Plant Science 162 (2002) 733 /743
www.elsevier.com/locate/plantsci
Effects of phenolic compounds on Agrobacterium vir genes and gene
transfer induction* a plausible molecular mechanism of phenol
binding protein activation
/
Philippe Joubert a, Daniel Beaupère b, Philippe Lelièvre b, Anne Wadouachi b, Rajbir
S. Sangwan a,*, Brigitte S. Sangwan-Norreel a
a
UPRES EA 2084, Laboratoire Androgenèse et Biotechnologie, Faculte des Sciences, Université de Picardie Jules Verne, 33 rue Saint-Leu, 80039
Amiens Cedex, France
b
Laboratoire de Chimie Organique, Université de Picardie Jules Verne, 33 rue Saint-Leu, 80039 Amiens Cedex, France
Received 31 October 2001; received in revised form 8 January 2002; accepted 8 January 2002
Abstract
We have studied the effects of nine new phenolic compounds: 4-(4-hydroxy-3-methoxyphenyl)-but-3-en-2-one 4, 4-(4-hydroxy-3methoxyphenyl)-but-3-en-2-one 5, 4-hydroxy-3,5-dimethoxycinnamyl alcohol or sinapyl alcohol 6, 4-hydroxy-3,5-dimethoxy-N ,N dimethylcinnamamide 7, N -ethyl-4-hydroxy-3,5-dimethoxycinnamamide 8, 1-(4-hydroxy-3,5-dimethoxyphenyl)-3-(4-hydroxy-3,5dimethoxyphenyl)-prop-2-en-1-one 11, 1-(4-hydroxyphenyl)-3-(4-hydroxyphenyl)-prop-2-en-1-one 12, 4-(4-hydroxy-3,5-dimethoxyphenyl)-butan-2-one 13, and 1-(4-hydroxy-3,5-dimethoxyphenyl)-3-(4-hydroxyphenyl)-propan-1-one 14, on Agrobacterium
virulence gene induction, on Agrobacterium -mediated gene transfer, and on both transient and stable transformation rates on
Petunia and tobacco. We confirmed that virulence induction and transformation rates are increased by the use of phenolic vir
inducers bearing an unsaturated lateral chain. However, we found that the presence of at least one ortho-methoxy group in the
phenolic compounds is required to increase the vir gene induction. Moreover, we synthesized phenolic compounds 13 and 14 with a
saturated lateral chain from the corresponding compounds 4-(4-hydroxy-3,5-dimethoxyphenyl)-but-3-en-2-one 2 and 1-(4-hydroxy3,5-dimethoxyphenyl)-3-(4-hydroxyphenyl)-prop-2-en-1-one 9, in order to stop conjugation between hydroxyl and carbonyl groups
in these molecules. These compounds showed a lower efficiency of both vir induction and gene transfer. Furthermore, we also
synthesized two amides derivatives 7 and 8 from sinapinic acid. The nature of alkyl groups on nitrogen is essential to the vir
induction and gene transfer, and we observed that N -dimethylamide compound 7 is less active than N -ethyl compound 8. This may
be due to the difference of the electron-donating effect between methyl and ethyl groups. We present a model for the molecular
mechanism of VirA activation by phenols. # 2002 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Petunia (Petunia hybrida ); Tobacco (Nicotiana tabacum ); Agrobacterium tumefaciens ; Vir gene induction; Transformation induction;
VirA; Phenolic compounds
1. Introduction
Among the different techniques used for the introduction of foreign genes in plants, Agrobacterium
tumefaciens -mediated transfer remains the most popular
and a powerful one [1 /4]. It is accepted that Agrobacterium- mediated transfer requires the activation of two
Agrobacterium gene families: chv and vir genes. The
* Corresponding author. Tel.: 33-3-22-82-76-49; fax: 33-3-2282-76-12.
E-mail address: [email protected] (R.S. Sangwan).
chv */chromosomal virulence*/genes are involved in
the recognition and immobilization of the bacteria on
the epidermal plant cell surface [5]. The vir genes are
located on an extra-chromosomal DNA replicon, the
Tumour-inducing plasmid (Ti). The induction of vir
genes leads to the transfer of a part of the Ti plasmid:
the transferred-DNA (T-DNA). Comprehensive reviews
on the DNA transfer process from the bacteria to the
plant cells are available [4,6,7].
The vir genes are specifically induced by phenolic
compounds. First, intermediates of lignin synthesis or
phenolic precursors, such as acetosyringone 1 (AS) and
0168-9452/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 1 6 8 - 9 4 5 2 ( 0 2 ) 0 0 0 1 2 - 2
734
P. Joubert et al. / Plant Science 162 (2002) 733 /743
hydroxy-AS [8,9], are chemoattractants at very low
concentrations, but become vir inducers at high concentrations [10 /13]. A variety of other phenols have
been described and their vir gene induction abilities
investigated [14 /16]. Furthermore, other families of
phenolic compound such as hydroxycinnamides are
known to act as vir gene inducers [17]. It has also
been reported that monosaccharides such as D-glucose,
D-galactose and D-xylose, which are ‘long distance’
attractors of Agrobacteria, work in synergy with phenolic compounds in order to increase vir gene induction
[18,19] and enlarge the recognition profiles of sensitive
Agrobacteria strains to phenolic compounds [20].
