Virus Research 112 (2005) 4251

Emergence of Rsv-resistance breaking Soybean mosaic virus

isolates from Korean soybean cultivars

B.K. Choi
a
, J.M. Koo
a
, H.J. Ahn
a
, H.J. Yum
a
, C.W. Choi
a,b,
,
K.H. Ryu
c
, P. Chen
d
, S.A. Tolin
e
a
Department of Biology and Medicinal Science, Pai Chai University, Daejeon 302-735, Republic of Korea
b
Bio-medicinal Research Center (RRC), Pai Chai University, Daejeon 302-735, Republic of Korea
c
Plant Virus GenBank, Division of Life and Environmental Sciences, Seoul Womens University, Seoul 139-774, Republic of Korea
d
Department of Crop, Soil, and Environmental Sciences, University of Arkansas, Fayetteville, AR 72701, USA
e
Department of Plant Pathology, Physiology and Weed Science, Virginia Tech, Blacksburg, VA 24061, USA
Received 14 July 2004; received in revised form 9 March 2005; accepted 9 March 2005
Available online 24 May 2005
Abstract
Twelve Rsv resistance-breaking (RB) isolates of Soybean mosaic virus (SMV) were obtained from eld-grown soybean plants showing
mosaic symptoms and subsequently examined biologically and molecularly. All of these RB isolates were identied as SMV based on
serological and infectivity assays, and the amplication of P1 gene products by reverse transcription-polymerase chain reaction (RT-PCR).
Differential soybean cultivars, lines or accessions Lee 68 (rsv), PI 96983, York, Marshall, Ogden, Kwanggyo, Suweon 97 (Rsv1 alleles),
L29 (Rsv3), and V94-5152 (Rsv4), following inoculation with each RB isolate, showed similar systemic symptoms suggesting that these
RB isolates can overcome Rsv resistance at three loci. To differentiate the 12 RB isolates molecularly, the P1 coding region for each isolate
was amplied, cloned, sequenced and compared to known SMV strains. The P1 region from the RB isolates shared 8690% and 9099%
similarities in amino acid (aa) and nucleotide sequence, respectively, with known SMV strains. Comparison of aa sequences indicated that
these RB isolates are newly emerging isolates capable of breaking Rsv resistance. Phylogenetic analysis further suggested that the RB isolates
can be classied as three major types. However, recombination was not observed within the coding region of P1 protein among the types.
This is the rst report on the emergence of SMV isolates capable of overcoming all of the known resistance alleles at the Rsv1 locus, as well
as distinct resistance genes at Rsv3 and Rsv4.
2005 Elsevier B.V. All rights reserved.
Keywords: Rsv resistance-breaking (RB) isolates; Soybean mosaic virus (SMV); P1 protein; Amino acid sequences; Phylogenetic analysis
1. Introduction
Soybean mosaic virus (SMV), a member of the genus
Potyvirus, is one of the most prevalent viral pathogens of

The nucleotide sequences reported in this paper have been submitted to

the EMBL database and assigned accession numbers AJ639655 for SMV-
CN1, AJ639654 for SMV-CN2, AJ558194 for SMV-CN3, AJ639653 for
SMV-CN7, AJ639652 for SMV-CN9, AJ639651 for SMV-CN10, AJ639650
for SMV-CN12, AJ639649 for SMV-CN13, AJ639648 for SMV-CN15,
AJ290450 for SMV-CN18, AJ639647 for SMV-CN31, and AJ639646 for
SMV-CN36.

Corresponding author. Tel.: +82 42 520 5617; fax: +82 42 520 5617.
E-mail address: choicw@mail.pcu.ac.kr (C.W. Choi).
soybean [Glycine max (L.) Merr.] worldwide. The virus
has a positive-sense single-stranded RNA genome of 9588
nucleotides with VPg at the 5

