Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Molecular Sensing of Bacteria in Plants

Molecular Sensing of Bacteria in Plants THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 8, Issue of February 21, pp. 6201–6208, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. THE HIGHLY CONSERVED RNA-BINDING MOTIF RNP-1 OF BACTERIAL COLD SHOCK PROTEINS IS RECOGNIZED AS AN ELICITOR SIGNAL IN TOBACCO*□ S Received for publication, September 26, 2002, and in revised form, November 21, 2002 Published, JBC Papers in Press, December 5, 2002, DOI 10.1074/jbc.M209880200 Georg Felix‡ and Thomas Boller From the Friedrich Miescher-Institute, P. O. Box 2543, CH-4002 Basel, Switzerland perception systems for patterns characteristic for entire groups To detect microbial infection multicellular organisms have evolved sensing systems for pathogen-associated or classes of microorganisms, and they respond to these general molecular patterns (PAMPs). Here, we identify bacterial elicitors with activation of signaling pathways that initiate cold shock protein (CSP) as a new such PAMP that acts defense mechanisms (5). This is highly reminiscent of innate as a highly active elicitor of defense responses in to- immunity in animals and humans. Among the elicitors that bacco. Tobacco cells perceive a conserved domain of represent patterns characteristic for fungi are cell wall compo- CSP and synthetic peptides representing 15 amino acids nents like glucans, chitin and chitosan oligosaccharides, pep- of this domain-induced responses at subnanomolar con- tides and proteins with fungal-specific N-glycosylation, and the centrations. Central to the elicitor-active domain is the membrane component ergosterol (6, 7). Similarly, cells of many RNP-1 motif KGFGFITP, a motif conserved also in many plant species have a perception system for the common bacte- RNA- and DNA-binding proteins of eukaryotes. Csp15- rial surface protein flagellin, the building block of the flagella Nsyl, a peptide representing the domain with highest (8). Perception of flagellin by Arabidopsis thaliana was shown homology to csp15 in a protein of Nicotiana sylvestris to depend on FLS2, a membrane-bound receptor kinase protein exhibited only weak activity in tobacco cells. Crystallo- with an extracellular leucine-rich repeat (9). Bacterial flagellin graphic and genetic data from the literature show that has recently also been identified as one of the “pathogen-asso- the RNP-1 domain of bacterial CSPs resides on a pro- ciated molecular patterns” (PAMPs) that activate the innate truding loop and exposes a series of aromatic and basic immune system of humans and animals (10) via the Toll-like side chains to the surface that are essential for the nu- receptor 5 (11, 12). Thus, perception of general elicitors in cleotide-binding activity of CSPs. Similarly, these side plants resembles perception of PAMPs in the innate immune chains were also essential for elicitor activity and re- system of animals with respect to the type of molecules per- placement of single residues in csp15 with Ala strongly ceived, the characteristics of pattern recognition receptors in- reduced or abolished activity. Most strikingly, csp15- Ala10, a peptide with the RNP-1 motif modified to volved, as well as some of the signaling mechanisms and de- KGAGFITP, lacked elicitor activity but acted as a com- fense responses induced (13). petitive antagonist for CSP-related elicitors. Bacteria Flagellin was the predominant if not only elicitor present in commonly have a small family of CSP-like proteins in- crude bacterial extracts that activated elicitor responses in the cluding both cold-inducible and noninducible members, tomato cells used in our previous experiments. Extracts from and Csp-related elicitor activity was detected in ex- bacteria without flagella or with flagellins that are strongly tracts from all bacteria tested. Thus, the CSP domain divergent in the elicitor-active domain represented by the oli- containing the RNP-1 motif provides a structure char- gopeptide flg22 proved inactive in the tomato cells (8). These acteristic for bacteria in general, and tobacco plants observations with one particular cell line, grown in vitro for have evolved a highly sensitive chemoperception system several years, do not exclude the existence of chemoperception to detect this bacterial PAMP. 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 A key aspect of active defense against invading microbial for the innate immune system of animals. Similarly, redun- pathogens is the ability to discriminate between self and infec- dancy of chemoperception systems for a variety of molecular tious nonself (1). In plants, recognition-dependent disease re- patterns characteristic for fungi has also been observed in sistance has been studied most thoroughly and most success- plants (6). Therefore, we set out to search for additional che- fully in cases that depend on the presence of specific resistance moperception systems of plants sensing molecular patterns genes that confer immunity to particular races of plant patho- characteristic for bacteria. Suspension cultured tobacco cells gens. Several of these resistance genes were shown to be in- have long been known to respond with a rapid K efflux, a volved in the chemoperception of factors specifically attributed concomitant medium alkalinization and an oxidative burst with particular strains of pathogens (2– 4). In addition, plants when treated with bacterial preparations containing either have a broader, more basal, surveillance involving sensitive living or heat-killed bacteria (14) but the bacterial factors elic- iting these responses have not been identified. In initial exper- iments we tested commercial preparations containing pepti- * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked doglycan from Micrococcus lysodeikticus (Staphylococcus “advertisement” in accordance with 18 U.S.C. Section 1734 solely to aureus) for induction of responses in cultured tobacco cells. indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains a supplementary figure. ‡ To whom correspondence should be addressed. Fax: 41-61-697-45- The abbreviations used are: PAMP, pathogen-associated molecular 27; E-mail: [email protected]. pattern; CSP, cold shock protein; CSD, cold shock domain. This paper is available on line at http://www.jbc.org 6201 This is an Open Access article under the CC BY license. 6202 Bacterial Cold Shock Proteins as Elicitors in Tobacco Peptidoglycan has long been known as a PAMP signaling pres- ence 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 total lyophilized M. lysodeikticus bacteria proved to be a far more potent 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 Materials—Peptides were synthesized by F. Fischer (Friedrich Mi- escher-Institute, Basel) or by Bio-Synthesis Inc. (Lewisville, TX). Pep- tides were dissolved in H O (stock solutions of 1 to 10 mM) and diluted in a solution containing 0.1% bovine serum albumin 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, Braunschweig, BRD) and grown in King’s B broth at 26 °C on a rotary shaker. Bacteria were harvested by centrifugation, washed once with H O, and resuspended in H O (10% of original volume). Crude bacterial elicitors were prepared by boiling the bacterial suspensions for 5–10 min and removing of bacterial debris by centrifugation. Lyophilized bacteria of M. lysodeik- ticus (Sigma) and the peptidoglycan fraction from M. lysodeikticus (Fluka, Buchs, Switzerland) were applied as suspensions in H O. The bacterial preparation “messenger” was obtained from EDEN Bioscience (Bothell, WA). Purification of Elicitor from M. lysodeikticus—Elicitor activity was purified from the lyophilized preparation of M. lysodeikticus (Sigma). Ten g of the lyophilisate was suspended in 100 ml of H O and heated for 10 min at 95 °C. After centrifugation (30 min at 10,000  g) the FIG.1. Extracellular alkalinization of tobacco cells in re- supernatant was mixed with 1 volume of acetone and the precipitate sponse to treatment with preparations from M. lysodeikticus. A, formed after overnight incubation at 20 °C was removed by centrifu- alkalinization in response to treatment with 10 g/ml lyophilized gation. The acetone concentration was brought to 80% (v/v) and the M. lysodeikticus cells or to 100 g/ml of the peptidoglycan fraction precipitate formed after4hat 20 °C was collected by centrifugation. derived from M. lysodeikticus (peptidoglycan). B, effect of protease K This precipitate was dissolved in 20 mM Tris-HCl, pH 7.5, and passed treatment on alkalinization-inducing activity of lyophilized M. lysodeik- ticus bacteria and the peptidoglycan preparation. Different doses of over an anion-exchange column with diethylaminoethyl-cellulose (DE- lyophilized M. lysodeikticus bacteria (closed circles), bacteria after pre- cellulose, Whatman) equilibrated with 20 mM Tris-HCl, pH 7.5. Activity treatment with protease K (overnight incubation with 1 mg/ml protease eluting in the flow-through was concentrated by acetone precipitation K, open circles), peptidoglycan (closed triangles), and peptidoglycan (80% acetone) and separated on a Sephacel C8 reversed phase column after pretreatment with protease K (open triangles) were added to (Amersham Biosciences AB) at pH 6.5 (10 mM phosphate buffer, pH 6.5, aliquots of the cell culture and the pH change measured after 15 min as solvent A and 80% acetonitrile, 20% phosphate buffer as solvent B). (initial pH 4.8). The two fractions containing the highest elicitor activity were pooled, pH adjusted to 3.5, and rerun on a Sephacel C8 reversed phase column at pH 3.5 (0.1% trifluoroacetic acid in H O at pH 3.5 as solvent A and RESULTS 80% acetonitrile, 20% H O with 0.1% trifluoroacetic acid as solvent B). Extracellular Alkalinization in Cultured Tobacco Cells Plant Cell Cultures—The tobacco (Nicotiana tabacum L.) cell culture line 275N, originally derived from pith tissue of Havanna 425 plants, Treated with Preparations from M. lysodeikticus—Peptidogly- was maintained and subcultured as described before (15) in a Murash- can, an essential cell wall component of all bacteria, acts as one ige-Skoog based medium. Cells were maintained as suspension cultures of the PAMPs signaling the presence of Gram-positive bacteria and used 4 to 10 days after subculture for experiments. Cell cultures of to the innate immune system in animals (1, 10). In initial tomato (“line Msk8” (16)), potato (17), Lycopersicon peruvianum (18), experiments we tested preparations containing peptidoglycan and A. thaliana (19) were cultured as described elsewhere. for induction of extracellular alkalinization in plant cells cul- Alkalinization Response—To measure alkalinization of the growth medium (the alkalinization response), 3-ml aliquots of the cell suspen- tured in liquid medium. Medium alkalinization, occurring as a sions were placed in open, 20-ml vials on a rotary shaker at 120 to 150 consequence of altered ion fluxes across the plasma membrane, cycles per min. Using small combined glass electrodes (Metrohm, Heri- can serve as a convenient, rapid, sensitive and quantitative sau, Switzerland) extracellular pH values were either recorded contin- bioassay to study elicitor perception by plant cells (16). As a uously with a pen recorder or measured after 15 or 20 min of treatment. source of peptidoglycan we used preparations from M. lysodeik- Oxidative Burst and Ethylene Biosynthesis in Leaf Tissue—Fully ex- ticus (S. aureus) because lyophilized bacteria and a peptidogly- panded leaves of different plant species were cut in 2-mm slices and floated on H O overnight. For measuring the oxidative burst, active