Download - Wiley Online Library

Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 2005? 200557615221531Original ArticleAgrobacterium colonization of unwounded plantsA. Brencic, E. R. Angert and S. C. Winans
Molecular Microbiology (2005) 57(6), 1522–1531
doi:10.1111/j.1365-2958.2005.04763.x
Unwounded plants elicit Agrobacterium vir gene
induction and T-DNA transfer: transformed plant cells
produce opines yet are tumour free
Anja Brencic, Esther R. Angert and
Stephen C. Winans*
Department of Microbiology, Cornell University, Ithaca, NY
14853, USA.
Summary
Agrobacterium tumefaciens is well known to cause
crown gall tumours at plant wound sites and to benefit
from this plant association by obtaining nutrients
called opines that are produced by these tumours.
Tumourigenesis requires expression of the vir regulon in response to chemical signals that are thought
to be released from wound sites. Here, we examine
chemical interactions between A. tumefaciens and
unwounded plants. To determine whether unwounded
plants can release significant amounts of vir gene
inducers, we constructed an A. tumefaciens strain
carrying a PvirB–gfp fusion. This fusion was strongly
induced by co-culture with tobacco seedlings that
have been germinated without any intentional wounding. The release of phenolic vir gene inducers was
confirmed by GC/MS analysis. We also constructed a
strain containing the gfp reporter located on an artificial T-DNA and expressed from a plant promoter. A.
tumefaciens efficiently transferred this T-DNA into
cells of unwounded plants in the absence of exogenous vir gene inducers. Many cells of seedlings colonized by the bacteria also produced octopine, which
was detected using a Pocc-gfp reporter strain. This
indicates transfer of the native T-DNA. However, these
transformed plant cells did not form tumours. These
results suggest that successful colonization of plants
by A. tumefaciens, including T-DNA transfer and opine
production, does not require wounding and does not
necessarily cause cell proliferation. Transformation of
plant cells without inciting tumours may represent a
colonization strategy for this pathogen that has
largely been overlooked.
Accepted 8 June, 2005. *For correspondence. E-mail scw2@
cornell.edu; Tel. (+1) 607 255 2413; Fax (+1) 607 255 3904.
© 2005 Blackwell Publishing Ltd
Introduction
Many pathogens of plants or animals can colonize their
hosts in several different ways with a range of consequences. In particular, many can colonize their hosts for
extended periods of time without causing any clinical manifestations. For example, Mycobacterium tuberculosis can
infect the human lung in the absence of symptoms, and
colonization causes disease only in a subset of those
infected (Parrish et al., 1998; Stewart et al., 2003; Monack
et al., 2004). Helicobacter pylori, species of Neisseria,
Staphylococcus, Streptococcus, and many other bacteria
that we generally consider to be serious pathogens, can
frequently establish long-term infections in humans that
are clinically asymptomatic (Monack et al., 2004).
Agrobacterium tumefaciens has long been known to
induce crown gall tumours at the wound sites of a variety
of plants. These tumours can seriously impair the host
(Smith and Townsend, 1907), and result from the transfer
of oncogenic DNA fragments (T-DNA) from the tumourinducing (Ti) megaplasmid to the chromosomes of plant
cells (Binns and Constantino, 1998; Tzfira et al., 2000;
Zhu et al., 2000; Zupan et al., 2000; Gelvin, 2003; Tzfira
et al., 2004). Some of the transferred genes cause uncontrolled cell division, while others direct the synthesis of
bacterial nutrients called opines. A. tumefaciens, however,
is not an obligate pathogen, as it can grow vigorously as
a saprophyte. It can also colonize plants without causing
any symptoms, forming relatively benign biofilms on roots
(Matthysse and Kijne, 1998; Matthysse and McMahan,
1998). These commensal interactions between A. tumefaciens and unwounded plants have received relatively
little attention, and are the topic of the present study.
T-DNA transfer requires products of the vir regulon,
whose transcription is induced by specific plant-released
phenolic compounds in combination with several other
stimuli, including plant-released monosaccharides, acidic
pH and temperatures below 30∞C (Heath et al., 1995).
These signals are detected by the VirA-VirG two component system, and by the periplasmic sugar binding protein
ChvE (Stachel et al., 1985; 1986; Stachel and Zambryski,
1986; Heath et al., 1995). Generally, it is believed that vir
gene-inducing signals are released from plants only at
wound sites. Phenolics are involved in lignification and
Agrobacterium colonization of unwounded plants 1523
healing of the wound, and may also be released from the
wound site as antimicrobials (Dixon and Paiva, 1995).
However, there are very few if any data suggesting that
wounding is required for the production or release of these
compounds. In one study, two phenolic inducers, acetosyringone (AS) and a-hydroxyacetosyringone, were
released in higher amounts from tobacco leaf discs than
from unwounded leaves (Stachel et al., 1985). However,
these compounds were first isolated from cultured plant
cells or roots, neither of which had been wounded
(Stachel et al., 1985; 1986). While the cultured cells and
tissues used in those studies may not reflect the exudates
of intact plants, it nevertheless seems plausible that
unwounded plants might release sufficient vir gene inducers to stimulate the vir regulon.
It has long been known that inoculating A. tumefaciens
on plants rarely if ever causes tumours unless the plant
is wounded at the site of inoculation (14). However, the
precise role of wounding in this process is unclear. As
described above, plant injury may be important for the
release of vir gene inducers. Wounding of plant tissues
may also be important to compromise any physical barriers that might block T-DNA transfer. The waxy cuticle that
coats plant epidermis may block productive physical contacts between the bacterium and the host cell envelope
that are required for T-DNA transfer (Wagner and Matthysse, 1992; Swart et al., 1994; Zhu et al., 2003). However, leaf mesophyll cells and root epidermal cells are not
protected by a cuticle, because they must be permeable
to gasses, minerals, and water. Indeed, one study showed
that T-DNA can be transferred to mesophyll cells of
unwounded plants (Escudero and Hohn, 1997), although
transfer in that study occurred only when the bacteria
were pre-induced with a synthetic phenolic compound.
