Cellular Response of pea plants to cadmium toxicity: cross-talk between reactive oxygen species, nitric oxide and calcium

and cross-talk under toxicity ABSTRACT Cadmium (Cd) toxicity has been widely studied in different plant species, however the mechanism involved in its toxicity as well as the cell response against the metal have not been well established yet. In this work, using pea plants we have studied the effect of Cd on antioxidants, reactive oxygen species (ROS) and nitric oxide (NO) metabolism of leaves by using different cellular, molecular and biochemical approaches. The growth of pea plants with 50 µM CdCl 2 affected differentialy the expression of superoxide dismutase isozymes at both transcriptional and post-transcriptional level, giving rise to an SOD activity reduction. The CuZn-SOD down-regulation was apparently due to the calcium (Ca) deficiency induced by the heavy-metal. In these circunstances, the overproduction of ROS (H 2 O 2 and O 2.- ) could be observed in vivo by confocal laser microscopy, mainly associated with vascular tissue, epidermis, and mesophyll cells and production of superoxide radicals was prevented by exogenous Ca. On the other hand, the nitric oxide synthase (NOS)-dependent NO production was strongly depressed by Cd and the treatment with Ca prevented this effect. Under these conditions, the pathogenesis-related proteins (PRPs) PrP4A, chitinase, and the HSP 71.2, were up-regulated; probably to protect cells against damages induced by Cd. The regulation of these proteins could be mediated by jasmonic acid and ethylene, whose contents increased by Cd treatment. A model is proposed for the cellular response to long-term Cd exposure consisting in a cross-talk between Ca, ROS and NO. acid, and ethylene, was studied, as well as the expression of the antioxidative enzymes superoxide dismutases and pathogen-related proteins of pea leaves. All these pieces of information are very important to understand the mechanisms involved in the defense of plant cells against different types of abiotic stress. 1). These results suggest that, under Cd stress conditions, Mn-SOD is regulated at transcriptional level, while Cu,Zn-SOD and Fe-SOD are regulated at both transcriptional and postranscriptional level. To investigate the involvement of Ca in the Cd-dependent regulation of SOD expression, pea plants were supplemented with Ca(NO 3 ) 2 during the Cd treatment, and the CuZn-SOD transcript levels were analyzed. The exogenous Ca supply reversed the effect of Cd on CuZn-SOD expression, reaching the same levels as in that develop during


INTRODUCTION
Cadmium (Cd) is a toxic element whose presence in the environment is mainly due to industrial processes and phosphate fertilizers, and then is transferred to the food chain (Pinto et al., 2004). Cd is rapidly uptaken by plant roots and then can be loaded into the xylem for its transport into leaves. Most plants are sensitive to low Cd concentrations which inhibit plant growth as consequence of alterations in the photosynthesis rate and the uptake and distribution of macro and micronutrients (Lozano-Rodríguez et al., 1997;Sandalio et al., 2001;Benavides et al., 2005). It is known that the content of polyvalent cations can be affected by the presence of Cd through competition for binding sites of proteins or transporters (Gussarson et al., 1996). Thus, Cd produced a decrease of calcium (Ca) content in different plant species (Gussarson et al., 1996;Sandalio et al., 2001). Ca is involved in the regulation of plant cell metabolism and signal transduction (Yang and Poovaiah, 2002;Rentel and Knight, 2004) and modulates cellular processes by binding proteins such as calmodulin (CaM), which in turn regulates the activity of target proteins (Roberts and Harmon, 1993).
Cd can be detoxified by phytochelatins, whose synthesis is induced by Cd and other metals and is accompanied by a decrease in the concentration of glutathione (Zenk 1996). In addition, Cd produces disturbances in the plant antioxidant defences producing an oxidative stress (Somashekaraiah et al., 1992;Shaw, 1995;Gallego et al., 1996;Sandalio et al., 2001;Dixit et al., 2001;Schützendübel et al., 2001;Romero-Puertas et al., 2002;2007;Rodríguez-Serrano et al., 2006). Recently, the cellular production of reactive oxygen species (ROS) in leaves from pea plants under Cd stress has been reported (Romero-Puertas et al., 2004). ROS were detected in epidermic, transfer and mesophyll cells, being plasma membrane the main source of ROS, although mitochondria and peroxisomes were also involved (Romero-Puertas et al., 2004). Concerning the mechanism of ROS production, Cd does not participate in Fenton-type reactions (Stoch and Bagchi, 1995) but can indirectly favour the production of different ROS (H 2 O 2 , O 2 .-, ·OH) by unknown mechanisms, giving rise to an oxidative burst

