Báo cáo khoa học: Interaction of Sesbania mosaic virus movement protein with the coat protein – implications for viral spread Soumya Roy Chowdhury and Handanahal Subbarao Savithri docx

Interaction of Sesbania mosaic virus movement protein
with the coat protein – implications for viral spread
Soumya Roy Chowdhury and Handanahal Subbarao Savithri
Department of Biochemistry, Indian Institute of Science, Bangalore, India
Introduction
Plants have an elaborate communication system that
permits transport of macromolecules from one cell to
another. Plant viruses have evolved mechanisms to
manipulate the same resident communication system
and redirect it in such a way that the viral genome is
transported from one cell to another, leading to spread
of infection. The virus-encoded movement protein
(MP), in association with other viral and host factors
called ancillary proteins, plays a central role in this
process. The MP–genome complex, or, in some cases,
assembled virus particles, interacts with the compo-
nents of plasmodesmata and dilates the openings to
permit passage through the cell wall [1,2]. Although
MPs are not conserved across genera, they perform
similar functions [3]. In terms of the nature of the
nucleoprotein complex that moves from cell to cell,
plant viruses may be broadly divided into two types
[4]. In the first type, MPs interact with viral RNA to
form a movement complex (M complex), which is
transported from cell to cell, as in tobacco mosaic
Keywords
coat protein; protein–protein interaction;
recombinant movement protein;
sobemovirus
Correspondence
H. S. Savithri, Department of Biochemistry,

Indian Institute of Science, Bangalore
560012, India
Fax: +91 8023600814
Tel: +91 8022932310
E-mail:
(Received 10 June 2010, revised 28
September 2010, accepted 1 November
2010)
doi:10.1111/j.1742-4658.2010.07943.x
Sesbania mosaic virus (SeMV) is a single-stranded positive-sense RNA
plant virus belonging to the genus Sobemovirus. The movement protein
(MP) encoded by SeMV ORF1 showed no significant sequence similarity
with MPs of other genera, but showed 32% identity with the MP of South-
ern bean mosaic virus within the Sobemovirus genus. With a view to
understanding the mechanism of cell-to-cell movement in sobemoviruses,
the SeMV MP gene was cloned, over-expressed in Escherichia coli and
purified. Interaction of the recombinant MP with the native virus (NV)
was investigated by ELISA and pull-down assays. It was observed that
SeMV MP interacted with NV in a concentration- and pH-dependent man-
ner. Analysis of N- and C-terminal deletion mutants of the MP showed
that SeMV MP interacts with the NV through the N-terminal 49 amino
acid segment. Yeast two-hybrid assays confirmed the in vitro observations,
and suggested that SeMV might belong to the class of viruses that require
MP and NV ⁄ coat protein for cell-to-cell movement.
Structured digital abstract
l
MINT-8050243: p53 (uniprotkb:P02340) physically interacts (MI:0915) with T-Ag (uni-
protkb:
P03070)bytwo hybrid (MI:0018)
l

MINT-8050226: MP (uniprotkb:Q9EB09) physically interacts (MI:0915) with CP (uni-
protkb:
Q9EB06)bytwo hybrid (MI:0018)
Abbreviations
CP, coat protein; CPMV, cowpea mosaic virus; GnHCl, guanidine hydrochloride; GST–MP, recombinant MP expressed in E. coli with an
N-terminal glutathione sulfur transferase tag; NV, native virus; M complex, movement complex formed by MP with viral genomic RNA;
MP, movement protein; rMP, recombinant MP expressed in E. coli with an N-terminal histidine tag; SBMV, Southern bean mosaic virus;
SeMV, Sesbania mosaic virus; TMV, tobacco mosaic virus; Y2H, yeast two-hybrid.
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 257
virus (TMV). TMV has been shown to be transported
as a replication complex that contains MP, viral repli-
case and genomic RNA [5]. In the second type, intact
virus particles are transported through MP-containing
tubules, as observed in cowpea mosaic virus (CPMV)
[6]. However, MPs that are known to form an M com-
plex can also form tubules [7], and MPs that form
tubules can also bind to RNA [8,9].
An extensive analysis of the function of MPs of
viruses from various genera has shown that taxonomi-
cally different viruses may use the same strategy, while
closely related viruses may use different strategies, and
some may use more than one strategy for the spread
of infection [3]. It is also possible that viruses may use
different strategies depending on the host they infect
[10]. The mechanism of virus movement is therefore
diverse and complex, involving several factors [11].
Sobemoviruses are plant RNA viruses that are
named after their type species, Southern bean mosaic
virus (SBMV). Viruses belonging to this genus are ico-
sahedral particles of approximately 30 nm in diameter.

The viral capsid is made up of 180 identical coat pro-
tein (CP) subunits organized with T = 3 icosahedral
symmetry, and Encapsidates a single molecule of geno-
mic RNA. The genomic RNA is a single-stranded mes-
senger-sense molecule, approximately 4–4.5 kb in size.
The 5¢ terminus of the RNA has a genome-linked pro-
tein (VPg), and the 3¢ end lacks a poly(A) tail. Sobem-
oviruses infect plants from 15 families, including
dicotyledonous and monocotyledonous species [12].
The first sobemovirus, SBMV, was reported from Lou-
isiana and California, USA, in 1943 [13]. Later it was
reported that sobemoviruses occur all over the world,
infecting plants in countries from Scandinavia to New
Zealand and throughout tropical Africa, North Amer-
ica and South East Asia. In susceptible hosts, sobem-
oviruses cause severe diseases with recurrent economic
losses. For example, rice yellow mottle virus is respon-
sible for the most rapidly spreading disease of rice in
Africa [12]. Previous studies on sobemoviruses have
shown that the protein coded by ORF1 of rice yellow
mottle virus [14] and Southern cowpea mosaic virus
[15] is essential for cell-to-cell movement. More
recently, it was reported that the ORF1-encoded pro-
tein of Cocksfoot mottle virus (CfMV), designated P1,
is essential for systemic spread of the virus [16]. Fur-
ther, the ORF1-encoded product of rice yellow mottle
virus has been implicated as a RNA silencing suppres-
sor [17,18]. These results suggest that ORF1-encoded
proteins function as MPs in sobemoviruses. However,
the molecular mechanism of cell-to-cell movement in

sobemoviruses has not been investigated. Other virus-
encoded ancillary proteins that may interact with
sobemoviral MPs and assist in cell-to-cell or systemic
movement of the virus have not yet been identified. It
is not known whether sobemoviruses use TMV-type or
CPMV-type movement strategies. Functional charac-
terization of sobemoviral MPs and understanding of
the role of ancillary proteins ⁄ domains in cell-to-cell
movement may assist in identification of genome seg-
ments that could be targeted for developing antiviral
strategies for this particular virus group.
Sesbania mosaic virus (SeMV) belongs to the Sob-
emovirus genus, and was first identified from infected
Sesbania grandiflora pers agathi on farms around
Tirupati, Andhra Pradesh, India. The 3D structure of
the purified virus has been determined, and it was
shown to be an icosahedral virus with a diameter of
30 nm comprising 180 identical CP subunits [19,20].
The SeMV genome is 4149 nucleotides long, and
encodes four potential overlapping ORFs [21]. Com-
parison of the nucleotide and the deduced amino acid
sequences of SeMV ORFs with those of other sobem-
oviruses revealed that SeMV is closest to the South-
ern bean mosaic virus Arkansas isolate (SBMV-Ark)
[21]. The mechanisms of SeMV assembly and poly-
protein processing have been reported previously
[22–25].
In the present study, the ORF1 gene encoding the
SeMV MP was cloned and over-expressed in Escheri-
chia coli in fusion with a hexahistidine or a glutathione

sulfur transferase (GST) tag. The recombinant proteins
were shown to interact with native virus (NV) using
pull-down assays and ELISA. Further, studies on dele-
tion mutants of MP were performed to determine the
domain responsible for the interaction of MP with CP
using ELISA and yeast two-hybrid (Y2H) assays.
Deletion of the N-terminal 49 amino acids of the
SeMV MP drastically reduced the interaction between
the two proteins. To our knowledge, this is the first
report demonstrating the interaction between MP and
CP of a sobemovirus, suggesting that SeMV might
belong to the class of viruses that require both MP
and CP for cell-to-cell movement.
Results
In silico analysis of MPs
In order to examine the similarity among the MPs of
sobemoviruses and other plant viruses, in silico analysis
was performed. The gene sequences were obtained from
the National Center for Biotechnology Information and
GenBank, and multiple sequence alignment was per-
formed using Clustal W2 ( />clustalw2/index.html). The results obtained are shown
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
258 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
in Table 1. The SeMV MP showed no significant
sequence similarity with MPs of viruses from other gen-
era. Within sobemoviruses, the sequence of the SeMV
MP was closest to that of the SBMV-Ark MP (32%
sequence identity), and the identity with MPs of other
sobemoviruses was not significant. The secondary struc-
ture of SeMV MP was predicted using the PredictPro-

