Portal de Periódicos da CAPES

Systemic Signaling in Tomato Plants for Defense against Herbivores

2003; Elsevier BV; Volume: 278; Issue: 32 Linguagem: Inglês

10.1074/jbc.m304159200

1083-351X

Gregory Pearce, Clarence A. Ryan,

Plant Stress Responses and Tolerance

An 18-amino acid peptide in tomato leaves called systemin is a primary signal released at wound sites in response to herbivory that systemically signals the activation of defense genes throughout the plants. We report here the isolation of three hydroxyproline-rich glycopeptides from tomato leaves, of 20, 18, and 15 amino acids in length, that signal the activation of defense genes, similar to the activity of the systemin peptide. The three new peptides cause an alkalinization of suspension-cultured cells and induce the synthesis of defensive proteinase inhibitor proteins when supplied at fmol levels to young tomato plants through their cut stems. This suggests that they are part of the wound signaling of tomato plants that activates defense against herbivores and pathogens. Isolation of cDNAs coding for the tomato peptides revealed that they are all derived from the same pre-proprotein precursor that is systemically wound-inducible. The peptides are considered members of the functionally characterized systemin family of defense signals from plants that are synthesized both in wounded leaves and in distal, unwounded leaves in response to herbivory or other mechanical wounding. The precursor deduced from the cDNA exhibits a leader sequence, indicating that it is synthesized through the secretory pathway, where it is hydroxylated and glycosylated. The amino acid sequence of the precursor exhibited weak identity to the precursor of two hydroxyproline-rich defense signals recently found in tobacco, suggesting that the two pre-protein precursors have evolved from a common ancestral protein. The identification of hydroxyproline-rich glycoprotein systemins in tomato indicates that the initiation of wound signaling is more complex than previously thought and appears to involve multiple peptide signals. An 18-amino acid peptide in tomato leaves called systemin is a primary signal released at wound sites in response to herbivory that systemically signals the activation of defense genes throughout the plants. We report here the isolation of three hydroxyproline-rich glycopeptides from tomato leaves, of 20, 18, and 15 amino acids in length, that signal the activation of defense genes, similar to the activity of the systemin peptide. The three new peptides cause an alkalinization of suspension-cultured cells and induce the synthesis of defensive proteinase inhibitor proteins when supplied at fmol levels to young tomato plants through their cut stems. This suggests that they are part of the wound signaling of tomato plants that activates defense against herbivores and pathogens. Isolation of cDNAs coding for the tomato peptides revealed that they are all derived from the same pre-proprotein precursor that is systemically wound-inducible. The peptides are considered members of the functionally characterized systemin family of defense signals from plants that are synthesized both in wounded leaves and in distal, unwounded leaves in response to herbivory or other mechanical wounding. The precursor deduced from the cDNA exhibits a leader sequence, indicating that it is synthesized through the secretory pathway, where it is hydroxylated and glycosylated. The amino acid sequence of the precursor exhibited weak identity to the precursor of two hydroxyproline-rich defense signals recently found in tobacco, suggesting that the two pre-protein precursors have evolved from a common ancestral protein. The identification of hydroxyproline-rich glycoprotein systemins in tomato indicates that the initiation of wound signaling is more complex than previously thought and appears to involve multiple peptide signals. In 1972, a herbivore-induced systemic defense response was first documented biochemically by the demonstration that insects chewing on tomato leaves caused the release of a systemic signal that induced the synthesis of proteinase inhibitor proteins throughout the plants (1Green T.R. Ryan C.A. Science. 