Furthermore, opines [21] and flavonoid compounds
[22] may be involved in vir gene induction.
It was demonstrated [23] that if the two basic elements
required for phenol-mediated induction of virulence
gene expression, VirA and VirG are encoded on the Ti
plasmid, they are dependent on the chromosomal background for even the very first stage for signal perception.
In the vir gene induction process, the first stage is the
recognition of phenols by a transmembrane VirA
protein encoded by the virA gene [24 /26]. The recognition and autophosphorylation of VirA [27] leads to the
phosphorylation of the VirG cytoplasmic protein [28]
which, in turn, induces the expression of other vir genes.
Moreover, a direct molecular mechanism which may
explain the action of AS on VirA was put forward
[16,29,30]. It was suggested that an acidic VirA residue
may protonate the carbonyl group of phenolic compounds. The electron resonance delocalization leads to
an electron transfer from the phenol to the acidic
residue. Then the phenol can protonate a VirA basic
residue, leading to the autophosphorylation of VirA and
then to the phosphorylation of VirG.
It was suggested [31] that the VirA protein may not be
directly involved in the bonding of phenolic compounds.
Indeed, two chromosome-encoded proteins (p10 and
p21) were discovered which could act as receptors of
phenols and mediate interaction between phenolics and
VirA. Turk et al. [32] demonstrated that the periplasmic
domain of VirA does not react with AS, whereas its
cytoplasmic domain (aa283 /aa304) is able to sense the
signal whenever AS is present in the medium. Recently,
Dyé and Delmotte [33] discovered another Agrobacterium-phenol binding protein, Pbp39, but whether this
new protein is a vir gene inducer has not yet been
determined.
In order to understand the molecular interactions
between phenolic compounds and the VirA complex
protein or any other receptor proteins such as p10 and
p21, we have developed new phenolic compounds and
tested their effects on vir induction. Furthermore, these
new phenols were tested on tobacco and Petunia
transformation and, as the results compared favourably
with their vir induction, this has enabled us to suggest a
mechanism for possible molecular interactions between
phenolics and VirA complex protein site.
2. Material and methods
2.1. Chemical structures, synthesis and description
The structures and details of different chemical
functions used are reported on Fig. 1 and Table 1.
Simple phenolic compounds 1, 3 and 6 were purchased
from Aldrich (Strasbourg, France). Typical experimental procedures for synthesis of all other derivatives.
2.1.1. General procedure for synthesis of benzalacetones
(2, 4, 5) and chalcones (9/12)
An ether solution of borontrifluoride (1 eq.) and
ketone (R /CO /CH3: 2 eq.) were added successively
dropwise to a solution of benzaldehyde derivative
(HO,R1,R2-PhCHO: 8.2 /10 3 mol) in CH2Cl2 (10
ml). The resulting solution was stirred for 24 h at
room temperature, then neutralized with NaOAc and
extracted with CH2Cl2. The combined extracts were
washed with water, dried over Na2SO4 and filtered. The
organic layers were concentrated under reduced pressure
and the crude product purificated by column chromatography on silica gel (n-hexane-etOAc). a,b-ethylenic
ketones were obtained in 60/75% yield. The same
aldolisation procedure was applied to acetophenone
derivative (HO,R1,R2 /PhCOCH3) for synthesis of a,bethylenic ketone 9.
2.1.2. General procedure for ketones 13 and 14
They were obtained by the hydrogenation of a,bethylenic ketones 2 and 9 [34]. In this procedure, a
toluene solution of unsaturated compound 2 and 9 (0.25
mol l 1) and 1-phenylethanol (0.25 mol l 1) and hydridotetrakis (tris tri-phenylphosphine) rhodium I [35]
(RhH(PPh3)4: 4.10 2 eq.) was heated at 50 8C for 2 h.
After solvent evaporation, the purification by chromatography on silica gel led to products 13 and 14 in 92
and 96% yield respectively.
2.1.3. General procedure for amides 7 and 8
N ,N?-dicyclohexylcarbodiimide (1.09 eq.) was added
to a suspension of 4-hydroxy-3,5-dimethoxycinnamic
acid in dry CH2Cl2 under N2 atmosphere. The resulting
suspension was stirred for 30 min at room temperature.
Then a solution of amine (dimethylamine 0/7 or
ethylamine 0/8) 2M in THF (1.2 eq.) was added dropwise. The mixture was stirred at room temperature for
12 h. The dicyclohexylurea (DCU) was filtered off,
washed with CH2Cl2 and the filtrate was concentrated
under reduced pressure. The purification of the residue
by flash chromatography on silica gel (EtOAc, EtOAc/
P. Joubert et al. / Plant Science 162 (2002) 733 /743
735
Fig. 1. General structures of tested phenolic compounds, which were AS, benzalacetones and chalcones derived. Arrows indicate ethylenic bonds,
which were hydrogenated on the 13 and 14 compounds.