It has been proposed that cell division during wound
healing may play a role in tumourigenesis (Kahl, 1982;
Binns and Thomashow, 1988). Dividing cells may have
less rigid cell walls that can be more easily breached. In
addition, DNA replication may facilitate T-DNA integration,
and the dividing cells may be more sensitive to the mitogenic effects of the T-DNA encoded phytohormones. Supporting these ideas is the finding that application of
exogenous auxin prior to infection stimulates plant cell
transformation (Sangwan et al., 1992; Stover et al., 1997;
Chateau et al., 2000). Other studies reported that Agrobacterium-induced tumours were limited to the cambium,
which is a meristematic tissue and thus predisposed to
cell division (Sangwan et al., 1991; Ghorbel et al., 2000).
In another study, the apical meristem of maize seedlings
was a preferred site for transformation (Shen et al., 1993).
Wounding can cause virtually any plant cell to dedifferentiate and become meristematic (Taiz and Zeiger, 2002).
This normally occurs during the process of wound healing
and is followed by rapid proliferation of cells at the site of
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
injury. Therefore, wounding may be required for differentiated plant cells to begin neoplastic cell division. Wounding
may also provide A. tumefaciens physical access to internal meristematic tissues such as the cambium.
This study focuses on interactions between A. tumefaciens and unwounded plants. Based on the studies
described above, it seemed plausible that unwounded
plants could release sufficient amounts of vir gene inducers to cause strong vir expression in agrobacteria colonizing these plants. Furthermore, it seemed plausible that at
least some cells of these plants could serve as targets of
T-DNA transformation. Wounding might, however, be
required for neoplastic proliferation of transformed plant
cells, although these cells may be able to produce and
release opines that can be consumed by the bacteria. To
test these hypotheses, we constructed a strain of A. tumefaciens containing a PvirB–gfp fusion, and used it to assay
the induction of vir genes in bacteria co-cultured with
tobacco seedlings that had been germinated without any
intentional wounding. We also constructed a strain containing a gfp gene expressed from a plant promoter and
located within an artificial T-DNA. We used this strain to
test whether unwounded plants can be transformed in the
absence of exogenously added phenolic compounds.
Finally, we used A. tumefaciens strains containing either
a PoccQ–gfp or an ooxA–gus reporter, both of which are
octopine-inducible, to test for production and release of
octopine by transformed tobacco cells.
Results
Construction and characterization of a P virB–gfp reporter
To observe in planta induction of vir genes, we constructed an A. tumefaciens strain (AB601) that contains a
wild-type copy of the virB operon and a PvirB–gfp fusion
carried in cis. To characterize this fusion, the strain was
cultured in AB minimal medium (pH 5.8) in the presence
of different concentrations of AS, and fluorescence of
bacterial cultures was quantified by flow cytometry. The
average fluorescence of all cultures increased with
increasing AS concentrations and with time, and the
fusion reached its maximal induction after two to 3 days
of incubation (Fig. 1A). In the culture where no AS was
added, cells were uninduced for the entire length of the
experiment (Fig. 1A). Expression of the reporter was abolished by culturing the cells at neutral pH rather than acidic
medium or at 37∞C rather than 27∞C (data not shown). For
all AS concentrations tested, all cells in a culture eventually became fluorescent, however, each culture showed a
broad distribution of fluorescence, and some cells within
each culture became fluorescent considerably earlier than
others (Fig. 1B). Cells incubated for 12 h with AS
appeared to have a bimodal distribution of fluorescence
1524 A. Brencic, E. R. Angert and S. C. Winans
vir-inducing compounds at concentrations sufficient to
activate the vir regulon.
To quantify the in planta vir induction, we periodically
sampled six seedlings that had been inoculated as
described above and vortexed them in liquid MS medium
to recover the bound bacteria. Bacterial samples were
then frozen at -80∞C and later assayed for vir gene induction by flow cytometry. Induction of the PvirB–gfp reporter
was first detected in the sample obtained on day 3 after
inoculation, and was strongest in the sample recovered
on the final day (day 8) of the experiment (Fig. 3A). Bacteria in these samples showed a bimodal distribution of
fluorescence (Fig. 3B). On day 8, approximately half of the
cells showed a fairly uniform level of fluorescence, while
the other half remained uninduced. The induced subpopulation by day 8 showed a fluorescence intensity of 100–
200 fluorescence units (Fig. 3B), which is similar to that
obtained using AB liquid medium supplemented with
10 mM AS (Fig. 1B, middle column), although induction on
plant surfaces required a considerably longer time interval
(8 days vs. 2 days). Overall, these results confirm our
hypothesis that vir genes of agrobacteria colonizing
unwounded tobacco seedling are strongly induced. This
was later further substantiated by our ability to detect TDNA transfer under these conditions (see below).
Fig. 1. Expression of PvirB–gfp of strain AB601 cultured in AB
medium (pH 5.8) in the presence of different AS concentrations.
Fluorescence was detected by flow cytometry for 10 000 cells per
sample.
A. Average fluorescence of samples cultured with 100 mM (), 10 mM
(), 1 mM (), or no AS ().
B. Histograms showing distribution of fluorescence among individual
cells of each sample at different time points during incubation. No
fluorescent cells were detected in samples cultured in the absence
of AS (data not shown).