end and poly(A) tail at the

3

end. The genome of potyviruses encodes a single large

polyprotein that is subsequently cleaved by virus-encoded
proteases (Dougherty and Semler, 1993). The rst protein in
the N-terminal region of polyprotein, P1 protease, catalyzes
the cleavage at a Tyr-Ser dipeptide between itself and
helper component (HC-Pro, protease) protein (Mavankal
and Rhoads, 1991; Verchot et al., 1992), and is essential
for genome amplication (Verchot and Carrington, 1995).
The diversity of potyviral P1 protein was proposed as
an important parameter to distinguish different strains of
0168-1702/$ see front matter 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.virusres.2005.03.020
B.K. Choi et al. / Virus Research 112 (2005) 4251 43
a virus and to investigate the evolutionary relationships
between different isolates (Lin et al., 2001; Domier et al.,
2003). The P1 protein may also affect symptomdevelopment,
host range, and geographical distribution (Lee and Wong,
1998).
Genetic resistance of soybean cultivars has been the most
effective means of managing the disease caused by SMV.
A single dominant gene for SMV resistance was rst iden-
tied in soybean PI 96983 and designated as Rsv1 (Kiihl
and Hartwig, 1979). Additional alleles at the same locus are
carried by cultivars York, Marshall, Kwanggyo, Ogden, and
Suweon 97 (Chen et al., 1991; Chen et al., 2002a). The Rsv1
allele-carrying genotypes exhibit different reaction combina-
tions in response to SMV strain groups G1 through G7 de-
scribed in the USA (Cho and Goodman, 1979, 1982), Japan
(Takahashi et al., 1980), China (Pu et al., 1982), and Korea
(Cho and Chung, 1986). The system used to assign SMV
isolate to pathotype is thus based on Rsv1-containing differ-
ential cultivars, with more virulent pathotypes overcoming a
greater number of Rsv1 alleles (Chen et al., 1991, 2002a). In
addition, the Rsv3 resistance gene has been identied in the
Williams isoline L29 (Buss et al., 1999), and found to con-
fer susceptibility to pathotypes of lesser virulence (G1G3)
but resistance to more virulent strains (G5G7) (Gunduz et
al., 2002). The Rsv4 resistance gene conferring resistance to
SMV G1G7 has been found in V94-5152 soybean derived
from PI 486355 (Buss et al., 1997; Ma et al., 1995), as well
as in PI 88788 and Peking (Gunduz et al., 2004).
In Korea, SMV was initially recognized as a mosaic-
causing pathogen in soybean in the early 1970s (Chung et al.,
1974). Soybean cultivars resistant to SMV were developed
and within a short time, a severe necrotic strain, SMV-N,
was found in some of the resistant cultivars, causing serious
yield losses in a large production area (Cho et al., 1977).
Since then, strains in soybean in Korea have been moni-
tored by pathotype. More than seven SMV strains existed
during 1980s, among which G5H was the most prevalent
(Cho et al., 1983). However the relative incidence of G5H
decreased gradually due to the utilization of resistant soybean
cultivars such as Ilpumgeomjeongkong, Jangmikong, Jang-
sukong, Jinpumkong, Muhankong, Myeongjunamulkong,
Pungsannamulkong, Sodamkong, and Soyangkong (Lee et
al., 1992; Cho, 1995). In 1999 and 2000, SMV-G5H was
estimated to represent only 4% of total SMV population
in the Suweon area (Kim, 2000). In the 1990s, another se-
vere isolate emerged, causing necrosis on cultivars Kwang-
gyo and Suweon 97, similar to the biological properties of
SMV-G7 (Cho, 1995; Kim, 2000; Kim et al., 2003). Desig-
nated as SMV-G7H, this isolate was found to share 94% nu-
cleotide (nt) sequence homology with G7 (Lim et al., 2003).
In 19992000, SMV-G7H was reported as a predominant
strain in Suweon, Korea, as it was isolated from more than
50% of SMV-infected plants, including those resistant to
SMV-G5H (Kim, 2000).
The susceptibility of resistant cultivars in Korea is
attributed to the emergence of resistance-breaking SMV
strains, most likely from selection imposed on this seed-
borne virus by the widespread use of Rsv resistance
(Harrison, 2002). It is unclear, however, what evolutionary
or genetic process changed SMV to enable the selected virus
to overcome Rsv-mediated resistance. Therefore, this work
was undertaken to systematically isolate and document the
occurrence of new RB SMV strains in Korea and to examine
their biological diversity based on reactions of several
Rsv-containing differential soybean lines. Among the RB
isolates are the rst reported SMV strains that overcome
resistance to SMV in soybean at all three of the known loci,
Rsv1, Rsv3, and Rsv4. The genetic diversity of the new RB
