Cell wall oligogalacturonides increase cytosolic free calcium in carrot protoplasts

Pectin is an acidic polysaccharide constituent of the plant cell wall. It is a polymer of α-(1,4)-linked galacturonic acid residues with neutral sugars, including a-(1,2)-linked rhamnosyl residues. Pectin can exist in either a sol or gel form and the type of gel it forms may be concerned with the regulation of cell wall growth. Cell wall pectin is structurally associated with Ca²⁺, which can crosslink pectin chains, causing them to cluster and gel (Jarvis, 1982). Binding of Ca²⁺ and the extent and type of gel formation are modified in a sensitive fashion by either acetylation or methylesterification of uronic acid, or by the attachment of longer, neutral side chains.

Many physical studies have shown that binding of Ca²⁺ leads to conformational changes in pectin chains. This is generally believed to result in the “egg box” structure, with Ca²⁺ (the “eggs”) sandwiched between egg box pectin chains in dimeric or multimeric forms (Fig. 1A,B and C). This conformation can also be induced by Mn²⁺. Structural analyses have indicated that the egg box conformation requires oligopectates with >10–14 residues (Thibault, 1980; Morris et al., 1982; Powell et al., 1982; Irwin et al., 1984; Thibault and Renaudo, 1985). Monoclonal antibodies have recently been raised to this supramolecular structure (Liners et al., 1989).

Invasion of a resistant plant host by a pathogen triggers a set of well-defined defence mechanisms (Horsfall and Cowling, 1980; Lamb et al., 1989). These host responses may involve the production of mechanical barriers, hydrolytic enzymes or specific chemicals such as phytoalexins designed to impair pathogen growth (Lamb et al., 1989; Mieth et al., 1986; Dietrich et al., 1990; Forrest and Lyon, 1990). In a number of cases, the eliciting signal is believed to be oligogalacturonides released from the plant cell wall. Optimal elicitation activity has been derived from oligogalacturonides with 9-14 residues (Bruce and West, 1989; Davis et al., 1986; Jin and West, 1984; Horn et al., 1989; Forrest and Lyon, 1990). Oligogalacturonides of the same size have also been found to result in protein phosphorylation (Farmer et al., 1991), as well as signalling other biological responses (Aldington et al., 1991).

How signals provided by cell wall fragments are transduced to initiate host defense mechanisms is still unknown, but recent information suggests that cytosolic free Ca²⁺ ([Ca²⁺]_i) might be involved (Cramer et al., 1985; Kurosaki et al., 1987; Stäb and Ebel, 1987). The critical analysis (i.e. measurements of [Ca²⁺]_i in living cells after elicitor treatment) is still unavailable. As a result we decided to examine the effects of oligogalacturonides on [Ca²⁺]_i using fluorescence ratio imaging of the dye Indo-1. Because of difficulties in loading walled cells with Ca²⁺-sensitive dyes (Callaham and Hepler, 1991; Read et al., 1992), we used carrot protoplasts to study the effects of cell wall oligogalacturonides produced by Erwinia carotovora atroseptica endopectate lyase. Since only small quantities of some of the elicitor preparations were available, our measurements were limited to just fluorescence ratio imaging of [Ca²⁺]_i in single protoplasts. Our results indicate that oligogalacturonides with a degree of polymerization (DP) >9 induce a prolonged increase in [Ca²⁺]_i. Furthermore, the data show that the oligogalacturonides need to have a Ca²⁺-induced conformation in order to elevate [Ca²⁺]_i.

4.4 g l^-1 Murashige and Skoog salts (Sigma Chemical Co.), 30 g l^-1 sucrose and 1 mg l-1 2,4-dichlorophenoyxacetic acid (pH 5.7). 50 ml cell suspensions were grown at 25°C in 250 ml flasks under constant agitation (150 revs per min) in the dark. The cells were subcultured every week by a 1:4 dilution into fresh medium.

The culture medium was supplemented with 100 g l-1 mannitol, 15 g l-1 pectinase (Sigma) and 5 g l-1 cellulase RS (Onozuka) for protoplast isolation. 1 g of D. carota cells per 10 ml of enzyme cocktail was used for each experiment. The enzymatic digestion of the cell walls was carried out overnight, with mild agitation at 25°C. Protoplasts were then successively filtered through 200 mm and 80 mm nylon filters and centrifuged at 100 g for 5 min. The pellet was resuspended in enzyme-free medium and centrifuged at 200 g for 3 min. The supernatant was discarded and the washing procedure repeated 3 times to remove the enzymes.