tein server ( [26]. As
shown in Fig. 1, the SeMV MP was predicted to be a
primarily a-helical protein. The predicted percentages
of a helix, b sheet and coil were 49%, 25% and 26%,
respectively. The potential involvement of post-transla-
tional modification of viral MPs in regulation of their
transport mechanism was first suggested in view of
finding that the 30 kDa MP of TMV is phosphorylated
within host cells. Other viral MPs, such as those of
tomato mosaic virus and potato leafroll virus, were sub-
sequently also shown to be phosphorylated during the
infection process [27,28]. One consequence of the phos-
phorylation event on MP is that it could result in the
unloading of the viral genome from the M complex
after it enters the neighbouring cell through plasmodes-
mata [29–31]. However, other reasons for phosphoryla-
tion of MPs could also exist that have yet to be
identified. Nevertheless, the fact remains that MPs are
sometimes post-translationally modified by phosphory-
lation. Therefore, a search for the presence of potential
phosphorylation sites in the SeMV MP was performed
using the netphos 2.0 server ( />services/NetPhos/). RNA binding sites and other motifs
were searched using the Block search program (http://
blocks.fhcrc.org/blocks/blocks_search.html). The results
suggested the presence of a nucleic acid binding domain
in the C-terminal segment of the SeMV MP (Fig. 1,
grey box) and a high density of predicted phosphoryla-
tion sites at the C-terminus of the protein (Fig. 1,
yellow box). However, no conserved motif was found
when the SeMV MP sequence was compared with other

well characterized MPs.
Over-expression of the SeMV MP in E. coli and
purification under denaturing conditions
The SeMV MP gene was amplified and cloned into the
pRSET C vector (Invitrogen, Carlsbad, CA, USA) and
over-expressed in E. coli as described in Experimental
procedures. The recombinant histidine-tagged MP was
designated rMP. The 20 kDa rMP, although expressed
in at a high level (Fig. 2A), was mostly present in the
insoluble fraction (Fig. 2B). The rMP was therefore
purified from the insoluble fraction under denaturing
conditions using 6 m guanidine hydrochloride (GnHCl),
and refolded by stepwise dialysis as described in Experi-
mental procedures. The purity of the protein was deter-
mined by 12% SDS ⁄ PAGE (Fig. 2C). The refolded rMP
was soluble and was used for further characterization.
Table 1. Sequence comparison of SeMV MP with other MPs from
various genera.
Genus Virus species
Percentage
identity
with SeMV
Percentage
similarity
with SeMV
Sobemovirus Sesbania mosaic virus 100 100
Southern bean mosaic
virus
32.4 50
Southern cowpea

mosaic virus
17.8 27.5
Lucerne transient streak
virus
17.1 32.6
Rice yellow mottle virus 9.1 12.3
Cocksfoot mottle virus 8 15.1
Tobamovirus Tobacco mosaic virus 11 19.9
Alfamovirus Cowpea chlorotic mottle
virus
11.5 19.3
Alfalfa mosaic virus 8 14.7
Cucumovirus Cucumber mosaic virus 9.8 17.6
Comovirus Cowpea mosaic virus 4.7 6.9
Bromovirus Brome mosaic virus 2.7 6.1
Fig. 1. Prediction of the secondary structure of the SeMV MP. S,
sequence; P, secondary structure predicted using the PredictPro-
tein server ( The grey boxes repre-
sents the RNA binding motif. The red boxes indicate cysteine
residues and the yellow box indicates a high-propensity phos-
phorylation site predicted by the NetPhos 2.0 server (http://www.
cbs.dtu.dk/services/NetPhos/).
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 259
Circular dichroism (CD) spectroscopy
The far-UV CD spectrum of the purified and refolded
rMP showed minima at 210 and 222 nm, suggesting
that the protein was folded and adopted a largely
a-helical conformation (Fig. 3A). Analysis of the CD
spectrum using K2D2 software ( />projects/k2d2/) showed that rMP comprises more than

84% a-helical structure, compared to the predicted
helix content of 49%. The thermal stability of the rMP
was monitored by measuring the molar ellipticity at
210 nm as a function of temperature. The rMP had a
T
m
of 65 °C (Fig. 3A, inset).
Fluorescence spectroscopic analysis
The intrinsic fluorescence spectrum of rMP showed
maximum emission at 345 nm upon excitation at
280 nm, typical of a folded protein (Fig. 3B). The
emission maximum showed a red shift to 365 nm upon
addition of 8 m urea due to exposure of the aromatic
residues to the solvent as the protein unfolded [30].
Another broad peak between 305 and 315 nm was also
observed upon urea denaturation (Fig. 3B). Generally,
the fluorescence emission induced in proteins by
280 nm excitation is dominated by tryptophan fluores-
cence and the tyrosine emission is nearly undetectable.
The tyrosine emission (305–315 nm) is observed only
when the protein is in the denatured state [32].
MP–NV interaction
NV or the CP is an important ancillary protein for the
movement of many viruses within the host. To deter-
mine whether SeMV MP and NV interact with each
other, pull-down assays were performed as described
in Experimental procedures. A distinct band corre-
sponding to CP was seen together with rMP in the
eluted fraction (data not shown).
To confirm the interaction of MP with NV, a modi-

fied ELISA was performed as described in Experimen-
tal procedures. ELISA plates were coated with NV,
blocked, and rMP was added. The interaction between
the two proteins was monitored by using antibodies
against rMP (Fig. 4A). In a reverse experiment, rMP
was immobilized on ELISA platea and NV was used
as the probe protein (Fig. 4B). In both the experi-
ments, BSA was used as a control. Cross-reaction of
the primary antibody to the immobilized protein was
also tested by ELISA in the absence of the interacting
proteins. rMP interacted with NV in both the experi-
ments (Fig. 4A,B).
Nature of the MP–NV interaction
To determine the concentration dependence and nature
of the interaction between MP and NV, the same
ELISA-based approach was used. ELISA plates were
coated with NV (5 l g), and, after blocking, increasing
AB C
Fig. 2. Expression, solubility analysis and purification of rMP. (A) The pRSET C-MP clone was transformed into E. coli BL21 (DE3) cells. The
total lysate after isopropyl-b-
D-thiogalactopyranoside induction was analysed by 12% SDS ⁄ PAGE. Lanes U and I correspond to uninduced
and induced total lysate, respectively. The arrow indicates the position of the rMP band in the induced sample (lane I). (B) SDS ⁄ PAGE (12%)
of soluble (S) and pellet (P) fractions of rMP-expressing cells. The arrow indicates the position of the rMP band. (C) SDS ⁄ PAGE (12%) show-
ing rMP purified by Ni-NTA affinity chromatography under denaturing conditions using 6
M GnHCl (lanes 1 and 2). The arrow indicates the
position of purified protein. Lane M, protein molecular mass markers. The gels were stained with Coomassie brilliant blue R250.
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
260 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
concentrations of rMP were added, and the ELISA
was performed. The absorbance at 450 nm was plotted

as a function of rMP concentration (Fig. 5A). The
results show that the interaction between NV and rMP
is concentration-dependent.
Next, we wished to investigate the effect of pH on
MP–NV interaction. Previously, it was shown that
SeMV particles are stable over a pH range of 3–10.4
[33]. Similar studies with rMP showed that the protein
was stable over a broad pH range and precipitated at
pH values > 9 [34]. Therefore, to determine the opti-
mal pH for interaction between the two proteins, rMP
was dissolved in buffers at various pH values (Fig. 5B),
and incubated with NV-bound plates for 1 h. Then the
unbound rMP was removed and an ELISA was per-
formed. The reactions were performed in triplicate.
After the reaction, absorbance values obtained at
450 nm were plotted as a function of the pH of the
A
B
Fig. 3. Biophysical characterization of refolded rMP. (A) Far-UV CD
spectrum of rMP. The molar ellipticity of rMP (0.5 mgÆmL
)1
) was
recorded from 190 to 300 nm in a 0.2 cm path-length cuvette with
a band width of 1 nm and response time of 1 s. The thermal stabil-
ity of rMP was monitored by measuring the molar ellipticity at
210 nm at various temperatures (inset). (B) Intrinsic fluorescence
spectra of rMP. The intrinsic fluorescence spectrum of rMP was
measured by exciting the sample at 280 nm and observing the fluo-
rescence emission from 300 to 400 nm in the absence (solid line)
and presence of 8

