1. veterinary
microbiology
Veterinary Microbiology 44 ( 1995) 77-92
Development of a PCR amplification assay as a
screening test using bulk milk samples for
identifying dairy herds infected with bovine viral
diarrhea virus
Gamal S. Radwan a, Kenny V. Brock a**, Joseph S. Hogan b,
K. Larry Smith b
aFood Animal Health Research Program, Ohio Agricultural Research and Development Center,
1680 Madison Ave. 44691 Wooster 44691, OH, USA
’ Department of Dairy Science. Ohio Agricultural Research and Development Center, Wooster. OH, USA
Received 9 December 1994; accepted 30 September 1994
Abstract
The approach of cDNA synthesis followed by polymerase chain reaction (PCR) amplification was
used to develop a rapid screening test for the detection of bovine viral diarrhea virus (BVDV) in bulk
tank milk samples. The initial development of this detection method was done using lactating Holstein
cows; 1 acutely infected with BVDV following experimental inoculation and 2 persistently infected
(PI) with BVDV. Viral RNA was extracted from somatic cells purified from whole milk using a
guanidinium isothiocyanate and phenol/chloroform extraction method. Oligonucleotide primers were
selected from the S’untranslated region (S’UTR) and p80 region of BVDV genome. In the acutely
infected cow, BVDV RNA was identified from days 6 to 10 postinoculation. Viral RNA extracted
from somatic cells of milk from PI cows was detected by PCR using both S’UTR and ~80 primer sets.
The sensitivity of PCR detection was determined by preparing dilutions of whole milk obtained from
the BVDV persistently infected animals with milk from a BVDV-negative cow followed by purifi-
cation of somatic cells and RNA extraction. BVDV was detected in milk serially diluted to 1:640
using PCR amplification. In addition, PCR amplification was 14.6 times more sensitive than virus
isolation in detecting BVDV RNA in purified milk somatic cells. PCR detected BVDV RNA from a
minimum of 580 somatic cells while the detection limit of virus isolation was 8500 cells. The
sensitivity and specificity of BVDV amplification were confirmed by Southern hybridization analysis.
BVDV RNA was detected using PCR in 33 out of 136 bulk milk samples collected from 124 individual
herds using the S’UTR primer set. These results indicate that PCR analysis of bulk tank milk samples
may provide a rapid and sensitive method of screening herds for the presence of BVDV infections.
Keyords: PCR; Herd screening; Bovine viral diarrhea virus; BVDV: Bulk milk: Somatic cells
* Corresponding author. Tel 216-263-3744, Fax 216-263-3677
0378-l 135/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved
.SSL)IO378-I 135(94)00121-9

2. 78 G.S. Radwan et al. / Veterinary Microbiology 44 (1995) 77-92
1. Introduction
Bovine viral diarrhea virus (BVDV) is one of the most important viral pathogens of
cattle, causing considerable economic losses throughout the world (Brownlie, 1985; Duffel1
and Harkness, 1985; Meyling et al., 1990). Three defined disease syndromes caused by
BVDV have been recognized: bovine viral diarrhea (BVD), mucosal disease, and fetal
infection resulting in persistent infections (Brownlie, 1985; Duffel1 and Harkness, 1985;
Duffel1 et al., 1984). Most primary infections are subclinical, but explosive outbreaks of
BVD may occur (Ames, 1986; Harkness, 1987). BVD is characterized by diarrhea, fever,
salivation, leukopenia, and erosion of the oral mucosa (Olafson et al., 1946). BVDV has
been demonstrated to be immunosuppressive (Potgieter et al., 1984) and as a result has the
potential to enhance disease by other pathogens or opportunistic organisms. Transplacental
infection is a common sequel in persistently infected (PI) cattle or after BVDV infection
of susceptible, pregnant heifers and cows (Liess et al., 1984; Orban et al., 1983). Infection
with noncytopathic virus during the first trimester of gestation may result in abortion,
stillbirth, or the birth of PI, immunotolerant calves (Brownlie, 1985). PI animals are the
main source of infectious virus to herdmates as they continually shed large quantities of
virus in body secretions and excretions (Duffel1 and Harkness, 1985; Roeder and Harkness,
1986; Werdin et al., 1989). Such animals are apparently healthy, but some die prematurely,
often after chronic illness, and all are at risk of developing mucosal disease (Brownlie,
1990). Mucosal disease is distinguished from BVD by its more sporadic occurrence, lower
morbidity, longer course, and higher mortality (Perdrizet et al., 1987).
Due to the obvious economic impact of BVDV infections, attempts are made world-wide
to conduct effective control measures; the aims of which are to break the cycle of transmis-
sion by identifying and eliminating PI animals and by preventing transplacental infections
(Duffel1 and Harkness, 1985; Duffel1 et al., 1984; Harkness, 1987; Radostits and Littlejohns,
1988). The effectiveness of such control measures is dependent on maintaining closed herds
and rigorous screening of animals for BVDV infection (Bolin, 1990; Harkness, 1987).