1972; 175: 776-777Crossref PubMed Scopus (859) Google Scholar). Proteinase inhibitors are among the front line defenses of plants that can upset protein digestion of herbivores that consume them (2Ryan C.A. Annu. Rev. Phytopathol. 1990; 28: 425-449Crossref Google Scholar). These observations led to a search for the "proteinase inhibitor-inducing factor" in leaves of tomato plants, resulting in the discovery that an 18-amino acid peptide called systemin is a key signal for the systemic activation of defense genes in response to pest attacks (3Pearce G. Strydom D. Johnson S. Ryan C.A. Science. 1991; 253: 895-898Crossref PubMed Scopus (757) Google Scholar).Tomato systemin is a proline-rich peptide that is derived from the C-terminal region of a 200-amino acid proprotein precursor called prosystemin (4McGurl B. Pearce G. Orozco-Cardenas M. Ryan C.A. Science. 1992; 255: 1570-1573Crossref PubMed Scopus (325) Google Scholar) in response to herbivore attacks or other severe wounding. Expression of the prosystemin gene in its antisense orientation in tomato plants effectively blocks the systemic synthesis of defense genes, demonstrating a key role for the peptide in distal signaling (4McGurl B. Pearce G. Orozco-Cardenas M. Ryan C.A. Science. 1992; 255: 1570-1573Crossref PubMed Scopus (325) Google Scholar). Systemin and its precursor have been identified in several Solanaceae species, including black nightshade, potato, and bell pepper from the tribe Solaneae (5Hunziker A.T. The Genera of Solanaceae. A. R. G. Gantner Verlag K. G., Berlin2001Google Scholar). Tobacco, another member of the Solanaceae family but from the subtribe Nicotianae, does not express a gene homologous to tomato systemin and does not have a strong leaf-to-leaf systemic signaling system (6Constabel C.P. Yip L. Ryan C.A. Plant Mol. Biol. 1998; 26: 55-62Crossref Scopus (105) Google Scholar). It does, however, have a strong defense response in leaves that are wounded (7Pearce G. Johnson S. Ryan C.A. Plant Physiol. 1993; 102: 639-644Crossref PubMed Scopus (98) Google Scholar) and a strong leaf-to-root signaling system (8Zhang Z.-P. Baldwin I.T. Planta. 1997; 203: 436-441Crossref Scopus (161) Google Scholar). A search for the localized defense signals in tobacco leaves resulted in the discovery of two 18-amino acid hydroxyproline-rich glycopeptides that are potent inducers of proteinase inhibitors in leaves of tobacco plants (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). The peptides were recently named tobacco hydroxyproline-rich systemins (TobHypSys I and II) 1The abbreviations used are: TobHypSys, tobacco hydroxyproline-rich systemin(s); TomHypSys, tomato hydroxyproline-rich systemin(s); TFA, trifluoroacetic acid; HPLC, high pressure liquid chromatography; MALDI-MS, matrix-assisted laser desorption ionization-mass spectrometry; RALF, rapid alkalinization factor.1The abbreviations used are: TobHypSys, tobacco hydroxyproline-rich systemin(s); TomHypSys, tomato hydroxyproline-rich systemin(s); TFA, trifluoroacetic acid; HPLC, high pressure liquid chromatography; MALDI-MS, matrix-assisted laser desorption ionization-mass spectrometry; RALF, rapid alkalinization factor. (10Pearce, G., and Ryan, C. A. (2003) Proc. Natl. Acad. Sci. U. S. A., in pressGoogle Scholar) as their discovery functionally broadened the definition of systemins to include any signaling peptide from plants that activates plant defensive genes. The two tobacco peptides are processed from a single pre-protein precursor, analogous to the origins of some peptide hormones in animals.Evidence for localized defensive wound signals in tomato plants that are independent of systemin was obtained when analyzing tomato plants expressing an antisense tomato prosystemin gene (4McGurl B. Pearce G. Orozco-Cardenas M. Ryan C.A. Science. 