
CAPIC probe 3' TTCTCGCGTTAGGCCT

...AATATAATGATGATAAGAGCGCAATAAGATTTACAATGGATTACCAACATGGAC 3'
Internal
control
...TTATATTACTACTATTCTCGCGTTATTCTAAATGTTACCTAATGGTTGTACCTG 5'

Fig. 1. Internal control (IC) production by polymerase chain reaction (PCR). Porcine Parvovirus (PPV)-DNA was
subjected to PCR with 3708-30 and SYIC primers. SYIC introduced the 3906-28 primer sequence, a 6 b deletion and a
11 b rearranged sequence at the 5′ end of the internal control (IC). This construct allowed subsequent amplification of
both viral and IC templates with 3708-30B and 3906-28 primers, and the discrimination of the two PCR product species
by hybridization to either CAPV or CAPIC probes. The sequences of these probes are shown here at their respective
hybridization positions. 5 b (∗) were added to their 5′ end in order to distance the hybridization region from the
microtiter plate bottom.

NH-Covalink 96-well Microtiter plates (Nunc) were The ELISA Amplification System kit (Gibco BRL-Life
prepared by covalent binding of the 5′-phos- Technologies) was used for the detection as indicated
phorylated capture probes (one PPV-sequence spe- in the supplier’s instructions except for the con-
cific coded CAPV and one IC-sequence specific coded centration of the streptavidine-alkaline phosphatase
CAPIC as indicated in Fig. 1). In each well, 625 ng conjugate in TBS (0·05  Tris-HCl pH 7·5, 0·15 
of a single specific probe was introduced, and the NaCl) which was adjusted to 500 ng ml−1. Final results
coating was performed using the method described were obtained by OD measurement at 492 nm. Coat-
by Legeay et al.27 with a minor modification for ing of the probes was assessed in each mitroplate by
the final washing solution which contained 1X SSC the assay of 5·1010 copies of the 5′ biotin-labelled
(0·15  NaCl, 0·015  sodium citrate, pH 7·0) instead ASCAPV and ASCAPIC probes added to six wells
of 6X SSC. After washing, the microplates were filled containing the CAPV and CAPIC probes, respectively.
with TE buffer and stored at 4°C for several months. Their sequences (ASCAPV: 5′-Biot-GAA CCA AAT
Prior to use, TE buffer was replaced by 100 ll of GG-3′, ASCAPIC: 5′-Biot-AAG AGC GCA AT-3′) are
hybridization buffer (1X SSC, 5X Denhardt’s, complementary to the sequences of CAPV and CAPIC
100 lg ml−1 of denatured sonicated salmon sperm probes, respectively.
DNA). Preincubation was performed for 30 min at All steps, from the hybridization to the OD reading,
37°C. The solution was discarded and the wells were were performed using the Microlab-FAME automate
filled with 5 ll of PCR products (previously diluted (Hamilton, Switzerland).
1/5 in water, denatured for 10 min at 95°C and placed
on ice for 10 min) added to 95 ll of fresh hybridization
buffer. Each PCR product was dispensed in two wells RESULTS
(one CAPV and one CAPIC specific). The hy-
bridization was performed for 10 min at 37°C. The PCR sensitivity and specificity
microwells were washed twice with 300 ll of solution
I (0·5X SSC, 0·1% Tween 20) with a 2 min soaking Ten-fold serial dilutions of NADL-8 strain DNA in TE
time at room temperature (RT), then once with 300 ll buffer were used to test the optimized conditions of
of solution II (0·1 X SSC, 0·1% Tween 20) with a the PCR assay. As little as one to ten copies of
3 min soaking time at RT, and finally with 300 ll of double stranded DNA have been detected (Fig. 2).
solution II with a 3 min incubation time at 37°C. No contamination in the PCR negative controls was
 PCR-based method for the detection of Porcine parvovirus 411

RF-DNA copies). Nucleic acid extractions and amp-

lifications were performed as described in Materials
and Methods. Up to 103 copies (6 fg) of RF-DNA
added to 20 mg of lung and 500 copies added to
100 ll of plasma or serum could be detected (data
not shown).