Analysis of vir-inducing compounds in seedling exudates
To identify phenolic inducers released from unwounded
seedlings, exudates of three types of plants (tobacco,
squash and tomato) were analysed using bioassays
(Miller, 1972) and by mass spectroscopy. Induction of the
virB–lacZ reporter of A. tumefaciens strain A136(pCH114)
intensities, with approximately half of the cells remaining
uninduced. At later time points, most cultures showed
greater homogeneity of GFP expression (Fig. 1B).
Induction of P virB–gfp in bacteria bound to unwounded
tobacco seedlings
To test whether the vir regulon is induced in the presence
of intact whole plants, we inoculated hydroponically germinated tobacco seedlings with strain AB601, and placed
them on solidified medium containing MS salts and lacking any carbon source. Care was taken not to injure these
seedlings in any way. Periodically, individual seedlings
were examined by epifluorescence microscopy for the
presence of fluorescent bacteria attached to the surfaces
of these seedlings. Bound bacteria began to express GFP
approximately 3 days after inoculation (Fig. 2). This was
true for root-bound bacteria (Fig. 2A and B), as well as for
bacteria colonizing the cotyledons (Fig. 2C and D). These
observations show that intact tobacco seedlings release
Fig. 2. Induction of PvirB–gfp of strain AB601 on root surfaces (A
and B) and cotyledons (C and D) of tobacco seedlings, as detected
by fluorescence microscopy 6 days after inoculation.
A and C. Bright field images.
B and D. Fluorescence images.
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Agrobacterium colonization of unwounded plants 1525
Hydroponically germinated tobacco seedlings were
inoculated with strain AB602 and transferred to solidified
MS medium. Periodically, individual seedlings were
removed and examined by epifluorescence microscopy.
After 3–4 days of incubation, we observed fluorescent
plant cells, with the predicted distribution of fluorescence
within each transformed cell (Fig. 4) (Haseloff et al., 1997;
Haseloff, 1999). This distribution was very similar to that
of transgenic Arabidopsis thaliana plants expressing the
same transgene (Haseloff, 1999). The numbers of fluorescent cells per seedling varied considerably among individual seedlings, but in many seedlings the fluorescent cells
were quite numerous. Transformed cells were largely limited to the leaf mesophyll (Fig. 4B and D), although occasionally we also observed transformation of leaf epidermal
cells (Fig. 4F), stomatal cells, and root epidermal cells
(data not shown). No fluorescent plant cells were
observed using a bacterial strain lacking pBIN-mgfp5-ER
or using a virG mutant strain. We do not know whether
the gfp transgene was integrated into the plant genome.
Fig. 3. In planta expression of the PvirB–gfp fusion of strain AB601.
Bacteria were recovered from tobacco seedlings at different time
points after inoculation, and fluorescence was detected by flow
cytometry for 10 000 cells per sample.
A. Average fluorescence of samples recovered from seedlings at
different time points after inoculation.
B. Histograms showing distribution of fluorescence among individual
cells of each sample at different time points during incubation.
(pCH116) (Chang et al., 1996), was detected with exudates of tobacco and squash, and was highest in the
sample from the final day of the experiment (day 15),
when activity of tobacco exudates was equivalent to 1–
10 mM AS, while the activity of squash exudates was
approximately threefold lower (Table S1). Exudates of
tomato exhibited no vir-inducing activity (Table S1). GC/
MS analysis revealed AS in the exudates of tobacco
(Fig. S1A and B), while exudates of squash contained two
other vir inducers, syringate and vanillate (Fig. S1C–E).
Tomato exudates did not contain detectable levels of any
known vir gene inducer (data not shown).
Transformation of intact tobacco seedlings
Based on the above results, it seemed possible that A.
tumefaciens could transfer T-DNA to the cells of intact
tobacco seedlings without AS pretreatment. To test this,
we introduced plasmid pBIN-mgfp5-ER into A. tumefaciens strain R10, resulting in strain AB602. pBIN-mgfp5ER is a binary vector with gfp fused to the 35S CaMV
promoter and inserted between the T-DNA borders of the
vector (Haseloff et al., 1997). The GFP protein encoded
by this plasmid is modified for transport to the endoplasmic reticulum, which tends to be localized near the nuclei
and cell periphery (Haseloff et al., 1997).
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Fig. 4. Expression of a T-DNA-localized gfp gene in cells of tobacco
seedlings, as viewed by fluorescence microscopy 8 days after inoculation with strain AB602.
A and B. Transformed mesophyll cells of a cotyledon.
C and D. Shoot apex surrounded by two cotyledons, with transformed
cells located in the shoot apex and near the apical meristem.
E and F. Edge of a cotyledon with one transformed epidermal cell
and numerous transformed cells in the mesophyll layer.
G and H. Close-up views of single transformed cells shown in part B.
A, C and E. Bright field images.
B, D, F, G and H. Fluorescence images.
1526 A. Brencic, E. R. Angert and S. C. Winans
The fluorescent tobacco cells described above never
initiated rapid cell division that would have resulted in a
tumour. In the entire course of this study, tumours were
not observed in unwounded plants, regardless of the bacterial strain. In each experiment, some seedlings were left
to grow on MS medium for at least 3 weeks. These seedlings did not develop tumours and the great majority
lacked any obvious disease symptoms.
Detection of octopine by bacteria bound to
unwounded seedlings
Although the transformed cells detected in this study did
not proliferate, it seemed likely that many of them were
transformed by the native T-DNA. These cells would
express opine synthase genes, and would release opines
in amounts that might be sufficient for detection by the
colonizing bacteria. To test this, we inoculated seedlings
with strain AB603, which is a derivative of strain R10 with
a PoccQ–gfp fusion and an intact occ operon carried in
cis. Eight days after inoculation, we observed fluorescent
bacteria on two out of nine examined seedlings. By day
14, fluorescent cells were observed on virtually all of the
seedlings (Fig. 5A), though the induced bacteria were limited to cotyledons and their numbers varied greatly from
seedling to seedling. Although the roots were heavily colonized, none of the root-colonizing bacteria were fluorescent (data not shown). This is consistent with the
extremely low level of root cell transformation described
above.
