Journal
of General Virology (2001), 82, 2615–2620. Printed in Great Britain
..........................................................................................................................................................................................................
SHORT COMMUNICATION
High frequency RNA recombination in porcine reproductive
and respiratory syndrome virus occurs preferentially between
parental sequences with high similarity
Joke J. F. A. van Vugt,† Torben Storgaard,‡ Martin B. Oleksiewicz and Anette Bøtner
Danish Veterinary Institute for Virus Research, Lindholm, DK-4771 Kalvehave, Denmark
Two types of porcine reproductive and respiratory
syndrome virus (PRRSV) exist, a North American
type and a European type. The co-existence of both
types in some countries, such as Denmark, Slovakia
and Canada, creates a risk of inter-type recombination. To evaluate this risk, cell cultures were
co-infected with either a North American and a
European type of PRRSV or two diverse types of
European isolate. Subsequently, an approximately
600 bp region of the PRRSV genome was tested for
recombination by quantitative real-time RT–PCR.
Between 0n1 and 2n5 % RNA recombination was
found between the European isolates, but no
recombination was detected between the European
and North American types. Calculation of the maximum theoretical risk of European–American recombination, based on the sensitivity of the RT–PCR
system, revealed that RNA recombination between
the European and North American types of PRRSV is
at least 10 000 times less likely to occur than RNA
recombination between diverse European isolates.
After the initial peak of porcine reproductive and respiratory syndrome (PRRS) outbreaks in 1987–1991 (Keffaber,
1989 ; Wensvoort et al., 1991), the disease is now endemic in
North America, Europe and Asia. Until recently, the European
type of PRRS virus (PRRSV) was restricted to Europe, while
the American type of PRRSV was restricted to North America
and Asia. Now, however, European wild-type PRRSV has been
Author for correspondence : Torben Storgaard.
Fax j45 44 43 45 58. e-mail TStS!novonordisk.com
† Present address : Laboratory of Entomology, Bode 49, Postbus
9101, 6700 HB Wageningen, The Netherlands.
‡ Present address : Novo Nordisk A/S, Novo Nordisk Park, DK-2760,
Ma/ løv, Denmark.
The GenBank accession nos of the sequences reported in this paper
are AF315699–AF315711 and AY034879.
0001-7901 # 2001 SGM
isolated from pigs in North America (Dewey et al., 2000) and
American wild-type PRRSV has been isolated in Slovakia
(Psikal et al., 1999). Furthermore, in several European countries,
a modified, live vaccine based on the American type of PRRSV
has been used and, at least in Denmark, this live vaccine has
reverted to wild-type virus (Bøtner et al., 1997 ; Nielsen et al.,
2001 ; Storgaard et al., 1999). Recombination between the
American isolates of PRRSV (intra-type recombination) occurs
(Yuan et al., 1999), so the co-existence of the European and the
American type of PRRSV within herds and even within single
animals, as seen in some countries such as Denmark (Anette
Bøtner, personal communication), creates a potential risk for
inter-type recombination. This could result in serious problems
with virus typing and vaccination and might potentially even
generate a virus with new biological properties.
To examine inter- and intra-type PRRSV RNA recombination, MARC-145 cells were co-infected (m.o.i. of 0n01
TCID per cell) with the Danish 111\92 isolate (Madsen et al.,
&!
