Effects of short-term NaCl stress on calmodulin transcript levels and calmodulin-dependent NAD kinase activity in two species of tomato

Authors

  • Olivier Delumeau,

    1. Groupe de Biochimie et de Biologie Moléculaire Végétales – EA 917, UFR Sciences, Université d’Angers, 2 Boulevard Lavoisier, 49045 Angers Cedex 01, France
    Search for more papers by this author
  • Marie-Christine Morère-Le Paven,

    1. Groupe de Biochimie et de Biologie Moléculaire Végétales – EA 917, UFR Sciences, Université d’Angers, 2 Boulevard Lavoisier, 49045 Angers Cedex 01, France
    Search for more papers by this author
  • Françoise Montrichard,

    1. Groupe de Biochimie et de Biologie Moléculaire Végétales – EA 917, UFR Sciences, Université d’Angers, 2 Boulevard Lavoisier, 49045 Angers Cedex 01, France
    Search for more papers by this author
  • Danielle L. Laval-Martin

    1. Groupe de Biochimie et de Biologie Moléculaire Végétales – EA 917, UFR Sciences, Université d’Angers, 2 Boulevard Lavoisier, 49045 Angers Cedex 01, France
    Search for more papers by this author

Correspondence: Danielle Laval-Martin Email: Danielle.Laval-martin@univ-angers.fr

ABSTRACT

NAD kinase is thought to play an important role in the plant cellular responses to biotic and abiotic stress as one of the isoforms of the enzyme is activated by the Ca2+–calmodulin (CaM) complex. NAD kinase activity was measured after short-term NaCl stress applied to isolated cells from Lycopersicon esculentum, var. Volgogradskij (NaCl-sensitive tomato) and L. pimpinellifolium, acc. PE2 (NaCl-tolerant species). NAD kinase activity remained constant in the sensitive species, whereas a sharp decrease was observed in the tolerant one. After salt treatment, an induction of the calmodulin gene(s) was observed in the two species, together with a 30–50% decrease in ‘active’ CaM content, i.e. CaM able to activate purified NAD kinase, in L. pimpinellifolium. The decrease in NAD kinase activity could not, however, be fully explained by this decrease in active CaM content. A similar decrease in NAD kinase activity was also recorded with other ionic stresses and exposure to high temperatures, but not in the case of drought, exposure to low temperatures, hormonal (indole-3-acetic acid and abscisic acid) or H2O2 treatments. External Ca2+ certainly plays a role in the biochemical mechanism(s) leading to NAD kinase inhibition, while no role could be shown for intracellular Ca2+. In addition, after salt stress, a modification of the redox state of NAD kinase seems to be responsible for the inhibition of the enzyme.

INTRODUCTION

Salinity is one of the major limiting environmental factors in crop production. Under salt stress, plants have to cope with water stress imposed by the low external water potential and with ion toxicity due to accumulation inside the plant ( Greenway & Munns 1980). Although many studies have indicated that salt stress induces the expression of specific genes and metabolic modifications (for reviews see Bray 1993; Bohnert, Nelson & Jensen 1995) – particularly the synthesis of osmoprotectants such as glycine-betaine – the mechanisms of salt tolerance in plants are not fully understood ( Munns 1993). More recently, saline stress has been demonstrated to also induce an oxidative stress ( Hernandez et al. 1993 , 1995; Gueta-Dahan et al. 1997 ).

The NAD kinase (EC 2·1.7·23) catalysing the phosphorylation of NAD+ to NADP+ plays an important role in metabolism since its activity alters the pyridine nucleotide ratios (NADP(H)/NAD(H)) which have been known for a long time to modulate metabolic pathways ( Yamamoto 1969). While NAD kinase activity has been detected in many organisms ( McGuinness & Butler 1985), its physical structure and characteristics remain unclear. Two loci have been identified in Salmonella ( Cheng & Roth 1994), but the corresponding genes have not yet been cloned.

