Systemin triggers an increase of cytoplasmic calcium in tomato mesophyll cells: Ca2+ mobilization from intra- and extracellular compartments


C. Moyen Laboratoire de Sciences Végétales, Université de Franche-Comté, Place Maréchal Leclerc, 25030 Besançon, France. Fax: 00 33 3 81 66 56 98


We show here that, within 1–2 min of application, systemin triggers a transient increase of cytoplasmic free calcium concentration ([Ca2+]c) in cells from Lycopersicon esculentum mesophyll. The systemin-induced Ca2+ increase was slightly but not significantly reduced by L-type Ca2+ channel blockers (nifedipine, verapamil and diltiazem) and the Ca2+ chelator [ethylene glycol tetraacetic acid (EGTA)], whereas inorganic Ca2+ channel blockers (LaCl3, CdCl2 and GdCl3) and compounds affecting the release of intracellular Ca2+ from the vacuole (ruthenium red, LiCl, neomycin) strongly reduced the systemin-induced [Ca2+]c increase. By contrast, no inhibitory effect was seen with the potassium and chloride channel blockers tested. Unlike systemin, other inducers of proteinase inhibitor (PI) and of wound-induced protein synthesis, such as jasmonic acid (JA) and bestatin, did not trigger an increase of cytoplasmic Ca2+. The systemin-induced elevation of cytoplasmic Ca2+ which might be an early step in the systemin signalling pathway, appears to involve an influx of extracellular Ca2+ simultaneously through several types of Ca2+ permeable channels, and a release of Ca2+ from intracellular stores sensitive to blockers of inositol 1,4,5-triphosphate (IP3)- and cyclic adenasine 5’-diphosphoribose (cADPR)-mediated Ca2+ release.


Synthesis and accumulation of proteinase inhibitors (PIs) is one of the responses induced in plants by wounding. This response occurs both in the wounded leaf and in distal unwounded leaves within a few hours of the initial damages (Pearce et al. 1991; O’Donnell 1995). This implies the systemic propagation of a signal from the injury site to distal tissues. Putative chemical messengers for this systemic wound signal include systemin. Systemin is an 18 amino acid polypeptide which was originally isolated from tomato (Lycopersicon esculentum) leaves and which, like wounding itself, elicits the systemic synthesis of PIs in tomato plants (Pearce et al. 1991). Systemin was named after its property of being systemically mobile in the phloem (Pearce et al. 1991). Following its application to wound sites on leaves, the polypeptide moves throughout the leaf and is found within 1–2 h in the phloem exudate (Pearce et al. 1991; Narvaez-Vasquez et al. 1995).

Systemin, the first plant polypeptide hormone, behaves in a strikingly similar manner to polypeptide hormones in animal and yeast cells (Schaller & Ryan 1995). However, until now the initial stages in stimulus–response coupling at the plasma membrane have not been clear. The signal transduction pathway is thought to involve the binding of systemin to a membrane receptor (Farmer & Ryan 1992) which has not yet been clearly identified. Schaller & Ryan (1994) found a 50 kDa plasma membrane protein which binds systemin. However, this proteolytic protein does not appear to be a receptor, but has the properties of a furin-like proteinase able to cleave systemin to smaller polypeptides (Schaller & Ryan 1994).

The binding of systemin to a plasma membrane receptor would lead to the activation of a lipase and then to a subsequent intracellular release of linolenic acid and synthesis of methyl jasmonate (Farmer & Ryan 1990, 1992). These two components of the octadecanoid signal transduction pathway are known to have a powerful ability to induce Pin II gene expression (Farmer & Ryan 1990, 1992). Systemin and jasmonic acid (JA) have also now been shown to trigger other cell responses such as synthesis of wound-inducible polyphenol oxidase (Constabel, Bergey & Ryan 1995) and aspartic protease [Schaller & Ryan 1996; see also Schaller & Ryan (1995)].

Moyen & Johannes (1996) recently showed that some of the early events triggered by systemin within a few minutes of application on Lycopersicon esculentum mesophyll tissue are a depolarization of the plasma membrane and a transient H+ efflux followed by a longer lasting H+ influx. Moreover, systemin also triggers an increase of extracellular K+ paralleled by an alkalization in the extracellular medium of suspension-cultured cells of Lycopersicon peruvianum (Felix & Boller 1995). Systemin-induced ion fluxes are affected in the presence of fusicoccin, which has been shown to antagonize the PI synthesis induced by systemin (O’Donnell 1995) and by oligogalacturonides (OG) (Doherty & Bowles 1990). This could suggest that ion fluxes play an important role in the early phases of the systemin signal transduction pathway leading to PI synthesis and probably to some other systemin-induced cell responses.

