View
0
Download
0
Category
Preview:
Citation preview
Felix et al. - 1
Molecular sensing of bacteria in plants: The highly conserved RNA-binding motif RNP-1 of bacterial cold shock proteins is recognized as an elicitor signal in tobacco
Georg Felix1 and Thomas Boller Friedrich Miescher-Institute, P.O. B. 2543, CH-4002 Basel, Switzerland
Running title: Bacterial cold-shock proteins as elicitors in tobacco
Keywords : Cold shock protein, Cold shock domain, peptidoglycan, elicitor activity, PAMP,
Pattern recognition receptor (PRR)
1 To whom correspondence should be addressed.
E-mail: Felix@fmi.ch; fax: +41 61 697 45 27
Copyright 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
JBC Papers in Press. Published on December 5, 2002 as Manuscript M209880200 by guest on February 18, 2020
http://ww
w.jbc.org/
Dow
nloaded from
Felix et al. - 2
Summary
To detect microbial infection multicellular organisms have evolved sensing systems for
pathogen-associated molecular patterns (PAMPs). Here, we identify bacterial cold shock
protein (CSP) as a new such PAMP that acts as a highly active elicitor of defense responses in
tobacco. Tobacco cells perceive a conserved domain of CSP and synthetic peptides
representing 15 amino acids of this domain induced responses at subnanomolar concentrations.
Central to the elicitor-active domain is the RNP-1 motif KGFGFITP, a motif conserved also in
many RNA- and DNA-binding proteins of eukaryotes. Csp15-Nsyl, a peptide representing the
domain with highest homology to csp15 in a protein of Nicotiana sylvestris exhibited only
weak activity in tobacco cells. Crystallographic and genetic data from the literature show that
the RNP-1 domain of bacterial CSPs’ resides on a protruding loop and exposes a series of aro-
matic and basic side chains to the surface that are essential for the nucleotide-binding activity
of CSPs’. Similarly, these side chains were also essential for elicitor activity and replacement
of single residues in csp15 with Ala strongly reduced or abolished activity. Most strikingly,
csp15-Ala10, a peptide with the RNP-1 motif modified to KGAGFITP, lacked elicitor activity
but acted as a competitive antagonist for CSP-related elicitors. Bacteria commonly have a
small family of CSP-like proteins including both cold-inducible and non-inducible members,
and Csp-related elicitor activity was detected in extracts from all bacteria tested. Thus, the
CSP domain containing the RNP-1 motif provides a structure characteristic for bacteria in
general, and tobacco plants have evolved a highly sensitive chemoperception system to detect
this bacterial PAMP.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 3
Introduction
A key aspect of active defense against invading microbial pathogens is the ability to discrimi-
nate between self and infectious nonself (1). In plants, recognition-dependent disease resistance
has been studied most thoroughly and most successfully in cases that depend on the presence of
specific resistance-genes which confer immunity to particular races of plant pathogens. Several
of these resistance genes were shown to be involved in the chemoperception of factors specifi-
cally attributed with particular strains of pathogens (2-4). In addition, plants have a broader,
more basal, surveillance involving sensitive perception systems for patterns characteristic for
entire groups or classes of microorganisms, and they respond to these general elicitors with
activation of signaling pathways that initiate defense mechanisms (5). This is highly reminis-
cent of innate immunity in animals and humans. Among the elicitors that represent patterns
characteristic for fungi are cell wall components like glucans, chitin and chitosan oligosaccha-
rides, peptides and proteins with fungal-specific N-glycosylation and the membrane component
ergosterol (6;7). Similarly, cells of many plant species have a perception system for the com-
mon bacterial surface protein flagellin, the building block of the flagella (8). Perception of
flagellin by Arabidopsis thaliana was shown to depend on FLS2, a membrane-bound receptor
kinase protein with an extracellular leucine rich repeat (LRR) (9). Bacterial flagellin has
recently also been identified as one of the ‘pathogen associated molecular patterns’ (PAMPs’)
that activate the innate immune system of humans and animals (10) via the toll-like receptor 5
(TLR5) (11;12). Thus, perception of general elicitors in plants resembles perception of
PAMPs’ in the innate immune system of animals with respect to the type of molecules per-
ceived, the characteristics of pattern recognition receptors (PRRs) involved, as well as some of
the signaling mechanisms and defense responses induced (13).
Flagellin was the predominant if not only elicitor present in crude bacterial extracts that acti-
vated elicitor responses in the tomato cells used in our previous experiments. Extracts from
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 4
bacteria without flagella or with flagellins that are strongly divergent in the elicitor-active
domain represented by the oligopeptide flg22 proved inactive in the tomato cells (8). These
observations with one particular cell line, grown in vitro for several years, do not exclude the
existence of chemoperception systems for other bacterial PAMPs’ in tomato or other plant
species. Perception of several different PAMPs’, indicative for the same class of microbial
pathogens, appears characteristic for the innate immune system of animals. Similarly, redun-
dancy of chemoperception systems for a variety of molecular patterns characteristic for fungi
has also been observed in plants (6). Therefore, we set out to search for additional chemoper-
ception systems of plants sensing molecular patterns characteristic for bacteria. Suspension
cultured tobacco cells have long been known to respond with a rapid K+ efflux, a concomitant
medium alkalinization and an oxidative burst when treated with bacterial preparations con-
taining either living or heat-killed bacteria (14) but the bacterial factors eliciting these
responses have not been identified. In initial experiments we tested commercial preparations
containing peptidoglycan from Micrococcus lysodeikticus (Staphilococcus aureus) for
induction of responses in cultured tobacco cells. Peptidoglycan has long been known as a
PAMP signaling presence of gram positive bacteria in the innate immune systems of animals
(10). The peptidoglycan preparation indeed induced significant and rapid responses in tobacco
but, surprisingly, a preparation of lyophilized M. lysodeikticus bacteria proved far more potent
as source of elicitor-activity. We concentrated on the purification and characterization of this
latter activity and, in the present work, identified it as a small protein belonging to the family of
so-called cold shock proteins.
Experimental Procedures
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 5
Materials
Peptides were synthesised by F. Fischer (Friedrich Miescher-Institute, Basel) or by Bio-
Synthesis Inc. (Lewisville, Tx, USA). Peptides were dissolved in H2O (stock solutions of 1 to
10 mM) and diluted in a solution containing 0.1% BSA and 0.1 M NaCl. Agrobacterium
tumefaciens (strain C58 T) , Rhizobium meliloti and Xanthomonas campestris were obtained
from Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSM GmbH, Braun-
schweig, BRD) and grown in King’s B broth at 26°C on a rotary shaker. Bacteria were
harvested by centrifugation, washed once with H2O and resuspended in H2O (10% of original
volume). Crude bacterial elicitors were prepared by boiling the bacterial suspensions for 5 to
10 min and removing of bacterial debris by centrifugation. Lyophilized bacteria of M.
lysodeikticus (Sigma, St Louis, MO, USA) and the peptidoglycan fraction from M.
lysodeikticus (Fluka, Buchs, Switzerland) applied as suspensions in H2O. The bacterial
preparation ‘messenger’ was obtained from EDEN Bioscience (Bothell, WA, USA).