In plants, one isoform of NAD kinase has been shown to be dependent on activation by the Ca2+–calmodulin (CaM) complex ( Anderson et al. 1980 ). Because variations of the cytosolic Ca2+ concentration are known to be a common feature of the early events occurring in cells after the application of various stimuli, the ability of NAD kinase to be CaM-activated in plant cells makes this enzyme a potential primary target of these Ca2+ variations. For example, it was recently demonstrated that, after a pathogen attack, the NADPH content was increased through an enhancement of NAD kinase activity ( Harding, Oh & Roberts 1997). Studies have reported changes in NAD kinase activity when plants were facing abiotic stress ( Slaski 1989; Maciejewska 1990; Slaski 1990; Zagdanska 1990), but no study has ever dealt with possible changes in NAD kinase activity in the case of NaCl stress. On the other hand, evidence exists that CaM transcripts increase in plant cells after biotic or abiotic stresses, including salt stress ( Botella & Arteca 1994; for recent reviews see Snedden & Fromm 1998; Zielinski 1998).

In response to stress, changes of Ca2+ concentration and/or CaM concentration might affect NAD kinase activity, and thus provide the plant with the ability to adapt its metabolism to the newly imposed conditions. The aim of this work was to measure NAD kinase activity, levels of CaM transcripts and of active CaM (i.e. CaM able to activate NAD kinase) in isolated tomato cells exposed to NaCl for a short term. Two species were studied and compared: the salt-sensitive and domestic Lycopersicon esculentum, var. Volgogradskij, and the salt-tolerant L. pimpinellifolium, acc. PE2 ( Guerrier 1996).

MATERIALS AND METHODS

Cells, culture medium and treatments

Cell suspensions were obtained from white and friable root calli of tomato, and grown as described by Delumeau, Montrichard & Laval-Martin (1998) for Lycopersicon pimpinellifolium, acc. PE2. Cells were inoculated in 1·5 L of medium in Erlenmeyer flasks (5 L) placed on a gyratory shaker (100 rev min−1). After 7 d of culture in darkness at 23 °C, and 24 h before treatment, the cell suspension was distributed in 40 mL aliquots into Erlenmeyer flasks (100 mL). Supplementation with NaCl, CaCl2 or MgCl2 was performed with autoclaved stock solutions whose concentrations were 5 M, 100 m M and 100 m M, respectively; 1,2-bis(2-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid (BAPTA), caffeine, abscisic acid (ABA) and indole-3-acetic acid (IAA) 100 m M stock solutions were sterilized through a 0·22 μm filter; mannitol was directly added as a powder to the culture medium, only a few minutes being necessary for complete dissolution. Control cells received an equal volume of distilled water. Treatments were applied for 1 h to the cells, which were then collected by centrifugation, drained on a filter paper and lyophilized. Treatments were applied to two cell samples and all the experiments have been repeated three times.

Lyophilized cells were ground in a mortar and homogenized in a cold buffer (10 mg of dry matter per cm3), consisting of 50 m M Tris–Cl (pH 7·5), 10% glycerol, 1 m M dithiothreitol (DTT) and 1 m M ethylene glycol-bis(β-aminoethyl ether) N,N,N′,N′-tetraacetic acid (EGTA). After centrifugation (30 min at 39 000 g), the supernatant was used to determine NAD kinase and G6P dehydrogenase activities, as well as the active CaM levels.

Calmodulin gene(s) expression analysis

Cells were collected 15, 30 and 45 min, and 1, 2 and 3 h after salt stress, frozen with liquid nitrogen and kept at –80°C until extraction. Total RNA was extracted from L. esculentum or L. pimpinellifolium cells (2 g of fresh matter) in 2 mL 100 m M Tris–Cl (pH 8) and 2 mL phenol. After centrifugation (20 min at 10 000 ×g), the aqueous phase was washed with 1 vol chloroform. Total RNA was then precipitated by 1/10 vol Na-acetate (3 M) and 2 vol ethanol for 30 min at − 80 °C. After centrifugation (20 min at 10 000 ×g), the pellet was dissolved in 100 μL of water and treated with 5 units of RNAase-free DNAase I for 30 min at 37 °C to remove the contaminating DNA. DNAase I was then inactivated by phenol/chloroform (1/1 v/v) treatment. Total RNA was recovered by ethanol precipitation and the pellet was resuspended in 100 μL of water.

Reverse transcription polymerase chain reaction was performed on total RNA extracted from L. esculentum, var. Volgogradskij, with primers corresponding to the 5′ and 3′ ends of the coding region of the L. esculentum calmodulin, TOMCALM1LE, accession number M67472. The sequence of the upper primer was 5′-GCAGAGCAGCTGACGGAGGA-3′ and that for the lower primer was 5′-TCACTTGGCAAGCATCATAC-3′. The amplified fragment was cloned with a cloning kit (pGEMt-easy from Promega Corp.) according to the manufacturer’s recommendations. The nucleotidic sequence of the cloned fragment displayed 100% identity with the TOMCALM1LE. The probe used for Northern blot analyses was the amplification product of the cloned fragment inserted in the plasmid with the same primers.

