BRIEF REPORT
Nucleotide sequence analysis of peanut stunt virus Rp strain
suggests the role of homologous recombination in cucumovirus
evolution
La´szlo´ Kiss Æ Endre Sebestye´n Æ Emese La´szlo´ Æ
Pa´l Salamon Æ Ervin Bala´zs Æ Katalin Sala´nki
Received: 29 January 2008 / Accepted: 19 May 2008 / Published online: 4 June 2008

Springer-Verlag 2008
Abstract The complete nucleotide (nt) sequence of pea-
nut stunt virus Robinia strain (PSV-Rp) was determined
and compared to other PSV strains and to representatives
of the genus Cucumovirus. Nt sequence comparison
showed 74.1–84.6% identity with the known PSV strains.
Phylogenetic analysis revealed the different origin of the
two genes encoded by RNA3. While the 3a gene clustered
with PSV-W, the coat protein gene clustered with PSV-Mi.
Recombination breakpoint analysis revealed two recombi-
nation points on RNA3. Based on these results, the
establishment of a fourth PSV subgroup is proposed. This
work revealed that homologous recombination occurred
during the evolution of PSV.
Peanut stunt virus (PSV) was first described in the United
States in 1966 and later reported all over the world except
Australia. PSV is an economically important pathogen of
legumes worldwide, causing severe symptoms on alfalfa
(Medicago sativa L.), bean (Phaseolus vulgaris L.), pea
(Pisum sativum L.), peanut (Arachis hypogaea L.), various
types of clover (Trifolium spp.) and black locust (Robinia
pseudoacacia L.) [14].
PSV is a member of the genus Cucumovirus in the
family Bromoviridae, which belongs to the alphavirus-like
superfamily. Other members of the genus are tomato
aspermy virus (TAV) and the type member cucumber
mosaic virus (CMV) [5]. As in the other members of the
genus Cucumovirus, the genome consists of three genomic
RNAs with positive polarity, designated RNA1, RNA2 and
RNA3, in order of decreasing size. RNA1 and RNA2 code
for the viral components of the replicase complex. RNA2
also codes for a small protein called 2b, which is respon-
sible for the suppression of posttranscriptional gene
silencing and also functions in host-specific long-distance
movement. This small protein overlaps with the carboxyl
terminus of the 2a protein and is translated from the sub-
genomic RNA4A. RNA3 is dicistronic and encodes two
proteins, the movement protein (3a protein) and the coat
protein (CP), which is translated from the subgenomic
RNA4 [10, 11, 15].
For a long time, PSV strains were divided into two
subgroups based on a wide range of comparative analyses,
including serology, competition hybridization and
sequence comparison. These two groups were named after
the origin of the type strains as E (eastern) and W (wes-
tern), corresponding to the current subgroups I and II,
respectively. Recently, the complete nucleotide (nt)
sequence of a strain from China was determined, and based
on nt sequence diversity, the establishment of a third
subgroup was proposed [21, 22]. However, several publi-
cations suggest that the phylogenetic relationships between
PSV strains are more complex. Richter et al. [18] proposed
up to six serogroups in Europe, and molecular hybridiza-
tion analysis showed that the Robinia strain of PSV,
L. Kiss
E. La´szlo´
K. Sala´nki (&)
Agricultural Biotechnology Center, Szent-Gyo¨rgyi Albert u. 4,
2100 Go¨do¨ll}o, Hungary
e-mail: salanki@abc.hu
L. Kiss
Department of Plant Pathology, University of Budapest,
Villa´nyi u. 29-43, 1118 Budapest, Hungary
E. Sebestye´n
E. La´szlo´
E. Bala´zs
Agricultural Research Institute of the Hungarian Academy
of Sciences, Brunszvik u. 2, 2462 Martonva´sa´r, Hungary
P. Salamon
Berkesz, Hungary
123
Arch Virol (2008) 153:1373–1377
DOI 10.1007/s00705-008-0120-z

previously named Robinia mosaic virus cannot be classi-
fied in either subgroup I or subgroup II of PSV [12]. A
natural reassortant strain was also identified (BV-15),
bearing the RNA1 characteristics of subgroup II, while
RNA2 and 3 were derived from a subgroup I strain [9].
