INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plasma membrane depolarization, external alkalinization, and the monitoring of ion fluxes have proven to be valuable specificity indicators of Nod factors (Erhardt et al., 1992; Felle et al., 1995, 1998; Kurkdjian, 1995), N-acetylchitooligosaccharides (Kuchitsu et al., 1993), and different kinds of other elicitors (Boller, 1995). So far, studies to demonstrate a common perception site for N-acetylchitooligosaccharides and lipochitooligosaccharides (Nod factors) comprising the same number of glucosamine residues have not been conclusive (Ryan and Farmer, 1991). Although tetramers and pentamers of the N-acetylchitooligosaccharides at 10⁷ M had very little effect on membane potential and external pH of alfalfa (Medicago sativa; Felle et al., 1995, 1998), in rice, Kuchitsu et al. (1995, 1997) demonstrated that chitoheptaose was in fact very effective in triggering membrane depolarization and ion fluxes, as well as intracellular pH changes. These agents, initiating either symbiotic responses or defense reactions, have the capacity to evoke these early effects. However, it is not understood how plants (or root hairs in our study) distinguish between their symbiotic partners and pathogenic organisms upon encounter or at which stage of the signal transduction this discrimination is accomplished. In an attempt to answer these questions, the functional perception of N-acetylchitooligosaccharides and Nod factors was compared using ion-selective microelectrodes intra- and extracellularly. By following up the propagation of the elicited responses in alfalfa root hairs at different stages, we demonstrate that the discrimination of symbiotic and defense-related signals occurs at the perception site as well as downstream during their transduction by Ca²⁺ and/or pH.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Differential Depolarization Response of Alfalfa to N-Acetylchitooligosaccharides and to NodRm-IV(C16:2,S)

Transient plasma membrane depolarizations, as a measure for early responses of plants to Nod factors or elicitors, have been found to be powerful indicators of symbiotic and defense reactions (e.g. Mathieu et al., 1991; Erhardt et al., 1992; Felle et al., 1995; Kurkdjian, 1995; Kikuyama et al., 1997). Figure 1 compares the initial maximal depolarization responses of alfalfa root hair plasma membranes with NodRm-IV(C16:2,S) and to N-acetylchitooligosaccharides of different chain lengths. Chitotetraose [(GlcNAc)₄], the glucosamine backbone of NodRm-IV(C16:2,S), did not depolarize the root hairs at 10⁷ M, whereas at this concentration the response to NodRm-IV(C16:2,S) was already saturated. Significant depolarizations started to emerge in the presence of 10⁹ M chitooctaose [(GlcNAc)₈], 10⁸ M chitoheptaose [(GlcNAc)₇], and 10⁷ M chitohexaose ((GlcNAc)₆); 10⁶ M chitooctaose evoked about one-half of the saturation value induced by NodRm-IV(C16:2,S).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Dose-effect relationships of the maximal depolarizations of alfalfa root hairs induced by NodRm-IV(C16:2,S) or (GlcNAc)n, as indicated. Mean values ± SE calculated from at least three experiments.

In Figure 2 the kinetics of the depolarization responses to (GlcNAc)₈ and (GlcNAc)₇ are compared with those to NodRm-IV(C16:2,S). (GlcNAc)₈ caused a slow depolarization of the root hairs, which peaked after about 5 min and recovered with a hyperpolarization. Following pre-incubation at 10⁷ M (GlcNAc)₈, the subsequent joint additions of NodRm-IV(C16:2,S) and (GlcNAc)₈ resulted in reduced depolarizations, whereby the Nod factor response was less affected by the presence of the chitooligosaccharides than vice versa. Thus taking the mean depolarizations for NodRm-IV(C16:2,S) and (GlcNAc)₈ from Figure 1 as a basis, pre-incubation with (GlcNAc)₈ reduced the response to NodRm-IV(C16:2,S) by about 25% (Fig. 2A) and pre-incubation with NodRm-IV(C16:2,S) reduced the response to (GlcNAc)₈ by 50% (Fig. 2B). On the other hand, when both agents were added jointly without pre-incubation, in four out of eight experiments the depolarization was enhanced by up to 20%, indicating different perception sites (Fig. 2C).