Northern blotting was conducted according to Sambrook, Fritsch & Maniatis (1989). Total RNA (15 μg), separated on a 1·2% (w/v) formaldehyde–agarose gel, was transferred to a nylon membrane by an overnight capillary transfer in 20 × saline sodium citrate (SSC). The RNA was cross-linked on the membrane by 3 min exposure to ultraviolet. The filter was then pre-hybridized at 42 °C in a solution of 5 × SSC, 5 × Denhardt’s reagent, 0·5% sodium dodecyl sulphate (SDS) and 50% formamide for 1 h. Hybridizations were performed overnight at 42 °C in the same solution supplemented with the α-32P-dCTP-labelled probe using the random primer labelling Ready-To-Go system (Amersham Pharmacia Biotech Inc.). Filters were washed twice at 60 °C in 2 × saline sodium phosphate EDTA (SSPE) plus 0·1% SDS for 15 min, twice in 1 × SSPE plus 0·1% SDS for 15 min and then twice in 0·1 × SSPE plus 0·1% SDS at room temperature.

Purification of CaM-dependent NAD kinase

The purification procedure of the CaM-dependent NAD kinase has already been described ( Delumeau et al. 1998 ). Briefly, proteins of a soluble extract of Lycopersicon pimpinellifolium cells were precipitated with (NH4)2SO4 (48% saturation) and were centrifuged (10 000 g, 10 min). The pellet was homogenized in extraction buffer at pH 8 and loaded onto a DEAE-cellulose column equilibrated in the same buffer. The fraction containing the NAD kinase, which was not retained on the column, was supplemented with 2 m M Ca2+ and loaded on a bovine CaM–agarose column. NAD kinase was eluted with a buffer consisting of 50 m M Tris–Cl pH 8, 10% glycerol, 1 m M DTT and 5 m M EGTA.

Measurements

Enzyme assays

The NAD kinase assays were performed as described in Delumeau et al. 1998 (modified from Goto 1984 and Pou de Crescenzo et al. 1997 ). Briefly, the concentration of NADP+ produced by NAD kinase is measured by a recycling procedure using glucose-6-phosphate (G6P) and G6P dehydrogenase and an oxido-reductant system, phenazine etho-sulphate (PES)-3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT). The rate of reduction of MTT, monitored at 570 nm, is then proportional to the amount of NADP+.

The G6P dehydrogenase assays were performed as described previously ( Delumeau et al. 1998 ).

CaM content

The CaM contents of soluble extracts were determined by their ability to activate the CaM-dependent NAD kinase partially purified from tomato cells. The soluble extracts were heated for 3 min at 90–95 °C and centrifuged (10 000 g, 20 min). After an overnight dialysis of the supernatants against distilled water, various volumes of a given supernatant were added to 3 μL of purified CaM-dependent NAD kinase in the presence of 1 m M Ca2+. These assays of NAD kinase activation were performed in the presence of 0·1 mg mL−1 of bovine serum albumin, which stabilizes the NAD kinase activity.

Protein concentrations

Protein concentrations in the extracts were determined by the method of Bradford (1976) using bovine serum albumin as standard.

RESULTS

Effect of NaCl stress and other abiotic stresses on NAD kinase activity

Under control conditions of growth, the total NAD kinase activity measured in soluble extracts from L. esculentum cells (salt-sensitive species) was about 10-fold lower than that of L. pimpinellifolium (salt-tolerant species) ( Fig. 1). When NaCl (in the range 25–200 m M) was added to the culture medium of L. esculentum cells, NAD kinase activity remained constant whatever the salt concentration used ( Fig. 1a). By contrast, addition of 100 or 200 m M NaCl to the culture medium of L. pimpinellifolium cells induced an important and rapid decrease in NAD kinase activity ( Fig. 1b), reaching about 10% of the control value 30 min after the addition of NaCl; this low level of activity persisted for at least 24 h (experiment duration). When L. pimpinellifolium cells were treated with only 25 m M NaCl, NAD kinase activity remained constant during the course of the experiment, while an increase of about 25% was observed in control cells. In these experiments with L. pimpinellifolium cell cultures, the extent of the decrease in NAD kinase activity varied from 50% up to 95%; the salt concentration inducing the maximal decrease in NAD kinase activity also varied from 100 to 200 m M NaCl.