M urea (dotted line). The emission maximum
showed a red shift to 365 nm upon addition of 8
M urea, and a
broad peak between 305 and 315 nm was also observed due to
emission by tyrosine residues that are exposed in the protein in the
denatured state.
A
B
Fig. 4. Interaction of rMP with NV. (A) ELISA of rMP and NV inter-
action. ELISA plates coated with NV (5 lg) (P1) were blocked with
10% skimmed milk in 1% NaCl ⁄ Pi (block) followed by addition of
5 lg of rMP (P2). The ELISA was performed using an anti-MP poly-
clonal antibody (pAb to P2) and developed using anti-rabbit IgG con-
jugated to horseradish peroxidise and DMB H
2
O
2
(sAb + Sub). The
steps involved and the controls used are indicated on the figure.
BSA was used as a negative control. (B) Reverse experiment in
which ELISA plates were coated with rMP (P1) and probed with
NV (P2). Primary polyclonal antibodies against NV (pAb to P2) were
used in this reaction, with controls similar to those used in (A).
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 261
buffer in which rMP was dissolved. The optimal pH
for interaction between the two proteins was between
pH 6.5 and 7.5, with a sharp decrease in both the
acidic and alkaline ranges. These observations suggest
that the interaction between the two proteins is opti-

mal near physiological pH.
To monitor the effect of NaCl on the interaction of
the two proteins, ELISA was performed as before,
except that the rMP was dissolved with various concen-
trations of NaCl and added to NV-bound ELISA plates.
Incubation of the SeMV MP or the NV with 1 m NaCl
did not affect solubility or stability. The absorbance at
450 nm was plotted as a function of NaCl concentration
(Fig. 5C). An increase in NaCl concentration did not
affect the interaction of NV with rMP, suggesting that
the interaction between the two proteins is strong.
Expression of GST–MP
The experiments described so far were performed using
refolded rMP. The MP-encoding gene was also cloned
and over-expressed with an N-terminal GST tag as
described in Experimental procedures. The GST–MP
fusion protein expressed in E. coli BL21 (DE3) was sol-
uble and of the expected size (45.4 kDa). The protein
was purified using glutathione affinity chromatography,
and was found to be homogeneous (Fig. 6A, lanes E).
Protein–protein interaction between NV and
GST–MP
To determine whether the soluble GST–MP interacts
with NV in a manner similar to that of refolded rMP,
A
B
C
Fig. 5. Analysis of the biochemical nature of the interaction between
rMP and NV. (A) Effect of rMP concentration on the NV–MP interac-
tion. NV-coated ELISA plates were incubated with increasing concen-

trations of rMP after a blocking step. The absorbance values
obtained at 450 nm by ELISA with anti-MP polyclonal antibody were
plotted as a function of rMP concentration. (B) Effect of pH on the
NV–MP interaction. ELISA plates coated with NV (P1) were incu-
bated with rMP (P2) in 50 m
M buffers at various pH as indicated in
the figure. After incubation, the wells were washed, ELISA was
performed using anti-MP polyclonal antibody (pAb to P2), and
absorbance values at 450 nm were plotted as a function of pH. rMP
in NaCl ⁄ Pi (pH 7.4) was used as a positive control. The other controls
used are indicated in the figure. (C) Effect of NaCl on the NV–MP
interaction. ELISA plates coated with NV (P1) were blocked with
10% skimmed milk in PBS and incubated with rMP (P2) dissolved in
50 m
M Tris ⁄ HCl buffer (pH 7.4) containing various concentrations of
NaCl as shown in the figure and incubated for 1 h. After extensive
washing of the wells, ELISA was performed as described in
Experimental procedures using anti-MP polyclonal antibody (pAb to
P2). The absorbance values obtained at 450 nm were plotted as a
function of NaCl concentration.
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
262 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
ELISA and GST pull-down experiments were per-
formed as described previously. GST–MP was found
to interact with NV both in ELISA (Fig. 6B) and pull-
down assays (data not shown). To rule out the possi-
bility that the interaction between GST–MP and NV is
due to interaction between GST and NV, a GST
blocking step was introduced in the ELISA-based
interaction assay (Fig. 6B, last two columns). There

was no significant difference in the binding of GST–
MP to NV in the presence or the absence of GST, sug-
gesting that the interaction is indeed between MP and
NV and not between GST and NV.
Generation of deletion mutants of GST–MP
The results presented so far clearly demonstrated that
MP interacts with NV. It was of interest to map the
region of MP that is responsible for this interaction.
For this purpose, a number of deletion mutant clones
were constructed. As described earlier, SeMV MP is
primarily an a-helical protein. The three predicted
N-terminal three helices (Fig. 1) were systematically
deleted to generate ND16, ND35 and ND49 deletion
mutant clones. In addition, three C-terminal deletion
mutant clones were also constructed: CD3, in which
a high-propensity phosphorylation site was removed,
CD19 in which the predicted RNA binding domain
was removed, and CD38, in which three cysteines
and the nucleic acid binding domain were deleted
(Fig. 1).
The GST–MP deletion mutants were over-expressed
in E. coli BL21(DE3)pLysS as described for the full
length GST–MP. All the mutant proteins were soluble
and of the expected size. The proteins were purified
using glutathione affinity chromatography (Fig. 7A).
Mapping of the SeMV MP domain necessary for
interaction with NV
To determine which domains are involved in the inter-
action of MP with NV, ELISA was performed as
described previously. ELISA plates were coated with

NV (5 lg) and blocked with 10% milk in NaCl ⁄ P
i
,
followed by addition of various mutants as probe pro-
teins. The ELISA was performed using anti-rMP as
the primary antibody. In parallel, subsequent wells
were incubated with GST and probed using polyclonal
antibodies against GST to rule out the possibility of
GST–NV interaction. Determination of the absor-
bance at 450 nm clearly showed that the N-terminal
deletions have a pronounced effect on MP–NV inter-
action. Successive deletion of one, two and three pre-
dicted helices from the N-terminus of MP reduced the
interaction with NV by 51.5%, 66.4% and 80.1%,
respectively, compared with the interaction between
GST–MP and NV. However, the interaction was not
affected when C-terminal amino acids were deleted. It
may therefore be concluded that MP interacts with
NV via the N-terminal domains. Similar observations
were also made in pull-down experiments (data not
shown).
A
B
Fig. 6. Purification of GST–MP, and determination of the interac-
tion between GST–MP and NV by ELISA. (A) Coomassie brilliant
blue-stained 12% SDS ⁄ PAGE gel showing GST–MP and GST puri-
fied by glutathione affinity chromatography. Lanes are marked as
unbound protein (U), washed samples (W), eluted GST–MP (E) and
purified GST (G). Lane M, protein molecular mass markers. Arrows
indicate the position of purified proteins. (B) Protein–protein interac-

tion between NV and GST–MP. An ELISA to measure the protein
interaction between NV (P1) and GST–MP (P2) was performed as
described in the legend to Fig. 4B using anti-MP polyclonal antibody
(pAb to P2). Experimental steps and controls are indicated in the
figure. An additional GST blocking step (GST) was included to rule
out the possibility of GST–NV interaction.
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 263
Y2H assays
All the results obtained so far were from in vitro exper-
iments. To confirm these results, a Y2H assay was
performed using the Matchmaker system. pGAD T7
(CP) and pGBK T7 (MP and deletion mutants) clones
were obtained as described in Experimental procedures.
The pGAD T7 and pGBK T7 clones were transformed
into the Saccharomyces cerevisiae AH109 strain in pairs
as indicated in Fig. 7. pGBKT7-P53 (murine p53 fused
to GAL4 DNA BD) and pGADT7-T Ag (the SV40
large T-antigen fused to GAL4 DNA AD) that have
been previously reported to interact in a Y2H assay [35]
were used as a positive control in these experiments.
After transformation, growth was monitored on syn-
thetic drop-out (SD) –Leu ⁄ –Trp plates to confirm that
both the plasmids were transformed into AH109 cells.
Subsequently, the transformed colonies were replica-
plated on –Leu ⁄ –Trp ⁄ –His (medium stringency), –Leu ⁄
–Trp ⁄ –His ⁄ –Ade (high stringency), –Leu ⁄ –Trp ⁄ –His ⁄ +
a-X-Gal (medium stringency with a-galactosidase) and
–Leu ⁄ –Trp ⁄ –His ⁄ –Ade ⁄ +a-X-Gal (high stringency
with a-galactosidase) SD plates to determine the quality

