Structure-dependent modulation of a pathogen response in plants by synthetic O-antigen polysaccharides

Article abstract of DOI:10.1021/ja0452166

Many phytopathogenic bacteria display lipopolysaccharides (LPS) with the O-chain repeating unit [α-L-Rha-(1→3)-α-L-Rha-(1→3)- α-L-Rha-(1→2)]_n. This trisaccharide unit was synthesized and oligomerized to obtain hexa- and nonasaccharides. The deprotected rhamnans were effective in suppressing the hypersensitive response (HR) and in inducing PR-1 gene expression in Arabidopsis thaliana. Conformational analysis of the oligorhamnans by NMR spectroscopy and molecular dynamics calculations revealed that a coiled structure develops with increasing chain length of the oligosaccharide. This is associated with increasing efficacy in HR suppression and PR-1 gene expression. We therefore infer that the coiled structure of phytopathogenic bacteria is a plant-recognizable pathogen-associated molecular pattern (PAMP).

Full text of DOI:10.1021/ja0452166

Published on Web 02/04/2005
Structure-Dependent Modulation of a Pathogen Response in
Plants by Synthetic O-Antigen Polysaccharides
Emiliano Bedini,† Cristina De Castro,† Gitte Erbs,‡ Lorenzo Mangoni,†
J. Maxwell Dow,§ Mari-Anne Newman,‡ Michelangelo Parrilli,*,† and
Carlo Unverzagt|
Contribution from the Dipartimento di Chimica Organica e Biochimica Complesso,
UniVersitario Monte Sant’Angelo, Via Cintia 4, 80126, Napoli, Italy, Section for Plant
Pathology, Royal Veterinary and Agricultural UniVersity, ThorValdsensVej 40,
1871 Frederiksberg, Denmark, BIOMERIT Research Centre, Department of Microbiology,
BioSciences Institute, National UniVersity of Ireland, Cork, Ireland, and Bioorganische Chemie,
Geba¨ude NWI, UniVersita¨t Bayreuth, D-95440 Bayreuth, Germany
Received August 8, 2004; E-mail: parrilli@cds.unina.it
Abstract: Many phytopathogenic bacteria display lipopolysaccharides (LPS) with the O-chain repeating
unit [R-L-Rha-(1f3)-R-L-Rha-(1f3)-R-L-Rha-(1f2)]n. This trisaccharide unit was synthesized and oligo-
merized to obtain hexa- and nonasaccharides. The deprotected rhamnans were effective in suppressing
the hypersensitive response (HR) and in inducing PR-1 gene expression in Arabidopsis thaliana.
Conformational analysis of the oligorhamnans by NMR spectroscopy and molecular dynamics calculations
revealed that a coiled structure develops with increasing chain length of the oligosaccharide. This is
associated with increasing efficacy in HR suppression and PR-1 gene expression. We therefore infer that
the coiled structure of phytopathogenic bacteria is a plant-recognizable pathogen-associated molecular
pattern (PAMP).
Introduction
Rha-(1 f 3)-R-L-Rha-(1 f 2)] motif. An efficient synthetic
path was developed for this trisaccharide, which was oligomer-
The infection of crop plants by phytopathogenic bacteria can
either lead to disease, which may cause great economic losses
in the crop, or to a rapid activation of plant resistance responses
that contain or eliminate the pathogen. During infection, the
recognition of bacterial components by plants plays a funda-
mental role in the outcome of the interaction. Since lipopolysac-
charides (LPS)1 cover almost 80% of the cell surface of Gram-
negative bacteria, their role in bacterial interactions with
eukaryotic hosts is undoubtedly crucial. LPS are one of a group
of general elicitors that can be recognized by plants to trigger
defense responses. The mechanism of plant defense activation
by LPS is unknown but is suggested to be analogous to the
innate immunity system of animals,2 which is based on the
recognition of pathogen-associated molecular patterns (PAMPs),3
characteristic structures of the pathogen indispensable for its
growth within the host.
ized to hexa- and nonasaccharides.5
The deprotected oligosaccharides were tested for their ability
to induce PR-1 gene expression, a plant immune response, and
in suppressing the hypersensitive response (HR) in Arabidopsis
thaliana caused by an avirulent strain of Pseudomonas syringae.