isolates was assessed at the molecular level by comparison
of sequences of the P1 region with those of thirteen previous
reported SMV strains.
2. Materials and methods
2.1. Viral origin and differential cultivar response
The 12 Chungnam SMV isolates (designated as CN plus
an isolate number) used in this study were collected fromnat-
urally infected soybean showing typical mosaic symptoms in
an experimental eld where national soybean variety testing
was conducted in 19982000 (Table 1). The eld was lo-
cated at the Chungnam Agricultural Research and Extension
Services (CARES Station), Daejeon, Korea. Two additional
isolates, CN31 and CN36, were collected fromsoybean elds
adjacent to the station. All collected isolates listed in Table 1
were identied as SMV by a positive reaction with SMV-G1
antiserum (ATCC PVAS-94) in agarose gel double diffusion
assays (Hunst and Tolin, 1982), subjected to successive isola-
tions fromsingle lesiontransfers onbeanTopcrop (Milbrath
and Soong, 1976), and maintained in soybean cv. Lee 68 (rsv)
by mechanical inoculation. To determine biological proper-
ties of those isolates, several differential cultivars, accessions,
and lines (hereafter referred to as cultivars), including Lee
68, PI 96983, York, Marshall, Kwanggyo, Ogden, Suweon
97, L29, and V94-5152, were inoculated with each isolate,
monitored for symptomdevelopment for 23 weeks in a tem-
perature controlled greenhouse, and conrmed for SMV sys-
temic infection by RT-PCR(see below) for the presence of P1
gene sequences. Soybean seeds of differential cultivars were
obtained from Dr. J.K. Moon of National Crop Experimental
Station in Suweon, Korea and Dr. P. Chen at University of
Arkansas, USA.
2.2. Preparation of virus, RNA extraction, and RT-PCR
amplication
Within 23 weeks post inoculation, each viral isolate was
puried from symptomatic leaf tissues of Lee 68 soybean
(strain/isolate maintenance host) according to the method of
Hunst and Tolin (1982). Viral RNA was extracted from pu-
ried virus by disruption with SDS and proteinase K, fol-
44 B.K. Choi et al. / Virus Research 112 (2005) 4251
Table 1
Origin and biological properties of new resistance-breaking isolates of Soybean mosaic virus collected from eld-grown soybean in Daejeon, Korea
SMV Isolates Cross reaction
a
RT-PCR
b
Collected cultivars Cultivar origin (genotype) or Reaction to SMV strain
c
CN1 + + Kangwon 2 (Rsv1) NC
CN2 + + Chungnam 1 NC
d
CN3 + + Suweon 195 Suweon97 (Rsv1-h) G1(R), G3(R), G4(R), G5(R), G6 (R), G7 (R),
G5H (R), G7H (R)
CN7 + + Sinpaldakong 2 Kwanggyo (Rsv1-k) G1 (R), G3 (R), G4 (R), G5 (N), G6 (R), G7
(N), G5H (R), G7H (N/R)
CN9 + + Iksannamulkong Essex (rsv1) x Hill G1(M), G3(R), G4(R), G5(M), G6 (M), G7 (M),
G5H (M), G7H (M)
CN10 + + Sobaegnamulkong Williams (rsv) G1(M), G3(M), G4(M), G5(M), G6 (M), G7 (M),
G5H (M), G7H (M)
CN12 + + Danbaegkong Dongsan69 x D76-8070 G1(M), G3(R), G4(M), G5(M), G6 (M), G7
(M), G5H (M), G7H (M)
CN13 + + Myeongjunamulkong Jangyeobkong x Baegunkong G1(R), G3(R), G4(R), G5(R), G6 (R),
G5H (R), G7H (R)
CN15 + + Pereunkong L78-379 (Rsv1) or Raiden (Rsv1-r) G1(R), G3(R), G4(N), G5(R),
G6 (R), G7 (N),G5H(R), G7H (R)
CN18 + + Taeankong NC
CN31 + + Unknown NC
CN36 + + Unknown NC
G7 + + ATCC/PV723 (purchased)
a
+ indicates positive reaction to SMV-G1 antiserum (ATCC PVAS-94) on agarose gel double diffusion.
b
+ indicates the presence of amplied P1 gene by reverse transcription-polymerase chain reaction.
c
Reference: Kim (2000); R, resistant; N, necrotic; M, mosaic.
d
Genotype not characterized or no data available.
lowed by serial phenol/chloroform extraction. The primers
for P1 sequence detection (forward primer: AGTCAAATG-
GCAACAATCATGandreverse primer: GGGATTAGTGCT-
GAATATCC) were synthesized according to the conserved
nucleotide (nt) sequences in the same region of SMV-G2, -
G7 (Jayaram et al., 1992) and -N (Eggenberger et al., 1989).
Reverse transcription (RT) was performed using 100 ng of
viral RNA in a reaction volume of 20 l containing 50 mM
TrisHCl (pH 8.3), 75 mM KCl, 10 mM DTT, 3 mM MgCl
2
,
1 mMdNTP, 50 pmol of reverse primer, 20 units of RNase in-
hibitor (Takara, Japan), and 200 U Molony murine leukemia
virus reverse transcriptase (Promega, USA). For PCR am-
plication, 20 l of the RT product was added to 80 l
of reaction mixture containing 10 mM TrisHCl (pH 8.3),
50 mM KCl, 1.54.5 mM MgCl
2
, 2.5 U Taq polymerase
(Takara, Japan), 0.25 mM dNTP, and 50 pmol of each for-
ward and reverse primer. The thermocycler (Bio-Rad, Gene
Cycler, USA) was programmed for 35 cycles of template
denaturation at 94