Current methods for the detection of BVDV such as virus isolation and various immu-
noassays either lack optimal sensitivity or rapidity for consistent and large scale testing for
virus detection in animal specimens. Therefore, there is need for a rapid, specific, and
sensitive screening test which can identify herds containing infected animals to allow culling
and maintenance of a BVDV-free status. The purpose of this study was to investigate the
adaptation of polymerase chain reaction (PCR) amplification assay for the detection of
BVDV in bulk milk samples to identify herds infected with BVDV.
2. Materials and methods
2.1. Animals, sampling, and processing of samples
Three adult, lactating Holstein cows were used in the initial study. One cow was acutely
infected with BVDV strain SD- 1 by experimental intranasal inoculation with 2 ml of serum
( 103CCID,,/ml) from a PI animal (Brock, 1991) . The other 2 animals were PI with BVDV

3. G.S. Radwan et al. / Veterinar?, Microbiology 44 (1995) 77-92 79
(PI animals # 1 and #2) and were obtained from 2 private herds. Cows were milked twice
daily using a bucket milker assembly. Milk samples were collected from the individual cow
buckets following milking. Milk was collected from tbe acutely infected cow for 14 days
postinoculation. Milk and whole blood samples were obtained from the 2 PI animals. Milk
somatic cells were purified according to the methods described by Paape et al. ( 1990).
Briefly, the whole milk sample (25 ml) was centrifuged at 1000 g for 15 min at 4°C to
pellet somatic cells and the supematant was removed. The cell pellet was resuspended in
15 ml of PBS and centrifuged at 200 g for I5 min at 4°C and the supematant was removed.
Somatic cell pellets were processed in duplicate. One somatic cell pellet was processed for
RNA extraction and the other pellet was resuspended in 0.5 ml of Dulbecco’s minimum
essential medium (DMEM) and stored at - 70°C until processed for virus isolation.
2.2. Cell culture and quantitation of virus
A 50 ~1 volume of the 0.5 ml of somatic cell suspension in DMEM containing 10% horse
serum was inoculated onto BVDV negative secondary bovine turbinate (BT) cell mono-
layers in 96-well microtiter plates in replicates of 3. The inoculum was removed after 1 h
and replaced with fresh medium. Cell cultures were incubated for 3-4 days at 37°C in 5%
CO,. Following 2 passages, the cell culture supematant was removed and plates were air
dried and cells were fixed with 10% acetone and 0.02% bovine serum albumin (BSA) for
10 min. BVDV antigen was detected using anti-BVDV monoclonal antibody followed by
indirect immunoperoxidase staining as modified from the methods of Afshar et al. ( Afshar
et al., 1989).
2.3. RNA extraction from milk somatic cells
Total RNA extraction was done using the guanidinium isothiocyanate method as previ-
ously described (Brock, 1991). Briefly, somatic cell pellets were mixed with 5 ml of
guanidinium solution (4 M guanidinium isothiocyanate, 25 mM sodium citrate, pH 7.0,
0.5% w/v sarcosyl, and 0.1 M 2-mercaptoethanol). Next, 0.5 ml of 2 M sodium acetate,
pH 4.0; 5 ml of phenol; and 1 ml of chloroform/isoamyl alcohol ( 24: 1) were sequentially
added. The mixture was shaken for 10 seconds, cooled on ice for 15 mitt, and centrifuged
at 10 000 g for 20 min at 4°C. The aqueous phase was removed, precipitated with an equal
volume of isopropanol, and placed at - 20°C overnight. The precipitated RNA was pelleted
by centrifugation at 10 000 g for 30 min at 4°C. The RNA pellet was dried and then
resuspended in 25 ~1 of diethylpyrocarbonate (DEPC) treated water.
2.4. Primer-directed amplification
Primer selection was made using a computer program (Oligo, National Biosciences,
Hamel, CA) designed to construct optimal oligonucleotide primers for use in PCR assays
(Rychlik and Rhoads, 1989). Primers were designed based on the published NADL (Collett
et al., 1988)) Osloss (Renard et al., 1987), and SD-1 (Deng and Brock, 1992) BVDV
nucleotide sequence data. Primers were selected from regions of high conservation of
nucleotide and amino acid sequences within the 5’ untranslated region (S’UTR) and p80

4. 80 G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92
Table 1
Sequences and location of the oligonucleotide primers used for PCR amplification of BVDV
Primer 5’ to 3’ sequence” Product size (bp)
S’UTR I ‘“GGCTAGCCATGCCC’ITAG’17 246
S’UTR 2 ‘45GCCTCTGCAGCACCCTAT3?8
~80 I h3’ZCTGCCAAATGCCTCAACCAAAGCT-4s 1,153
~80 2 7474GGACAACCCGGTCACTTGCTTCAG’4s’
“Numbers in superscript denote nucleotide position along BVDV NADL genome.