and strength of interaction between the SeMV MP or
the deletion mutants and CP (Fig. 8A,B).
AH109 cells co-transformed with pGBK T7 MP and
pGAD T7 CP grew on all nutritional selection medium
up to the final level of selection (–Leu ⁄ –Trp ⁄ –His ⁄
–Ade ⁄ +a-X-Gal), similar to the positive control com-
prising p53 and T-Ag (first two rows from the top in
Fig. 8A,B), suggesting that MP and CP also interact
with each other under the ex vivo conditions of Y2H
system. However, the AH109 strain transformed with
either the pGAD T7 MP clone or the pGBK T7 CP
clone alone did not form colonies, ruling out the possi-
bility of de novo activation of the reporter gene in the
presence of the expressed proteins. Similarly, untrans-
formed AH109 S. cerevisiae alone did not form any
colonies (data not shown).
To identify the domain in MP that is involved in
interaction with CP, MP mutant gene products
obtained by PCR were cloned into the pGBK T7 vec-
tor, and the mutants were tested for their interaction
with CP expressed from the pGAD T7 vector. Fig. 8
shows that all the mutants exhibited positive Y2H
interaction with CP. The interaction between ND16
and CP (third row from top, Fig. 8) was observed for
growth under medium stringency conditions (–Leu ⁄ –
Trp ⁄ –His ⁄ +a-X-Gal), but no interaction was observed
under high stringency conditions (–Leu ⁄ –Trp ⁄ –His ⁄
–
Ade). For ND35 (fourth row from top in Fig. 8A,B),
the level of interaction was comparable to that

between MP ND16 and CP. pGAD T7 CP and
pGBK T7 ND49 clones co-transformed into AH109
cells only formed white colonies in –Leu ⁄ –Trp ⁄ –
His ⁄ +a-X-Gal plates (fifth row, column 4, in
Fig. 8A,B). These results show that the interaction of
A
B
Fig. 7. Mapping of the SeMV MP domain responsible for interaction
of MP with NV. (A) Purification of GST–MP deletion mutants. GST–MP
deletion mutants were over-expressed in E. coli BL21 (DE3) and
purified using glutathione affinity chromatography. The Coomassie
brilliant blue-stained gel after 12% SDS ⁄ PAGE of purified GST–MP
deletion mutant proteins is shown. (B) ELISA-based protein–protein
interaction assay between NV and GST–MP deletion mutants. ELISA
plates coated with NV (P1) and blocked with 10% skimmed milk in
PBS and were incubated with GST–MP deletion mutants (P2). The
remaining stages of the reaction were performed as described in the
experimental procedure section using anti-MP polyclonal antibody
(pAb to P2) (hatched bars). Subsequent NV-coated wells were incu-
bated with GST and probed using anti-GST polyclonal antibody to rule
out the possibility of GST–NV interaction (white bars). Experimental
steps and controls are indicated in the figure. The absorbance at
450 nm for each of the conditions is shown. The percentage
decrease in absorbance for the mutants as compared to GST–MP
and NV interaction is indicated above the bars.
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
264 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
MP with CP is greatly reduced by deletion of 49
amino acids from the N-terminus. C-terminal deletions
of MP (CD3, CD19 and CD38) (last three rows in

Fig. 8) had a minimal effect on the interaction between
MP and CP. However, none of the deletion mutants
grew on –Leu ⁄ –Trp ⁄ –His ⁄ –Ade plates, indicating that
there was some loss of interaction (rows 3–8, column
5, in Fig. 8A,B) .
b-galactosidase assay with ortho-nitrophenyl-b-
galactopyranoside as substrate
Transformed colonies that showed a positive Y2H
interaction were grown in liquid culture (–Leu ⁄ –Trp ⁄ –
His medium), and were assayed for b-galactosidase
activity to validate and quantify the results of the two-
hybrid interactions. A single colony was picked from
each of the SD plates, and the b-galactosidase assay
was performed as described in Experimental proce-
dures. The results are presented as a percentage of
arbitrary units of b-galactosidase activity (values indi-
cated on the top of each bar) relative to that obtained
with transformants expressing p53 and T-antigen
(100%). Values are the means of at least three separate
experiments (Fig. 9A). There was no difference in the
b-galactosidase activity between transformants express-
ing CP with MP, ND16, CD3, CD19 or CD38. How-
ever, an appreciable decrease in activity was seen in
transformants expressing CP with ND35 or ND49.
Interaction between the SeMV MP and CP resulted in
71% activity compared with the interaction between
p53 and the T-antigen (100%). The interaction
between the DN49 mutant MP and CP resulted in
27% activity. This corresponds to a reduction in activ-
ity of 60% compared with interaction of the wild-type

SeMV MP and CP. This reduction in interaction could
also be due to a difference in the level of expression of
the interacting proteins. To rule out this possibility,
the levels of all the bait and prey proteins were quanti-
fied by ELISA. The CP expressed from pGADT7 vec-
tor has a haemagglutinin tag fused to its N-terminus,
and the MP and the mutants expressed from
pGBK T7 have cMyc epitope tags. The co-transformed
colonies were grown overnight in 5 mL of appropriate
selection medium, and cells were lysed to release the
proteins. ELISA was performed with the total lysate
using haemagglutinin polyclonal antibody or cMyc
monoclonal antibody at 1 : 10 000 dilution (Fig. 9B).
All the proteins were expressed at comparable levels.
Thus the drastic reduction in interaction of ND35 and
AB
Fig. 8. Y2H interaction between MP and CP. (A) pGBK T7 (MP, MP deletion mutants and p53) and pGAD T7 (CP and T Ag) clones were
transformed in pairs into the AH109 strain, plated onto –Leu ⁄ –Trp SD transformant selection plates, and incubated for 96 h. Colonies that
grew were marked, and replica-plated onto various nutritional marker SD plates with various stringencies for reporter gene expression. To
assess a-galactosidase activity, colonies were plated onto SD plates containing a-X-Gal. (B) Schematic representation of the results in (A).
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 265
ND49 with CP is because of deletion of the interacting
domain.
Discussion
MPs are a diverse group of non-structural proteins of
plant viruses that are involved in the spread of infection
from cell to cell and systemically within the host
plant [11]. The present study comprised biochemical
characterization of the MP encoded by ORF1 of

SeMV, a member of the genus Sobemovirus. Analysis
of the deduced amino acid sequence of the SeMV MP
showed that it is predominantly an a-helical protein. It
has a C-terminal nucleic acid binding domain and a
predicted phosphorylation site as expected of a protein
involved in viral movement (Fig. 1).
The interaction between MPs and virus- or host-
encoded ancillary proteins is important for transport
of the viral genome from one cell to another. The
results presented here clearly show that the purified
SeMV MP interacts with NV and CP, suggesting that
SeMV might belong to the class of viruses that require
MP and NV ⁄ CP for cell-to-cell movement.
An inherent characteristic of MPs is their ability to
interact with plasmodesmata and components of the
cellular vasculature. Hence, they tend to form inclusion
bodies when expressed in vitro [36–38]. The rMP over-
expressed in E. coli was also present in the insoluble
fraction and was purified under denaturing conditions
(Fig. 2), but could be successfully refolded. Upon
denaturation with 8 m urea, in addition to the shift of
the fluorescence emission maxima from 345 to 365 nm
due to exposure of tryptophan residues, an additional
broad peak at 305–315 nm was observed. There is a
single tryptophan (position 84) and eight tyrosines in
the SeMV MP. It is possible that the fluorescence
emission of these tyrosines is quenched by energy
transfer to tryptophan or charged amino groups or
protonated carboxylates in their vicinity in the refolded
protein. However, upon denaturation, fluorescence due