Optimization of the microtiter plate

hybridization
Fig. 2. Sensitivity of the polymerase chain reaction
(PCR) test. PCR products were amplified from serial The hybridization was optimized in order to improve
dilutions of PPV RF-DNA (104 copies to 1 copy) and detection sensitivity and specificity. Optimal strin-
subjected to electrophoresis on a 2·5% agarose gel. (0):
PCR negative control. (M): molecular weight marker PBR
gency of the hybridization mix was chosen by testing
322/Hae III. five solutions with final SSC concentrations ranging
from 6X to 1X. Significant enhancement of the OD
observed. None of the viral strains tested for specificity was obtained with the mixture containing 1X SSC
(B19, MVM, CPV, DPV and PRV) was amplified (data (data not shown).
not shown). Different amounts of the CAPV probe were used
to coat microwells (125, 250, 375, 500 and
625 ng well−1). As shown in Fig. 3, high specific
Evaluation of the extraction and amplification signals (i.e. OD for PCR products from virus in CAPV
protocols wells) were obtained with 250 and 625 ng well−1.
Six-hundred and twenty-five ng of probe per well was
Lung and plasma samples from SPF swine were spiked selected because it permitted an important decrease
with 10-fold serial dilutions of viral DNA (104–10 of the cross-reaction signals (i.e. OD for IC-PCR

2.5

2
OD 492 nm

1.5

0.5

0
125 250 375 500 625
–1
CAPV probe (ng well )

V-PCR (1) V-PCR (2) IC-PCR (1) IC-PCR (2)

Fig. 3. Test of various amounts of virus-specific probe per well for optimization of microtiter plate hybridization.
Polymerase chain reaction (PCR) products obtained from amplification of 104 copies of virus (V-PCR) or internal control
(IC-PCR) were tested in separate wells coated with variable amounts of virus specific probe (CAPV). The results of two
independent experiments (1,2) are traducted in optical density (OD) measurement at 492 nm. The data show that
625 ng well−1 yielded high specific signal (V-PCR in CAPV wells) and low cross-reaction signal (IC-PCR in CAPV wells).
412 C. Arnauld et al.

2.5

–57%
1.5
OD 492 nm

–39%
1
–33%

0.5 –81% –80%

–71%
–12%
0
104 103 102 T-PCR
Initial virus or IC copy number

Pure V-PCR 1/5 V-PCR Pure IC-PCR

1/5 IC-PCR Pure T-PCR 1/5 T-PCR

Fig. 4. Test of diluted polymerase chain reaction (PCR) products for optimization of microtiter plate hybridization. PCR
products obtained from amplification of 104, 103 and 102 copies of virus (V-PCR) or internal control (IC-PCR) were
tested pure or 1/5 diluted in water in wells coated with virus specific probe (CAPV). The results are traducted in optical
density (OD) measurement at 492 nm. 1/5 dilution of PCR products allowed a more significant decrease of the cross-
reaction signal (IC-PCR in CAPV wells) than the decrease of the specific signal (V-PCR in CAPV wells). T-PCR: negative
PCR control.

products in CAPV wells) in two independent ex- 0·130, respectively, in wells containing the CAPV
periments. The same results were obtained when probe. For nine assays of PCR products issued from
working with CAPIC wells. 100 IC copies, OD values of 0·098±0·100 were
Detection in CAPV wells of pure and 1/5 diluted obtained with the CAPV probe, and 0·519±0·162
PCR products obtained by amplification of 104, 103 with the CAPIC probe. This latter result determined
and 102 copies of viral RF-DNA and IC was also the choice of 100 IC copies as the optimal quantity
compared. Despite a loss of specific signal (57, 39 to add to each PCR tube.
and 33%, respectively), dilution of the PCR products
allowed to strongly decrease the cross-reaction signal
(81, 80 and 71%, respectively) (Fig. 4).
The time requirement for hybridization was op-
timized by testing for 2 h, 1 h, 30 min, 20 min and Data analysis
10 min. The shortest time (10 min) clearly improved
both signal sensitivity and specificity (data not shown). A cut-off value (0·135) was determined as the mean
Finally, the washing parameters (saline concentrat- plus three standard deviations of 23 individual OD
ion of the solutions, temperatures and soaking times) values obtained from extraction controls in CAPV
were optimized by testing increasing conditions of wells. Samples were considered positive when the
stringency. The most stringent protocol was selected OD exceeded the cut-off in the CAPV well, negative
(data not shown). when the OD was higher than the cut-off in the
Gathering optimized conditions, it was possible to CAPIC well and less than it in the CAPV well, and
detect PCR products amplified from 104 to 10 copies non-interpretable when the OD was below 0·135 in
of viral DNA, with OD values ranging from 2·500 to both wells.
 PCR-based method for the detection of Porcine parvovirus 413

