Plant Cell Physiol. 49(2): 242–250 (2008)
doi:10.1093/pcp/pcm181, available FREE online at www.pcp.oxfordjournals.org
ß The Author 2008. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists.
All rights reserved. For permissions, please email: [email protected]
Efficient and High-Throughput Vector Construction and
Agrobacterium-Mediated Transformation of Arabidopsis thaliana
Suspension-Cultured Cells for Functional Genomics
Yoichi Ogawa 1, 4, Tomoko Dansako 1, 4, Kentaro Yano 1, Nozomu Sakurai 1, Hideyuki Suzuki 1, Koh Aoki 1,
Masaaki Noji 2, Kazuki Saito 1, 3 and Daisuke Shibata 1, *
1
Kazusa DNA Research Institute, 2-6-7 Kazusa-Kamatari, Kisarazu, Chiba, 292-0818 Japan
Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima, 770-8514 Japan
3
Graduate School of Pharmaceutical Sciences, Chiba University, 1-33 Yayoi-cho, Inage, Chiba, 263-8522 Japan
Editor-in-Chief’s choice
2
of genes in the coordinated process. Large-scale T-DNA
insertion lines have been produced as genetic resources
for loss-of-function mutation analyses. Efforts have been
devoted to locating the position(s) of the T-DNA within
individual genes in these lines, and such defined tagged lines
have been contributing significantly to gene discovery in
Arabidopsis research (Alonso et al. 2003). However, when
the gene of interest has functionally redundant paralogous
genes in the genome, no phenotype is expected in the
loss-of-function lines. Such cases are often seen, as multigene
families are prevalent in the Arabidopsis genome
(Arabidopsis Genome Initiative 2000). In these cases,
gain-of-function mutant lines have complementary roles
for functional analysis of these genes. As resources for
gain-of-function mutations, transgenic plants carrying
random insertion(s) of a strong promoter sequence on
chromosomes, called activation tagging lines, have been
generated at large scales (for a review, see Tani et al. 2004).
However, the use of these lines is limited to a certain extent as
the position of the promoter insertion on chromosomes has
not been located in most lines. Alternatively, large sets of
full-length cDNA clones have become available from several
plants such as Arabidopsis (Seki et al. 2002), rice (Kikuchi
et al. 2003), poplar (Nanjo et al. 2004) and tomato (Tsugane
et al. 2005). These full-length cDNA clones made it feasible
to generate gain-of-function transgenic plants which carry
the cDNAs driven by a strong promoter, using ordinary
vector construction and transformation protocols.
Plant suspension-cultured cells, maintained in sterile
medium under artificial environments, are suitable for gene
function analysis as well as for physiological study. These
cells are advantageous for collecting large amounts of
uniform cell samples that are grown under strictly
controlled conditions, such as of temperature and biotic/
abiotic stress treatment, facilitating obtaining reproducible
results. The Nicotiana tabacum BY-2 cell line (Nagata et al.
1992) is one of the most used cell lines for various types of
such research. However, genomic information about the
We established a large-scale, high-throughput protocol to
construct Arabidopsis thaliana suspension-cultured cell lines,
each of which carries a single transgene, using Agrobacteriummediated transformation. We took advantage of RIKEN
Arabidopsis full-length (RAFL) cDNA clones and the
Gateway cloning system for high-throughput preparation of
binary vectors carrying individual full-length cDNA sequences.
Throughout all cloning steps, multiple-well plates were used
to treat 96 samples simultaneously in a high-throughput
manner. The optimal conditions for Agrobacterium-mediated
transformation of 96 independent binary vector constructs
were established to obtain transgenic cell lines efficiently.
We evaluated the protocol by generating transgenic
Arabidopsis T87 cell lines carrying individual 96 metabolismrelated RAFL cDNA fragments, and showed that the protocol
was useful for high-throughput and large-scale production
of gain-of-function lines for functional genomics.
Keywords: Agrobacterium-mediated transformation —
Arabidopsis thaliana — Functional genomics — Gateway
cloning system — High-throughput — Suspension-cultured
cells.
Abbreviations: CaMV, cauliflower mosaic virus; GST,
glutathione S-transferase; GUS, b-glucuronidase; Hm, hygromycin; Km, kanamycin; MEPM, meropenem; NAA, 1-naphthaleneacetic acid; RAFL cDNA, RIKEN Arabidopsis full-length cDNA;
SATase, serine O-acetyltransferase.
Introduction
Transgenic approaches are crucial for understanding
gene function in molecular biological research. Recent
advances in DNA array technology allow researchers to
find a set of genes that function coordinately in the biological
process of interest (Gachon et al. 2005). Therefore, there is
increasing demand for the analysis of multiple mutant lines
showing a loss or gain of function in order to clarify the roles
4
These authors contributed equally to this work.
*Corresponding author: E-mail, [email protected]; Fax, þ81-438–52-3948.
242
Transgenic Arabidopsis cultured cell lines
species is still limited only to expressed sequence tags (ESTs)
(Matsuoka et al. 2004). Arabidopsis is a model plant for a
wide range of plant research because of its unique
characteristics, i.e. small plant size, short generation time
and small genome size (Meinke et al. 1998), and its genome
has been completely sequenced (Arabidopsis Genome
Initiative 2000). Among several established Arabidopsis
cell lines, the suspension-cultured cell line T87 is unique in
photosynthetic ability under light irradiation (Axelos et al.
1992). Therefore, T87 cells have been widely used for
genetic analyses (e.g. Callard et al. 1996, Uno et al. 2000,
Takahashi et al. 2001, Stolc et al. 2005). Thus, the cells are
suitable as a host for generating transgenic lines. However,
a high-throughput protocol for large-scale production of
transgenic cultured cell lines has not been established yet.
In this study we established a high-throughput
protocol for vector construction and transformation of
Arabidopsis suspension-cultured T87 cells. As a gene source,
we chose cDNA clones from the RIKEN Arabidopsis fulllength (RAFL) cDNA clone set (Seki et al. 2002). We
designed the high-throughput procedure for introduction of
RAFL cDNA fragments into binary vectors. The condition
for Agrobacterium-mediated transformation with individual
96 binary vector constructs into Arabidopsis T87 cells was
optimized for high-throughput processing. We validated
the protocol by generating transgenic cell lines that overexpress each of the 96 full-length cDNA clones encoding
metabolism-related genes. We discuss the usefulness of the
protocol for large-scale production of individual gain-offunction lines for functional genomics.