1998) and the American PRRS live vaccine virus (Ingelvac
PRRS Vet) (Boehringer Ingelheim) or co-infected with the
Danish and Dutch Lelystad isolates of PRRSV (Meulenberg et
al., 1993). The Lelystad and Danish isolates were adapted to
MARC-145 cells by serial passage. For control infections,
MARC-145 cells were mock-infected with medium or infected
solely with the Danish, Lelystad or American isolate. After 3
days of incubation at 37 mC, cells were lysed and total RNA
was isolated. cDNA was synthesized using PRRSV-specific
primers, essentially as described previously (Oleksiewicz et al.,
1998). Detection of viral RNA recombination by PCR requires
primers that are absolutely specific for each of the three virus
isolates. Danish-specific forward (5h GCGTCACTTTCAACAAGCCATCTC) and reverse (5h TTTGATGGTAACAAGGTCGCTGC) primers that amplified an expected fragment of
894 bp only when used on cDNA from Danish PRRSVinfected cells were designed (Fig. 1, lanes 2–5). Lelystadspecific forward (5h GGTCTCAGCAGCGCAAGAGAA) and
reverse (5h CAAATCCTGCAGTGGATACAGCG) primers
that amplified an expected fragment of 687 bp only when used
on cDNA from Lelystad PRRSV-infected cells were also
designed (Fig. 1, lanes 6–8). Finally, American-specific forward
(5h GCGATAGGGACACCTGTGTATGTT) and reverse (5h
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19
CGBF
J. J. F. A. van Vugt and others
Fig. 1. PCR using homologous or heterologous primer pairs. Lanes 1 and 18 are marker lanes containing ΦX174 HaeIIIdigested DNA. The size (bp) of the marker bands is indicated on the left. The primers used for PCR were the Danish (lanes
2–5), Lelystad (Dutch) (lanes 6–8) or North American (lanes 9–11) forward and reverse primer pairs. The Danish forward and
Lelystad reverse primers (lanes 12–14) and the Danish forward and North American reverse primers (lanes 15–17) were also
used. The results of PCR on cDNA from cells infected with the Danish (lanes 2, 7 and 10), Lelystad (lanes 3 and 6) and
American (lanes 4 and 9) isolates are shown. The results of PCR on cDNA from cells co-infected with the Danish and Lelystad
isolates (lane 12), a lysate mixture from cells singly infected with the Danish and Lelystad isolates (lane 13), cells co-infected
with the Danish and American isolates (lane 15) or a lysate mixture from cells singly infected with the Danish and American
isolates (lane 16) are shown. Lanes 5, 8, 11, 14 and 17 show the result of PCR on cDNA from mock-infected cells. The PCR
products are 894 bp (lane 2), 687 bp (lane 6), 626 bp (lane 9) and 667 bp (lane 12) in size. DK, Danish ; NL, Lelystad ; US,
North American ; Neg., negative control ; Mix cont., mixture control.
GGTAGACACAGTGACTAAAGCGACT) primers that
amplified an expected fragment of 626 bp only when used on
cDNA from American PRRSV-infected cells were designed
(Fig. 1, lanes 9–11). All three forward primers were located
within the 3h end of open reading frame (ORF) 3 and all three
reverse primers were located within the 3h end of ORF 5 (Fig.
2). PCR was carried out using AmpliTaq Gold polymerase
according to the manufacturer’s instructions (Applied Biosystems), starting with a hot-start activation step of 5 min at
95 mC, followed by 45 cycles of 95 mC for 15 s, 60 mC for 15 s
and 72 mC for 15 s. To exclude the possibility that artificial
recombination might have occurred during RNA isolation,
cDNA synthesis or PCR amplification, Danish–American and
Danish–Lelystad mixture controls were produced by combining cell lysates from the single PRRSV control infections in
a 1 : 1 ratio before RNA isolation. To detect European intratype recombination, PCR was carried out using the Danishspecific forward primer and the Lelystad-specific reverse primer
on cDNA from Danish–Lelystad co-infected cells and from the
Danish–Lelystad mixture control. A band with the expected
size of 667 bp was seen only on cDNA from co-infected cells
and not from the corresponding mixture control (Fig. 1, lanes
12–14). When the equivalent experiment was performed to
detect European–American inter-type recombination, no specific band was detected in any of the reactions (Fig. 1, lanes
15–17). It was concluded, therefore, that intra- but not intertype recombination occurred under the in vitro cell culture
conditions used.
To determine sites of recombination, the Danish–Lelystad
recombinant PCR product was gel-purified and cloned into the
CGBG
pCRII vector (Invitrogen), according to the manufacturer’s
instructions. Plasmid DNA purified from 13 individual clones
was sequenced using BigDye chain terminators, as recommended by the manufacturer (Applied Biosystems), and the
result was compared to both the sequence of the Danish
111\92 isolate (GenBank accession no. AY034879) and the
sequence of the Lelystad isolate (GenBank accession no.