Purification of elicitor from Micrococcus lysodeikticus
Elicitor activity was purified from lyophilized preparation of M. lysodeikticus (Sigma). Ten g
of the lyophilizate were suspended in 100 ml H2O and heated for 10 min at 95° C. After
centrifugation (30 min 10’000 x g) the supernatant was mixed with 1 volume of acetone and
the precipitate formed after overnight incubation at –20° C was removed by centrifugation. The
acetone concentration was brought to 80 % (v/v) and the precipitate formed after 4 h at –20° C
was collected by centrifugation. This precipitate was dissolved in 20 mM Tris-HCl pH 7.5 and
passed over an anion-exchange column with diethylaminoethyl-cellulose (DE-cellulose,
Whatman, Maidstone, England) equilibrated with 20 mM Tris-HCl pH 7.5. Activity eluting in
the flow through was concentrated by acetone precipitation (80 % acetone) and separated on a
Sephasil C8 reversed phase column (Pharmacia, Uppsala, Sweden) at pH 6.5 (10 mM phos-
phate buffer pH 6.5 as solvent A and 80 % acetonitrile / 20 % phosphate buffer as solvent B).
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 6
The two fractions containing highest elicitor activity were pooled, pH adjusted to 3.5 and rerun
on Sephasil C8 reversed phase column at pH 3.5 ( 0.1 % TFA in H2O at pH 3.5 as solvent A
and 80 % acetonitrile / 20 % H2O with 0.1 % TFA as solvent B).
Plant cell cultures
The tobacco (Nicotiana tabacum L.) cell culture line 275N, originally derived from pith tissue
of a Havanna 425 plants, was maintained and subcultured as described before (15) in a
Murashige-Skoog based medium. Cells were maintained as suspension cultures and were used
4- to 10-days after subculture for experiments. Cell cultures of tomato (“line Msk8“, (16)),
potato (17), Lycopersicon peruvianum (18) and Arabidopsis thaliana (19) were cultured as
described elsewhere.
Alkalinization response
To measure alkalinization of the growth medium (the alkalinization response), 3 ml aliquots of
the cell suspensions were placed in open, 20-ml vials on a rotary shaker at 120 to 150 cycles
per min. Using small combined glass electrodes (Metrohm, Herisau, Switzerland) extracellular
pH values were either recorded continuously with a pen recorder or measured after 15 or 20
min of treatment.
Oxidative burst and ethylene biosynthesis in leaf tissue
Fully expanded leaves of different plant species were cut in 2-mm slices and floated on H2O
overnight. For measuring the oxidative burst, active oxygen species released by the leaf tissue
were measured by a luminol-dependent assay (20). Slices were transferred to assay tubes (2-4
slices corresponding to ~20 mg fresh weight) containing 0.1 ml of H2O supplied with 20 µM
luminol and 1 µg horseradish peroxidase (Fluka). Luminescence was measured in a LKB 1250
luminometer (LKB Wallac, Turku, Finland) for 20 min after the addition of the test solution.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 7
For assaying ethylene production, leaf slices (~50mg fresh per assay) were transferred to
6-ml glass tubes containing 1 ml of an aqueous solution of the peptide being tested. Vials were
closed with rubber septa and ethylene accumulating in the free air space was measured by gas
chromatography after 2 to 2.5 h of incubation.
Reproducibility
The results shown in the Figures represent single experiments which are representative
for several independent repetitions.
Results
Extracellular alkalinization in cultured tobacco cells treated with preparations from M.
lysodeikticus
Peptidoglycan, an essential cell wall-component of all bacteria, acts as one of the PAMPs
signaling presence of gram-positive bacteria to the innate immune system in animals (1;10). In
initial experiments we tested preparations containing peptidoglycan for induction of
extracellular alkalinization in plant cells cultured in liquid medium. Medium alkalinization,
occurring as a consequence of altered ion fluxes across the plasma membrane, can serve as a
convenient, rapid, sensitive and quantitative bioassay to study elicitor perception by plant cells
(16). As a source of peptidoglycan we used preparations from M. lysodeikticus (Staphylococcus
aureus) since lyophilized bacteria and a peptidoglycan fraction are commercially available.
Also, as deduced from the genomes of the three fully sequenced strains of Staphylococcus
aureus (M. lysodeikticus) that do not encode proteins resembling flagellin, these preparations
should be free of elicitor-active flagellin that could interfere in the assays. No alkalinization
was observed in the tomato cells of the line Msk8 after treatment with lyophilized M. lysodeik-
ticus bacteria or the peptidoglycan fraction derived from these bacteria (data not shown). While
these negative results confirmed the absence of elicitor-active flagellin they did not provide
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 8
evidence for a chemoperception system responding to peptidoglycan in the tomato cells. When
tested on tobacco cells, however, both preparations of M. lysodeikticus caused rapid and strong
medium alkalinization (Figure 1). As shown in the examples in Figure 1A, extracellular pH
started to increase after a lag of ~3 to 5 min and reached a maximum after ~10 to 15 min.
Depending on the cell density and the initial pH of different batches of the cell culture the
amplitude of the alkalinization response (∆pHmax) varied from 1.2 to 2 pH units for lyophilized
bacteria and from 0.6 to 1.4 pH units for the peptidoglycan fraction, respectively. In aliquots
from a given batch of cells, however, ∆pHmax was highly reproducible and consistently showed
a bigger response for the preparation of total bacteria than for the peptidoglycan fraction. The
responses of the cells to both preparations of M. lysodeikticus were dose-dependent and lower,
non-saturating doses led to prolonged lag phases, smaller maximal pH-increases and shortened
durations of medium alkalinization. The pH-change occurring within 15 min (∆pH15min) of
treatment was a steady function of the dose applied and was used as a parameter to compare the
relative strength of the two preparations of M. lysodeikticus (Figure 1B). Half-maximal stimu-
lation was observed with 30 µg/ml of the peptidoglycan (EC50) and <1 µg/ml with the lyophi-
lized bacteria, respectively. Treatment with protease K strongly affected the activity of the bac-
terial preparation resulting in a 200-fold higher EC50 value (200 µg/ml) but led only to a 3-fold
increase for the EC50 value of the peptidoglycan fraction (Figure 1B). These results provided
preliminary evidence for the presence of two distinct elicitor-activities in M. lysodeikticus: a
non-proteinaceous elicitor in the peptidoglycan fraction and a second, potent, proteinaceous
elicitor predominating in the total bacteria preparation. On a per weight basis the
proteinaceous factor was more than 100-fold more active than the peptidoglycan factor and
further work focussed on the characterization of this new protein elicitor.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 9
Purification of an elicitor-active protein from M. lysodeikticus and its identification as
bacterial cold shock protein
The elicitor activity, extracted from the crude preparation of M. lysodeikticus (Staphylo-
coccus aureus), was heat-stable (5 min, 95°C), passed ultra-filters with a molecular weight cut-
off of 10 kDa and was inactivated by treatment with trypsin (data not shown), indicating that
the elicitor activity was attributable to a peptide or small protein. Activity was purified on a
Sephasil C8 reversed phase column (Figure 2A). In the first chromatography at pH 6.5 activity
eluted as a single peak (Figure 2B). The two fractions containing most of the activity were
pooled and rerun on the C8 column at pH 3.5. The peak of activity eluting from this second run
correlated with a single peak of OD214. Separation by SDS-PAGE (14 % (w/v) acrylamide)
showed a band migrating with an apparent molecular weight of 7 to 9 kDa and elicitor activity,
detected in eluates of the sliced gel pieces, was found to co-migrate with this band (data not
shown). N-terminal sequencing of the protein and sequence information obtained from some of
the peptides after tryptic digestion identified the protein as a cold shock protein (CSP). In
Figure 3 the sequence information from the purified protein was aligned with the sequence of
the major cold shock protein from M. luteus and a consensus sequence obtained from >150
bacterial cold shock proteins present in the data bank.