Results
High-throughput vector construction
To prepare binary vector constructs for individual
genes in a high-throughput manner, we took advantage of
RAFL cDNA clones (Seki et al. 2002) and the Gateway
cloning system (Walhout et al. 2000) (Fig. 1). As RAFL
cDNA fragments were cloned between distinct SfiI recognition sequences, 50 -GGCCAAATCGGCC-30 and 50 -GGCC
CTTATGGCC-30 , in vectors (Seki et al. 1998) and the
recognition sequences are rarely found in the Arabidopsis
genome, SfiI digestion excises, in most cases, single cDNA
fragments. Even if SfiI recognition sites exist in the middle
of the cDNA, proper cloning into the entry vector is
expected in most of the cases, although such cases are rare
(see below and Discussion). Once a target DNA fragment is
cloned in the entry vector of the Gateway system, the
fragment can be transferred to destination vectors via
recombination in a single step. Therefore, we designed a
new entry vector, named pRAFLENTR, which carries the
same SfiI recognition sites as in RAFL cDNA vectors
between the recombination site sequences attL1 and attL2
243
Construction of vector
Cloning of RAFL cDNA
1. Sfil digestion of 96 RAFL cDNA plasmids
2. Ligation of Sfil-digested RAFL cDNA and
pRAFLENTR
3. Transformation of E. coli with cDNA-carrying
pRAFLENTR
4. Extraction of pRAFLENTR
Preparation of binary vector
1. LR recombination of pRAFLENTR and pGWB2
2. Transformation of E. coli with cDNA-carrying
pGWB2
3. Extraction of pGWB2
Preparation of Agrobacterium
1. Transformation of Agrobacterium by freeze-thaw
2. Storage at −80°C as glycerol stock
Transformation of T87 cells
Day 1: Pre-culture of T87 cells
1. Suspending of T87 cells with B5 medium + 1µM
NAA at 0.05 g wet weight per 10 ml
2. Culturing at 22°C under light with shaking for 2 d
Day 2: Pre-culture of Agrobacterium
1. Culturing of Agrobacterium in LB medium +
selective agent at 28°C with shaking for overnight
Day 3: Inoculation and co-cultivation
1. Adding of 1/1000 vol. of Agrobacterium culture
2. Culturing at 22°C under light with shaking for 2 d
Day 5: Post-co-cultivation
1. Collection of T87 cells by centrifugation
2. Washing of cells with 10 ml of mJPL3 medium
for 30 sec with vigorous shaking
3. Collection of cells by centrifugation
5. Resuspending of cells with 3 ml of mJPL3 medium
6. Spreading of cell suspension over two selection
media (mJPL3 + 25 mg l−1 MEPM + 20 mg l−1
Hm + 3 g l−1 gellan gum)
7. Culturing at 22°C under light for ≥ 2 w
2–3 weeks: Culture of transgenic T87 cells
1. Transferring of resistant calli onto fresh selection
medium.
2. Subculture of calli every 3 to 4 w
3. Re-establishment of suspension culture by
transferring 0.5 g wet weight of calli into 100 ml of
mJPL3 medium (+ 12.5 mg l−1 MEPM + 5 mg l−1
Hm)
Fig. 1 The scheme for high-throughput construction of vector and
transformation of Arabidopsis T87 cells.
Transgenic Arabidopsis cultured cell lines
High-throughput transformation for Arabidopsis T87 cells
We
established
a
procedure
for
efficient
Agrobacterium-mediated transformation of Arabidopsis
suspension-cultured T87 cells (Fig. 1). First, we replaced
the micronutrient, vitamins and buffer of JPL medium
(Jouannean and Péaud-Lenoël 1967), which was established
by modifying MS medium (Murashige and Skoog 1962) and
has been used for maintaining the T87 cell line (Axelos et al.
1992), by one-third strength MS micronutrient, full-strength
MS vitamins and 2-(N-morpholino)ethanesulfonic acid
(MES), respectively. The modification allows easy preparation of the medium, named mJPL3, by using commercially
available pre-mixed stock chemicals without altering the
major properties of T87 cells, such as fast growth and
uniform small cell clumps. We looked for suitable
transformation conditions for T87 cells cultured in mJPL3
medium. We inoculated Agrobacterium tumefaciens
EHA101 carrying the binary vector pIG121-Hm [carrying
uidA under the cauliflower mosaic virus (CaMV) 35S
promoter] into T87 cells at early- to mid-log phases (1–5 d
of pre-culture), which are generally known to be suitable for
efficient transformation (An 1985). However, no transgenic
calli were obtained under the conditions tested (Fig. 2).
Alternatively, we cultured the cells in B5 medium
(Gamborg et al. 1968) supplemented with 1 mM 1-naphthaleneacetic acid (NAA) before Agrobacterium inoculation
(pre-culture) and following 2 d co-cultivation. Hygromycinresistant (Hmr) calli were obtained reproducibly 2–3 weeks
after Agrobacterium inoculation from cells pre-cultured in
B5 medium for 1–3 d (Fig. 2). Interestingly, the cells
A
Pre-culture medium
B5
mJPL3
(see Materials and Methods). As a destination vector, we
used a Gateway-based binary vector, pGWB2 (a gift from
Dr. Tsuyoshi Nakagawa of Shimane University), which can
be used for Agrobacterium-mediated plant transformation
with hygromycin selection of transgenic cells.
Throughout the process of vector construction, we
used multiwell plates for enzymatic reactions and DNA
preparations as described in Materials and Methods. The
protocol established allows one person to produce 96
Agrobacterium strains that carry individual binary vector
constructs within 2 weeks, when the person works 5 h per
day, 5 d a week.
To validate whether the protocol is still useful for the
RAFL clones that carry an internal SfiI site, we applied the
same protocol to such a RAFL cDNA clone, pda07581
(At5g58778). We found that the cDNA was separated into
two fragments upon SfiI digestion. However, as the 30
overhang sequence of the internal SfiI site differed from
those of the SfiI site of the vector, the fragments ligated in
the intact cDNA form when inserted into the binary vector.