oxy- can fraction are commercially available. Also, as deduced from gen species released by the leaf tissue were measured by a luminol-de- the genomes of the three fully sequenced strains of S. aureus pendent assay (20). Slices were transferred to assay tubes (2– 4 slices (M. lysodeikticus) that do not encode proteins resembling corresponding to 20 mg of fresh weight) containing 0.1 ml of H O flagellin, these preparations should be free of elicitor-active supplied with 20 M luminol and 1 g of horseradish peroxidase (Fluka). flagellin that could interfere in the assays. No alkalinization Luminescence was measured in a LKB 1250 luminometer (LKB Wallac, was observed in the tomato cells of the line Msk8 after treat- Turku, Finland) for 20 min after the addition of the test solution. For assaying ethylene production, leaf slices (50 mg of fresh per ment with lyophilized M. lysodeikticus bacteria or the pepti- assay) were transferred to 6-ml glass tubes containing 1 ml of an doglycan fraction derived from these bacteria (data not shown). aqueous solution of the peptide being tested. Vials were closed with Whereas these negative results confirmed the absence of elic- rubber septa and ethylene accumulating in the free air space was itor-active flagellin they did not provide evidence for a chemo- measured by gas chromatography after 2 to 2.5 h of incubation. perception system responding to peptidoglycan in the tomato Reproducibility—The results shown in the figures represent single cells. When tested on tobacco cells, however, both preparations experiments that are representative for several independent repetitions. of M. lysodeikticus caused rapid and strong medium alkalin- Bacterial Cold Shock Proteins as Elicitors in Tobacco 6203 phy at pH 6.5 activity eluted as a single peak (Fig. 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 A . Separa- tion by SDS-PAGE (14% (w/v) acrylamide) showed a band migrating with an apparent molecular weight of 7,000 to 9,000 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 iden- tified the protein as a cold shock protein (CSP). In Fig. 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 bac- terial cold shock proteins present in the data bank. Identification of the “Cold Shock Domain” (CSD) as the Elic- itor-active Epitope—In attempts to localize the elicitor activity FIG.2. Purification of the alkalinization-inducing activity by to a particular domain of the protein the purified CSP was reversed phase chromatography. An extract from lyophilized subjected to peptide cleavage. Digestion with trypsin, Lys-C, or M. lysodeikticus bacteria, pre-purified by ion-exchange chromatography Glu-C (V8 protease) abolished the activity and did not result in as described under “Experimental Procedures,” was fractionated on a C8 reversed phase column at pH 6.5. Fractions with highest activity, smaller fragments with elicitor activity (data not shown). As in eluting between 24 and 28 min, were re-chromatographed on the C8 previous work with bacterial flagellin (8) we speculated that column at pH 3.5. Upper panel shows elution profile (OD ) of the first plant cells might have a perception system for the most char- run at pH 6.5 (10 mM phosphate buffer) and the second run at pH 3.5 acteristic and most conserved domain of the CSPs. Although (0.1% trifluoroacetic acid). Lower panel shows extracellular alkaliniza- tion in tobacco cells (pH ) induced by aliquots of the fractions 15min these small bacterial proteins show a high overall homology eluting in the first (open bars) and second (open bars) runs. they are particularly conserved in a domain close to the N terminus. Based on the consensus sequence of bacterial CSPs, ization (Fig. 1). As shown in the examples in Fig. 1A, extracel- a 22-amino acid peptide spanning this domain was synthesized lular pH started to increase after a lag of 3 to 5 min and (Fig. 3, underlined sequence) and tested for induction of alka- reached a maximum after 10 to 15 min. Depending on the cell linization in tobacco cells. This peptide, termed csp22, proved density and the initial pH of different batches of the cell culture even more active than the intact CSP purified from M. lyso- the amplitude of the alkalinization response (pH ) varied max deikticus and induced medium alkalinization with an EC of from 1.2 to 2 pH units for lyophilized bacteria and from 0.6 to 0.1 nM (Fig. 4). 1.4 pH units for the peptidoglycan fraction, respectively. In To further delineate the epitope that activates responses in aliquots from a given batch of cells, however, pH was max the plant cells, peptides lacking varying numbers of amino acid highly reproducible and consistently showed a bigger response residues from the N- or C-terminal end were synthesized and for the preparation of total bacteria than for the peptidoglycan assayed for activity in dose-response curves as described above fraction. The responses of the cells to both preparations of for csp22 and CSP. The amino acid sequences and the EC M. lysodeikticus were dose-dependent and lower, nonsaturat- values are summarized in (Fig. 5). Omitting 5 amino acid ing doses led to prolonged lag phases, smaller maximal pH residues from the N terminus of csp22 reduced activity only increases, and shortened durations of medium alkalinization. slightly (EC of 1.2 nM) but removal of the Lys residue at The pH change occurring within 15 min (pH ) of treat- 15min position 6 showed a much stronger effect (EC of 220 nM) and ment was a steady function of the dose applied and was used as further trimming by 4 amino acid residues resulted in an a parameter to compare the relative strength of the two prep- inactive peptide. The peptide termed csp15, comprising the 15 arations of M. lysodeikticus (Fig. 1B). Half-maximal stimula- amino acid residues central to csp22, was nearly as active as tion was observed with 30 g/ml of the peptidoglycan (EC ) csp22 (EC of 0.3 nM) and served as a core peptide for testing and 1 g/ml with the lyophilized bacteria, respectively. Treat- structural analogues with replacements of single amino acid ment with protease K strongly affected the activity of the residues with alanine. Csp15-Ala3, csp15-Ala4, csp15-Ala8, bacterial preparation resulting in a 200-fold higher EC value and csp15-Ala12 all exhibited at least 1000-fold reduced activ- (200 g/ml) but led only to a 3-fold increase for the EC value ity compared with csp15. Csp15-Ala10 was inactive even at the of the peptidoglycan fraction (Fig. 1B). These results provided highest concentration of 100 M tested (Fig. 5). In contrast, preliminary evidence for the presence of two distinct elicitor substitution of Phe at position 10 with a Tyr residue resulted in activities in M. lysodeikticus: a nonproteinaceous elicitor in the a peptide with full activity in the tobacco cells (Fig. 5). Among peptidoglycan fraction and a second, potent, proteinaceous elic- the peptides with single substitutions with Ala only csp15-Ala7 itor predominating in the total bacteria preparation. On a per showed no significant decrease in activity. Interestingly, the weight basis the proteinaceous factor was more than 100-fold Glu at this position also shows the least conservation in the more active than the peptidoglycan factor and further work different sequences of bacterial CSPs (Fig. 3). focused on the characterization of this new protein elicitor. The three-dimensional structure has been determined for Purification of an Elicitor-active Protein from M. lysodeikti- the major bacterial cold shock proteins CspB from Bacillus cus and Its Identification as Bacterial Cold Shock Protein—The subtilis (Molecular Modeling Database number 3622; Protein elicitor activity, extracted from the crude preparation of Data Bank number 1CSP) (21) and CspA (CS7.4) from E. coli M. lysodeikticus (S. aureus), was heat-stable (5 min, 95 °C), passed ultrafilters with a molecular weight cut-off of 10,000, (Molecular Modeling Database number 1677; Protein Data Bank number 1MJC) (22). CspB forms a dimer whereas CspA and was inactivated by treatment with trypsin (data not shown), indicating that the elicitor activity was attributable to occurs as monomer. Besides this difference of dimerization the structures of both proteins are very similar, forming compact a peptide or small protein. Activity was purified on a Sephacal C8 reversed phase column (Fig. 2A). In the first chromatogra- -barrel structures built up from five antiparallel -strands 6204 Bacterial Cold Shock Proteins as Elicitors in Tobacco FIG.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 data base. Partial sequence for M. lysodeikticus represents information obtained from the purified protein after tryptic digest and Edman degradation of some of the peptides. Csp22 and csp15 denote peptides synthesized according to the consensus sequence. FIG.4. Elicitor activity of the purified Csp and synthetic pep- tides 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)or15 (csp15) amino acid residues of the conserved region from bacterial CSPs as indicated in the legend to Fig. 3. with connecting turns and loops. Fig. 6 shows models (second- ary structure and a three-dimensional ribbon model) of the FIG.5. Alkalinization-inducing activity of csp-related pep- molecular structure of CspB, highlighting the domain spanned tides. EC values were determined from dose-response curves ob- by the csp15 peptide. Clearly, elicitor activity can be attributed tained for the different peptides. Specific activity relative to the activity of the most active peptide csp22 (hatched bar,EC of 0.1 nM). Loga- to the domain formed by antiparallel strands 1 and 2 and the rithmic scaling was used to indicate residual activity in some of the loop L1. This domain includes a RNA-binding motif known as peptides. No activity could be detected with peptides denoted with RNP-1 (also termed RNP-CS) and exposes a cluster of aromatic asterisks (relative activity 10 ). 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 stitutions leading to strong or complete reduction in affinity of the interaction of the protein with nucleic acids (23). In Table I, CspB for nucleic acids exhibited strongly reduced elicitor ac- these single amino acid replacements and their effects on nu- tivity in tobacco cells (higher EC values). The substitution of cleic acid binding were compared with the corresponding amino Phe by Tyr at the position that corresponds to residue 10 in acid changes in csp15 and their effects on elicitor activity in csp15 did not affect affinity of CspB for nucleic acids and also tobacco. All the substitutions in csp15 that correspond to sub- did not alter elicitor activity. Bacterial Cold Shock Proteins as Elicitors in Tobacco 6205 FIG.6. Structure of bacterial CSPs. A, schematic scheme for secondary struc- ture of CspB from B. subtilis from Schin- delin et al. (21) with domains spanned by csp15 (hatched part) and the RNA-bind- ing motifs RNP-1 (KGFGFITP) and RNP-2 (VFVHF) indicated. B, structure of CspB monomer (Molecular Modeling Da- tabase number 3622; Protein Data Bank number 1CSP (21)) drawn with WebLab ViewerLite (Molecular Simulations Inc., Cambridge, UK) with side groups exposed in the domain spanned by csp15. TABLE I and rice (data not shown). Negative results with particular Comparison of mutations in cspB affecting DNA-binding and single lines of cell cultures do not allow concluding on the absence of amino acid changes in csp15 on elicitor activity a perception system in the corresponding plant species because -Fold reduction of this perception system might be not expressed or might have Substitution Substitution in -Fold increase for a affinity for a in CspB csp15 EC oligonucleotide