To confirm these results and to quantify the in planta
induction of the occQ operon, we inoculated seedlings
with strain KYC16, which carries an octopine-inducible
ooxA–gusA7 fusion (Cho et al., 1996). This strain detects
octopine at 1000-fold lower concentrations than does
AB603 (data not shown), probably because AB603
degrades octopine, while KYC16 does not, so the added
octopine is not depleted (Cho et al., 1996). At regular
intervals, bacteria were recovered from the seedlings by
maceration and assayed for b-glucuronidase (Gallagher,
1992). This activity was first detected on day 5 after inoculation, and continued to increase for the remaining
11 days of the experiment (Fig. 5B). The maximal
detected activity was similar to the activity obtained when
strain KYC16 was cultured in AT medium in the presence
of 10–100 nM octopine (Fig. 5C). However, based on our
microscopic observations, we predict that many bacteria
were not induced, especially those bound to the roots. As
the b-glucuronidase assay measures the average gene
expression of the entire population, within each sample,
activity of the induced subpopulation was probably higher.
No Gus activity was detected in the samples processed
from uninoculated seedlings (data not shown). These
results indicate that cells of these seedlings were trans-
Fig. 5. A. Expression of the PoccQ–gfp fusion of strain AB603 during
colonization of tobacco seedling cotyledons, as viewed by fluorescence microscopy 8 days after inoculation. The same strain incubated
in defined broth was not detectably fluorescent (data not shown).
B. GUS specific activity of ooxA–gus of strain KYC16 recovered from
tobacco seedlings at different time points after inoculation. Before
inoculation of seedlings, KYC16 cells were cultured overnight in AB
minimal medium (pH 5.8) with 25 mM AS. Expression is represented
in units of b-glucuronidase specific activity (Gallagher, 1992). No Gus
activity was detected in the samples processed from uninoculated
seedlings (data not shown).
C. GUS specific activity of the ooxA–gus of strain KYC16 cultured in
AT minimal medium at 27∞C in the presence of 10 mM octopine (),
1 mM (X), 100 nM (D), 10 nM (), or no octopine (). Samples were
analysed for b-glucuronidase specific activity as described previously
(Gallagher, 1992).
formed by the native T-DNA, resulting in the production of
octopine, which was detected and presumably catabolized by the bacteria.
Discussion
To our knowledge, this is the first report of the use of a
Pvir–gfp fusion to study the induction of the A. tumefaciens vir regulon, though others have made AS-inducible
vir–gfp fusions (Rashkova et al., 2000; Atmakuri et al.,
2003). We established that vir genes are induced during
colonization of intact tobacco seedlings, and we detected
AS in the exudates of axenic tobacco seedlings. We also
detected syringate and vanillate in the exudates of squash
seedlings. These results unequivocally demonstrate that
intentional wounding is not required for the release of vir© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Agrobacterium colonization of unwounded plants 1527
gene inducers, and that these compounds are released
from unwounded plants in amounts sufficient for induction
of vir genes. Although we cannot rule out the possibility
that small wound-like lesions occur during normal plant
growth, it may be better to think of vir-gene inducing
signals as ‘plant-released’, rather than ‘wound-released’.
However, it still needs to be established whether these
findings can be extended to other plant species and/or to
other stages of growth.
When PvirB–gfp-expressing bacteria were treated with
AS in broth cultures, at early stages of induction, only a
small proportion of them were induced. This could be due
to the fact that virA and virG are positively autoregulated,
and that increased amounts of VirG enhance induction
(Heath et al., 1995). It seems possible that some cells in
a population might reach a threshold at which the concentration of VirG is sufficiently high to initiate a positive
amplification loop, while other cells in the same population
may not reach this threshold until later times. Such a
phenomenon could help explain why in one study with
cultures of AS-treated agrobacteria, only 10% of the bacteria displayed the virB-encoded pilus (Fullner et al.,
1996). Cell to cell variation in the response to environmental stimuli has been described for numerous biological
systems (McAdams and Arkin, 1997; 1999). These variations are commonly observed even under the most uniform experimental conditions, such as in well-mixed
bacterial cultures, and are thus believed to arise from
variations in internal cellular processes rather than from
differing microenvironments. The molecular-level phenomena that produce these differences are thought to be
rooted in the statistical mechanical behaviour of smallscale chemical systems, where concentrations of reacting
species are extremely low and, consequently, stochastic
fluctuations in reaction rates are large (McAdams and
Arkin, 1997; 1999).
We noted a difference in induction kinetics between
bacteria cultured in broth with AS and bacteria growing
on plant surface. This could be due to the plants releasing
inducers only in the presence of bacteria (Stachel et al.,
1985), however, this cannot be the only explanation
because we have detected these compounds in exudates
of axenic plants (Fig. S1, Table S1). It could also be
because in the former experiment (Fig. 1A), glucose was
provided for bacterial growth and to enhance vir gene
induction, while in the latter experiment (Fig. 3), the bacteria obtained all nutrients from their hosts. Furthermore,
many of the bacteria growing on seedlings appeared to
form microcolonies or biofilms and may have been less
metabolically active than planktonic cells.
We further showed that A. tumefaciens can transfer TDNA to cells of unwounded tobacco seedlings without
pretreatment of the bacteria with AS. T-DNA transfer to
cells of non-wounded tobacco has been documented
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
before, however, transformation in that study required that
the bacteria be pre-induced with AS (Escudero and Hohn,
1997). Differences in experimental design, such as the
age of the seedlings and the fact that in our study seedlings were grown hydroponically until inoculation, must
account for the different results of the two studies.