M96262). All 13 sequences obtained have been submitted to
GenBank (accession nos AF315699–AF315711). Seven of the
clones had recombined in a 104 bp region in ORF 5, with
complete identity between the parental strains (Fig. 2). Two
clones had recombined in a region of 52 bp in ORF 4, with
complete identity between the parental strains. Three clones
had recombined in 5 (ORF 4), 8 (ORF 4) and 26 (ORF 5) bp
regions, with complete sequence identity between the parental
strains. Finally, one clone differed from the rest in that two
deletions had been introduced at the site of recombination (Fig.
2), suggesting that the mechanism of recombination is not
always precise, as described previously for other RNA viruses
(Nagy & Bujarski, 1995). The Danish and Lelystad PRRSV
isolates have 94 % overall identity in the region examined for
recombination (621 bp), while the Danish and American
PRRSV isolates only have 60 % overall identity in the region
examined for recombination (579 bp). Despite the low level of
overall identity between the Danish and American isolates in
the 579 bp area, small regions (up to 11 bp) of complete
identity do exist. The presence of recombination between the
two European isolates in short stretches of identity (less than
11 bp) and the absence of recombination between the Danish
and American isolates suggest that short stretches of perfect
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19
PRRSV recombination
Fig. 2. Identification of recombination hot spots between Danish and Lelystad PRRSV. The positions of the Danish-specific
forward primer (M) and the Lelystad-specific reverse primer (J) are shown in relation to PRRSV ORFs 3–5. The identity of the
two parental sequences (Danish and Lelystad) is shown using a sliding window of 20 bp and a step size of 10 bp. Each of the
13 sequenced clones (A–M corresponding to AF315699–AF3115711) is drawn. Solid lines indicate sequences originating
from the Danish parental strain, while dashed lines indicate sequences originating from the Lelystad parental strain. Boxes
connecting the solid and the dashed lines indicate regions of complete identity between the two parental strains. The actual site
of RNA recombination is within these boxes.
identity should be flanked by longer regions with a high
percentage of identity in order to recombine (Fig. 2). The total
absence of RNA recombination between the Danish and
American PRRSV isolates could, however, be caused by
insufficient sensitivity of our RT–PCR assay or by non-optimal
experimental conditions for virus recombination. To evaluate
these two possibilities, a quantitative real-time RT–PCR was
established for the different viral RNA targets. This was carried
out by adding SYBR Green I (375 000i concentration, diluted
in the final PCR) (Molecular Probes) to the PCR set-up. SYBR
Green I increases its fluorescence upon binding to the PCR
product (double-stranded DNA) and this increase in fluores-
cence was measured in each PCR elongation step using an ABI
7700 thermal cycler (Applied Biosystems). For this real-time
PCR, the specificity of the primers was further increased by the
addition of Perfect Match PCR Enhancer (Stratagene) to a final
concentration of 1 mU\µl. To determine the quantitative
nature of the five different PCR tests (Danish, Lelystad,
American, Danish–Lelystad and Danish–American), PCR products were produced for each of the target sequences using
their respective primer set and cloned into the pCRII vector. As
no Danish–American PCR product was produced under the
experimental conditions used, this PCR product was made
artificially using internal primers with overlap extension, as
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19
CGBH
J. J. F. A. van Vugt and others
Fig. 3. Real-time PCR for quantification of RNA recombination. Real-time amplification plots are given for a representative
experiment to detect (A) European intra-type or (B) European–American inter-type recombination. (A) cDNA synthesized from
a supernatant sample taken 72 h post-infection from a cell culture co-infected with the Danish (m.o.i. of 0n01 TCID50 per cell)
and Lelystad (m.o.i. of 0n01 TCID50 per cell) isolates gave a CT value of 26 when using the Danish-specific forward and reverse
primers (5) and when using the Lelystad-specific forward and reverse primers (). Using the Danish-specific forward primer
and the Lelystad-specific reverse primer to detect European intra-type recombinant RNA resulted in a CT value of 34 (=). The
equivalent primer pairs used on cDNA from the European intra-type mixture control resulted in CT values of 26 (#, Danish), 26
(j, Lelystad) and 40 (i, recombinant). The latter CT value of 40 (i) was shown to represent non-specific amplification, as
determined by subsequent melting-point analysis (data not shown). (B) cDNA synthesized from a cell lysate sample taken 48 h
post-infection from a cell culture co-infected with the Danish (m.o.i. of 0n01 TCID50 per cell) and American (m.o.i. of 0n01
TCID50 per cell) isolates gave a CT value of 18 when using the Danish-specific forward and reverse primers (5) and a CT value
of 20 when using the American-specific forward and reverse primers (). Using the Danish-specific forward primer and the
American-specific reverse primer to detect inter-type recombinant RNA resulted in no amplification at all (=). The equivalent
primer pairs used on cDNA from the inter-type mixture control resulted in CT values of 21 (#, Danish) and 18 (j, American)
or no amplification (i, recombinant).