The Ca²⁺-sensitive dye Indo-1 was loaded by electroporation using a gene pulser (BioRad) (Gilroy et al., 1986). Protoplasts (106 ml-1) were permeabilized in 0.4 cm cuvettes in Murashige and Skoog medium supplemented with 100 g 1-1 mannitol, pH 6.5, and containing 50 mM Indo-1 (free acid, Molecular Probes Inc., Eugene, OR, USA). For electroporation, the field strength was 1200 V cm-1, the time constant 100 ms and 3 pulses were delivered at 5 s intervals. The protoplasts were kept on ice for 10 min before and after electroporation. The electroporated protoplasts were maintained in the dark for 90 min at room temperature before analysis. After this period no long term effects of protoplasting, low temperature or electroporation were evident and, unless stimulated, the protoplasts maintained constant [Ca²⁺]_i levels for fur- ther periods in excess of 30 min. Further details of dye loading and validation of [Ca²⁺]_i measurements in carrot protoplasts can be found in Gilroy et al. (1986, 1987, 1989).

Fluorescence ratio imaging was performed using equipment and techniques previously described in detail (Gilroy et al., 1991; Read et al., 1992). Loaded cells were analysed using a Nikon CF Fluor DL 40× 0.85 NA objective on a Nikon Diaphot inverted micro- scope equipped with epifluorescence optics. Samples were excited at 350 nm with a 75 W xenon light source. Fluorescence was detected at 405 nm and 480 nm using 10 nm bandwidth interfer- ence filters mounted in a computer-controlled filter wheel (New- castle Photonic Systems, Newcastle-Upon-Tyne, UK). Digitized images were captured at 10 min intervals using an Extended ISIS CCD camera (Photonic Science, Kent, UK) connected to a Synapse frame store (Synoptics Ltd., Cambridge, UK). Image cap- ture was controlled by the Semper 6 Plus image processing lan- guage run on a Dell 310 microcomputer equipped with a 80386 microprocessor, maths coprocessor and 90 Mb hard disc. Auto- fluorescence and dark current were subtracted from each image prior to ratio image calculation. Because autofluorescence could not be imaged in the same cell before and after electroporation, an average pixel value of autofluorescence at each wavelength was obtained from 10 unloaded protoplasts. Ratio images were subse- quently colour-coded to represent different Ca²⁺ concentrations determined after in vitro calibration. This was carried out using a set of buffers containing 3 mM Indo-1, 100 mM KCl, 1 mM MgCl₂, 10 mM HEPES (pH 7.0), 10 mM EGTA and variable amounts of CaCl₂ to give appropriate free Ca²⁺ concentrations (Grynkievicz et al., 1985). Hard copies of images were produced with a Mitsubishi CP 100B video copy processor.

Vital staining of Indo-1 loaded protoplasts with trypan blue demonstrated that they retained viability before, during and after [Ca²⁺]_i imaging (data not shown).

Dye loading was assessed by correlative fluorescence microscopy and phase contrast light microscopy using a Zeiss Photomicro- scope III. A 350/10 nm excitation filter, 380 nm dichroic mirror and a 400 nm long pass barrier filter were used for fluorescence microscopy. Photographic images were recorded on Kodak TMAX 400 35 mm film.

Protoplasts were mixed with polygalacturonic acid (PGA, Sigma Chemical Co.) or oligogalacturonides of different lengths with dif- ferent conformations.

Oligogalacturonides (Fig. 1B) were produced by digestion of PGA using a purified endopectate lyase from Erwinia carotovora atroseptica (Stack et al., 1980). Digestion products were purified by gel permeation on Bio-gel P6 (Thibault, 1980). Two fractions were used in experiments: oligogalacturonides ≤9 residues long and oligogalacturonides >9 residues long. The choice of these two fractions was dictated by the ability of these oligogalacturonides to be recognized by the anti-pectic dimer 2F4 monoclonal anti- body described by Liners et al. (1989, 1992).

The dimeric conformation of either PGA or oligogalacturonides was produced in 10% mannitol with either 150 mM NaCl and 0.5 mM CaCl₂ or 150 mM NaCl and 0.5 mM MnCl₂ (Irwin et al., 1984). The single chain conformation was maintained in solutions of 10% mannitol containing 150 mM NaCl.