C for 1 min, primer annealing at 50

C
for 2 min, and DNA synthesis at 72

C for 3 min. Strain

SMV-G7 (ATCC PV-723) was included in RT-PCR as a
reference. To minimize the experimental errors associated
with the reverse transcriptase and Taq DNA polymerase
activities, the amplication process was repeated several
times. Identical results were achieved from each ampli-
cation.
2.3. Cloning and sequencing RT-PCR products
PCR-amplied DNA fragments of the P1 gene were gel
puried using a Gel Extraction Kit (QIAGEN, Germany) and
ligated into pGEM-T Easy Vector (Promega, USA) accord-
ing to the manufacturers instructions, and the ligated re-
combinants were transformed into E. coli JM109 cells. Plas-
mids were prepared from the transformed bacterial cells us-
ing QIA Plasmid Prep Kit (QIAGEN, Germany) and used
for sequencing reactions. Nucleotide sequences were deter-
mined in both directions with an automated DNA sequencer
(ABI 377, USA) and then veried by the ABI3700 se-
quencer, and the conrmed sequences were registered on the
EMBL database. For sequence comparisons, P1 gene prod-
ucts of the reference strains or isolates were obtained from
the NCBI data library including SMV-G2 (S42280) and -G7
(AF241739) (Jayaram et al., 1992), SMV-G7d (AY216987)
(Hajimorad et al., 2003), SMV-N (NC002634) (Eggenberger
et al., 1989), SMV-G5b (AY294044: then designated G5 in
Korea) and -G7H (AY294045) (Lim et al., 2003), SMV-
HH5 (AJ310200: then designated Huanghuai) and HZ
(AJ312439: then designated Severe) (Chen et al., 2004), as
well as web-published isolates SMV-G1 (AF200558), -G3
(AF200561), -G5a (AF200564:then designated G5 in USA),
-G6 (AF200567), and -Aa (AB100442).
2.4. Phylogenetic analysis
Distance matrices for complete P1 sequences were
calculated from the multiple sequence alignments by
the CLUSTALW service (http://www.ebi.ac.uk/clustalw)
(Higgins et al., 1994) and by DNAMAN version 5.2.9 (Lyn-
nonBiosoft, Quebec, Canada). For bothtypes of phylogenetic
analyses, the same coding region (NC003224) of Zucchini
yellow mosaic virus (ZYMV) was included as an outgroup
B.K. Choi et al. / Virus Research 112 (2005) 4251 45
for the phylogenetic comparisons. Phylogenetic analysis was
rst performed using the generated matrices as an input in
DNAMAN to build an unrooted tree, and the statistical sig-
nicance of branching was estimated by bootstrap resam-
pled data sets based on 1000 replications. A second phyloge-
netic tree was derived by the neighbor-joining (NJ) method
(Saitou and Nei, 1987), distances were corrected by the two-
parameter method of Kimura (1983), and inferred gaps in
the multiple alignments were ignored. Systematic screening
for the presence of recombination patterns was performed
using nucleotide alignments and the UPGMA-based Recom-
bination Detection Program (RDP), which used a pairwise-
distance scanning approach coupled with an UPGMA-based
reconstruction algorithm (Martin and Rybicki, 2000). The
probability that the nucleotide arrangement in a recombinant
region may have occurred by chance was tested using a bino-
mial distribution. Sliding windows ranging from 5 to 100 nt
were tested.
3. Results and discussion
3.1. Response of differential cultivars and lines
It was evident that Rsv RB isolates emerged in nature be-
cause eld-grown, SMV-resistant soybean cultivars possess-
ing Rsv1 alleles were naturally infected with SMV. Of the 12
isolates collected, only CN10 was collected from a cultivar
susceptible to all strains, and two, CN9 and CN12, were from
sources resistant only to SMV-G3 (Table 1). The source cul-
tivars of isolates CN1, CN3, CN7, and CN15 were known to
have Rsv1 or an allele of Rsv1. The genotype or the source
cultivar for the remaining isolates was either uncharacter-
ized or unknown (Table 1). To determine the pathotypes of
collected isolates, each was inoculated to a set of SMV dif-
ferential soybean cultivars with Rsv1 alleles, or with an Rsv3
or Rsv4 gene. As shown in Table 2, all 12 isolates gave rise
to typical systemic mosaic symptoms in all cultivars tested,
even though these cultivars readily differentiated SMV type
isolates and variants of the pathotypes G4 through G7. The
responses to the RB isolates were more like that of Lee 68 or
of York, both of which are susceptible to SMV G5 through
G7. The same results were obtained in several repeated inoc-
ulations, conrming the collected RB SMV isolates cannot
be differentiated on these soybean cultivars. Even Suweon
97 (Rsv1-h) and V94-5152 (Rsv4), which have been demon-
strated to be valuable sources of resistance in the USA as
they resist all identied SMV strains (Chen et al., 2002a;
Buss et al., 1997), were susceptible to the 12 newRBisolates
in this study. The differential cultivars carrying Rsv1, Rsv3
or Rsv4 failed to inhibit viral replication and spread. Thus,
these newly emerged RB SMV isolates represent new and
unique SMV pathotypes that cannot be classied to strain
group in the SMV classication system of Cho and Good-
man (1979, 1982). We do not yet know the prevalence of
these RB isolates in other areas of Korea, and whether this
prevalence will stay at the same or increase, even in the ab-
sence of resistant cultivars. If these RB isolates were to be-
come prevalent, they will likely dominate the once preva-
lent strains of lesser virulence such as SMV-N, -G2, G3, -
G5, -G5H, and -G7H, to which resistances are available in
germplasm.
Table 2
Differential responses of soybean cultivars inoculated with eld-collected isolates and reference strains of Soybean mosaic virus
SMV isolates Cultivar response to SMV isolates
a
Lee68
(rsv)
York
(Rsv1-y)
Marshall
(Rsv1-m)
Ogden
(Rsv1-t)
Kwanggyo
(Rsv1-k)
PI96983
(Rsv1)
Suweon97
(Rsv1-h)
L29
(Rsv3)
V94-5152
(Rsv4)
CN1 S S S S S S S S S
CN2 S S S S S S S S S
CN3 S S S S S S S S S
CN7 S S S S S S S S S
CN9 S S S S S S S S S
CN10 S S S S S S S S S
CN12 S S S S S S S S S
CN13 S S S S S S S S S
CN15 S S S S S S S S S
CN18 S S S S S S S S S
CN31 S S S S S S S S S
CN36 S S S S S S S S S
G7
b
S S N N N N R R R
G7H
c
S N N N N
G6
b
S S N R N R R R R
G5
b
S S R R N R R R R
G5H
c
S N R N N
G4
b
S N R R R R R S R
a
R, resistant; N, local necrotic lesion and systemic necrosis; S, susceptible systemic mosaic; , not determined.
b
Reference strain: Ma et al. (2003).
c
Reference strain: Kim (2000); Kim et al. (2003).
46 B.K. Choi et al. / Virus Research 112 (2005) 4251
3.2. Sequence analysis and comparison
To differentiate the 12 biologically indistinguishable iso-
lates and characterize their genetic variability, complete nt
sequences and amino acid (aa) sequences encoding the P1
gene were obtained and analyzed by computer-based pro-
grams to compare with known strains. The P1 protein was
known as the least conserved region of the entire polyprotein
of potyvirus, especially in the N-terminal half having hy-
pervariation in length and in sequence (Vance et al., 1992).