These experiments were conducted in order to evaluate the role
of O-chain oligosaccharides in plant innate immunity. Informa-
tion regarding the secondary structure of the synthesized
oligosaccharide was obtained through NMR spectroscopy and
molecular modeling and correlated to the biological activities.
Results and Discussion
The synthesis of the three rhamnans was based on the central
trisaccharide 1,5 which was converted to a glycosyl acceptor
by selective removal of the chloroacetyl moiety or to a glycosyl
donor by activation as a trichloroacetimidate.6
The O-chain of the LPS from phytopathogenic bacteria
generally comprises a rhamnan backbone, with single monosac-
charide branches differentiating each serotype.4 In particular the
most frequent backbone consists of a [R-L-Rha-(1 f 3)-R-L-
Reaction of the trisaccharide acceptor with the trisaccharide
donor gave hexasaccharide 3, which was further elongated to
the nonasaccharide 5 (see Supporting Information). Zemple`n
deacylation of the protected oligorhamnans 1, 3, and 5 furnished
the desired oligosaccharides 2, 4, and 6 in good yields (com-
pounds 3 and 5 required higher reaction temperatures than those
† Universitario Monte Sant’Angelo.
‡ Royal Veterinary and Agricultural University.
§ National University of Ireland.
| Universita¨t Bayreuth.
(1) Caroff, M.; Karibian, D. Carbohydr. Res. 2003, 338, 2431-2447.
(2) Parker, J. Trends Plant Sci. 2003, 8, 245-247.
(3) Nu¨rnberger, T.; Brunner, F. Curr. Opin. Plant Biol. 2002, 5, 318-324.
(4) Corsaro, M. M.; De Castro, C.; Molinaro, A.; Parrilli, M. Recent Res. DeVel.
Phytochem. 2001, 5, 119-138
(5) Bedini, E.; Parrilli, M.; Unverzagt, C. Tetrahedron Lett. 2002, 43, 8879-
8882
(6) Bedini, E.; Barone, G.; Parrilli, M.; Unverzagt, C. Carbohydr. Res. 2004,
339, 393-400.
9
2414
J. AM. CHEM. SOC. 2005, 127, 2414-2416
10.1021/ja0452166 CCC: $30.25 © 2005 American Chemical Society

Modulation of a Pathogen Response in Plants
A R T I C L E S
Figure 2. 1H NMR comparison of the anomeric region of oligosaccharides
2, 4, and 6.
Figure 1. (a) NaOMe, MeOH, rt, 4 h, 74%. (b) NaOMe, MeOH, 40 °C,
16 h, 84%. (c) NaOMe, MeOH, 40 °C, 3 days, 88%; CA ) chloroacetyl.
trisaccharide 2 PR-1 mRNA increased 4-fold 20 h after treat-
ment, with a decrease at 24 h. This treatment was also only
moderately effective in prevention of HR. The hexasaccharide
4 displayed a complete HR suppression at 100 µg per mL and
a partial response at 50 µg per mL. PR-1 expression in response
to 50 µg per mL of compound 4 was already increased 20-fold
at 12 h after treatment and 29-fold at 20 h, the timespan where-
after the bacteria were introduced into the tissue for the HR
suppression test. Nonasaccharide 6 caused complete suppression
at both concentrations used and a very marked increase of the
PR-1 gene expression, to 52- and 86-fold at 12 h and 20 h after
treatment, respectively. Although it is already established that
LPS can prevent the HR response9,10 and induce plant defense-
related genes,9 nothing is known about the recognition mech-
anisms of LPS in plants. To our knowledge this is the first dem-
onstration that short oligosaccharides can induce defense related
gene expression and prevent HR in a size-dependent manner.
To probe whether the biological activity of the oligomers 2,
4, and 6 is based on structural features associated with the
growing chain length, we analyzed the oligosaccharides by NMR
and both molecular mechanics and dynamics calculations.
Comparison of the anomeric regions (Figure 2) revealed distinct
H-1 signals for the residues A and B of all compounds, most
likely due to the anomeric benzyl moiety.