View larger version (27K): [in this window] [in a new window]   Figure 2.   Membrane potential response of alfalfa root hairs to NodRm-IV(C16:2,S), (GlcNAc)7, or (GlcNAc)8 at the indicated molar concentrations added either jointly or successively, as indicated. A, Pre-incubations with (GlcNAc)7 and (GlcNAc)8, followed by joint additions of NodRm-IV(C16:2,S) and (GlcNAc)8 or (GlcNAc)7. B, Pre-incubation with NodRm-IV(C16:2,S), followed by joint additions of NodRm-IV(C16:2,S) and (GlcNAc)7 or (GlcNAc)8. C, Joint addition of NodRm-IV(C16:2,S) and (GlcNAc)8 without pre-incubation. Representative kinetics of at least five equivalent experiments each.

Chitotetraose and chitohexaose had no significant effect on the NodRm-IV(C16:2,S) response (data not shown), but obviously affected the response to chitooctaose (Fig. 3). Although evoking no or minor depolarizations on their own, (GlcNAc)₄ and (GlcNAc)₆ markedly reduced the depolarization response to chitooctaose. When added at equal concentrations (10⁷ M) the mean depolarization response to (GlcNAc)₈ of 10.5 mV (see Fig. 1) was reduced to 8 mV (approximately 76%) by (GlcNAc)₄ or (GlcNAc)₆, to 5 mV (approximately 47%) in the presence of 10⁶ M (GlcNAc)₄, and to 4 mV (approximately 38%) in the presence of (GlcNAc)₆. Joint additions of (GlcNAc)₄ plus (GlcNAc)₈ or (GlcNAc)₆ plus (GlcNAc)₈ without pre-incubation did not result in an enhancement, but in a response reduction.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Effect of chitooctaose [(GlcNAc)8] on the membrane potential of alfalfa root hairs as affected by pre-incubation and joint additions with chitotetraose [(GlcNAc)4] or chitohexaose [(GlcNAc)6] at the indicated molar concentrations. Representative kinetics of at least four equivalent experiments each.

External Alkalinization

Apart from the depolarization, external alkalinization is also a typical response to Nod factors (Felle et al., 1998) and to elicitors (Felix et al., 1993). Figure 4 shows that the pH responses to (GlcNAc)₈ and NodRm-IV(C16:2,S) essentially reflect those of the depolarization kinetics. Whereas the pH response to NodRm-IV(C16:2,S) is a transient alkalinization of 0.3 to 0.4 units (Fig. 4B), (GlcNAc)₈ at low concentrations (10¹⁰ M) hardly affected external pH, whereas at higher concentrations (10⁸ and 10⁷ M) the external space first alkalized and then acidified. Pre-incubation with (GlcNAc)₈ reduced the alkalinization response to 10⁷ M NodRm-IV(C16:2,S) in a concentration-dependent manner (Fig. 4A), but not when both agents were added jointly without pre-incubation (Fig. 4B, lower curve). Moreover, pre-incubation with NodRm-IV(C16:2,S) reduced the initial alkalinization response by (GlcNAc)₈, but not the acidification (Fig. 4B, upper curve).

View larger version (15K): [in this window] [in a new window]   Figure 4.   pH of the alfalfa root hair space responding to (GlcNAc)8 and to NodRm-IV(C16:2,S), added on their own or jointly at the indicated molar concentrations. A, Response to 107 M NodRm-IV(C16:2,S) plus (GlcNAc)8 following pre-incubation with different concentrations of (GlcNAc)8. B, Effect of (GlcNAc)8 plus NodRm-IV(C16:2,S) following pre-incubation with NodRm-IV(C16:2,S) (top curve) and joint addition of NodRm-IV(C16:2,S) and (GlcNAc)8 without pre-incubtation (bottom curve). Measured with a blunt pH-sensitive microelectrode; see "Materials and Methods." For the sake of alignment some traces were interrupted. The different timing in addition of the second compound NodRm-IV(C16:2,S) had no influence on the response. Time and pH scale apply for A and B. Representative kinetics of at least five equivalent experiments each.