The P1 region is variable among SMVisolates (Domier et al.,
2003), andhas alsobeenimplicatedinrecombinational events
(Chen et al., 2004), The 308 aa of P1 protein for SMV-G1, -
G2 -G7 and -Nwere aligned with the corresponding 309 aa of
other isolates by DNAMAN version 5.2.9 (Fig. 1). Another
input in the fasta format was done with CLUSTALW and
showed same multiple alignments (data not shown). Interest-
ingly, four SMV strains (G1, G2, G7 and N) have one less
amino acid, at the position 198 from the N-terminus, as com-
pared to others. Signicant N-terminal variations were found
in aa sequences of G2 and G7 strains, particularly in the 16 aa
residues located 92107 from the N-terminus. However, the
aa variability in the N-terminus of other SMVisolates was not
frequent, but rather was conservatively distributed over the
entire coding region. The C-terminal region of the P1 protein
in potyviruses contains highly conserved residues that are
responsible for its self-cleaving protease activity (Verchot et
al., 1992). Essential aa known for proteolysis were also found
in all SMV isolates for catalytic triad composed of His
222
,
Ser
263
and Phe-Val-Val-Arg-Gly between positions 282 and
286, except substitution of Ile for rst Val (Fig. 1).
The sequence alignments for strain comparison showed
that the similarities were variable among 25 SMV isolates
with 9099% for nt and 8699% for aa. A minimum aa sim-
ilarity was observed between SMV-G7 and -CN3, -CN7, -
CN10, -CN31 or -G6 (Table 3). In all pairwise comparisons,
the P1 aa sequence similarities in most cases were lower than
the nt sequence. SMV-G7H has been a predominant strain in
Korea (Kim, 2000; Kim et al., 2003) and four RB isolates in
this study (CN1, CN9, CN12 and CN13) showed close relat-
edness (97% aa similarity) with G7H. The overall order of
the % aa similarity with G7H was G5b >HH5 >CN1, CN9,
CN12, CN13, G1, G5a, HZ, and N>CN18 >CN2 >CN10,
G6, Aa, and G7d >CN15 >CN36 and G3 >G2 >CN3 and
CN31 >CN7 >G7 (Table 3). It should be noted that the aa
similarity between G7H and G5b was higher (99%, with
99%nt similarity) than that between G7Hand G7 (89%, with
95%nt similarity). The results support the previous sugges-
tion that there has been a possible genetic change fromSMV-
G5b to the more virulent strain -G7H overtime (Kim, 2000).
3.3. Phylogenetic analysis
Phylogenetic analysis is useful in studying evolutional
and epidemiological history of a virus, especially in tracing
the origin of emerging virus (Tomimura et al., 2003).
Two different analyses, DNAMAN version 5.2.9 and
CLUSTALW, were performed in this study to determine the
relationships between the P1 proteins of 25 SMVisolates and
known strains. Both dendrogram trees revealed that although
considerable divergence exists among P1 protein of known
SMV strains and RB isolates, all these diverse SMV isolates
can be grouped into three major types by similarity clustering
(Fig. 2Aand B). In both trees, type I included four subgroups
in common: A (SMV-CN3, -CN7 and -CN31), B (SMV-Aa,
-CN15, -CN36, -G3, and -G7d), C (SMV-G1 and-N and/or
-G2 and -G7) and D (SMV-HH5 and HZ). However, a minor
difference was found in clustering between two trees. A
single clade (subtype C) formed among SMV-G1, -G2, -G7,
and -Nin the NJ tree (Fig. 2B), while they were separated into
different branches in the DNAMAN tree (Fig. 2A). Three
RB isolates (SMV-CN7, -CN10 and -CN13) in subtype A
were in the same clade and appeared to be clonal variants
as indicated by 100% bootstrap value (Fig. 2A). In subtype
B, a Japanese isolate SMV-Aa and two Korean RB isolates
-CN15 and -CN36 formed a clade with -G3, as implied by
bootstrap value of 76%. The values of bootstrap analysis
greater than 70% are normally considered as supporting
the grouping (Hillis and Bull, 1993). Type II represents two
subtypes E and F. Although SMV-G6 formed a clade with
-CN9, -G5a and -CN13 as a subtype F, the bootstrap value
of this clade was low (<50%). Type III consisted of Korean
isolates SMV-CN1, -CN12, -G5b and -G7H, suggesting that
RB isolates -CN1 and -CN12 share the same origin with
predominant SMV strains in Korea. The data presented here
suggest that genetic diversity exists among the emerging
pathotypes of SMV, and among subtypes within a specic
pathotype.
It is logical to hypothesize that the emergence of the RB
SMV isolates in this study may have been due to genetic re-
combination. Recently, recombination has been shown to be
a signicant phenomenon of genetic evolution of potyviruses
(Bousalem et al., 2000; Glais et al., 2002; Tomimura et al.,
2003; Moreno et al., 2004), as well as other RNAplant viruses
(Aaziz and Tepfer, 1999; Nagy et al., 1999; Garca-Arenal et
al., 2001; Chen et al., 2002b). To determine whether recombi-
nation occurred specically in the P1 coding region between
isolates, we performed RDP program among a set of aligned
nt sequences. No recombinant sequence in the P1 coding re-
gion was discovered between SMV isolates. The result is in
agreement with the observation of lacking recombination in
the P1 gene of Watermelon mosaic virus (WMV) (Moreno
et al., 2004), but conicts with the study of the P1 gene of
Turnip mosaic virus (TuMV) in which the recombination is
common (Tomimura et al., 2003). However, two Korean iso-
lates (SMV-G5b and -G7H) have high aa similarities with two
Chinese isolates (SMV-HH5 and HZ) (Chen et al., 2004), but
fall into different types (Fig. 2), suggesting a possibility of a
recombinational event in the coding regions of SMVgenome
between isolates. Further analysis of full-length genomic se-
quences of multiple SMV isolates is needed to determine
whether variation is truly due to genetic recombination.
B.K. Choi et al. / Virus Research 112 (2005) 4251 47
Fig. 1. Comparison of P1 amino acid sequences of new resistance-breaking isolates to those of known strains of SMV. The rst and second boxes indicate
variable regions; third to sixth boxes indicate conserved motifs; * indicates lacking amino acid; - indicates identical amino acid residues in the alignment.
The phylogenetic analysis in the present study does not
support geographic clustering of the RB isolates in compar-
ison with known strains, indicating that the clustering was
independent of country origin. The known type isolates of the
seven SMV strain groups as described by Cho and Goodman
(1979) were seed-borne isolates from soybean germplasm
introduced into the USA from different countries of origin:
G1 from Japan, G2 (60) from Liberia, G3 from Pakistan, G5
fromThailand, and G6 and G7 fromKorea (Jain et al., 1992).
At that time isolates from US sources were G1 N/G2 (75-16-
1) or G3 (Cho and Goodman, 1979; Hunst and Tolin, 1982;
Jain et al., 1992). Interestingly, there were certain sequence
4
8
B
.
K
.
C
h
o
i
e
t
a
l
.
/
V
i
r
u
s
R
e
s
e
a
r
c
h
1
1
2
(
2
0
0
5
)
4
2