It is well known that Ca2+ is one of the major ions implicated in the initiation of many signal transduction pathways in living organisms. Following various kinds of stimuli, such as cold-shock (Knight et al. 1991), light (McAinsh et al. 1995; Moyen et al. 1995), touch (Knight et al. 1991), elicitors (Knight et al. 1991; Messiaen et al. 1993; Messiaen & Cutsem 1994) and hormones (Irving, Gehring & Parish 1992; McAinsh, Brownlee & Hetherington 1992), a rise of intracellular Ca2+ occurs and triggers numerous cellular processes by modulation of ion channels, protein kinases and other cellular proteins. Recent results from Fisahn (1997) suggest that Ca2+ ions are among the initial components that mediate the systemic induction of Pin II gene expression in potato plants.

This work was undertaken to investigate whether calcium could be implicated as an intracellular messenger in the systemin signalling pathway leading to at least some of the systemin-induced cell responses. Therefore, the first goal of this work was to see if systemin triggers calcium fluxes and then to identify from which pool (intra- or extracellular) the calcium is mobilized. However, in this first study we did not intend to elucidate the effects of the systemin-induced calcium changes on systemin-induced cell responses.

To determine the systemin-induced calcium changes, tomato plants genetically transformed to express apoaequorin were used and the calcium-sensitive luminescent protein aequorin was reconstituted in vivo.


Apoaequorin binary construction, plant transformation and selection of elite transgenic tomato lines

A DNA fragment containing the apoaequorin coding sequence under the control of the CaMV 35S promoter and CaMV transcriptional termination sequence was excised from a pUC18 based plasmid (Knight et al. 1991) and ligated into the binary vector SLJ44024 (Jones et al. 1992) digested with EcoR1. In the resulting binary vector pMAQ the direction of transcription of the 35S:AEQ gene is parallel to that of the neomycin phosphotransferase gene, NPT, used for selecting transformed plants (Fig. 1).

Figure 1.

. Schematic representation of CaMV 35S PMAQ fusion. Aequorin Xba1-Pst 1 fragment was placed under the control of a 550 bp 35S promoter. The fusion also contains a 200 bp 35S terminator. This 1·4 kb EcoRI fragment was inserted into the SLJ44024 vector conferring resistance to kanamycin in genetically transformed plants.

Plant transformations were performed with the kanamycin-sensitive tomato cv. Moneymaker (Cf-0). The binary vector pMAQ was mobilized into Agrobacterium tumefaciens strain LBA4404 (Hoekema et al. 1983) and transgenic tomato plants were regenerated as described by Fillatti et al. (1987). Ten independent transformants were selected for further analysis. RNA gel blot analysis was performed to identify plants expressing the highest levels of apoaequorin gene transcript. From these, primary transformants were selected that appeared from DNA gel blot analysis and kanamycin segregation ratios to contain a single T-DNA integration site. Two lines PMAQ B and PMAQ G, each homozygous for the T-DNA, were used in this study.

Plant material

The transgenic tomato plants [Lycopersicon esculentum Mill. cv Moneymaker (Cf-0)] expressing apoaequorin used to measure cytoplasmic free calcium were grown in a controlled environment (16 h/22 °C day, 8 h/18 °C night: humidity 70%, light 200 μmol photons m–2 s–1). Fully expanded leaves were excised from these tomato plants. To ensure a rapid penetration of the various effectors used in this study, the lower epidermis of these leaves was removed with forceps. The pieces of mesophyll used to measure the calcium-sensitive luminescence were then allowed to recover several hours from excision. It has been shown previously that in such conditions, mesophyll cells are able to react to systemin (Moyen & Johannes 1996).