Up till now, four complete PSV nt sequences have been
deposited in the GeneBank database, far fewer than in the
case of CMV, while the heterogeneity and taxonomic
relationships of the PSV strains seem to be rather complex.
The goal of the present study was to make a further
contribution to the phylogenetic analysis of taxonomical
relationships between the PSV strains. The complete nt
sequence of a Robinia strain of PSV was determined.
Based on the nt sequence of the Rp strain, the establish-
ment of a fourth PSV subgroup was proposed, and the
detailed sequence analysis revealed two ancient recombi-
nation points in RNA3.
PSV-Rp was isolated from black locust (R. pseudoaca-
cia L.) in Go¨do¨ll}o, Hungary. The virus was propagated on
Nicotiana clevelandii, and the virions were purified as
described by Francki et al. [6]. RNA was extracted with
phenol/SDS and then polyadenylated with poly(A) poly-
merase (Bethesda Research Laboratories, Bethesda,
Maryland, USA). The first-strand cDNA synthesis was
primed with oligo dT, and the second strand was synthe-
sized in the presence of ribonuclease H and DNA
polymerase I, as described by Gubler and Hoffman [7],
using a cDNA synthesis kit (Amersham, Buckinghamshire,
England). The cDNA obtained was blunted and ligated into
EcoRV-digested pBluescript SK(+). The clones were
selected by Northern hybridization to the PSV RNA fol-
lowed by colony hybridization. Clones were identified for
all three genomic RNAs of almost nearly full length and
sequenced from both directions using the universal M13
and M13 reverse sequencing primers and internal primers
designed according to the specific sequences. The 50 ter-
minal sequence of RNA 1, 2 and 3 was determined using
the 50 Rapid Amplification of cDNA Ends (RACE) kit
(Invitrogen, Carlsbad, CA, USA) according to the
manufacturer’s protocol. The primers were designed based
on the previously determined nt sequences (RNA1: 50GGC
TGCAGTTCAGTATGGCGCATTTTGTCCCGGACG30;
RNA2: 50GTGGAAGGCTCTTCCCGCGC30; RNA3: 50C
GGCGCGCATCCAGTGGTGTC30).
The nt and deduced amino acid (aa) sequences were
analyzed using the Emboss-Align [17] and Clustal X (using
default parameters) programs [20]. These programs were
also used for neighbor-joining analysis and genetic dis-
tance determination. Bootstrap analysis consisted of 1,000
replications. Recombination break point analysis was car-
ried out using the PDM analysis module of TOPALi v2
[13] with the following parameters: window size = 200 nt,
step size = 10 nt, adjust threshold (percentile)
significance = 0.95.
The complete nt sequence of PSV-Rp was determined
and deposited in the GeneBank database (AccNo:
AM905353, AM905354, AM905355). The length of
RNA1, RNA2 and RNA3 was found to be 3,325 nt, 2,942
nt and 2,208 nt, respectively. The five open reading frames
(ORF) characteristic of cucumoviruses were identified. All
the typical features of the cucumoviruses were present in
the determined nt sequences. The sequence comparison of
the three RNAs, the coding regions and the predicted
proteins are shown in Table 1. The comparison included
different PSV strains with known complete nt sequences
and other representatives of the genus Cucumovirus,
Table 1 Sequence identities between the RNAs, ORFs and encoded proteins of the PSV-Rp strain and those of other PSV strains, CMV and
TAV
Viruses RNA1 RNA2 RNA3
Complete ORF1a Complete ORF2a ORF2b Complete ORF3a ORF3b
%nta %nt %aab %nt %nt %aa %nt %aa %nt %nt %aa %nt %aa
PSV-J (I)c 79.6 79.6 88.9 75.0 74.4 79.2 67.3 66.0 83.9 86.2 83.8 83.3 82.9
PSV-W (II) 84.6 84.6 90.1 82.3 82.5 83.3 75.5 66.7 82.6 87.6 82.1 78.8 71.9
PSV-MI (III) 79.7 79.8 88.7 74.1 73.6 75.5 66.6 67.3 80.9 83.3 84.5 83.2 86.6
CMV-Fny 67.2 68.0 72.2 60.1 59.4 54.0 54.8 33.1 60.6 66.5 64.1 56.7 49.1
CMV-Trk 68.1 69.0 73.2 61.6 61.6 57.4 57.8 39.3 61.4 68.0 66.6 57.9 50.5
TAV-KC 66.2 67.2 68.7 60.6 60.6 55.1 52.1 30.6 66.6 74.2 75.5 68.1 72.1
The nucleotide sequences were obtained from GeneBank. Accession numbers: PSV-ER: U15728, U15729, U15730; PSV-J: D11126, D11127,
D00668; PSW-W: U33145, U33146, U31366; PSV-Mi: AY429431, AY429430, AY775057; CMV-Fny: D00356, D00355. D10538; CMV-Trk7:
AJ007933, AJ007934, L15336; TAV-KC: AJ320273, AJ320274, AJ237849
a Nucleotide
b Amino acid
c PSV strain subgroup assignment is given in brackets
1374 L. Kiss et al.
123

namely CMV-Fny, CMV-Trk7 and TAV-KC. The nt
sequence identities for the complete genomic RNAs among
the PSV-Rp and other PSV strains were between 74.1 and
84.6% (corresponding to 74.1–84.6% similarity), and it
was in the range of 60.1–68.1% with other cucumoviruses.
According to the ICTV criteria for demarcating species in
the genus Cucumovirus, subgroups generally have at least
65% sequence similarity [19], and thus the values proved
that PSV-Rp is a strain of PSV. In the case of RNA1, 2 and
proteins 1a, 2a PSV-Rp showed the highest degree of
identity with PSV-W, but in the case of the 2b protein, the
identity was greater with PSV-Mi. This protein was the
most diverse of all the PSV proteins, showing only 66.0–
67.3% identity. It is worth noting that aa identity values
extended in a narrow range, especially in the case of the 1a
and 2b proteins. In the case of RNA3, the level of identity
was fairly similar with all the isolates (80.9–83.9%), but
the identity relationships varied for different regions of
RNA3. While the complete nt sequence showed the highest
degree of identity with PSV-J (83.9%), the 3a ORF showed
the highest degree of identity with PSV-W (87.6%), while
the CP ORF was most similar to that of PSV-Mi and PSV-
J. For the CP ORF, the identity was almost the same with
PSV-J (83.3%) and PSV-Mi (83.2%) at the nt level, while
at the aa level, sequence identity was greater with the PSV-
Mi strain (86.6%).
Phylogenetic analysis based on protein coding regions
was also carried out including the different PSV strains and
representatives of various cucumoviruses (data not shown).
In the case of proteins 1a, 2a and 3a, PSV-Rp clustered with
PSV-W, but in the case of the CP it clustered with PSV-Mi.
The different phylogenetic backgrounds and the previous
sequence comparisons of the two proteins coded by RNA3
suggested the presence of recombination in the evolutionary
history of RNA3. To identify potential recombination
events, recombination breakpoint analysis was carried out
on the whole RNA3. Two clear peaks on the phylogenetic
correlation profile (Fig. 1) indicated recombination break-
points around nt 1,199 and 1,873 (numbers refer to the
PSV-Rp sequence). One of the recombination points was
localized in the intergenic region, while the other was
identified at the 30 end of the CP coding region. The phy-
logenetic analysis of the three regions revealed two sister
groups in each region. While the 50 region and the 30 end of
PSV-Rp clustered with PSV-W, the region consisting
mainly of the CP of PSV-Rp clustered with PSV-Mi,
Fig. 1 Phylogenetic analysis of the different regions of PSV RNA3.
At the bottom, the probabilistic divergence measures (PDM) statistic
of the TOPALi v2 program package (sliding window size = 200 nt,
step size = 10 nt) is shown. Peaks in PDM values above the dashed
line indicate proposed recombination points significant at the 5%
level. Above it, a schematic representation of PSV RNA3 is
presented. Thick arrowheads show recombination points. Square
braces show the genomic regions corresponding to the three
phylograms
Nucleotide sequence of peanut stunt virus Rp strain 1375
123

proving the different origin of the three regions. This result
was in a good agreement with the data on nucleic acid
identity. Although the relationship between the branch
length and real time is far from straightforward, the length
corresponds to the evolutionary distance between the two
nodes they connect. In the present case, the branch length
referred to an ancient recombination event.