EtOH 14:1) led to compounds 7 or 8 as yellow solids in
80 and 82% yield respectively.
2.1.4. Physical data
The physical characteristics of all known compounds
were in perfect accordance with the literature values. All
unknown compounds were characterised by elementary
analysis and NMR Spectroscopy.
2.2. Plant material
Leaves of Petunia hybrida and Nicotiana tabacum
were cultured on a basal medium (MS) that contained
salts and vitamins of Murashige and Skoog [36], 3% (w/
v) sucrose and 0.8% Difco agar, pH adjusted to 5.6
before autoclaving (120 8C for 20 min). The cultures
were maintained in a culture room at 27 8C under a
light regime of fluorescent light (:/32 E m 2 s 1) 16 h
a day.
2.3. Bacterial strains
A. tumefaciens A348 pSM219 virH ::lacZ (pinF ::lacZ )
was used to evaluate vir gene induction. It was kindly
supplied by B. Hohn (FMI, Basel, Switzerland). A.
tumefaciens EHA105 (pG3.3) carrying a binary vector
with NPTII and GUS genes was kindly provided by S.
Hamon-Poirier (CNRS Gif/Yvette, France). Bacteria
Table 1
Description of various phenolic compounds tested
No.
R1
R2
R3
Usual name
1
OCH3
OCH3
COCH3
Acetosyringone (AS)
Benzalacetones derived
2
3
4
5
6
7
8
OCH3
OCH3
OCH3
H
OCH3
OCH3
OCH3
OCH3
OCH3
H
H
OCH3
OCH3
OCH3
COCH3
CO2H
COCH3
COCH3
CH2OH
CON(CH3)2
CONHCH2CH3
/
Sinapinic acid
/
/
Sinapyl alcohol
/
/
Chalcones derived
No.
R1
R2
R1?
R2?
9
10
11
12
H
OCH3
OCH3
H
H
OCH3
OCH3
H
OCH3
H
OCH3
H
OCH3
H
OCH3
H
Reduced compounds
13
14
Same 10 with reduced ethylenic bond
Same 12 with reduced ethylenic bond
Showing the substitutions at R1, R2, R1?, R2? and R3. Compounds 1, 2, 3, 9 and 10: Joubert et al. [40], and this work. Compounds 4, 5, 6, 7, 8, 11, 12,
13 and 14: this work.
736
P. Joubert et al. / Plant Science 162 (2002) 733 /743
colonies were picked from selection plates (Luria Broth
medium/carbenicillin 100 mg/l (A348) or kanamycin
100 mg/l (EHA105)) and grown overnight in Luria
Broth liquid medium [37] with the appropriate antibiotics added, at 28 8C, in a rotary shaker (210 rpm).
éales et des Fourrages (Paris, France). In each case, we
proceeded to a variance analysis. This test gave us
probable differences between phenolic treatments. Additional investigations with a Newman /Kuels test
classified these treatments in decreasing order with a
95% safety margin.
2.4. Determination of vir gene expression
To test vir gene induction, an overnight colony of
Agrobacterium A348 (pSM219) virH ::lacZ was centrifuged and suspended in 15 ml of mannitol-phosphate
buffer (0.6 M mannitol 0.01 M NaPi, pH 5.2). A 100 ml
aliquot of this bacterial suspension was added to a 900
ml MSPS liquid medium (MS medium supplemented
with 62.5 mM NaPi and 3% sucrose, pH 5.25), to which
the phenolic compound to be tested was also added, and
grown for 18 h at 28 8C. The expression of vir genes
was monitored through the measurement of b-galactosidase activity of the fusion gene. Bacterial cells were
harvested and b-gal activity [37] determined as described
by Sambrook et al. [38].
2.5. Inoculation of explants
Young leaves were collected from 2 months-old
tobacco and petunia plants. Leaf pieces (1 cm2) were
dipped into a suspension EHA105 strain (1 ml overnight
culture of bacteria/9 ml of MS medium) for 20 min,
blotted dry on sterile paper, and incubated on MS basal
medium with the appropriate phenolic compound for 3
days. Cocultured explants were washed in liquid MS
medium in the presence of 500 mg/l cefotaxim and
blotted dry on sterile paper. They were then plated on
MS medium containing 300 mg/l cefotaxim, 250 mg/l
kanamycin and: (i) 4.44 mM BAP/0.57 mM IAA for
tobacco plants; and (ii) 4.44 mM BAP/2.68 mM NAA
for petunia plants, for direct selection and regeneration
of transformed callus and buds.
2.6. Transient and stable transformation results
Transient transformation was assessed 3 days after
coculture. Leaf explants were ground and extracts used
to measure b-glucuronidase activity, as described by
Jefferson [39]. Four or five weeks after coculture,
regenerated transformed (kanamycin resistant) callus
and buds of tobacco and petunia, respectively, were
counted. The average number of regenerates per explant
was then established. Forty five explants were cocultured per series and all experiments repeated three times.