CGBI
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19
PRRSV recombination
described by others (Horton et al., 1989). Tenfold dilutions
were then produced for each of the five plasmid controls,
spanning the range of 1–10) DNA copies per reaction. Realtime PCR was performed in triplicate and the logarithm of the
number of DNA copies per sample was plotted as a function of
the PCR cycle at which the fluorescence increased above the
background level (CT value) (Ririe et al., 1997). This resulted in
a nearly perfect linear correlation (r# 0n99) for all five
different primer combinations, showing that all primer combinations resulted in a quantitative PCR with a linear range of
eight orders of magnitude (data not shown). Moreover, the CT
value versus the copy number plots allowed us to compensate
for the slightly different amplification efficiencies of these five
PCR tests (data not show). We then used the quantitative PCR
set-up to calculate the percentage of recombination in coinfected cell cultures. The percentage of recombination was
calculated by dividing the number of recombinant molecules
by the average number of non-recombinant molecules in the
sample. To find the optimal conditions for RNA recombination,
cell cultures were infected with different virus ratios and
harvested at different time-points. In a total of 43 experiments,
the percentage of European intra-type RNA recombination in
the cell lysate ranged from 0n1 to 2n5 %, with no apparent
correlation to the amount or ratio of virus used. Also, the cell
supernatant was investigated for the presence of intra-type
recombinant RNA. At 6 and 24 h after infection, no RNA
recombination between the two European isolates could be
detected in the supernatant. At 48 and 72 h after infection,
between 0n1 and 0n9 % RNA recombination was detected in the
supernatant from cells co-infected with the Danish and Lelystad
isolates (Fig. 3 A). This percentage of European intra-type
recombination found in the current study might be slightly less
than the percentage of recombination reported previously for
other arteriviruses. Among two North American PRRSV
isolates, up to 10 % RNA recombination was found in a
1182 bp region (Yuan et al., 1999) and for lactate dehydrogenase-elevating virus, up to 5 % RNA recombination was
found in a 1276 bp region (Li et al., 1999). The current study is
the first to show recombination in the European type and also,
it is the only one that has quantified precisely the level of RNA
recombination. The fragment investigated for recombination
in the current study is almost half the size of those reported
previously for arterivirus recombination (Li et al., 1999 ; Yuan
et al., 1999) and it would be expected that the relative amount
of recombination is proportional to the length of the fragment
studied. It is interesting to speculate that if there is up to 2n5 %
recombination in a 621 bp fragment, the level of RNA
recombination for the complete genome of 15 000 bp would
be up to 60 %, assuming an evenly distributed level of sequence
identity across the whole viral genome. That does not even
take into account double crossovers undetected by the current
assay (Baric et al., 1990) and also does not account for the fact
that the present study only measured Danish–Lelystad recombination and not Lelystad–Danish recombination. Of
course, it is highly speculative, but, nevertheless, the calculations suggest that, on average, more than one recombination
event might take place per viral genome per cell culture
passage.