Ion Fluxes from and into the Root Hair Space

We have recently demonstrated that in alfalfa the temporal sequence of the early responses to NodRm-IV(C16:2,S) started with a Ca²⁺ influx, followed by an anion and a further delayed K⁺ efflux (Felle et al., 1998). We argued that the anion loss from the cells was triggered by the increased cytosolic Ca²⁺ activity and that the anion efflux was the most likely cause of the depolarization and possibly of the external alkalinization. Figure 5 shows that (GlcNAc)₈ induced similar responses, although with less pronounced Cl and K⁺ fluxes. The Ca²⁺ response, however, differed considerably from that observed with Nod factors. Whereas the response to NodRm-IV(C16:2,S) was a slow Ca²⁺ loss from the root hair space never exceeding 0.1 pCa, the response to (GlcNAc)₈ was fast and transient with peaks 0.3 to 0.4 pCa above the starting activity, indicating a substantial Ca²⁺ influx.

View larger version (23K): [in this window] [in a new window]   Figure 5.   Ca2+, K+, and Cl activities measured in the root hair space of alfalfa with ion-selective microelectrodes responding to 107 M (GlcNAc)8; see "Materials and Methods." For comparison the membrane potential response (Em) and the Ca2+ response to NodRm-IV(C16:2,S) are given. Representative kinetics of at least four equivalent experiments each.

Cytosolic pH and Ca²⁺ Activity

In an earlier work we demonstrated that the cytoplasmic pH of alfalfa was alkalized by Nod factors, whereas chitotetraose had no such effect (Felle et al., 1996). As shown in Figure 6A, (GlcNAc)₈ elicited a completely different response. Following a weak initial alkalinization, the cytosolic pH slowly acidified by 0.2 to 0.3 pH unit within 30 to 45 min. Recovery of the cytosolic pH within the first hour after incubation with (GlcNAc)₈ was not observed. Following a 50 min pre-incubation of the root hairs with (GlcNAc)₈, addition of 10⁷ M NodRm-IV(C16:2,S) caused a rapid alkalinization by 0.1 pH (accompanied by a reduced depolarization), showing that NodRm-IV(C16:2,S) indeed caused a response in the inverse direction. As shown in Figure 6B (lower curve), addition of the Nod factor first yielded an initial alkalinization, which, as soon as chitooctaose was given, turned into an acidification. When both agents were added jointly the typical Nod factor alkalinization was observed, which after about 20 min spontaneously turned into an acidification, typical for the chitooctaose response (Fig. 6B, upper curve).

View larger version (18K): [in this window] [in a new window]   Figure 6.   Cytosolic pH (pHc) of alfalfa root hairs measured with a pH-sensitive microelectrode and membrane potential (Em). A, Simultaneous recordings of membrane potential and cytosolic pH responding to 107 M (GlcNAc)8 first and then to NodRm-IV(C16:2,S). B, Cytosolic pH responding to joint addition of (GlcNAc)8 and NodRm-IV(C16:2,S) at 107 M each (top curve), or to NodRm-IV(C16:2,S) first, followed by (GlcNAc)8 (bottom curve). Representative kinetics of at least six equivalent measurements each.

In contrast to the differential pH responses, NodRm-IV(C16:2,S) and (GlcNAc)₈ elicited a Ca²⁺ increase, albeit with kinetics of different shapes and amplitudes. As shown in Figure 7, (GlcNAc)₈ caused a rapid increase in cytosolic free [Ca²⁺], most of which recovered within 10 min (Fig. 7, curves a and b) to a level approximately 0.1 pCa below the control. The loss of Ca²⁺ from the root hair space followed a similar, but inverse kinetics. The increase in Ca²⁺ activity elicited by NodRm-IV(C16:2,S) was somewhat slower and less pronounced, but recovered to about the same level as observed with (GlcNAc)₈.