Identification of the ‘cold shock domain’ (CSD) as the elicitor-active epitope
In attempts to localize the elicitor activity to a particular domain of the protein the puri-
fied CSP was subjected to peptide cleavage. Digestion with trypsin, Lys-C or Glu-C (V8 prote-
ase) abolished the activity and did not result in smaller fragments with elicitor activity (data not
shown). As in previous work with bacterial flagellin (8) we speculated that plant cells might
have a perception system for the most characteristic and most conserved domain of the CSPs’.
Although these small bacterial proteins show a high overall homology they are particularly
conserved in a domain close to the N-terminus. Based on the consensus sequence of bacterial
CSPs’, a 22 amino acid peptide spanning this domain was synthesized (Figure 3, underlined
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 10
sequence) and tested for induction of alkalinization in tobacco cells. This peptide, termed
csp22, proved even more active than the intact CSP purified from M. lysodeikticus and induced
medium alkalinization with an EC50 of ~0.1 nM (Figure 4).
To further delineate the epitope that activates responses in the plant cells, peptides lacking
varying numbers of amino acid residues from the N-terminal or C-terminal end were synthe-
sized and assayed for activity in dose response curves as described above for csp22 and CSP.
The amino acid sequences and the EC50 values are summarized in (Figure 5). Omitting 5 amino
acid residues from the N-terminus of csp22 reduced activity only slightly (EC50 of ~1.2 nM)
but removal of the Lys residue at position 6 showed a much stronger effect (EC50 of ~220 nM)
and further trimming by 4 amino acid residues resulted in an inactive peptide. The peptide
termed csp15, comprising the 15 amino acid residues central to csp22, was nearly as active as
csp22 (EC50 of 0.3 nM) and served as a core peptide for testing structural analogues with
replacements of single amino acid residues with alanine. Csp15-Ala3, csp15-Ala4, csp15-Ala8
and csp15-Ala12 all exhibited at least 1000-fold reduced activity compared to csp15. Csp15-
Ala10 was inactive even at the highest concentration of 100 µM tested (Figure 5). In contrast,
substitution of Phe at position 10 with a Tyr residue resulted in a peptide with full activity in
the tobacco cells (Figure 5). Among the peptides with single substitutions with Ala only csp15-
Ala7 showed no significant decrease in activity. Interestingly, the Glu at this position also
shows least conservation in the different sequences of bacterial CSPs’ (Figure 3).
The three dimensional structure has been determined for the major bacterial cold shock
proteins CspB from B. subtilis (MMDB Id: 3622 PDB Id: 1CSP , (21) and CspA (CS7.4) from
E. coli (MMDB Id: 1677 PDB Id: 1MJC (22). CspB forms a dimer while CspA occurs as
monomer. Besides this difference of dimerization the structures of both proteins are very
similar, forming compact β-barrel structures built up from five antiparallel β-strands with
connecting turns and loops. Figure 6 shows models (secondary structure and a 3-D ribbon
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 11
model) of the molecular structure of CspB, highlighting the domain spanned by the csp15
peptide. Clearly, elicitor activity can be attributed to the domain formed by the antiparallel
strands β1 and β2 and the loop L1. This domain includes a RNA-binding motif known as RNP-
1 (also termed RNP-CS) and exposes a cluster of aromatic and basic side chains to the surface
of the protein. An analysis using site-directed mutagenesis of CspB from B. subtilis has
demonstrated that these conserved residues are essential for the interaction of the protein with
nucleic acids (23). In Table 1, these single amino acid replacements and their effects on nucleic
acid binding were compared to the corresponding amino acid changes in csp15 and their effects
on elicitor activity in tobacco. All the substitutions in csp15 that correspond to substitutions
leading to strong or complete reduction in affinity of CspB for nucleic acids exhibited strongly
reduced elicitor activity in tobacco cells (higher EC50 values). The substitution of Phe by Tyr at
the position that corresponds to residue 10 in csp15 did not affect affinity of CspB for nucleic
acids and also did not alter elicitor activity.
Activity of peptides representing homologous domains occurring of proteins from plants and
animals
Csp-related proteins are common to all eubacteria and they usually form a small family of
proteins that include both cold-inducible and non-inducible members (24). The domain con-
taining the RNP-1 motif is conserved also in many eukaryotic proteins that bind to RNA or
DNA. Examples for proteins with this so-called cold shock domain (CSD) include human and
animal transcription factors recognizing the Y-box sequence and glycine rich RNA-binding
proteins occurring in plants (see supplementary data for gene structure and alignment with bac-
terial CSPs’). Peptides corresponding to the homologues of a human Y-box protein and a Gly-
rich protein from N. sylvestris were synthesized and tested for activity. The csp15 homologue
from the human Y-box protein was inactive whereas the peptide representing N. sylvestris
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 12
sequence induced responses with an EC50 of 300 nM and was thus ~1000-fold less active than
the csp15 representing the bacterial sequence (Figure 5).
Responses induced by CSP in different plant species
We examined cell cultures derived from other plant species for alkalinization in response
to CSP-related elicitors. Responses with characteristics similar to the ones of the tobacco line
275N were observed also with a second line of tobacco, originating from a plant of the variety
SR1, with a cell line derived from potato and a cell culture from Lycopersicon peruvianum
(data not shown). In contrast, no responses could be detected in the cell culture line msk8,
originally derived from a cross Lycopersicon esculentum with L. peruvianum, and in cell lines
from A. thaliana and rice (data not shown). Negative results with particular lines of cell cul-
tures do not allow concluding on the absence of a perception system in the corresponding plant
species since this perception system might be not expressed or might have been lost during the
years of growth in vitro.