For PGA treatments the protoplasts were adhered to a poly- lysine-coated slide before treatment and perfused with the inducing solution. Since only very small amounts of the purified oli- gogalacturonides were available, in these experiments protoplasts were mixed first and then allowed to settle before imaging. Thus in these experiments the earliest images were only obtained after 3-4 min. No effect of immobilization or of 150 mM NaCl and 0.5 mM Ca²⁺ on [Ca²⁺]_i was detected in controls (data not shown). In some cases, the sucrose concentration was reduced to keep osmolality constant. Again, we found this had no effect on [Ca²⁺]_i (data not shown).

Photobleaching and dye leakage from the cells limited the avail- able time for imaging to approximately 30 min. We considered this satisfactory for our purposes because other studies have demonstrated that oligogalacturonides induce membrane depolar- ization during this period (Thain et al., 1990; Mathieu et al., 1991).

All experiments were repeated at least four times to establish a common pattern of response. Representative experiments only are illustrated here.

To demonstrate that cells remained viable and capable of respond- ing to elicitor treatment, we measured PAL activity according to the assay method described by Lamb et al. (1979). The concen- trations of elicitors used were the same as those used for imag- ing [Ca²⁺]_i. For measuring PAL activity, 0.5 ml of crude proto- plast extract was incubated for 1 h at 40°C in a final reaction mixture of 3.0 ml containing 30 μM sodium borate, pH 8.8, and 30 μM phenylalanine. Absorbance at 286 nm was read against an identical mixture except that 30 mM of D-phenylalanine was sub- stituted for L-phenylalanine.

The protoplasts used in this investigation were highly vac- uolate (Fig. 2A). Only cells in which dye was clearly loaded in the cytoplasm (amounting to approx. 10% of pro- toplasts) were used in the analyses. Organelles such as mitochondria, the endoplasmic reticulum and vacuoles (Bush et al., 1989; Schroeder and Hedrich, 1989; Rizzuto et al., 1992) possess free Ca²⁺ concentrations >5 μM, which are significantly higher than that found in the cytosol. Because Indo-1 has a high affinity for free Ca²⁺ (K_d ∼ 250 nM, Grynkievicz et al., 1985), any dye com- partmentalized within these organelles will be Ca²⁺-satu- rated and thus maximally fluorescent. Punctate fluores- cence or localized high ratio values were not evident in unstimulated cells (Figs 2A, 3B,C,D and F), indicating that significant dye sequestration within organelles was absent. Cells which exhibited vacuolar loading were obvious (Fig. 2B) and were rejected for analysis. Autofluorescence detected with the narrow 10 nm bandwidth emission fil- ters used for fluorescence ratio imaging was negligible but was, nevertheless, subtracted at each emission wavelength during the production of the ratio images. The resting level of [Ca²⁺]_i ranged from 180 nM to 300 nM in different pro- toplasts, which is in good agreement with measurements we have made previously on carrot protoplasts (Gilroy et al., 1986, 1987).

Fig. 3A shows that incubation of Indo-1-loaded proto- plasts with 1 mg ml-1 PGA plus Ca²⁺ (i.e. with PGA in the dimeric conformation) leads to increases in [Ca²⁺]_i. The increase in [Ca²⁺]_i typically commenced between 5 and 10 min and reached concentrations over 1 mM by 30 min. No [Ca²⁺]_i increase was detected in any protoplast before 5 min following exposure to PGA (n = 6). The observed increases in [Ca²⁺]_i varied between protoplasts, even within the same preparation (Fig. 3A). The highest concentration of [Ca²⁺]_i commonly appeared around the periphery of a protoplast with a gradient (red-yellow-green-blue) which seemed to propogate towards the protoplast interior (e.g. see Fig. 3C). This is circumstantial evidence for Ca²⁺ entering the cytosol from outside rather than being mobilised from internal Ca²⁺ reserves. Further evidence that Ca ²⁺ channels located in the plasma membrane may be involved in the [Ca²⁺]_i increase is shown in Fig. 3B. In the presence of the Ca²⁺ channel blocker verapamil, PGA plus Ca²⁺ failed to induce the [Ca²⁺]_i increase shown in Fig. 3A. The increase in [Ca²⁺]_i was observed at a concentration ranging from 0.2 mg ml-1 up to 1 mg ml-1 (data not shown). Below 0.2 mg ml-1 no increase was evident. Concentrations higher than 1 mg ml-1 were not tested.

Fig. 3C and D further indicates that PGA must have the dimeric conformation to be active. PGA plus Mn²⁺ increases [Ca²⁺]_i with a similar time course (Fig. 3C) to PGA plus Ca²⁺ (Fig. 3A). Again the location of [Ca²⁺]_i increase is clearly in the cytoplasm with the concentration gradient in [Ca²⁺]_i decreasing away from the plasma mem- brane in the focal plane imaged. Fig. 3D shows that the presence of Ca²⁺ or Mn²⁺ in solution is necessary for the PGA-induced increase in [Ca²⁺]_i. PGA chains which are not in a dimeric conformation are therefore inactive.