been lost during the years of growth in vitro. Induced release of active oxygen species, an oxidative burst, CspB (wt) csp15 1 1 and increased biosynthesis of the stress hormone ethylene are K7Q csp15-Ala2 3 1000 W8A csp15-Ala3 5 1000 responses characteristic for plants under attack by pathogens F15A csp15-Ala10 Inactive 10,000 or treated by elicitor preparations (6, 25). We used these re- F17A csp15-Ala12 Inactive 2000 sponses to monitor responsiveness toward CSP-derived elici- F15Y csp15-Tyr10 1 1 tors in leaf tissues from different plant species. As exemplified F17Y 6 Not tested in Fig. 7 for leaf tissue from tomato, a rapid, significant in- Csp15 comprises amino acids 5 to 20 in CspB. EC values for the crease in ethylene biosynthesis and in active oxygen species induction of the alkalinization response from Fig. 5. Effect of site-directed point mutations in CspB on the binding of was observed after treatment with csp15 but not after treat- single stranded DNA oligonucleotides containing Y-box motif. Data ment with the same dose of csp15-Nsyl. Similarly, clear CSP- from Table I in Schro ¨ der et al. (23). dependent induction of ethylene biosynthesis and oxidative burst was observed in tobacco and several other solanaceous plants including potato (Solanum tuberosum), Solanum dulca- Activity of Peptides Representing Homologous Domains Oc- mara, Scopolia carniolica, and Mandragora officinarum.In curring of Proteins from Plants and Animals—Csp-related pro- contrast, no response could be detected from leaf tissue and cell teins are common to all eubacteria and they usually form a cultures of A. thaliana, cucumber, and rice. Also, no signs of a small family of proteins that include both cold-inducible and hypersensitivity response could be detected after injection of noninducible members (24). The domain containing the RNP-1 CSP peptides to leaves of tobacco or tomato (data not shown). motif is conserved also in many eukaryotic proteins that bind to In summary, a perception system for CSP-related elicitors is RNA or DNA. Examples for proteins with this so-called CSD common to solanaceous plants but has not yet been found include human and animal transcription factors recognizing outside of this plant family. the Y-box sequence and glycine-rich RNA-binding proteins oc- The Inactive Peptide csp15-Ala10 Antagonizes Elicitor Activ- curring in plants (see Supplementary Materials for gene struc- ity of CSP—Peptides lacking either 4 amino acids from the ture and alignment with bacterial CSPs). Peptides correspond- C-terminal part (csp11) or 6 amino acids from the N-terminal ing to the homologues of a human Y-box protein and a Gly-rich part spanned by csp15 lacked activity even when applied in protein from Nicotiana sylvestris were synthesized and tested micromolar concentrations (data not shown). No response was for activity. The csp15 homologue from the human Y-box pro- observed also by application of these two peptides in combina- tein was inactive whereas the peptide representing N. sylves- tion (data not shown). Truncated forms of the biologically ac- tris sequence induced responses with an EC of 300 nM and tive peptides systemin and flg22 were previously found that was thus 1000-fold less active than the csp15 representing the bacterial sequence (Fig. 5). showed characteristics of competitive antagonists for the re- spective nontruncated agonistic peptides (26 –28). No antago- Responses Induced by CSP in Different Plant Species—We examined cell cultures derived from other plant species for nistic activity could be observed for the two truncated CSP peptides described above (data not shown). In contrast, csp15- alkalinization in response to CSP-related elicitors. Responses with characteristics similar to the ones of the tobacco line 275N Ala10, also inactive as agonist (Fig. 5), did exhibit antagonistic activity and suppressed responses induced by csp15 (Fig. 8). were observed also with a second line of tobacco, originating from a plant of the variety SR1, with a cell line derived from When added concomitantly with 3 nM csp15, a concentration of 3 M strongly inhibited induction of the alkalinization response potato and a cell culture from L. peruvianum (data not shown). In contrast, no responses could be detected in the cell culture (Fig. 8A, “0 min”). Complete inhibition was observed when line msk8, originally derived from a cross of Lycopersicon escu- csp15-Ala10 was added 30 s before the agonist but progres- lentum with L. peruvianum, and in cell lines from A. thaliana sively weaker effects were observed when the antagonist was 6206 Bacterial Cold Shock Proteins as Elicitors in Tobacco FIG.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 M csp15 or 1 M csp15-Nsyl as indicated. Bars and error bars show mean  S.D. of n 4 replicates. B, luminescence of leaf slices in a solution with luminol and peroxidase after treatment with 1 M concentrations of csp15 or csp15-Nsyl as indicated. Light emission at the very beginning of the experiments is caused by phos- phorescence of the green tissue. 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 FIG.8. The peptide csp15-Ala10 acts as competitive suppressor observed with unrelated elicitors like flagellin and chitin frag- of csp-related elicitors. A, extracellular alkalinization in suspension- ments (data not shown). Inhibition of CSP-related activity by cultured tobacco cells treated with 3 nM csp15 and 3 M csp15-Ala10 at csp15-Ala10 was competitive and could be overcome by increas- the time points indicated (slanted arrows). Extracellular pH of un- treated cells was 4.9 and addition of 3 M csp15-Ala10 alone did not ing concentrations of active peptide or intact CSP. As shown in cause significant pH changes. B, alkalinization induced by different the example in Fig. 8B, this resulted in an increase of the EC doses of M. lysodeikticus (circles) and peptidoglucan preparation (trian- for the CSP containing preparation of M. lysodeikticus bacteria gles) in cells without pretreatment (closed symbols) or cells pretreated from 1 g/ml in the absence of the antagonist, to 20 g/ml in for 3 min with 10 M csp15-Ala10 (open symbols). C and D, alkaliniza- tion induced by the harpin-containing preparation messenger (1 g/ml, the presence of 3 M csp15-Ala10, respectively. In contrast, no C) and by a crude extract from A. tumefaciens (1 l/ml, D) in cells shift in dose response was observed with the peptidoglycan without pretreatment or cells pretreated for 3 min with 3 M csp15- fraction (Fig. 8B). These results confirm predominance of the Ala10 as indicated. CSP-related elicitor in the crude bacterial preparation and the presence of an activity unrelated to CSP in the peptidoglycan rence of CSP-related elicitor activity in extracts from different, preparation. if not all, bacteria. Tobacco cells were found to respond to crude extracts from all DISCUSSION bacterial species tested (n  20). The antagonist csp15-Ala10 could serve as a diagnostic tool to test for the presence of Various types of living bacteria as well as preparations of csp-related activity. For example, cells responded with strong heat-killed bacteria can trigger rapid responses in plant cell alkalinization when treated with messenger, an extract from cultures and defense responses in intact plant tissues (20, 30, E. coli expressing transgenic harpin from Erwinia amylovora 31). Flagellin (8) and lipopolysaccharides (32) have been iden- (Fig. 8C). Interestingly, at least at limiting doses of messenger tified as common bacterial determinants or PAMPs that act as applied, activity was fully antagonized by csp15-Ala10. This elicitors of defense responses in plant cells. In this report we indicated that a CSP-related stimulus and not harpinEa, pre- add CSPs as a further bacterial PAMP that acts as an elicitor viously reported to act as an inducer of alkalinization in tobacco of defense responses in plants. (29), was the activity predominating in this preparation. As Cold shock proteins were named based on the original obser- shown for the example of an extract from A. tumefaciens in Fig. vation that rapid cooling with a T of 10 °C (cold shock) 8D, csp15-Ala10 antagonized also the alkalinization inducing induces accumulation of specific proteins in many bacterial activity of crude extracts from the plant-associated species species. The major CSPs are small, 7.4 kDa proteins that A. tumefaciens, R. meliloti, and X. campestris, extracts that belong to a family of highly conserved proteins commonly oc- were previously found to be devoid of elicitor-active flagellin curring in all bacteria. At least some members of this family (8). In summary, these results demonstrate the common occur- are (also) constitutively expressed or are induced under stress Bacterial Cold Shock Proteins as Elicitors in Tobacco 6207 conditions different from cold shock (24). For example, the mediated process for both elicitors. Thus, we hypothesize that family of CspA-like proteins in E. coli consists of eight members csp perception occurs via a csp receptor that functions in a manner similar to the receptor for flagellin. Activation of this (CspA through CspH) and only CspA, CspB, and CspG are putative CSP receptor might also involve two consecutive steps cold-inducible. Thus, despite their name, members of the CSP family occur also in bacteria not subjected to a cold shock 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 treatment. CSPs are implicated in various cellular processes, of csp15, Phe in csp15, or Tyr in csp15-Tyr10, appears neces- including cellular growth and adaptation to low temperatures, sary for this second step to activate the receptor. The antago- nutrient stress, and stationary phase. CSPs bind to nucleic nist csp15-Ala10, apparently, does not undergo the second, acids and appear to function as RNA chaperones and anti- locking step and interacts with the receptor site in a more terminators of translation (33). readily reversible manner. This could explain the high excess of A shift to low temperature induces also a set of specific pro- antagonist csp15-Ala10 over csp15 required to block elicitor teins in plants (34). Some of these cold-regulated proteins are action completely and the apparent inefficiency of the antago- small hydrophilic proteins of 6.6 kDa (35), like the major bacte- nist when applied subsequent to the csp agonists (Fig. 8). rial CSP but they are nonhomologous in sequence, and their Proteins with a cold shock domain comprising the RNP-1 physiological function in the cold acclimation process remains motif are conserved also in eukaryotes and have been identified unknown. However, many eukaryotes including plants and ani- also in genes of A. thaliana and N. sylvestris. Although clearly mals have proteins with a nucleic acid-binding domain that homologous, the sequences corresponding to the elicitor-active shows a strikingly high homology and similar RNA-binding prop- epitope show some differences in comparison to the bacterial erties to bacterial CSPs (36). It is this universally conserved consensus. The synthetic peptide csp15-Nsyl, representing the domain, also termed CSD, that contains the RNP-1 motif and the least divergent form of this domain in genes known from epitope found to act as elicitor of tobacco cells. N. sylvestris, indeed did show some activity in the bioassay The elicitor activity of bacterial CSPs could be localized to a with tobacco cell. However, the specific activity of this peptide stretch of 15 amino acid residues that forms a loop with two was 1000-fold lower than that of csp15 representing the bac- antiparallel -strands and exposes a series of aromatic and terial epitope. A lower specific