Induced release of active oxygen species, an oxidative burst, and increased biosynthesis
of the stress hormone ethylene are responses characteristic for plants under attack by pathogens
or treated by elicitor preparations (6;25). We used these responses to monitor responsiveness
towards CSP-derived elicitors in leaf tissues from different plant species. As exemplified in
Figure 7 for leaf tissue from tomato, rapid, significant increase in ethylene biosynthesis and in
active oxygen species was observed after treatment with csp15 but not after treatment with the
same dose of csp15-Nsyl. Similarly, clear CSP-dependent induction of ethylene biosynthesis
and oxidative burst was observed in tobacco and several other solanaceous plants including
potato (Solanum tuberosum), Solanum dulcamara, Scopolia carniolica and Mandragora offici-
narum. In contrast, no response could be detected leaf tissue and cell cultures of A. thaliana,
cucumber and rice. Also, no signs of a hypersensitivity response (HR) could be detected after
injection of CSP-peptides to leaves of tobacco or tomato (data not shown).
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 13
In summary, a perception system for CSP-related elicitors is common to solanaceous
plants but has not yet been found outside of this plant family.
The inactive peptide csp15-Ala10 antagonizes elicitor activity of CSP
Peptides lacking either 4 amino acids from the C-terminal part (csp11) or 6 amino acids from
the N-terminal part spanned by csp15 lacked activity even when applied in micromolar concen-
trations (data not shown). No response was observed also by application of these two peptides
in combination (data not shown). Truncated forms of the biologically active peptides systemin
and flg22 were previously found that showed characteristics of competitive antagonists for the
respective non-truncated agonistic peptides (26-28). No antagonistic activity could be
observed for the two truncated CSP-peptides described above (data not shown). In contrast,
csp15-Ala10, also inactive as agonist (Figure 5), did exhibit antagonistic activity and
suppressed responses induced by csp15 (Figure 8). When added concomitantly with 3 nM
csp15, a concentration of 3 µM strongly inhibited induction of alkalinization response (Figure
8A, “0 min”). Complete inhibition was observed when csp15-Ala10 was added 30 s before the
agonist but progressively weaker effects were observed when the antagonist was added after
the agonist and an addition after 3.5 min remained without apparent effect on the ongoing
response. Inhibition by csp15-Ala10 was specific for CSP-derived elicitors and was not
observed with unrelated elicitors like flagellin and chitin fragments (data not shown). Inhibition
of CSP-related activity by csp15-Ala10 was competitive and could be overcome by increasing
concentrations of active peptide or intact CSP. As shown in the example in Figure 8B, this
resulted in an increase of the EC50 for the CSP containing preparation of M. lysodeikticus
bacteria from 1 µg/ml in the absence of the antagonist, to 20 µg/ml in the presence of 3 µM
csp15-Ala10, respectively. In contrast, no shift in dose-response was observed with the
peptidoglycan fraction (Figure 8B). These results confirm predominance of the CSP-related
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 14
elicitor in the crude bacterial preparation and the presence of an activity unrelated to CSP in the
peptidoglycan preparation.
Tobacco cells were found to respond to crude extracts from all bacterial species tested (n>20).
The antagonist csp15-Ala10 could serve as a diagnostic tool to test for the presence of csp-
related activity. For example, cells responded with strong alkalinization when treated with
‘messenger’, an extract from E. coli expressing transgenic harpin from Erwinia amylovora
(Figure 8C). Interestingly, at least at limiting doses of ‘messenger’ applied, activity was fully
antagonized by csp15-Ala10. This indicated that a CSP-related stimulus and not harpinEa,
previously reported to act as an inducer of alkalinization in tobacco (29), was the activity pre-
dominating in this preparation. As shown for the example of an extract from Agrobacterium
tumefaciens in Figure 8D, csp15-Ala10 antagonized also the alkalinization-inducing activity of
crude extracts from the plant-associated species Agrobacterium tumefaciens, Rhizobium
meliloti and Xanthomonas campestris, extracts that were previously found to be devoid of
elicitor-active flagellin (8). In summary, these results demonstrate the common occurrence of
CSP-related elicitor-activity in extracts from different, if not all, bacteria.
Discussion
Various types of living bacteria as well as preparations of heat-killed bacteria can trigger rapid
responses in plant cell cultures and defense responses in intact plant tissues (20;30;31).
Flagellin (8) and lipopolysaccharides (32) have been identified as common bacterial determi-
nants or PAMPs’ that act as elicitors of defense responses in plant cells. In this report we add
CSPs’ as further bacterial PAMP which acts as an elicitor of defense responses in plants.
Cold shock proteins were named based on the original observation that rapid cooling with
a ∆T of >-10 °C (cold shock) induces accumulation of specific proteins in many bacterial
species. The major CSPs’ are small, ~7.4 kD, proteins that belong to a family of highly con-
served proteins commonly occurring in all bacteria. At least some members of this family are
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 15
(also) constitutively expressed or are induced under stress conditions different from cold shock
(24). For example, the family of CspA-like proteins in E. coli consists of eight members (CspA
through CspH) and only CspA, CspB and CspG are cold-inducible. Thus, despite their name,
members of the CSP family occur also in bacteria not subjected to a cold shock treatment.
CSPs’ are implicated in various cellular processes, including cellular growth and adaptation to
low temperatures, nutrient stress and stationary phase. CSPs’ bind to nucleic-acids and appear
to function as RNA-chaperones and anti-terminators of translation (33).
A shift to low temperature induces also a set of specific proteins in plants (34). Some of
these cold-regulated proteins are small hydrophilic proteins of 6.6 kD (35), like the major bac-
terial CSP but they are non-homologous in sequence, and their physiological function in cold
acclimation process remains uknown. However, many eukaryotes including plants and animals
have proteins with a nucleic-acid-binding domain that shows a strikingly high homology and
similar RNA-binding properties to bacterial CSPs’ (36). It is this universally conserved
domain, also termed cold-shock domain (CSD), that contains the RNP-1 motif and the epitope
found to act as elicitor of tobacco cells.
The elicitor activity of bacterial CSPs’ could be localized to a stretch of ~15 amino acid
residues that forms a loop with two antiparallel β-strands and exposes a series of aromatic and
basic amino side chains to the surface of the protein. It is this epitope that exhibits highest con-
servation between the different bacterial CSPs’ and, as has been demonstrated by site directed
mutagenesis of CspB from B. subtilis, is essential for the interaction of the protein with nucleic
acids (23). Most notably, synthetic peptides with amino acid sequences reflecting the changes
leading to reduced or abolished binding to nucleic acids of CspB were also strongly affected in
elicitor activity (Table 1). This strong correlation raises the question whether some sort of
nucleic acid might be involved in the perception process by the plant cell. However, intact
CSPs’ and the csp-derived peptide elicitors have characteristics of molecules that are not
permeable for membranes and the first responses to subnanomolar concentrations of csp-de-
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 16
rived elicitors occur after a lag-phase of less than 2 min. These characteristics rather suggest a
chemoperception system with a specific, high-affinity primary interaction site in the apoplast,
most likely the plasmamembrane, of the plant cells. The strong correlation of nucleic acid
binding and elicitor activity might thus reflect evolution of a chemoperception system directed
at a particular surface epitope of the bacterial CSPs’ that is under a high selective pressure for
retaining functionality of the protein.