Table 1. Analysis of 24 piglet tissue samples by The presence of PPV sequences was detected in six
polymerase chain reaction (PCR) and microplate of the 15 samples collected during the treatment
detection. period from four of the five patients studied.
CAPV CAPIC PPV test

+ − + − + − NI
DISCUSSION
1 23 24 0 1 23 0
Until now, PPV has been diagnosed by cell
Porcine Parvovirus (PPV) test was considered positive for
CAPV+ result [i.e. optical density (OD) higher than the cut-off culture,29 haemagglutination (HA)3 or immuno-
in CAPV well], negative for CAPV− and CAPIC+ result (i.e. fluorescence (IF).8 The PCR test developed by Molitor
OD lower than the cut-off in CAPV well and higher in CAPC et al.,17 significantly improved PPV detection in both
well) (non-interpretable) was noticed when both OD were
below the cut-off. mammalian cell lines and foetal tissue samples, but
did not include essential procedures for the prevention
of PCR carry-over contamination and enzymatic in-
hibition. Routine use of this PCR test may therefore
Table 2. Analysis of 35 human serum or plasma lead to false positive or negative results. In addition,
samples by polymerase chain reaction (PCR) and conventional techniques, specially IF, are still used
microplate detection.
as reference methods in diagnostic laboratories.
CAPV CAPIC PPV test In the present study, we focused on the de-
velopment of a PCR assay coupled with hybridization
+ − + − + − NI and colorimetric detection of the amplification prod-
ucts in 96 microwell plates. Like Molitor et al.,17 the
Before 0 6 6 0 0 6 0
treatment
target sequence was chosen within the structural gene
During 6 9 15 0 6 9 0 VP2, because of its low rate of nucleotidic homology
treatment between Parvovirus of the same group (PPV, MVM,
After 0 14 13 1 0 13 1 CPV).21 Initial experiments using candidate primers
treatment selected upon physical and thermodynamic criteria31
Samples were separated in three groups corresponding to led to non-specific amplifications from porcine gen-
collection before, during and after treatment with porcine Factor omic sequences. The authors believe that this lack
VIII. Porcine Parvovirus (PPV) tests were read as described for of specificity was not only linked to low melting
Table 1.
temperatures of the primers due to the PPV genome
composition (37% GC), but also to their own
sequence design compared with porcine sequences.
Assay of clinical samples An extensive research for specific primers was there-
fore carried out using a statistical strategy as previously
The results obtained for samples of porcine and described.32,33 In addition, the selected primers were
human origin are presented in Tables 1 and 2, re- lengthened with a few PPV specific bases at their 5′
spectively. end. These two modifications resulted in significant
It could be deduced from OD higher than the cut- improvement of both the specificity and sensitivity of
off in wells containing the IC specific probe, that the amplification. Efficient amplification of one to 10
there was no inhibition of the PCR amplification in copies of the PPV NADL-8 strain, and the absence
the porcine tissue samples, whereas inhibition was of amplification of B19, MVM, CPV and PRV DNA,
apparent in one plasma sample which was therefore confirmed the specificity of this couple of primers.
considered non-interpretable. Internal control systems have been introduced by
Immunofluorescence test results from the LVD 29 several authors into diagnostic tests based on PCR,
were available for 16 of the 24 porcine samples. With to monitor amplification efficiency.26,27,34–36 A similar
our method, one positive sample was found out of system was developed to identify any false negative
the four positive results obtained by immuno- results. The IC construction presented similarities with
fluorescence. other previously described constructs.26,27 A short 6 b
Samples of human origin were divided into three deletion and a specifically rearranged 11 b sequence
groups corresponding to collection before, during and were located between the two PCR primers. Hence,
after treatment with porcine factor VIII, respectively the construction allowed (1) amplification of both
(Table 2). All the specimens taken before and after viral and IC targets with a unique pair of primers, (2)
treatment were found to be negative with our method. evaluation of respective amplification of the two target
414 C. Arnauld et al.