2.7. Statistical studies
All experiments were controlled through statistical
analysis carried out with the STAT-ITCF computer
program provided by the Institut Technique des Cér-
3. Results
3.1. Action of benzalacetones and chalcones on vir gene
expression and transformation rates
Benzalacetone derivatives 2, 4 and 5, which differ by
the number of methoxy groups, have been tested (Table
1). As reported in our previous work [40], we noted (Fig.
2) that a compound without methoxy groups, benzalacetone 5, does not induce vir functions, whereas the
presence of one of them, in benzalacetone 4, leads to less
active (2/5, 3/2) vir inducers. Similarly, the chalcone
derivative 12, lacking methoxy group as compared to
the structure of compound 11, does not induce vir genes,
whereas 11 had an effect on the vir induction similar to
AS effect.
However, the tetra-methoxylated compound 11
showed a significant decrease of vir induction ability
compared to its analogue derivatives 9 and 10. We
suggest that the presence of four methoxy groups
increases the molecular bulk of the phenolic molecule
and consequently reduce its affinity to the VirA receptor
site.
In all cases, we noted that activity of 2, 9 and 10 on vir
induction was very high and greater than AS. From the
hypothesis of Hess et al. [16], the conjugated structures
led to a delocalization of electrons by mesomeric effect
on the whole molecule (Fig. 3). In comparison with AS,
the additional double bond on R3 alkyl chain should
contribute to a better stabilisation of the mesomeric
form, thereby allowing the phenolic proton to leave
more easily and then activate phenol binding protein
before speeding up the VirA.
Following tests of these compounds as vir inducers in
Agrobacterium , we tested them for tobacco transformation efficiency. Most compounds displayed positive
effects on transformation results (Fig. 4). Phenolics 2,
9 and 1 (AS) appeared to be the strongest vir inducers
with the production of twice as many transformed callus
than in control experiment (control: leaf explants
cocultured without any phenolic) while compound 3,
of a limited vir induction ability, did not enhance the
transformation rate. These results confirm the importance of these compounds and their effects, on the
virulence of Agrobacterium , and hence on T-DNA
transfer.
We also checked the GUS gene expression by
quantification of b-glucuronidase in order to evaluate
P. Joubert et al. / Plant Science 162 (2002) 733 /743
737
Fig. 2. Action of phenolics on virH gene. All structures were tested at a concentration of 100 mM. Values are average (n 9)9s.e. Probability
showing differences between media was equal to 1.00 and Newman /Kuels test classified medium effects in five classes noted A to E.
Fig. 3. Induction model of phenol binding protein site (VirA) with a
electronic transfer from basic to acidic residue (proposed by Hess et al.
[16]).
the relationship between vir gene induction, and transient and stable gene transfer. This experiment was
aimed at linking vir induction results (Fig. 2) to stable
transformation occurrences (Fig. 4) generated by the
same molecules. Fig. 5 shows that the most efficient vir
inducers molecules (2, 9, 1 (AS), 10 and 11) also result in
a high gene transfer frequency in Petunia and tobacco.
The gradient of gene transfer and stable transformation
seems to follow vir gene induction. Nevertheless, we
noted the particularly high level of gene transfer induced
by compound 9 on Petunia . Conversely, we tested a non
vir -inducing compound in order to find out whether it
may also induce a low frequency of gene transfer on
Petunia and tobacco. The molecule 5 tested showed a
Fig. 4. Influence of phenolics on tobacco transformation. The number of callus regenerated per leave explant after 3 weeks’ culture on selection
medium is represented. Explants had been cocultured with bacteria for 3 days in the presence of 100 mM tested phenolic compound. The control was
a cocultured medium without phenolic. Values are average (n 6)9s.e. Probability showing differences between media was 0.95 and Newman /
Kuels test classified the phenolic effect in two homogeneous groups noted A and B.
738
P. Joubert et al. / Plant Science 162 (2002) 733 /743
Fig. 5. Action of phenolic compounds on Agrobacterium -mediated gene transfer. Transient expression of GUS gene in leaf explants of Petunia and
tobacco after 3 days of A. tumefaciens cocultivation, in the presence of 100 mM of the tested phenolic compounds. The control represents GUS
activity in non-cocultivated explants; the control is GUS activity in explants cocultivated without any phenolic. GUS activities are measured
kinetic (nm) apparition of fluorescent 4-MU substrate per minute and per gram of fresh weight. Values are average (n 3)9s.e. Probability showing
differences between media was equal to 1.00 and Newman /Kuels test classified the phenolic effect in six homogeneous Petunia groups and five
Tobacco groups noted A, B, C, D, E and F.
total inability to induce gene transfer and, furthermore,
a low inhibition of this transfer, a feature presumably
related to the general toxicity of phenols.