Aequorin reconstitution and luminescence measurements

In vitro reconstitution of the aequorin was performed as described previously (Knight et al. 1991). Briefly, 1 mg of mesophyll tissue was homogenized in 500 mm3 of a reconstitution buffer containing 0·5 M NaCl, 5mM ethylene diamine tetraacetic acid (EDTA), 5mMβ-mercaptoethanol, 0·1% gelatin (w/v), 10 mM Tris–HCl, pH 7·4. After spinning the tubes for 10 min in a microfuge, 100 mm3 of supernatant was taken, to which 1 μM coelenterazine (Molecular Probes Inc, Eugene, OR, USA) was added. Reconstitution was performed for 5 h in darkness. The luminescence was then discharged in the luminometer by adding 500 mm3 of 1 M CaCl2. Luminescence measurements were made using a chemiluminometer LB 9501/16 Berthold (Berthold System Inc, Aliquippa, PA, USA). Luminescence counts were recorded continuously and integrated every 1 s and then stored on a personal computer.

In vivo reconstitution of the calcium-sensitive photoprotein aequorin was effected by incubating the pieces of tomato leaves overnight with 2·5 μM coelenterazine in water. Each mesophyll piece was then transferred into a tube (Röhren-Sarstedt Ltd, Leicester, UK) containing 500 mm3 of a medium consisting of 10 mM Mes adjusted to pH 5·6 with 1 M Tris.

Usually 5 min after the start of the recording, 500 mm3 of 200 nM systemin in buffer was added into the tube with a light-tight syringe through a luminometer port, to reach a final concentration of 100 nM (except as otherwise stated). The various inhibitors used in these experiments were added at the chosen concentrations usually 30 min before the addition of systemin (except as otherwise stated). Various types of controls were carried out with ruthenium red to check that the effect of this compound on the calcium changes measured was not due to luminescence quenching (data not shown).

Calibration of calcium measurements

At the end of the experiments, the remaining aequorin was discharged by the addition of 1 cm3 of 2 M CaCl2 and 20% ethanol (v/v). The Ca2+ concentration was calculated from the percentage of the luminescence emitted, divided by the total luminescence, by using the equation pCa=0·332588(– log k) + 5·5593, where k is a rate constant equal to luminescence counts per second divided by total remaining counts (Knight, Trewavas & Knight 1996).

The results shown in Tables 1–3 are expressed as the mean [± standard deviation (SD) with n number of repeats] of the change of resting cytoplasmic calcium concentration (Δ[Ca2+]c) evoked by systemin, with: Δ[Ca2+]c = [Ca2+] at the maximum of the peak –[Ca2+] just before the addition of systemin. The curves shown are the mean curves of the several replicates used to calculate the numerical values given in Tables 1–3, and it should be noted that the magnitude of the resulting peak can sometimes appear smaller than the mean calculated in Tables 1–3. This occurred when following systemin addition, the maximum [Ca2+]c was not reached after the same time period in all the replicates. Therefore, the figures also give an insight into the variability of the lag period between systemin addition and cell response.

Table 1.  . Effect of inorganic and organic calcium channel blockers on the magnitude of the systemin-induced [Ca2+]c increase Thumbnail image of


Effect of systemin on [Ca2+]c in tomato mesophyll

By using tomato plants expressing apoaequorin and reconstituting aequorin in vivo, we found that the resting cytoplasmic calcium concentration ([Ca2+]c) of mesophyll cells was 94 ± 53 nM (mean ± SD, n = 304). When systemin was applied in the bathing medium of tomato mesophyll fragments 22–26 h after their excision from the plants, the addition of systemin elicited an increase of [Ca2+]c. With 100 nM systemin, [Ca2+]c increased after 1 min, reached a peak in ≈ 2 min, and then decreased slowly during the following 30 min. The amplitude of the peak induced was 152 ± 66 nM (mean ± SD, n = 131) showing a variability between the different batches of plants used. Therefore, for every set of experiments, individual sets of controls were performed to take into account the variability of the systemin-induced response.

As shown in 2Fig. 2a, the magnitude of the [Ca2+]c increase evoked by systemin was dependent on the systemin concentration. With experimental batches of plants which showed a Δ[Ca2+]c of 196 ± 64 nM (mean ± SD, n = 23) with 100 nM systemin, the changes of [Ca2+]c evoked by 10 nM, 1 nM, 0·1 nM and 0·01 nM systemin were 108 ± 79 (mean ± SD, n = 14), 83 ± 41 (mean ± SD, n = 15), 64 ± 33 (mean ± SD, n = 14), and 59 ± 40 (mean ± SD, n = 13) nM, respectively. In controls where medium alone was added without systemin, a small apparent [Ca2+]c increase of 14 ± 15 nM (mean ± SD, n = 9) was recorded.