species by electrophoresis owing to the deletion and of the specific signal intensity, therefore leading to a
(3) routine discrimination of PCR products originating loss of detection sensitivity.
from IC or virus, by hybridization of either mutated Finally, our method was tested using clinical speci-
or wild type 11 b sequences, respectively. mens of porcine or human origin. One of the four
The results obtained with clinical samples dem- piglet tissue samples found positive by IF also gave
onstrated the efficiency of our IC system, and espe- a positive result with PCR. It is hypothesized that this
cially the absence of amplification of IC spiked to disparity between the results of the two methods may
one particular plasma sample (Table 2). Polymerase be due: (1) to the interpretation of IF results which is
chain reaction carry-over contaminations were pre- known to be difficult and often subjective or (2) to a
vented by the addition of UDG in combination with lack of reliability of the tissue sampling step in our
dUTP instead of dTTP into the PCR mixes. As already method, because the extaction kit allowed the treat-
noted in a previous report,18 the destruction of un- ment of small tissue samples in which the homo-
desirable PCR products proved to be very efficient geneity of the viral location is unknown. However,
here because the high AT content of the PPV matrix our protocol permitted the detection of 6 fg of viral
resulted in a high degree of dUTP incorporation and, RF-DNA in 20 mg of tissue sample, whereas 100 fg
consequently, of UDG activity sites. Thus, sim- was detected by Molitor et al.,17 who also showed
ultaneous application of UDG/dUTP and IC systems that their method was 100–1000 fold more sensitive
was essential to ensure the maximum reliability of than IF.
this PCR assay intended for routine use. Polymerase chain reaction analysis of the human
A microwell hybridization assay was chosen for samples revealed the presence of PPV in 40% of
the final detection of PCR products. The advantages the blood specimens collected within the period of
of this methodology compared with agarose gel contaminated Factor VIII infusion. In contrast, samples
electrophoresis have already been emphasized as collected from 9 days to several months later (data
regards specificity, security, easy treatment of numer- not shown) were all negative. The presence of PPV
sequences could be the consequence of the dilution
ous samples and use of equipment and automates
of the initial viral load in the whole plasmatic volume.
dedicated to immunoassays.37–40 In addition, this
However, we could not definitely exclude the hypo-
method provides a simple protocol for the dis-
thesis of PPV persistence in human, up to the 9th day
crimination of PCR products amplified from either
post infusion.
virus or IC. In several cases, microwell hybridization
Other applications, such as screening of permanent
has been reported to be more sensitive than visu-
cell lines and biological products of porcine origin,
alization in agarose gel.36,41,42 The authors found that
could be developed. This method could also be
the cut-off value determined for objective readings
considered in xenotransplantation studies in which
(0·135) allowed detection of 10–100 initial viral cop-
biosecurity is a permanent concern.43
ies while 1–10 copies could be detected by agarose
gel electrophoresis. This result is thought to be a
consequence of the hybridization parameters which, ACKNOWLEDGEMENTS
in order to avoid cross-reactions and to improve the
dynamic range of the colorimetric signal, had to be The authors are grateful to F. Salingardes (LVD 29, Quim-
strongly optimized. These cross-reactions and the low per, France) who supplied piglet tissue samples.
intensity signal could be due: (1) to the low GC
content of the target sequence (a similar method
developed for the detection of a virus with a GC REFERENCES
content of 50% was optimized with much less strin- 1. Cartwright, S. & Huck, R. A. (1967). Viruses isolated
gent conditions;27 (2) to the short length of the specific in association with herd infertility, abortions and still-
probes (11 b) [it has recently been proved that even births in pigs. Veterinary Research 81, 196–97.
a small extension of the probes (2 b) may significantly 2. Mengeling W. L., Cutlip, R. C., Wilson, R. A., Parks,
B. & Marshall, R. F. (1975). Fetal mummification as-
enhance the hybridization specificity];26 or (3) to
sociated with porcine parvovirus infection. Journal of
competitive hybridization reactions between the dif- the American Veterinary Medical Association 166,
ferent species of PCR product strands and the probe. 993–95.
It was observed that a simultaneous decrease in hy- 3. Van Leengoed, L. A., Vos, J., Gruys, E., Rondhuis, P. &
bridization duration, an increase in probe con- Brand, A. (1983). Porcine parvovirus infection: review
and diagnosis in a sow herd with reproductive failure.
centration and the dilution of PCR products was The Veterinary Quarterly 5, 131–41.
essential to control these competitive reactions. How- 4. Molitor, T. W. & Joo, H. S. (1990). Clinical and patho-
ever, the latter parameter also resulted in a decrease logical features of porcine-parvovirus-related disease
 PCR-based method for the detection of Porcine parvovirus 415