Perception of csp-related elicitors resembles the chemoperception system for flagellin-de-
rived elicitors studied before (8;28). In both cases, elicitor activity could be attributed to an
epitope spanning an epitope of ~15 amino acids representing the most conserved part of the
respective protein. Both elicitors are active at subnanomolar doses and activity is highly de-
pendent on the genuine amino acid sequence of the conserved domain. ‘Mutational’ analysis
using structural analogs of the elicitors allowed identification of peptides lacking elicitor
activity but exhibiting properties of competitive antagonists. Perception of flagellin was could
be shown to involve a specific, high-affinity binding site and the membrane bound receptor
kinase FLS2 (9;37). A model involving a two-step process for receptor activation was proposed
to explain the effects of agonistic and antagonistic peptides (28). At present, experiments that
directly demonstrate a receptor site for the csp-elicitors are lacking. Nevertheless, CSP- and
flagellin-derived elicitors induce the same set of responses with similar kinetics, indicating a
similar, receptor-mediated process for both elicitors. Thus, we hypothesize that csp-perception
occurs via a csp-receptor that functions in a manner similar to the receptor for flagellin.
Activation of this putative CSP-receptor might also involve two consecutive steps with binding
of the elicitor as a first step and activation of the receptor as a second step. An aromatic side
chain on residue 10 of csp15, Phe in csp15 or Tyr in csp15-Tyr10, appears necessary for this
second step to activate the receptor. The antagonist csp15-Ala10, apparently, does not undergo
the second, locking step and interacts with the receptor site in a more readily reversible
manner. This could explain the high excess of antagonist csp15-Ala10 over csp15 required to
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 17
block elicitor action completely and the apparent inefficiency of the antagonist when applied
subsequent to the csp agonists (Figure 8).
Proteins with a cold shock domain comprising the RNP-1 motif are conserved also in
eukaryotes and have been identified also in genes of A. thaliana and N. sylvestris. Although
clearly homologous, the sequences corresponding to the elicitor-active epitope show some
differences in comparison to the bacterial consensus. The synthetic peptide csp15-Nsyl, repre-
senting the least divergent form of this domain in genes known from N. sylvestris, indeed did
show some activity in the bioassay with tobacco cell. However, the specific activity of this pep-
tide was ~1000-fold lower than that of csp15 representing the bacterial epitope. A lower spe-
cific activity could be counterbalanced by the presence of high local concentrations of the
stimulus. In initial attempts with extracts of tobacco plants or cells from tissue culture we failed
to detect factors with CSP-like activity in bioassays (data not shown). Thus, at present, we do
not have evidence for endogenous factors stimulating tobacco via the CSP perception system
described in this report. Endogenous factors of tobacco, capable of stimulating medium alka-
linization in cultured cells, have recently been described (38) but these peptidic factors show no
apparent homology to CSPs’.
Bacterial CSPs’ can be regarded as molecules that are highly characteristic for bacteria in
general and could thus serve as PAMPs’ signaling the presence of bacteria to the plant cells.
An obvious problem with the hypothesis that bacterial CSPs’ serve as PAMP signaling the
presence of bacteria to the plant cells is the localization of these proteins, which are generally
assumed to function in the cytoplasm of the bacteria. So far, CSPs’ have not been reported to
be exported or exposed to the surface by intact bacteria and, consequently, CSPs’ are probably
not directly detectable by a chemoperception system assumed to reside on the surface of the
plant cells. Further studies will be required to test whether CSPs’ are released from bacteria
during invasion of their plant hosts. A release could be based on a bacterial export system
activated in the course of the infection process, or it could result from bacterial or plant
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 18
processes causing a general leakiness of the bacteria. As demonstrated for bacteria under mild
osmotic shock (39), leakiness of bacteria leading to release of small cytoplasmic proteins might
be more common than suggested by studies under optimal media conditions used to grow
bacterial cells in the laboratory. Precedence’s for cytoplasmic components of bacteria that act
as PAMPs’ and stimulate the innate immune responses via toll-like receptors in animals
include ‘non-secreted’ components such as the heat shock protein HSP60 (40) and bacterial
DNA (12). Bacterial DNA is recognized via its content of non-methylated CpG
oligonucleotides and this PAMP was successfully applied as a potent immuno-stimulatory
factor (41;42) but the process leading to release of the DNA from the bacteria has not been
elucidated. Similarly, no process that secretes HSP60 from intact bacterial cells has been
described. HSP60 is well conserved from microbes to humans and HSP60 from both
mammalian and microbial sources can trigger inflammatory responses via TLR4 (40),
suggesting that TLR4 may detect both endogenous and exogenous ligands as alarm signals.
Exposure of endogenous HSP60 to the TLR4 receptor could be envisaged to occur via release
from wounded or injured cells and perception by the receptor on different, intact cells. As
discussed above, it remains to be seen whether the chemoperception system for CSPs’
described in this report might similarly react to both endogenous and exogenous ligands.
Peptidoglycan consists of a glycan backbone with alternating β-1-4 linked residues of N-
acetyl-D-glucosamine and muramic acid and forms the major component of the cell wall in
gram-positive bacteria. Peptidoglycans, sensed by a family of peptidoglycan recognition
proteins (PGRPs’) that are conserved from insects to humans (43), are important PAMPs’ for
the innate immunity of animals. The peptidoglycan preparation of M. lysodeikticus also
triggered elicitor responses in tobacco cells. We have not yet characterized this activity in
detail and cannot exclude that it is due to a minor component or a ‘contaminant’ of the
peptidoglycan fraction. Nevertheless, this non-proteinaceous factor is perceived as a quality of
stimulus distinct from CSPs’ and provides evidence for a further bacterial PAMP with elicitor-
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 19
activity in tobacco cells. Thus, similar to the chemoperception systems for a variety of fungal-
derived PAMPs’, perception of bacteria by plant cells appears not to depend on a single
bacterial factor but rather involves several different factors, including at least flagellin (8),
lipopolysaccharides (32), peptidoglycan and CSP. This redundancy, characteristic also for the
recognition mechanisms in the innate immune system of animals, points at possible difficulties
with approaches to demonstrate a direct physiological role for any particular of these
chemoperception systems for plant defense. Inhibition or knock-out of only one of the systems
might be without strong effect on the overall recognition system. In contrast to bacterial
‘avirulence factors’, which act as elicitors that are specific and unique for a particular
pathogen, the structures recognized as PAMPs’ are essential or ‘vital factors’ for the
functioning of the bacterial organisms in general and cannot easily be changed, removed or
mutated for probing their role in plant defense.
