Virus-induced gene silencing in plants.
ABSTRACT Virus-induced gene silencing (VIGS) is a technology that exploits an RNA-mediated antiviral defense mechanism. In plants infected with unmodified viruses the mechanism is specifically targeted against the viral genome. However, with virus vectors carrying inserts derived from host genes the process can be additionally targeted against the corresponding mRNAs. VIGS has been used widely in plants for analysis of gene function and has been adapted for high-throughput functional genomics. Until now most applications of VIGS have been in Nicotiana benthamiana. However, new vector systems and methods are being developed that could be used in other plants, including Arabidopsis. Here we discuss practical and theoretical issues that are specific to VIGS rather than other gene "knock down" or "knockout" approaches to gene function. We also describe currently used protocols that have allowed us to apply VIGS to the identification of genes required for disease resistance in plants. These methods and the underlying general principles also apply when VIGS is used in the analysis of other aspects of plant biology.
- SourceAvailable from: G. P. Lee[Show abstract] [Hide abstract]
ABSTRACT: We developed a reassortant RNA virus vector derived from (CMV), which has advantages of very wide host range and can efficiently induce gene silencing in a few model plants. Certain CMV isolates, however, show limited host ranges presumably because they naturally co-evolved with their own hosts. We used a reassortant comprised of two strains of CMV, Y-CMV and Gn-CMV, to broaden the host range and to develop a virus vector for virus-induced gene silencing (VIGS). Gn-CMV could infect chili pepper and tomato more efficiently than Y-CMV. Gn-CMV RNA1, 3 and Y-CMV RNA2-A1 vector were newly reconstructed, and the transcript mixture of RNA1 and 3 genomes of Gn-CMV and RNA2 genome of Y-CMV RNA2 containing portions of the endogenous phytoene desaturase (PDS) gene (CMV2A1::PDSs) was inoculated onto chili pepper (cv. Chung-yang), tomato (cvs. Bloody butcher, Tigerella, Silvery fir tree, and Czech bush) and . All the tested plants infected by the reassortant CMV vector showed typical photo-bleaching phenotypes and reduced expression levels of mRNA. These results suggest that the reassortant CMV vector would be a useful tool for the rapid induction of the RNA silencing of endogenous genes in chili pepper and tomato plants.The plant pathology journal 01/2012; 28(1). · 0.67 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Virus-induced gene silencing (VIGS) is a recently developed gene transcript suppression technique for characterizing the function of plant genes. However, efficient VIGS has only been studied in a few plant species. In order to extend the application of VIGS, we examined whether a VIGS vector based on TRV would produce recognizable phenotypes in soybean. Here, we report that VIGS using the Tobacco rattle virus (TRV) viral vector can be used in several soybean cultivars employing various agro-inoculation methods including leaf infiltration, spray inoculation, and agrodrench. cDNA fragments of the soybean phytoene desaturase(PDS) was inserted into TRV RNA-2 vector. By agrodrench, we successfully silenced the expression of PDS encoding gene in soybean. The silenced phenotype of PDS was invariably obvious 3 weeks after inoculation with the TRV-based vector. Real-time RT-PCR analyses showed that the endogenous level of GmPDS transcripts was dramatically reduced in the silenced leaf tissues. These observations confirm that the silenced phenotype is closely correlated with the pattern of tissue expression. The TRV-based VIGS using agrodrench can be applied to functional genomics in a soybean plants to study genes involved in a wide range of biological processes. To our knowledge, this is the first high frequency VIGS method in soybean plants.The plant pathology journal 01/2005; 21(2). · 0.67 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Lycopene ε-cyclase (ε-LCY) is a key enzyme that catalyzes the synthesis of α-branch carotenoids through the cyclization of lycopene. Two cDNA molecules encoding ε-LCY (designated Ntε-LCY1 and Ntε-LCY2) were cloned from Nicotiana tabacum. Ntε-LCY1 and Ntε-LCY2 are encoded by two distinct genes with different evolutionary origins, one originating from the tobacco progenitor, Nicotiana sylvestris, and the other originating from Nicotiana tomentosiformis. The two coding regions are 97% identical at the nucleotide level and 95% identical at the amino acid level. Transcripts of Ntε-LCY were detectable in both vegetative and reproductive organs, with a relatively higher level of expression in leaves than in other tissues. Subcellular localization experiments using an Ntε-LCY1-GFP fusion protein demonstrated that mature Ntε-LCY1 protein is localized within the chloroplast in Bright Yellow 2 suspension cells. Under low-temperature and low-irradiation stress, Ntε-LCY transcript levels substantially increased relative to control plants. Tobacco rattle virus (TRV)-mediated silencing of ε-LCY in Nicotiana benthamiana resulted in an increase of β-branch carotenoids and a reduction in the levels of α-branch carotenoids. Meanwhile, transcripts of related genes in the carotenoid biosynthetic pathway observably increased, with the exception of β-OHase in the TRV-ε-lcy line. Suppression of ε-LCY expression was also found to alleviate photoinhibition of Potosystem II in virus-induced gene silencing (VIGS) plants under low-temperature and low-irradiation stress. Our results provide insight into the regulatory role of ε-LCY in plant carotenoid biosynthesis and suggest a role for ε-LCY in positively modulating low temperature stress responses.International Journal of Molecular Sciences 01/2014; 15(8):14766-14785. · 2.46 Impact Factor