The efficiency was equivalent to that for cDNAs that do not
have an internal SfiI site.
1
2
3
4
Pre-culture period (d)
5
B 600
mJPL3
B5
500
No. of Hmr calli
244
400
300
200
100
0
1
2
3
4
Pre-culture period (d)
5
Fig. 2 Effects of the medium and period of pre-culture on
transformation of Arabidopsis T87 cells. Arabidopsis T87 cells
(0.05 g wet weight) were suspended in 10 ml of mJPL3 medium
or B5 medium supplemented with 1 mM NAA, and cultured for
1–5 d. An overnight culture of A. tumefaciens EHA101
(pIG121-Hm) was inoculated into cell suspensions and co-cultured
for 2 d. The inoculated T87 cells were plated onto selection
medium after washing with mJPL3 medium and then cultured
for 2 weeks. Pictures are shown in (A) and the numbers of Hmr
calli obtained are shown in (B) as the mean SD of three
independent experiments.
pre-cultured for 2 d exhibited the highest transformation
efficiency (500 Hmr calli per 10 ml of culture) among the
culture conditions tested, whereas 1 and 3 d pre-culture
yielded only 510 and 200 Hmr calli, respectively. In the
1–4 d co-culture period tested, co-cultivation for 2 d in
combination with 2 d pre-culture was optimal to obtain the
highest transformation efficiency (data not shown).
Expression of uidA, transferred from the binary vector
pIG121-Hm, in the transgenic calli was confirmed by
5-bromo-4-chloro-3-indolyl-b-D-glucuronic acid (X-gluc)
staining (data not shown).
For high-throughput processing, we simplified the step
of cell washing after Agrobacterium co-cultivation, which is
laborious and time-consuming in transformation procedures. Most transformation protocols include three or more
cell washings to suppress overgrowth of Agrobacterium
during subsequent culture periods. In fact, overgrowth
of Agrobacterium, which leads to plant cell death,
was occasionally observed in our experiments, when
the inoculated T87 cells were washed only once and
Transgenic Arabidopsis cultured cell lines
600
3- to 4-week intervals of subculture. To re-establish the
suspension culture for large-scale preparation of transgenic
cells, 0.5 g wet weight of 4-week-old calli were suspended in
100 ml of liquid mJPL3 medium and subcultured at
biweekly intervals as described in Materials and Methods.
No. of Hmr calli
500
400
300
200
100
0
245
0
1
2
3
No. of cell washing after co-cultivation
Fig. 3 Effects of washing of Arabidopsis T87 cells after
co-cultivation with Agrobacterium on transformation efficiency.
Arabidopsis T87 cells (0.05 g wet weight) were pre-cultured in
10 ml of B5 medium supplemented with 1 mM NAA for 2 d,
inoculated with A. tumefaciens EHA101 (pIG121-Hm) and then
co-cultured for 2 d. T87 cells were washed 0–3 times with mJPL3
medium, plated on selection medium, and then cultured for
2 weeks. The numbers of Hmr calli obtained are shown as the
mean SD of three independent experiments.
carbenicillin or timentin was used for elimination of
Agrobacterium during culture. Such overgrowth was
seen even if a high concentration of the antibiotic
(50–200 mg l1) was used. Thus, we replaced the antibiotic
by meropenem (MEPM) at 25 mg l1, which possesses high
antibacterial activity and low phytotoxicity (Ogawa and Mii
2007). By using MEPM, a single cell washing was enough to
obtain transgenic cells without reducing the transformation
efficiency (Fig. 3), which resulted in saving of labor and
time in the whole process. Further simplification could be
possible by omitting the washing process. We tested
transformation without cell washing after co-cultivation
and found that 150 transgenic Hmr calli per 10 ml of culture
were still obtained (Fig. 3), which would be enough for most
experimental purposes.
In addition to simplification of the washing process, we
simplified the step of Agrobacterium inoculation and
co-cultivation. Most transformation protocols have used
Agrobacterium cells after washing and resuspending with
fresh co-cultivation medium (e.g. supplementation with
acetosyringone). However, our experiments showed that
direct addition of overnight-grown stationary phase
Agrobacterium to pre-cultured T87 cells followed by further
co-cultivation produced as many transgenic calli as the
conventional protocols. Concentration of Agrobacterium
cells in ratios from 1/100 to 1/1,000 in T87 suspension
culture only slightly affected the transformation efficiency
(data not shown).
The transgenic T87 calli obtained by the protocol were
transferred onto a semi-solidified mJPL3 plate containing
the antibiotics MEPM (for elimination of Agrobacterium)
and Hm (for selection of plant cells), and maintained by
Evaluation of the high-throughput protocol
We evaluated the high-throughput protocol conditions
in this study using 96 metabolism-related RAFL cDNA
clones (Supplementary Table S1). The 96 clones were
processed simultaneously and the fragments were introduced into the binary vector pGWB2 downstream of the
CaMV 35S promoter. The processes from vector construction to transformation of Agrobacterium were independently repeated three times using the same starting
materials. All plasmid constructs produced were analyzed
by restriction enzyme digestion to check for proper cloning
of the fragments in the vector. Except for pda08414
(At4g35300, sugar transporter-like protein), all clones
produced proper constructs at least once in three time
trials (Supplementary Table S1). The pda08414 clone was
cloned successfully in another set of trials. The results
suggest that bulk treatment using multiwell plates causes
cloning failure of some clones in a single trial, but most can
be cloned in other rounds of trials, unless the cloned
constructs are incompatible with Escherichia coli. Actually,
in our trials of 4800 RAFL clones, the clones that were not
successfully processed in the first round were included in the
next round of trials and most of them (499%) were
successfully processed in the second- or third-round trials
(unpublished data). All of the binary vectors carrying the
RAFL cDNA fragments were successfully introduced into
Agrobacterium in single trials.
We prepared 96 transformation-ready Agrobacterium
strains carrying individual binary vectors, and the strains
were divided into four sets, each of which contains 24
strains. Each set of Agrobacterium was independently
subjected to transformation trials of Arabidopsis T87 cells.