Figure 2.

. Effect of systemin on the resting [Ca2+]c of tomato mesophyll cells. (a) Concentration dependence of the systemin-induced [Ca2+]c increase. (b) Example of oscillations triggered by 100 nM systemin. Systemin was added at t = 300 s.

Following the initial Ca2+ peak elicited by systemin, some oscillations could also be seen which were more or less pronounced between replicates. 2Figure 2b shows an example of a recording where these oscillations were particularly marked. With 100 nM systemin, the mean period of these oscillation was 3·6 ± 0·63 min (mean ± SD, n = 39 from 18 plants).

A small transient elevation of Ca2+ could be seen when fresh medium (with or without systemin) was added in the bathing medium of mesophyll tissues. This peak is far shorter (few seconds) and smaller than the one specifically elicited by systemin and can be attributed to the mechanical shock resulting from the addition of medium. Such a response to mechanical disturbance has already been described by Knight et al. (1991).

In order to investigate whether the calcium involved in the systemin-induced [Ca2+]c increase was extracellular Ca2+ flowing into the cells or Ca2+ released from intracellular stores (such as organelles, e.g. vacuoles), various compounds known to affect cellular processes and ion transport at the plasma membrane and tonoplast were tested in this study.

Effect of ethylene glycol tetraacetic acid (EGTA) on the systemin-induced [Ca2+]c increase

The addition of 1 mM EGTA to the extracellular medium 30 min before the addition of systemin did not significantly affect the systemin-induced [Ca2+]c increase. In the control the Δ[Ca2+]c was 182 ± 28 nM (mean ± SD, n = 15), whereas in EGTA-pretreated plants, the systemin-induced Δ[Ca2+]c was 165 ± 27 nM (mean ± SD, n = 8).

Effect of Ca2+ channel blockers on the systemin-induced [Ca2+]c increase

A 10 min pretreatment with 1 mM of trivalent and divalent cations (LaCl3, GdCl3 and CdCl2) known to be inorganic blockers of calcium channels (Tsien et al. 1987; Hess 1988) caused a 70–80% inhibition of the systemin-induced [Ca2+]c increase (Fig. 3 & Table 1). With 0·1 mM LaCl3 no significant inhibition was seen.

Figure 3.

. Effect of calcium channel blockers on the systemin-induced [Ca2+]c increase. Inorganic blockers such as LaCl3, GdCl3 and CdCl2 were used at 1 mM and organic blockers such as nifedipine (NIF), verapamil (VERA) and diltiazem (DILT) at 0·1 mM. Systemin was added at a final concentration of 100 nM at t = 300 s.

Compared with these inorganic channel blockers, organic channel blockers such as nifedipine, verapamil and diltiazem only exerted a small inhibitory effect on the systemin-induced [Ca2+]c increase (17–26% inhibition) which was not statistically significant [Fig. 3 (insert) & Table 1].

Effect of putative intracellular Ca2+ release modulators on the systemin-induced [Ca2+]c increase

As shown in 4Fig. 4a, when applied 30 min prior to systemin, 50 μM ruthenium red induced a strong inhibition (75%) of the systemin-induced [Ca2+]c increase (Table 2). A 1 h pretreatment with 10 mM LiCl also resulted in a clear reduction (56%) of the magnitude of the systemin-induced [Ca2+]c increase (Fig. 4b & Table 2). When applied at 100 μM for 30 min, neomycin alone evoked a small increase in resting [Ca2+]c (36 ± 22 nM, n = 6) and inhibited the systemin-induced [Ca2+]c rise by 63% (Fig. 4c & Table 2). Ten millimolar caffeine alone evoked a small increase (33 ± 10 nM, n = 4) of the resting [Ca2+]c beginning ≈ 2 min after its addition to the bathing medium of the tomato mesophyll tissue (Fig. 4d insert). Then, after the addition of systemin, [Ca2+]c increased and reached a maximum slightly higher than the one reached by the control without caffeine pretreatment (Fig. 4d & Table 2). Thus, in the caffeine-pretreated plants, the systemin-induced Δ[Ca2+]c itself is smaller, but occurs from a higher ‘resting’[Ca2+]c.

Figure 4.