Virus-induced gene silencing in plants
Rui Lu, Ana Montserrat Martin-Hernandez, Jack R. Peart, Isabelle Malcuit,
and David C. Baulcombe*
The Sainsbury Laboratory, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK
Accepted 7 February 2003
Virus-induced gene silencing (VIGS) is a technology that exploits an RNA-mediated antiviral defense mechanism. In plants
infected with unmodified viruses the mechanism is specifically targeted against the viral genome. However, with virus vectors
carrying inserts derived from host genes the process can be additionally targeted against the corresponding mRNAs. VIGS has been
used widely in plants for analysis of gene function and has been adapted for high-throughput functional genomics. Until now most
applications of VIGS have been in Nicotiana benthamiana. However, new vector systems and methods are being developed that
could be used in other plants, including Arabidopsis. Here we discuss practical and theoretical issues that are specific to VIGS rather
than other gene ‘‘knock down’’ or ‘‘knockout’’ approaches to gene function. We also describe currently used protocols that have
allowed us to apply VIGS to the identification of genes required for disease resistance in plants. These methods and the underlying
general principles also apply when VIGS is used in the analysis of other aspects of plant biology.
? 2003 Elsevier Science (USA). All rights reserved.
Keywords: Functional genomics; Posttranscriptional gene silencing; Virus vectors
When a plant virus infects a host cell it activates an
RNA-based defense that is targeted against the viral
genome . By analogy with RNA interference in ani-
mals it is thought that this mechanism involves pro-
cessing of double-stranded (ds) RNA into short
interfering (si) RNAs . An RNase complex is then
guided by base pairing of the siRNAs so that it specif-
ically targets single-stranded (ss) target RNA that is
similar to the dsRNAs.
The dsRNA in virus-infected cells is thought to be the
replication intermediate that causes the siRNA/RNase
complex to target the viral single-stranded RNA. In the
initially infected cell the viral ssRNA would not be a
target of the siRNA/RNase complex because this repli-
cation intermediate would not have accumulated to a
high level. However, in the later stages of the infection,
as the rate of viral RNA replication increases, the viral
dsRNA and siRNA would become more abundant.
Eventually, the viral ssRNA would be targeted inten-
sively and virus accumulation would slow down .
Many plant viruses encode proteins that are sup-
pressors of this RNA silencing process [4,5]. These
suppressor proteins would not be produced until after
the virus had started to replicate in the infected cell so
they would not cause complete suppression of the RNA-
based defense mechanism. However, these proteins
would influence the final steady-state level of virus ac-
cumulation. Strong suppressors would allow virus ac-
cumulation to be prolonged and at a high level.
Conversely, if a virus accumulates at a low level it could
be due to weak suppressor activity.
Virus-induced gene silencing—VIGS—is a virus vector
technology that exploits this RNA defense. The dsRNA
replication intermediate would be processed so that the
siRNA in the infected cell would correspond to parts of
the viral vector genome, including any nonviral insert.
Thus, if the insert is from a host gene, the siRNAs would
target the RNase complex to the corresponding host
mRNA and the symptoms in the infected plant would
reflect the loss of the function in the encoded protein.
Methods 30 (2003) 296–303
*Corresponding author. Fax: 44-0- 1603-450011.
1046-2023/03/$ - see front matter ? 2003 Elsevier Science (USA). All rights reserved.
There are now several examples that validate this
approach to suppression of gene expression. Thus, when
tobacco mosaic virus (TMV) or potato virus X (PVX)
vectors were modified to carry inserts from the plant
phytoene desaturase gene the photobleaching symptoms
on the infected plant reflected the absence of photo-
protective carotenoid pigments that require phytoene
desaturase [6,7]. Similarly when the virus carried inserts
of a chlorophyll biosynthetic enzyme  there were
chlorotic symptoms and, with a cellulose synthase insert,
the infected plant had modified cell walls .
Genes other than those encoding metabolic enzymes
can also be targeted by VIGS. For example, if the viral
insert corresponded to genes required for disease resis-
tance, the plant exhibited enhanced pathogen suscepti-
bility. In one such example the insert in a tobacco rattle
virus (TRV) vector was from a gene (EDS1) that is re-
quired for N-mediated resistance to TMV. The virus
vector-infected N-genotype plant exhibited compro-
mised TMV resistance .The outcome of this experi-
ment is illustrated in Fig. 1. Other defense-related genes
have also been targeted by VIGS [11–15]
The symptoms of a TRV vector carrying a leafy insert
demonstrate how VIGS can be used to target genes that
regulate development. Leafy is a gene required for flower
development. Loss-of-function leafy mutants produce
modified flowers that are phenocopied in the TRV-leafy-
infected plants . Similarly the effects of tomato golden
mosaic virus vectors carrying parts of the gene for a
cofactor of DNA polymerase illustrate how VIGS can be
used to target essential genes . The plants infected
with this geminivirus vector were suppressed for division
growth in and around meristematic zones of the shoot.