activity could be counterbal- basic amino side chains to the surface of the protein. It is this anced by the presence of high local concentrations of the stim- epitope that exhibits highest conservation between the differ- ulus. In initial attempts with extracts of tobacco plants or cells ent bacterial CSPs and, as has been demonstrated by site- from tissue culture we failed to detect factors with CSP-like directed mutagenesis of CspB from B. subtilis, is essential for activity in bioassays (data not shown). Thus, at present, we do the interaction of the protein with nucleic acids (23). Most not have evidence for endogenous factors stimulating tobacco notably, synthetic peptides with amino acid sequences reflect- via the CSP perception system described in this report. Endog- ing the changes leading to reduced or abolished binding to enous factors of tobacco, capable of stimulating medium alka- nucleic acids of CspB were also strongly affected in elicitor linization in cultured cells, have recently been described (38) activity (Table I). This strong correlation raises the question of but these peptidic factors show no apparent homology to CSPs. whether some sort of nucleic acid might be involved in the Bacterial CSPs are molecules highly characteristic for bac- perception process by the plant cell. However, intact CSPs and teria in general and could thus serve as PAMPs signaling the the csp-derived peptide elicitors have characteristics of mole- presence of bacteria to the plant cells. An obvious problem with cules that are not permeable for membranes and the first this hypothesis that is the localization of these proteins, which responses to subnanomolar concentrations of csp-derived elic- are generally assumed to function in the cytoplasm of the itors occur after a lag phase of less than 2 min. These charac- bacteria. So far, CSPs have not been reported to be exported or teristics rather suggest a chemoperception system with a spe- exposed to the surface by intact bacteria and, consequently, cific, high affinity primary interaction site in the apoplast, most CSPs are probably not directly detectable by a chemopercep- likely the plasma membrane, of the plant cells. The strong tion system assumed to reside on the surface of the plant cells. correlation of nucleic acid binding and elicitor activity might Further studies will be required to test whether CSPs are thus reflect evolution of a chemoperception system directed at released from bacteria during invasion of their plant hosts. A a particular surface epitope of the bacterial CSPs that is under release could be based on a bacterial export system activated in a high selective pressure for retaining functionality of the the course of the infection process, or it could result from protein. bacterial or plant processes causing a general leakiness of the Perception of csp-related elicitors resembles the chemoper- bacteria. As demonstrated for bacteria under mild osmotic ception system for flagellin-derived elicitors studied before (8, shock (39), leakiness of bacteria leading to release of small 28). In both cases, elicitor activity could be attributed to an cytoplasmic proteins might be more common than suggested by epitope of 15 amino acids representing the most conserved studies under optimal media conditions used to grow bacterial part of the respective protein. Both elicitors are active at sub- cells in the laboratory. Precedence for cytoplasmic components nanomolar doses and activity is highly dependent on the gen- of bacteria that act as PAMPs and stimulate the innate im- uine amino acid sequence of the conserved domain. “Mutation- mune responses via Toll-like receptors in animals include “non- al” analysis using structural analogues of the elicitors allowed secreted” components such as the heat shock protein HSP60 identification of peptides lacking elicitor activity but exhibiting (40) and bacterial DNA (12). Bacterial DNA is recognized via properties of competitive antagonists. Perception of flagellin its content of nonmethylated CpG oligonucleotides and this was shown to involve a specific, high affinity binding site and PAMP was successfully applied as a potent immunostimula- the membrane-bound receptor kinase FLS2 (9, 37). A model tory factor (41, 42) but the process leading to release of the involving a two-step process for receptor activation was pro- DNA from the bacteria has not been elucidated. Similarly, no posed to explain the effects of agonistic and antagonistic pep- process that secretes HSP60 from intact bacterial cells has tides (28). At present, experiments that directly demonstrate a been described. HSP60 is well conserved from microbes to receptor site for the csp elicitors are lacking. Nevertheless, humans and HSP60 from both mammalian and microbial CSP- and flagellin-derived elicitors induce the same set of sources can trigger inflammatory responses via Toll-like recep- responses with similar kinetics, indicating a similar, receptor- tor 4 (40), suggesting that Toll-like receptor 4 may detect both 6208 Bacterial Cold Shock Proteins as Elicitors in Tobacco protein sequencing services and mass spectrometry, and Martin endogenous and exogenous ligands as alarm signals. Exposure Regenass for maintaining the cell cultures and technical assistance. of endogenous HSP60 to the Toll-like 4 receptor could be en- visaged to occur via release from wounded or injured cells and REFERENCES perception by the receptor on different, intact cells. As dis- 1. Medzhitov, R., and Janeway, C. A., Jr. (2002) Science 296, 298 –300 2. Yang, Y., Shah, J., and Klessig, D. F. (1997) Genes Dev. 11, 1621–1639 cussed above, it remains to be seen whether the chemopercep- 3. Dangl, J. L., and Jones, J. D. (2001) Nature 411, 826 – 833 tion system for CSPs described in this report might similarly 4. Schneider, D. S. (2002) Cell 109, 537–540 5. Heath, M. C. (2000) Curr. Opin. Plant Biol. 3, 315–319 react to both endogenous and exogenous ligands. 6. Boller, T. (1995) Annu. Rev. Plant Physiol. Plant Mol. Biol. 46, 189 –214 Peptidoglycan consists of a glycan backbone with alternating 7. Hahn, M. G. (1996) Annu. Rev. Phytopathol. 34, 387– 412 1– 4-linked residues of N-acetyl-D-glucosamine and muramic 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 acid and forms the major component of the cell wall in Gram- 10. Aderem, A., and Ulevitch, R. J. (2000) Nature 406, 782–787 positive bacteria. Peptidoglycans, sensed by a family of pepti- 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., doglycan recognition proteins that are conserved from insects Hatakeyama, T., Aoyagi, H., Kurazono, H., Moss, J., and Hirayama, T. to humans (43), are important PAMPs for the innate immunity (2001) J. Biol. Chem. 276, 30521–30526 of animals. The peptidoglycan preparation of M. lysodeikticus 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, also triggered elicitor responses in tobacco cells. We have not 1099 –1103 yet characterized this activity in detail and cannot exclude that 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 it is because of a minor component or a “contaminant” of the 15. Felix, G., and Meins, F., Jr. (1987) Planta (Heidelberg) 172, 386 –392 peptidoglycan fraction. Nevertheless, this nonproteinaceous 16. Felix, G., Regenass, M., and Boller, T. (1993) Plant J. 4, 307–316 17. Schweizer, P., Felix, G., Buchala, A., Muller, C., and Me ´ traux, J. P. (1996) factor is perceived as a quality of stimulus distinct from CSPs Plant J. 10, 331–341 and provides evidence for a further bacterial PAMP with elic- 18. Felix, G., and Boller, T. (1995) Plant J. 7, 381–389 itor activity in tobacco cells. Thus, similar to the chemopercep- 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, tion systems for a variety of fungal-derived PAMPs, perception 974 –978 of bacteria by plant cells appears not to depend on a single 21. Schindelin, H., Marahiel, M. A., and Heinemann, U. (1993) Nature 364, 164 –168 bacterial factor but rather involves several different factors, 22. Schindelin, H., Jiang, W., Inouye, M., and Heinemann, U. (1994) Proc. Natl. including at least flagellin (8), lipopolysaccharides (32), pepti- Acad. Sci. U. S. A. 91, 5119 –5123 23. Schro ¨ der, K., Graumann, P., Schnuchel, A., Holak, T. A., and Marahiel, M. A. doglycan, and CSP. This redundancy, characteristic also for the (1995) Mol. Microbiol. 16, 699 –708 recognition mechanisms in the innate immune system of ani- 24. Thieringer, H. A., Jones, P. G., and Inouye, M. (1998) Bioessays 20, 49 –57 mals, points at possible difficulties with approaches to demon- 25. Lamb, C., and Dixon, R. A. (1997) Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 251–275 strate a direct physiological role for any particular of these 26. Pearce, G., Johnson, S., and Ryan, C. A. (1993) J. Biol. Chem. 268, 212–216 chemoperception systems for plant defense. Inhibition or 27. Meindl, T., Boller, T., and Felix, G. (1998) Plant Cell 10, 1561–1570 28. Meindl, T., Boller, T., and Felix, G. (2000) Plant Cell 12, 1783–1794 knockout of only one of the systems might be without a strong 29. Wei, Z. M., Laby, R. J., Zumoff, C. H., Bauer, D. W., He, S. Y., Collmer, A., and effect on the overall recognition system. In contrast to bacterial Beer, S. V. (1992) Science 257, 85– 88 “avirulence factors,” which act as elicitors that are specific and 30. Atkinson, M., Bina, J., and Sequeira, L. (1993) Mol. Plant-Microbe Interact. 6, 253–260 unique for a particular pathogen, the structures recognized as 31. Jakobek, J. L., and Lindgren, P. B. (1993) Plant Cell 5, 49 –56 PAMPs are essential or “vital factors” for the functioning of the 32. Meyer, A., Puhler, A., and Niehaus, K. (2001) Planta (Heidelberg) 213, 214 –222 bacterial organisms in general and cannot easily be changed, 33. Bae, W., Xia, B., Inouye, M., and Severinov, K. (2000) Proc. Natl. Acad. Sci. removed, or mutated for probing their role in plant defense. U. S. A. 97, 7784 –7789 In summary, our results provide evidence for novel bacterial 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, elicitors, cold shock protein, for which tobacco and other So- 293–299 lanaceae have evolved specific and sensitive chemoperception 36. Graumann, P. L., and Marahiel, M. A. (1998) Trends Biochem. Sci. 23, 286 –290 systems. The accuracy and sensitivity of the perception system 37. Bauer, Z., Gomez-Gomez, L., Boller, T., and Felix, G. (2001) J. Biol. Chem. 276, for the CSP domain comprising the RNP-1 motif detailed in 45669 – 45676 38. Pearce, G., Moura, D. S., Stratmann, J., and Ryan, C. A., Jr. (2001) Proc. Natl. this report indicate a receptor mechanism involving a high Acad. Sci. U. S. A. 98, 12843–12847 affinity binding site on the surface of the plant and should 39. Berrier, C., Garrigues, A., Richarme, G., and Ghazi, A. (2000) J. Bacteriol. 182, provide the basis for further work to identify the protein acting 248 –251 40. Seo, S., Okamoto, M., Seto, H., Ishizuka, K., Sano, H., and Ohashi, Y. (1995) as pattern recognition receptor for CSP. Science 270, 1988 –1992 41. Krieg, A. M. (2002) Trends Immunol. 23, 64 – 65 Acknowledgments—We thank Franz Fischer (Friedrich Miescher- 42. Krieg, A. M. (2002) Annu. Rev. Immunol. 20, 709 –760 Institute, Basel) for the synthesis of various peptides, Renate Matthies, 43. Kang, D., Liu, G., Lundstrom, A., Gelius, E., and Steiner, H. (1998) Proc. Natl. Daniel Hess, and Jan Hofsteenge (Friedrich Miescher-Institute) for Acad. Sci. U. S. A. 95, 10078 –10082 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Molecular Sensing of Bacteria in Plants

Felix, Georg; Boller, Thomas
Journal of Biological Chemistry – Feb 1, 2003
Download PDF
8 pages

Loading next page...