Most of the transformed cells described in this study
were located within the mesophyll layer of the cotyledons,
although some were located in the leaf epidermis. Transformation of these cells might seem surprising given the
existence of the waxy cuticle. The study by the Hohn lab
showed that plant transformation required opening of stomata (Escudero and Hohn, 1997), suggesting that the
bacteria used stomata to enter the apoplast. In our experiments, the stomata were frequently observed to be fully
open (data not shown). Thus, the mesophyll and epidermal cells may have been transformed by the bacteria that
entered the apoplast, which would not require penetrating
the cuticle.
Only rarely did we observe transformation of the root
epidermal cells. Further evidence that roots were rarely
transformed was obtained using a PoccQ–gfp reporter,
which was induced in bacteria colonizing the cotyledons,
but not in those colonizing roots or root hairs. It is not clear
why roots were not more readily transformed. In particular,
root hairs lack a cuticle, and the nuclei of root epidermal
cells are frequently in the root hair itself, so they should
in principle be highly susceptible to transformation. One
trivial explanation is that transformed root cells do not
express detectable levels of GFP or release octopine at
levels sufficient for detection.
Although A. tumefaciens is a soil inhabitant, it should
not be surprising that these bacteria can colonize leaves,
especially those of seedlings, which grow among soil particles, often in very wet conditions. Seedlings are readily
colonized by various species of soil bacteria, including
Pseudomonas spp., which is one of the most frequent
colonizers of leaf surfaces (Andrews and Harris, 2000;
Hirano and Upper, 2000). What surprised us, however,
was that these bacteria detected signals generally thought
to be released only from wound sites, that they were so
readily able to transfer T-DNA, and that T-DNA transfer did
not cause tumours.
Tumours were never observed in the absence of wounding with any of the strains used in this study. Neoplastic
proliferation of transformed cells might require wounding
and/or might occur only in particular cell types. In the
previous study using unwounded plants, cell proliferation
was occasionally detected (Escudero and Hohn, 1997).
Others, however, have shown that tumours initiate only
from meristematic tissues, such as the vascular cambium
(Sangwan et al., 1991; Ghorbel et al., 2000). As wounding
induces plant cell dedifferentiation and division, it is possible that in tissues that are not intrinsically meristematic,
1528 A. Brencic, E. R. Angert and S. C. Winans
the wound response is needed to initiate tumours (Kahl,
1982; Binns and Thomashow, 1988). Another possible
explanation is that tumours may occur only when a sufficiently large number of plant cells are transformed in a
confined space, analogous to bacterial ‘quorum sensing’.
Without exogenous wounding, the number of transformed
plant cells might be below this critical threshold. It is also
possible that in our experiments, T-DNA did not integrate
into the plant chromosome and hence that the amount of
plant hormones produced during transient expression of
T-DNA genes was not sufficient to initiate cell proliferation.
The infected seedlings, however, produced octopine at
levels that were readily detectable by the bacteria. This
indicates that successful colonization of plants by A. tumefaciens, resulting in opine production by the colonized
plants, might occur in the absence of tumours. As
described above, phytohormone production might have
been too low in transformed cells to initiate their proliferation, or, alternatively, these differentiated cells might not
proliferate in response to phytohormones. However, one
of these hormones, cytokinin, stimulates transport of nutrients from the rest of the plant to the cells that release it
(Davies, 1990), and might therefore enhance the release
of nutrients to the bacteria. Transport of nutrients to the
transformed cells might also enhance the production of
opines.
In this study, we have demonstrated that A. tumefaciens
can detect vir gene inducers from unwounded seedlings,
that it can efficiently transfer T-DNA to the cells of these
seedlings, and that these transformed seedlings release
octopine but do not show any obvious signs of disease.
Taken together, these results suggest that transformation
of cells of non-wounded plants could be an entirely new
strategy for A. tumefaciens plant colonization that has
largely been overlooked, and this discovery could lead to
an entirely new paradigm for A. tumefaciens–plant interactions. Based on the results of our study, it is plausible
that transformation of plant cells without causing their
proliferation could be a common outcome of T-DNA transfer. Studies of this and most other host–pathogen interactions have mainly been focused on catastrophic
interactions, while commensal interactions are often overlooked or less emphasized. However, minimizing the harm
to the host may be a sound strategy for many pathogens,
and commensalism may represent the ideal interaction for
both host and parasite (Ewald, 1983).
Experimental procedures
Strains and plasmids
Plasmid pJS202 is a derivative of pVIK165 (Kalogeraki and
Winans, 1997), which was designed to construct fusions
between bacterial promoters and gfp(S65T). In pSJ202, the
gfp gene of pVIK165 was replaced with gfp(S65G, S72A)
(Cormack et al., 1996), by digesting pVIK165 with XbaI
(which cuts within the multiple cloning site) and MfeI (which
cuts within the gfp gene) and introducing an XbaI-MfeI fragment containing gfp(S65G, S72A) from plasmid pKEN (Cormack et al., 1996).
To make a PvirB–gfp fusion, a 420 bp DNA fragment that
contains the virB promoter was synthesized by polymerase
chain reaction (PCR) amplification using primers 5¢-GCAC
GAATTCCCGAGCGAAGGTTTTCGC-3¢ and 5¢-GCAGAG
CTCCGAATGCCGTGCATCGAGACGG-3¢, and was introduced into pSJ202 as an EcoRI-SacI fragment. The resulting
plasmid, pSSC103, was mobilized from Escherichia coli
strain S17-ë pir to A. tumefaciens strain R10 by conjugation,
where it recombined into the Ti plasmid by Campbell integration, resulting in a PvirB–gfp fusion. In the resulting strain,
AB601, the native virB operon is preserved downstream of
the integrated plasmid. Strain R10 is a wild-type A. tumefaciens strain with octopine-type Ti plasmid pTiR10 (Sciaky
et al., 1978).