The effects of oligogalacturonides, with chain lengths shorter and greater than 9 residues, were each tested (Fig. 3E and F). Changes in [Ca²⁺]_i were only observed with oli- gogalacturonides >9 residues in length (for DP >9, n = 4; for DP ≤9, n = 6). The time course of [Ca²⁺]_i increase in Fig. 3E commenced within 20 min but the changes in [Ca²⁺]_i were not as dramatic as those observed after the addition of PGA dimers (Fig. 3A and C).

PAL activity increased in carrot protoplasts treated with oligogalacturonides with DP >9, or dimeric PGA in the presence either Ca²⁺ or Mn²⁺. Large increases in PAL activity were not observed until 18 h after elicitor addition. Significant increases in PAL activity did not occur after treatment with oligogalacturonides with DP ≤9, monomeric PGA, or dimeric PGA in the presence of verapamil (Fig. 4).

The data shown here indicate that PGA and oligogalactur- onides released from the plant cell wall increase [Ca²⁺]_i in a conformation-dependent manner. The addition of PGA in the presence of either Ca²⁺ or Mn²⁺ initiates [Ca²⁺]_i increases commencing some 5-20 min after exposure. Cal- cium and Mn ²⁺ are known to induce a supramolecular con- formation in PGA (Morris et al., 1982; Powell et al., 1982; Irwin et al., 1984). Evidence supporting the inductive char- acter of the dimeric conformation was obtained by treating protoplasts with oligogalacturonides, greater or less than 9 residues in length, with appropriate concentrations of mono- valent and divalent cations to induce the dimeric confor- mation (Fig. 1; Liners et al., 1989, 1992). Our data show that only oligomers of DP >9 in the dimeric conformation (i.e. the pectic fraction recognized by the 2F4 antibody) induced an elevation in [Ca²⁺]_i within 30 min of addition. Because the dye photobleaching and leakage prevented [Ca²⁺]_i imaging beyond 30 min, we cannot exclude the pos- sibility that dimeric oligomers of DP ≤9 do not induce a delayed [Ca²⁺]_i elevation. Whether the multimer has the same inductive effect on [Ca²⁺]_i is not known. However, we consider it possible that the multimer would not be active in plants because the large size of multimers would restrict their diffusion across the cell wall.

The inhibition of a dimeric PGA-induced [Ca²⁺]_i eleva- tion by verapamil suggests a requirement for Ca²⁺ channel activity for the response to occur. Although we considered using LaCl₃ as an additional channel blocker, La3+ can dis- place Ca²⁺ from oligogalacturonides and results would therefore be difficult to interpret. Several of the images we presented indicate a gradient of [Ca²⁺]_i from the plasma membrane inwards. This suggests the involvement of plasma membrane Ca²⁺-channels. The opening of Ca²⁺- channels might be induced by membrane depolarization, as has been described previously (Thain et al., 1990; Mathieu et al., 1991), especially since verapamil blocks L-type, volt- age-gated Ca ²⁺-channels (Hess, 1990). From our study we are unable rule out the possibility of Ca²⁺-induced Ca²⁺- release (Berridge and Irvine, 1989), whereby a modest amount of Ca²⁺ entering the cell from the outside is itself the trigger for release of Ca²⁺ from internal stores.

Mn²⁺ can act as a surrogate for Ca²⁺ and enters through open Ca²⁺ channels (Hess, 1990). Because the fluorescence of Indo-1 is quenched by Mn²⁺, this effect can be used to monitor Ca²⁺-channel activity in the plasma membrane. The reason why dye quenching was not observed after adding PGA in the presence of Mn²⁺ is because the concentration of free Mn²⁺ used was far too low. In our experience 5 mM Mn²⁺ is necessary to produce an effect with plant proto- plasts exhibiting significant plasma membrane Ca²⁺ chan- nel activity (Shacklock et al., 1992).