 
/lp/unpaywall/molecular-sensing-of-bacteria-in-plants-aqsSdrL05t

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m209880200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 8, Issue of February 21, pp. 6201–6208, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. THE HIGHLY CONSERVED RNA-BINDING MOTIF RNP-1 OF BACTERIAL COLD SHOCK PROTEINS IS RECOGNIZED AS AN ELICITOR SIGNAL IN TOBACCO*□ S Received for publication, September 26, 2002, and in revised form, November 21, 2002 Published, JBC Papers in Press, December 5, 2002, DOI 10.1074/jbc.M209880200 Georg Felix‡ and Thomas Boller From the Friedrich Miescher-Institute, P. O. Box 2543, CH-4002 Basel, Switzerland perception systems for patterns characteristic for entire groups To detect microbial infection multicellular organisms have evolved sensing systems for pathogen-associated or classes of microorganisms, and they respond to these general molecular patterns (PAMPs). Here, we identify bacterial elicitors with activation of signaling pathways that initiate cold shock protein (CSP) as a new such PAMP that acts defense mechanisms (5). This is highly reminiscent of innate as a highly active elicitor of defense responses in to- immunity in animals and humans. Among the elicitors that bacco. Tobacco cells perceive a conserved domain of represent patterns characteristic for fungi are cell wall compo- CSP and synthetic peptides representing 15 amino acids nents like glucans, chitin and chitosan oligosaccharides, pep- of this domain-induced responses at subnanomolar con- tides and proteins with fungal-specific N-glycosylation, and the centrations. Central to the elicitor-active domain is the membrane component ergosterol (6, 7). Similarly, cells of many RNP-1 motif KGFGFITP, a motif conserved also in many plant species have a perception system for the common bacte- RNA- and DNA-binding proteins of eukaryotes. Csp15- rial surface protein flagellin, the building block of the flagella Nsyl, a peptide representing the domain with highest (8). Perception of flagellin by Arabidopsis thaliana was shown homology to csp15 in a protein of Nicotiana sylvestris to depend on FLS2, a membrane-bound receptor kinase protein exhibited only weak activity in tobacco cells. Crystallo- with an extracellular leucine-rich repeat (9). Bacterial flagellin graphic and genetic data from the literature show that has recently also been identified as one of the “pathogen-asso- the RNP-1 domain of bacterial CSPs resides on a pro- ciated molecular patterns” (PAMPs) that activate the innate truding loop and exposes a series of aromatic and basic immune system of humans and animals (10) via the Toll-like side chains to the surface that are essential for the nu- receptor 5 (11, 12). Thus, perception of general elicitors in cleotide-binding activity of CSPs. Similarly, these side plants resembles perception of PAMPs in the innate immune chains were also essential for elicitor activity and re- system of animals with respect to the type of molecules per- placement of single residues in csp15 with Ala strongly ceived, the characteristics of pattern recognition receptors in- reduced or abolished activity. Most strikingly, csp15- Ala10, a peptide with the RNP-1 motif modified to volved, as well as some of the signaling mechanisms and de- KGAGFITP, lacked elicitor activity but acted as a com- fense responses induced (13). petitive antagonist for CSP-related elicitors. Bacteria Flagellin was the predominant if not only elicitor present in commonly have a small family of CSP-like proteins in- crude bacterial extracts that activated elicitor responses in the cluding both cold-inducible and noninducible members, tomato cells used in our previous experiments. Extracts from and Csp-related elicitor activity was detected in ex- bacteria without flagella or with flagellins that are strongly tracts from all bacteria tested. Thus, the CSP domain divergent in the elicitor-active domain represented by the oli- containing the RNP-1 motif provides a structure char- gopeptide flg22 proved inactive in the tomato cells (8). These acteristic for bacteria in general, and tobacco plants observations with one particular cell line, grown in vitro for have evolved a highly sensitive chemoperception system several years, do not exclude the existence of chemoperception to detect this bacterial PAMP. 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 A key aspect of active defense against invading microbial for the innate immune system of animals. Similarly, redun- pathogens is the ability to discriminate between self and infec- dancy of chemoperception systems for a variety of molecular tious nonself (1). In plants, recognition-dependent disease re- patterns characteristic for fungi has also been observed in sistance has been studied most thoroughly and most success- plants (6). Therefore, we set out to search for additional che- fully in cases that depend on the presence of specific resistance moperception systems of plants sensing molecular patterns genes that confer immunity to particular races of plant patho- characteristic for bacteria. Suspension cultured tobacco cells gens. Several of these resistance genes were shown to be in- have long been known to respond with a rapid K efflux, a volved in the chemoperception of factors specifically attributed concomitant medium alkalinization and an oxidative burst with particular strains of pathogens (2– 4). In addition, plants when treated with bacterial preparations containing either have a broader, more basal, surveillance involving sensitive living or heat-killed bacteria (14) but the bacterial factors elic- iting these responses have not been identified. In initial exper- iments we tested commercial preparations containing pepti- * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked doglycan from Micrococcus lysodeikticus (Staphylococcus “advertisement” in accordance with 18 U.S.C. Section 1734 solely to aureus) for induction of responses in cultured tobacco cells. indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains a supplementary figure. ‡ To whom correspondence should be addressed. Fax: 41-61-697-45- The abbreviations used are: PAMP, pathogen-associated molecular 27; E-mail: [email protected]. pattern; CSP, cold shock protein; CSD, cold shock domain. This paper is available on line at http://www.jbc.org 6201 This is an Open Access article under the CC BY license. 6202 Bacterial Cold Shock Proteins as Elicitors in Tobacco Peptidoglycan has long been known as a PAMP signaling pres- ence 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 total lyophilized M. lysodeikticus bacteria proved to be a far more potent 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 Materials—Peptides were synthesized by F. Fischer (Friedrich Mi- escher-Institute, Basel) or by Bio-Synthesis Inc. (Lewisville, TX). Pep- tides were dissolved in H O (stock solutions of 1 to 10 mM) and diluted in a solution containing 0.1% bovine serum albumin 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, Braunschweig, BRD) and grown in King’s B broth at 26 °C on a rotary shaker. Bacteria were harvested by centrifugation, washed once with H O, and resuspended in H O (10% of original volume). Crude bacterial elicitors were prepared by boiling the bacterial suspensions for 5–10 min and removing of bacterial debris by centrifugation. Lyophilized bacteria of M. lysodeik- ticus (Sigma) and the peptidoglycan fraction from M. lysodeikticus (Fluka, Buchs, Switzerland) were applied as suspensions in H O. The bacterial preparation “messenger” was obtained from EDEN Bioscience (Bothell, WA). Purification of Elicitor from M. lysodeikticus—Elicitor activity was purified from the lyophilized preparation of M. lysodeikticus (Sigma). Ten g of the lyophilisate was suspended in 100 ml of H O and heated for 10 min at 95 °C. After centrifugation (30 min at 10,000  g) the FIG.1. Extracellular alkalinization of tobacco cells in re- supernatant was mixed with 1 volume of acetone and the precipitate sponse to treatment with preparations from M. lysodeikticus. A, formed after overnight incubation at 20 °C was removed by centrifu- alkalinization in response to treatment with 10 g/ml lyophilized gation. The acetone concentration was brought to 80% (v/v) and the M. lysodeikticus cells or to 100 g/ml of the peptidoglycan fraction precipitate formed after4hat 20 °C was collected by centrifugation. derived from M. lysodeikticus (peptidoglycan). B, effect of protease K This precipitate was dissolved in 20 mM Tris-HCl, pH 7.5, and passed treatment on alkalinization-inducing activity of lyophilized M. lysodeik- ticus bacteria and the peptidoglycan preparation. Different doses of over an anion-exchange column with diethylaminoethyl-cellulose (DE- lyophilized M. lysodeikticus bacteria (closed circles), bacteria after pre- cellulose, Whatman) equilibrated with 20 mM Tris-HCl, pH 7.5. Activity treatment with protease K (overnight incubation with 1 mg/ml protease eluting in the flow-through was concentrated by acetone precipitation K, open circles), peptidoglycan (closed triangles), and peptidoglycan (80% acetone) and separated on a Sephacel C8 reversed phase column after pretreatment with protease K (open triangles) were added to (Amersham Biosciences AB) at pH 6.5 (10 mM phosphate buffer, pH 6.5, aliquots of the cell culture and the pH change measured after 15 min as solvent A and 80% acetonitrile, 20% phosphate buffer as solvent B). (initial pH 4.8). The two fractions containing the highest elicitor activity were pooled, pH adjusted to 3.5, and rerun on a Sephacel C8 reversed phase column at pH 3.5 (0.1% trifluoroacetic acid in H O at pH 3.5 as solvent A and RESULTS 80% acetonitrile, 20% H O with 0.1% trifluoroacetic acid as solvent B). Extracellular Alkalinization in Cultured Tobacco Cells Plant Cell Cultures—The tobacco (Nicotiana tabacum L.) cell culture line 275N, originally derived from pith tissue of Havanna 425 plants, Treated with Preparations from M. lysodeikticus—Peptidogly- was maintained and subcultured as described before (15) in a Murash- can, an essential cell wall component of all bacteria, acts as one ige-Skoog based medium. Cells were maintained as suspension cultures of the PAMPs signaling the presence of Gram-positive bacteria and used 4 to 10 days after subculture for experiments. Cell cultures of to the innate immune system in animals (1, 10). In initial tomato (“line Msk8” (16)), potato (17), Lycopersicon peruvianum (18), experiments we tested preparations containing peptidoglycan and A. thaliana (19) were cultured as described elsewhere. for induction of extracellular alkalinization in plant cells cul- Alkalinization Response—To measure alkalinization of the growth medium (the alkalinization response), 3-ml aliquots of the cell suspen- tured in liquid medium. Medium alkalinization, occurring as a sions were placed in open, 20-ml vials on a rotary shaker at 120 to 150 consequence of altered ion fluxes across the plasma membrane, cycles per min. Using small combined glass electrodes (Metrohm, Heri- can serve as a convenient, rapid, sensitive and quantitative sau, Switzerland) extracellular pH values were either recorded contin- bioassay to study elicitor perception by plant cells (16). As a uously with a pen recorder or measured after 15 or 20 min of treatment. source of peptidoglycan we used preparations from M. lysodeik- Oxidative Burst and Ethylene Biosynthesis in Leaf Tissue—Fully ex- ticus (S. aureus) because lyophilized bacteria and a peptidogly- panded leaves of different plant species were cut in 2-mm slices and floated on H O overnight. For measuring the oxidative