In these trials, 94 Agrobacterium strains (97.9%) gave
4100 Hmr calli. Although the remaining two strains gave
520 Hmr calli, these two strains also produced 4100 Hmr
calli in the second or third round of trials. The current
protocol allows one person (3 h maximum a day) to treat 24
different Agrobacterium strains for inoculation to T87 cells
within a week (four working days), or one person
(6 h maximum a day) to treat 96 strains within 2 weeks,
following the 2 weeks of the selection period. Usually, we
maintained 10–20 independent transgenic calli for further
analyses, as the expression levels of transgenes varied
among transgenic lines (described below). Practically,
although some Agrobacterium strains did not produce
sufficient numbers of calli, the simple and robust transformation protocol established in this study allow us to re-try
246
Transgenic Arabidopsis cultured cell lines
WT
A
WT
1
2
3
4
5
6
7
8
9
10
kDa
1
2
SATase
3
4
1
2
3
4
5
6
7
8
9
37
11
B
12
WT
9
13
1
14
2
10
15
3
11
12
16
4
17
5
13
18
6
19
7
20
8
14
25
Fig. 5 Western blot analysis of total soluble protein extracted
from transgenic Arabidopsis T87 cell lines. Cell lines that were
transformed with a serine O-acetyltransferase (SATase; At5g56760,
pda00783) under a CaMV 35S promoter were analyzed. Wild-type
T87 cells (WT) were also included. Numbers indicate independent
cell lines. Arrowheads show signals of 34 kDa SATase.
wild-type T87 cells (Fig. 5). These results show that the
protocol employed for transformation is suitable for
production of large-scale gain-of-function lines.
Discussion
Fig. 4 Northern blot analysis of total RNA extracted from
transgenic Arabidopsis T87 cell lines. Cell lines that were transformed with either (A) a glutathione S-transferase (GST; At1g02920,
pda04245) or (B) a serine O-acetyltransferase (SATase; At5g56760,
pda00783), both under a CaMV 35S promoter, were analyzed.
Wild-type T87 cells (WT) were also included. Different numbers
given in the figure indicate independent cell lines.
transforming those strains in the next round of our
transformation routine, and we could finally obtain
sufficient numbers of calli for most of the strains.
We examined the expression levels of transgenes in
transgenic cells by Northern blot analysis. A total of 584
transgenic T87 lines, which contained 31 randomly selected
vector constructs and 14–20 transgenic T87 lines generated
from each construct, were subjected to expression analysis.
Fig. 4A and 4B shows examples of two of 31 transgenes,
glutathione S-transferase (GST; At1g02920, pda04245) and
serine O-acetyltransferase (SATase; At5g56760, pda00783),
respectively. Endogenous expression levels of the 31 genes
examined were not high or were illegible on Northern blots.
When compared with the endogenous expression level, 526
(90.1%) of 584 lines tested exhibited largely increased
expression of the transgenes. In some cases, the expression
levels in transgenic cells were apparently lower than the
endogenous level, suggesting the occurrence of posttranscriptional gene silencing, which is often seen in
transgenic plants when the transgene is driven by a strong
constitutive promoter such as CaMV 35S (Jorgensen 2003).
We further examined protein expression levels of SATase
transgenic lines. Nine independent transgenic lines showing
high transgene expression levels were selected for Western
blot analysis. Signals indicating 34 kDa SATase proteins
were detected in all nine transgenic lines examined, and
these signals were apparently stronger than signals in
Here we established an efficient high-throughput
protocol for production of vector constructs and subsequent Agrobacterium-mediated transformation of
Arabidopsis suspension-cultured T87 cells with the constructs. We took advantage of RAFL cDNA clones, the
Gateway cloning system and multiple-well plates for highthroughput binary vector construction. The protocol
provides a means for producing a large-scale population
of transgenic cell lines each of which carries a single
transgene controlled under the proper promoter. The
cDNA clones provided by RIKEN BioResource Center
comprise 15,295 clones that cover 460% of expressed
genes, which belong to diverse gene families (Seki et al.
2002). The number of RAFL clones is still increasing
(20,997 clones at October 2007). Therefore, the RAFL
cDNA clone collection is useful for gene function analyses
in various research areas. The cell lines that overexpress
RAFL cDNA fragments controlled under a strong promoter serve as gain-of-function lines. As the Gateway cloning
system is versatile, once pRAFLENTR vector constructs
with RAFL fragments are prepared, they can be transferred
into various destination vectors other than pGWB2, such as
for gene silencing experiments, in a single step. The protocol
described in this report allows laboratories that are
equipped with ordinary molecular biology instruments to
produce several hundred transgenic cell lines within a few
months by a single person. Thus, the transgenic cell lines
produced using the protocol will contribute to accelerating
Arabidopsis functional genomics studies. Moreover, the
high-throughput vector construction protocol can also be
applied to full-length cDNA clones of other plant species
such as rice (Kikuchi et al. 2003) and tomato (Tsugane et al.
2005), as some (rice) or all (tomato) cDNA fragments are
cloned in the SfiI site.
Transgenic Arabidopsis cultured cell lines
Transferring full-length cDNA fragments into the
Gateway entry vector pRAFLENTR from RAFL clones
using the restriction enzyme SfiI would be less problematic,
even if the fragments contain multiple SfiI sites. If a single
SfiI site exists inside a RAFL clone and SfiI digestion
produces 3-nt long 30 overhang sequences that differ from
those of the SfiI sites of pRAFLENTR, such as pda07581
(At5g58778), the two divided fragments can be cloned
properly into pRAFLENTR with enough efficiency, as
shown in the case of pda07581. If the overhang sequence
produced is the same as one of the overhang sequences of
the SfiI sites of pRAFLENTR, the cloning step produces
two possible clones, which carry the fragment in the proper
or reverse orientation. Restriction fragment analysis of such
clones may determine which one is the proper one. Our
search for SfiI sites in 15,295 RAFL clones revealed that
71 clones had single SfiI sites, but no clones had more than
two sites. In these clones, all of the 3-nt long 30 overhang
sequences differed from those of the SfiI sites of
pRAFLENTR. Therefore, it is expected that these RAFL
clones can be cloned properly into pRAFLENTR with our
protocol. Another search showed that the frequency of the
occurrence of SfiI sites in all Arabidopsis expressed
sequences (0.3 105) is much less than that expected in
random nucleotide sequences (1.5 105). Thus, it would
be very rare to encounter cDNA clones that have more than
two SfiI sites, and it is still advantageous to use SfiI sites as
the cloning sites.