and its diagnosis. In Handbook of Parvoviruses, vol Casal, J. I. (1989). Porcine parvovirus: DNA sequence
II. (Tjissen P., ed.) Pp. 135–150. Boca Raton: CRC and genome organization. Journal of General Virology
Press. 70, 2541–553.
5. Mengeling, W. L. & Cutlip, R. C. (1976). Reproductive 22. Crawford, L. V. (1966). A minute virus of mice. Virology
disease experimentally induced by exposing pregnant 29, 605–12.
gilts to porcine parvovirus. American Journal of Vet- 23. Hornsleth, A., Carlsen, K. M., Christensen, L. S., Gun-
erinary Research 37, 1393–394. destrup, M., Heegaard, E. D. & Myhre, J. (1994).
6. Jordan, E. K. & Sever J. L. (1994). Fetal damage caused Estimation of serum concentration of parvovirus B19
by parvoviral infections. Reproductive Toxicology 8, DNA by PCR in patients with chronic anaemia. Re-
161–89. search in Virology 145, 379–86.
7. Vannier, P., Tillon, J. P., Cariolet, R. & Madec, F. (1984). 24. Cariolet, R, (1986). Bilan de 10 années d’utilisation de
A seroepizootiological study of parvovirus in pig herds. porcs exempts d’organismes pathogènes spécifiques
Zentralblatt für Veterinar medizin 31, 36–45. (EOPS) à la station de pathologie porcine de Plou-
8. Mengeling, W. L., Lager, K. M., Zimmerman, J. K., fragan. Journées de la Recherche Porcine en France
Samarikermani, N. & Beran, G. W. (1991). A current 18, 321–30.
assessment of the role of porcine parvovirus as a cause 25. Thiery, R., Pannetier, C., Rziha, H.-J. & Jestin, A.
of fetal porcine death. Journal of Veterinary Diagnostic (1996). A fluorescence-based quantitative PCR method
Investigation 3, 33–35. for investigation of pseudorabies virus latency. Journal
9. Huysman, C. N., Van Leengoed, L. A. M. G., de Jong, of Virological Methods 61, 79–87.
M. C. M. & Van Osta, A. L. M. (1992). Reproductive 26. Denis, M., Soumet, C., Legeay et al. (1997). De-
failure associated with porcine parvovirus in an en- velopment of a semiquantitative PCR assay using in-
zootically infected pig herd. The Veterinary Record ternal standard and colorimetric detection on
131, 503–506. microwell plate for Pseudorabies virus. Molecular and
10. Hallauer, C., Siegl, G. & Kronauer, G. (1971). Parvo- Cellular Probes 11, 439–48.
viruses as contaminants of permanent human cell 27. Legeay, O., Bounaix, S., Denis, et al. (1997). De-
lines. III. Biological properties of the isolated viruses. velopment of a RT-PCR test coupled with a microplate
Archiv für die Gesamte Virusforschung 38, 366–82. colorimetric assay for the detection of a swine Arteri-
11. Croghan, D. L., Matchett, A. & Koski, T. A. (1973). virus in boar semen. Journal of Virological Methods
Isolation of porcine parvovirus from commercial tryp- 68, 65–80.
sin. Applied Microbiology 26, 431–33. 28. Johnson, R. H. (1973). Isolation of swine parvovirus in
12. Eis-Hübinger, A. M., Sasowski, U., Brackmann, H. Queensland. Australian Veterinary Journal 49, 157–59.
H., Kaiser, R., Matz, B. & Schneweis, K. E. (1996). 29. Mengeling, W. L. & Paul, P. S. (1986). Interepizootic
Parvovirus B19 DNA is frequently present in re- survival of porcine parvovirus. Journal of the American
combinant coagulation Factor VIII products. Throm- Veterinary Medical Association 188, 1293–295.
bosis and Haemostasis 76, 1120. 30. Joo, H. S., Donaldson-Wood, C. R. & Johnson, R. H.
13. Prowse, C., Ludlam, C. A. & Yap, P. L. (1997). Human (1976). Rapid diagnostic techniques for detection of
parvovirus B19 and blood products. Vox Sanguinis 72, porcine parvovirus infection in mummified foetuses.
1–10. Australian Veterinary Journal 52, 51–52.
14. Santagostino, E., Mannucci, P. M., Gringeri, A. et al. 31. Persing, D. H. (1993). Target selection and op-
(1997). Transmission of parvovirus B19 by coagulation timization of amplification reactions. In Diagnostic
factor concentrates exposed to 100°C heat after lyo- Molecular Biology, Principles and Applications. (Per-
philisation. Transfusion 37, 517–22. sing, D. H., Smith, T. F., Tenover, F. C. & White, T. J.,
15. Mengeling, W. L., Paul, P. S. & Lager, K. M. (1993). eds) Pp. 88–104. Washington DC: American Society
Virus-induced maternal reproductive failure of swine. for Microbiology.
Journal of the American Veterinary Medical As- 32. Griffais, R., André, P. M. & Thibon, M. (1991). K-tuple
sociation 203, 1268–272. frequency in the human genome and polymerase chain
16. Oraveerakul, K., Choi, C. S. & Molitor, T. W. (1990). reaction. Nucleic Acid Research 19, 3887–891.
Detection of porcine parvovirus using nonradioactive 33. Chenal, V., Souque, P., Markovits, A. & Griffais, R.
nucleic acid hybridization. Journal of Veterinary Diag- (1996). Choosing highly specific primers for the poly-
nostic Investigation 2, 85–91. merase chain reaction using the octamer frequency
17. Molitor, T. W., Oraveerakul, K, Zhang, Q. Q., Choi, disparity method: application to Chlamydia trach-
C. S. & Ludermann, L. R. (1991). Polymerase chain omatis. Gene 176, 97–101.
reaction (PCR) amplification for the detection of por- 34. Mulder, J., McKinney, N., Christopherson, C., Sninsky,
cine parvovirus. Journal of Virological Methods 32, J., Greenfield, L. & Kwok, S. (1994). Rapid and simple
201–11. PCR assay for quantitation of human im-
18. Longo, M. C., Berninger, M. S. & Hartley, J. L. (1990). munodeficiency virus type 1 RNA in plasma: ap-
Use of uracyl DNA glycosylase to control carry-over plication to acute retroviral infection. Journal of
contamination in polymerase chain reactions. Gene Clinical Microbiology 32, 292–300.
93, 125–28. 35. Gerard, B., Peponnet, C., Brunie, et al. (1996). Fluoro-
19. Molitor, T. W., Joo, H. S. & Collett, M. S. (1984). metric detection of HIV-1 genome through use of
Characterization of the genomic and replicative form an internal control, inosine-substituted primers, and
DNA of two virus isolates. Virology 137, 241–54. microtiter plate format. Clinical Chemistry 42, 696–
20. Sambrook, Fritsch & Maniatis (1989). Molecular Clon- 703.
ing: A Laboratory Manual. 2nd edn, Pp. 6·3–6·15. 36. Kox, L. F. F., Noordhoek, G. T., Kunakorn, M., Mulder,
Cold Spring Harbor: Laboratory Press. S., Sterrenburg, M. & Kolk, A. H. J. (1996). Microwell
21. Ranz, A. I., Manclus, J. J., Diaz-Aroca, E. & hybridization assay for detection of PCR products
416 C. Arnauld et al.