3.2. Action of hydrogenated benzalacetones and
chalcones on vir gene and transformation induction
In relation to the hypothesis of Hess et al. [16], we
synthesized molecules 13 and 14 and tested them on vir
gene induction. The double bonds of compounds 2 and
9 were treated under catalytic hydrogenation conditions
to give products 13 and 14. According to the above
mentioned hypothesis, the electron transfer between the
carbonyl and phenolic groups of molecules 13 and 14
was not possible. The acidity of the hydroxyl group was
only enhanced by attractive effect of the phenyl group
(Fig. 6). As regards vir gene induction, compounds 13
and 14 showed a limited efficiency compared to
compounds 2 and 9, respectively (Fig. 7). Such reduced
activities accounted for 1/2 and 1/3 of that of the
original products, but remained much higher than the
activities observed in the control experiment.
Fig. 6. Example of electronic delocalization which can affect the aromatic cycle of phenolic compounds and explains the vir activity of compound 13.
P. Joubert et al. / Plant Science 162 (2002) 733 /743
739
Fig. 7. Action of hydrogenated compounds */13 and 14 */on virH gene induction. Compounds were tested at 100 mM. Values are average
(n 9)9s.e. Probability showing differences between media was equal to 1.00 and Newman /Kuels test classified medium effects in five classes noted
A to E.
We tested these four molecules 2, 13, 9 and 14 in the
presence of Agrobacteria throughout cocultivation with
explants of Tobacco and Petunia . After a few weeks on
selection medium, we calculated that explants cocultivated with 13 and 14 showed a significant decrease in the
regenerated callus number. Therefore, compound 14 did
not enhance the transformation induction. However,
explants cocultivated with 13 maintained a higher level
of transformation than the control.
From both the molecular mechanism suggested by
Hess et al. [16] and the above results, we believe that the
following hypothesis is well-founded. On the one hand,
a non-conjugated lateral chain R3 might explain the
relative decrease in the vir and transformation induction
(Fig. 7 and Fig. 8). On the other hand, the hydrogenation of this chain changes the initial planar structure of
the molecule and induces, in comparison with 2 and 9, a
new freedom of movement for R3. Therefore, the
Fig. 8. Influence of phenolics 13 and 14 on Petunia and Tobacco transformation. The number of callus regenerated per leaf explant after 3 weeks’
culture on selection medium for Tobacco and 4 weeks for Petunia is represented. Explants had been cocultured with bacteria for 3 days in the
presence of 100 mM tested phenolic compound. The control was a cocultured medium without phenolic. Values are average (n 6)9s.e. Probability
showing differences between media was 0.95 and Newman /Kuels test classified phenolic effect in three homogeneous groups noted A, B and C.
740
P. Joubert et al. / Plant Science 162 (2002) 733 /743
Fig. 9. Action of nitrogenated compounds */7 and 8 */on virH gene induction, compared to the corresponding benzalacetones. Compounds were
tested at 100 mM. Values are average (n 6)9s.e. Probability showing differences between media was equal to 1.00 and Newman /Kuels test
classified medium effects in five classes noted A, B, C, D and E.
probability of simultaneous recognition of both the
chemical phenol and carbonyl functions in the VirA site
decreased.
3.3. Action of amino compounds on vir and
transformation induction
It was reported by Dyé et al. [41,42] that different
syringamides are new vir inducers. These compounds
appeared to have 2 /20 times stronger action than the
syringic acid, depending on the Agrobacterium lacZ
fusion strains. We decided to test amides of the sinapinic
acid 3 as vir inducers. So, we synthesized two amide
derivatives of 3: the 4-hydroxy-3,5-dimethoxy-N ,N dimethylcinnamamide 7 and the N -ethyl-4-hydroxy3,5-dimethoxycinnamamide 8.
Fig. 9 shows the vir induction of these two products (7
and 8) compared with the analogous cinnamic derivatives: 2, 3 (sinapinic acid) and 6 (sinapyl alcohol) and
with the two controls (negative control without any
phenolic, and positive control with AS in the same
concentration as all other phenolics, 100 mM). In both
cases the induction ratios of 3 (sinapinic acid) and 6
(sinapyl alcohol) compared with 2 were equal to 0.6.
Compound 8 had the same effect as compound 3 (90%
of it). Compound 7 showed a lower vir induction
activity: 43% that of compound 3 and 26% that of
compound 2. The activity of the two amides did not
reach that of 3 as expected, and as reported by Dyé et al.
[41] for the ethyl amide of syringic acid. However, we
noted that the replacement of the two methyl groups of
the nitrogen atom (7) by only one ethyl chain (8)
induced an increase in the vir gene induction (8/7 /2).
The spatial size of the dimethyl groups on nitrogen (7) is
higher than in the ethyl group (8). This may explain why
it is so difficult for the carbonyl group to reach the VirA
acidic receptor site.
We noted that compound 6 has a higher vir action
while its R3 chain has an alcohol function and not a
carbonyl function. This molecule could be oxidized by
the oxygen in the medium and transformed into
aldehyde, then acid, which are reliable vir inductors.
4. Discussion
Hess et al. [16] first suggested a hypothetical direct
mechanism of VirA activation, when the phenolic
molecule is placed in the receptor protein site VirA.