. Inhibition of the systemin-induced [Ca2+]c increase by compounds affecting the release of Ca2+ from intracellular stores. (a) Effect of 50 μM ruthenium red; (b) effect of 10 mM LiCl; (c) effect of 100 μM neomycin; (d) effect of 10 mM caffeine (insert: effect of caffeine on the resting [Ca2+]c). Systemin was added at a final concentration of 100 nM at t = 300 s.

Table 2.  . Effect of compounds affecting the release of Ca2+ from intracellular stores on the magnitude of the systemin-induced [Ca2+]c increase Thumbnail image of

Effect of K+ and Cl channel blockers on the systemin-induced [Ca2+]c increase

When mesophyll pieces were incubated for 30 min with K+ channel blockers such as tetraethylammonium (TEA) (10 mM) or Cl channel blockers such as anthracene-9-carboxylic acid (A9C) (300 μM), niflumic acid (100 μM) or 4,4’-diisothiocyanatostilbene-2,2’-disulfonic acid (DIDS) (300 μM), no significant effect on the magnitude of the systemin-induced [Ca2+]c increase was observed (Table 3).

Table 3.  . Effect of potassium and chloride channel blockers on the magnitude of the systemin-induced [Ca2+]c increase Thumbnail image of

Effects of PI synthesis inducers (JA and bestatin) on the systemin-induced [Ca2+]c increase

5Figure 5a and b shows that JA and bestatin which, like systemin, are known to induce PI synthesis (Farmer & Ryan 1990; Schaller, Bergey & Ryan 1995), had no effect on the resting [Ca2+]c. On batches of plants where 100 nM systemin evoked Ca2+ increases of, respectively, 91±30 nM (mean ± SD, n=5) and 102±33nM (mean±SD, n = 5), 100 μM JA and 100 μM bestatin only triggered [Ca2+]c rises of, respectively, 10±7nM (mean±SD, n = 6) and 9±11 nM (mean±SD, n=5) which were not significantly different from the effect seen in the control (addition of medium).

Figure 5.

. Effect of inducers of proteinase inhibitor (PI) synthesis on the resting [Ca2+]c. (a) Effect of 100 μM jasmonic acid (JA); (b) effect of 100 μM bestatin. For control, medium alone was added. Systemin was added at a final concentration of 100 nM. Systemin, JA and bestatin were added at t = 120 s.


Systemin-induced cytoplasmic Ca2+ increase

Systemin is a polypeptide known to elicit the systemic synthesis of PIs in tomato plants (Pearce et al. 1991). Up to now the early events evoked by systemin at the plasma membrane surface are far from understood. It is supposed that systemin binds to a putative receptor and activates a lipase which leads to the release of JA. However, the steps between the binding of systemin to its receptor and the activation of the octadecanoic pathway are unclear. It has been shown that one of the first events triggered by systemin in Lycopersicon esculentum mesophyll cells is a membrane depolarization and complex effects on H+ fluxes (Moyen & Johannes 1996). Increases of extracellular pH and K+ concentration have also been reported following the addition of systemin to the bathing medium of Lycopersicon peruvianum suspension-cultured cells (Felix & Boller 1995).

In this study, we showed that in Lycopersicon esculentum mesophyll cells, systemin also triggers, within 2 min of its application, an increase in cytoplasmic Ca2+ concentration ([Ca2+]c). The choice of using tomato plants expressing apoaequorin was made to allow us to monitor [Ca2+]c changes non-invasively on mesophyll tissue because the use of protoplasts, which are required for [Ca2+]c measurements with calcium binding fluorescent dyes, has several disadvantages. For example, during protoplast isolation, the systemin receptor could be damaged. Furthermore, the cell wall digesting enzymes release elicitors such as OG which are thought to trigger defence reactions through the same signalling pathway as systemin (Farmer & Ryan 1992) and could, therefore, affect the systemin response. It has also been shown that cell cultures show a less complex response to systemin (Moyen & Johannes 1996) and to hormones such as auxin (Felle, Peters & Palme 1991) than tissues. Moreover, the cultured cell response to systemin differs with the cell line used (Felix & Boller 1995). For our experiments the pieces of mesophyll used did not contain any major vascular bundles and minor ones were usually removed when the lower epidermis was stripped off. Previous data (Moyen & Johannes 1996) have shown that in a few hours the pieces of mesophyll have recovered from the stripping of their lower epidermis and are able to react again to systemin. The number of epidermal cells is small compared with the number of mesophyll cells. Therefore, although we cannot entirely rule out that part of the response to systemin might occur in the upper epidermal cells, it can be assumed that the main effect seen represents the mesophyll cell response. A consequence of not having all the cells reacting to systemin would be an underestimation of the systemin-induced [Ca2+]c change occurring in the responding cells.