In summary, our results provide evidence fora novel bacterial elicitors, cold shock
protein, for which tobacco and other Solanaceae have evolved specific and sensitive chemo-
perception systems. The accuracy and sensitivity of the perception system for the CSP domain
comprising the RNP-1 motif detailed in this report indicate a receptor mechanism involving a
high-affinity binding site on the surface of the plant and should provide the basis for further
work to identify the protein acting as pattern recognition receptor for CSP.
Acknowledgements
We thank Franz Fischer (Friedrich Miescher-Institute, Basel) for the synthesis of various pep-tides, Renate Matthies, Daniel Hess and Jan Hofsteenge (Friedrich Miescher-Institute) for protein sequencing services and mass spectrometry, and Martin Regenass for maintaining the cell cultures and technical assistance.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 20
Reference List
1. Medzhitov, R. and Janeway, C. A. Jr. (2002) Science 296, 298-300
2. Yang, Y., Shah, J., and Klessig, D. F. (1997) Genes & Development 11, 1621-1639
3. Dangl, J. L. and Jones, J. D. (2001) Nature 411, 826-833
4. Schneider, D. S. (2002) Cell 109, 537-540
5. Heath, M. C. (2000) Curr.Opin.Plant Biol. 3, 315-319
6. Boller, T. (1995) Annu.Rev.Plant Physiol.Plant Mol.Biol. 46, 189-214
7. Hahn, M. G. (1996) Annu.Rev.Phytopathol. 34, 387-412
8. Felix, G., Duran, J. D., Volko, S., and Boller, T. (1999) Plant J. 18, 265-276
9. Gomez-Gomez, L. and Boller, T. (2000) Mol.Cell 5, 1003-1011
10. Aderem, A. and Ulevitch, R. J. (2000) Nature 406, 782-787
11. Ogushi, K., Wada, A., Niidome, T., Mori, N., Oishi, K., Nagatake, T., Takahashi, A., Asakura, H., Makino, S., Hojo, H., Nakahara, Y., Ohsaki, M., Hatakeyama, T., Aoyagi, H., Kurazono, H., Moss, J., and Hirayama, T. (2001) J.Biol.Chem. 276, 30521-30526
12. Hayashi, F., Smith, K. D., Ozinsky, A., Hawn, T. R., Yi, E. C., Goodlett, D. R., Eng, J. K., Akira, S., Underhill, D. M., and Aderem, A. (2001) Nature 410, 1099-1103
13. Nurnberger, T. and Scheel, D. (2001) Trends Plant Sci. 6, 372-379
14. Atkinson, M. M., Huang, J.-S., and Knopp, J. A. (1985) Plant Physiol. 79, 843
15. Felix, G. and Meins, F., Jr. (1987) Planta 172, 386-392
16. Felix, G., Regenass, M., and Boller, T. (1993) Plant J. 4, 307-316
17. Schweizer, P., Felix, G., Buchala, A., Muller, C., and Métraux, J. P. (1996) Plant J. 10, 331-341
18. Felix, G. and Boller, T. (1995) Plant J. 7, 381-389
19. May, M. J. and Leaver, C. J. (1993) Plant Physiol. 103, 621-627
20. Keppler, L. D., Baker, C. J., and Atkinson, M. M. (1989) Phytopathology 79, 974-978
21. Schindelin, H., Marahiel, M. A., and Heinemann, U. (1993) Nature 364, 164-168
22. Schindelin, H., Jiang, W., Inouye, M., and Heinemann, U. (1994) Proc.Natl.Acad.Sci.U.S.A 91, 5119-5123
23. Schröder, K., Graumann, P., Schnuchel, A., Holak, T. A., and Marahiel, M. A. (1995) Mol.Microbiol. 16, 699-708
24. Thieringer, H. A., Jones, P. G., and Inouye, M. (1998) Bioessays 20, 49-57
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 21
25. Lamb, C. and Dixon, R. A. (1997) Annu.Rev.Plant Physiol.Plant Mol.Biol. 48, 251-275
26. Pearce, G., Johnson, S., and Ryan, C. A. (1993) J.Biol.Chem. 268, 212-216
27. Meindl, T., Boller, T., and Felix, G. (1998) PC 10, 1561-1570
28. Meindl, T., Boller, T., and Felix, G. (2000) PC 12, 1783-1794
29. Wei, Z. M., Laby, R. J., Zumoff, C. H., Bauer, D. W., He, S. Y., Collmer, A., and Beer, S. V. (1992) Science 257, 85-88
30. Atkinson, M., Bina, J., and Sequeira, L. (1993) Mol.Plant-Microbe Interact. 6, 253-260
31. Jakobek, J. L. and Lindgren, P. B. (1993) PC 5, 49-56
32. Meyer, A., Puhler, A., and Niehaus, K. (2001) Planta 213, 214-222
33. Bae, W., Xia, B., Inouye, M., and Severinov, K. (2000) Proc.Natl.Acad.Sci.U.S.A 97, 7784-7789
34. Guy, C. L. (1990) Annu.Rev.Plant Physiol.Plant Mol.Biol. 41, 187-223
35. Gilmour, S. J., Lin, C., and Thomashow, M. F. (1996) Plant Physiol 111, 293-299
36. Graumann, P. L. and Marahiel, M. A. (1998) Trends Biochem.Sci. 23, 286-290
37. Bauer, Z., Gomez-Gomez, L., Boller, T., and Felix, G. (2001) J.Biol.Chem.
38. Pearce, G., Moura, D. S., Stratmann, J., and Ryan, C. A., Jr. (2001) Proc.Natl.Acad.Sci.U.S.A 98, 12843-12847
39. Berrier, C., Garrigues, A., Richarme, G., and Ghazi, A. (2000) J.Bacteriol. 182, 248-251
40. Seo, S., Okamoto, M., Seto, H., Ishizuka, K., Sano, H., and Ohashi, Y. (1995) Science 270, 1988-1992
41. Krieg, A. M. (2002) Trends Immunol. 23, 64-65
42. Krieg, A. M. (2002) Annu.Rev.Immunol. 20, 709-760
43. Kang, D., Liu, G., Lundstrom, A., Gelius, E., and Steiner, H. (1998) Proc.Natl.Acad.Sci.U.S.A 95, 10078-10082
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 22
Table 1. comparison of mutations in cspB affecting DNA-bindinga and single amino acid changes in csp15b on elicitor activity
substitution in CspB substitution in
csp15b fold-reduction of affinity for oligonucleotidea
fold-increase for EC50 b
CspB (wt) csp15 1 1 K7Q csp15-Ala2 3 1000 W8A csp15-Ala3 5 1000 F15A csp15-Ala10 inactive >10000 F17A csp15-Ala12 inactive 2000 F15Y csp15-Tyr10 1 1 F17Y 6 not tested a effect of site-directed point mutations in CspB on the binding of single stranded DNA
oligonucleotides containing Y-box motif. Data from Table 1 in Schröder et al. (23). b csp15 comprises amino acids 5 to 20 in CspB. EC50-values for the induction of the
alkalinization response from Fig. 5.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 23
Figure 1. Extracellular alkalinization of tobacco cells in response to treatment with preparations from M. lysodeikticus. (A) Alkalinization in response to treatment with 10 µg/ml lyophilized M. lysodeikticus
cells or to 100 µg/ml of the peptidoglycan fraction derived from M. lysodeikticus
(peptidoglycan).