Another hypothetical mechanism implies mediated-protein such as p10 and p21 or even the Pbp39 [31,33],
which leads to an indirect mechanism of VirA activation. Then, transfer of phenolic proton to the protein
basic residue and transfer to the phenol carbonyl group
from the protein acidic residue proton, are possible
through the conjugation of the molecules.
Here we confirmed that the induction of the Agrobacterium vir genes was essential for an effective gene
Table 2
Comparison of phenolic pKa for some of the tested molecules on
Agrobacteria virulence with their vir induction rates
Phenolic compounds
Phenol pKa
vir induction
3
7.91
1 (AS)
8.04
10
9.81
2
9.94
P. Joubert et al. / Plant Science 162 (2002) 733 /743
741
Table 3
Comparison of inductor and mesomeric donor effects of R radicals (Fig. 5) for some of the tested molecules on Agrobacteria virulence, which all have
ethylenic bonds on phenol para with their vir induction rates [43]
R radicals
Donor effects
Compounds
Vir induction
CH3
2
(C6H4)OH
10
transfer to the plants. Previously, we had shown that vir
active molecules must have two methoxy groups in
ortho position of phenolic function and one carbonyl
group on the R3 chain [40].
The presence of strong active vir gene phenolic
molecules such as AS 1, benzalacetone 2 and chalcones
9, 10 and 11, increased the gene transfer in Tobacco and
Petunia , and transformation rates 2 /3 times higher than
the control. Therefore, we confirmed that phenolic
molecules are essential in the T-DNA transfer process.
The presence of the double bonds in the phenolic
compounds enhanced the molecular conjugation and
enable the electronic transfer from phenol to carbonyl
and finally activated the protein receptor (Fig. 3).
According to the above hypothesis of Hess et al. [16],
it is suggested that leaving-ability of the phenolic proton
should be correlated with the vir induction power of
molecules. In order to find out if the ability of ionisation
of phenol function was the only induction factor on the
VirA site, we calculated various pKa of some tested
molecules and compared them to their vir induction
power. The values were predicted according to the
method described by Perrin et al. [43]. We check this
method experimentally by measurement and calculation
pKa of AS and 2. Tables 2 and 3 shows the pKa and vir
induction of these molecules. The less active compounds
had the lowest pKa, and the pKa of molecules increased
along with the vir induction power. Thus, it seems that
effective induction of phenol binding (VirA) protein
decreases with the phenol dissociation.
From the above suggested model (Fig. 3), a R radical
can effect the induction power of phenols. This R radical
might actively have the donor electronic effect, antagonist to the electronic flow from the phenol donor
residue. Thus, we compared donor inductor and mesomeric effects of some R radicals with the vir induction
powers of the corresponding molecules (Tables 2 and 3).
For example, we noted that the compound 2, with a high
vir induction, has a weak electron-donating group (/
CH3) whereas the amide 7 with a high electron-donating
effect (/N(CH3)2) has a low vir activity. Moreover, the
literature showed [14] that three methyl esters of ferulic,
sinapinic and syringic acids have a far higher action on
vir genes than the corresponding free acids, but in this
OH
3
NHEt
8
()
N(Me)2
7
case the acidity of the acid group was hidden by the
esterification.
From the comparison of pKa of phenol derivatives
with the vir induction of some molecules (i), comparing
the electron-donating effects with the vir induction (ii),
and seeing that there is no possible conjugation between
hydroxyl and carbonyl groups in compound 13 (iii), we
believe that the model of phenol binding protein
molecular activation proposed by Hess et al. [16] is
highly significant. However, the physical distance between the phenol and carbonyl functions accounts for
20% with the two active compounds AS 1 and 2 (5.9 and
7.5 Å respectively). In this case, we cannot easily
imagine that there should be only one phenol receptor
type. Recently an A. tumefaciens protein surface*/
Pbp39 */which binds phenols was discovered [33]. We
also know that the two proteins p10 and p21 have
similar properties. Moreover, phenols of flavones class,
such as kaempferol [22] showed good vir induction.
These data suggest that VirA may be composed as a
protein complex made of several phenol receptors, each
of them specialized with one phenol class indirectly
inducing VirA.
Finally, we report here that the three compounds with
the most important conjugation fields (2, 9, and 10), also
have the higher vir induction effect. Their great stability
could result in an easier electron delocalization on the
whole molecule, and therefore an easier (more efficient)
VirA activation.
Acknowledgements
We thank B. Hohn and A. de Laat for generous gifts
of A. tumefaciens strains; Y. Dessaux and S. Millam for
critical reading and improving the English style of the
manuscript.
References
[1] P.J.J. Hooykaas, R.A. Schilperoort, Agrobacterium and plant
genetic engineering, Plant Mol. Biol. 19 (1992) 15 /38.
[2] P.J.J. Hooykaas, A.G.M. Beijersbergen, The virulence system of
Agrobacterium tumefaciens , Annu. Rev. Phytopathol. 32 (1994)
157 /179.
742
P. Joubert et al. / Plant Science 162 (2002) 733 /743
[3] B. Tinland, The integration of T-DNA into plant genomes,
Trends Plant Sci. 1 (1996) 178 /184.