region of BVDV genome. The sequence and location of oligonucleotide primers are given
in Table 1. PCR assay was done using reagents supplied in a kit ( Perkin-Elmer Cetus Corp.,
Norwalk, CT), and a programmable DNA thermal cycler (Coy Laboratory Products Inc.,
Ann Arbor, MI). First strand complementary DNA (cDNA) synthesis was done using
reverse transcriptase and random primers. The first strand reaction contained 1 X RT buffer
(5 X RT buffer contains 250 mM Tris-HCl [ pH 831,375 n&l KCl, and 15 rnM MgCl,),
10 mM dithiothreitol, 1 mM of each deoxynucleotide (dATP, dGTP, dTTP, dCTP), 20
units of RNasin, 300 ng of random hexanucleotide, 200 units of moloney murine leukemia
virus reverse transcriptase (MMLV-RT) , and 8.5 ~1 of extracted BVDV RNA in a reaction
volume of 20 ~1. The reaction mixture was incubated at 37°C for 1 h, heat inactivated at
75°C for 10 mins, and 80 ~1 of the following reaction mixture was added: 1 X PCR buffer
( 10 X PCR buffer contains 500 mM KCl, 100 mM Tris-HCl [ pH 8.31)) 1.25 PM of each
upstream and downstream primers, 2.5 units of Ampli Taq polymerase, and 2 mM of MgCl?.
The reaction mixture was initially denatured at 94°C for 4 min followed by 30 cycles of
reaction parameters: template denaturation at 94°C for 1 min primer annealing at 55°C for
1.5 min, and extension at 72°C for 3 min. An additional incubation was done at 72°C for 7
min to complete the extension of open (5’ overhangs) templates.
2.5. Identijkation of the PCR products
Following amplification, 10 ~1 of the PCR product was examined by 1% agarose gel
electrophoresis in Tris acetate EDTA (TAE) buffer using a 1 Kb ladder as molecular weight
standard. The gels were stained by ethidium bromide (0.5 pg/ml) (Sambrook et al., 1989).
The specificity of PCR products was determined by observing the expected size of the
product on the gel and by hybridization with BVDV specific probes.
2.6. Southern transfer
After the PCR products were separated on electrophoresis gels, the amplified DNA bands
were blotted to 1.2 micron nylon membranes by an alkaline vacuum transfer method using
a VacuGene XL vacuum blotting system (Pharmacia LKB Biotechnology, Piscataway, NJ)
according to the manufactures’ instructions. The nylon membrane filters were air dried and
baked at 80°C for 1 h.

5. G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92 81
2.7. Preparation of probes and hybridization
Two plasmids, pBV-18 and pBV4-p80 derived from BVDV-NADL (kindly provided by
Dr. Marc Collett, Seattle, WA) were used to generate PCR-derived probes. The plasmids
contained inserts encompassing nucleotides 24 to 1308 ( pBV- 18) and nucleotides 5644 to
7949 (pBV-~80). These cDNA inserts were used as templates in PCR amplification using
the S’UTR and ~80 primer sets as described above. Amplified DNA products were concen-
trated and purified using Centricon- 100 microconcentrators ( Amicon, Beverly, MA). The
purified PCR products were radiolabelled with [ a3’P] dCTP to a specific activity of approx-
imately 10’ cpm/ pg by nick translation ( Rigby et al., 1977 ) . Labelled DNA was separated
from unincorporated nucleotides by centrifugation through Sephadex G-50 according to
methods previously described (Maniatis et al., 1982). Prior to hybridization, probes were
denatured by heating at 100°C for 5 min followed by rapid cooling on ice. Blots were
prehybridized at 42°C for 4-6 h in hybridization buffer (50% formamide, 6 X SSC [ 1 X SSC
is 0.15 M NaCl. 0.015 M trisodium citrate, pH 7.01, 0.5% SDS, 5 XDenhardt’s solution
[ 1 X Denhardt’s solution is 0.02% ficoll, 0.02% polyvinylpyrorollidone and 0.02% BSA] ,
and 100 pg/ml of sheared salmon sperm DNA) with gentle agitation. Denatured probes
were added to fresh hybridization buffer and hybridization was done at 42°C for 16-24 h.
Hybridized blots were washed 4 times in 2 X SSC and 0.1% SDS for 5 min at room
temperature and twice in 0.1 X SSC and 0.1% SDS for 15 min at 50°C. After washing, blots
were air dried at room temperature, and then exposed to radiographic film with an intensi-
fying screen at - 70°C for 24 h.