1992; 255: 1570-1573Crossref PubMed Scopus (325) Google Scholar). The plants lack systemic signaling but do exhibit a strong localized defense response. We report herein that a search for additional peptide signals in tomato leaves has resulted in the isolation of three hydroxyproline-rich, glycosylated peptides, of 20, 18, and 15 amino acids in length, that are powerful inducers of defense responses in tomato cell cultures and leaves and appear to have roles in the amplification of mobile wound signals. The isolation of cDNAs coding for the three peptides revealed that all three peptides are derived from a single wound-inducible pre-proprecursor protein of 146 amino acids in length. This precursor shares weak amino acid identities with a pre-proprecursor of tobacco defense peptides in its N- and C-terminal regions, suggesting that the tobacco and tomato precursors may have been derived from a common ancestral gene. This is the first report in plants of three peptide signals being derived from a single precursor, a scenario common to animals and yeast, and indicates that defensive wound signaling in tomato plants may involve multiple peptide signals.MATERIALS AND METHODSAlkalinization Assay—Tomato suspension cells were maintained in Murashige and Skoog media as described previously (11Scheer J. Ryan C.A. Plant Cell. 1999; 11: 1525-1535Crossref PubMed Scopus (116) Google Scholar), but excluding buffer, with the media adjusted to pH 5.6 with KOH. Cultures were maintained by transferring 3 ml of cells to 45 ml of media every 7 days with shaking at 160 rpm. Tomato cells were used for alkalinization assays 4–7 days after transfer. One h before assaying for alkalinating activity, 1-ml aliquots of cells were transferred into each well of 24-well cell-culture cluster plates and allowed to equilibrate at 160 rpm. Fractions from various separation procedures (1–10 μl) were added to the cells, and the increase in pH of the media was recorded after 20 min.Proteinase Inhibitor-inducing Assay—Tomato plants were grown in growth chambers under 17-h light (28 °C) and 7-h dark (17 °C) conditions. Light intensity was 200 microeinsteins m–2s–1. Purified peptides were assayed by immunoradial diffusion (12Ryan C.A. Anal. Biochem. 1967; 19: 434-440Crossref PubMed Scopus (112) Google Scholar, 13Trautman R. Cowan K.M. Wagner G.G. Immunochemistry. 1971; 8: 901-916Crossref PubMed Scopus (54) Google Scholar) for their ability to induce proteinase inhibitor II synthesis in excised tomato plants as described previously (3Pearce G. Strydom D. Johnson S. Ryan C.A. Science. 1991; 253: 895-898Crossref PubMed Scopus (757) Google Scholar) using rabbit antiserum prepared by injections of pure tomato inhibitor II protein (14Plunkett G. Senear D.F. Zuroske G. Ryan C.A. Arch. Biochem. Biophys. 1982; 213: 463-472Crossref PubMed Scopus (103) Google Scholar). Solutions of peptides were supplied to excised tomato plants through their cut stems for 45 min, and the plants were transferred to 20-ml vials containing water and incubated in closed Plexiglas boxes for 24 h at 200 microeinsteins m–2s–1. The juice was expressed from the leaves using a mortar and pestle and then assayed for proteinase inhibitor II content.Polypeptide Isolation—Ten tomato plants (Lycopersicon esculentum var. Castlemart) were grown under greenhouse conditions for approximately 8 weeks. Before the onset of fruit, the plants were sprayed with methyl jasmonate as described previously (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). Fifteen h later, the leaves were collected, ground in liquid nitrogen, and stored at –20 °C. Frozen leaf material (2 kg) was homogenized in a 4-liter blender for 2 min using 2.4 liters of 1% trifluoroacetic acid (TFA) as the extraction solvent. The liquid was squeezed through four layers of cheesecloth and one layer of Miracloth (Calbiochem) and centrifuged at 10,000 × g for 20 min. The acidic supernatant was adjusted to pH 4.5 with 10 n NaOH and centrifuged at 10,000 × g for 20 min. After re-adjusting the supernatant to a pH of 2.5 with TFA, the liquid was applied to a 40-μm 3 × 25 mm C18 reversed-phase flash column (Bondesil, Varian Analytical Instruments, Walnut Creek, CA) equilibrated with 0.1% TFA/H2O. Elution was performed at 8 p.s.i. with compressed nitrogen gas. The column was eluted with 100 ml of 0.1% TFA/H2O and then 250 ml of 40% methanol/0.1% TFA. The fraction eluting with 40% methanol was vacuum-evaporated to remove methanol and then lyophilized to dryness, yielding 11.5 g from five duplicated preparations (50 plants total). One-third of the dry powder was