Fig. 2 illustrates VIGS phenotypes due to silencing of
other essential genes from a survey of approximately
5000 different Nicotiana benthamiana cDNAs.
Until now most of the applications of VIGS have been
in the plant virologist?s model plant—N. benthamiana.
For reasons that are not fully understood this plant is
susceptible to an unusually wide range of viruses and the
VIGS symptoms in N. benthamiana are generally much
more pronounced and more persistent than in other
plants, including N. tabacum. However, VIGS can be
applied in otherspecies. For example,with abarley stripe
mosaic virus vector, VIGS of PDS has been demon-
strated in a monocot—barley . In addition there are
TRV vectors with the ability to support VIGS in Ara-
bidopsisandtomato aswellasinN. benthamiana.
Why is VIGS particularly good in N. benthamiana?
One possible explanation is the plasmodesmatal ex-
clusion limit that may be greater in N. benthamiana
than in other species . Alternatively this Australian
native plant could be defective in defense components
that normally restrict the rate or extent of virus accu-
mulation. The ability of viruses to approach meristem
cells may also be relevant. Most viruses are excluded
from the meristem and adjacent cells  and cannot
silence gene expression in these tissues. Consequently,
if the target mRNA of VIGS encodes a stable protein,
the silencing phenotype in mature differentiated tissue
could be masked by protein produced in the immature
cells that are in and adjacent to the meristem. How-
ever, if the virus penetrates close to the meristem, an
mRNA will be silenced before the encoded protein has
accumulated to a high level and the silencing pheno-
type would be enhanced.
It was surprising that meristem-expressed genes could
be targeted because viruses are normally excluded from
this region of the plant, as discussed below. In these
instances it may be that the silencing phenotype is
mediated by an intercellular signal of silencing that
spreads from the infected cell [17,23,24]. This signal may
be a nucleic acid—either double-stranded RNA or one
of the classes of short RNA associated with silencing.
Unfortunately there are complications for the appli-
cation of VIGS because it involves the use of genetically
Fig. 1. VIGS of NbEDS1 compromises N-mediated resistance to TMV.
(a) Schematic representation of the VIGS procedure to test the
requirement of NbEDS1 for N-mediated resistance. Four- to five-
week-old N-transgenic seedlings were inoculated with TRV vectors by
agroinfiltration and, approximately 21 days later, upper leaves were
challenge inoculated with TMV:GFP. Accumulation of TMV:GFP
was monitored by GFP fluorescence under UV illumination. (b)
TRV:00, TRV:N, or TRV:EDS plants were challenge inoculated with
TMV:GFP sap and accumulation of TMV:GFP was monitored by
GFP fluorescence at 5dpi on inoculated leaves (top) and at 15dpi in
systemic organs (bottom). White arrows indicate foci of GFP. Re-
produced with permission from Peart et al., Plant J. 29 pp 569–579,
R. Lu et al. / Methods 30 (2003) 296–303
modified plant pathogens and must be carried out in
strict containment. Moreover the VIGS phenotype may
be influenced by the effects of the virus vector in the
infected plant. It is an approach that is not suitable for
analysis of many traits associated with reproductive bi-
ology. However, notwithstanding these practical com-
plications there are several advantages to VIGS over
gene-silencing technologies involving transgenic plants
with inverted repeat constructs . First, the constructs
can be assembled by direct cloning in the virus vector
and do not involve assembly of inverted repeats that
may be unstable during propagation in bacterial hosts.
Second, the procedure is rapid—virus vector constructs
can be assembled in a few days and the VIGS phenotype
develops within 1 or 2 weeks. Consequently it is feasible
to carry out high-throughput VIGS of many genes in a
target plant genome. A further advantage of VIGS is its
conditional nature. The target mRNA is not silenced
until the virus vector infects the plant. It is possible
therefore to suppress genes, for example PCNA ,
that are essential for host cell growth and development.
Most mutations in such genes would be lethal and dif-
ficult to retrieve. Similarly, transgene silencing could not
be used to investigate essential genes. The silencing
phenotype would cause the cells to die during regener-
ation of transformed plants.