burst, active oxy- can fraction are commercially available. Also, as deduced from gen species released by the leaf tissue were measured by a luminol-de- the genomes of the three fully sequenced strains of S. aureus pendent assay (20). Slices were transferred to assay tubes (2– 4 slices (M. lysodeikticus) that do not encode proteins resembling corresponding to 20 mg of fresh weight) containing 0.1 ml of H O flagellin, these preparations should be free of elicitor-active supplied with 20 M luminol and 1 g of horseradish peroxidase (Fluka). flagellin that could interfere in the assays. No alkalinization Luminescence was measured in a LKB 1250 luminometer (LKB Wallac, was observed in the tomato cells of the line Msk8 after treat- Turku, Finland) for 20 min after the addition of the test solution. For assaying ethylene production, leaf slices (50 mg of fresh per ment with lyophilized M. lysodeikticus bacteria or the pepti- assay) were transferred to 6-ml glass tubes containing 1 ml of an doglycan fraction derived from these bacteria (data not shown). aqueous solution of the peptide being tested. Vials were closed with Whereas these negative results confirmed the absence of elic- rubber septa and ethylene accumulating in the free air space was itor-active flagellin they did not provide evidence for a chemo- measured by gas chromatography after 2 to 2.5 h of incubation. perception system responding to peptidoglycan in the tomato Reproducibility—The results shown in the figures represent single cells. When tested on tobacco cells, however, both preparations experiments that are representative for several independent repetitions. of M. lysodeikticus caused rapid and strong medium alkalin- Bacterial Cold Shock Proteins as Elicitors in Tobacco 6203 phy at pH 6.5 activity eluted as a single peak (Fig. 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 A . Separa- tion by SDS-PAGE (14% (w/v) acrylamide) showed a band migrating with an apparent molecular weight of 7,000 to 9,000 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 iden- tified the protein as a cold shock protein (CSP). In Fig. 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 bac- terial cold shock proteins present in the data bank. Identification of the “Cold Shock Domain” (CSD) as the Elic- itor-active Epitope—In attempts to localize the elicitor activity FIG.2. Purification of the alkalinization-inducing activity by to a particular domain of the protein the purified CSP was reversed phase chromatography. An extract from lyophilized subjected to peptide cleavage. Digestion with trypsin, Lys-C, or M. lysodeikticus bacteria, pre-purified by ion-exchange chromatography Glu-C (V8 protease) abolished the activity and did not result in as described under “Experimental Procedures,” was fractionated on a C8 reversed phase column at pH 6.5. Fractions with highest activity, smaller fragments with elicitor activity (data not shown). As in eluting between 24 and 28 min, were re-chromatographed on the C8 previous work with bacterial flagellin (8) we speculated that column at pH 3.5. Upper panel shows elution profile (OD ) of the first plant cells might have a perception system for the most char- run at pH 6.5 (10 mM phosphate buffer) and the second run at pH 3.5 acteristic and most conserved domain of the CSPs. Although (0.1% trifluoroacetic acid). Lower panel shows extracellular alkaliniza- tion in tobacco cells (pH ) induced by aliquots of the fractions 15min these small bacterial proteins show a high overall homology eluting in the first (open bars) and second (open bars) runs. they are particularly conserved in a domain close to the N terminus. Based on the consensus sequence of bacterial CSPs, ization (Fig. 1). As shown in the examples in Fig. 1A, extracel- a 22-amino acid peptide spanning this domain was synthesized lular pH started to increase after a lag of 3 to 5 min and (Fig. 3, underlined sequence) and tested for induction of alka- reached a maximum after 10 to 15 min. Depending on the cell linization in tobacco cells. This peptide, termed csp22, proved density and the initial pH of different batches of the cell culture even more active than the intact CSP purified from M. lyso- the amplitude of the alkalinization response (pH ) varied max deikticus and induced medium alkalinization with an EC of from 1.2 to 2 pH units for lyophilized bacteria and from 0.6 to 0.1 nM (Fig. 4). 1.4 pH units for the peptidoglycan fraction, respectively. In To further delineate the epitope that activates responses in aliquots from a given batch of cells, however, pH was max the plant cells, peptides lacking varying numbers of amino acid highly reproducible and consistently showed a bigger response residues from the N- or C-terminal end were synthesized and for the preparation of total bacteria than for the peptidoglycan assayed for activity in dose-response curves as described above fraction. The responses of the cells to both preparations of for csp22 and CSP. The amino acid sequences and the EC M. lysodeikticus were dose-dependent and lower, nonsaturat- values are summarized in (Fig. 5). Omitting 5 amino acid ing doses led to prolonged lag phases, smaller maximal pH residues from the N terminus of csp22 reduced activity only increases, and shortened durations of medium alkalinization. slightly (EC of 1.2 nM) but removal of the Lys residue at The pH change occurring within 15 min (pH ) of treat- 15min position 6 showed a much stronger effect (EC of 220 nM) and ment was a steady function of the dose applied and was used as further trimming by 4 amino acid residues resulted in an a parameter to compare the relative strength of the two prep- inactive peptide. The peptide termed csp15, comprising the 15 arations of M. lysodeikticus (Fig. 1B). Half-maximal stimula- amino acid residues central to csp22, was nearly as active as tion was observed with 30 g/ml of the peptidoglycan (EC ) csp22 (EC of 0.3 nM) and served as a core peptide for testing and 1 g/ml with the lyophilized bacteria, respectively. Treat- structural analogues with replacements of single amino acid ment with protease K strongly affected the activity of the residues with alanine. Csp15-Ala3, csp15-Ala4, csp15-Ala8, bacterial preparation resulting in a 200-fold higher EC value and csp15-Ala12 all exhibited at least 1000-fold reduced activ- (200 g/ml) but led only to a 3-fold increase for the EC value ity compared with csp15. Csp15-Ala10 was inactive even at the of the peptidoglycan fraction (Fig. 1B). These results provided highest concentration of 100 M tested (Fig. 5). In contrast, preliminary evidence for the presence of two distinct elicitor substitution of Phe at position 10 with a Tyr residue resulted in activities in M. lysodeikticus: a nonproteinaceous elicitor in the a peptide with full activity in the tobacco cells (Fig. 5). Among peptidoglycan fraction and a second, potent, proteinaceous elic- the peptides with single substitutions with Ala only csp15-Ala7 itor predominating in the total bacteria preparation. On a per showed no significant decrease in activity. Interestingly, the weight basis the proteinaceous factor was more than 100-fold Glu at this position also shows the least conservation in the more active than the peptidoglycan factor and further work different sequences of bacterial CSPs (Fig. 3). focused on the characterization of this new protein elicitor. The three-dimensional structure has been determined for Purification of an Elicitor-active Protein from M. lysodeikti- the major bacterial cold shock proteins CspB from Bacillus cus and Its Identification as Bacterial Cold Shock Protein—The subtilis (Molecular Modeling Database number 3622; Protein elicitor activity, extracted from the crude preparation of Data Bank number 1CSP) (21) and CspA (CS7.4) from E. coli M. lysodeikticus (S. aureus), was heat-stable (5 min, 95 °C), passed ultrafilters with a molecular weight cut-off of 10,000, (Molecular Modeling Database number 1677; Protein Data Bank number 1MJC) (22). CspB forms a dimer whereas CspA and was inactivated by treatment with trypsin (data not shown), indicating that the elicitor activity was attributable to occurs as monomer. Besides this difference of dimerization the structures of both proteins are very similar, forming compact a peptide or small protein. Activity was purified on a Sephacal C8 reversed phase column (Fig. 2A). In the first chromatogra- -barrel structures built up from five antiparallel -strands 6204 Bacterial Cold Shock Proteins as Elicitors in Tobacco FIG.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 data base. Partial sequence for M. lysodeikticus represents information obtained from the purified protein after tryptic digest and Edman degradation of some of the peptides. Csp22 and csp15 denote peptides synthesized according to the consensus sequence. FIG.4. Elicitor activity of the purified Csp and synthetic pep- tides 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)or15 (csp15) amino acid residues of the conserved region from bacterial CSPs as indicated in the legend to Fig. 3. with connecting turns and loops. Fig. 6 shows models (second- ary structure and a three-dimensional ribbon model) of the FIG.5. Alkalinization-inducing activity of csp-related pep- molecular structure of CspB, highlighting the domain spanned tides. EC values were determined from dose-response curves ob- by the csp15 peptide. Clearly, elicitor activity can be attributed tained for the different peptides. Specific activity relative to the activity of the most active peptide csp22 (hatched bar,EC of 0.1 nM). Loga- to the domain formed by antiparallel strands 1 and 2 and the rithmic scaling was used to indicate residual activity in some of the loop L1. This domain includes a RNA-binding motif known as peptides. No activity could be detected with peptides denoted with RNP-1 (also termed RNP-CS) and exposes a cluster of aromatic asterisks (relative activity 10 ). 