Our results show that the systemin-induced [Ca2+]c increase occurred within 1 min and reached a peak after ≈ 2 min. Our data also showed that in many cases oscillations occurred following the initial Ca2+ increase triggered by systemin. These oscillations, which have a particularly high magnitude on the example shown in 2Fig. 2b, can also be seen in many of the other figures but with a lower magnitude. One of the hypotheses for this discrepancy could be that in some recordings most of the individual cell oscillating responses occurred in phase leading to marked overall oscillations, while in other recordings the individual cell oscillating responses were not in phase or had different frequencies, therefore leading to less pronounced oscillations. Another hypothesis could be that only some types of mesophyll cells show oscillating responses. To address this question, individual cell responses would need to be recorded with a different method more sensitive than aequorin-reported calcium changes.

Oscillations of intracellular Ca2+ concentration have already been described in response to diverse stimuli and are thought to be a way of encoding information (Berridge 1990; McAinsh et al. 1995). Previous studies have also shown that some elicitors and hormones can evoke a rise of cytoplasmic Ca2+ in plant cells. Elicitors such as OG have been reported to evoke an increase in cytoplasmic Ca2+ in carrot protoplasts (Messiaen et al. 1993; Messiaen & Van Cutsem 1994) and a decrease of the Ca2+ concentration of the bathing medium of tobacco suspension cells (Mathieu et al. 1991). However, the kinetics of these OG-induced [Ca2+] changes differ strongly from the one evoked by systemin in tomato mesophyll tissue. In carrot protoplasts, the OG-induced [Ca2+]c increase started between 5 and 10 min after OG addition, and reached a concentration of more than 1 μM in 30 min (Messiaen & Van Cutsem 1994). With tobacco suspension cells, the OG-induced decrease of the external [Ca2+] reached a maximum after 50 min (Mathieu et al. 1991). Plant hormones such as abscisic acid (ABA) (Gilroy et al. 1991; Irving et al. 1992; McAinsh et al. 1992) and indol acetic acid (IAA) (Irving et al. 1992) have also been shown to trigger cytosolic [Ca2+] rises in plant cells. As shown here with systemin, ABA and IAA evoked a 1·5–3-fold increase in resting [Ca2+]c (Irving et al. 1992). Therefore, as in the cases of ABA and IAA, the systemin-induced [Ca2+]c change could be expected to be sufficient to trigger other cell responses.

JA and bestatin are ineffective on [Ca2+]c

JA and bestatin are known to induce some of the cell responses also evoked by systemin, in particular PI synthesis (Farmer & Ryan 1990; Schaller et al. 1995). More recently it has been shown that, like systemin, JA and bestatin also induce aspartic protease synthesis [Schaller & Ryan 1996; see also Schaller & Ryan (1995)] and wound-inducible polyphenol oxidase synthesis (Constabel et al. 1995). Our data show that under our experimental conditions, unlike systemin, JA and bestatin had no effect on the [Ca2+]c of tomato mesophyll cells. JA is one of the end products of the octadecanoid pathway by which the systemin signal is transduced within the plant (Farmer & Ryan 1992). Our results therefore indicate that the systemin-induced [Ca2+]c increase is a step occurring upstream of JA biosynthesis.

Bestatin is an inhibitor of aminopeptidase which induces the same group of defence genes that are activated by systemin (Schaller & Ryan 1995, 1996; Schaller et al. 1995). The absence of a bestatin effect on [Ca2+]c agrees with the conclusion of Schaller et al. (1995) that bestatin appears to be exerting its effect close to the transcriptional control of the defence gene, i.e. late in the signalling pathway.

Ca2+ mobilization from different cellular pools upon systemin addition

The systemin-induced [Ca2+]c increase is inhibited partially by compounds such as ruthenium red, LiCl and neomycin, which are three compounds thought to interfere with the release of calcium from intracellular compartments such as the vacuole. Neomycin and LiCl both affect IP3-mediated Ca2+ release (Berridge 1984; Frankling-Tong et al. 1996), whereas ruthenium red is known to block the cADPR-mediated Ca2+ release from vacuoles (Allen, Muir & Sanders 1995). The marked inhibition of the systemin-induced Ca2+ increase by these compounds argues for the involvement of a Ca2+ release from intracellular compartments which might be part of the systemin signalling pathway.