to tyrosine is observed as a broad peak at 305–315 nm
[32,39].
rMP eluted in the void volume when analysed by
size-exclusion chromatography, suggesting that the
refolded MP formed large oligomers (data not shown).
Most MPs form soluble aggregates, probably because
of their inherent ability to form M complexes or
tubules for transport across plasmodesmata. In vitro,
some MPs are known to form heteromeric complexes
with various cytoskeletal elements and host factors
such as the major nucleolar protein fibrillarin [40].
Together with the MP, the CP plays a pivotal role
in cell-to-cell movement of certain viruses [3]. In order
to examine interactions between the CP and the MP,
in vitro and ex vivo studies were performed. The inter-
action of rMP with NV was dependent on the concen-
tration of rMP (Fig. 5A). However, the stoichiometry
of interaction could not be estimated as the refolded
rMP formed soluble aggregates. Also, as the NV was
immobilized on ELISA plates, not all the sites would
A
B
Fig. 9. Quantification of the MP–CP Y2H interaction by b-galactosi-
dase assay and estimation of the level of protein expression. (A) A
b-galactosidase assay of transformed colonies that showed positive
Y2H interaction was performed as described in Experimental proce-
dures. The results are presented as a percentage of arbitrary units
of b-galactosidase activity (values are indicated on the top of each
bar) relative to the interaction between p53 and T-antigen (100%).
Values are means of at least three separate experiments. (B) Direct

antigen-coating ELISA for estimation of the interaction between CP
and MP or its deletion mutants. Total protein isolated from AH109
cells transformed with pGBK T7 MP or MP deletion clones and
pGAD T7 CP was coated on to ELISA plates. The amount of MP
and the deletion mutants was quantified using cMyc monoclonal
antibody (open bars). The amount of CP was estimated using hae-
magglutinin polyclonal antibody (closed bars).
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
266 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
be available for interaction with MP. However, the
ELISA results unambiguously demonstrated that rMP
and NV interact with each other.
An insight into the nature of interaction between the
rMP and NV was obtained by monitoring the effect of
pH and NaCl concentration (Fig. 5B,C). The forma-
tion of salt bridges and hydrogen bonds crucial for
protein–protein interactions is dependent on pH. Vari-
ation in pH may also lead to conformational changes
that may hinder interactions. The interaction between
rMP and NV was optimal near physiological pH, sug-
gesting that these interactions might be relevant for
in vivo functions of MP. It is likely that the interaction
between the two proteins is largely ionic, as rMP and
CP have opposite computed net charges ()4.8 and
+13.3, respectively). High ionic strength can reduce
such interactions due to shielding of ionizable groups.
However, the interaction between the proteins was
unaffected when rMP was allowed to bind to immobi-
lized NV in the presence of 1 m NaCl, suggesting that
the interaction between the two proteins is quite strong

and is probably not dominated by ionic interactions.
The SeMV MP was soluble when expressed with a
GST tag. The purified GST–MP also interacted with
NV in a manner similar to rMP (Fig. 6). In order to
map the domains of MP that interacted with NV, vari-
ous deletion mutants of GST–MP (ND16, ND35, ND49,
CD3, CD19 and CD38) were generated. The purified
GST–MP and the deletion mutant MPs showed similar
CD spectra (data not shown), suggesting that there were
no gross conformational changes due to the mutations.
Analysis of the ELISA data obtained with the dele-
tion mutants showed that the C-terminal deletions did
not affect interaction between the two proteins.
However, N-terminal deletions resulted in decreased
interactions, and the ND49 mutant showed 80% loss
of interaction with NV. These observations suggest
that there is a major interacting domain within the
N-terminal 49 amino acids of the SeMV MP. This is
in contrast to observations on CPMV MP, in which
the C-terminal domain was shown to be involved in
NV recognition and binding [41,42].
To substantiate the in vitro results, Y2H assays were
performed. Similar studies have been performed with
Bhendi yellow vein mosaic virus [43]. Interaction
between SeMV MP and CP was also observed in the
Y2H assays. However, the interaction was not as
strong as that observed between p53 and T-antigen.
The MP–CP interaction resulted in 71.3% b-galactosi-
dase activity compared to that observed for the p53–T-
antigen interaction (Figs 8A and 9A). This could have

important implications for the plant virus lifecycle.
The interaction between MP and CP must be transient.
The M complex must disassemble once the genome is
translocated to an adjacent cell so that further steps in
the lifecycle, i.e. replication of the viral genome, may
proceed. Dissociation of viral genome from the com-
plex has been attributed to phosphorylation of MP
[44].
Similar to the in vitro results, deletion of amino
acids from the N-terminus resulted in a decreased
interaction between MP and CP in Y2H assays. Dele-
tion of the N-terminal 16 amino acids or the C-termi-
nal 3, 19 and 38 amino acids had a marginal effect.
However, deletion of 35 amino acids from the N-ter-
minus (first two helices) reduced the interaction by
50%, and deletion of 49 amino acids (first three heli-
ces) reduced the interaction by 60% as monitored by
measuring the b-galactosidase activity (Fig. 9A), sug-
gesting that the location of the interacting domain is
between residues 17 and 49. Further, it should be
noted that there was no a-galactosidase activity on SD
plates (Fig. 8A, total absence of blue colonies in –
Leu ⁄ –Trp ⁄ –His ⁄ +a-X-Gal plates) for the interaction
between CP and the deletion mutant DN49. However,
the colonies grew in –Leu ⁄ –Trp ⁄ –His plates, demon-
strating that a degree of interaction between the pro-
teins remains. It is possible that another site of
interaction beyond the N-terminal 49 residues may
contribute to these weak interactions.
In conclusion, the results presented here clearly dem-

onstrate that the recombinant SeMV MP interacts with
NV ⁄ CP via the N-terminal domain. To our knowledge,
this is the first report identifying CP as an interacting
partner for any sobemoviral MP. SeMV may require
CP and MP for cell-to-cell movement within the host.
However, at this stage, it is not possible to classify
SeMV as a type II virus (where the CP acts as an
ancillary protein to the MP) or a type III virus (where
the complete virion moves across the plasmodesmata)
with respect to movement. Further studies on the nat-
ure and composition of the nucleoprotein complex
translocated from cell to cell via the plasmodesmata
are required to clarify the mechanism of viral move-
ment in sobemoviruses.
Experimental procedures
Construction of rMP and GST–MP clones
The ORF1 gene of SeMV was amplified using high-fidelity
fusion polymerase (New England Biolabs, Ipswich, MA,
USA) using sense and antisense primers (MP sense and MP
anti, Table 2) corresponding to the N- and C-termini of the
MP, respectively, and a full-length SeMV cDNA clone
(AY004291) as the template. The PCR product was cloned
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 267
into the PvuII site of the pRSET C vector (Invitrogen,
Carlsbad, CA, USA). The additional nucleotides (87 bp)
derived from the vector backbone as a result of the cloning
strategy employed were removed by NheI digestion fol-
lowed by re-ligation. The recombinant clone was designated
pRSET C-MP. Expression of SeMV MP from the

pRSET C-MP clone resulted in a protein with additional
16 amino acids at the N-terminus, including the hexahisti-
dine tag, due to the cloning strategy used. In order to
express MP as an N-terminally GST-tagged protein, the
PCR product was cloned into the EcoRI site of the
pGEX 4T1 vector (GE Healthcare, Uppsala, Sweden). The
identity of both clones was confirmed by PCR using T7
sense and MP antisense primers (Table 1) and DNA
sequencing.
Construction of GST–MP deletion mutant clones
MP deletion mutant genes were amplified separately using
high-fidelity fusion polymerase, with sense and antisense
primers (marked with the prefix E, Table 2) corresponding
to the N- and C-termini for each deletion mutant protein
and the pGEX 4T1 MP clone as the template. The PCR
products were cloned into the EcoRI site of the pGEX 4T1
vector. The identity of all clones was confirmed by PCR
using T7 sense and mutant specific antisense primers
(Table 2) and DNA sequencing.
Expression and purification of the recombinant
proteins under normal and denaturing conditions
The pRSET C-MP clone and the pGEX 4T1-MP clone were
transformed separately into E. coli BL21(DE3) pLysS cells
(Novagen, Darmstadt, Germany). A single colony was inoc-
ulated into 20 mL of Luria–Bertani medium containing
50 lgÆmL
)1
ampicillin, and allowed to grow overnight at
37 °C. The overnight culture was inoculated into 500 mL of
Terrific broth containing 50 lgÆmL