Figure 1.

NAD kinase activity measured in the salt sensitive L. esculentum (a) and the salt tolerant L. pimpinellifolium (b) cells after salt stress. The 8-day-old culture of isolated cells received 25 m M NaCl (□), 100 m M NaCl (◊) or 200 m M NaCl (▴), or distilled water (●). Cells were collected after the salt treatment at the time intervals indicated; the values are the means ± SE of three measurements.

Inhibition of NAD kinase activity by NaCl was tested in vitro using purified NAD kinase from L. pimpinellifolium. In the range 0–150 m M NaCl, the enzyme activity was not affected, indicating that a direct inhibition of NAD kinase by NaCl was not responsible for the decrease in NAD kinase activity observed under salt stress.

The hydric component of the salt stress was tested by comparing the effect of NaCl and of mannitol (twice the NaCl concentration to induce iso-osmotic stress) ( Fig. 2). In this experiment, the maximal effect of NaCl was observed for concentrations of 200 m M or more and resulted in a 90% decrease in enzyme activity, while only a 10% decrease was observed for the corresponding mannitol concentration (400 m M). The highest tested concentration of mannitol (800 m M) induced only a 40% decrease in enzyme activity. This result indicated that the enzymatic response was due to the ionic component of the salt. Figure 2 also displays the G6P dehydrogenase activity measured in saline samples, showing no significant variations.

Figure 2.

NAD kinase activity measured in L. pimpinellifolium cells after a NaCl stress (closed bars) or a mannitol stress (open bars) for 1 h. Mannitol is twice the NaCl concentration to produce equivalent osmotic stress. G6P dehydrogenase activity (●) was measured in the saline samples.

The decrease in NAD kinase activity could be due to the presence of Na+ or of Cl. Different combinations of ions were applied to L. pimpinellifolium cells. Whatever the ionic species (KCl, NaNO3 or KNO3) added to the culture medium, decreases in NAD kinase activity similar to that induced by NaCl stress were recorded ( Table 1).

Table 1.  Different stresses applied to L. pimpinellifolium cells and their effects on NAD kinase activity
StressNAD kinase activity (%)
  1. H2O2 (1 m M)

  2. 124

Control100
Ionic
NaCl (100 m M) 49
KNO3 (100 m M) 52
NaNO3 (100 m M) 43
KCl (100 m M) 52
Temperature
0 °C98
42 °C15
Hormonal
IAA89
ABA95
Oxidative

Other stresses were then tested, and a decrease in NAD kinase activity was also observed for high temperature (42 °C), while low temperature (0 °C) had no effect. Neither ABA (from 0·1 to 10 μM) nor IAA (from 60 to 480 μM) induced any change in NAD kinase activity. A slight increase (24%) in NAD kinase activity was recorded when L. pimpinellifolium cells were exposed to 1 m M H2O2 ( Table 1).

Effect of NaCl stress on calmodulin

The effects of salt treatment on the expression of Cam gene(s) are shown in Fig. 3. In the two species of tomato, the Cam transcript levels remained nearly constant in control cells during the course of the experiment. In contrast, increases in Cam transcript levels were observed after salinization: these increases were detectable after only 15 min of treatment; they appeared to be maximum at 30–45 min; then, after 2–3 h of treatment, the Cam transcript levels returned to control levels.

Figure 3.

Time course of NaCl-induced calmodulin gene expression. Cells receiving distilled water served as control and the remainder were 200 m M NaCl-stressed. At the time intervals indicated, cells were collected by centrifugation and immediately frozen in liquid nitrogen. Aliquots of 15 μg of total RNA were loaded on 1·2% formaldehyde–agarose gels, transferred to nylon membrane and fixed under UV. Standardization of RNA loading was controlled by the intensity of ethidium bromide coloration of the ribosomal RNA.

Because of these changes in Cam transcript levels, the ability of endogenous control and salinized CaM to activate purified NAD kinase were tested as described in Materials and methods. This experiment ( Fig. 4) indicated that after 1 h of treatment, and whatever the salt concentration added to the L. pimpinellifolium cells, the calmodulin extracted from salinized cells was about 35–50% less efficient at activating the purified NAD kinase than that from control cells. A similar result was obtained when the activation of purified NAD kinase was performed with CaM extracted from cells after a 6 h treatment.