2.8. Sensitivity of PCR detection
Two methods were used to determine the sensitivity of PCR amplification. Both methods
were performed twice from individual milk samples taken 1 week apart. First, whole milk
obtained from PI animal #2 was serially diluted 2-fold ( 1:20 through 1:1280) with whole
milk from a BVDV-negative cow in 25 ml volumes. Somatic cells purified from milk
dilutions were processed for RNA extraction, PCR amplification, and Southern hybridiza-
tion analysis as previously described. Second, the sensitivity of PCR amplification for
detection of viral RNA extracted from somatic cells obtained from a PI animal was compared
with that of virus isolation. One hundred ~1 of a somatic cell suspension purified from milk
obtained from a PI animal #2 (adjusted to 1.7 X 10” cells/ 100 ~1) was serially diluted lo-
fold in duplicate with DMEM. One series was processed for virus isolation as previously
described using an inoculation volume of 50 ~1. RNA was extracted from the other series
and the pellet resuspended in 25 ~1 using 8.5 ~1 for first strand synthesis and PCR ampli-
fication as previously described. Somatic cells purified from milk obtained from the BVDV-
negative cow were included as a negative control.
2.9. Screening of bulk milk samples for BVDV
A total of 136 bulk milk samples were collected from 124 individual dairy herds. Dupli-
cate samples were collected from 12 herds at different times. Of these samples, 95 were
collected from 83 herds suspected of being BVDV-infected based on herd history and

6. 82 G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92
clinical manifestation. The remaining 41 samples were obtained from randomly selected
dairy herds. Samples were collected in sterile or clean containers and held on ice for transport
to the laboratory. Upon arrival, somatic cells were purified from 175 ml of the bulk milk
sample and total RNA was extracted as previously described. cDNA was synthesized from
RNA extracted from somatic cells and PCR amplification was done using the S’UTR primer
set. A negative control sample containing all the PCR reagents with no template added was
included in each amplification round. Size specific amplification products were identified
by agarose gel electrophoresis and analyzed by Southern blot hybridization of the PCR
products using the pBV- 18 PCR-derived DNA probe. Virus isolation was done for 2
passages with subsequent detection of BVDV antigen using the microtiter plate immuno-
peroxidase test as previously described.
3. Results
3.1. Quantitation of BVDV in milk and serum
The 50% endpoint of cell culture infective doses/ml (CCIDJml) was calculated using
the results of virus isolation from ten-fold serum and milk dilutions. BVDV titers in milk
and serum from experimentally (acutely infected) and PI animals are shown in Table 2.
Milk from the experimentally, acutely infected animal contained 102.5 CCIDso of BVDV/
ml while PI animals #l and #2 contained 106.5 and 105.5 CCID5,/ml, respectively. Levels
of BVDV in serum from PI animals #l and #2 were 104.5 and 104.’ CCIDS,/ml, respec-
tively. The average somatic cell counts for PI animal # 1 was approximately 225 000 cells/
ml ( 12 samples) and 125 000 cells/ml for PI animal #2 (4 samples) during the sampling
period.
3.2. PCR ampli$cation of BVDV cDNA from milk somatic cells
The product length of primer-directed amplification using the S’UTR and p80 primer
sets, based on the published sequence data of BVDV strains NADL, Osloss, and SD- 1, was
calculated to be 246 and 1153 base pairs (bp), respectively (Table 1 and Fig. 1). In the
experimentally infected animal, PCR amplification using the S’UTR primer set detected
BVDV RNA extracted from milk somatic cells collected on postinoculation days 6,7, 8,9,
Table 2
BVDV titers (CCID50/ml)” in milk and serum from experimentally and persistently infected (PI) lactating
animals
Animal CCID,,/ml milk CCID,,/ml serum
Experimentally infected lo*5 NDh
PI animal # 1 106.5 IO45
PI animal #2 1os5 10J’
“Values represent the mean titers of 3 replicates.
bND = not determined.

7. G.S. Radwan et al. / Veterinary Microbioiogy 44 (1995) 77-92 83
5’ untranslated region 3’ untranslated region
I structural
proteins nonstructural proteins
u246 bp
5'UTRI 5'UTR2
Fig. 1. Schematic representation of BVDV genome indicating the position of the primer pairs and the expected
lengths of the PCR products.
and 10 (Fig. 2). Table 3 compares the results of BVDV detection in milk somatic cells
purified from whole milk of the experimentally infected animal using PCR amplification
and virus isolation assay. BVDV was isolated from somatic cells purified from milk collected
on postinoculation days 7,8, and 9 with peak titers occuring on days 7 and 8 postinoculation
at 1O2.5CCID,,/ml.
PCR amplification identified BVDV RNA extracted from somatic cells purified from
whole milk from PI animals #l and #2 using the S’UTR and ~80 primer sets (Fig. 3 A
and B, respectively). Southern blot analysis of the agarose gel (Fig. 3 C and D) demon-
strated that the amplified products following amplification with the S’UTR and p80 primer
sets were BVDV specific by hybridization with probes prepared from amplified cDNA from
BVDV NADL clones pBV- 18 and pBV4-~80.