)1
ampicillin and allowed
to grow at 37 ° C till the attenuance at 600 nm reached 0.6.
The expression of MP was induced using 0.3 mm isopropyl-
b-d-thiogalactopyranoside (Sigma-Aldrich, St Louis, MO,
USA), and grown for 10 h at 15 °C. The cells were
harvested by centrifugation at 10 000 g for 10 min and
resuspended in 30 mL of lysis buffer (100 mm Tris ⁄ HCl,
200 mm NaCl, 10 mm b-mercaptoethanol, pH 8). Resuspended
cells were sonicated for 15 min at an amplitude o f 30 in a
Vibra-Cell sonicator ( So nics & M ater ials Inc., N ewtown, CT,
USA), and the lysate was centrifuged at 10 000 g for 10 min
at 4 °C. The solubility of the expressed protein was checked
using SDS ⁄ PAGE [45]. The protein bands were visualized
by staining with Coomassie brilliant blue (Sigma-Aldrich).
The rMP was purified from the insoluble fraction by
Ni-NTA chromatography in the presence of 6 m GnHCl.
The cell pellet was resuspended in 50 mL of lysis buffer
containing 6 m GnHCl and sonicated. Ni-NTA agarose
beads pre-equilibrated with lysis buffer containing 6 m
GnHCl were added to the supernatant obtained after cen-
trifugation at 10 000 g for 10 min, and the protein was
allowed to bind to the beads for 2 h at 4 °C. The beads
were then packed in a column and washed with 50 mL and
20 mL each of washing buffers A (lysis buffer containing
10 mm imidazole and 6 m GnHCl) and B (lysis buffer con-
taining 50 mm imidazole and 6M GnHCl), respectively, for
30 min each. The protein was then eluted in elution buffer
(lysis buffer containing 300 mm imidazole and 6 m GnHCl).
The denatured purified protein was refolded by stepwise

dialysis against lysis buffer (not containing imidazole) with
decreasing concentrations of GnHCl (6, 4, 2, 1 and 0 m), and
stored at 4 °C. The refolded rMP was us ed to raise antibod-
ies in rabbit as described previously [22].
Table 2. Oligonucleotide primers used in the study.
Name Sequence (5¢fi3¢) Description
MP sense
MP anti
CCG
GCTAGCG
GAATTC
ATGATGGTAATGCAAGCTCAGCATACT
CCGG
GAATTC
GGAGGAGGACATAGCCCT
Primers for amplification of the MP gene. The EcoRI site is
indicated in bold and the NheI site is underlined.
E.CP sense
E.CP anti
CCG
CATATGG
GAATTC
ATGATGGCGAAAAGGCTTTCG
CCG
CATATGG
GAATTC
GTTGTTCAGGGCTGAGGC
Primers for amplification of the CP gene. The EcoRI site is
indicated in bold and the NdeI site is underlined
E.MP.N35 sense CCG

CATATGG
GAATTC
ATGATGGTATGTGAAGTGGAATTTGAT Primer for amplification of the MP ND35 gene. The EcoRI
site is indicated in bold and the NdeI site is underlined
E.MP.N16 sense CCG
CATATGG
GAATTC
ATGATGGTATTCATTGGTTTTGAGGAC Primer for amplification of the MP ND16 gene. The EcoRI
site is indicated in bold and the NdeI site is underlined
E.MP.N49 sense CCG
CATATGG
GAATTC
ATGATGGTAGTGAGAGCCCACAACCAA Primer for amplification of the MP ND49 gene. The EcoRI
site is indicated in bold and the NdeI site is underlined
E.MP.C3 anti CCG
CATATGG
GAATTC
CATAGCCCTTGCAGCTCG Primer for amplification of the MP CD3 gene. The EcoRI
site is indicated in bold and the NdeI site is underlined
E.MP.C 19 anti CCGA
AAGCTT
GAATTCCGGACACGAATAGAAGTATTC Primer for amplification of the MP CD19 gene. The HindIII
site is indicated in bold and the EcoRI site is underlined
E.MP C38 anti CCGA
AAGCTT
GAATTCCGGCCCGTTTTCACAAGGAGC Primer for amplification of the MP CD38 gene. The HindIII
site is indicated in bold and the EcoRI site is underlined
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
268 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
For purification of soluble GST–MP, GST binding resin

was added to the soluble fraction of the cell lysate obtained
as described above and incubated for 2 h at 4 °C. The resin
was packed in a column and washed thoroughly with wash
buffer (20 m m Tris ⁄ HCl pH 7.5 containing 200 mm NaCl,
10 mm imidazole, 0.1% Nonidet-P40 ((Sigma-Aldrich,
St Louis, MO, USA) and 10 mm b-mercaptoethanol). The
bound protein was eluted using 20 mm reduced glutathione
in wash buffer. The purified proteins were extensively dialy-
sed against storage buffer (50 mm Tris ⁄ HCl, pH 8, contain-
ing 100 mm NaCl, 10 mm b-mercaptoethanol and 10%
glycerol0 to remove the reduced glutathione, and stored at
)20 °C. The same procedure was used for the purification
of GST–MP deletion mutants.
CD spectroscopy
CD spectra were recorded using a Jasco-815 spectropola-
rimeter (Jasco Analytical Instruments, Easton, MD, USA).
The molar ellipticity was monitored from 190 to 250 nm
using 0.5 mgÆmL
)1
protein in a 0.2 cm path-length cuvette
with a bandwidth of 1 nm and response time of 1 s. All
spectra were corrected using respective buffer controls. The
stability of the protein was monitored by CD spectroscopy
in a PTC-423S Peltier thermal control system (Jasco), by
observing the change in the molar ellipticity at 210 nm due
to loss of secondary structure with increase in temperature.
The temperature was increased at a rate of 1 °CÆmin
)1
, and
the ellipticity was monitored from 20–100 °C. The melting

temperature (T
m
, the temperature at which 50% of the pro-
tein is denatured) of the protein was calculated from the
curve obtained.
Fluorescence spectroscopy
Fluorescence experiments were performed using a Perkin-
Elmer LS5S luminescence spectrometer (Perkin-Elmer, Wal-
tham, MA, USA). The intrinsic fluorescence spectrum was
monitored from 300 to 400 nm upon excitation at 280 nm
in a 1 cm path-length cuvette. Fluorescence of the protein
(0.2–0.4 mgÆmL
)1
)in20mm Tris ⁄ HCl buffer, pH 8.0, was
determined in the absence and presence of 8 m urea.
Purification of NV
SeMV was purified from infected Sesbania grandiflora
(21 days post-infection) as described previously [21].
Pull-down assay and ELISA for monitoring
protein–protein interactions
rMP was dialysed against assay buffer (20 mm Tris ⁄ HCl, pH
8). The stability of the protein in the assay buffer was
checked by storing the protein overnight at 4 °C. There was
no visible precipitation. The sample was centrifuged at 100 g
and the supernatant was used for the pull-down assay. rMP
and NV were mixed in a microcentrifuge tube (100 l g each
in 100 lL) and vortexed at low speed in an end-to-end
rotor at 4 °C. After 1 h incubation, 20 lL of Ni-NTA resin
(Novagen) was added, and the mixture was incubated at 4 °C
for 1 h with intermittent mixing. The mixture was subse-