View larger version (22K): [in this window] [in a new window]   Figure 7.   Cytosolic free [Ca2+] of alfalfa root hairs measured with a Ca2+-selective microelectrode (see "Materials and Methods") responding to (GlcNAc)8 (a and b) or NodRm-IV(C16:2,S). For comparison, the Ca2+ response to (GlcNAc)8 of the root hair space is shown (see text). a and b are the same conditions, demonstrating the variability of the response from two different cells. Representative kinetics of at least four similar experiments. Ca2+ microelectrodes measured 20 to 50 µm behind the root hair tip.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In the rhizosphere the legume roots or root hairs encounter numerous signal molecules, such as Nod factors, that induce nodule organogenesis, as well as a variety of cell wall fractions from other organisms, some of which are chitooligosaccharides-eliciting defense reactions. It can be assumed that signal molecules of both groups may be present at the same time and site on the plant roots. Thus the plants are challenged to respond to both signals without losing the ability to react to either one or the other. In plant defense there are two typical phases: phase 1, an immediate reaction that involves responses like oxidative burst and/or ion fluxes, reactions that probably serve to buy time by making the conditions unfavorable for the potentially attacking microorganism and phase 2, as a long term reaction, involves an array of measures to actively fight an intrusion. The data presented here would thus be characteristics of the phase 1.

The Early Events

Nod factor-induced plasma membrane depolarization and external alkalinization are two features of the same event, namely the activation of anion channels, triggered by incresased cytosolic Ca²⁺, much of which probably has entered the cells from outside (Felle et al., 1998, 1999a; Figs. 5 and 7). Anions rapidly leaving the cell and depolarize the plasma membrane; however, a fraction thereof, namely the organic acid anions, alkalize the external space by binding protons (H.H. Felle, É. Kondorosi, Á. Kondorosi, and M. Schultze, unpublished data). Efflux of K⁺ starts after its driving force has changed direction during membrane depolarization; it thus compensates the negative charges and initiates repolarization. Although the comparison of the fluxes evoked by NodRm-IV(C16:2,S) or (GlcNAc)₈ indicates similar cascades of events, analysis of the kinetics reveals some differences. NodRm-IV(C16:2,S) and (GlcNAc)₈ transiently depolarized the plasma membrane of alfalfa root hairs and transiently alkalized the root hair space (Fig. 2). These responses differ in that their recoveries of the Nod factor responses are incomplete, whereas in the presence of (GlcNAc)₈ the plasma membrane hyperpolarized and the root hair space finally acidified, indicating an increase in the activity of the plasma membrane proton pump. We suggest that this stimulation is induced by the cytosolic acidification (Fig. 6), whereas the Nod factor-induced cytosolic alkalinization (Felle et al., 1996) might reduce the pump activity temporarily to some extent, resulting in only partial recovery of membrane potential and external pH (Felle et al., 1998; Fig. 4). Felix et al. (1998) have proposed that the transient nature of the external alkalinization observed in suspension-cultured tomato cells might be due to desensitization of the perception system. Because the (GlcNAc)₈-induced alkalinization was followed by a substantial acidification, this argument would not hold for the observations in our system.

Different Perception Sites for Nod Factors and N-Acetylchitinoligosaccharides

The responses to NodRm-IV(C16:2,S) and (GlcNAc)₈ show signal interferences when added jointly following pre-incubation (Figs. 2 and 3). Because depolarization and external pH changes are delayed secondary responses, and hence are neither direct membrane effects nor immediate reactions of one or more putative receptors to the binding of the respective ligands, any reduction of these responses can likewise not be attributed to binding interference, but must have occurred downstream of perception. Although chitotetraose, structurally closest to NodRm-IV(C16:2,S) and the most effective Nod factor in alfalfa, neither depolarized (Figs. 1 and 3) nor interfered significantly with any responses to various Nod factors (Felle et al., 1995), chitoheptaose (Kuchitsu et al., 1997) and chitooctaose did elicit such effects (Figs. 1-3). The observation that jointly added Nod factor and chitooctaose caused response amplification (Fig. 2C) suggests different peception sites and cannot be explained by the increase in total substrate concentration, as doubling the Nod factor concentration from 10⁷ M to 2 × 10⁷ M did not increase the depolarization (Fig. 1). The explanation of such a "stimulatory" effect is difficult, but could be due to an incomplete activation of the involved ion channels by the individual compounds or could result from an activation of different channels. Thus two suggestions follow: (a) Nod factors and oligochitins are recognized by different sites, and (b) the reduced responses could be due to desensitization, but very likely indicate interference downstream of the perception sites; it could be that signal propagation is mediated through joint elements such as G-proteins (Hebe et al., 1999; Pingret et al., 1998) or protein phosphatases (H.H. Felle, É. Kondorosi, Á. Kondorosi, and M. Schultze, unpublished data). Moreover, since cytosolic pH and Ca²⁺ are affected by either compound (Figs. 6 and 7), and apparently both are involved in the subsequent activation of the ion fluxes across the plasma membrane, we suggest that this may be one bottleneck responsible for the mutual interferences observed.