The present data also show that the systemin-induced [Ca2+]c increase was slightly but not significantly inhibited by organic blockers of voltage-gated Ca2+ channels (nifedipine, verapamil and diltiazem), and not inhibited by Cl channel blockers (DIDS, A9C, niflumic acid) or by K+ channel blockers (TEA, CsCl; data not shown). By contrast, LaCl3, CdCl2 and GdCl3 which are known to be very potent calcium channel blockers (Tsien et al. 1987; Hess 1988) abolished almost completely the [Ca2+]c increase triggered by systemin. In our study these three channel blockers were only applied at 1 mM for 10 min before systemin was added. Considering that mesophyll tissues and not protoplasts were used as an experimental system, the short incubation time and low inhibitor concentration suggest that these antagonists exert their effect mainly on the plasma membrane and the possibility that these ions enter the cells and have an additional effect on Ca2+ mobilization from internal stores is unlikely. It should also be kept in mind that no changes in the resting [Ca2+]c have been observed during the inhibitor treatments (data not shown) as could be expected if the calcium antagonists were entering the cell and displacing Ca2+ from its intracellular binding sites. The stronger effect triggered by inorganic Ca2+ channel blockers compared with organic blockers could be attributed to the blockade of more types of channels by these cations. Indeed it has been shown in animal cells, that Cd2+ and La3+ block at least two types of voltage-activated Ca2+ channels (L- and N-types), whereas organic antagonists such as nifedipine and verapamil only block the L-type (Tsien et al. 1987; Hess 1988). It should be acknowledged that like most inhibitors, calcium channel blockers are not perfectly specific and have been shown to block some other channel types such as K+ channels in hypocotyls and root cells of Arabidopsis (Lewis & Spalding 1998). However, it should be kept in mind that even if these blockers block channels that are not specific Ca2+ channels, this would not affect our conclusions because we have measured their effect on the aequorin-reported intracellular [Ca2+]. An argument in support of that point comes from the fact that blockers of K+ channels and of Cl channels tested in this study showed no effect on the systemin-induced [Ca2+] changes.

Thus, taken together these results suggest that some of the calcium mobilized by systemin comes from the apoplast through a non-selective Ca2+ permeable channel or simultaneously through several types of channels. It has recently been shown in parsley protoplasts that a Ca2+ permeable channel sensitive to LaCl3 is activated upon addition of a polypeptide elicitor (Zimmermann et al. 1997). Activation of non-selective Ca2+ permeable channels by abscisic acid has also been reported (Schroeder & Hagiwara 1990).

The observation that the systemin-induced Ca2+ increase occurred even when the bathing pH buffered medium of the mesophyll pieces contained EGTA does not rule out the possibility that part of the Ca2+ involved comes from the extracellular space. Indeed, the measurement of free [Ca2+] in the bathing medium of mesophyll tissues (by spectrophotometry using Arsenazo III) showed that in our experimental conditions, after a 30 min incubation with 1 mM EGTA, or after a 30 min or 60 min incubation with 10 mM EGTA the free [Ca2+] reached several micromolar (data not shown). Therefore, it can be assumed that during an incubation with EGTA, the chelation of the external calcium maintains a calcium gradient between the bathing medium and the tissues, and leads to a release of calcium from the tissues. The tissue released enough Ca2+ to get an external Ca2+ concentration near the plasma membrane which was low but sufficient to allow Ca2+ influx through a high-affinity permeable Ca2+ channel similar to the one recently described by Pineros & Tester (1997).

From the data obtained with Ca2+ channel blockers we could suggest that the systemin-induced Ca2+ mobilization occurred through a non-selective Ca2+ permeable channel or simultaneously through several types of channels. The results with EGTA would be consistent with the hypothesis that the [Ca2+] increase results from Ca2+ fluxes through several types of calcium permeable channels, one of which must be a high-affinity Ca2+ channel. The contribution of several types of Ca2+ permeable channels (and of several cellular Ca2+ pools) could also explain why the inhibition of the systemin-induced [Ca2+] changes seen in the presence of the various Ca2+ antagonists tested was never complete.