2. Future refinements of VIGS
Although VIGS has been used widely for analysis of
gene function, particularly for genes involved in defense,
there is still scope for refining the virus vector technol-
ogy. One of the factors that may be important is the
presence of subgenomic promoters in the virus vector. In
the first described version of a tobacco rattle VIGS
vector the nonviral insert was not coupled to a viral
subgenomic promoter . However, in a later version
that is effective on a broader host range and that causes
more rapid VIGS , there is a subgenomic promoter
upstream of the nonviral insert. These data suggest that
a systematic analysis of subgenomic promoters and their
effects on VIGS could lead to production of more effi-
A second approach to refinement of VIGS vectors is
illustrated by the development of an efficient TMV
vector for protein expression . This refined vector
directs severalfold higher levels of foreign protein ex-
pression than its progenitor and was developed by DNA
shuffling of the viral movement protein gene. In princi-
ple a similar ‘‘directed evolution’’ approach could also
be used to adapt VIGS virus vectors for plant hosts
other than N. benthamiana. One or more of the viral
genes could be shuffled and the resulting mutants
screened for enhanced VIGS.
In most of the reported examples of VIGS the target
sequence was the transcribed region of a gene and ex-
pression was suppressed posttranscriptionally. However,
as exemplified by the use of a TRV vector carrying a
Fig. 2. Selected VIGS phenotypes. From a VIGS survey of 5000 dif-
ferent N. benthamiana cDNAs approximately 15% produced pro-
nounced symptoms resulting in suppression of plant growth or
development. These images illustrate three of the symptom types due
to silencing of ubiquitin (a), magnesium chelatase (b), or an unknown
R. Lu et al. / Methods 30 (2003) 296–303
transgene promoter, VIGS is not restricted to post-
transcriptional mechanisms. Plants infected with this
vector exhibited specific transcriptional silencing of the
target promoter . The transcriptional VIGS by this
construct was accompanied by methylation of the target
DNA and, surprisingly, the imprint was inherited in the
progeny of the infected plant although the virus was not.
Clearly this observation raises many questions about the
potential mechanisms of VIGS. However, it also indi-
cates that VIGS could be used to silence genomic pro-
moters. A report that transgene dsRNA constructs
direct transcriptional silencing of endogenous gene
promoters is consistent with this proposed use of VIGS
to target promoter sequences .
3. Practicalities of VIGS
3.1. The vector insert
There are several issues to be addressed concerning
the vector insert in a VIGS experiment, including length.
In test experiments we established that phytoene desat-
urase could be silenced with inserts that were as short as
23 nucleotides . However, the silencing was less ex-
tensive and more transient than with larger inserts and
we routinely use sequence elements of 150–500 nucleo-
tides. Larger inserts can be used but they are genetically
unstable in the virus vector and do not cause enhance-
ment of the silencing phenotype.
A second consideration is location of the inserted
sequence within the target gene. If the insert is from a
region that is unique to the target gene the silencing
phenotype will be highly gene specific. However, if the
insert includes regions that are similar in related genes or
other members of the gene family, the specificity may be
compromised. In some situations, for example if there is
functional redundancy in a gene family, this sequence—
rather than genetic locus—specificity is an advantage
because conventional genetic approaches would not be
Until now there has not been a systematic analysis of
the sequence similarity needed for cross-silencing of re-
lated genes. However, based on our analysis with short
inserts of phytoene desaturase  and on the under-
standing of RNA silencing mechanisms, it seems likely
that VIGS could target any genes that have more than
about 20 nucleotides perfectly matched to the insert in
the VIGS vector. At present we do not know whether
VIGS is better with coding or noncoding targets.
However, introns do not direct VIGS .
3.2. The virus vector
Having identified suitable inserts for a VIGS experi-
ment the next decision is the virus vector. To some
extent this issue is determined by the plant of interest.
TRV is effective on N. benthamiana, tomato, Arabidop-
sis, and other related species [16,19,20]. PVX induces
strong silencing on N. benthamiana and N. clevelandii 
and barley stripe mosaic virus on barley . Gemini-
virus vectors have also been used for N. benthamiana 
and Arabidopsis .
3.3. Vectors for in vitro transcription and ‘‘agroinocula-
RNA virus vectors were originally developed in a
form that required in vitro transcription of infectious
RNA from a linearized plasmid DNA template [31–33].
A protocol for such a PVX vector is described in Pro-
tocol I. However, an alternative procedure referred to as
agroinoculation has been adopted because it is easier to
use than in vitro transcription. Agroinoculation involves
the use of Ti plasmid vectors of Agrobacterium tum-
efaciens in which a region—the T-DNA—is transferred
to the genome of infected plant cells. This process can be
exploited for inoculation of plant viruses if the full-
length viral cDNA is present in the T-DNA and between
a promoter and a transcriptional terminator that is ac-
tive in plant cells .
Agroinoculation has advantages over in vitro tran-
scription because the virus vector cDNA does not have
to be isolated, digested, or transcribed. The Agrobacte-
rium cells carrying the insert can be inoculated directly
into the plant. Presumably there are cells at the site of
inoculation that are transformed to be infected with the
virus vector that is represented in the T-DNA. These
cells would then serve as a reservoir of infection that
spreads systemically throughout the plant. Agroinocu-
lation is well established for DNA viruses and it was
initially thought not to be efficient for TMV, an RNA
virus . However, our experience with PVX is that
agroinoculation is the most potent way of introducing
cDNA-derived viral RNA into plants (R. Lu and D.C.