All tested virus ratios between the Danish and Lelystad
isolate resulted in RNA recombination, showing that the
current assay was robust at generating recombinant RNA. The
same quantitative RT–PCR methodology was therefore applied to study whether there was recombination between the
European and North American types of PRRSV. Under none of
the conditions tested was recombination observed (Fig. 3 B).
Nevertheless, the quantitative real-time PCR allowed us to
determine the sensitivity of the assay to detect inter-type
recombination by comparing the minimal amount of recombinant RNA that could have been detected with the amount of
the parental virus RNA present in the co-infected cells. It was
calculated that if inter-type recombination was present at a low
level, below the sensitivity of the RT–PCR, recombination
occurred at a frequency of less than 10−'. The maximum
theoretical risk of inter-type recombination seems therefore to
be at least 10 000 times less likely to occur than European intratype recombination.
Because recombination destroys molecular clocks (Schierup
& Hein, 2000), the recent demonstration of a perfect molecular
clock in the European type of PRRSV (Forsberg et al.,
2001 ; Oleksiewicz et al., 2000) indicates that recombination
among Danish isolates of PRRSV has been a rare event in the
field. Taken together with the result of the current study, the
risk of inter-type recombination in the field might seem very
low. Furthermore, even if inter-type RNA recombinants are
generated at a low level under field conditions, there is a good
chance that such recombinants are not viable due to the low
level of similarity between the parental strains (Godeny et al.,
1993) or that inter-type recombination would be out competed
by one of the original virus isolates, as seen for the
recombinants of the American PRRSV isolates in cell culture
(Yuan et al., 1999). However, if an inter-type recombinant virus
had a selective advantage due to, for example, pre-existing
immunity to the parental viruses in the pig population, even an
extremely rare non-homologous recombination event could
rapidly spread in the pig population. It is important, therefore,
that the risk of new inter-type PRRSV recombinants is kept in
mind constantly when evaluating current and new diagnostic
procedures.
Preben Normann is thanked for the introduction to the automatic
DNA sequencer. This project was partially supported by a grant
(BIOT99-2) from the Danish Ministry of Food, Agriculture and Fishery
given to Torben Storgaard. Joke van Vugt did the experimental work as
part of her Masters degree in biology at the Wageningen University in
The Netherlands.
References
Baric, R. S., Fu, K., Schaad, M. C. & Stohlman, S. A. (1990). Establishing
a genetic recombination map for murine coronavirus strain A59
complementation groups. Virology 177, 646–656.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19
CGBJ
J. J. F. A. van Vugt and others
Bøtner, A., Strandbygaard, B., Sørensen, K. J., Have, P., Madsen,
K. G., Madsen, E. S. & Alexandersen, S. (1997). Appearance of acute
Nielsen, H. S., Oleksiewicz, M. B., Forsberg, R., Stadejek, T., Bøtner, A.
& Storgaard, T. (2001). Reversion of a live porcine reproductive and
PRRS-like symptoms in sow herds after vaccination with a modified live
PRRS vaccine. Veterinary Record 141, 497–499.
respiratory syndrome virus vaccine investigated by parallel mutations.
Journal of General Virology 82, 1263–1272.
Dewey, C., Charbonneau, G., Carman, S., Hamel, A., Nayar, G.,
Friendship, R., Eernisse, K. & Swenson, S. (2000). Lelystad-like strain
Oleksiewicz, M. B., Bøtner, A., Madsen, K. G. & Storgaard, T. (1998).
of porcine reproductive and respiratory syndrome virus (PRRSV)
identified in Canadian swine. Canadian Veterinary Journal 41, 493–494.
Sensitive detection and typing of porcine reproductive and respiratory
syndrome virus by RT–PCR amplification of whole viral genes. Veterinary
Microbiology 64, 7–22.
Forsberg, R., Oleksiewicz, M. B., Petersen, A.-M. K., Hein, J., Bøtner,
A. & Storgaard, T. (2001). A molecular clock dates the common ancestor
Oleksiewicz, M. B., Bøtner, A., Toft, P., Grubbe, T., Nielsen, J.,
Kamstrup, S. & Storgaard, T. (2000). Emergence of porcine re-
of European-type porcine reproductive and respiratory syndrome virus at
more than 10 years before the emergence of disease. Virology (in press).