If we were to suppose that the systemin-induced calcium changes are one of the first steps of the signalling pathway leading to important defence reactions following wounding, it would not seem unreasonable for a cell to have several ways of raising its intracellular [Ca2+] as a safe mechanism to trigger vital defence processes.

This influx of external Ca2+ might then induce the release of Ca2+ from intracellular stores. Some evidence for a Ca2+-induced Ca2+ release via slowly activated vacuolar (SV) channels has been shown in plant vacuoles (Ward & Schroeder 1994; Allen & Sanders 1995). In the ABA-induced Ca2+ increase observed in guard cells, it has also been postulated that both influx of external calcium and release of calcium from intracellular stores occurred, as Ca2+ channel blockers could only partially inhibit the calcium increase triggered by ABA (McAinsh, Brownlee & Hetherington 1991; McAinsh et al. 1995). A model describing the mobilization of Ca2+ from two different pools, one IP3-sensitive and one IP3-insensitive has been proposed to explain the initiation of Ca2+ oscillations (Berridge 1990) and could account for the oscillations triggered by systemin in tomato mesophyll cells.

It is likely that the systemin-induced changes in [Ca2+]c play an important role in the signalling pathway. This is suggested by recent findings of Fisahn (1997) which implicate that elevation of [Ca2+]c is involved in the induction of wound-inducible Pin II genes. This was concluded from simultaneous recordings of membrane potential and of aequorin-emitted luminescence changes performed to correlate the electrical response and the calcium component of the long distance signal mediating the systemic induction of Pin II gene expression.

The present report does not intend to elucidate the effects of [Ca2+]c changes on the induction of wound-inducible genes. However, the direct target of the systemin-induced [Ca2+]c elevations and the physiological significance of this systemin-induced calcium increase will need to be identified in future studies keeping in mind that local changes in [Ca2+]c (e.g. in the vicinity of the plasma membrane and tonoplast) are likely to be much higher than the average values obtained from a large number of mesophyll cells.

Taken together our data suggest that systemin triggered an influx of extracellular Ca2+ simultaneously through several types of channels such as non-specific Ca2+ permeable channels and a high-affinity Ca2+ channel, and a release of Ca2+ from intracellular stores sensitive to blockers of IP3- and cADPR-mediated Ca2+ release.

Our findings of a systemin effect on tomato mesophyll cell membrane potential (Moyen & Johannes 1996) and on [Ca2+]c could also contribute to the understanding of the nature of the long distance signal involved in wounding response. Besides chemical signalling via the mobile polypeptide systemin (Bergey, Hoi & Ryan 1996), there is also evidence for electrical (Wildon et al. 1992; Stankovic & Davies 1996) and hydraulic signalling (Malone, Alarcon & Palumbo 1994). From our results, it can be expected that systemin propagation across mesophyll tissue leads to an electrical (depolarizing) wave and a calcium wave. Therefore, it is possible that in cells containing voltage-gated and Ca2+-activated channels action potential could be triggered. In this way it could be envisaged that a combination of chemical and electrical signalling is involved.

The kinetics of the systemin-induced [Ca2+] changes and depolarization (Moyen & Johannes 1996) are quite similar and could suggest a close link between these two parameters. So far it is difficult to speculate whether the depolarization is caused by calcium influx, or rapidly follows as a consequence of the changes in [Ca2+] (e.g. by Ca2+ activation of anion channels). Furthermore, other systemin-induced ion fluxes – such as H+ (Moyen & Johannes 1996) and K+ (Felix & Boller 1995) – have been shown to occur early following the systemin application and are, therefore, also very likely to contribute to the systemin-induced depolarization. The relationship between calcium uptake and membrane depolarization is an issue that will need to be specifically addressed in future experiments.


The work with systemin, performed at the University of York, was supported by a grant from the BBSRC (ICS AI 87/537). The work on tomato plant transformation was performed at the Sainsbury Laboratory and supported by the Gatsby Foundation. We are very grateful to Dr R. Hall (University of York) for giving us access to his luminometer and Kate Harrison (Sainsbury Laboratory) for carrying out the tomato transformation. MRK is a Royal Society Research Fellow.


  1. Present address: C. Moyen, Laboratoire de Sciences Végétales, Université de Franche-Comté, Place Maréchal Leclerc, 25030 Besançon, France. Fax: 00 33 3 81 66 56 98.

  2. Present address: Department of Botany, North Carolina State University, Raleigh, NC 27695-7612, USA.