(B) Effect of protease K treatment on alkalinization-inducing activity of lyophilized M.
lysodeikticus bacteria and the peptidoglycan preparation. Different doses of lyophilized M.
lysodeikticus bacteria (closed circles), bacteria after pretreatment with protease K
(overnight incubation with 1 mg/ml protease K, open circles), peptidoglycan (closed
triangles) and peptidoglycan after pretreatment with protease K (open triangles) were
added to aliquots of the cell culture and pH-change measured after 15 min (initial pH 4.8).
M. lysodeikticus, µg/ml10-2 10-1 100 101 102 103
∆pH
at 1
5 m
in
0.0
0.4
0.8
1.2
1.6
2.0 bacteria
" protease digest
" protease digest
peptidoglycan
control
M. lysodeikticus
Time, min0 10 20 30 40 50 60
6.4
6.2
6.0
5.8
5.6
5.4
5.2
5.0
4.8
peptidoglycan
Extra
cellu
lar p
H
A
B
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 24
Figure 2. Purification of the alkalinization-inducing activity by reversed phase
chromatography
An extract from lyophilized M. lysodeikticus bacteria, pre-purified by ion-exchange
chromatography as described in methods, was fractionated on a C8 reversed phase column
at pH 6.5. Fractions with highest activity, eluting between 24 and 28 min, were re-
chromatographed on the C8 column at pH 3.5. Upper panel shows elution profile (OD280)
of first run at pH 6.5 (10 mM phosphate buffer) and of second run at pH 3.5 (0.1 % TFA).
Lower panel shows extracellular alkalinization in tobacco cells (∆pH15min) induced by
aliquots of the fractions eluting in the first run (open bars) and second run (open bars)
Sephasil C8 reversed phase
Time, min0 10 20 30 40 50 60
OD
280n
m
0.00
0.05
0.10
0.15
0.20
60% B
100% B1st run, pH 6.5
2nd run, pH 3.5
∆pH
max
0.0
0.5
1.0 1st run, pH 6.5 2nd run, pH 3.5
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 25
Figure 3. Alignment of bacterial cold shock proteins.
Sequence alignment of some CSPs’, representative for different bacteria species. Letters
indicate positions that differ from the consensus sequence. Consensus sequence and
percentage of conservation for the amino acid residues were calculated from >150
bacterial CSP-sequences present in the Swissprot database. Partial sequence for M.
lysodeikticus represents information obtained from the purified after tryptic digest and
Edman degradation of some of the peptides. Csp22 and csp15 denote peptides
synthesized according to the consensus sequence.
mekgt*****na******c****ge*i*a*y*t**m**yrt*ka**s*q*dvh**pk*nh*svivpveveaava meigi*****na******sa**v*a*i*a*y*v*em**yr**ka****q**vlhsdk*sh*tkiipi*dtqe matgt*************aqd**gp*****y***nat**r****n*v*n*dvth*e.****e**spa masgt*****s*******aqd**gp***a*y*n*naq*yre*q***a*t*d*t**qk****e*i*pa skikgn*****es******t**d*s***********tn***t*a***r*e***tn*ak**s****ia* akikgq*****es******t*ad*s************n***t*a***n*e***qd*qk**a*v***ai sgkmtgi******d******t*dd*s***********n**y***d*******t*es*ak**a*g***s* sgkmtgi******d******t*dd*s***********n**y***d*******t*es*ak**a*****s* msnkmtgl******d******s*vd*s***********n*nyrt*f*****t*s*es*ak**a****iitd maqgt*************t*dds*g*****y*e**tg***t*d*nar*q***g**ak****tg**lv meqgt**************r*n**.**********s******d*******dve**a**a*****q*a mnmeqgt**************r*n**.*****************d***a*t*dvee*q********q*a mqgr**************r*d**.**********q**y*******q*e*d*vd*a********v** mqrgk*****n***y****v***s.******t****e***t*****e*****v***********v** mqrgk*****n***y****v***s.******t****e*********e*****v***********v** mqngk*****n********v****.******t**e***y*******e*****ve*******s**v** mlegk*****s********v**q*.***********e***t*****a*****ve************e* i****s********v**q*.***********e***c*****a*****ve********* *****s*****l**v**q*.***********e***t***s*a*****ve********* mqtgk*****g********v***e.***************t*****e*****vd************n mngk*****n********m**se.**********s**y*a*****e***d*te**********a** mavgt*********y***a**dnsa***********n***e*q*ndrve**t*dgpk*l******** avgtv wfnaek elqendrvefetqdgpk
E. coli cspDH. influenzae cspD
Str. clavuligerus csp7Str. coelicolor cspF
E. coli cspEE. coli cspCE. coli cspA
S. typhimurium cspAE. coli cspB
Arthr. globiformisB. subtilis cspC
L. monocytogenes csp1B. cereus cspC
B. caldolyticus cspBB. stearothermophilus cspB
B. subtilis cspDB. subtilis cspB
B. globisporus cspBB. globigii cspBB. cereus cspDB. cereus cspBM. luteus cspA
M. lysodeikticus (purified protein)
CSP consensus ---TGTVKWFNAEKGFGFITPDGGDKDVFVHFSAIQGDGFKSLEEGQKVSFEI-QGNRGPQAANVTKLA ______________________
csp22 AVGTVKWFNAEKGFGFITPDGG identity : x 60-90%, x >90% csp15 VKWFNAEKGFGFITP
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 26
Figure 4. Elicitor activity of the purified Csp and synthetic peptides spanning the
conserved N-terminal domain of bacterial CSPs’.
Dose response curves for alkalinization induced by intact Csp (Csp7.4kDa) and synthetic
peptides representing 22 (csp22) or 15 (csp15) amino acid residues of the conserved region
from bacterial CSPs’ as indicated in Fig. 3 .
Concentration, nM10-2 10-1 100 101 102 103
∆pH
at 1
5 m
in
0.0
0.2
0.4
0.6
0.8
1.0
1.2
csp15
Csp7.4kDa
csp22
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 27
Figure 5. Alkalinization-inducing activity of csp-related peptides
EC50 values were determined from dose response curves obtained for the different peptides.
Specific activity relative to the activity of the most active peptide csp22 (hatched bar, EC50 of
0.1 nM). Logarithmic scaling was used to indicate residual activity in some of the peptides. No
activity could be detected with peptides denoted with asterisks (relative activity <10-5).