Gallego et al. (1999) have reported Agrobacteriummediated transformation of T87 cells. However, the
transformation protocol would need laborious and timeconsuming manipulations when large-scale production of
transgenic cell lines is required, and the protocol uses
B5-based AT medium for culture of the T87 cells, instead of
JPL medium which has been used as the original medium
for the T87 cells (Axelos et al. 1992). We developed
modified JPL medium, named mJPL3 medium, which is
easier to prepare than JPL medium. The choice of culture
medium for T87 cells, B5 medium or mJPL3 medium,
affected the conditions and appearance of suspensioncultured cells. It has been shown that transformation of
T87 cells was successful when B5-based medium was used
for routine subculture and transformation steps (Gallego
et al. 1999), while we failed to obtain transgenic cells when
mJPL3 medium was used throughout the culture including
the pre-culture and co-culture periods (Fig. 2). On the other
hand, B5 medium tended to induce large cell clumps with a
prolonged culture period, which are not suitable for
re-establishment of suspension culture (see Fig. 2A), while
mJPL3 maintained uniform and small cell clumps. These
results implied that B5 medium is suitable for cells to enter
into the transformation-competent state, while mJPL3
medium is suitable for keeping cells in a less aggregated
247
state even with a prolonged culture period. Therefore, we
used mJPL3 medium for subculture and re-establishment of
suspension culture, and B5 medium for pre-culture and
co-cultivation. Our transformation protocol could also be
utilized successfully in other laboratories for T87 cells
cultured in the original JPL medium (Dr. S. Takahashi,
Tohoku University, Dr. D. Ohta, Osaka Prefectural
University, and Dr. Kanamaru, Kobe University, personal
communication).
Transgenic suspension-cultured cells have advantages
over transgenic plants. First, transgenic cells are obtained
within several weeks after Agrobacterium inoculation,
while at least several months are needed to obtain
transgenic plants. Secondly, uniform cells with the same
genetic background can be obtained readily from transgenic
cultured cells at a large scale. In contrast, it takes several
generations to obtain Arabidopsis plants with homozygous
transgenes. Thirdly, strict control of the environment
of cultured cells is possible, while control of the environment of plants is not easy even in a controlled growth
chamber. Our statistical analyses of the Arabidopsis
transcriptome detected by microarrays showed higher
reproducibility in the data obtained from three biological
replicates of Arabidopsis suspension-cultured T87 cells
than those from 4-week-old Arabidopsis plants grown
in a well-controlled growth chamber (N.S., H.S. and
D.S., unpublished data). However, transgenic suspensioncultured cells have a couple of disadvantages. First,
maintenance of a large number of transgenic cell lines is
laborious, while storage of seeds derived from transgenic
plants is easy. Thus, we recently developed a protocol
for cryopreservation of transgenic T87 cells (unpublished
data). Secondly, cultured cells are not applicable to
phenomena specific to differentiated tissues and organs.
Taken together, transgenic cultured cell lines and plant lines
will play complementary roles in accelerating functional
genomics.
Recently, a system for random introduction of RAFL
cDNA fragments driven by a strong promoter into plants,
called the FOX hunting system, was developed (Ichikawa
et al. 2006). In this system, a pool of RAFL cDNA
plasmids was subjected to SfiI digestion and subsequently
to ligation with a binary vector. The binary vector DNAs
were pooled and introduced into Agrobacterium. Using
pooled Agrobacterium strains and in planta Agrobacteriummediated transformation, many transgenic Arabidopsis
plants were generated. Several dominant phenotypes were
seen in the population, indicating that the population serves
as a gain-of-function mutant resource. As the chromosomal
locations of transgenes in the population at a large scale
have not been determined yet, genes are identified solely
from the lines that exhibit phenotypes. Therefore, our
approach to generate gain-of-function lines for individual
248
Transgenic Arabidopsis cultured cell lines
targeted genes is different and complementary to the FOX
hunting system.
Using the protocol established in this study, we are
currently working on generating a large population of gainof-function T87 cell lines with 41,000 individual genes for
metabolism, which will contribute to plant metabolome
research. Our transformation protocol for Arabidopsis T87
cells has been successfully used by our collaborators
to elucidate the functions of uncharacterized genes (Hirai
et al. 2007).
Materials and Methods
DNA materials, vectors and bacterial strains
All full-length cDNA clones of Arabidopsis used in this study
(RAFL cDNA clones) were obtained from RIKEN BioResource
Center (Tsukuba, Japan). The list of the RAFL cDNA clones is
shown in Supplementary Table S1.
We modified the Gateway donor vector pDONR201
(Invitrogen, Carlsbad, CA, USA) to accept SfiI fragments excised
from RAFL cDNA clones. From a RAFL cDNA clone, pda01738,
the entire cDNA fragment and SfiI recognition sequences flanking
both ends of the cDNA were amplified using a set of PCR primers,
RAFLB1 (50 -GGGGACAAGTTTGTACAAAAAAGCAGGCT
GGCCAAATCGGCCGAGCTCGAATTC-30 ) and RAFLB2
(50 -GGGGACCACTTTGTACAAGAAAGCTGGGTGGCCCT
TATGGCCGGATCCAAGAGC-30 ) (the SfiI recognition
sequences found in pda01738 are underlined). As the primers
contained the recombination sites attB1 and attB2 for Gateway
cloning, the PCR-amplified fragment was introduced into the
donor vector pDONR201 using BP clonase (Invitrogen) at the
cloning sites via recombination according to the supplier’s
protocol. The resultant vector, which carried the two SfiI sites
between the recombination sites attL1 and attL2, was digested
with the restriction enzyme SfiI. The vector fragment with the
SfiI recognition sequences at both ends, named pRAFLENTR,
was purified by gel electrophoresis, and used to clone the RAFL
cDNA fragments in this study.