In contrast to CMV, the taxonomical relationships
between the various PSV strains are still in question.
Previous results classified the Robinia strain of PSV
(beforehand called Robinia mosaic virus, RoMV) to the
PSV subgroup I, based on serological reaction, but
hybridization analysis suggests that RoMV could not be
unequivocally classified to PSV subgroup I or II [12]. The
final classification of RoMV strains and the assessment of
their relationship with known PSV subgroups awaited
determination of the nt sequence. In the present study, the
complete nt sequence of PSV-Rp was determined. Hajim-
orad et al. [8] proposed that cucumovirus strains classified
within the same subgroup should have at least 90% nt
sequence identity for the whole genome, while the nt
sequence identity between the different subgroups should
be less than 80%. PSV-Rp is clearly distinct from other
PSV isolates, since the nt sequence identity with the most
closely related strain is only slightly above 80%. PSV-Rp
has the greatest identity to PSV-W for RNA 1 and 2 (84.6
and 82.3%, respectively), but in the case of RNA3, all the
strains are fairly similar (80.9–83.9%). Since these identity
values are far from the expected 90% between the members
of a subgroup, it is proposed that a fourth PSV subgroup
should be established.
An analysis of RNA3 gave clear evidence that homol-
ogous recombination played a role in the evolution of PSV.
One of the recombination points was located 50 of the CP
coding region, while the other was located within this
region, 47 nt ahead of the stop codon, thus nearly the
complete CP was exchanged. Previously, natural recom-
bination points were identified between different
cucumoviruses (CMV and TAV) in doubly infected plants,
and more recently, between CMV subgroups under
greenhouse conditions. In these cases, homologous
recombination points were distributed throughout RNA3,
including the non-coding regions, the movement protein
and the CP [3, 16], but the resulting recombinants were
counter-selected. On the other hand, Bonnet et al. [1]
concluded that the emergence of recombinant proteins
involves a fitness cost, so the exchange of whole genes is
more probable if the infectious nature of the recombinant is
preserved as in the present.
The role of recombination in the evolution of field CMV
strains was analyzed in detail [1]. Recombinants between
the IA and IB subgroup strains were identified by ribonu-
clease protection assay, and the genetic structure of the
virus population suggested that recombinant genotypes are
counter-selected [1, 4] In most cases, genetic exchange
seems to have a fitness cost, and hybrid genotypes disap-
pear from the population [1, 16]. Nevertheless, in some
special cases, the recombinants might have an advantage in
virus evolution and host adaptation, as in the case of CMV
infecting Peruvian lily (Alstromeria) [2] or the lily infect-
ing CMV from Korea and Japan in which recombination
points have also been identified [1]. The present findings
suggest that homologous recombination occurred during
the evolution of PSV. Competition experiments on differ-
ent hosts could answer the question if evolutionary
advantages resulted in the survival of this recombinant
strain.
In summary, these results present molecular evidence
that PSV-Rp belongs to a distinct PSV subgroup and
indicate that an ancient recombination event plays a role in
the evolution of PSV.
Acknowledgments This work was funded by the projects OTKA
T048683 and TS61023. Katalin Sala´nki was the recipient of a Bolyai
Ja´nos fellowship from the Hungarian Academy of Sciences.
References
1. Bonnet J, Fraile A, Sacristan S, Malpica JM, Garcia-Arenal F
(2005) Role of recombination in the evolution of natural popu-
lations of Cucumber mosaic virus, a tripartite RNA plant virus.
Virology 332:359–368
2. Chen YK, Goldbach R, Prins M (2002) Inter- and intramolecular
recombinations in the cucumber mosaic virus genome related to
adaptation to alstroemeria. J Virol 76:4119–4124
3. de Wispelaere M, Gaubert S, Trouilloud S, Belin C, Tepfer M
(2005) A map of the diversity of RNA3 recombinants appearing
in plants infected with Cucumber mosaic virus and Tomato
aspermy virus. Virology 331:117–127
4. Escriu F, Fraile A, Garcı´a-Arenal F (2007) Constraints to genetic
exchange support gene coadaptation in a tripartite RNA virus.