N-term
relative activity, log scale10
-3 -2 -1 010 10 1010
-410
-5
VKWFNAEKGFG
VKWFNAEKGFGF
VKWFNAEKGFGFI
VKWFNAEKGFGFITP
csp22 AVGTVKWFNAEKGFGFITPDDG
KWFNAEKGFGFITPDDG
WFNAEKGFGFITPDDG
EKGFGFITPDDG
csp15 VKWFNAEKGFGFITP
VKAFNAEKGFGFITP
VKWANAEKGFGFITP
VKWFNAAKGFGFITP
VKWFNAEAGFGFITP
VKWFNAEKGAGFITP
VKWFNAEKGFGAITP
csp15Ybox VKWFNVRNGYGFITP
csp15Nsyl VKWFSDQKGFGFITP
VKWFNAEKGYGFITP
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 28
N
C
2
46
54 57
65
6263
β1
β3
β4β5
V
KWFNADK
GFGFITP
β2
csp15
VFV
HF
RNP-2 : VFVHFRNP-1 : KGFGFITP
A B
Figure 6. Structure of bacterial CSPs’
(A) Schematic scheme for secondary structure of CspB from B. subtilis from Schindelin et al.
(21) with domain spanned by csp15 (hatched part) and the RNA-binding motifs RNP-1 and
RNP-2 indicated.
(B) Structure of CspB monomer (MMDB Id: 3622 PDB Id: 1CSP , (21)) drawn with WebLab
ViewerLite (Molecular Simulations Inc., Cambridge, UK) with side groups exposed in the
domain spanned by csp15.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 29
5 min
10 uV (RLU)
control
csp15-Nsyl
csp15
A
B
Figure 7. Induction of ethylene biosynthesis and oxidative burst in tomato leaf tissue.
(A) Ethylene biosynthesis in tomato leaf slices treated for 2 h with 1 µM csp15 or 1 µM csp15-
Nsyl as indicated. Bars and error bars show mean and standard deviation of n=4 replicates.
(B) Luminescence of leaf slices in a solution with luminol and peroxidase after treatment with
1 µM concentrations of csp15 or csp15-Nsyl as indicated. Light emission at the very beginning
of the experiments is caused by phosphorescence of the green tissue.
Ethylene, nmol / (g*h)0 1 2 3 4
csp15
csp15-Nsylv
control
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 30
M. lysodeikticus, µg/ml10-2 10-1 100 101 102 103
∆ pH
at 1
5 m
in
0.0
0.4
0.8
1.2
1.6 +csp15-Alal10bacteria
peptidoglycan +csp15-Ala10
∆pH 0.2
2 min
at -0.5 min
at 0 min
at 1 min
at 3.5 min+ csp15-Ala10:
csp1
5
controlA
B
D
∆pH 0.2
10 min
∆pH 0.2
10 min
C
csp15-Ala10+ 'messenger'
'messenger'
csp15-Ala10 + A. tum.
A. tumefaciens
csp15-Ala10
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 31
Figure 8. The peptide csp15-Ala10 acts as competitive suppressor of csp-related elicitors.
(A) Extracellular alkalinization in suspension-cultured tobacco cells treated with 3 nM csp15
and 3 µM csp15-Ala10 at the time points indicated (slanted arrows). Extracellular pH of
untreated cells was 4.9 and addition of 3 µM csp15-Ala10 alone did not cause significant pH-
changes.
(B) Alkalinization induced by different doses of M. lysodeikticus (circles) and peptidoglucan
preparation (triangles) in cells without pretreatment (closed symbols) or cells pretreated for 3
min with 10 µM csp15-Ala10 (open symbols).
(C) and (D) Alkalinization induced by the harpin-containing preparation ‘messenger’ (1 µg/ml,
C ) and by a crude extract from A. tumefaciens (1 µl/ml, D ) in cells without pretreatment or
cells pretreated for 3 min with 3 µM csp15-Ala10 as indicated.
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Felix et al. - 32
Supplementary data:
Comparison of proteins with cold shock domains from animals and plants with
bacterial CSP
Schematic representation of Y-box binding proteins with CSD (top) and glycine rich proteins
containing CSDs’ from plant (bottom). Alignment of CSD sequences with consensus
sequence of bacterial CSPs’ (middle, with * denoting residues conserved in Y-box proteins or
grp2 proteins, respectively. Shaded sequences indicate RNP-1 (left) and RNP-2 (right).
EKKIIASQVS GTVKWFNVKS GYGFINRDDT KEDVFVHQTA IVKNNPRKYL RSVGDGEKVE FDVVEGEKG NEAANVTGPE GSNVQGEKKVLATKVL GTVKWFNVRN GYGFINRNDT KEDVFVHQTA IKKNNPRKYL RSVGDGETVE FDVVEGEKG AEAANVTGPD GVPVEGDKKVIATKVL GTVKWFNVRN GYGFINRNDT KEDVFVHQTA IKKNNPRKYL RSVGDGETVE FDVVEGEKG EEAANVTGPG GVPVQGDKKVIATKVL GTVKWFNVRN GYGFINRNDT KEDVFVHQTA IKKNNPRKYL RSVGDGETVE FDVVEGEKG AEAANVTGPE GVPVQG
******* * *** * ***** * * * * * * * * *****
Aplysia californicaMus musculusHomo sapiens
Xenopus laevis
bacterial CSP (consensus)
Y box transcription factors
N-term C-termcold shock domain (CSD) ++++ ---- ++++ ---- ++++ ---- ++++
MSNMXT GTVKWFNAEK GFGFIEPEDG SKDVFVHFSA IQGDG....F KSLEEGQKVS FEIEQGARGP QAANVTKL ****** * ***** * ** * *** * * * * ** * ** * * *
MAEESGQRAK GTVKWFSDQK GFGFITPDDG GEDLFVHQSG IRSEG....F RSLAEGETVE FEVESGGDGR TKAVDVTGPD GAAVQGVNMSGGDRRK GTVKWFDTQK GFGFITPSDG GDDLFVHQSS IRSEG....F RSLAAEESVE FDVEVDNSGR PKAIEVSGPD GAPVQGGDNGGGERRK GSVKWFDTQK GFGFITPDDG GDDLFVHQSS IRSEG....F RSLAAEEAVE FEVEIDNNNR PKAIDVSGPD GAPVQG
N. sylvestris (grp2)A. thaliana (grp2b)A. thaliana (grp2)
MSGGGDMS
glycine rich proteins (grp2) in plants
alternating regions with positive and negative charges
N-term C-termcold shock domain (CSD) glycine rich domain
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Georg Felix and Thomas BollerRNP-1 of bacterial cold shock proteins is recognized as an elicitor signal in tobaccoMolecular sensing of bacteria in plants: The highly conserved RNA-binding motif
published online December 5, 2002J. Biol. Chem.
10.1074/jbc.M209880200Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
by guest on February 18, 2020http://w
ww
.jbc.org/D
ownloaded from
Recommended