A strain expressing a PoccQ–gfp fusion was constructed
by the same procedure except that a 260 bp occQ-occR
intergenic region including the occQ promoter was cloned
from pLW131 (Wang and Winans, 1995), into pSJ202 as an
EcoRV fragment, resulting in plasmid pRA400. This plasmid
was mobilized from E. coli S17-ë pir to A. tumefaciens R10
by conjugation, resulting in strain AB603. AB603 retained the
wild-type copy of the occQ operon and a PoccQ–gfp fusion
carried in cis. Strain KYC16 is a derivative of strain R10 with
Tn5gusA7 integrated at the ooxA gene (Cho et al., 1996).
Plasmid pBIN-mgfp5-ER is a binary vector that carries a
fusion between the 35S CaMV promoter and gfp, inserted
between the T-DNA borders (Haseloff et al., 1997). The plasmid was introduced into A. tumefaciens strain R10 via electroporation, resulting in strain AB602.
Agrobacterium tumefaciens reporter strain A136(pCH114)
(pCH116) was used in bioassays to test for the presence of
vir inducers in plant exudates. The strain expresses virA–,
virG– and virB–lacZ (Chang et al., 1996).
Bioassays
For bioassays of seedling exudates, seeds of tobacco (Nicotiana tabacum cv. Samsun NN), tomato (Lycopersicum
esculentum cv. Heinz 1370) and squash (Cucurbita maxima)
were surface sterilized and transferred to sterile Erlenmeyer
flasks containing 50 ml of deionized water. The flasks were
incubated with mild shaking at room temperature for 15 days.
Every 3 days, 10 ml of samples of the culture supernatants
were withdrawn, acidified by addition of 100% phosphoric
acid to a final concentration of 1.5% v/v, and extracted with
an equal volume of ethyl acetate. Ethyl acetate was evaporated using nitrogen gas, and pellets were resuspended in
5 ml of AB minimal medium (pH 5.8), filter sterilized, and
inoculated with strain A136(pCH114) (pCH116) (Chang
et al., 1996). Cells were cultured overnight at 27∞C for 16 h
and assayed for b-galactosidase activity (Miller, 1972).
GUS specific activity of ooxA–gus fusion of strain
KYC16 was measured and represented in units of bglucuronidase specific activity as described previously
(Gallagher, 1992).
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Agrobacterium colonization of unwounded plants 1529
Seed germination and seedling inoculation
Acknowledgements
Seeds of N. tabacum cv Samsun NN were surface-sterilized
by soaking in 50% bleach and 0.1% SDS for 10 min, followed
by extensive washing. Seeds were then transferred to sterile
water in Petri dishes and incubated for 48 h at 4∞C in the
dark, after which the seedlings were maintained at room
temperature with exposure to natural and artificial lighting.
Unless indicated otherwise, before inoculation, bacteria
were cultured overnight in Luria–Bertani medium with antibiotics. Cells were then washed with sterile water, and
suspended to OD600 = 0.5 in 5% sucrose and 0.005%
TRITON-X100. Approximately 100 hydroponically grown
seedlings were dipped into the bacterial suspensions for
5 min, washed gently in 5% sucrose, and transferred to solidified MS medium lacking any carbon source (Murashige and
Skoog, 1962). At the time of inoculation, seedlings were 1–
2 cm long and approximately 2 weeks old. After inoculation,
seedlings were incubated at room temperature on a bench
top with exposure to ambient fluorescent and natural lighting.
We would like to acknowledge Stephen S. Chung, and Drs
Samina Jafri and Reiko Akakura for constructing some of the
strains used in the study. We thank Dr Rod Getchell for help
with flow cytometry, Dr Anatol Eberhard for help with GC/MS,
and Ms. Nahid Shaikh for help with plant inoculations and
microscopic observations. We also would like to thank Dr
Larry Walker for help with fluorescence spectrophotometry.
Microscopy
Individual seedlings were transferred to a glass slide, overlaid
with a cover slip, and viewed on an Olympus BX61 microscope using 20¥ and 40¥ phase-contrast objectives. For
detection of GFP, samples were viewed by epifluorescence
microscopy using a Chroma 41017 (bandpass emission) filter
cube [HQ 470/40, Q 495LP, HQ 525/50] or a Chroma 41012
(longpass emission) wide blue filter cube [HQ 480/40, Q
505LP, HQ 510LP]. Images were acquired using a Cooke
SensiCam with a Sony Interline chip. Image acquisition was
performed using SlideBook Software package (Intelligent
Imaging) and figures were assembled using Photo-Paint
(Corel, Salinas, CA).
Recovery of bacteria from seedlings
For each sample, six seedlings were transferred to liquid MS
medium in a culture tube, which was then vortexed vigorously
to remove the bound bacteria. In experiments involving the
ooxA–gus fusion, six seedlings were transferred to liquid MS
medium in an Eppendorf tube, and macerated. The liquid was
separated from plant fragments by pipetting and centrifugation. Pellets were resuspended in 200 ml of 20% glycerol and
stored at -80∞C until analysis.
Analysis of vir gene induction by flow cytometry
Strain AB601 was cultured at 27∞C in AB minimal medium
(pH 5.8) in the presence of various concentrations of the vir
gene inducer AS. Culture samples were collected at different
time points and frozen in 25% glycerol at -80∞C. On the day
of analysis, samples were thawed, diluted approximately 25fold in AB buffer, and analysed in a FACSCalibur flow cytometer (Beckton Dickinson) equipped with an argon laser
emitting at 488 nm. Fluorescence data were collected using
logarithmic amplifiers for 10 000 cells per sample (as determined by light side-scattering), and analysed using CellQuest
software.