The Gateway-based binary vector pGWB2 was obtained
from Dr. Tsuyoshi Nakagawa (Shimane University, Japan).
Disarmed A. tumefaciens strain EHA101 (Hood et al. 1986)
was used in the present study. A binary vector pIG121-Hm (Ohta
et al. 1990; a gift from Dr. Kenzo Nakamura of Nagoya
University), which has two selective marker genes, hpt and nptII,
and a reporter gene, uidA, was used for experiments to establish the
transformation protocol. For large-scale production of transgenic
cell lines, binary vectors carrying individual genes were constructed
as described below. Agrobacterium was cultured in LB medium
supplemented with 50 mg l1 kanamycin (Km) and 25 mg l1 Hm
at 288C with shaking overnight before inoculation into Arabidopsis
T87 cells.
High-throughput vector construction
A set of 96 RAFL cDNA plasmids (80 ng per 17.5 ml of
distilled water) was mixed individually with 2.5 ml of buffer and 5 U
of the restriction enzyme SfiI (TAKARA SHUZO CO. LTD,
Ohtsu, Japan) in a 96-well PCR plate (Thermo-Fast 96 NonSkirted, ABgene, Epsom, UK). The plate was incubated in a
thermal cycler (ASTEC, Fukuoka, Japan) at 508C for 5 h for SfiI
digestion and then at 708C for 15 min to denature the enzyme.
A 2 ml aliquot of ligation mixture (TOYOBO, Osaka, Japan)
and 1 ml (10 ng) of SfiI-digested pRAFLENTR were mixed in a
new 96-well PCR plate, and the plate was incubated at 168C for
2 h. The contents of each well, which had been kept at 808C,
were placed on ice to thaw the frozen cells. The ligation mixtures
(3 ml each) were added and mixed into 50 ml of competent cells of
Escherichia coli DH5a (Inoue et al. 1990) in a 96-well PCR plate
placed on ice, and the plate was further incubated for 20 min. To
introduce DNA into E. coli by heat shock, the plate was heated at
428C in a thermal cycler for 1 min, and then cooled to 48C. The E.
coli mixtures were mixed individually with 600 ml of SOC medium
in a 96-deep-well plate (2.2 ml Storage Plate, ABgene) and cultured
at 378C for 20 min. E. coli cultures (30 ml each) were spread
individually over SOC agar containing 50 mg l1 Km in eight
12-well plates (3815-012, Asahi Techno Glass, Tokyo, Japan) and
kept at 378C overnight. Single colonies from the wells were
inoculated in 600 ml of 2 YT medium containing 50 mg l1 Km
and incubated at 378C for 16 h. Plasmid DNAs were prepared from
the cultured cells using CosMCPrep (Agencourt Bioscience,
Beverly, MA, USA) according to the manufacturer’s instructions,
and dissolved in 40 ml of distilled water in a 96-well plate. To check
if the inserts were properly inserted into the cloning site of
pRAFLENTR, 10 ml of the plasmid DNA solutions were digested
with the restriction enzyme BanII in 20 ml reaction solutions in a
96-well PCR plate at 378C for 1 h. The sizes of DNA fragments
were determined by agarose gel electrophoresis and were compared
with those expected from the cDNA sequences, which we
determined using an in-house Perl program to calculate the
fragment sizes from a set of cDNA sequences.
LR in vitro recombination was carried out using a Gateway
cloning system kit according to the instruction manual. Briefly,
plasmid DNA of 96 pRAFLENTRs (4 ml each) was mixed
individually with 2 ml of pGWB2, 2 ml of LR clonase and 2 ml of
5 buffer in a 96-well PCR plate and incubated at 208C for 2 h.
The reaction was stopped by adding 1 ml of proteinase K and
keeping it at 378C for 30 min. The reaction mixtures (1 ml each)
were used for transformation of E. coli DH5a. The procedures for
E. coli transformation and plasmid DNA preparation were as
described above, although both 50 mg l1 Km and 50 mg l1 Hm
were used as selective antibiotics. The insertion of RAFL cDNA
fragments in the vector was confirmed by checking the sizes of the
fragments after PCR according to the standard protocol, followed
by restriction enzyme digestion and gel electrophoresis as described
above.
Transformation of A. tumefaciens EHA101 was performed by
the freeze–thaw method of An et al. (1988), but with a modification
for multiwell plate processing. DNA solutions (3 ml each) were
mixed with the competent Agrobacterium cells (50 ml each)
dispensed into a 96-deep-well plate and the mixture was frozen in
liquid nitrogen. The plate was warmed at 378C for 1 min, 300 ml of
LB medium was added, and then the plate was incubated at 288C
with shaking for 1 h. The entire mixture was spread individually
over 96 LB agar plates containing 50 mg l1 Km and 50 mg l1 Hm,
and cultured at 288C for 2 d. Single colonies were transferred onto
fresh agar plates and further cultured at 288C for 2 d. The plates
were kept at 48C for up to 2 months. Single clones from each vector
construct were used for transformation of suspension-cultured
cells. Inserts in the binary vector were confirmed by checking the
sizes of the fragments by colony PCR, following restriction enzyme
digestion and gel electrophoresis as described above. The clones,
which were used for transformation of Arabidopsis cells, were
dispensed into 96-well-formatted tubes (BIOTUBE system T101,
Transgenic Arabidopsis cultured cell lines
Simport Plastics, Beloeil, Canada) and stored at 808C as glycerol
stocks for long-term storage.
Transformation of Arabidopsis T87 cells
Arabidopsis thaliana (L.) Heynh. ecotype Columbia suspension-cultured T87 cells (Axelos et al. 1992) were obtained from
RIKEN BioResource Center. T87 cells were maintained in mJPL3
medium [JPL macronutrient (stock A0 of Axelos et al. 1992), onethird strength MS micronutrient (M-0529, Sigma, St Louis, MO,
USA), MS vitamins (M-3900, Sigma), 0.1 g l1 casamino acids,
15 g l1 sucrose, 1 mM NAA and 1% (v/v) 250 mM MES (pH 5.9)]
under continuous illumination (100 mmol m2 s1) at 228C with
rotary shaking at 120 r.p.m. Two-week-old cells were sieved
through a 500 mm stainless mesh to collect fine cell clumps, and
0.5 g wet weight of cells were transferred to a 300 ml flask
containing 100 ml of fresh mJPL3 medium for subculture.