5
1
Table 3
Matrix of similarities of P1 gene nucleotide and amino acid sequences between new resistance-breaking (RB) isolates and known strains of Soybean mosaic virus
Similarities (%) to isolates and known strains
a
CN 1 CN 2 CN 3 CN 7 CN 9 CN 10 CN 12 CN 13 CN 15 CN 18 CN 31 CN 36 G1 G2 G3 G5a G5b G6 G7 G7d G7H Aa HH5 HZ N
CN1 94 89 89 96 93 99 95 92 95 89 91 95 90 92 95 96 93 87 93 97 93 96 96 95
CN2 95 89 89 95 94 94 95 91 97 89 90 94 89 91 94 95 92 87 92 95 91 95 94 95
CN3 91 91 98 91 88 90 90 89 89 98 90 91 87 90 90 90 88 86 90 91 90 92 91 92
CN7 92 91 99 91 88 89 90 89 89 99 89 91 87 89 90 90 88 86 90 90 89 91 91 92
CN9 96 96 92 93 94 96 98 93 96 91 91 96 91 92 99 97 94 88 93 97 93 97 96 96
CN10 94 96 90 91 95 93 93 91 94 88 90 93 88 90 93 94 91 86 91 94 91 94 93 93
CN12 99 95 92 92 96 94 96 93 95 90 92 96 91 92 96 97 93 88 93 97 93 97 96 96
CN13 95 95 92 92 99 94 96 93 95 90 91 95 90 92 98 96 94 88 93 97 93 96 96 95
CN15 94 94 92 92 95 93 94 95 92 89 95 94 89 94 92 93 91 88 93 93 97 94 94 94
CN18 96 98 92 92 97 96 96 96 95 89 91 95 90 92 95 96 93 88 93 96 92 96 95 96
CN31 92 91 99 99 93 91 92 92 92 92 90 91 87 90 90 90 88 86 90 91 90 92 91 92
CN36 93 93 92 92 94 92 93 93 96 94 92 92 88 92 90 92 90 87 92 93 95 92 92 93
G1 96 95 93 93 96 94 96 96 95 96 93 95 94 94 95 96 93 89 95 97 94 97 96 99
G2 95 94 92 92 95 93 95 95 94 95 92 94 98 89 90 91 89 93 90 92 89 92 91 94
G3 94 94 92 92 95 93 94 94 95 94 92 94 95 94 92 93 90 89 93 93 94 94 94 94
G5a 96 96 92 92 99 95 96 98 95 96 92 93 96 95 94 96 94 87 93 97 93 96 96 95
G5b 97 95 92 93 97 94 97 96 94 96 93 94 96 95 94 96 94 89 94 99 93 98 97 97
G6 95 95 91 92 97 94 95 96 94 96 92 93 95 94 94 96 95 86 91 94 91 94 93 94
G7 93 93 92 92 94 92 94 94 94 94 92 93 95 94 94 94 94 93 93 89 88 89 89 90
G7d 93 93 92 92 94 92 94 94 94 94 92 93 95 94 94 94 95 93 98 94 93 95 94 95
G7H 97 95 93 93 97 94 97 96 95 96 93 94 96 95 95 96 99 95 95 95 94 98 97 97
Aa 95 94 92 93 96 93 95 95 98 95 93 97 96 95 96 95 95 94 94 94 95 94 94 94
HH5 98 96 93 93 97 95 98 96 95 97 93 94 97 96 95 97 98 96 95 95 98 95 98 97
HZ 97 95 92 92 96 94 97 96 95 96 92 93 96 95 95 96 97 95 94 94 97 95 99 97
N 95 95 93 93 96 94 96 96 95 96 93 95 99 98 95 96 96 95 95 95 96 96 97 96
a
Numbers in bold type at upper right corner refer amino acid sequence similarities and numbers in plain type at lower left corner refer to nucleotide sequence similarities.
B.K. Choi et al. / Virus Research 112 (2005) 4251 49
Fig. 2. Phylogenetic analyses for amino acid (aa) sequences of P1 protein of 12 new resistance-breaking SMV isolates and those of 13 known strains and
isolates obtained from the NCBI database. (A) The phlogenetic tree (unrooted) was constructed by the DNAMAN version 5.2.9. The grouping occurred after
bootstrapping the data (only values >50% are shown) and the numbers on the branches indicate bootstrap percentages based on 1000 resamplings. (B) The
phlogenetic tree (unrooted) was constructed by the neighbor-joining method by CLUSTALW and the numbers at the end of branches indicate genetic distance
corrected by Kimuras method (1983). The P1 aa sequences of ZYMV was used as the outgroup.
differences between SMV-G5b (AY294044) from Korea and
-G5a (AF200564) from USA (or between SMV-G7 and -
G7H), even though they shared similar biological characteris-
tics. Perhaps SMVisolates of the same origin could diversify
due to selection pressure overtime imposed by host cultivars
as well as by aphid and seed transmission. The emergence
of a new variant from the same strain group, SMV-G7d, by
serial inoculation of a culture of SMV-G7 originating from
a cDNA clone in the resistant cultivar PI 96983 possess-
ing Rsv1 (Hajimorad et al., 2003), supports this hypothesis.
Similar evidence was reported from rapid emergence of RB
isolates in Rice yellow mottle Sobemovirus (Fargette et al.,
2002). The emergence of SMV RB isolates in Korean soy-
bean cultivars can most likely be attributed to release and
utilization of resistant germplasm and cultivars, to the use of
a narrow genetic base for SMV resistance, to selection pres-
sure overtime on SMV populations, and to mutation and/or
genetic recombination, all common events in plant virus evo-
lution (Roosinck, 1997; Garca-Arenal et al., 2001; Harrison,
2002).
50 B.K. Choi et al. / Virus Research 112 (2005) 4251
Acknowledgement
This workwas partlysupportedbya grant (R05-2003-000-
10293-0) fromthe Basic Research Programof the Korea Sci-
ence and Engineering Foundation (KOSEF), and by a grant
(R12-1999-002-03003-02003) from KOSEF, MOST and the
city of Daejeon through the Research Center for Biomedici-
nal Resources (RRC). We appreciate the ChungnamAgricul-
tural Research and Extension Services for collecting infected
soybean plants and National Crop Experimental Station for
supplying soybean seeds of differential cultivars.
