1 Introduction

Cadmium (Cd) is an extremely toxic heavy metal that is naturally present in low concentrations in soil (Maksimović et al. 2007; Gratão et al. 2012). High concentrations of this metal have been found in some soils due to human activities such as mining, industrial activities, and increasing use of fertilizers and pesticides (Lux et al. 2011; Andresen and Küpper 2013).

Mechanic and chemical treatments are the most traditional methods for treating heavy metal-contaminated soils; the first removes the soil with heavy equipment, while chemical methods are based on induced chemical reactions in soils making heavy metal unavailable (Yao et al. 2012). However, these treatments may affect soil aggregation or are not efficient to remove completely the heavy metal from the soil, besides that both methods are expensive choices.

Phytoremediation is a promising technique which is based on the plant ability to uptake and accumulate toxic elements in its tissues; however, the “ideal” plant species characteristics for soil phytoremediation is still under debate, although some basic aspects are known such as high biomass production (Zhu et al. 2012; Souza et al. 2013a).

Most of the available studies targeted hyperaccumulator plant species such as Noccaea caerulescens and Arabidopsis hallery (Huguet et al. 2012; Lovy et al. 2013). However, these plant species are slow growing and produce small amounts of biomass, making them unfeasible for field application when considering remediation purpose. On the other hand, some crop plant species whose ability to uptake toxic elements are lower than the hyperaccumulator species, have moderate to high tolerance to heavy metals, exhibit rapid growth, and produce large amounts of biomass and therefore are more suitable for phytoremediation studies (Andrade et al. 2008; de Souza et al. 2012b).

The leguminous plant species Dolichos lablab L. has appropriate morphological and growth characteristics for phytoremediation, but there is no information about its capacity to tolerate Cd and use in heavy metal phytoremediation. This plant species has already been shown to be tolerant to drought and saline stresses (D’Souza and Devaraj 2010; Younis 2010) and a potential phytoremediator for the herbicide trifloxysulfuron sodium (Procópio et al. 2004). Additionally, D. lablab L. is widely used as green manure due to its capability of nitrogen fixation.

The approach of using non-hyperaccumulator plant species that tolerate moderate concentrations of heavy metals is a very interesting strategy mainly when considering crop species that produce high amount of biomass (Bhargava et al. 2012). There are few studies carried out in this context (Andrade et al. 2008; Adhikari and Kumar 2012), especially using leguminous plants.

The potential of D. lablab L. for Cd phytoremediation was investigated in this work, and three distinct experiments were performed: the first to determine D. lablab L. ability to grow in Cd-contaminated soil, while the second evaluated its antioxidant metabolism response when exposed to Cd and the third to investigate the response in terms of protein changes in roots and leaves using a two-dimensional electrophoresis-based proteomic approach.

Proteomic analysis has been extensively used to investigate protein production patterns under several abiotic stresses (Bona et al. 2011; Arruda et al. 2011) and contributed for the understanding of how plants cope with stressful situations (Farinati et al. 2009; Ahsan et al. 2012; Arruda et al. 2013). This approach has been carried out mainly for hyperaccumulator plant species (Farinati et al. 2009; Zhao et al. 2011), whereas in the case of non-hyperaccumulator plant species, this approach may, for instance, contribute for comprehension of which proteins are being modulated by Cd presence in soil and consequent uptake and translocation in the plant system.

Therefore, the central objective of this study was to assess how D. lablab L. responds at biochemical and physiological levels to increasing concentrations of Cd, which may also provide information on its potential use in phytoremediation of Cd-contaminated soils.

2 Materials and methods

2.1 Experiment I

2.1.1 Plant material, soil preparation, and experimental design

Seeds of D. lablab L. were provided by “Piraí Sementes”. The seeds were surface sterilized in 10 % hypochloride solution for 20 min, washed in running water, immersed in pure water for 60 min, and finally sown in pots filled with Cd-contaminated soil previously prepared as described below.

This experiment was carried out in soil in order to simulate a realistic contaminated environment. The soil in each pot was contaminated with CdCl2 solution and allowed to rest for 15 days for stabilization. A loamy sand soil was used and its chemical characteristics were as follows: pH, 4.5; base saturation (V%), 24; organic matter, 2.1 %; H + Al, 47 mmolc dm−3; P, 5 mg kg−1; K, 45 mg kg−1; Ca, 200 mg kg−1; Mg, 48 mg kg−1; Cu, 0.6 mg kg−1; Fe, 86 mg kg−1; Mn, 5.8 mg kg−1; and Zn, 2.5 mg kg−1. After application of 0, 5, 10, and 15 mg kg−1 of Cd, the final available Cd concentrations were 0.05, 3.3, 7.0, and 14.7 mg kg−1.