Ml 2345 67
Fig. 2. Agarose gel electrophoresis of PCR products following amplification of cDNA of BVDV, RNA extracted
from milk somatic cells of experimentally infected animal using the S’UTR primer set. Lane M contains the
molecular weight size marker. Lane l-7 represent PCR products from somatic cells obtained on postinoculation
days 5,6,7,8,9, 10, and 12, respectively. The amplified product of 246 bp is indicated by an arrow on the right.

8. 84 G.S. Radwan et al. / Veterinary Microbiology 44 (1995) 77-92
Table 3
Detection of BVDV in milk somatic cells from experimentally infected animal by PCR amplification and virus
isolation
Assay used Days post inoculation
5 6 7 8 9 10 12
PCR amplification” _ + + + + + -
Virus isolation _ _ + + + - _
“Using S’UTR primer set.
A Ml 2
B
Ml 2
1,=
1,018. .- 1,153 bp
C D
,in:
Fig. 3. Agarose gel electrophoresis of PCR products following amplification of cDNA of BVDV RNA extracted
from milk somatic cells of 2 persistently infected (PI) animals using the 5’UTR (Fig. 3A) and ~80 (Fig. 38)
primer sets. Lane M is the standard size marker. Lane 1, PI animal # 1; lane 2, PI animal #2. The amplified product
is indicated by an arrow on the right of each panel. Fig. 3C and 3D represent Southern blot hybridization of
agarose gels described in Fig. 3A and 3B, respectively. Hybridization was done with pBV-18 PCR-derived probe
(Fig. 3C) and with pBV4-~80 PCR-derived probe (Fig. 3D). Autoradiography was done for 24 hours at - 70°C.

9. G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92 85
3.3. Sensitivity of PCR detection
PCR amplification detected BVDV RNA in somatic cells purified from milk obtained
from PI animal #2 serially diluted to 1:640 with negative whole milk using the S’UTR
primer set (Fig. 4). No amplification was observed with RNA extracted from somatic cells
purified from BVDV-negative whole milk (data not shown). The sensitivity of PCR ampli-
fication and virus isolation assay for BVDV detection in milk somatic cells is shown in Fig.
5 and Table 4. Both PCR amplification and virus isolation detected BVDV in somatic cells
diluted to lo-* (corresponds to 8500 cells for virus isolation and 5800 cells for PCR).
However, PCR amplification using the 5’UTR primer set detected BVDV RNA from milk
somatic cells diluted to 10e3 (corresponds to 580 cells). In this case, the PCR amplification
assay was 14.6 times more sensitive than virus isolation in detecting BVDV in milk somatic
cells. No specific amplification product was detected with cDNA from somatic cells purified
from BVDV-negative milk used as a negative control (Fig. 5, lane 6). The sensitivity and
specificity of PCR amplification for BVDV RNA detection in the 2 dilution experiments
A
Fig. 4. A, Agarose gel’electrophoresis of PCR amplification products of BVDV cDNA of RNA purified from
whole milk dilutions with the S’UTR primer set. Lane M is the molecular weight size marker. Lanes l-7 represent
PCR products from somatic cells purified from two-fold dilutions of whole milk from I:20 to 1: 1280, respectively.
Fig. 4B; Southern blot hybridization of the agarose gel as described (Fig. 4A) with the pBV-18 FCR-derived
probe. Autoradiography was done for 24 hours at - 70°C.

10. 86 G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92
Fig. 5. A. Agarose gel electrophoresis of PCR amplification products of BVDV cDNA from RNA purified from
milk somatic cells using the S’UTR primer set. Lane M is the molecular weight size marker. Lanes l-5 represent
ten-fold somatic cell dilutions from 10’ to lo-“, respectively. Lane 6 represents somatic cells purified from milk
of a BVDV-negative cow. Amplified product is indicated by an arrow on the right. Fig. 5B, Southern blot
hybridization of the agarose gel described in Fig. 5A with pBV-I8 PCR-derived probe. Autoradiography was
done for 24 hours at - 70°C.
Table 4
Comparison of the sensitivity of virus isolation and PCR amplification in the detection of BVDV in milk somatic
cells
Ten fold dilution Virus isolation PCR”
100 + +
10’ + +
102 +b +
103 - fL
IO4 - _
lo5 - _
*Usinn the S’UTR mimer set.
bCorr&onds to 8,500 somatic cells
‘Corresponds to 580 somatic cells

11. G.S. Radwan et al. / Veterinary Microbiology 44 (1995) 77-92 87
Table 5
PCR screening of bulk milk samples for BVDV
Origin of bulk milk sample Total no. tested” No. positive by PCRb and Southern blot
hybridization”
Herds suspected of being BVDV infected 95 (83) 31
Randomly selected herds 41 (41) 2
Total 136 (124) 33
“Numbers in parentheses denotes total number of individual herds tested.
%sing the S’UTR primer set.
‘Hybridization of the PCR products was done using pBV-18 PCR-derived probe.
were confirmed by Southern blot hybridization. PCR products blotted following amplifi-
cation using the S’UTR primer set hybridized specifically with the corresponding pBV- 18
PCR-derived probe (Fig. 4 and 5, B) . The results for the sensitivity of detection were
identical for both serially collected samples from PI animal #2.