When one repeating unit was added as in 4, the new anomeric
signals, D, E, and F appeared near the shifts of the O-chain
polysaccharide isolated from Pseudomonas syringae pv.
Coronafaciens,11 which is comprised of the same repeating unit
as the synthetic rhamnans, used in this study. This evidence
supports that a regular secondary structure is building up
together with the chain elongation process. On the basis of the
observed proton chemical shifts, this regular conformational
motif was present in compounds 4 and 6 but not in oligosac-
charide 2.
Table 1. Ability of Compounds 2 (Trisaccharide), 4
(Hexasaccharide), and 6 (Nonasaccharide) To Induce PR-1 Gene
Expression in Arabidopsis thalianaa
time after treatment (h)
2
4
6
12
20
24
+1.9 ns
+4.2 ***
no change
+20.3 ***
+29.9 ***
+28.7 ***
+52.6 ***
+86.8 ***
+26.9 ***
a +: fold upregulated compared to water treated tissue. After normaliza-
tion to 18SrRNA. ns: not significant. *** ) P < 0.001.
Table 2. Ability of Compounds 2 (Trisaccharide), 4
(Hexasaccharide), and 6 (Nonasaccharide) To Suppress the HR in
Arabidopsis thaliana Caused by Pst AvrRPM1a
compound
2
4
6
control (H2O)
50 µg/mL
100 µg/mL
+
+
++
+++
+++
+++
-
-
a +++: complete suppression in inoculated area. ++: suppression in
25-80% of inoculated area. +: suppression in <25%. -: no suppression.
for 1 and an unpolar cosolvent because of their reduced
solubility in methanol) (Figure 1).
The series of compounds 2, 4, and 6 were tested for their
ability to induce PR-1 gene expression (Table 1) and to modulate
the hypersensitive response (HR), a programmed cell death of
plants triggered by avirulent bacteria (Table 2). The trisaccharide
2, the hexasaccharide 4 or the nonasaccharide 6 were dissolved
in water (50 µg ml-1) and infiltrated into 6 week old leaves of
Col-O. A further set of leaves was inoculated with water. The
plants were placed in a growth cabinet at 25 °C with 16 h of
light. The leaves were harvested 12, 20, and 24 h after
inoculation. The changes in PR-1 gene expression were followed
using quantitative real-time RT-PCR analysis (see Supporting
Information).
For the HR suppression test, the oligosaccharide 2, 4, or 6
was infiltrated into 6 week old leaves of Col-O as aqueous
solutions (50 or 100 µg mL-1) and water as a control. After
20 h in a growth cabinet these leaves were inoculated with
Pst/aVrRPM1 in the pretreated and the control area, and
the elicitation of HR was monitored over 24 h (Table 2). HR
can be induced in Arabidopsis thaliana accession Columbia
(Col-O) after inoculation with Pseudomonas syringae pv.
tomato strain DC3000 carrying the avirulence gene aVrRPM1
(Pst/aVrRPM1)7 as described.8
Similarly, in compound 6, the three new residues G, H, and
I displayed proton resonances similar to those of the polysac-
charide and almost coincident to those of D, E, and F,
respectively, suggesting that residues in homologous position
in the sequence experience a similar environment.
(7) Debener, T.; Lehnackers, H.; Arnold, M.; Dangl, J. L. Plant J. 1991, 1,
289-302.
(8) Newman, M.-A.; von Roepenack-Lahaye, E.; Parr, A.; Daniels, M. J.; Dow,
J. M. Plant J. 2002, 29, 487-495.
(9) Dow, M.; Newman, M.-A.; von Roepenack-Lahaye, E. Annu. ReV.
Phytopathol. 2000, 38, 241-261
The panel of synthetic oligorhamnans showed a pronounced
chain-length-dependent induction of the PR-1 gene expression,
and in prevention of the HR in A. thaliana. In response to the
(10) Erbs, G.; Newman, M.-A. Mol. Plant Pathol. 2003, 4, 421-425
(11) Zdorovenko, E. L.; Zatonsky, G. V.; Zdorovenko, G. M.; Pasichnyk, L.
A.; Shashkov, A. S.; Knirel, Y. A. Carbohydr. Res. 2001, 336, 329-336.