quently centrifuged at 100 g for 60 s to pellet the bound
protein with the resin. The supernatant was discarded. The
pellet fraction was washed five times for 5 min each using
1 mL each of buffer W1 (20 mm Tris ⁄ HCl, 100 mm NaCl,
5mm imidazole, pH 8), buffer W2 (20 mm Tris ⁄ HCl,
200 mm NaCl, 10 mm imidazole, pH 8) and buffer W3
(20 mm Tris ⁄ HCl, 200 mm NaCl, 15 mm imidazole, pH 8).
Finally, the bound proteins were eluted using elution buffer
(20 mm Tris ⁄ HCl, 200 mm NaCl, 250 mm imidazole, pH 8).
The eluate, together with 50 lL aliquots of unbound and
wash fractions, were separated by SDS ⁄ PAGE followed by
silver staining to check for the presence of proteins.
The interaction between MP, MP mutants and NV was
also tested by ELISA as described previously with minor
modifications [46–49]. The wells of the ELISA plate (F8
Nunc Maxisorp loose, Nunc, Roskilde, Denmark) were
coated with 5.0 lg of first protein (100 lL per well) at
37 °C for 2 h. The protein was diluted with 1· NaCl ⁄ P
i
(pH 7.4). The unadsorbed protein was removed, and the
wells were blocked with 10% skimmed milk in 1· NaCl ⁄ P
i
for 1 h at 37 °C. The plates were then incubated with the
second protein for 2 h at 37 °C. BSA was used as a control.
The wells were washed three times for 3 min each with 1·
NaCl ⁄ P
i
containing 0.05% Triton X-100, and then three
times with 1· NaCl ⁄ P
i

for three min. The bound second
protein was detected using polyclonal antibodies specific to
it by incubating the wells with the antibodies (1 : 5000 dilu-
tion) for 1 h at room temperature. Washes were repeated as
before, and the wells were further incubated with goat anti-
rabbit IgG conjugated to horseradish peroxidase
(1 : 10 000) in 5% skimmed milk in 1· NaCl ⁄ P
i
(100 lL
per well) for 45 min, followed by washing for 15 min, as
described above, and addition of 1· substrate 3,3¢,5,5¢-tetra-
methylbenzidine (TMB) ⁄ H
2
O
2
(diluted from 20· stock
solution in distilled water). The reaction was stopped by
the addition of 2 N H
2
SO
4
(50 lL per well). Interactions
were quantified by determining the absorbance at 450 nm
using a SpectraMax 340PC384 absorbance microplate
reader (Molecular Devices Inc. Sunnyvale, CA, USA). All
experiments were performed in triplicate, and standard
deviations were calculated.
Yeast two-hybrid interaction
The yeast strain AH109 and plasmids pGBKT7 and
pGADT7 were from Matchmaker Two-Hybrid System 3

(Clontech Laboratories Inc., Mountain View, CA, USA),
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 269
conferring HIS3, ADE2, MEL1 and lacZ reporters and
allowing high-stringency assay s. For ba it and pr ey construc-
tion, the oligodeoxynucleotide primers shown in Table 2
(marked with prefix E) were used for PCR amplification of
MP, MP deletion mutant and CP genes. Amplicons were
sub-cloned into the EcoRI site of pGBKT7 (MP, ND16, ND35,
ND49, CD3, CD19 and CD38) and pGADT7 (CP) vectors.
The yeast strains were transformed with the constructs,
and colonies were grown according to the Clontech Yeast
Protocols Handbook (Protocol No. PT3024-1. 4. Version
No. PR973283, Clontech www.clontech.com/images/pt/
PT3024-1.pdf). Plasmid selection within yeast cells were
maintained by growing cells in minimal medium (0.67%
yeast nitrogen base, 2% glucose) with appropriate omission
of amino acids (–Leu and –Trp for yeast transformed with
both bait and prey plasmids). Replica plating was per-
formed under conditions of increasing stringency according
to the manufacturer’s suggestions, whereby interacting pro-
teins were sequentially analysed for growth on nutritional
selection plates containing –Leu ⁄ –His ⁄ –Trp or –Ade ⁄ –
Leu ⁄ –His ⁄ –Trp, with or without 5-bromo-4-chloro-3-indol-
yl-a-d-galactopyranoside (a-X-Gal) to monitor MEL1
reporter construct expression directly. Images were captured
after 4–6 days of growth at 30 °C. AH109 yeast cells trans-
formed with pGBKT7-P53 (murine p53 fused to GAL4
DNA BD) and pGADT7-T Ag (SV40 large T-antigen fused
to GAL4 DNA AD) that had previously been reported to

interact in a yeast two-hybrid assay [35] were used as a
positive control in these experiments.
To determine the strength of protein–protein interactions,
b-galactosidase solution assays were performed using ortho-
nitrophenyl-b-galactopyranoside as the substrate (Sigma-
Aldrich). b-galactosidase activity was quantified using
the following formula: 1000 · [D
420
) (1.75)] ⁄ (T · V ·
D
600
), where the attenuance at 420 nm (D
420
) is due to
product formation, D
600
is the cell density of the culture, T
is the reaction time (min) and V is the volume (mL). All
b-galactosidase assays were performed in triplicate using
constructs in both the GAL4 activating and DNA binding
domain fusion proteins. Expression of the interacting pro-
teins was measured by direct antigen-coating ELISA. The
cells from co-transformed colonies grown in 5 mL –Leu ⁄ –
Trp ⁄ –His SD liquid medium were lysed, and the lysate was
coated on to ELISA plates, blocked and washed as before.
Haemagglutinin polyclonal antibodies were used to check
expression of CP from the pGAD T7 vector, and cMyC
monoclonal antibody was used to monitor the expression
of MP and the deletion mutant proteins.
Acknowledgements

We thank Professor N. Appaji Rao and Professor
M.R.N. Murthy (Molecular biophysics unit, Indian
Institute of Science) for valuable discussions. We thank
the Department of Biotechnology and the Department
of Science and Technology, New Delhi, India, and the
Indian Institute of Science, Bangalore, India, for finan-
cial support. S.R.C. thanks the University Grant Com-
mission, New Delhi, India, for the senior research
fellowship.
References
1 Waigmann E, Lucas WJ, Citovsky V & Zambryski P
(1994) Direct functional assay for tobacco mosaic virus
cell-to-cell movement protein and identification of a
domain involved in increasing plasmodesmal permeabil-
ity. Proc Natl Acad Sci USA 91, 1433–1437.
2 Lazarowitz SG (1999) Probing plant cell structure and
function with viral movement proteins. Curr Opin Plant
Biol 2, 332–338.
3 Scholthof HB (2005) Plant virus transport: motions of
functional equivalence. Trends Plant Sci 10, 376–382.
4 Carrington JC, Kasschau KD, Mahajan SK & Schaad
MC (1996) Cell-to-cell and long-distance transport of
viruses in plants. Plant Cell 8, 1669–1681.
5 Kawakami S, Watanabe Y & Beachy RN (2004)
Tobacco mosaic virus infection spreads cell to cell as
intact replication complexes. Proc Natl Acad Sci USA
101, 6291–6296.
6 Kasteel D, Wellink J, Verver J, van Lent J, Goldbach R
& van Kammen A (1993) The involvement of cowpea
mosaic virus M RNA-encoded proteins in tubule forma-

tion. J Gen Virol 74, 1721–1724.
7 Canto T & Palukaitis P (1999) The hypersensitive
response to cucumber mosaic virus in Chenopodium
amaranticolor requires virus movement outside the
initially infected cell. Virology 265, 74–82.
8 Citovsky V, Knorr D & Zambryski P (1991) Gene I,
a potential cell-to-cell movement locus of cauliflower
mosaic virus, encodes an RNA-binding protein. Proc
Natl Acad Sci USA 88, 2476–2480.
9 Carvalho CM, Pouwels J, van Lent JW, Bisseling T,
Goldbach RW & Wellink J (2004) The movement pro-
tein of cowpea mosaic virus binds GTP and single-
stranded nucleic acid in vitro. J Virol 78, 1591–1594.
10 Nurkiyanova KM, Ryabov EV, Kalinina NO, Fan Y,
Andreev I, Fitzgerald AG, Palukaitis P & Taliansky M
(2001) Umbravirus-encoded movement protein induces
tubule formation on the surface of protoplasts and
binds RNA incompletely and non-cooperatively. J Gen
Virol 82, 2579–2588.
11 Lucas WJ (2006) Plant viral movement proteins: agents
for cell-to-cell trafficking of viral genomes. Virology
344, 169–184.
12 Tamm T & Truve E (2000) Sobemoviruses. J Virol 74,
6231–6241.
13 Zaumeyer WJ & Harter LL (1943) Inheritance of symp-
tom expression of bean mosaic virus 4. J Agric Res 67,
295–300.
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
270 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS
14 Bonneau C, Brugidou C, Chen L, Beachy RN &