Godeny, E. K., Chen, L., Kumar, S. N., Methven, S. L., Koonin, E. V. &
Brinton, M. A. (1993). Complete genomic sequence and phylogenetic
analysis of the lactate dehydrogenase-elevating virus (LDV). Virology
194, 585–596.
Horton, R. M., Hunt, H. D., Ho, S. N., Pullen, J. K. & Pease, L. R.
(1989). Engineering hybrid genes without the use of restriction
enzymes : gene splicing by overlap extension. Gene 77, 61–68.
Keffaber, K. K. (1989). Reproductive failure of unknown etiology.
American Association of Swine Practitioners 1, 1–10.
Li, K., Chen, Z. & Plagemann, P. (1999). High-frequency homologous
genetic recombination of an arterivirus, lactate dehydrogenase-elevating
virus, in mice and evolution of neuropathogenic variants. Virology 258,
73–83.
Madsen, K. G., Hansen, C. M., Madsen, E. S., Strandbygaard, B.,
Bøtner, A. & Sørensen, K. J. (1998). Sequence analysis of porcine
reproductive and respiratory syndrome virus of the American type
collected from Danish swine herds. Archives of Virology 143, 1683–1700.
Meulenberg, J. J. M., Hulst, M. M., de Meijer, E. J., Moonen, P. L. J. M.,
den Besten, A., de Kluyver, E. P., Wensvoort, G. & Moormann, R. J. M.
(1993). Lelystad virus, the causative agent of porcine epidemic abortion
and respiratory syndrome (PEARS), is related to LDV and EAV. Virology
192, 62–72.
Nagy, P. D. & Bujarski, J. J. (1995). Efficient system of homologous
RNA recombination in brome mosaic virus : sequence and structure
requirements and accuracy of crossovers. Journal of Virology 69, 131–140.
CGCA
productive and respiratory syndrome virus deletion mutants : correlation
with the porcine antibody response to a hypervariable site in the ORF 3
structural glycoprotein. Virology 267, 135–140.
Psikal, I., Moutelikova, R., Kosinova, E., Mojzis, M., Smid, B., Nejedla,
E., Indik, S. & Rodak, L. (1999). Molecular identification and genotyping
of porcine reproductive and respiratory syndrome virus (PRRSV) strains
in the Czech and Slovak Republics. Proceedings of the 3rd International
Symposium of PRRSV and Aujeszky 1, 175–176.
Ririe, K. M., Rasmussen, R. P. & Wittwer, C. T. (1997). Product
differentiation by analysis of DNA melting curves during the polymerase
chain reaction. Analytical Biochemistry 245, 154–160.
Schierup, M. H. & Hein, J. (2000). Recombination and the molecular
clock. Molecular Biology of Evolution 17, 1578–1579.
Storgaard, T., Oleksiewicz, M. & Bøtner, A. (1999). Examination of the
selective pressures on a live PRRS vaccine virus. Archives of Virology 144,
2389–2401.
Wensvoort, G., Terpstra, C., Pol, J. M. A., ter Laak, E. A., Bloemraad,
M., de Kluyver, E. P., Kragten, C., van Buiten, L., den Besten, A.,
Wagenaar, F., Broeckhuijsen, J. M., Moonen, P. L. J. M., Zetstra, T., de
Boer, E. A., Tibben, H. J., de Jong, M. F., van’t Veld, P., Groenland,
G. J. R., van Gennep, J. A., Voets, M., Verheijden, J. H. M. &
Bramskamp, J. (1991). Mystery swine disease in the Netherlands : the
isolation of Lelystad virus. Veterinary Quarterly 13, 121–130.
Yuan, S., Nelsen, C. J., Murtaugh, M. P., Schmitt, B. J. & Faaberg,
K. S. (1999). Recombination between North American strains of porcine
reproductive and respiratory syndrome virus. Virus Research 61, 87–98.
Received 1 June 2001 ; Accepted 3 August 2001
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 16 Jun 2017 04:21:19