Effect of Cd on macro-and micro-nutrient content of pea leaves
In a previous work it was found that Cd induced a strong reduction in the Ca content of leaves (Sandalio et al., 2001). Ca is an important signalling component in biotic and abiotic stress and disturbances in its content has been associated with toxicity by Cd, Zn, Cu, or Al (Kinraide et al., 2004), althouhgt the mechanisms involved are not well known yet. The growth of pea plants in full-nutrient solutions containing 50 µM CdCl 2 for 15 days produced an accumulation of Cd in the leaves of about 13 µg per g of dry weight (Table S1). In these conditions, a reduction in the content of the following nutrients was observed: Ca (27%), Cu (30%), Fe (19%), Mn (47%), Mg (20%) and Zn (41%). On the contrary, Cd produced a 3-fold increase in the S content, while the Na contents were not affected by the heavy-metal treatment (Table S1).
To get more insights into the role of the Cd-induced Ca deficiency in the heavy-metal toxicity, plants were suplemented with 10 mM Ca(NO 3 ) 2. The addition of Ca to the nutrient solution produced an increase in the content of this element in both control and Cd-treated plants, and a 30% reduction in the Cd accumulation in the leaves of Cd-treated plants, without affecting the content of the remaining elements, except the Mn which increases and Mg that decreases slightly, in control plants (Table S1).

Response of superoxide dismutases to Cd
The growth of pea plants for a long period of time with 50 µM CdCl 2 produced reductions in the activities of Mn-SOD, Fe-SOD and Cu,Zn-SOD of 60, 80 and 90%, respectively control plants (Fig 1). The enzymatic analysis of CuZn-SOD showed that this activity was also recovered by Ca (data not shown).

ROS and NO in the plant response to Cd
Confocal laser scanning microscopy (CLSM) has been widely used to study fluorescent probes distribution in fixed and living plant tissues (Fricker and Meyer, 2001;Sandalio et al., 2008). To image ROS and NO accumulation in leaves from pea plants treated with Cd, specific fluorescent probes were used. DCF-DA was used to detect H 2 O 2 /peroxides, DHE for O 2 .-, and DAF-2 DA for NO, and samples were observed by confocal laser microscopy (CLSM) (Sandalio et al., 2008).  in the inner side of cell wall ( Fig S3).
The NO-derived DAF-2DA green fluorescence was found in xylem vessels, sclerenchyma, and epidermic cells of control plants (Fig 3, panel A) but, in contrast with ROS generation, Cd treatment produced a significant reduction of NO-dependent fluorescence observed in control leaves (Fig 3, panel B). The incubation of control leaves with aminoguanidine or L-NAME (not shown), two well known inhibitors of animal NOS, also produced a strong reduction of DAF-2DA fluorescence (Fig 3, panel C), which is indicative of the involvement of a NOS-like activity in the production of the NO detected. As a positive control, Cd-treated pea plants were incubated with 10 µM sodium nitroprusside (SNP), a NO donor, and the NO-dependent fluorescence was observed by confocal laser microscopy. In these conditions, a NO-dependent increase in DAF-2DA fluorescence in the leaf tissue was observed (Fig 3, panel D), showing the specificity of DAF-2DA for NO.  Fig S4). This suggests that the NO decrease by Cd could be in part due to an inactivation of the NOS activity as consequence of the Cdinduced of Ca deficiency in leaves. A higher magnification of images from Ca-Cd-treated plants shows the production of NO associated to the apoplast in xylem vessels, and also in sclerenchyma cells ( Figure S5.