Fauquet C (1998) Expression of the rice yellow mottle
virus P1 protein in vitro and in vivo and its involvement
in virus spread. Virology 244, 79–86.
15 Sivakumaran K, Fowler BC & Hacker DL (1998)
Identification of viral genes required for cell-to-cell
movement of southern bean mosaic virus. Virology 252,
376–386.
16 Meier M, Paves H, Olspert A, Tamm T & Truve E
(2006) P1 protein of Cocksfoot mottle virus is indis-
pensable for the systemic spread of the virus. Virus
Genes 32, 321–326.
17 Voinnet O, Pinto YM & Baulcombe DC (1999) Sup-
pression of gene silencing: a general strategy used by
diverse DNA and RNA viruses of plants. Proc Natl
Acad Sci USA 96, 14147–14152.
18 Sarmiento C, Gomez E, Meier M, Kavanagh TA &
Truve E (2007) Cocksfoot mottle virus P1 suppresses
RNA silencing in Nicotiana benthamiana and Nicotiana
tabacum. Virus Res 123, 95–99.
19 Bhuvaneshwari M, Subramanya HS, Gopinath K,
Savithri HS, Nayudu MV & Murthy MR (1995) Struc-
ture of sesbania mosaic virus at 3 A
˚
resolution. Structure
3, 1021–1030.
20 Subramanya HS, Gopinath K, Nayudu MV,
Savithri HS & Murthy MR (1993) Structure of Sesbania
mosaic virus at 4.7 A
˚
resolution and partial sequence

of the coat protein. J Mol Biol 229, 20–25.
21 Lokesh GL, Gopinath K, Satheshkumar PS &
Savithri HS (2001) Complete nucleotide sequence of
Sesbania mosaic virus: a new virus species of the genus
Sobemovirus. Arch Virol 146, 209–223.
22 Satheshkumar PS, Lokesh GL, Sangita V, Saravanan V,
Vijay CS, Murthy MR & Savithri HS (2004) Role
of metal ion-mediated interactions in the assembly and
stability of Sesbania mosaic virus T = 3 and T = 1
capsids. J Mol Biol 342, 1001–1014.
23 Satheshkumar PS, Lokesh GL, Murthy MR &
Savithri HS (2005) The role of arginine-rich motif
and b-annulus in the assembly and stability of
Sesbania mosaic virus capsids. J Mol Biol 353, 447–
458.
24 Nair S & Savithri HS (2009) Processing of SeMV poly-
proteins revisited. Virology 396, 106–117.
25 Savithri HS & Murthy MRN (2010) Structure and
assembly of Sesbania mosaic virus. Curr Sci 98, 346–
351.
26 Rost B, Yachdav G & Liu J (2004) The PredictProtein
server. Nucleic Acids Res 32, W321–W326.
27 Kawakami S, Padgett HS, Hosokawa D, Okada Y,
Beachy RN & Watanabe Y (1999) Phosphorylation
and ⁄ or presence of serine 37 in the movement protein
of tomato mosaic tobamovirus is essential for intracellu-
lar localization and stability in vivo. J Virol 73, 6831–
6840.
28 Sokolova M, Prufer D, Tacke E & Rohde W (1997)
The potato leafroll virus 17K movement protein is

phosphorylated by a membrane-associated protein
kinase from potato with biochemical features of protein
kinase C. FEBS Lett 400, 201–205.
29 Rhee Y, Tzfira T, Chen MH, Waigmann E &
Citovsky V (2000) Cell-to-cell movement of tobacco
mosaic virus: enigmas and explanations. Mol Plant
Pathol 1, 33–39.
30 Matsushita Y, Hanazawa K, Yoshioka K, Oguchi T,
Kawakami S, Watanabe Y, Nishiguchi M & Nyunoya
H (2000) In vitro phosphorylation of the movement
protein of tomato mosaic tobamovirus by a cellular
kinase. J Gen Virol 81, 2095–2102.
31 Waigmann E, Chen MH, Bachmaier R, Ghoshroy S &
Citovsky V (2000) Regulation of plasmodesmal trans-
port by phosphorylation of tobacco mosaic virus
cell-to-cell movement protein. EMBO J 19,
4875–4884.
32 Ruan K, Li J, Liang R, Xu C, Yu Y, Lange R &
Balny C (2002) A rare protein fluorescence behavior
where the emission is dominated by tyrosine: case of
the 33-kDa protein from spinach photosystem II.
Biochem Biophys Res Commun 293, 593–597.
33 Lokesh GL (2002) Molecular insight into the genome
organization and assembly of Sesbania mosaic virus. PhD
thesis, Indian Institute of Science, Bangalore, India.
34 Chowdhury SR (2010) Molecular characterization of
movement protein encoded by ORF 1 of Sesbania mosaic
virus. PhD thesis, Indian Institute of Science, Bangalore,
India.
35 Li B & Fields S (1993) Identification of mutations in

p53 that affect its binding to SV40 large T antigen by
using the yeast two-hybrid system. FASEB J 7, 957–
963.
36 Rojas MR, Noueiry AO, Lucas WJ & Gilbertson RL
(1998) Bean dwarf mosaic geminivirus movement
proteins recognize DNA in a form- and size-specific
manner. Cell 95, 105–113.
37 Brill LM, Dechongkit S, DeLaBarre B, Stroebel J,
Beachy RN & Yeager M (2004) Dimerization of recom-
binant tobacco mosaic virus movement protein. J Virol
78, 3372–3377.
38 Brill LM, Nunn RS, Kahn TW, Yeager M &
Beachy RN (2000) Recombinant tobacco mosaic virus
movement protein is an RNA-binding, a-helical
membrane protein. Proc Natl Acad Sci USA 97, 7112–
7117.
39 Lakowicz J (1999) Principles of Fluorescence Spectros-
copy. Kluwer Academic ⁄ Plenum Publishers, New York.
40 Canetta E, Kim SH, Kalinina NO, Shaw J, Adya AK,
Gillespie T, Brown JW & Taliansky M (2008) A plant
virus movement protein forms ringlike complexes with
the major nucleolar protein, fibrillarin, in vitro. J Mol
Biol 376, 932–937.
S. R. Chowdhury and H. S. Savithri Interaction of SeMV MP with viral coat protein
FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS 271
41 Lekkerkerker A, Wellink J, Yuan P, van Lent J,
Goldbach R & van Kammen AB (1996) Distinct func-
tional domains in the cowpea mosaic virus movement
protein. J Virol 70, 5658–5661.
42 van Lent J, Storms M, van der Meer F, Wellink J &

Goldbach R (1991) Tubular structures involved in
movement of cowpea mosaic virus are also formed in
infected cowpea protoplasts. J Gen Virol 72, 2615–2623.
43 Kumar PP, Usha R, Zrachya A, Levy Y, Spanov H &
Gafni Y (2006) Protein–protein interactions and nuclear
trafficking of coat protein and bC1 protein associated
with Bhendi yellow vein mosaic disease. Virus Res 122,
127–136.
44 Lee JY & Lucas WJ (2001) Phosphorylation of viral
movement proteins – regulation of cell-to-cell traffick-
ing. Trends Microbiol 9, 5–8.
45 Laemmli UK (1970) Cleavage of structural proteins
during the assembly of the head of bacteriophage T4.
Nature 227, 680–685.
46 Carvalho CM, Wellink J, Ribeiro SG, Goldbach RW &
Van Lent JW (2003) The C-terminal region of the
movement protein of Cowpea mosaic virus is involved
in binding to the large but not to the small coat protein.
J Gen Virol 84, 2271–2277.
47 Goodfellow I, Chaudhry Y, Gioldasi I,
Gerondopoulos A, Natoni A, Labrie L, Laliberte JF &
Roberts L (2005) Calicivirus translation initiation
requires an interaction between VPg and eIF4E. EMBO
Rep 6, 968–972.
48 Kaiser WJ, Chaudhry Y, Sosnovtsev SV &
Goodfellow IG (2006) Analysis of protein–protein
interactions in the feline calicivirus replication complex.
J Gen Virol 87, 363–368.
49 Leonard S, Plante D, Wittmann S, Daigneault N,
Fortin MG & Laliberte JF (2000) Complex formation

between potyvirus VPg and translation eukaryotic initi-
ation factor 4E correlates with virus infectivity. J Virol
74, 7730–7737.
Interaction of SeMV MP with viral coat protein S. R. Chowdhury and H. S. Savithri
272 FEBS Journal 278 (2011) 257–272 ª 2010 The Authors Journal compilation ª 2010 FEBS