9
J. AM. CHEM. SOC. VOL. 127, NO. 8, 2005 2415

A R T I C L E S
Bedini et al.
In agreement with proton NMR data, oligosaccharide 2 does
not present a real secondary motif if compared with the other
two species; this could be due to the small number of glycidic
residues as well as to the presence of the benzyl moiety.
Looking at oligosaccharide 4, a cleft is defined between resi-
dues C and E: residue C apparently starts the cavity that is
completed from E; this pattern is repeated in 6 as well, where
the cavities are two and are defined by the residues C-E and
F-H; in addition, these two conformational motifs display a
different orientation with respect to the chain elongation
direction.
Application of these results to the conformation of the whole
polysaccharide suggests the presence of a helical pattern, al-
though this assumption needs to be confirmed with different
experimental approaches.
Conclusions
The combined NMR and molecular modeling data show that
the long synthetic oligomers 4 and 6 adopt a secondary structure
defined by the presence of one and two coils, respectively.
The bioactivity of the synthetic oligorhamnans correlates with
their chain length. Compound 2 (with no coil) has low biological
impact, whereas 4 and 6 are quite effective both in PR-1 gene
expression, a plant immune response, and in HR suppression
in Arabidopsis thaliana caused by an avirulent strain of
Pseudomonas syringae.
On the basis of these findings we propose that the coiled
structure exposed from dimeric and trimeric oligomers 4 and
6, of the repeating unit [R-L-Rha-(1f3)-R-L-Rha-(1f3)-R-L-
Rha-(1f2]n, peculiar for many O-chains of phytopathogenic
bacteria, is a PAMP recognized by plants and leads to the
elicitation of specific responses including the suppression of
HR and PR-1 gene expression.
Figure 3. Two different views of the Connelly surface models of
oligorhamnans 2, 4, and 6 (reported from the left to the right, respectively).
On top of the picture, the clefts induced from the rhamnose residues are
evidenced.
Further information was obtained by molecular modeling.
According to the molecular mechanics approach, the optimal
dihedral angles for each glycosidic junction were evaluated using
the MM3* force field:12 the relaxed maps reveal that the
rhamnosidic linkage, independent of its connectivity, displays
two minima at different energies (Figure S3, Supporting In-
formation).
The dihedral angle values of the lowest minimum were used
for the construction of the starting structures for 2, 4, and 6
which were submitted to molecular dynamics analysis. The three
oligosaccharides were kept in a thermal bath at 303 K (4000 ps
for 2 and 4; 8000 ps for 6), and ensemble average interproton
distances for each molecule were extracted from the simulations
and translated into NOE contacts according to a full matrix
relaxation approach. The predicted NOEs showed good agree-
ment with those collected experimentally, proving the reliability
of the simulation data. Analysis of the dynamics simulation data
revealed that each glycosidic angle occupies the preferred values
most of the time (Figure S4, Supporting Information) and that
the lifetime of the second conformation is short. These results
showed that the three oligosaccharides spend most of the time
in one main conformational state depicted from the optimal
dihedral angles found during the molecular mechanics approach.
To gain additional information regarding the overall shape of
these molecules, the surface of the three oligosaccharides 2, 4,
and 6 was visualized (Figure 3) according to the Connelly
method.
Acknowledgment. The authors thank the “Centro di Metod-
ologie Chimico-Fisiche” of the University Federico II of Naples
for NMR and computer modeling facilities and MIUR (Progetti
di Ricerca di Interesse Nazionale, M. P., 2004) for financial
support. The work at Biomerit Research Centre is supported
by the Science Foundation of Ireland through an Investigator
award (03/IN3/B373) to J.M.D. The work at KVL is supported
by the Danish Agricultural and Veterinary Research Council.
We also acknowledge financial support to M.-A. N. from
Carlsbergfondet, Copenhagen, Denmark. We acknowledge
funding by the Fonds der Deutschen Chemischen Industrie.
Supporting Information Available: Full experimental pro-
cedures. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA0452166
(12) Allinger, N. L.; Yuh, Y. H.; Lii, J. H. J. Am. Chem. Soc. 1989, 111, 8551-
8566.
9
2416 J. AM. CHEM. SOC. VOL. 127, NO. 8, 2005