Figure 4.

Activation of the purified NAD kinase by CaM extracted from control cells (●), from cells treated with 25 m M NaCl (□), with 100 m M NaCl (◊) and with 200 m M NaCl (▴). Cells were lyophilized after salt treatment and 20 mg of dry matter were extracted with the same volume of buffer; identical protein concentrations were obtained in control and salted samples; after heating for 3 min at 90 °C, coagulated proteins were removed by centrifugation (20 min at 10 000 ×g) and the supernatant was dialysed overnight against distilled water. Activation assays of the purified NAD kinase were performed with indicated volumes of the CaM-containing dialysate.

To test the hypothesis of a change in NAD kinase affinity towards CaM due to the salt treatment, NAD kinase extracted from control or salinized cells (100 m M NaCl for 1 h) was precipitated with 48% ammonium sulphate, which resulted in an almost total separation of the NAD kinase from the endogenous CaM. The NAD kinases from control or salinized cultures were then activated by the addition of increasing amounts of bovine CaM ( Fig. 5). Although the maximal activity of the NAD kinase extracted from treated cells was lower than that of the control NAD kinase, the affinities of the two enzymes towards bovine CaM were found to be similar (control K0·5 = 16·4 ng mL−1 versus salinized K0·5 = 15·1 ng mL−1). Therefore, NaCl treatment did not change the affinity of NAD kinase towards CaM.

Figure 5.

Activation of both control NAD kinase (●) and of NAD kinase extracted from salted cells (□) by bovine CaM. Proteins were extracted from control and 100 m M NaCl-treated cells of L. pimpinellifolium. After precipitation of the proteins with 48% saturation of ammonium sulphate, the two preparations of NAD kinase were activated with increasing amounts of a commercial bovine CaM.

Role of Ca2+ in the enzymatic response to salt stress

The effects of simultaneous additions of 200 m M NaCl and of various concentrations of CaCl2 in the culture medium were compared to the effects of CaCl2 alone. Table 2 indicates that the presence of Ca2+ (in the range 3–9 m M) partially counteracted the effect of NaCl stress on NAD kinase activity. In this range of Ca2+, the salt stress induced only a 25% decrease, against 70% in the control. This effect seemed to be specific to Ca2+, since 3 m M Mg2+ was much less efficient to counteract the effect of salt stress (not shown). It should be noted that the addition of a high Ca2+ concentration (18 m M) to the control medium induced a 40% decrease in NAD kinase activity and that, in this case, no additional decrease could be observed after NaCl addition. Conversely, when a decrease in Ca2+ concentration of the culture medium was realized by the addition of 10 m M BAPTA, a decrease in NAD kinase activity, comparable to or even more intense that the one induced by 200 m M NaCl, was observed. Thus, the Ca2+ activity in the culture medium appeared to be an important factor for the decrease in NAD kinase activity induced by salt stress.

Table 2.  Effect of Ca2+, BAPTA (Ca2+ chelator), Li+, La3+ and Gd3+ (Ca2+ channel blockers) and caffeine on the NAD kinase activity after salt stress in L. pimpinellifolium cells
 NAD kinase activity
 (nmol h−1 mg−1 protein)
 0 m M NaCl200 m M NaCl
  1. Caffeine (10 m M)

  2. 295 ± 11

  3. 86 ± 8

  4. Values are means ± SE (n = 3). In the culture medium, the initial Ca2+ concentration is 3 m M and Ca2+ was added with NaCl at the same time; Li+, La3+ and Gd3+ were added 1 h prior to the addition of NaCl. The NaCl stress was applied for 1 h.

Control condition321 ± 1093 ± 10
Ca2+ (3 or 9 m M) 315 ± 5258 ± 7
Ca2+ (18 m M) 187 ± 9185 ± 5
BAPTA (10 m M) 54 ± 4
Li+ (10 m M) 327 ± 1580 ± 4
La3+ (1 m M) 333 ± 588 ± 6
Gd3+ (1 m M) 336 ± 12108 ± 6

Since the intra-cytoplasmic Ca2+ concentration was shown to be increased by salt stress ( Lynch & Läuchli 1988; Lynch, Polito & Läuchli 1989), we then tested whether Ca2+ could act as a second messenger in the signal transduction leading to the decrease in NAD kinase activity during salt stress. Neither the addition of La3+ and Gd3+ (Ca2+ channel blockers) impairing the external Ca2+ entry, nor the addition of Li+ inhibiting the Ca2+ release from internal stores by the inositol-tris-phosphate system, 1 h before the salt stress application, could prevent the effect of salt stress on NAD kinase activity ( Table 2). Conversely, caffeine, used to induce the Ca2+ release from the internal stores, neither mimicked nor counteracted the NaCl effect on NAD kinase activity. Thus, changes in the intra-cytoplasmic Ca2+ concentration did not seem to be involved in the NAD kinase response to salt stress.