Baulcombe, unpublished data). Agroinoculation is use-
ful for many VIGS applications. However, it is partic-
ularly useful in high-throughput applications when, for
example, cDNA libraries or collections of ESTs are to
Cloning of inserts into the viral vector exploits stan-
dard ligation and transformation procedures. Protocol
II provides a procedure for efficient transformation of
A. tumefaciens that we have used with PVX and TRV
Inoculation of in vitro transcripts is by manual
inoculation following standard virological procedures,
as described in Protocol III. For agroinoculation with
PVX vectors a simple toothpick method is used. An
R. Lu et al. / Methods 30 (2003) 296–303
Agrobacterium colony carrying the viral vector construct
is simply spiked with a toothpick that is then used to
pierce a hole in the leaf or petiole of the plant. With
TRV a modified agroinoculation procedure has to be
used because the genome has two RNA components
that are carried separately on Ti plasmid vectors in
different strains of Agrobacterium. It is therefore neces-
sary to ensure simultaneous agroinfection with these two
strains and, in our experience, the toothpick method is
not efficient for this purpose. The modified procedure
involves infiltration of a mixture of the Agrobacterium
strains into a leaf of a young seedling following a pro-
cedure that is described in the Protocol IV.
Unfortunately neither direct inoculation of tran-
scripts nor agroinoculation has been effective in our
hands on Arabidopsis with a TRV vector. We have
therefore resorted to a more involved inoculation pro-
cedure that is not applicable in high-throughput appli-
cations but which can be used for targeting silencing of
selected genes. The procedure, described in Protocol V,
exploits the ability of TRV to infect both N. benthami-
ana and Arabidopsis. It involves an intermediate step in
which a high-titer inoculum is prepared from agroin-
oculated leaves of N. benthamiana and then mechani-
cally inoculated to Arabidopsis.
3.5. Assessment of VIGS
The VIGS symptoms appear several days postinoc-
ulation (dpi) and become extensive throughout the in-
fected plant. At later times they may fade. The precise
timing of the maximum VIGS is very dependent on
environmental conditions and it is advisable to set up
tests with well-characterized vector constructs to moni-
tor the progress of the infection process. The phytoene
desaturase vectors are useful for this purpose because
they produce photobleaching symptoms that are very
obvious. However, it may be appropriate also to use
vectors that are targeted against genes involved in the
trait of interest. For example, to identify genes required
for disease resistance it would be appropriate to estab-
lish the progression of VIGS using a control vector that
targets a known defense-related gene.
Several points should be borne in mind when inter-
preting a VIGS phenotype. The first is that the absence
of a phenotype does not necessarily rule out involve-
ment of the target gene in the trait of interest. VIGS is
never complete and it is always possible that a silencing
phenotype was not observed because the target gene
function was supported by the residual low level of
mRNA in the virus vector-infected plants. A second
issue in the interpretation of VIGS data is the possibility
that chance sequence similarity between the insert and
an unknown mRNA is responsible for the phenotype.
To rule out this artifact we routinely use a second no-
noverlapping insert from the same target gene. If the
target gene has been correctly identified this second in-
sert would reproduce the original VIGS phenotype. If
the target gene is a member of a multigene family it is
necessary to target conserved and nonconserved regions,
as discussed above, to determine whether the silencing
phenotype was influenced by one or several family
members. We also assess the abundance of the intended
target RNA or protein to confirm directly that its ex-
pression has been silenced.
Pleiotropy is also an issue that needs to be considered
when interpreting VIGS experiments. We encountered
this complication in a large-scale survey of genes re-
quired for disease resistance in plants. Of approximately
5000 genes tested more than 100 had a phenotype that
was manifested as loss of a cell death response that is
often associated with disease resistance. However, fewer
than 10 of these candidate genes showed a loss of disease
resistance following from VIGS (R. Lu et al., unpub-
lished data). The result was interesting in that it showed
how disease resistance could be separated from the cell
death response. It thus confirmed that these are separate
branches of disease resistance. However, the VIGS assay
did not by itself indicate whether the candidate genes
were directly implicated in the cell death response. It is
likely, in some instances at least, that these genes pro-
duce secondary factors that are not directly involved in
the cell death pathway.
4. Concluding comments
In the analysis of gene function the results of VIGS
are a starting point. They provide an indication, but not
proof, that a gene is implicated in a particular trait.
Further tests including biochemical analyses of protein
function are required for full understanding. Neverthe-
less the speed and ease of application of VIGS mean that
in many instances this approach is a logical first step in
the characterization of an unknown gene.