[4] J. Zupan, T.R. Muth, O. Draper, P. Zambryski, The transfer of
DNA from Agrobacterium tumefaciens into plants: a feast of
fundamental insights, The Plant J. 23 (2000) 11 /28.
[5] C.J. Douglas, W. Halperin, E.W. Nester, Agrobacterium tumefaciens mutants affected in attachment to plants cells, J. Bacteriol.
152 (1982) 1265 /1269.
[6] G. Kahl, K. Weising, Genetic engineering of plant cells, in: H.J.
Rehm, G. Reed (Eds.), Biotechnology, genetic fundamentals and
genetic engineering, vol. 2, 1993, pp. 547 /625.
[7] K. Weising, G. Kahl, Natural genetic engineering of plant cells:
the molecular biology of crown gall and hairy root disease, World
J. Microbiol. Biotech. 12 (1996) 327 /351.
[8] S.E. Stachel, E. Messens, M. Van Montagu, P.C. Zambryski,
Identification of the signal molecules produced by wounded plant
cells that activate T-DNA transfer in Agrobacterium tumefaciens ,
Nature 318 (1985) 624 /629.
[9] S.E. Stachel, P.C. Zambryski, virA and virG control the plantinduced activation of the T-DNA transfer process of Agrobacterium tumefaciens , Cell 46 (1986) 325 /333.
[10] A.M. Ashby, M.D. Watson, C.H. Shaw, A Ti-plasmid determined
function is responsible for chemotaxis of Agrobacterium tumefaciens towards the plant wound product acetosyringone, FEMS
Microbiol. Lett. 41 (1987) 189 /192.
[11] G.J. Loake, A.M. Ashby, C.H. Shaw, Attraction of Agrobacterium tumefaciens C58C1 towards sugars involves a highly sensitive
chemotaxis system, J. Gen. Microbiol. 134 (1988) 1427 /1432.
[12] N. Shimoda, A. Toyoda-Yamamoto, J. Nagamine, S. Usami, M.
Katayama, Y. Sakagami, Y. Machida, Control of expression of
Agrobacterium vir genes by synergestic actions of phenolic signal
molecules and monosaccharides, Proc. Nat. Acad. Sci. USA 87
(1990) 6684 /6688.
[13] C.H. Shaw, Swimming against the tide: chemotaxis in Agrobacterium, Bioassays 13 (1991) 25 /29.
[14] P.A. Spencer, G.H. Towers, Specificity of signal compounds
detected by Agrobacterium tumefaciens , Phytochemistry 27 (9)
(1988) 2781 /2785.
[15] L.S. Melchers, A.J.G. Regensburg-Tuink, R.A. Schilperoort,
P.J.J. Hooykaas, Specificity of signal molecules in the activation
of Agrobacterium virulence gene expression, Mol. Microbiol. 3
(1989) 969 /977.
[16] K.M. Hess, M.W. Dudley, D.G. Lynn, R.D. Joerger, A.N. Binns,
Mechanism of phenolic activation of Agrobacterium virulence
genes: development of a specific inhibitor of bacterial sensor/
response systems, Proc. Natl. Acad. Sci. USA 88 (1991) 7854 /
7858.
[17] K. Berthelot, D. Buret, B. Guérin, D. Delay, J. Negrel, F.M.
Delmotte, Vir -gene-inducing activities of hydroxycinnamic acid
amides in Agrobacterium , Phytochemistry 49 (1998) 1537 /1548.
[18] G.A. Cangelosi, R.G. Ankenbauer, E.W. Nester, Sugars induce
the Agrobacterium virulence genes through a periplasmic binding
protein and a transmembrane signal protein, Proc. Natl. Acad.
Sci. USA 87 (1990) 6708 /6712.
[19] M.-L.W. Huang, G.A. Cangelosi, W. Halperin, E.W. Nester, A
chromosomal Agrobacterium tumefaciens gene required for effective plant signal transduction, J. Bacteriol. 172 (1990a) 1814 /
1822.
[20] W.-T. Peng, Y.-W. Lee, E.W. Nester, The phenolic recognition
profiles of the Agrobacterium tumefaciens VirA protein are
broadened by a high level of the sugar binding protein Chv E, J.
Bacteriol. 180 (1998) 5632 /5638.
[21] K. Veluthambi, M. Krishnan, J.H. Gould, R.H. Smith, S.B.
Gelvin, Opines stimulate induction of the vir genes of the
Agrobacterium tumefaciens Ti plasmid, J. Bacteriol. 171 (1989)
3696 /3703.
[22] R. Zerback, K. Dressler, D. Hess, Flavonoid compounds from
pollen and stigma of Petunia hybrida: inducers of the vir region of
the Agrobacterium tumefaciens Ti plasmid, Plant Science 62
(1989) 83 /91.
[23] A.M. Campbell, J.B. Tok, J. Zhang, Y. Wang, M. Stein, D.G.
Lynn, A.N. Binns, Xenognosin sensing in virulence: is there a
phenol receptor in Agrobacterium tumefaciens ?, Chem. Biol. 7
(2000) 65 /76.