dissolved in 10 ml of 0.1% TFA/H2O, centrifuged at 10,000 × g for 10 min, and applied to a Sephadex G-25 column (4 × 40 cm) equilibrated with 0.1% TFA/H2O. Fractions (8 ml) were collected, and the alkalinating activity was assayed as described above using 10 μl/fraction. The activity was found in fractions at or near the void, which were pooled and lyophilized. The yield from three duplicate runs was about 400 mg of dry weight. This preparation was called the "lyophilized crude extract." This material was dissolved in 10 ml of 0.1% TFA/H2O (40 mg/ml) for semipreparative reversed-phase C18-HPLC. After centrifugation and filtration, the components from 10 sequential 40-mg samples were separated on the column (218TP510, 5-μm 10 × 250 mm column, Vydac, Hesperia, CA) with a flow rate of 2 ml/min. After 2 min, a gradient was applied from 0–40% acetonitrile/0.1% TFA over 90 min. The absorbance was monitored at 220 nm. Fractions eluting at 1-min intervals were collected, and 10-μl aliquots were used with 1 ml of tomato suspension-cultured cells to determine the alkalinating activity in each fraction. After further purification, six peaks were identified, and partial amino acid sequence analyses of the individual components were determined. Two peaks corresponded to the systemin sequence (3Pearce G. Strydom D. Johnson S. Ryan C.A. Science. 1991; 253: 895-898Crossref PubMed Scopus (757) Google Scholar), whereas a later eluting, large peak was the previously identified RALF peptide (15Pearce G. Moura D.S. Stratmann J. Ryan C.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12843-12847Crossref PubMed Scopus (282) Google Scholar). This peptide is not involved with defense but with cell division and elongation. The remaining three peaks were composed of multiple peptide components and were further purified. Fractions 29–33 (peaks 1 and 2) and fractions 49–52 (peak 3) were pooled and lyophilized. The combined yields for peaks 1 and 2, called Fraction I, was 33 mg, and the yield for peak 3, called Fraction II, was 22 mg. Each fraction was subjected to strong cation exchange chromatography on a PolySULFOETHYL Aspartamide™ column (5-μm 4.6 × 200 mm column, The Nest Group, Southborough, MA) equilibrated with 5 mm potassium phosphate, pH 3, in 25% acetonitrile. After dissolving Fraction I in 3 ml of buffer and Fraction II in 2 ml of buffer, the solutions were analyzed separately by sequentially loading 1-ml samples onto the column (three separations for Fraction I and two for Fraction II). Two min after applying the sample to the column, a 90-min gradient was applied to 40% elution buffer consisting of 5 mm potassium phosphate, 500 mm potassium chloride, pH 3, in 25% acetonitrile. Absorbance was monitored at 220 nm at a flow rate of 1 ml/min, and fractions were collected at 1-min intervals. A 2-μl aliquot from each fraction was assayed with the suspension-cultured cells. Two peaks of alkalinization activity were identified among the components eluting from the separation profile of Fraction I. An initial peak, called Peak 1, eluted at 56–61 min, and a smaller peak, called Peak 2, eluted at 79–82 min. Fraction II contained one major activity peak that eluted at 59–63 min and was called Peak 3. Fractions from each peak were pooled, lyophilized, and further purified after dissolving each in 1-ml column equilibration buffer, 10 mm potassium phosphate, pH 6, and clarifying by centrifugation. The centrifugates were individually applied to a reversed-phase C18 HPLC column (218TP54, 5-μm 4.6 × 250 mm column, Vydac) with a flow rate of 1 ml/min. After 2 min, a 90-min gradient was applied to 40% elution buffer consisting of 10 mm potassium phosphate, pH 6, in 50% acetonitrile for Peaks 1 and 2 and was applied to 60% elution buffer for Peak 3. Absorbance was monitored at 220 nm. Two-μl aliquots were used to determine alkalinating activity; Peak 1 eluted at 45–52 min, Peak 2 at 44–47 min, and Peak 3 at 46–50 min. These were pooled and lyophilized. Further purification was carried out on the analytical C18 column used above but using TFA/methanol as the elution solvent. The lyophilized peaks were dissolved in 1 ml of 0.1% TFA/H2O and centrifuged, and the supernatants were loaded