The experiment was carried out in a greenhouse in a completely randomized design with four CdCl2 concentrations applied to the soil (0, 5, 10, and 15 mg kg−1) with four biological repetitions. Seeds were planted and 45 days old plants were harvested, washed in running water, and dried with soft paper. Shoots and roots were separated and dried at 60 °C for 72 h.

2.1.2 Growth measurements and mineral analysis

After drying, shoots and roots were weighted for dry mass determination. Shoots and roots were ground and subjected to nitric-perchloric digestion for determination of P, S, Mg, Ca, Fe, Cu, Zn, Mn, and Cd by ICP-OES. K was determined in the same extract in flame photometer equipment. For total N determination, shoots and roots were digested in sulfuric acid followed by quantification by the method of semi-micro Kjeldahl. For determination of Cd concentration in soil, the metal was extracted with a DTPA solution as described by Lindsay and Norvell (1978) and determined in an atomic absorption spectrometer.

2.1.3 Phytoremediation potential evaluation

The phytoremediation potential was evaluated by analyzing tolerance index (TI = BMT/BMC), translocation index (TI % = 100*CdSH/CdWP) according to Rahman et al. (2013) and transfer factor (TF = [Cd]WP/[Cd]S) according to Lübben and Sauerbeck (1991), where BMT is total biomass produced in Cd treatments, BMC is total biomass produced in control treatment, CdSH is total microgram of Cd in shoots, CdWP is total microgram of Cd in whole plant, [Cd]WP is Cd concentration in whole plant (mg kg−1), and [Cd]S is Cd concentration in soil (mg kg−1).

2.1.4 Statistical analysis

Analysis of variance (ANOVA) and Tukey’s test at 5 % of significance using the software SISVAR® were performed.

2.2 Experiment II

2.2.1 Plant material and experimental design

This experiment was carried out in sand in order to have clear samples without interference by any type of biological contaminants. For the determination of antioxidant enzyme responses of D. lablab L. under increasing concentrations of Cd, a sand-pot experiment in greenhouse was conducted. Seeds were germinated in bioplant®/vermiculite (1:1) substrate. Seedlings were transferred to pots filled with autoclaved sand and received complete Hoagland nutritive solution for 20 days. After this period, 0, 50, 100, and 200 μM CdCl2 (corresponding to 0, 5.6, 11.2, and 22.4 mg L−1 Cd) were set and applied in complete Hoagland nutritive solution.

Plants were exposed to the Cd concentrations for 5 days, and complete Hoagland nutritive solution (Hoagland and Arnon 1950) without Cd was applied when necessary to supply nutrients. After 5 days of exposure, the second and third leaves as well as central root parts were collected and immediately frozen in liquid N2. The samples were stored at −80 °C for further enzyme extraction and analyses. The remaining vegetative tissues were dried at 60 °C for 72 h, weighed, and Cd was quantified.

2.2.2 Stress indicators measurements

Oxidative stress indicators, lipid peroxidation, and hydrogen peroxide, were measured in leaves and roots. Lipid peroxidation was determined as described by Heath and Packer (1968). Hydrogen peroxide was determined as described by Alexieva et al. (2001).

2.2.3 Enzyme extraction and protein quantitation

Leaves and roots were grinded separately in a mortar with a pistil using liquid N 2. The samples were then homogenized in 100 mM potassium phosphate buffer (pH 7.5) containing 1 mM EDTA, 3 mM DTT and 5 % (w/v) PVPP. The homogenate was centrifuged at 12000 g at 4 °C for 30 min. The supernatant was collected in 200 μL aliquots and stored at−80 °C for protein quantitation and enzyme activities. Total extracted protein was determined as described by Bradford (1976) using BSA as a standard.

2.2.4 Enzymes activities

Superoxide dismutase (SOD; EC 1.15.1.1) activity and isoenzymes identification were determined in non-denaturing polyacrylamide gel electrophoresis (PAGE) according to Laemmli (1970) followed by gel activity staining as described by Beauchamp and Fridovich (1971) with some modifications (Azevedo et al. 1998). Catalase (CAT; EC 1.11.1.6) activity was determined as described by Kraus et al. (1995) with some modifications (Azevedo et al. 1998). Guaiacol peroxidase (GPOX; EC 1.11.1.7) activity was determined as described by Matsuno and Uritani (1972). Ascorbate peroxidase (APX; 1.11.1.11) activity was determined as described by Nakano and Asada (1981). Glutathione reductase (GR; EC 1.6.4.2) activity was determined as described by Smith et al. (1988).

2.2.5 Statistical analysis

ANOVA and Tukey’s test at 5 % of significance using the software SISVAR® were performed.