To our knowledge it is not yet known whether VIGS
can be applied in animals. The enzymes required for the
RNA-mediated defense are certainly present in animals
and it is known that they can target viral RNAs [2,35]. A
further indication that the RNA-silencing-based antivi-
ral defense mechanism operates in animals is from the
recent observation that an insect virus encodes a protein
suppressor of the RNA-silencing mechanism . Pre-
sumably these suppressors would have evolved only if
RNA silencing is an antiviral defense in animals as in
plants. To test this idea it would be necessary to intro-
duce a host gene fragment into an animal virus vector
and to ask whether the corresponding mRNA was tar-
geted in the infected animal. This experiment may not
have been carried out. However, it would be interesting
to know the outcome: if VIGS could be adapted to
animals it would be possible to develop new strategies
R. Lu et al. / Methods 30 (2003) 296–303
for disease therapy and carry out high-throughput
analysis of gene function in whole animals.
5.1. Virus vectors
Virus vectors for VIGS in plants have been described
in several recent papers [16,18,20,30]. Those described
by us and colleagues can be obtained by application to
our Web site (www.sainsbury-laboratory.ac.uk). Recip-
ients of these vectors must indicate that they are au-
thorized to receive and to carry out experiments with
genetically modified plant viruses. Details of vectors and
procedures from the Dinesh–Kumar laboratory at Yale
are available at http://kumar8.biology.yale.edu/lab/Ge-
5.2. Protocol I—in vitro transcription of PVX cDNA
All solutions and equipment should be RNase free.
Solutions should be made up with DEPC-treated water;
all tubes and tips should be sterile and gel running
equipment should be treated with 0.5M NaOH to de-
(1) Linearize 25lg of template DNA in a volume of
50ll with 20U of the appropriate enzyme. To ensure
complete digestion the reaction should be carried out at
37?C for 16h. Complete linearization of the DNA is
essential as circular DNA is a much better template for
transcription than linear DNA.
(2) Extract with 30ll of phenol/chloroform. Take
45ll of the aqueous phase and spin dialyze through
Sepharose CL-6B in a 0.5-ml Eppendorf tube. Check
that the DNA has been efficiently linearized by electro-
phoresis of a 1-ll aliquot on a 1% agarose gel.
(3) Set up the transcription reaction as indicated be-
low. The order in which the reactants is added is im-
portant because the RNase inhibitor will release RNase
in the absence of DTT and because the spermidine in the
T7 buffer can precipitate DNA. To prevent DNA pre-
cipitation the reaction components should be mixed
together at room temperature.
(a) 10ll 5? T7 buffer (5? buffer is 200mM Tris–HCl,
pH 8.0, 40mM MgCl2, 10mM spermidine, 125mM
(b) 2.5ll 0.1M DTT
(c) 1ll placental RNase inhibitor (37–40 U/ll from
Boehringer or Pharmacia)
(d) 5ll 10? A/C/U/G mix (10? A/C/U/G mix is
20mM each ATP, CTP, and UTP and 2mM GTP)
(e) 5ll 5mM cap (m7GpppG from Boehringer
or Pharmacia; dissolve 5U in 60ll)
(f) 11.5ll water
(g) 10ll linear DNA
(4) Incubate for 5min at 37?C, add 5ll of T7 RNA
polymerase (50U/ll from GIBCO-BRL) and mix by
(5) Incubate reaction for 25min at 37?C and then
add 5ll of 20mM GTP. Do not mix vigorously.
(6) Incubate for 35min at 37?C. Carry out a phenol/
chloroform extraction with 30ll and recover 45ll of the
aqueous phase. Set aside 2ll of the transcripts for
running on a gel.
(7) Transcripts can be run on a 1% agarose gel
alongside RNA molecular weight standards. Running
the bromophenol blue dye half-way down the gel will
ensure separation of the DNA and RNA.
5.3. Protocol II—high efficiency transformation of Agro-
A. tumefaciens strains were made electrocompetent
for high efficiency transformation as follows: 500ml of
SOB medium (2% Bacto tryptone, 0.5% Bacto yeast
extract, 10mM NaCl, 2.5mM KCl) in a 2-L flask is
inoculated with 1.0ml of an overnight culture of bac-
teria and incubated for 5–6h at 28?C with vigorous
shaking. When the OD550 reaches 0.7, the culture is
chilled on ice for 30min. The cells are then harvested at
4000g for 15min at 4?C, washed four times with 250ml
10% glycerol, and resuspended to a final volume of
1.0ml in ice-cold 10% glycerol. The cells are used im-
mediately or stored at )70?C in 60-ll aliquots.
To 20ll of electrocompetent cells add 0.5–3ll of a
ligation which has previously been spin dialyzed. The
electroporation is performed using the GIBCO BRL
electroporator (Cell-Porator and Voltage Booster) in a
prechilled 0.15-cm cuvette according to conditions rec-
ommended by the manufacturer (capacitance 330lF,
resistance 4000X, voltage 380 V,1impedance low X,
charge rate fast). Cells are then transferred to 0.5ml of
SOC medium and incubated at an appropriate temper-
ature for 1h with gentle shaking (100rpm). The trans-
formed cells are selected on L plates supplemented with
the appropriate antibiotic.