Reversibility of the inhibition

To test for the reversibility of the NaCl effect on NAD kinase activity, L. pimpinellifolium cells were first exposed to 200 m M NaCl for 1 h, then cells were transferred into control medium. The NAD kinase activity, initially reduced to 15% of the control value after the salt treatment, increased after transfer into control medium and stabilized to 60% of the control value after 3 h (result not shown).

The NAD kinase activity extracted from salinized cells could be restored by increasing DTT concentrations during the extraction ( Fig. 6). The strengthening of DTT concentration from 1 m M to 25 m M induced a fourfold increase in NAD kinase activity extracted from salinized cells. In contrast, with 25 m M DTT in the extraction buffer, the control NAD kinase activity increased only by 15% compared to its initial value.

Figure 6.

NAD kinase activity as a function of DTT concentration added to the extraction buffer. Proteins from control (●) and 200 m M NaCl-treated (▴) cells were extracted with the extraction buffer containing increasing concentrations of DTT, and NAD kinase activities were measured as described in Materials and methods.

DISCUSSION

This study mainly dealt with the alterations in NAD kinase activity in tomato cells after a short-term NaCl stress. In control conditions of growth, NAD kinase activity was found to be much higher in the salt-tolerant species (L. pimpinellifolium) than in the salt-sensitive one (L. esculentum). Slaski (1990) had shown a positive correlation between the level of NAD kinase activity and aluminium tolerance in several monocotyledons, but reported an opposite result for dicotyledons. In contrast to the increase in NAD kinase activity observed after aluminium stress in monocotyledons ( Slaski 1989), our study indicates that a NaCl treatment induces a large decrease in this enzymatic activity in the salt-tolerant tomato. A decrease in NAD kinase activity had already been reported in wheat after a water stress ( Zagdanska 1990), whereas an increase had been shown in Brassica napus after cold treatment ( Maciejewska 1990). The alterations of NAD kinase activity seem then to be dependent on the nature of the stress and on the species.

In L. pimpinellifolium, NAD kinase activity is 95% dependent on activation by the Ca2+–CaM complex ( Delumeau et al. 1998 ), and the decrease in activity could reach 95% after salt stress. It was therefore considered important to study the possible changes in CaM content. Clearly, salt stress induced the expression of Cam gene(s) in the two species of tomato. These results are in agreement with data reported by Botella & Arteca (1994) on whole plants of Vigna radiata submitted to salt treatment, except that the mRNA level peaked 6 h after the addition of NaCl in Vigna radiata, whereas it peaked only 30–45 min after the addition of NaCl in tomato, which could be explained by the use of isolated cells in our study. Many studies have reported Cam expression in response to various stimuli ( Zielinski 1998) but none have studied NAD kinase activity in parallel. Our study shows that, unexpectedly, although an induction of Cam gene(s) did occur in tomato cells under salt stress, NAD kinase activity was unchanged or decreased. This suggests that either (i) Cam expression is not accompanied by an increase in CaM protein as suggested by Roberts & Harmon (1992), or (ii) supplementary calmodulins have no effect on NAD kinase activity. The last hypothesis is supported by determination of the in vivo CaM concentration as in the μM range ( Ling & Assman 1992) and thus in large excess considering the Kd for the interaction between CaM and NAD kinase which is in the nM range ( Harmon, Jarrett & Cormier 1984; Delumeau et al. 1998 ). Nevertheless, as pointed out by Zielinski (1998), CaM concentration could be considered to be at subsaturating levels compared with the concentration of its binding sites on the set of CaM-regulated proteins. Whether the in vivo CaM concentration is sufficient or not to fully activate the NAD kinase remains under debate, but this work demonstrates clearly that, after salt stress, the CaM was qualitatively modified regarding its capacity to activate the NAD kinase. Thus, salt stress could affect and modify the existing pool of CaM proteins or induce the synthesis of calmodulin isoform(s) unable to activate the NAD kinase. This last hypothesis is supported by the existence of a CaM isoform unable to activate the NAD kinase in soybean ( Lee et al. 1995 , 1997). Therefore, the results presented in Fig. 4 could also be explained by competition for NAD kinase binding sites due to the presence of this ‘inactive’ calmodulin. It would be interesting to know whether the cDNA encoding a calmodulin-related protein isolated from Dunaliella salina after salt stress ( Ko & Lee 1996) corresponds to a CaM able or unable to activate the NAD kinase.