3.4. Screening of bulk milk samples for BVDV
Following PCR amplification of RNA extracted from somatic cells purified from the bulk
milk samples, 33 out of 136 bulk milk samples were identified as positive for BVDV (Table
A B
29%
220-
Fig. 6. A, Agarose gel electrophoresis of a PCR amplification product of BVDV cDNA from RNA in bulk milk
samples using the S’UTR primer set. Lane 1 is the molecular weight size marker as indicated. The amplified
product (lane 2) is indicated by an arrow on the right. Fig. 6B, Southern blot hybridization of the agarose gel
described in Fig. 6A with the pBV-18 PCR-derived probe. Autoradiography was done for 24 hours at - 70°C.

12. 88 G.S. Radwan et al. / Veterinav Microbiology 44 (1995) 77-92
5). Thirty one of the 33 positive samples were collected from herds suspected to have
BVDV infections. On the other hand, the remaining 2 bulk milk samples found positive for
BVDV by PCR were collected from dairy herds randomly selected for PCR screening of
bulk milk for BVDV. PCR amplification was carried out using the S’UTR primer set (Fig.
6 A). Hybridization of southern blots of the electrophoresed PCR products (Fig. 6 B)
confirmed that the amplified fragment was complementary with BVDV cDNA sequences.
On the other hand, BVDV was not isolated from any of the 136 bulk milk samples tested.
4. Discussion
In this study, a screening assay using PCR amplification for the detection of BVDV RNA
in bulk milk samples was developed. In the initial phase of this study it was necessary to
determine the level of BVDV shedding in milk from experimental acutely and PI animals.
The aim was to verify if BVDV was present in milk in concentrations which would allow
detection of viral RNA by PCR amplification. It has been reported that in acute infections,
virus appears to be shed in low concentrations by infected cattle (Duffel1 and Harkness,
1985). In this study, the level of BVDV present in milk from the experimental acutely
infected animal was lower than in milk from the 2 PI animals. Somatic cells in milk are a
mixture of secretory epithelial cells and leukocytes (Heald, 1985) and epithelial cells and
cells belonging to the mononuclear system support the replication of BVDV (Bielefeldt
Ohmann, 1983). In addition, preliminary studies indicated that RNA extracted from milk
somatic cells provided better amplification than RNA extracted from whole milk. Therefore,
it was assumed that milk somatic cells would be a good source for BVDV RNA detection
using PCR amplification. It is also interesting to note that in the PI animals, levels of BVDV
in milk were higher than those in serum. This may be explained by the presence of higher
levels of cellular elements in milk as compared with serum.
Similar amplification products were consistently obtained from somatic cells from the
experimentally and PI animals. The expected size for the amplified products was examined
using electrophoresis, and the final identification of the amplified products was confirmed
as BVDV-specific by the hybridization assay using BVDV-specific PCR-derived probes.
The specificity of the PCR amplification to detect BVDV in somatic cells purified from
milk obtained from the PI animals was confirmed by successful amplification of 2 different
regions of the viral genome. The most highly conserved regions of the genome are within
the S’UTR and the p80 region (Collett et al., 1989; Deng and Brock, 1992) from which
both sets of primers used in this study were selected. The S’UTR primer set provided more
consistent and discrete amplification products as compared with the p80 primer set. The
results of PCR amplification of BVDV RNA from milk and somatic cell dilutions confirmed
the sensitivity of this detection method. It was determined that PCR amplification could
detect BVDV RNA extracted from somatic cells purified from 25 ml of whole milk sample
diluted to 1640. In the routine bulk milk testing, somatic cells were purified from 175 ml
samples. Therefore, this would extrapolate to detecting 1 PI animal in a herd of 5000 if
levels of virus shed are consistent with that of PI animal #2 and production levels of
persistent and noninfected animals were similar. Furthermore, PCR amplification using the

13. G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92 89
S’UTR primer set was 14.6 times more sensitive than virus isolation in detecting BVDV in
milk somatic cells.
Following determination of the sensitivity of PCR amplification, the assay was used to
detect BVDV RNA in somatic cells purified from bulk tank milk samples. The use of PCR
amplification allowed the detection of BVDV in bulk milk samples from 3 1 BVDV-suspect
herds and 2 herds randomly selected for PCR screening for BVDV. Bulk milk samples that
are confirmed positive by southern blot hybridization of the PCR product can be considered
positive for BVDV indicating active acute or persistent infection in lactating animals. In
herds identified as positive by preliminary PCR screening of a bulk milk sample, further
complete individual animal testing must be done to identify the individual source of virus.
Although the S’UTR primer set is from the most highly conserved regions of the genome,
false negative results may be obtained due to the potential genetic diversity among different
BVDV field isolates. Previous studies have demonstrated that genomic variability among
BVDVs affects the performance of PCR amplification (Boye et al., 1991; Hertig et al..