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
References
Andrews, J.H., and Harris, R.F. (2000) The ecology and
biogeography of microorganisms on plant surfaces. Annu
Rev Phytopathol 38: 145–180.
Atmakuri, K., Ding, Z., and Christie, P.J. (2003) VirE2, a type
IV secretion substrate, interacts with the VirD4 transfer
protein at cell poles of Agrobacterium tumefaciens. Mol
Microbiol 49: 1699–1713.
Binns, A.N., and Thomashow, M.F. (1988) Cell biology of
Agrobacterium infection and transformation of plants. Annu
Rev Microbiol 42: 575–606.
Binns, A.N., and Constantino, P. (1998) The Agrobacterium
oncogenes. In The Rhizobiaceae. Spaink, H.P., Kondorosi,
A., and Hooykaas, P.J.J. (eds). Dordrecht: Kluwer Academic Publishers, pp. 251–266.
Chang, C.H., Zhu, J., and Winans, S.C. (1996) Pleiotropic
phenotypes caused by genetic ablation of the receiver
module of the Agrobacterium tumefaciens VirA protein. J
Bacteriol 178: 4710–4716.
Chateau, S., Sangwan, R.S., and Sangwan-Norreel, B.S.
(2000) Competence of Arabidopsis thaliana genotypes and
mutants for Agrobacterium tumefaciens-mediated gene
transfer: role of phytohormones. J Exp Bot 51: 1961–1968.
Cho, K., Fuqua, C., Martin, B.S., and Winans, S.C. (1996)
Identification of Agrobacterium tumefaciens genes that
direct the complete catabolism of octopine. J Bacteriol 178:
1872–1880.
Cormack, B.P., Valdivia, R.H., and Falkow, S. (1996) FACSoptimized mutants of the green fluorescent protein (GFP).
Gene 173: 33–38.
Davies, P.J. (1990) Plant Hormones and Their Role in Plant
Growth and Development. Dordrecht: Kluwer Academic
Publishers.
Dixon, R.A., and Paiva, N.L. (1995) Stress-induced phenylpropanoid metabolism. Plant Cell 7: 1085–1097.
Escudero, J., and Hohn, B. (1997) Transfer and integration
of T-DNA without cell injury in the host plant. Plant Cell 9:
2135–2142.
Ewald, P.W. (1983) Host-parasite relations, vectors, and the
evolution of disease severity. Annu Rev Ecol Syst 14: 465–
485.
Fullner, K.J., Lara, J.C., and Nester, E.W. (1996) Pilus
assembly by Agrobacterium T-DNA transfer genes. Science 273: 1107–1109.
Gallagher, S.R. (1992) GUS Protocols. San Diego, CA: Academic Press.
Gelvin, S.B. (2003) Agrobacterium-mediated plant transformation: the biology behind the ‘gene-jockeying’ tool. Microbiol Mol Biol Rev 67: 16–37.
Ghorbel, R., Dominguez, A., Navarro, L., and Penna, L.
1530 A. Brencic, E. R. Angert and S. C. Winans
(2000) High efficiency genetic transformation of sour
orange (Citrus aurantium) and production of transgenic
trees containing the coat protein gene of citrus tristeza
virus. Tree Physiol 20: 1183–1189.
Haseloff, J. (1999) GFP variants for multispectral imaging of
living cells. Methods Cell Biol 58: 139–151.
Haseloff, J., Siemering, K.R., Prasher, D.C., and Hodge, S.
(1997) Removal of a cryptic intron and subcellular localization of green fluorescent protein are required to mark transgenic Arabidopsis plants brightly. Proc Natl Acad Sci USA
94: 2122–2127.
Heath, J.D., Charles, T.C., and Nester, E.W. (1995) Ti plasmid and chromosomally encoded two-component systems
important in plant cell transformation by Agrobacterium
species. In Two-Component Signal Transduction. Hoch,
J.A., and Silhavy, T.J. (eds). Washington, DC: American
Society for Microbiology Press, pp. 367–385.
Hirano, S.S., and Upper, C.D. (2000) Bacteria in the leaf
ecosystem with emphasis on Pseudomonas syringae: a
pathogen, ice nucleus, and epiphyte. Microbiol Mol Biol
Rev 64: 624–653.
Kahl, G. (1982) Molecular biology of wound healing: the
conditioning phenomenon. In Molecular Biology of Plant
Tumors. Kahl, G., and Schell, J.S. (eds). New York: Academic press, pp. 211–267.
Kalogeraki, V.S., and Winans, S.C. (1997) Suicide plasmids
containing promoterless reporter genes can simultaneously disrupt and create fusions to target genes of
diverse bacteria. Gene 188: 69–75.
McAdams, H.H., and Arkin, A. (1997) Stochastic mechanisms in gene expression. Proc Natl Acad Sci USA 94:
814–819.
McAdams, H.H., and Arkin, A. (1999) It’s a noisy business!
Genetic regulation at the nanomolar scale. Trends Genet
15: 65–69.
Matthysse, A.G., and Kijne, J.W. (1998) Attachment of Rhizobiaceae to plant cells. In The Rhizobiaceae. Spaink, H.P.,
Kondorosi, A., and Hooykaas, P.J.J. (eds). Dordrecht: Kluwer Academic Publishers, pp. 235–249.
Matthysse, A.G., and McMahan, S. (1998) Root colonization
by Agrobacterium tumefaciens is reduced in cel, attB,
attD, and attR mutants. Appl Environ Microbiol 64: 2341–
2345.
Miller, J. (1972) Experiments in Molecular Genetics. Cold
Spring Harbor, NY: Cold Spring Harbor Laboratory
Press.