References
Aaziz, R., Tepfer, M., 1999. Recombination between genomic RNAs of
two cucumoviruses under conditions of minimal selection pressure.
Virology 263, 282289.
Bousalem, D., Douzery, E.J.P., Fargette, D., 2000. High genetic diver-
sity, distant phylogenetic relationships and intraspecies recombination
events among natural populations of Yam mosaic virus: a contribu-
tion to understanding potyvirus evolution. J. Gen. Virol. 81, 243
255.
Buss, G.R., Ma, G., Chen, P., Tolin, S.A., 1997. Registration of V94-5152
soybean germplasm resistant to soybean mosaic potyvirus. Crop Sci.
37, 19871988.
Buss, G.R., Ma, G., Kristipati, S., Chen, P., Tolin, S.A., 1999. A new
allele at the Rsv3 locus for resistance to soybean mosaic virus. In:
H.E. Kauffman (Ed.), Proceedings of the World Soybean Research
Conference VI, Superior Printing, Chicago, IL, p. 490.
Chen, P., Buss, G.R., Roane, C.W., Tolin, S.A., 1991. Allelism among
genes for resistance to soybean mosaic virus in strain-differential soy-
bean cultivars. Crop Sci. 31, 305309.
Chen, P., Buss, G.R., Tolin, S.A., Gunduz, I., Cicek, M., 2002a. A valu-
able gene in Suweon 97 soybean for resistance to soybean mosaic
virus. Crop Sci. 42, 333337.
Chen, Y.-K., Goldbach, R., Prins, M., 2002b. Inter- and intramolecu-
lar recombinations in the Cucumber mosaic virus genome related to
adaptation to Alstroemeria. J. Virol. 76, 41194124.
Chen, J., Zheng, H.-Y., Lin, L., Adams, M.J., Antoniw, J.F., Zhao, M.-F.,
Shang, Y.-F., Chen, J.-P., 2004. A virus related to Soybean mosaic
virus from Pinellia ternate in China and its comparison with local
soybean SMV isolates. Arch. Virol. 149, 349363.
Cho, E.K., 1995. Identication of soybean mosaic virus strains and a
consideration on genetics of soybean for resistance to SMV strains.
Plant Dis. Agr. 1, 2225.
Cho, E.K., Chung, K.W., 1986. Strains of soybean mosaic virus caus-
ing soybean necrotic disease in Korea. Korean J. Breed. 18, 150
153.
Cho, E.K., Goodman, R.M., 1979. Strains of soybean mosaic virus: classi-
cation based on virulence in resistant soybean cultivars. Phytopathol-
ogy 69, 467470.
Cho, E.K., Goodman, R.M., 1982. Evaluation of resistance in soybeans
to soybean mosaic virus strains. Crop Sci. 22, 11331136.
Cho, E.K., Chung, B.J., Lee, S.H., 1977. Studies on identication and
classication of soybean virus disease in Korea. II. Etiology of a
necrotic disease of Glycine max. Plant Dis. Rep. 61 (4), 313317.
Cho, E.K., Choi, S.H., Cho, W.T., 1983. Newly recognized soybean mo-
saic virus mutants and sources of resistance in soybeans. Res. Rept.
ORD (S.P.M.U). 25, 1822.
Chung, B.J., Lee, S.H., Cho, E.K., Park, H.C., 1974. Identication of
soybean viruses and soybean varietal reactions. Annu. Rep. C.I.R.C.,
137145.
Domier, L.L., Latorre, I.J., Steinlage, T.A., McCoppin, N., Hartman,
G.L., 2003. Variability and transmission by Aphis glycines of North
American and Asian Soybean mosaic virus isolates. Arch. Virol. 148,
19251941.
Dougherty, W.G., Semler, B.L., 1993. Expression of virus-encoded pro-
teinases: functional and structural similarities with cellular enzymes.
Microbiol. Rev. 57, 781822.
Eggenberger, A.L., Stark, D.M., Beachy, R.N., 1989. The nucleotide se-
quence of a soybean mosaic virus coat protein-coding region and its
expression in Escherichia coli, Agrobacterium tumefaciens and to-
bacco callus. J. Gen. Virol. 70, 18531860.
Fargette, D., Pinel, A., Traor e, O., Ghesqui` ere, A., Konat e, G., 2002.
Emergence of resistance-breaking isolates of Rice yellow mottle virus
during serial inoculations. Eur. J. Plant Pathol. 108, 585591.
Garca-Arenal, F., Fraile, A., Malpica, J.M., 2001. Variability and ge-
netic structure of plant virus populations. Annu. Rev. Phytopathol.
39, 157186.
Glais, L., Tribodet, M., Kerlan, C., 2002. Genomic variability in Potato
potyvirus Y (PVY): evidence that PVY (N) W and PVY (NTN) vari-
ants are single to multiple recombinants between PVY (O) and PVY
(N) isolates. Arch. Virol. 147, 363378.
Gunduz, I., Buss, G.R., Chen, P., Tolin, S.A., 2002. Characterization of
SMV resistance genes in Tousan 140 and Hourei soybean. Crop Sci.
42, 9095.
Gunduz, I., Buss, G.R., Chen, P., Tolin, S.A., 2004. Genetic and pheno-
typic analysis of Soybean mosaic virus resistance in PI 88788 soy-
bean. Phytopathology 94, 687692.
Hajimorad, M.R., Eggenberger, A.L., Hill, J.H., 2003. Evolution of Soy-
bean mosaic virus-G7 molecularly cloned genome in Rsv1 genotype
soybean results in emergence of a mutant capable of evading Rsv1-
mediated recognition. Virology 314, 487509.
Harrison, B.D., 2002. Virus variation in relation to resistance-breaking in
plants. Euphytica 124, 181192.
Higgins, D., Thompson, J., Gibson, T., Thompson, J.D., Higgins, D.G.,
Ginson, T.J., 1994. CLUSTAL W: Improving the sensitivity of pro-
gressive multiple sequence alignment through sequence weighting,