199 1; Ward and Misra, 199 1) . The influence of genetic variation must be considered when
interpreting results of any PCR screening assay.
Attempts to isolate BVDV from bulk milk samples tested were unsuccessful. The negative
virus isolation results may be due to the presence of anti-BVDV antibodies in the bulk milk
samples. The presence of these antibodies does not affect PCR detection of viral RNA, as
it does virus isolation. It is likely that the majority of bulk tank milk samples contained
moderate to low levels of antibodies due to widespread use of BVDV vaccines. Recent
studies using indirect ELISA indicated good correlation between the level of antibodies to
BVDV in bulk tank milk and the prevalence of BVDV antibody positive cows (Niskanen
et al., 1991; Carlsson et al., 1993).
In conclusion, PCR analysis of bulk tank milk samples may provide a rapid and sensitive
method to screen herds for the presence of BVDV infections. By this detection assay results
are obtained within 2 to 3 days. The speed and sensitivity of the PCR amplification makes
it a useful method for testing large numbers of bulk milk samples in a relatively short time.
Screening of bulk milk samples by this assay may also be used to easily monitor the BVDV
infection status of dairy herds. The practical application of this screening method must be
determined by correlation with individual animal testing. The use of this screening assay
may be most beneficial as a method of focusing on or identifying BVDV positive herds for
further development of control strategies and not a definitive test to insure that a herd is
negative for BVDV. Positive PCR results would indicate infection, however negative PCR
results could not be interpreted as indicating the absence of infection.
Acknowledgements
We would like to thank Dr. MS. Collett for kindly providing the pBV-18 and pBV4-p80
BVDV clones. We appreciate the technical assistance of Sylva Riblet, Jerry Meitzler, and
Sue Romig. This study was supported in part by special research grant #2002007 1, Amer-
ican Veterinary Medical Association Foundation. Salaries and research support were pro-
vided by State and Federal Funds appropriated to the Ohio Agricultural Research and
Development Center, The Ohio State University. Journal article no. 219-93.

14. 90 G.S. Radwan et al. /Veterinary Microbiology 44 (1995) 77-92
References
Afshar, A., Dulac, G.C., and Bouffard, A., 1989. Application of peroxidase labelled antibody assay for detection
of porcine IgG antibodies to hog cholera and bovine viral diarrhea virus. J. Viral. Methods, 23: 253-262.
Ames, T.. 1986. The causative agent of BVD: its epidemiology and pathogenesis. Vet. Med., 81: 848-869.
Bielefeldt Ohmann, H., 1983. Pathogenesis of bovine viral diarrhea-mucosal disease: distribution and significance
of BVDV antigen. Res. Vet. Sci., 14: 5-10.
Bolin, S.R., 1990. Control of bovine virus diarrhea virus. Rev. Sci. Tech. Off. Int. Epizoot., 9: 163-171.
Boye, M., Kamstrup. S. and Dalsgaard, K., 1991. Specific sequence amplification of bovine virus diarrhea virus
(BVDV) and hog cholera virus and sequencing of BVDV nucleic acid. Vet. Microbial., 29: I-13.
Brock, K.V., 1991. Detection of persistent bovine viral diarrhea infections by DNA hybridization and polymerase
chain reaction assay. Arch. Virol. [Suppl. 31: 199-208.
Brownlie, J., 1985. Clinical aspects of the bovine virus diarrhea/mucosal disease in cattle. In Pratt., 7: 195-202.
Brownlie, J., 1990. The pathogenesis of bovine viral diarrhoea virus infections. Rev. Sci. Tech. Off. Int. Epizoot..
9: 43-54.
Carlsson, U., Niskanen, R., Alenius, S. and Larson, B., 1993. A strategy for elimination of ongoing infection with
BVDV in Dairy herds. Proc. European Society Vet. Viral. second symposium on pestivimses, pp. 243-245.
Collett, MS., Larson, R., Gold, C., Strick, D., Anderson, D.K. and Purchio. A.F., 1988. Molecular cloning and
nucleotide sequence of the pestivirus bovine viral diarrhea virus. Virology, 165: 191-199.
Collett, M.S., Moennig, V. and Horzinek, MC., 1989. Recent advances in pestivirus research. J. Gen. Virol., 70:
253-266.
Deng, R. and Brock, K.V., 1992. Molecular cloning and nucleotide sequence of a pestivirus genome, noncytopathic
bovine viral diarrhea strain SD-l. Virology, 191: 867-879.
Duffell, S.J. and Harkness, J.W., 1985. Bovine virus diarrhoea-mucosal disease infection in cattle. Vet. Rec., 117:
240-245.
Duffell, S.J., Sharp, M.W., Winkler, C.E., Terlecki, S., Richardson, C., Done, J.T., Roeder, P.L. and Hebert,C.N..