PLoS Pathog 3:e8. doi:10.1371/s0030008
5. Fauquet CM, Mayo MA (2001) The Bromoviridae. In: Virus
taxonomy, seventh report of the international committee on tax-
onomy of viruses
6. Francki RIB, Mossop DW, Hatta T (1979) Cucumber mosaic
virus. Descriptions of Plant Viruses. No. 213. (No. 1. rev.)
Common Mycol Inst/Assoc Appl Biol, Kew, Surrey, England
7. Gubler U, Hoffman BJ (1983) A simple and very efficient method
for generating cDNA libraries. Gene 25:263–269
8. Hajimorad MR, Hu CC, Ghabrial SA (1999) Molecular charac-
terization of an atypical old world strain of peanut stunt virus.
Arch Virol 144:1587–1600
9. Hu CC, Ghabrial SA (1998) Molecular evidence that strain
BV-15 of peanut stunt cucumovirus is a reassortant between
subgroup I and II strains. Phytopathology 88:92–97
10. Hu C-C, Aboul-Ata AE, Naidu RA, Ghabrial SA (1997) Evidence
for the occurrence of two distinct subgroups of peanut stunt cu-
cumovirus strains: molecular characterization of RNA3. J Gen
Virol 78:929–939
11. Hu C-C, Sanger M, Ghabrial SA (1998) Production of infectious
RNA transcripts from full length cDNA clones representing two
1376 L. Kiss et al.
123

subgroups of peanut stunt virus strains: mapping satellite RNA
support to RNA1. J Gen Virol 79:2013–2021
12. Militao V, Moreno I, Rodriguez-Cerezo E, Garcia-Arenal F
(1998) Differential interactions among isolates of peanut stunt
cucumovirus and its satellite RNA. J Gen Virol 79:177–184
13. Milne I, Wright F, Rowe G, Marshall DF, Hushmeier D, McGuire
G (2004) TOPALi: software for automatic identification of
recombinant sequences within DNA multiple alignments.
Bioinformatics 20:1806–1807
14. Mink GI (1972) Peanut stunt virus. CMI/AAB Description of
Plant Viruses No. 92
15. Palukaitis P, Garcia-Arenal F (2003) Cucumoviruses. Adv Virus
Res 62:241–323
16. Pierrugues O, Guilbaud L, Fernandez-Delmond I, Fabre F, Tepfer
M, Jacquemond M (2007) Biological properties and relative
fitness of inter-subgroup cucumber mosaic virus RNA 3 recom-
binants produced in vitro. J GenVirol 88:2852–2861
17. Rice P, Longden I, Bleasby A (2000) EMBOSS: the European
molecular open software suite. Trends Genet 16:276–277
18. Richter J, Haack M, Wesemann M, Beczner L (1987) Differen-
tiation of peanut stunt virus isolates in six serotypes. Arch
Phytopathol Pflanzenschutz 2:127–133
19. Roosinck MJ, Bujarski J, Ding SW, Hajimorad R, Hanada K,
Scott S, Tousgnant M (2006) Index of viruses—Bromoviridae.
In: Bu¨chen-Osmond C (ed) ICTVdB—the universal virus
database, version 4. Columbia University, New York, USA.
http://www.ncbi.nlm.nih.gov/ICTVdb/Ictv/fs_index.htm
20. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins
DG (1997) The ClustalX windows interface: flexible strategies
for multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res 24:4876–4882
21. Xu Z, Higgins C, Chen K, Dietzgen RG, Zhang Z, Yan L, Fang X
(1998) Evidence for a third taxonomic subgroup of peanut stunt
virus from China. Plant Dis 82:992–998
22. Yan LY, Xu ZY, Goldbach R, Kunrong C, Prins M (2005)
Nucleotide sequence analyses of genomic RNAs of Peanut stunt
virus Mi, the type strain representative of a novel PSV subgroup
from China. Arch Virol 150:1203–1211
Nucleotide sequence of peanut stunt virus Rp strain 1377
123