Monack, D.M., Mueller, A., and Falkow, S. (2004) Persistent
bacterial infections: the interface of the pathogen and the
host immune system. Nat Rev Microbiol 2: 747–765.
Murashige, T., and Skoog, F. (1962) A revised medium for
rapid growth and bioassays with tobacco tissue cultures.
Physiol Plant 15: 473–497.
Parrish, N.M., Dick, J.D., and Bishai, W.R. (1998) Mechanisms of latency in Mycobacterium tuberculosis. Trends
Microbiol 6: 107–112.
Rashkova, S., Zhou, X.R., Chen, J., and Christie, P.J. (2000)
Self-assembly of the Agrobacterium tumefaciens VirB11
traffic ATPase. J Bacteriol 182: 4137–4145.
Sangwan, R.S., Bourgeois, Y., and Sangwan-Norreel, B.S.
(1991) Genetic transformation of Arabidopsis thaliana
zygotic embryos and identification of critical parameters
influencing transformation efficiency. Mol Gen Genet 230:
475–485.
Sangwan, R.S., Bourgeois, Y., Brown, S., Vasseur, G., and
Sangwan-Norreel, B. (1992) Characterization of competent
cells and early events of Agrobacterium-mediated genetic
transformation in Arabidopsis thaliana. Planta 188: 439–
456.
Sciaky, D., Montoya, A.L., and Chilton, M.D. (1978)
Fingerprints of Agrobacterium Ti plasmids. Plasmid 1:
238–253.
Shen, W.H., Escudero, J., Schlappi, M., Ramos, C., Hohn,
B., and Koukolikova-Nicola, Z. (1993) T-DNA transfer to
maize cells: histochemical investigation of beta-glucuronidase activity in maize tissues. Proc Natl Acad Sci USA
90: 1488–1492.
Smith, E.F., and Townsend, C.O. (1907) A plant tumour of
bacterial origin. Science 25: 671–673.
Stachel, S.E., and Zambryski, P.C. (1986) VirA and VirG
control the plant-induced activation of the T-DNA transfer
process of A. tumefaciens. Cell 46: 325–333.
Stachel, S.E., Messens, E., Vanmontagu, M., and Zambryski,
P. (1985) Identification of the signal molecules produced
by wounded plant cells that activate T-DNA transfer in Agrobacterium tumefaciens. Nature 318: 624–629.
Stachel, S.E., Nester, E.W., and Zambryski, P.C. (1986) A
plant-cell factor induces Agrobacterium tumefaciens vir
gene expression. Proc Natl Acad Sci USA 83: 379–383.
Stewart, G.R., Robertson, B.D., and Young, D.B. (2003)
Tuberculosis: a problem with persistence. Nat Rev Microbiol 1: 97–105.
Stover, E., Swartz, H., and Burr, T. (1997) Crown gall formation in a diverse collection of vitis genotypes inoculated
with Agrobacterium vitis. Am J Enol Vitic 48: 26–32.
Swart, S., Logman, T.J.J., Smit, G., Lugtenberg, B.J.J., and
Kijne, J.W. (1994) Purification and partial characterization
of a glycoprotein from pea (Pisum sativum) with receptor
activity for rhicadhesin, an attachment protein of Rhizobiaceae. Plant Mol Biol 24: 171–183.
Taiz, L., and Zeiger, E. (2002) Plant Physiology. Sunderland,
MA: Sinauer Associates.
Tzfira, T., Rhee, Y., Chen, M.H., Kunik, T., and Citovsky, V.
(2000) Nucleic acid transport in plant–microbe interactions:
the molecules that walk through the walls. Annu Rev Microbiol 54: 187–219.
Tzfira, T., Li, J., Lacroix, B., and Citovsky, V. (2004) Agrobacterium T-DNA integration: molecules and models.
Trends Genet 20: 375–383.
Wagner, V.T., and Matthysse, A.G. (1992) Involvement of a
vitronectin-like protein in attachment of Agrobacterium
tumefaciens to carrot suspension culture cells. J Bacteriol
174: 5999–6003.
Wang, L., and Winans, S.C. (1995) The 60 Nucleotide occR
operator contains a subsite essential and sufficient for
OccR binding and a 2nd subsite required for ligandresponsive DNA bending. J Mol Biol 253: 691–702.
Zhu, J., Oger, P.M., Schrammeijer, B., Hooykaas, P.J.J.,
Farrand, S.K., and Winans, S.C. (2000) The bases of
crown gall tumorigenesis. J Bacteriol 182: 3885–3895.
Zhu, Y., Nam, J., Humara, J.M., Mysore, K.S., Lee, L.Y., and
Cao, H. (2003) Identification of Arabidopsis rat mutants.
Plant Physiol 132: 494–505.
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Agrobacterium colonization of unwounded plants 1531
Zupan, J., Muth, T.R., Draper, O., and Zambryski, P. (2000)
The transfer of DNA from Agrobacterium tumefaciens into
plants: a feast of fundamental insights. Plant J 23: 11–28.
Supplementary material
The following supplementary material is available for this
article online:
© 2005 Blackwell Publishing Ltd, Molecular Microbiology, 57, 1522–1531
Fig. S1. Gas chromatograms of tobacco (A) and squash (C)
seedling exudates, and mass spectra (B, D and E) of virinducing phenolic compounds detected in the exudates.
Table S1. Expression of virB–lacZ of strain A136(pCH114)
(pCH116) (Chang et al., 1996) in response to different concentrations of AS, and in response to exudates of seedlings
of tobacco (Nicotiana tabacum cv. Samsun NN), tomato
(Lycopersicum esculentum cv. Heinz 1370) and squash
(Cucurbita maxima).