[24] B. Leroux, M.F. Yanoksky, S.C. Winans, J.E. Ward, S.F. Ziegler,
E.W. Nester, Characterization of the virA locus of Agrobacterium
tumefaciens : a transtriptional regulator and host range determinant, EMBO J. 6 (1987) 849 /856.
[25] L.S. Melchers, D.V. Thompson, K.B. Idler, S.T.C. Neuteboom,
R.A. de Maagd, R.A. Schilperoort, P.J.J. Hooykaas, Molecular
characterisation of the virulence gene virA of the Agrobacterium
tumefaciens octopine Ti plasmid, Plant Mol. Biol. 11 (1988) 227 /
237.
[26] S.C. Winans, R.A. Kerstetter, J.E. Ward, E.W. Nester, A protein
required for transcriptional regulation of Agrobacterium virulence
genes spans the cytoplasmic membrane, J. Bacteriol. 171 (1989)
1616 /1622.
[27] Y. Huang, P. Morel, B. Powell, C.I. Kado, VirA, a corregulator
of Ti-specified genes is phosphorylated in vitro, J. Bacteriol. 172
(1990b) 1142 /1144.
[28] S. Jin, R.K. Prusti, T. Roitsch, R.G. Akenbauer, E.W. Nester,
Phosphorylation of the VirG protein of Agrobacterium tumefaciens by the autophosphorylated VirA protein: essential role in
biological activity of VirG, J. Bacteriol. 172 (1990) 4945 /4950.
[29] Y.-W. Lee, S. Jin, W.-S. Sim, E.W. Nester, Genetic evidence for
direct sensing of phenolic compounds by the VirA protein of
Agrobacterium tumefaciens , Proc. Natl. Acad. Sci. USA 92 (1995)
12245 /12249.
[30] Y.-W. Lee, S. Jin, W.-S. Sim, E.W. Nester, The sensing of plant
signal molecules by Agrobacterium : genetic evidence for direct
recognition of phenolic inducers by the VirA protein, Gene 179
(1996) 83 /88.
[31] K. Lee, M.W. Dudley, K.M. Hess, D.G. Lynn, R.D. Joerger,
A.N. Binns, Mechanism of activation of Agrobacterium virulence
genes: identification of phenol-binding proteins, Proc. Natl. Acad.
Sci. USA 89 (1992) 8666 /8670.
[32] S.C.H.J. Turk, R.P. van Lange, T.J.G. Regensburg-Tuı̈nk, P.J.J.
Hooykaas, Localization of the VirA domain involved in acetosyringone-mediated vir gene induction in Agrobacterium tumefaciens , Plant Mol. Biol. 25 (1994) 899 /907.
[33] F. Dyé, F.M. Delmotte, Purification of a protein from Agrobacterium tumefaciens strain A348 that binds phenolic compounds,
Biochem. J. 321 (1997) 319 /324.
[34] D. Beaupère, P. Bauer, R. Vrau, Réduction catalytique de cétones
a,b-éthyléniques à température modérée par RhH(Pf3)4, Can. J.
Chem. 57 (1979) 218 /221.
[35] J.J. Lewinson, J.D. Robinson, Transition-metal complexes containing phosphorus ligands. Part III. Convenient syntheses of
some triphenylphosphine complexes of the platinum metals, J.
Chem. Soc. A. (1970) 2947.
[36] T. Murashige, F. Skoog, A Revised Medium for Rapid Growth
and Bioessay with Tobacco Tissue Cultures, Physiol. Plant 15
(1962) 473 /497.
[37] J.H. Miller, Experiments in Molecular Genetics, Cold Spring
Harbor Laboratory Press, New York, 1972.
[38] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning, Cold
Spring Harbor Laboratory Press, New York, 1989.
[39] R.A. Jefferson, Assaying chimearic genes in plants: the GUS gene
fusion system, Plant Mol. Biol. Rep. 5 (1987) 387 /405.
[40] P. Joubert, R.S. Sangwan, M. El Arabi Aouad, D. Beaupère, B.S.
Sangwan-Norreel, Influence of phenolic compounds on Agrobacterium vir gene induction and onion gene transfer, Phytochemistry 40 (6) (1995) 1623 /1628.
P. Joubert et al. / Plant Science 162 (2002) 733 /743
[41] F. Dyé, D. Delay, K. Berthelot, F. Delmotte, Ethyl syringamide, a
new inducer of Agrobacterium tumefaciens virulence genes,
International Conference: ‘‘Agrobacterium and beyond’’ CNRS
Gif/Yvette, September 1995, 24 /26.
743
[42] F. Dyé, K. Berthelot, B. Griffon, D. Delay, F.M. Delmotte,
Alkylsyringamides, new inducers of Agrobacterium tumefaciens
virulence genes, Biochimie 79 (1997) 3 /6.
[43] D.D. Perrin, B. Dempsey, E.P. Serjeant, pKa predictions for Organic Acids and Bases, Chapman & Hall, London,
1981.