5.4. Protocol III—mechanical inoculation of PVX tran-
For inoculation of one plant use the transcripts of
1lg plasmid DNA in 20ll DEPC-treated water. Add
1M sodium phosphate, pH 7.0, to a final concentration
of 42.9mM, vortex, and spin briefly. Add bentonite (10–
20%), and mix. Mark two adjacent leaves of a four- to
six-leaf tobacco plant for inoculation and lightly dust
the leaves to be infected with carborundum from a
1Note. A voltage at 380V rather than 400V will reduce the
frequency of explosions encountered during electroporation of A.
tumefaciens strains without affecting the transformation efficiency.
R. Lu et al. / Methods 30 (2003) 296–303
plastic water bottle (silicon carbide, 600 grit; BDH).
Using a Gilsen pipette add 10–20ll of solution, in small
drops, onto each leaf to be inoculated. Wearing gloves,
gently rub the liquid across the surface of each leaf.
After 5min wash off the carborundum. It may help to
soften the leaves by leaving plants in the dark for 24h
prior to inoculation.
5.5. Protocol IV—agroinoculation of TRV vectors
The TRV vector comprises two Ti plasmid constructs
that are propagated separately in A. tumefaciens. pTV00
is based on RNA2 of TRV strain PPK20 with a poly-
linker of (50to 30) SpeI, BamHI, SmaI, XmaI, HindIII,
BspDI, ClaI, AccI, ApaI, and KpnI for insertion of
nonviral sequence. The sequence of TRV RNA2 is
available from our Web site (www.sainsbury-labora-
tory.ac.uk). pBINTRA6 is an RNA1 Ti plasmid clone
of TRV (GenBank Accession No. AF314165) that has
been described elsewhere.
These vectors are provided as DNA samples. To
propagate in Escherichia coli use 50lg/ml kanamycin
for both constructs. We recommend E. coli strain
DH5a, especially for pBINTRA6, since it tends to be
unstable in E. coli. For propagation in A. tumefaciens
use C58C1 for pBINTRA6 and GV3101 with pSa-rep
 for pTV00 and derivatives and select with a com-
bination of 50lg/ml kanamycin, 5lg/ml tetracycline,
and 50lg/ml rifampicin (rifampicin can be omitted).
Separate A. tumefaciens cultures (10 ml) containing
pBINTRA6 and the TRV RNA2 constructs are grown
overnight at 28?C, spun down at 3000rpm for 20min,
and resuspended in the same volume of 10mM MgCl2,
with 100lM acetosyringone and 1mM Mes, pH 5.6.
These cultures can be mixed (1:1) and infiltrated into
leaves by pressing a syringe against the lower surface.
The temperature should be below 24?C, because higher
temperatures may inhibit T-DNA transfer.
5.6. Protocol V—Inoculation of TRV vectors to Arabid-
For TRV infection of Arabidopsis we first inoculate
N. benthamiana by infiltration and extract sap from the
inoculated leaf after 3 days. The virions are then en-
riched by precipitation in polyethylene glycol. First, the
inoculated leaves are homogenized in 4ml 100mM
phosphate buffer, pH 7.2, and the homogenate is cen-
trifuged at 9000g for 1h. The supernatant is filtered
through Miracloth. PEG 6000 is added to 10% w/v and
NaCl to 0.34M. Virions are precipitated on ice over-
night and collected by centrifugation at 9000g for 1h.
The pellet is then resuspened in 0.5ml phosphate buffer
by vortex mixing in a cold room for 1h. Debris is re-
moved by centrifugation and the suspension is inocu-
lated by gentle rubbing into carborundum-dusted leaves
of young plants. We typically include a TRV construct
that silences the sulfur gene as a control in each exper-
iment. Sulfur is a magnesium chelatase required for
chlorophyll production, and the silenced plants turn
We are grateful to the Gatsby Charitable Foundation
for support of the Sainsbury Laboratory and the
BBSRC for a grant to develop TRV vectors. Use of
genetically modified virus vectors is licensed from the
Department of the Environment, Food and Rural Af-
fairs (PHL 161/4080). Many colleagues in the Sainsbury
Laboratory have contributed over the years to the de-
velopment of virus vectors and gene silencing technol-
ogy and we are indebted for their input and ideas.
 F. Ratcliff, S. MacFarlane, D.C. Baulcombe, Plant Cell 11 (1999)
 P.D. Zamore, Nat. Struct. Biol. 8 (2001) 746–750.
 O. Voinnet, Trends Genet. 17 (2001) 449–459.
 G. Brigneti, O. Voinnet, W.X. Li, L.H. Ji, S.W. Ding, D.C.
Baulcombe, EMBO J. 17 (1998) 6739–6746.
 O. Voinnet, Y.M. Pinto, D.C. Baulcombe, Proc. Natl. Acad. Sci.
USA 96 (1999) 14147–14152.
 M.H. Kumagai, J. Donson, G. Della-Cioppa, D. Harvey, K.