One of the main results of this paper is the rapid (30 min) decrease in NAD kinase activity in L. pimpinellifolium cells submitted to salt stress, which could not be entirely explained by alteration of the calmodulin. Such a decrease in NAD kinase activity was unexpected in response to salt stress. A possible explanation would be that, as ATP is a substrate of NAD kinase for NADP production, the enzyme inhibition would favour the use of ATP for possible new adaptation mechanisms. Salt stress is often divided into two components: osmotic and ionic stress ( Munns 1993). Here, the cellular response was demonstrated to be clearly associated with an ionic stress and not with an osmotic stress. One important element to understand how this NAD kinase response occurs in the salinized cells is revealed by the experiment of simultaneously adding calcium and NaCl. Enhancement of plant salt tolerance by calcium has been known for a long time ( LaHaye & Epstein 1969) and new advances have been recently reviewed ( Bressan, Hasegawa & Pardo 1998). Our experiments clearly indicate that a relationship exists between the external Ca2+ availability and the relative extent of NAD kinase inhibition. Displacement of the Ca2+ from the plasma membrane by NaCl had already been proposed to be a primary response to salt stress, and increases in cytosolic Ca2+ have also been recorded ( Cramer, Läuchli & Polito 1985; Lynch, Cramer & Läuchli 1987; Lynch & Läuchli 1988; Lynch, Polito & Läuchli 1989). A decrease in Ca2+ activity through the presence of ions such as Na+ or K+ in the culture medium has been already documented ( Cramer & Lauchli 1986). This could explain the decrease in NAD kinase activity observed whatever the nature of the ionic stress applied to the cells. On the other hand, experiments involving (i) Ca2+ release from the internal store (by the addition of caffeine), or blockage of this release by (ii) Li+ or (iii) calcium channel blockers, did not demonstrate that Ca2+ could act as a second messenger in the signal transduction pathway leading to inhibition of the NAD kinase in L. pimpinellifolium cells. Moreover, while cold, ABA and IAA have been shown to induce changes in cytosolic Ca2+ concentrations ( Felle 1988; Gilroy et al. 1991 ; Knight et al. 1991 ; Knight, Trewavas & Knight 1996), no change in NAD kinase activity was observed in this study. In Arabidopsis thaliana, Ca2+ transients of similar amplitude and duration were found in response to both mannitol and salt ( Knight, Trewavas & Knight 1997). If the same is true for tomato, our results suggest the existence of a mechanism independent from cytosolic Ca2+ for the inhibition of NAD kinase activity.

The last point of this work examined the biochemical changes possibly affecting the NAD kinase from L. pimpinellifolium after salt stress. Neither direct inhibition by NaCl nor a change in affinity for CaM were involved in the decrease in NAD kinase activity. Without rejecting a possible production of some form of tight-binding inhibitor of the CaM-dependent NAD kinase activity, the very interesting result of this study, reported in Fig. 6, is that the redox state of the NAD kinase seems to be an important modulator of its activity. Studies have already demonstrated that salt stress induced changes in the anti-oxidant enzymatic defences ( Comba, Benavides & Tomaro 1998; Gosset, Millhollon & Lucas 1994), and that the over-expression of such enzymes conferred ( Roxas et al. 1997 ), or not ( Van Camp et al. 1996 ), a tolerance to NaCl. Since NADPH is used for detoxifying the active oxygen species, the control of the activity of NAD kinase through its redox state could be relevant. We are currently testing the hypothesis of such a possible regulation of this enzyme by modifying the thiol groups of the purified protein.

ACKNOWLEDGMENTS

This work was supported by the European Project SALTO AIR 3-CT94-1508 ‘New strategies for improving salt stress tolerance in crop plants’ initiated by Professor G. Guerrier, University of Orléans, France. The authors are grateful to Dr Magnin and Professor Rees for correcting the English in this manuscript.

Ancillary