Two-week-old T87 cells were sieved through a 500 mm
stainless mesh, collected on a 50 mm nylon mesh and then
resuspended in either B5 medium (G-5893, Sigma) supplemented
with 1 mM NAA and 30 g l1 sucrose or mJPL3 medium at 0.5 g
wet weight per 100 ml of medium. A 10 ml aliquot of cell
suspension was transferred into a 50 ml flask and cultured under
continuous illumination at 228C with shaking at 120 r.p.m. for 1–
5 d. A 10 ml aliquot of overnight culture of Agrobacterium
(2 109 cfu ml1) harboring a RAFL cDNA-carrying pGWB2
was directly added to a pre-cultured T87 cell suspension, and they
were cultured for a further 1–3 d. After co-cultivation, the cell
suspension was transferred into a 15 ml tube and T87 cells were
collected by centrifugation at 200g for 1 min. Co-cultured T87
cells were washed with 10 ml of mJPL3 medium with shaking (60–
90 r.p.m.) for 30 s. The cells were collected by centrifugation,
resuspended with 3 ml of mJPL3 medium, and then spread over
two plates of selection medium, which is mJPL3 medium
supplemented with 25 mg l1 MEPM (Meropen, Dainippon
Sumitomo Pharma, Osaka, Japan), 20 mg l1 Hm and solidified
with 3 g l1 gellun gum (Phytagel, Sigma). After 2–3 weeks of
culture, Hmr calli formed on the selection medium were transferred
onto fresh selection medium using sterilized toothpicks, and
maintained by subculture every 3–4 weeks. Suspension cultures
of transformed calli were established by transferring cells into
mJPL3 medium containing 12.5 mg l1 MEPM and 5 mg l1 Hm at
a density of 0.5 g wet weight of cells per 100 ml of medium. They
were maintained by standard procedures as described above.
Histochemical b-glucuronidase (GUS) assay for T87 calli transformed with a binary vector pIG121-Hm was performed according
to Jefferson (1987).
RNA extraction and Northern blot analysis
RNA samples of transgenic and wild-type T87 calli were
extracted using an RNeasy Plant Mini kit (Qiagen, Hilden,
Germany). Probes were generated by PCR using pRAFLENTRs
carrying each cDNA as a template and oligonucleotide pairs
annealing to attL1 and attL2, 50 -ACAAGTTTGTACAAA
AAAGC-30 and 50 -ACCACTTTGTACAAGAAAGC-30 , respectively. Labeling of the probes and hybridization were performed
using the AlkPhos Direct Labeling and Detection System with
CDP-Star according to the instruction manual (GE Healthcare,
Buckinghamshire, UK).
Protein extraction and Western blot analysis
Protein samples of transgenic and wild-type T87 cells were
extracted in extraction buffer (250 mM potassium phosphate,
pH 8.0, 0.5 mM EDTA and 10 mM 2-mercaptoethanol). Extracted
249
proteins (10 mg each) were electrophoresed in a 12.5%
SDS–polyacrylamide gel and were stained with CBB Stain One
(Nacalai Tesque, Kyoto, Japan). For Western blot analysis, 30 mg
of protein samples were electrophoresed in a 12.5% SDS–
polyacrylamide gel and then transferred onto a PVDF membrane
(Immobilon-P, Millipore). Blots were hybridized with the rabbit
antiserum against SATase of watermelon (Saito et al. 1995) at a
1 : 300 dilution. The antibody was detected using an NBT/BCIP
color development substrate and a goat anti-rabbit secondary IgG
conjugated with alkaline phosphatase (Promega, Madison, WI,
USA).
Supplementary material
Supplementary material mentioned in the article is available
to online subscribers at the journal website www.pcp.
oxfordjournals.org.
Funding
New Energy and Industrial Technology Development
Organization (NEDO) as part of a project called ‘Development of Fundamental Technologies for Controlling the
Material Production Process of Plants’.
Acknowledgments
The authors thank T. Matsuura, M. Hasegawa, K. Mori,
K. Moriya, Y. Asami and M. Iwata for their technical assistance.
We thank RIKEN BioResource Center (Tsukuba, Japan) for
RAFL cDNA clones and Arabidopsis T87 cells, Dr. T. Nakagawa
(Shimane University) for his gift of pGWB2, and
Dr. K. Nakamura (Nagoya University) for his gift of pIG121-Hm.
References
Alonso, J.M., Stepanova, A.N., Leisse, T.J., Kim, C.J., Chen, H., et al.
(2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana.
Science 301: 653–657.
An, G. (1985) High efficiency transformation of cultured tobacco cells.
Plant Physiol. 79: 567–570.
An, G., Ebert, P.R., Mitra, A. and Ha, S.B. (1988) Binary vectors. In Plant
Molecular Biology Manual. Edited by Gelvin, S.B. and Shilperoort, R.A.
pp. 1–19. Kluwer Academic Publishers, Dordrecht.
Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of
the flowering plant Arabidopsis thaliana. Nature 408: 796–815.
Axelos, M., Curie, C., Mazzolini, L., Bardet, C. and Lescure, B. (1992)
A protocol for transient gene expression in Arabidopsis thaliana
protoplasts isolated from cell suspension culture. Plant Physiol.
Biochem. 30: 123–128.
Callard, D., Axelos, M. and Mazzolini, L. (1996) Novel molecular markers
for late phases of the growth cycle of Arabidopsis thaliana cell-suspension
cultures are expressed during organ senescence. Plant Physiol. 112:
705–715.
Gachon, C.M., Langlois-Meurinne, M., Henry, Y. and Saindrenan, P.
(2005) Transcriptional co-regulation of secondary metabolism enzymes in
Arabidopsis: functional and evolutionary implications. Plant Mol. Biol.
58: 229–245.
Gallego, M.E., Sirand-Pugnet, P. and White, C.I. (1999) Positive–negative
selection and T-DNA stability in Arabidopsis transformation. Plant Mol.
Biol. 39: 83–93.