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 stitutions leading to strong or complete reduction in affinity of the interaction of the protein with nucleic acids (23). In Table I, CspB for nucleic acids exhibited strongly reduced elicitor ac- these single amino acid replacements and their effects on nu- tivity in tobacco cells (higher EC values). The substitution of cleic acid binding were compared with the corresponding amino Phe by Tyr at the position that corresponds to residue 10 in acid changes in csp15 and their effects on elicitor activity in csp15 did not affect affinity of CspB for nucleic acids and also tobacco. All the substitutions in csp15 that correspond to sub- did not alter elicitor activity. Bacterial Cold Shock Proteins as Elicitors in Tobacco 6205 FIG.6. Structure of bacterial CSPs. A, schematic scheme for secondary struc- ture of CspB from B. subtilis from Schin- delin et al. (21) with domains spanned by csp15 (hatched part) and the RNA-bind- ing motifs RNP-1 (KGFGFITP) and RNP-2 (VFVHF) indicated. B, structure of CspB monomer (Molecular Modeling Da- tabase number 3622; Protein Data Bank number 1CSP (21)) drawn with WebLab ViewerLite (Molecular Simulations Inc., Cambridge, UK) with side groups exposed in the domain spanned by csp15. TABLE I and rice (data not shown). Negative results with particular Comparison of mutations in cspB affecting DNA-binding and single lines of cell cultures do not allow concluding on the absence of amino acid changes in csp15 on elicitor activity a perception system in the corresponding plant species because -Fold reduction of this perception system might be not expressed or might have Substitution Substitution in -Fold increase for a affinity for a in CspB csp15 EC oligonucleotide been lost during the years of growth in vitro. Induced release of active oxygen species, an oxidative burst, CspB (wt) csp15 1 1 and increased biosynthesis of the stress hormone ethylene are K7Q csp15-Ala2 3 1000 W8A csp15-Ala3 5 1000 responses characteristic for plants under attack by pathogens F15A csp15-Ala10 Inactive 10,000 or treated by elicitor preparations (6, 25). We used these re- F17A csp15-Ala12 Inactive 2000 sponses to monitor responsiveness toward CSP-derived elici- F15Y csp15-Tyr10 1 1 tors in leaf tissues from different plant species. As exemplified F17Y 6 Not tested in Fig. 7 for leaf tissue from tomato, a rapid, significant in- Csp15 comprises amino acids 5 to 20 in CspB. EC values for the crease in ethylene biosynthesis and in active oxygen species induction of the alkalinization response from Fig. 5. Effect of site-directed point mutations in CspB on the binding of was observed after treatment with csp15 but not after treat- single stranded DNA oligonucleotides containing Y-box motif. Data ment with the same dose of csp15-Nsyl. Similarly, clear CSP- from Table I in Schro ¨ der et al. (23). dependent induction of ethylene biosynthesis and oxidative burst was observed in tobacco and several other solanaceous plants including potato (Solanum tuberosum), Solanum dulca- Activity of Peptides Representing Homologous Domains Oc- mara, Scopolia carniolica, and Mandragora officinarum.In curring of Proteins from Plants and Animals—Csp-related pro- contrast, no response could be detected from leaf tissue and cell teins are common to all eubacteria and they usually form a cultures of A. thaliana, cucumber, and rice. Also, no signs of a small family of proteins that include both cold-inducible and hypersensitivity response could be detected after injection of noninducible members (24). The domain containing the RNP-1 CSP peptides to leaves of tobacco or tomato (data not shown). motif is conserved also in many eukaryotic proteins that bind to In summary, a perception system for CSP-related elicitors is RNA or DNA. Examples for proteins with this so-called CSD common to solanaceous plants but has not yet been found include human and animal transcription factors recognizing outside of this plant family. the Y-box sequence and glycine-rich RNA-binding proteins oc- The Inactive Peptide csp15-Ala10 Antagonizes Elicitor Activ- curring in plants (see Supplementary Materials for gene struc- ity of CSP—Peptides lacking either 4 amino acids from the ture and alignment with bacterial CSPs). Peptides correspond- C-terminal part (csp11) or 6 amino acids from the N-terminal ing to the homologues of a human Y-box protein and a Gly-rich part spanned by csp15 lacked activity even when applied in protein from Nicotiana sylvestris were synthesized and tested micromolar concentrations (data not shown). No response was for activity. The csp15 homologue from the human Y-box pro- observed also by application of these two peptides in combina- tein was inactive whereas the peptide representing N. sylves- tion (data not shown). Truncated forms of the biologically ac- tris sequence induced responses with an EC of 300 nM and tive peptides systemin and flg22 were previously found that was thus 1000-fold less active than the csp15 representing the bacterial sequence (Fig. 5). showed characteristics of competitive antagonists for the re- spective nontruncated agonistic peptides (26 –28). No antago- Responses Induced by CSP in Different Plant Species—We examined cell cultures derived from other plant species for nistic activity could be observed for the two truncated CSP peptides described above (data not shown). In contrast, csp15- alkalinization in response to CSP-related elicitors. Responses with characteristics similar to the ones of the tobacco line 275N Ala10, also inactive as agonist (Fig. 5), did exhibit antagonistic activity and suppressed responses induced by csp15 (Fig. 8). were observed also with a second line of tobacco, originating from a plant of the variety SR1, with a cell line derived from When added concomitantly with 3 nM csp15, a concentration of 3 M strongly inhibited induction of the alkalinization response potato and a cell culture from L. peruvianum (data not shown). In contrast, no responses could be detected in the cell culture (Fig. 8A, “0 min”). Complete inhibition was observed when line msk8, originally derived from a cross of Lycopersicon escu- csp15-Ala10 was added 30 s before the agonist but progres- lentum with L. peruvianum, and in cell lines from A. thaliana sively weaker effects were observed when the antagonist was 6206 Bacterial Cold Shock Proteins as Elicitors in Tobacco FIG.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 M csp15 or 1 M csp15-Nsyl as indicated. Bars and error bars show mean  S.D. of n 4 replicates. B, luminescence of leaf slices in a solution with luminol and peroxidase after treatment with 1 M concentrations of csp15 or csp15-Nsyl as indicated. Light emission at the very beginning of the experiments is caused by phos- phorescence of the green tissue. 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 FIG.8. The peptide csp15-Ala10 acts as competitive suppressor observed with unrelated elicitors like flagellin and chitin frag- of csp-related elicitors. A, extracellular alkalinization in suspension- ments (data not shown). Inhibition of CSP-related activity by cultured tobacco cells treated with 3 nM csp15 and 3 M csp15-Ala10 at csp15-Ala10 was competitive and could be overcome by increas- the time points indicated (slanted arrows). Extracellular pH of un- treated cells was 4.9 and addition of 3 M csp15-Ala10 alone did not ing concentrations of active peptide or intact CSP. As shown in cause significant pH changes. B, alkalinization induced by different the example in Fig. 8B, this resulted in an increase of the EC doses of M. lysodeikticus (circles) and peptidoglucan preparation (trian- for the CSP containing preparation of M. lysodeikticus bacteria gles) in cells without pretreatment (closed symbols) or cells pretreated from 1 g/ml in the absence of the antagonist, to 20 g/ml in for 3 min with 10 M csp15-Ala10 (open symbols). C and D, alkaliniza- tion induced by the harpin-containing preparation messenger (1 g/ml, the presence of 3 M csp15-Ala10, respectively. In contrast, no C) and by a crude extract from A. tumefaciens (1 l/ml, D) in cells shift in dose response was observed with the peptidoglycan without pretreatment or cells pretreated for 3 min with 3 M csp15- fraction (Fig. 8B). These results confirm predominance of the Ala10 as indicated. CSP-related elicitor in the crude bacterial preparation and the presence of an activity unrelated to CSP in the peptidoglycan rence of CSP-related elicitor activity in extracts from different, preparation. if not all, bacteria. Tobacco cells were found to respond to crude extracts from all DISCUSSION bacterial species tested (n  20). The antagonist csp15-Ala10 could serve as a diagnostic tool to test for the presence of Various types of living bacteria as well as preparations of csp-related activity. For example, cells responded with strong heat-killed bacteria can trigger rapid responses in plant cell alkalinization when treated with messenger, an extract from cultures and defense responses in intact plant tissues (20, 30, E. coli expressing transgenic harpin from Erwinia amylovora 31). Flagellin (8) and lipopolysaccharides (32) have been iden- (Fig. 8C). Interestingly, at least at limiting doses of messenger tified as common bacterial determinants or PAMPs that act as applied, activity was fully antagonized by csp15-Ala10. This elicitors of defense responses in plant cells. In this report we indicated that a CSP-related stimulus and not harpinEa, pre- add CSPs as a further bacterial PAMP that acts as an elicitor viously reported to act as an inducer of alkalinization in tobacco of defense responses in plants. (29), was the activity predominating in this preparation. As Cold shock proteins were named based on the original obser- shown for the example of an extract from A. tumefaciens in Fig. vation that rapid cooling with a T of 10 °C (cold shock) 8D, csp15-Ala10 antagonized also the alkalinization inducing induces accumulation of specific proteins in many bacterial activity of crude extracts from the plant-associated species species. The major CSPs are small, 7.4 kDa proteins that A. tumefaciens, R. meliloti, and X. campestris, extracts that belong to a family of highly conserved proteins commonly oc- were previously found to be devoid of elicitor-active flagellin curring in all bacteria. At least some members of this family (8). In summary, these results demonstrate the common occur- are (also) constitutively expressed or are induced under stress Bacterial Cold Shock Proteins as Elicitors in Tobacco 6207 conditions different from cold shock (24). For example, the mediated process for both elicitors. Thus, we hypothesize that family of CspA-like proteins in E. coli consists of eight members csp perception occurs via a csp receptor that functions in a manner similar to the receptor for flagellin. Activation of this (CspA through CspH) and only CspA, CspB, and CspG are putative CSP receptor might also involve two consecutive steps cold-inducible. Thus, despite their name, members of the CSP family occur also in bacteria not subjected to a cold shock 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 treatment. CSPs are implicated in various cellular processes, of csp15, Phe in csp15, or Tyr in csp15-Tyr10, appears neces- including cellular growth and adaptation to low temperatures, sary for this second step to activate the receptor. The antago- nutrient stress, and stationary phase. CSPs bind to nucleic nist csp15-Ala10, apparently, does not undergo the second, acids and appear to function as RNA chaperones and anti- locking step and interacts with the receptor site in a more terminators of translation (33). readily reversible manner. This could explain the high excess of A shift to low temperature induces also a set of specific pro- antagonist csp15-Ala10 over csp15 required to block elicitor teins in plants (34). Some of these cold-regulated proteins are action completely and the apparent inefficiency of the antago- small hydrophilic proteins of 6.6 kDa (35), like the major bacte- nist when applied subsequent to the csp agonists (Fig. 8). rial CSP but they are nonhomologous in sequence, and their Proteins with a cold shock domain comprising the RNP-1 physiological function in the cold acclimation process remains motif are conserved also in eukaryotes and have been identified unknown. However, many eukaryotes including plants and ani- also in genes of A. thaliana and N. sylvestris. Although clearly mals have proteins with a nucleic acid-binding domain that homologous, the sequences corresponding to the elicitor-active shows a strikingly high homology and similar RNA-binding prop- epitope show some differences in comparison to the bacterial erties to bacterial CSPs (36). It is this universally conserved consensus. The synthetic peptide csp15-Nsyl, representing the domain, also termed CSD, that contains the RNP-1 motif and the least divergent form of this domain in genes known from epitope found to act as elicitor of tobacco cells. N. sylvestris, indeed did show some activity in the bioassay The elicitor activity of bacterial CSPs could be localized to a with tobacco cell. However, the specific activity of this peptide stretch of 15 amino acid residues that forms a loop with two was 1000-fold lower than that of csp15 representing the bac- antiparallel -strands