Hanley, L.K. Grill, Proc. Natl. Acad. Sci. USA 92 (1995) 1679–
 M.T. Ruiz, O. Voinnet, D.C. Baulcombe, Plant Cell 10 (1998)
 S. Kjemtrup, K.S. Sampson, C.G. Peele, L.V. Nguyen, M.A.
Conkling, W.F. Thompson, D. Robertson, Plant J. 14 (1998)
 R.A. Burton, D.M. Gibeaut, A. Bacic, K. Findlay, K. Roberts, A.
Hamilton, D.C. Baulcombe, G.B. Fincher, Plant Cell 12 (2000)
 J.R. Peart, G. Cook, B.J. Feys, J.E. Parker, D.C. Baulcombe,
Plant J. 29 (2002) 569–579.
 Y. Liu, M. Schiff, G. Serino, X.W. Deng, S.P. Dinesh-Kumar,
Plant Cell 14 (2002) 1483–1496.
 Y.L. Liu, M. Schiff, R. Marathe, S.P. Dinesh-Kumar, Plant J. 30
 D.H. Slaymaker, D.A. Navarre, D. Clark, O. del Pozo, G.B.
Martin, D.F. Klessig, Proc. Natl. Acad. Sci. USA 99 (2002)
 H.L. Jin, M.J. Axtell, D. Dahlbeck, O. Ekwenna, S.Q. Zhang, B.
Staskawicz, B. Baker, Dev. Cell 3 (2002) 291–297.
 J.R. Peart, R. Lu, A. Sadanandom, I. Malcuit, P. Moffett, D.C.
Brice, L. Schauser, D.A.W. Jaggard, S. Xiao, M. Coleman, J.M.
Dow, J.D.G. Jones, K. Shirasu, D.C. Baulcombe, Proc. Natl.
Acad. Sci. USA 99 (2002) 10865–10869.
 F. Ratcliff, A.M. Martin-Hernandez, D.C. Baulcombe, Plant J. 25
 C. Peele, C.V. Jordan, N. Muangsan, M. Turnage, E. Egelkrout,
P. Eagle, L. Hanley-Bowdoin, D. Robertson, Plant J. 27 (2001)
 S. Holzberg, P. Brosio, C. Gross, G.P. Pogue, Plant J. 30 (2002)
R. Lu et al. / Methods 30 (2003) 296–303
 T. Dalmay, A.J. Hamilton, S. Rudd, S. Angell, D.C. Baulcombe,
Cell 101 (2000) 543–553.
 E. Waigmann, M.H. Chen, R. Bachmaier, S. Ghoshroy, V.
Citovsky, EMBO J. 19 (2000) 4875–4884.
 R.E.F. Matthews, Plant Virology, Academic Press, San Diego,
 O. Voinnet, D.C. Baulcombe, Nature 389 (1997) 553.
 J.-C. Palauqui, T. Elmayan, J.-M. Pollien, H. Vaucheret, EMBO
J. 16 (1997) 4738–4745.
 N.A. Smith, S.P. Singh, M.B. Wang, P.A. Stoutjesdijk, A.G.
Green, P.M. Waterhouse, Nature 407 (2000) 319–320.
 R.L. Toth, G.P. Pogue, S. Chapman, Plant J. 30 (2002) 593–600.
 L. Jones, F. Ratcliff, D.C. Baulcombe, Curr. Biol. 11 (2001)
 T. Sijen, I. Vijn, A. Rebocho, R. van Blokland, D. Roelofs,
J.N.M. Mol, J.M. Kooter, Curr. Biol. 11 (2001) 436–440.
 C.L. Thomas, L. Jones, D.C. Baulcombe, A.J. Maule, Plant J. 25
 M.A. Turnage, N. Muangsan, C.G. Peele, D. Robertson, Plant J.
30 (2002) 107–114.
 M. Janda, R. French, P. Ahlquist, Virology 158 (1987) 259–262.
 S.N. Chapman, T.A. Kavanagh, D.C. Baulcombe, Plant J. 2
 J. Donson, C.M. Kearney, M.E. Hilf, W.O. Dawson, Proc. Natl.
Acad. Sci. USA 88 (1991) 7204–7208.
 T.H. Turpen, A.M. Turpen, N. Weinzettl, M.H. Kumagai, W.O.
Dawson, J. Virol. Methods 42 (1993) 227–240.
 C.D. Novina, M.F. Murray, D.M. Dykxhoorn, P.J. Beresford, J.
Riess, S.-K. Lee, R.G. Collman, J. Lieberman, P. Shankar, P.A.
Sharp, Nat. Med. 8 (2002) 681–686.
 H. Li, W.X. Li, S.W. Ding, Science (2002) 1319–1321.
 R.P. Hellens, E.A. Edwards, N.R. Leyland, S. Bean, P.M.
Mullineaux, Plant Mol. Biol. 42 (2000) 819–832.
R. Lu et al. / Methods 30 (2003) 296–303