Jasmonic acid, salicylic acid and ethylene under Cd stress
Jasmonic acid is a component of the signalling processes under biotic and abiotic stress (Devoto and Turner, 2005). To determine whether JA was involved in the cell response to Cd toxicity, this molecule was analyzed by GC-MS in leaves from control and Cd-treated pea plants. Under Cd stress, an increase of two times in methyl jasmonate (MeJA) took place in pea leaves (Fig 4) and free JA was detected neither in control nor in Cd-treated plants. The analysis of SA content show that free SA is the main form present in pea leaves. On the contrary, Cd treatment did not produce any statistically significant effect on the SA levels, although the content of conjugated (MeSA) and free SA were slightly reduced in Cd-treated plants (Fig 4). Analysis of ethylene by gas chromatography showed an increase of two times in leaves from pea plants grown with 50 μM CdCl 2 ( Fig 5) and this increase was reversed by supplying Ca to the nutrient solution, although a slight increase of ET emission was also observed in control plants (Fig 5).  (Fig 6). The induction of PrP4A and HSP 71.2 was reverted by the supply of ASC, a H 2 O 2 scavenger, which suggests that both genes are at least partially regulated by ROS (Fig 7). However, Ca did not change the expression level of PrP4A in both control and Cd-treated plants (data not shown). To study if there was a differential response of different cell types to Cd, the expression of the PrP4A gene was observed in situ on cross-sections of pea leaves by fluorescence in situ hybridization (FISH) (Fig 8).

Cd produces disturbances in cations accumulation
Cd is well known to produce disturbances in both uptake and distribution of elements in pea plants (Hernández et al., 1998;Sandalio et al., 2001;Tsyganov et al., 2007), and other plant species (Gussarson et al., 1996;Rogers et al., 2000). In this work, long term growth with 50 µM Cl 2 Cd produced a decrease in the content of Ca, Cu, Fe, Mn, and Zn in pea leaves.
Similar results have been observed previously in the same species (Sandalio et al., 2001) as well as in other plant species (Salt et al., 1995;Gussarson et al., 1996;Shukla et al., 2003;Azevedo et al., 2005). On the contrary, S was accumulated 3-fold in Cd treated plants respect to the control plants. The induction of sulfur metabolism by Cd has been previously described and involves a coordinate transcriptional regulation of genes for sulfate uptake and its assimilation, as well as GSH and phytochelatins (PCs) biosynthesis (Howarth et al., 2003;Nocito et al., 2006). Induction of PCs is one of the main detoxification strategies against Cd, by chelating Cd ions and preventing its toxicity (Howarth et al., 2003;Nocito et al., 2006).

Differential expression of superoxide dismutases by Cd
Cd-dependent reduction of SOD activity has been reported in wheat (Milone et al., 2003), peas (Sandalio et al., 2001) and beans (Cardinaels et al., 1984), although the opposite effect was observed in Alyssum plants (Schickler and Caspi, 1999) sunflower (Laspina et al., 2005), coffee cells (Gomes-Junior et al., 2006) and radish roots (Vitória et al., 2001). These discrepances are due to differences in the metal concentration and also in the period of treatment used in each case, in addition to the plant tissue studied. Thus, in garlic plants, SOD increased at short time Cd treatment, but decreased after long-term exposure (Zang et al., 2005). In this work, long term exposure to high Cd concentrations produced in pea leaf the down-regulation of Mn-SOD and CuZn-SOD transcripts which is correlated with the reduction of their activities previously observed (Sandalio et al., 2001;Romero-Puertas et al., 2007). The plastidic Fe-SOD, in its turn, was up-regulated although its activity was previously observed to be reduced by the metal (Sandalio et al., 2001).