and exposes a series of aromatic and terial epitope. A lower specific activity could be counterbal- basic amino side chains to the surface of the protein. It is this anced by the presence of high local concentrations of the stim- epitope that exhibits highest conservation between the differ- ulus. In initial attempts with extracts of tobacco plants or cells ent bacterial CSPs and, as has been demonstrated by site- from tissue culture we failed to detect factors with CSP-like directed mutagenesis of CspB from B. subtilis, is essential for activity in bioassays (data not shown). Thus, at present, we do the interaction of the protein with nucleic acids (23). Most not have evidence for endogenous factors stimulating tobacco notably, synthetic peptides with amino acid sequences reflect- via the CSP perception system described in this report. Endog- ing the changes leading to reduced or abolished binding to enous factors of tobacco, capable of stimulating medium alka- nucleic acids of CspB were also strongly affected in elicitor linization in cultured cells, have recently been described (38) activity (Table I). This strong correlation raises the question of but these peptidic factors show no apparent homology to CSPs. whether some sort of nucleic acid might be involved in the Bacterial CSPs are molecules highly characteristic for bac- perception process by the plant cell. However, intact CSPs and teria in general and could thus serve as PAMPs signaling the the csp-derived peptide elicitors have characteristics of mole- presence of bacteria to the plant cells. An obvious problem with cules that are not permeable for membranes and the first this hypothesis that is the localization of these proteins, which responses to subnanomolar concentrations of csp-derived elic- are generally assumed to function in the cytoplasm of the itors occur after a lag phase of less than 2 min. These charac- bacteria. So far, CSPs have not been reported to be exported or teristics rather suggest a chemoperception system with a spe- exposed to the surface by intact bacteria and, consequently, cific, high affinity primary interaction site in the apoplast, most CSPs are probably not directly detectable by a chemopercep- likely the plasma membrane, of the plant cells. The strong tion system assumed to reside on the surface of the plant cells. correlation of nucleic acid binding and elicitor activity might Further studies will be required to test whether CSPs are thus reflect evolution of a chemoperception system directed at released from bacteria during invasion of their plant hosts. A a particular surface epitope of the bacterial CSPs that is under release could be based on a bacterial export system activated in a high selective pressure for retaining functionality of the the course of the infection process, or it could result from protein. bacterial or plant processes causing a general leakiness of the Perception of csp-related elicitors resembles the chemoper- bacteria. As demonstrated for bacteria under mild osmotic ception system for flagellin-derived elicitors studied before (8, shock (39), leakiness of bacteria leading to release of small 28). In both cases, elicitor activity could be attributed to an cytoplasmic proteins might be more common than suggested by epitope of 15 amino acids representing the most conserved studies under optimal media conditions used to grow bacterial part of the respective protein. Both elicitors are active at sub- cells in the laboratory. Precedence for cytoplasmic components nanomolar doses and activity is highly dependent on the gen- of bacteria that act as PAMPs and stimulate the innate im- uine amino acid sequence of the conserved domain. “Mutation- mune responses via Toll-like receptors in animals include “non- al” analysis using structural analogues of the elicitors allowed secreted” components such as the heat shock protein HSP60 identification of peptides lacking elicitor activity but exhibiting (40) and bacterial DNA (12). Bacterial DNA is recognized via properties of competitive antagonists. Perception of flagellin its content of nonmethylated CpG oligonucleotides and this was shown to involve a specific, high affinity binding site and PAMP was successfully applied as a potent immunostimula- the membrane-bound receptor kinase FLS2 (9, 37). A model tory factor (41, 42) but the process leading to release of the involving a two-step process for receptor activation was pro- DNA from the bacteria has not been elucidated. Similarly, no posed to explain the effects of agonistic and antagonistic pep- process that secretes HSP60 from intact bacterial cells has tides (28). At present, experiments that directly demonstrate a been described. HSP60 is well conserved from microbes to receptor site for the csp elicitors are lacking. Nevertheless, humans and HSP60 from both mammalian and microbial CSP- and flagellin-derived elicitors induce the same set of sources can trigger inflammatory responses via Toll-like recep- responses with similar kinetics, indicating a similar, receptor- tor 4 (40), suggesting that Toll-like receptor 4 may detect both 6208 Bacterial Cold Shock Proteins as Elicitors in Tobacco protein sequencing services and mass spectrometry, and Martin endogenous and exogenous ligands as alarm signals. Exposure Regenass for maintaining the cell cultures and technical assistance. of endogenous HSP60 to the Toll-like 4 receptor could be en- visaged to occur via release from wounded or injured cells and REFERENCES perception by the receptor on different, intact cells. As dis- 1. Medzhitov, R., and Janeway, C. A., Jr. (2002) Science 296, 298 –300 2. Yang, Y., Shah, J., and Klessig, D. F. (1997) Genes Dev. 11, 1621–1639 cussed above, it remains to be seen whether the chemopercep- 3. Dangl, J. L., and Jones, J. D. (2001) Nature 411, 826 – 833 tion system for CSPs described in this report might similarly 4. Schneider, D. S. (2002) Cell 109, 537–540 5. Heath, M. C. (2000) Curr. Opin. Plant Biol. 3, 315–319 react to both endogenous and exogenous ligands. 6. Boller, T. (1995) Annu. Rev. Plant Physiol. Plant Mol. Biol. 46, 189 –214 Peptidoglycan consists of a glycan backbone with alternating 7. Hahn, M. G. (1996) Annu. Rev. Phytopathol. 34, 387– 412 1– 4-linked residues of N-acetyl-D-glucosamine and muramic 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 acid and forms the major component of the cell wall in Gram- 10. Aderem, A., and Ulevitch, R. J. (2000) Nature 406, 782–787 positive bacteria. Peptidoglycans, sensed by a family of pepti- 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., doglycan recognition proteins that are conserved from insects Hatakeyama, T., Aoyagi, H., Kurazono, H., Moss, J., and Hirayama, T. to humans (43), are important PAMPs for the innate immunity (2001) J. Biol. Chem. 276, 30521–30526 of animals. The peptidoglycan preparation of M. lysodeikticus 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, also triggered elicitor responses in tobacco cells. We have not 1099 –1103 yet characterized this activity in detail and cannot exclude that 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 it is because of a minor component or a “contaminant” of the 15. Felix, G., and Meins, F., Jr. (1987) Planta (Heidelberg) 172, 386 –392 peptidoglycan fraction. Nevertheless, this nonproteinaceous 16. Felix, G., Regenass, M., and Boller, T. (1993) Plant J. 4, 307–316 17. Schweizer, P., Felix, G., Buchala, A., Muller, C., and Me ´ traux, J. P. (1996) factor is perceived as a quality of stimulus distinct from CSPs Plant J. 10, 331–341 and provides evidence for a further bacterial PAMP with elic- 18. Felix, G., and Boller, T. (1995) Plant J. 7, 381–389 itor activity in tobacco cells. Thus, similar to the chemopercep- 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, tion systems for a variety of fungal-derived PAMPs, perception 974 –978 of bacteria by plant cells appears not to depend on a single 21. Schindelin, H., Marahiel, M. A., and Heinemann, U. (1993) Nature 364, 164 –168 bacterial factor but rather involves several different factors, 22. Schindelin, H., Jiang, W., Inouye, M., and Heinemann, U. (1994) Proc. Natl. including at least flagellin (8), lipopolysaccharides (32), pepti- Acad. Sci. U. S. A. 91, 5119 –5123 23. Schro ¨ der, K., Graumann, P., Schnuchel, A., Holak, T. A., and Marahiel, M. A. doglycan, and CSP. This redundancy, characteristic also for the (1995) Mol. Microbiol. 16, 699 –708 recognition mechanisms in the innate immune system of ani- 24. Thieringer, H. A., Jones, P. G., and Inouye, M. (1998) Bioessays 20, 49 –57 mals, points at possible difficulties with approaches to demon- 25. Lamb, C., and Dixon, R. A. (1997) Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 251–275 strate a direct physiological role for any particular of these 26. Pearce, G., Johnson, S., and Ryan, C. A. (1993) J. Biol. Chem. 268, 212–216 chemoperception systems for plant defense. Inhibition or 27. Meindl, T., Boller, T., and Felix, G. (1998) Plant Cell 10, 1561–1570 28. Meindl, T., Boller, T., and Felix, G. (2000) Plant Cell 12, 1783–1794 knockout of only one of the systems might be without a strong 29. Wei, Z. M., Laby, R. J., Zumoff, C. H., Bauer, D. W., He, S. Y., Collmer, A., and effect on the overall recognition system. In contrast to bacterial Beer, S. V. (1992) Science 257, 85– 88 “avirulence factors,” which act as elicitors that are specific and 30. Atkinson, M., Bina, J., and Sequeira, L. (1993) Mol. Plant-Microbe Interact. 6, 253–260 unique for a particular pathogen, the structures recognized as 31. Jakobek, J. L., and Lindgren, P. B. (1993) Plant Cell 5, 49 –56 PAMPs are essential or “vital factors” for the functioning of the 32. Meyer, A., Puhler, A., and Niehaus, K. (2001) Planta (Heidelberg) 213, 214 –222 bacterial organisms in general and cannot easily be changed, 33. Bae, W., Xia, B., Inouye, M., and Severinov, K. (2000) Proc. Natl. Acad. Sci. removed, or mutated for probing their role in plant defense. U. S. A. 97, 7784 –7789 In summary, our results provide evidence for novel bacterial 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, elicitors, cold shock protein, for which tobacco and other So- 293–299 lanaceae have evolved specific and sensitive chemoperception 36. Graumann, P. L., and Marahiel, M. A. (1998) Trends Biochem. Sci. 23, 286 –290 systems. The accuracy and sensitivity of the perception system 37. Bauer, Z., Gomez-Gomez, L., Boller, T., and Felix, G. (2001) J. Biol. Chem. 276, for the CSP domain comprising the RNP-1 motif detailed in 45669 – 45676 38. Pearce, G., Moura, D. S., Stratmann, J., and Ryan, C. A., Jr. (2001) Proc. Natl. this report indicate a receptor mechanism involving a high Acad. Sci. U. S. A. 98, 12843–12847 affinity binding site on the surface of the plant and should 39. Berrier, C., Garrigues, A., Richarme, G., and Ghazi, A. (2000) J. Bacteriol. 182, provide the basis for further work to identify the protein acting 248 –251 40. Seo, S., Okamoto, M., Seto, H., Ishizuka, K., Sano, H., and Ohashi, Y. (1995) as pattern recognition receptor for CSP. Science 270, 1988 –1992 41. Krieg, A. M. (2002) Trends Immunol. 23, 64 – 65 Acknowledgments—We thank Franz Fischer (Friedrich Miescher- 42. Krieg, A. M. (2002) Annu. Rev. Immunol. 20, 709 –760 Institute, Basel) for the synthesis of various peptides, Renate Matthies, 43. Kang, D., Liu, G., Lundstrom, A., Gelius, E., and Steiner, H. (1998) Proc. Natl. Daniel Hess, and Jan Hofsteenge (Friedrich Miescher-Institute) for Acad. Sci. U. S. A. 95, 10078 –10082

Journal

Journal of Biological Chemistry – Unpaywall

Published: Feb 1, 2003

There are no references for this article.