position-specic gap penalties and weight matrix choice. Nucleic
Acids Res. 22, 46734680.
Hillis, D.M., Bull, J.J., 1993. An empirical test of bootstrapping as a
method of assessing condence in phlogenetic analysis. Syst. Biol.
42, 182192.
Hunst, P.L., Tolin, S.A., 1982. Isolation and comparison of two strains
of soybean mosaic virus. Phytopathology 72, 710713.
Jain, R.K., McKern, N.M., Tolin, S.A., Hill, J.H., Barnett, O.W., Tosic,
M., Ford, R.E., Beachy, R.N., Yu, M.H., Ward, C.W., Shukla, D.D.,
1992. Similarity of coat protein peptide proles of fourteen potyvirus
isolates from soybean conrms that they are strains of the one virus.
Phytopathology 82, 294299.
Jayaram, C., Hill, J.H., Miller, W.A., 1992. Complete nucleotide se-
quences of two soybean mosaic virus strains differentiated by re-
sponse of soybean containing the Rsv resistance gene. J. Gen. Virol.
73, 20672077.
Kiihl, R.A.S., Hartwig, E.E., 1979. Inheritance of reaction to soybean
mosaic virus in soybeans. Crop Sci. 19, 372375.
Kim, Y.H., 2000. Molecular classication of soybean mosaic virus strains
and their interaction with soybeans. Ph.D. Thesis, Kyungpook Natl.
Univ., Daegu, Korea.
Kim, Y.H., Kim, O.S., Lee, B.C., Moon, J.K., Lee, S.C., Lee, J.Y., 2003.
G7H, a new Soybean mosaic virus strain: its virulence and nucleotide
sequence of CI gene. Plant Dis. 87, 13721375.
Kimura, M., 1983. The natural theory of molecular evolution. Cambridge
University Press, Cambridge, England.
Lee, Y.H., Hwang, Y.H., Seong, Y.G., Kim, Y.H., Kim, W.H., 1992. Re-
sistance to disease, good in seed quality, high yielding and widely
adapted new soybean variety Taekwangkong. Res. Rep. RDA 34
(2), 1115.
Lee, K.C., Wong, S.M., 1998. Variability of P1 protein of zucchini yellow
mosaic virus for strain differentiation and phylogenetic analysis with
other potyviruses. DNA Seq. 9, 275293.
B.K. Choi et al. / Virus Research 112 (2005) 4251 51
Lim, W.-S., Kim, Y.-H., Kim, K.-H., 2003. Complete genome sequences
of the genomic RNA of soybean mosaic virus strains G7H and G5.
Plant Pathol. J. 19 (3), 171176.
Lin, S.S., Hou, R.F., Yeh, S.D., 2001. Complete genome sequence and
genetic organization of a Taiwan isolate of Zucchini yellow mosaic
virus. Bot. Bull. Acad. Sin. 42, 243250.
Ma, G., Chen, P., Buss, G.R., Tolin, S.A., 1995. Genetic characteristics
of two genes for resistance to soybean mosaic virus in PI 486355
soybean. Theor. Appl. Genet. 91, 907914.
Ma, G., Chen, P., Buss, G.R., Tolin, S.A., 2003. Genetic study of a lethal
necrosis to soybean mosaic virus in PI507389 soybean. J. Hered. 94,
205211.
Martin, D.P., Rybicki, E.P., 2000. RDP: detection of recombination
amongst aligned sequences. Bioinformatics 16, 562563.
Mavankal, G., Rhoads, R.E., 1991. In vitro cleavage at or near the N-
terminus of the helper component protein in the tobacco vein mottling
virus polyprotein. Virology 185, 721731.
Milbrath, G.M., Soong, M., 1976. A local lesion assay for soybean mo-
saic virus using Phaseolus vulgaris L. Top Crop. Phytopathol. Z. 87,
255259.
Moreno, I.M., Malpica, J.M., Daz-Pend on, J.A., Moriones, E., Fraile, A.,
Garca-Arenal, F., 2004. Variability and genetic structure of the popu-
lation of watermelon mosaic virus infecting melon in Spain. Virology
318, 451460.
Nagy, P.D., Christopher, O., Bujarski, J.J., 1999. Mapping sequences
active in homologous RNA recombination in brome mosaic
virus: prediction of recombination hot spots. Virology 254, 92
104.
Pu, Z., Cao, Q., Fang, D., Xi, B., Fang, C., 1982. Identication of strains
of soybean mosaic virus. Acta Phytophyl. Sin. 9, 1520.
Roosinck, M.J., 1997. Mechanisms of plant virus evolution. Annu. Rev.
Phytopathol. 35, 191209.
Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method
for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406425.
Takahashi, K., Tanaka, T., Iida, W., Tsuda, Y., 1980. Studies on virus
diseases and casual viruses of soybean in Japan. Bull. Tohoku Natl.
Agr. Exp. Sta. 62, 1130.
Tomimura, K., Gibbs, A.J., Jenner, C.E., Walsh, J.A., Ohshima, K., 2003.
The phylogeny of Turnip mosaic virus; comparisons of 38 genomic
sequences reveal a Eurasian origin and a recent emergence in Asia.
Mole. Ecol. 12, 20992111.
Vance, V.B., Moore, D., Turpen, T.H., Bracker, A., Hollowell, V.C., 1992.
The complete nucleotide sequence of pepper mottle virus genomic
RNA: comparison of the encoded polyprotein with those of other
sequence potyviruses. Virology 191, 1930.
Verchot, J., Carrington, J.C., 1995. Evidence that the potyvirus P1 pro-
teinase functions in trans as an accessory factor for genome ampli-
cation. J. Virol. 69, 36683674.
Verchot, J., Herndon, K.L., Carrington, J.C., 1992. Mutational analysis
of the tobacco etch potyviral 35-kDa-proteinase: identication of es-
sential residues and requirements for autoproteolysis. Virology 190,
298306.