Cd induces ROS accumulation and reduction of NO
The reduction observed in SOD activity and other antioxidants such as catalase, previously observed (Sandalio et al., 2001;Romero-Puertas et al., 2007)  peroxisomes, mitochondria and plasma membrane was demonstrated, being NADPH oxidase the main source of ROS in plasma membrane (Romero-Puertas et al., 2004). An oxidative burst has also been associated with Cd toxicity in Nicotiana tabacum cells suspensions, being an NADPH oxidase involved (Olmos et al., 2003;Garnier et al., 2006). The highest fluorescence detected was localized in the cell wall of the xylem vessels (Fig 2). Similar results have been observed in pea roots where the highest ROS production was associated to the vascular tissue (Rodríguez-Serrano et al., 2006). In different plant species ROS production has been associated with cell wall lignification in the xylem (Ogawa et al., 1997;Ros-Barceló, 1999). Apart from lignification, the production of ROS in vascular tissues could serve as a signal under stress conditions, such as it has been proposed in wounding damage (Orozco-Cardenas and Ryan, 1999). ROS overproduction was partially due to the Ca deficiency The protecting role of Ca can be also explained by the up-regulation of antioxidants such as CuZn-SOD.
The analysis of NO production by DAF2-DA fluorescence microscopy showed that fluorescence of control leaves was mainly due to a NOS-like activity, to judge by its inhibition by aminoguanidine (Corpas et al., 2004). But, in contrast with ROS, the production of NO was strongly reduced by Cd (Fig 3). The reduction of NO levels by Cd was also previously observed in pea roots and leaves under the same experimental conditions (Rodríguez-Serrano et al., 2006;Barroso et al., 2006). Aluminium treatment also led to a reduction of NO production in roots from Hibiscus (Tian et al., 2007)   In this work, the main production of ROS and NO took place in the xylem, sclerenchyma and epidermis. These results are consistent with reports from other authors showing that in cell wall lignification of xylem elements an oxidative burst is involved and a NO burst also participates in the programmed cell death associated to the differentiating vessels (Gabaldón et al., 2005). Moreover, ROS and NO production could be involved in signal tranduction pathways to activate the response to stress in other tissues.

The cellular response to Cd is mediated by jasmonic acid and ethylene
To get deeper insights into the mechanisms involved in the cell response to Cd toxicity, the analysis of jasmonic acid (JA), salicylic acid (SA) and ethylene (ET) contents was carried out. Cd induced an increase of JA and ethylene which suggest that theses molecules are involved in the cellular response to Cd toxicity. JA is an oxylipin which act as signalling compound in different defense situations such as response to pathogens and herbivore attack (Wastermack and Partnier, 1997). However, responses mediated by JA can be also triggered by diverse abiotic stresses (Devoto and Turner, 2005). JA is obtained from linolenic acid and its production is associated with lipid peroxidation and membrane damages.
In a previous work we have demonstrated that growth of pea plants with Cd induced lipid peroxidation in leaves (Sandalio et al., 2001)

Pathogenesis-related proteins (PRPs) could protect against Cd toxicity
The analysis of PRPs expression in pea plants under Cd stress showed the upregulation of chitinases, PrP4A and the HSP 71.2, while PAL did not change. Chitinase catalyse the hydrolitic cleavage of the β-1,4-glycoside bond of N-acetylglucosamine and is considered a defense mechanism against pathogens (Kasprzewska, 2003). An induction of chitinase activity has been observed in pea plants by Cd (Metwally et al., 2003), and other heavy-metals like lead and arsenic (Békésiova et al., 2007), and also by osmotic stress (Tateishi et al., 2001), low temperature (Stressmann et al., 2004) and wounding (Wu and Bradford, 2003). Chitinases are probably components of the general defense response program of cells, although they can also play unknown specific roles in heavy-metal stress.
Thus, transgenic plants expressing fungal chitinases showed enhanced tolerance to metals (Danna et al., 2006), and chitinase isoforms are differentially modified by different metals (Békésiova et al., 2007). Concerning the regulation of chitinases, Wu and Bradford (2003) demonstrated that they are regulated by ET and JA in tomato leaves; and Rakwal et al (2004) reported a regulation by ET, JA, and ROS in rice plants.
PrP4A is a hevein-related protein which binds chitin, can inhibit growth of fungus and belongs to chitinase I and II classes (Broekaert et al., 1990). To judge by the results obtained in this work, in pea plants Cd could generate a similar response induced by pathogen attack which is characterized by a ROS overproduction, NO reduction and PRP upregulation. However, unlike ozone or pathogen attack, Cd did not produce any visible symptom of local necrosis (Sandalio et al., 2001), although the formation of microlesions not visually detectable cannot be excluded. On the basis of the results obtained in this work and others previously reported, a cross-talk between ROS, NO and Ca in the regulation of cellular response to long-term Cd exposure is proposed (Fig. 9). Cd

Plant material and growth conditions
Pea (Pisum sativum L., cv Lincoln) plants were obtained from Royal Sluis (Enkhuizen, Holland). Plants were grown in the greenhouse in aerated full-nutrient media under optimum conditions during 14 days (Sandalio et al., 2001). Then, the media either remained unsupplemented (control plants) or were supplemented with 50 μM CdCl 2 (Cdtreated plants), and plants were grown for 14 days. To determine the effect of Ca, control and Cd-treated plants were supplemented with 10 mM Ca(NO 3 ) 2 one day before the addition of Cd and were grown for 14 days. The effect of ASC (a H 2 O 2 scavenger) was studied by infiltrating the leaves with 10 mM ASC.