Gamborg, O.L., Miller, R.A. and Ojima, K. (1968) Nutrient requirement of
suspension cultures of soybean root cells. Exp. Cell Res. 50: 151–158.
250
Transgenic Arabidopsis cultured cell lines
Hirai, M.Y., Sugiyama, K., Sawada, Y., Tohge, T., Obayashi, T., et al.
(2007) Omics-based identification of Arabidopsis Myb transcription
factors regulating aliphatic glucosinolate biosynthesis. Proc. Natl Acad.
Sci. USA 104: 6478–6483.
Hood, E.E., Helmer, G.L., Frakey, R.T. and Chilton, M.-D. (1986) The
hypervirulence of Agrobacterium tumefaciens A281 is encoded in a region
of pTiBo542 outside of T-DNA. J. Bacteriol. 168: 1291–1301.
Ichikawa, T., Nakazawa, M., Kawashima, M., Iizumi, H., Kuroda, H.,
et al. (2006) The FOX hunting system: an alternative gain-of-function
gene hunting technique. Plant J. 48: 974–85.
Inoue, H., Nojima, H. and Okayama, H. (1990) High efficiency
transformation of Escherichia coli with plasmids. Gene 96: 23–28.
Jefferson, R.A. (1987) Assaying chimeric genes in plants: the GUS gene
fusion system. Methods Enzymol. 5: 387–405.
Jorgensen, R.A. (2003) Sense cosuppression in plants: past, present, and
future. In RNAi: A Guide to Gene Silencing. Edited by Hannon, G.J.
pp. 5–21. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
NY.
Jouanneau, J.-P. and Péaud-Lenoël, C. (1967) Croissance et synthèse des
protéines de suspensions cellulaires de Tabac sensibles à al kinétine.
Physiol. Plant. 20: 834–850.
Kikuchi, S., Satoh, K., Nagata, T., Kawagashira, N., Doi, K., et al. (2003)
Collection, mapping, and annotation of over 28,000 cDNA clones from
japonica rice. Science 301: 376–379.
Matsuoka, K., Demura, T., Galis, I., Horiguchi, T., Sasaki, M., Tashiro, G.
and Fukuda, H. (2004) A comprehensive gene expression analysis toward
the understanding of growth and differentiation of tobacco BY-2 cells.
Plant Cell Physiol. 45: 1280–1289.
Meinke, D.W., Cherry, J.M., Dean, C.D., Rounsley, S. and Koornneef, M.
(1998) Arabidopsis thaliana: a model plant for genome analysis. Science
282: 662–682.
Murashige, T. and Skoog, F. (1962) A revised medium for rapid
growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15:
473–497.
Nagata, T., Nemoto, Y. and Hasegawa, S. (1992) Tobacco BY-2 cell line as
the ‘HeLa’ cell in the cell biology of higher plant. Int. Rev. Cytol. 132:
1–30.
Nanjo, T., Futamura, N., Nishiguchi, M., Igasaki, T., Shinozaki, K. and
Shinohara, K. (2004) Characterization of full-length enriched expressed
sequence tags of stress-treated poplar leaves. Plant Cell Physiol. 45:
1738–1748.
Ogawa, Y. and Mii, M. (2007) Meropenem and moxalactam: novel
b-lactam antibiotics for efficient Agrobacterium-mediated transformation. Plant Sci. 172: 564–572.
Ohta, S., Mita, S., Hattori, T. and Nakamura, K. (1990) Construction and
expression in tobacco of a b-glucuronidase (GUS) reporter gene
containing an intron within the coding sequence. Plant Cell Physiol. 31:
805–813.
Saito, K., Yokoyama, H., Noji, M. and Murakoshi, I. (1995) Molecular
cloning and characterization of a plant serine acetyltransferase playing a
regulatory role in cysteine biosynthesis from watermelon. J. Biol. Chem.
270: 16321–16326.
Seki, M., Carninci, P., Nishiyama, Y., Hayashizaki, Y. and Shinozaki, K.
(1998) High-efficiency cloning of Arabidopsis full-length cDNA by
biotinylated CAP trapper. Plant J. 15: 707–720.
Seki, M., Narusaka, M., Kamiya, A., Ishida, J., Satou, M., et al. (2002)
Functional annotation of a full-length Arabidopsis cDNA collection.
Science 296: 141–145.
Stolc, V., Samanta, M.P., Tongprasit, W., Sethi, H., Liang, S., et al. (2005)
Identification of transcribed sequences in Arabidopsis thaliana by using
high-resolution genome tiling arrays. Proc. Natl Acad. Sci. USA 102:
4453–4458.
Takahashi, S., Katagiri, T., Hirayama, T., Yamaguchi-Shinozaki, K. and
Shinozaki, K. (2001) Hyperosmotic stress induced a rapid and transient
increase in inositol 1,4,5-trisphosphate independent of abscisic acid in
Arabidopsis cell culture. Plant Cell Physiol. 42: 214–222.
Tani, H., Chen, X., Nurmberg, P., Grant, J.J., SantaMaria, M., Chini, A.,
Gilroy, E., Birch, P.R. and Loake, G.J. (2004) Activation tagging
in plants: a tool for gene discovery. Funct. Integr. Genomics 4: 258–66.
Tsugane, T., Watanabe, M., Yano, K., Sakurai, N., Suzuki, H. and
Shibata, D. (2005) Expressed sequence tags of full-length cDNA clones
from the miniature tomato (Lycopersicon esculentum) cultivar MicroTom. Plant Biotechnol. 22: 161–165.
Uno, Y., Furihata, T., Abe, H., Yoshida, R., Shinozaki, K. and
Yamaguchi-Shinozaki, K. (2000) Arabidopsis basic leucine-zipper transcription factors involved in an abscisic acid-dependent signal transduction pathway under drought and high-salinity conditions. Proc. Natl
Acad. Sci. USA 97: 11632–11637.
Walhout, A. J., Temple, G. F., Brasch, M. A., Hartley, J. L., Lorson, M. A.,
van den Heuvel, S. and Vidal, M. (2000) GATEWAY recombinational
cloning: application to the cloning of large numbers of open reading
frames or ORFeomes. Methods Enzymol. 328: 575–592.
(Received October 24, 2007; Accepted December 26, 2007)