onto the column at a flow rate of 1 ml/min. After 2 min, a 90-min gradient was applied from 0–40% methanol/0.05% TFA for Peak 1, from 0–35% methanol/0.05% TFA for Peak 2, and from 0–70% methanol/0.05% TFA for Peak 3. One-min fractions were collected, and the absorbance was monitored at 214 nm. Alkalinating activity was determined with 2 μl. Peak 1 eluted as a doublet at 41–49 min, which was called Peaks 1a and 1b. Peak 2 eluted at 65–68 min, and Peak 3 eluted at 56–59 min. Peak 2 was pure as judged by MALDI-MS analysis. Its yield was quantified from its peak area eluting from a narrow bore C18 HPLC column using known peptide standards. The yield was 190 pmol.The early eluting peak of the Peak 1 doublet, Peak 1a, eluted at 42–44 min, whereas Peak 1b eluted at 46–48 min. To further purify Peaks 1a and 1b as well as Peak 3, the analytical C18 column separation was repeated using the pH 6 buffer system but with shallower gradients of 0–30% for Peaks 1a and 1b and 0–40% for Peak 3. Alkalinization activity was assayed using 2 μl from each fraction, identifying Peak 1a in fractions 48–50 and Peak 1b in fractions 52–54. Peak 3 eluted in fractions 64–65. The peaks were pooled and lyophilized. The peaks were further purified and quantified by dissolving in 1 ml of 0.1% TFA/H20 and were loaded onto a narrow bore reversed-phase C18-HPLC column (218TP52, 5-μm 2.1 × 250 mm column, Vydac). After 2 min, a 90-min gradient was applied from 0–25% acetonitrile/0.1% TFA for Peaks 1a and 1b and from 0–35% acetonitrile/0.1% TFA for Peak 3. The flow rate was 0.25 ml/min, and absorbance was monitored at 210 nm. Alkalinization activity was assayed by adding 1 μl of each fraction to 1 ml of tomato cells. The activity was found in fractions 58–59 (Peak 1a), fractions 61–62 (Peak 1b), and fractions 72–73 (Peak 3). These were pooled and quantified by their peak areas relative to known quantities of synthetic TobHypSys I and II (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). The yields for Peaks 1a, 1b, and 3 were 680, 400, and 340 pmol, respectively.Peptide Sequence Analyses and Synthesis—N-terminal sequencing was performed using Edman chemistry on an Applied Biosystems (Foster City, CA) Procise Model 492 protein sequencer. MALDI spectra were obtained using a PerSeptive Biosystems (Framington, MA) Voyager time-of-flight mass spectrometer equipped with a nitrogen laser (337 nm). α-Cyano-4-hyroxycinnamic acid (Aldrich) was used as the matrix. Carboxypeptidase P was used to confirm the C-terminal HQ sequence of Peak 1 as described (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). Peptide synthesis was performed using Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry by solid-phase techniques with an Applied Biosystems Model 431 synthesizer as described previously (15Pearce G. Moura D.S. Stratmann J. Ryan C.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12843-12847Crossref PubMed Scopus (282) Google Scholar). Carbohydrate analysis was performed by mild acid hydrolysis and MALDI-MS analysis as described previously (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar).cDNA Isolation—An expressed sequence tag search of The Institute for Genomic Research tomato (L. esculentum) gene index revealed a protein that contained sequences for both Peaks 1 and 3, fused to a sequence of β-tubulin (BG629139). Oligonucleotide primers were designed to amplify the sequence coding for the two peptides by reverse-transcription polymerase chain reaction. This provided a probe to identify the cDNA in a tomato leaf cDNA library. The GenBank™ accession number of the TomHypSys precursor cDNA is AY292201.Northern Blot Analyses—Leaves of treated and control plants (12 days old) were removed and immediately frozen in liquid N2 and stored at –80 °C until extraction. Each treatment consisted of four leaves, one each from separate plants. The leaf material was ground to a fine powder in a mortar and pestle with liquid N2 and extracted using Triazol buffer (Invitrogen) according to the manufacturer's protocol. Total RNA was quantitated, and 15 μg of each sample was fractionated by electrophoresis in 1.4% agarose-formaldehyde gels, blotted on nylon membranes, and hybridized with [32P]dCTP-specific probes at 62 °C. An 18 S RNA was used as a loading control. Membranes were washed twice with 2× saline/sodium phosphate/EDTA for 20 min at room temperature followed by 40-min washes at 62 °C in 2× saline/sodium phosphate/EDTA, 1% SDS and 1× saline/sodium phosphate/EDTA, 1% SDS and then were exposed to x-ray film for 15 h to 3 days at –80 °C.RESULTS AND DISCUSSIONA crude soluble tomato leaf preparation enriched in small peptides was employed to search for defense-signaling peptides in tomato leaves. The search was initiated as a result of the development of a novel, rapid assay that had identified previously two defense-signaling peptides in tobacco leaf extracts (9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). The assay is based on the response of suspension-cultured cells to peptide ligands that interact with receptors leading to the inhibition of a membrane proton ATPase, causing an increase in pH of the culture medium (16Meindl T. Boller T. Felix G. Plant Cell. 1998; 10: 1561-1570Crossref PubMed Scopus (108) Google Scholar). The assay is called the "medium alkalinization assay," for which only a few μl of fractions eluting from columns are required.The peptides present in the tomato leaf extracts were separated on a C18 reversed-phase HPLC column and assayed for the presence of alkalinating activities in the eluted peaks. Several active peaks were obtained (Fig. 1). Two of the activity peaks contained systemin peptides, identified by their N-terminal sequences (3Pearce G. Strydom D. Johnson S. Ryan C.A. Science. 1991; 253: 895-898Crossref PubMed Scopus (757) Google Scholar), and another activity peak was RALF, a 5-kDa, receptor-mediated peptide that is not involved in plant defense (15Pearce G. Moura D.S. Stratmann J. Ryan C.A. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12843-12847Crossref PubMed Scopus (282) Google Scholar). The three other activity peaks (labeled 1, 2, and 3 in Fig. 1) were mixtures of peptides that required further purification. The active components in these peaks were purified with a series of HPLC columns, similar to the protocols utilized for the purification of tomato and tobacco defense peptides (see "Materials and Methods") (3Pearce G. Strydom D. Johnson S. Ryan C.A. Science. 1991; 253: 895-898Crossref PubMed Scopus (757) Google Scholar, 9Pearce G. Moura D.S. Stratmann J. Ryan C.A. Nature. 2001; 411: 817-820Crossref PubMed Scopus (202) Google Scholar). Peak 1 was separated into two bioactive peaks, Peaks 1a and 1b, whereas the other two peaks, Peaks 2 and 3, were composed of single components. The yields of peptides from 10 kg of tomato leaves, quantified by comparing peak areas with those of known quantities of a synthetic tobacco systemin standard, were estimated to be 680 pmol for Peak 1a, 400 pmol for Peak 1b, 190 pmol for Peak 2, and 340 pmol for Peak 3.MALDI-MS analysis of Peak 1a revealed a major peak at 4191 mass units with a ladder of lower mass peaks indicative of a peptide having different numbers of pentose residues attached (Δ132 mass units) (Fig. 2A, top). After mild acid hydrolysis to release carbohydrates from the main chain, one major peak was found having a mass of 2076 mass units (Fig. 2A, bottom). The mass difference between 4191 and 2076 mass units suggests that the largest glycosylated species in Peak 1 contains 16 pentose units. The masses found for the various other species present in Peak 1 indicate that a single peptide backbone is decorated with 12–17 pentose units.Fig. 2MALDI-mass spectral analysis of components in purified peaks. Each peak was analyzed before (upper panels) and after (lower panels) acid hydrolysis to remove attached carbohydrates. A, Peak 1a; B, Peak 1b; C, Peak 2; D, Peak 3.