The increase in [Ca²⁺]_i we observed here was very slow and prolonged. We observed no change in [Ca²⁺]_i within the first 5-10 min of adding the oligogalacturonides but thereafter high levels (over 1 mM) were observed by 20-30 min. These kinetics have been considered unusual, since it is thought that cells cannot tolerate prolonged high con- centrations of [Ca²⁺]_i (Hepler and Wayne, 1985). Further- more, activation of the plasma membrane Ca²⁺-ATPase might be expected to follow such [Ca²⁺]_i increases, return- ing the [Ca²⁺]_i to its resting level. However, measurements on Commelina guard cells, Funaria protonemata and barley aleurone protoplasts after different stimulatory treatments (Gilroy et al., 1991; Hahm and Saunders, 1991; Gilroy and Jones, 1992), have shown that this pattern of [Ca²⁺]_i kinet- ics may be relatively common in plant cells. Further sup- port for a prolonged increase in [Ca²⁺]_i has been provided by the recent results of Mathieu et al. (1991). They showed that oligogalacturonides (DP 10-16) induced a gradual decrease in the Ca²⁺ concentration of the culture medium of tobacco cells, over about 1 h after exposure. A prolonged elevation in [Ca²⁺]_i might provide a mechanism for the effects of the extracellular signal to be easily reversible within a certain time period. Thus a high level of oli- gogalacturonides may have to be maintained to enable the response to continue whilst a transient exposure to the stim- ulus might have little effect or be simply ignored by the treated cells.

A possible mechanism to explain a long term elevation of [Ca²⁺]_i could be inhibition of the Ca²⁺-ATPases of the cell which normally maintain a low resting [Ca²⁺]_i level. In that case, oligogalacturonides might have to enter the cell by some form of receptor-coupled endocytosis and the lag period in response could represent the time for sufficient oligogalacturonide to enter the cell to initiate inhibition. Endocytosis of labelled oligogalacturonides has been demonstrated in cultured soybean cells (Horn et al., 1989). The specificity required by the oligogalacturonide does indicate the likely necessity of a specific binding protein although such receptors have not yet been identified (Ryan and Farmer, 1991).

Elicitation in vitro is generally obtained at elicitor con- centrations of 10-4 to 10-6 M. In our case, [Ca²⁺]_i was only significantly elevated at 10-4 M. A similar oligogalactur- onide concentration induced membrane depolarization in tomato leaf cells (Thain et al., 1990). This apparently high concentration, however, may be similar to the local con- centration of oligogalacturonides released from the cell wall at the site of pathogen invasion.

The ability of oligogalacturonides to induce [Ca²⁺]_i ele- vation in a size- and conformation-dependent way suggests that Ca²⁺ signalling may be involved in the triggering of plant defence genes. Indeed, phytoalexin accumulation (Jin and West, 1984) and lignification (Bruce and West, 1989) are induced by the same pectic oligomers that elevate [Ca²⁺]_i levels in our carrot protoplasts. Furthermore, Kurosaki et al. (1987) have shown that the addition to cul- tured carrot cells of the Ca²⁺ ionophore A23187, in the pres- ence of extracellular Ca²⁺, induces the synthesis of the phy- toalexin 6-methoxymellein.

Dimeric PGA and polygalacturonides with a DP >9 both increased [Ca²⁺]_i and PAL activity in carrot protoplasts; single chain PGA, polygalacturonides with a DP ≤9, and dimeric PGA in the presence of verapamil, did not. PAL is the first enzyme involved in the synthesis of isoflavonoid phytoalexins (Lamb et al., 1989). Our results, therefore, support the hypothesis that Ca²⁺ is a second messenger in the signal transduction pathway which links oligogalactur- onide elicitors with the response of phytoalexin synthesis. However, we found that the quantitative effects of the active elicitors on [Ca²⁺]_i and PAL activity were not the same (e.g. dimeric PGA caused the greatest elevation in [Ca²⁺]_i whilst polygalacturonides with a DP >9 induced the highest level of PAL activity). The reason for this is not known but may indicate that more than one signalling pathway is involved. In this respect, Kurosaki et al. (1987) found that the addition of a crude elicitor preparation to cultured carrot cells increased intracellular cAMP levels. However, whether cAMP is a second messenger in plant cells is still some- what contraversial (Trewavas and Gilroy, 1991).

In a previous paper we showed that cell wall elicitors from yeast and the filamentous fungus Gliocladium both induced increases in [Ca²⁺]_i (Knight et al., 1991). However, further evidence is still necessary to establish clearly whether oligogalacturonides or other elicitors which regu- late the expression of specific genes (e.g. in a defense response) are directly mediated by [Ca²⁺]_i.

We would like to thank Dr Steve C. Fry for critically reading the manuscript. Support for this work was provided by the Insti- tut pour l’encouragement de la Recherche Scientifique dans l’Industrie et l’Agriculture (IRSIA, Belgium) to J.M. and by the Agricultural and Food Research Council to N.D.R. and A.J.T.