RT-PCR analysis of gene expression
Total RNA was isolated from leaves by the acid guanidine thiocyanate-phenol-chloroform method of Chomczynski and Sacchi (1987)  Taq polymerase (Eppendorf, Hamburg, Germany) and 0.4 µmol of each primers (see Table   2S)

Preparation of probes
The cDNA from Pisum sativum was obtained by the protocol described before, and used for PCR amplification with the specific primers for the PrP4A included in the Table 2. The amplified fragment were isolated from an agarose gel and cloned using a pGEMT-easy cloning system (Promega). Specific probes of single-stranded RNA were generated by in vitro transcription using a DIG RNA labelling kit according to the manufacturer protocols (Roche).

Fluorescence in situ hybridization (FISH)
Pea leaves pieces were fixed in 4% paraformaldehyde in PBS (pH 7.0) for 16 hours, at 4ºC, cryoprotected with 2.3 M sucrose, embedded in tissue freezing medium, frozen on dry ice and sectioned in a cryostat (Leica CM 1800, Vienna, Austria) at -30ºC. Cryostat sections (40-60 μm thickness) were tawed and pretreated to facilitate penetration of the labelling reagents: sections were dehydrated in a series of 30, 50, 70 and 100% methanol/water, then rehydrated in a series of 70, 50, 30% methanol/water, and finally in PBS, for 5 min each step. Then, the sections were treated with 2% (w/v) cellulase (Onozuka R-10) in PBS for 1 h at room temperature, washed in PBS and water, and dried. The hybridization was performed essentially as previously described (Massoneau et al. 2005). Sections were incubated with hybridization solution at 50°C overnight. The hybridization solution consisted of digoxigeninlabelled RNA probe diluted 1/40 in the hybridization buffer (50% formamide, 10% dextran sulfate, 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 300 mM NaCl, 2xSSC, at room temperature, and in 0.1xSCC at 50ºC (1x SSC: 150 mM sodium chloride, 15 mM sodium citrate, pH 7.2).
After washing in PBS; sections were incubated in 5% BSA for 5 min. Then, the hybridization signal was detected by incubation with mouse anti-digoxigenin antibodies (Sigma), diluted 1/5000 in 1% BSA in PBS for 90 min at room temperature, followed by a fluorescence anti- The temperature was as follows: from 40ºC (1 min) to 20ºC/min until 150ºC (3 min), and from 5ºC/min to 230ºC. Helium was used as carrier gas (1 ml/min) and split less injection (1μl).

Ethylene determination
For the determination of endogenous ethylene production, fresh leaves (45 g) were placed in 100 ml hermetic vials, flushed with ethylene-free air and incubated for 3 h at room temperature. The ethylene concentration was determined on a gas chromatograph Perkin-Elmer 8600, fitted with a flame-ionisation detector and a Poropak-R column. Nitrogen was used as carrier gas and a commercial standard mixture of ethylene was used for calibration of the gas chromatograph.

Macronutrients and Micronutrients determination
Leaves were dried and mineralized with H 2 O 2 /nitric acid and microwave and the content of the elements were assayed by inductively coupled plasma emission spectrometry (ICP-EOS) analysis.         Table S1. Effect of Cd and Ca treatment on nutrient contents of leaves of pea plants. Table S2. Oligonucleotides used in this work for the semi-quantitative PCR analysis.        Figure 9. Model proposed for cross-talk between calcium, ROS and NO and its role in the regulation of the plant response to Cd toxicity. +, upregulation; -, downregulation; , reduction; , increase. PRs, pathogenesis related proteins; MAT-1, methionine adenosyltransferase 1.