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Peak 1b revealed a major peak at 3531 mass units and a ladder of species containing various numbers of pentose residues (Fig. 2B, top). Upon hydrolysis, one major peak appeared at 2076 mass units, as with Peak 1a (Fig. 2B, bottom). The mass differences among the various species indicate that 11 pentose residues are present in the largest component of Peak 1b, with a range of 8 to 13 pentose units. In mass spectrometry profiles of both Peaks 1a and 1b, a fragment with a mass of 1947 was present that has not been identified. This mass is 129 mass units smaller than the major 2076 mass peak found with both Peaks 1a and 1b and may be the 2076-mass unit peptide lacking a C-terminal glutamine. Peaks 1a and 1b exhibited identical N-terminal amino acid sequences and identical activities in the alkalinization assay (data not shown). These two peaks were subsequently pooled and treated as a single peptide with varying quantities of carbohydrate attachments.MALDI-MS analysis of Peak 2 revealed a mass of 4513 mass units with a ladder of species ranging from 3985 to 4645 mass units, differing by Δ132 mass units (Fig. 2C, top). Upon acid hydrolysis, the spectra revealed a major peak at 2530 mass units and a minor peak at 2368 mass units, a difference of 162 mass units, which is the predicted mass of a hexose that remained after hydrolysis (Fig. 2C, bottom). Peak 2, by difference of 4513 and 2368 mass units, appears to be decorated with 12–16 pentose residues and one hexose residue.Peak 3 had a major mass peak of 2954 mass units before hydrolysis with minor peaks at 2822 and 2689 mass units, revealing minor species that represented a pentose ladder (Fig. 2D, top). Upon hydrolysis, a major peak resulted at 1651 mass units (Fig. 2D, bottom), a loss of 1303 mass units that does not provide an integer value when divided by the pentose molecular weight of 132 mass units. However, after a short 1-h hydrolysis, a series of mass peaks resulted (data not shown) that indicated the presence of nine pentoses of 132 mass units and one of 113 mass units. The identity of the 113-mass unit fragment is unknown but may be an anhydropentose.Specific activities of the three glycopeptides in the alkalinization assay were determined and compared with the specific activity of tomato systemin (Fig. 3A). Peaks 1 and 3 displayed virtually the same specific activities as systemin, with a half-maximal activity at concentrations of about 0.25 nm. Peak 2, however, induced a much weaker alkalinization response, with a half-maximal activity at concentrations near 2 nm. Comparisons of the time courses of alkalinization of the three peptides with systemin at 1-nm concentrations (Fig. 3B) also indicated that peptides from Peaks 1 and 3 more rapidly caused alkalinization of the cell medium than the peptide from Peak 2. To assess whether the three peptides could activate a defense response in planta, each peptide was supplied to young, excised tomato plants through their cut stems, and the induction of proteinase inhibitor II protein in leaves was assessed 24 h later. This assay has been routinely used to assay the activity of systemin and other elicitors of defense gene expression in tomato plants. Supplying as little as 2.5 pmol of each peptide to tomato plants induced a maximal synthesis and accumulation of inhibitor II protein in leaves of tomato plants, similar to the activity of systemin (Fig. 4). A similar induction of proteinase inhibitor I protein was observed (data not shown).Fig. 3Concentration dependence in the alkalinization activities of the three purified glycosylated peptides derived from Peaks 1, 2, and 3 (Fig. 1) compared with the alkalinization activity of systemin. A, the increase in pH of the tomato suspension cell medium in response to increasing concentrations of peptides was measured 20 min after the addition of different concentrations of peptide. Each bar represents the average of three separate experiments. B, the time course of the alkalinization in response to 1-nm concentrations of Peaks 1 (♦), 2 (▴), and 3 (▪) and systemin (□). Each data point represents the average of three separate experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig. 4Induction of the synthesis of proteinase inhibitor II in leaves of young, excised tomato plants supplied with Peak 1, 2, and 3 peptides and systemin (2.5 pmol/100 μl) through their cut stems. After imbibing the peptide solutions (∼45 min), the plants were transferred to 20-ml vials, placed in a closed Plexiglas box, and incubated under constant light for 24 h (see "Materials and Methods"). Inhibitor II in the expressed juice from