1984. Bovine virus diarrhoea-mucosal disease virus-induced fetopathy in cattle: efficacy of prophylactic
maternal pre-exposure. Vet. Rec., 114: 558-561.
Harkness, J.W., 1987. The control of bovine viral diarrhea virus infection. Ann. Rech. Vet., 18: 167-174.
Heald, C.W., 1985. Milk collection. In: B.L. Larson, (Editor), Lactation. Iowa State University Press, Ames, IA.
pp. 198-228.
Hertig, C., Pauli, V., Zanoni. R. and Peterhans, E., 1991. Detection of bovine viral diarrhea (BVD) virus using
the polymerase chain reaction. Vet. Micmbiol., 26: 65-76.
Liess, B.S., Orban, S., Frey, H.-R., Trautwein, G., Weifel, W. and Blindow, H., 1984. Studies on transplacental
transmissibility of a bovine viral diarrhoea virus (BVD) vaccine virus in cattle: II. Inoculation of pregnant
cows without detectable neutralizing antibodies to BVD virus 90 to 229 days before parturition (5 1st to 190th
day of gestation. Zentralbl. Vet. Med. B, 30: 669-68 1.
Maniatis. T., Fritsch, E.F. and Sambrook, J., 1982. Molecular cloning: a laboratory manual. Cold Spring Labo-
ratory, Cold Springs Harbor, NY, p. 446.
Meyling, A., Houe, H. and Jensen, A.M., 1990. Epidemiology of bovine virus diarrhoea virus. Rev. Sci. Tech.
Off. Int. Epizoot., 9: 75-93.
Niskanen, R., Aienius, S., Larson, 9. and Jacobsson, S-O., 1991. Determination of level of antibodies to bovine
virus diarrhoea virus (BVDV) in bulk tank milk as a tool in the diagnosis and prophylaxis of BVDV infections
in dairy herds. Arch. Virol., [ Suppl. 31: 245-25 1.
Olafson, P., MacCullum, A.D. and Fox, F.H., 1946. Apparently new transmissible disease of cattle. Cornell Vet.,
36: 205-213.
Orban, S., Hafez, S.M., Frey, H.-F., Blindow, H. and Sasse-Pastzer, B., 1983. Studies on transplacental transmis-
sibility of a bovine virus diarrhoea (BVD) vaccine virus: I. Introduction of pregnant cows 15 to 90 days before
parturition ( 190th to 265th day of gestation. Zentralhl. Vet. Med. B, 30: 619-634.
Paape, M.J.. Miller, R.H. and Ziv. G., 1990. Effects of flortenicol, chloramphenicol. and thiamphenicol on
phagocytosis, chemiluminescence, and morphology of bovine polynuclear neutrophil leukocytes. J. Dairy Sci.,
73: 1334-134-l.
Perdrizet, J.A., Rebhun, W.C., Dubovi, E.J., and Donis. R.O., 1987. Bovine virus diarrhea-clinical syndromes in
dairy cattle. Cornell Vet., 77: 46-74.

15. G.S. Radwan et al. / Veterinary Microbiology 44 (1995) 77-92 91
Potgieter. L.N.D., McCracken, M.D., Hopkins, F.M., Walker, R.D. and Guy, J.S., 1984. Experimental production
of bovine respiratory tract disease with bovine viral diarrhea virus. Am. J. Vet. Res., 45: 1582-1585.
Radostits, O.M. and Littlejohns, I.R., 1988. New concepts in the pathogenesis, diagnosis and control of diseases
caused by the bovine viral diarrhea virus. Can. Vet. J., 29: 513-528.
Renard, A.. Dina, D. and Martial, J.A., 1987. Complete nucleotide sequence of bovine viral diarrhea genome and
its fragment, useful for making antigenic proteins useful for therapy and diagnosis. European Patent Appli-
cation, No. 0208672.
Rigby, P.W.J.. Dieckmann. M., Rhodes, C., and Berg, P., 1977. Labelling deoxyribonucleic acid to high specific
activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol., 113: 237-251.
Roeder, P.L. and Harkness, J.W., 1986. BVD virus infection: prospects for control. Vet. Rec., 118: 143-147.
Rychlik, W. and Rhoads, R.E., 1989. A computer program for choosing optimal oligonucleotides for filter
hybridization, sequencing and in vitro amplification of DNA. Nucleic Acid Res., 17: 8543-8551.
Sambrook, J., Fritscb, E.F. and Maniatis, T., 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor
Press, Cold Spring Harbor, NY, p. 6.15.
Ward, P. and Misra, V., 1991. Detection of bovine viral diarrhea virus, using degenerate oligonucleotide primers
and the polymerase chain reaction. Am. J. Vet. Res., 52: 1231-1236.
Werdin, R.E., Ames, T.A., Goyal, S.M. and DeVries, G.P., 1989. Diagnostic investigation of bovine viral diarrhea
infection in a Minnesota dairy herd. J. Vet. Diagn. Invest.. 1: 5761.