from Mycobacterium tuberculosis complex and the viral DNA. Journal of Virological Methods 55, 185–97.
recombinant Mycobacterium smegmatis strain 1008 40. Nelson, S., Matlow, A., McDowell, et al. (1997). De-
used as an internal control. Journal of Clinical tection of Bordetella pertussis in clinical specimens by
Microbiology 34, 2117–120. PCR and a microtiter plate-based DNA hybridization
37. Boutin, P., Arnauld, C., Thiery, R., Costa, J. M., Vidaud, assay. Journal of Clinical Microbiology 35, 117–20.
M. & Jestin, A. (1994). Chemiluminescent detection 41. Mansy, F., Hoyois, B., De Vos, M.-J., Van Elsen, A.,
of amplified pseudorabies virus GP50 DNA with im- Bollen, A. & Godfroid, E. (1996). Colorimetric solid-
mobilized probes on microtiter wells. Acta Veterinaria phase capture hybridization assay for detection of
Hungerica 42, 381–85. amplified Borrelia burgdorferi DNA. BioTechniques
38. Soumet, C., Ermel, G., Boutin, P, Boscher, E. & Colin, 21, 122–25.
P. (1995). Chemiluminescent and colorimetric en- 42. Koskiniemi M., Mannonen, L., Kallio, A. & Vaheri,
zymatic assays for the detection of PCR-amplified A. (1997). Luminometric microplate hybridization for
Salmonella sp. products in microplates. BioTechniques detection of varicella-zoster virus PCR product from
19, 792–96. cerebrospinal fluid. Journal of Virological Methods 63,
39. Zammetteo, N., Moris, P., Alexandre, I., Vaira, D., 71–79.
Piette, J. & Remacle, J. (1995). DNA probe hy- 43. Murphy, F. A. (1996). The public health risk of animal
bridization in microwells using a new bioluminescent organ and tissue transplantation into humans. Science
system for the detection of PCR-amplified HIV-1 pro- 273, 746–47.