Changes in Cytosolic pH and Ca²⁺ Activity Indicate Signal Chain Forking

Although the early ion flux responses to NodRm-IV(C16:2,S), (GlcNAc)₈, or other elicitors (Nürnberger et al., 1994; Boller, 1995) to some extent follow a similar pattern, the changes in cytosolic pH and Ca²⁺ activity elicited by NodRm-IV(C16:2,S) and (GlcNAc)₈ were drastically different.

Cytosolic pH

Cytosolic Ca²⁺ Activity

6

6

Model Conceptions

Figure 8 shows a model that summarizes the observations presented, but also includes our previous findings (Felle et al., 1995, 1996, 1998, 1999a, 1999b). There are two caveats: (a) Although we have good indications of different perception sites for (GlcNAc)_4-8 and Nod factors, in spite of reports on high-affinity binding sites for N-acetychitooligosaccharides (Shibuya et al., 1996), little is known about the putative receptors, and (b) whereas cytosolic alkalinization may indeed be the key event to prevent activation of defense reactions, the changes in cytosolic Ca²⁺ activity and acidification must be regarded as necessary, but not sufficient events to activate them. It is clear that molecular work will have to be done to find out whether the treatments presented in this study will in fact have the impact on the cells that we propose.

View larger version (51K): [in this window] [in a new window]   Figure 8.   Model summarizing the findings described in this and earlier work (see text). It comprises the different stages of the discrimination of the Nod factor from the N-acecetylchitooligosaccharide signal: perception by different sites, changes in cytosolic Ca2+ activity, and pH changes, the latter of which are thought to play a key role in the activation of defense genes (see text).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

General Assay Conditions

Seeds of alfalfa (Medicago sativa subsp. sativa cv Sitel) were surface sterilized and prepared for treatment as described (Felle et al., 1995). Intact 2-d-old seedlings were fixed with candle wax on the bottom of a chamber that was constantly perfused with a solution containing (in millimolars) 0.5 MES [2-(N-morpholino)-ethanesulfonic acid]/Tris (mixed to pH 6.9), 1 KCl, 0.1 NaCl, and CaCl₂ each; conditions different from these are given in the figure legends. (GlcNAc)₇ and (GlcNAc)₈ were kindly provided by N. Shibuya and E. Minami (Tsukuba, Japan), whereas the other N-acetylchitooligosaccharides were purchased (pure grade, Sheikagu, Tokyo). NodRm-IV(C16:2,S) from Rhizobium meliloti (Schultze et al., 1992) was prepared from aqueous stock solutions of 1 mM. Prior to the tests the seedlings were incubated in the perfusion solution for approximately 1 h.

Ion-Selective Microelectrodes

The electrical set-up for the impalement of root hairs, the fabrication of ion-selective microelectrodes, and their intracellular application have been described previously (Felle and Bertl, 1986; Felle, 1996; Felle et al., 1998). The preparation of the ion-selective electrodes for extracellular use differed in that the tip was 2 to 5 µm in diameter, blunt, and heat polished. To give the sensor in the tip enough firmness to stay in place for extended use, the respective sensor cocktail (Fluka, Milwaukee, WI) was dissolved in a mixture of polyvinylchloride/tetrahydrofuran (40 mg/mL) at a ratio of 30:70 (v/v). After evaporation of the tetrahydrofuran, the remaining firm gel was topped up with the undiluted sensor cocktail, followed by the reference solution required for the respective ions. After equilibration, these electrodes gave stable responses for at least 2 weeks when stored in a dry chamber. The electrode tips were placed 10 µm from the root surface. To compare changes in ion concentrations occurring at the same location directly, electrodes were combined in double-barreled tips. The electrodes were connected to a high-impedance amplifier (FD 223, WP-Instruments, Sarasota, FL) that simultaneously measured and subtracted the signals coming from the ion-selective electrode and the voltage reference. Signals were recorded on a chart recorder (L 2200, Linseis, Germany).