2.3 Experiment III

2.3.1 Experimental design

This experiment was carried out in sand in order to have clear samples without interference by any type of biological contaminants. Seeds were germinated in substrate containing (1:1) bioplant®/vermiculite. Seedlings were transferred to pots filled with autoclaved sand and received complete Hoagland nutritive solution (Hoagland and Arnon 1950) and at the 35th day the nutrient solution was supplemented with CdCl2 at 100 μM final concentration. The plants were kept in Cd solution for 5 days. At the end of this period of exposure, leaf and root samples were collected and stored at−80 °C for proteomic analysis. Such as described for experiments I and II, the remaining vegetative tissues were dried at 60 °C for Cd determination.

2.3.2 Protein extraction, quantification and two-dimensional gel electrophoresis proteomic analysis

Foliar tissue proteins extraction was carried out with TCA/acetone solution as described by Xu et al. (2008). After precipitation, the pellet was vacuum-dried and 50 mg of dried material was added to a maximum of 1 mL of solubilization solution (7 M urea, 2 M thiourea, 2 % CHAPS and 0.4 % triton X-100). The sample was vigorously stirred using a vortex and kept in ultrasonic bath for 20 min and after 3 cycles of stirring and ultrasonic bath, the material was centrifuged at 12,100×g at 4 ºC for 20 min and the supernatant was collected for further analysis.

Root tissue protein extraction was carried out with 100 mM Tris buffer (pH 8.5), 5 mM DTT, 1 mM EDTA, and 1 mM PMSF as described by Lee et al. (2011) followed by a precipitation step for 18 h with TCA/acetone. The pellet was washed 3 times in 0.07 % β-mercaptoethanol in pure acetone. Protein solubilization was carried out in a maximum volume of 1 mL of solubilization solution (7 M urea, 2 M thiourea, 2 % CHAPS, and 0.4 % triton X-100). Protein concentration determination was carried out as described by Bradford (1976).

2.3.3 Protein focalization—1st dimension

Isoeletric focusing was performed in 18 cm strips “Immobiline Dry Strip” non-linear pH 4–7 from GE Healthcare Life Sciences. A total amount of 800 and 500 μg of extracted protein from leaves and roots were used, respectively. After 6 h of rehydration, the proteins were subjected to step isoeletric focusing according to the following program: 30 V for 12 h, 100 V for 1 h, 200 V for 1 h, 400 V for 1 h, 700 V for 1 h, 1000 V for 1 h, 5000 V for 10 h, 8000 V for 4 h, and 100 V for 3 h for leaf proteins and 200 V for 1 h, 500 V for 1 h, 1000 V for 1 h, 5000 V for 8 h, 8000 V for 3:30 h, and 200 V for 2 h for root proteins.

2.3.4 Protein separation by SDS-polyacrylamide gel electrophoresis—2nd dimension

Prior to electrophoretic protein separation, the proteins in the strips were reduced and alkylated as recommended by GE Healthcare (2004). The strips were placed on a 12.5 % polyacrylamide gel and proteins separated at 15 mA per gel for 20 min, and then at 30 mA per gel for 4:30 h. The gels were stained with colloidal Coomassie Blue G-250 as recommended by GE Healthcare (2004).

2.3.5 Gel image analysis

The gels were scanned in an ImageScanner using the software LabScan™ v 5.0 (Amershan Bioscience). The spots were detected by the software ImageMaster 2D Platinum 7.0 and the parameters contrast, smoothness, saliency, and minimal area were adjusted for automatic spot detection.

2.3.6 Protein sequencing

The selected spots were excised from the gels and processed for mass spectrometry analysis according to the procedures described by Shevchenko et al. (2007). Digested proteins were submitted to MALDI TOF-TOF MS/MS analysis for protein identification. The results were compared by BLAST against virideplantae NCBI databank through the software MASCOT®.

2.3.7 Statistical analysis

The analysis of the spots was carried out using the software ImageMaster 2D Platinum 7.0, followed by ANOVA and Student’s t test considering p < 0.05.

3 Results

3.1 Cadmium effects on mineral nutritional I: macronutrients

The shoot concentrations of P, S, K, and Ca were not affected by Cd presence, while the concentrations of N and Mg increased with the increase in Cd concentration in the soil (Table 1). In roots, Cd did not influence the concentrations of N, P, K, and S, but the concentration of Mg decreased as the concentration of Cd in soil increased, while Ca concentration exhibited the opposite pattern (Table 1).

Table 1 Macronutrients concentration (g kg−1) in shoots and roots of Dolichos lablab L. under increasing concentrations of cadmium
Full size table

3.2 Cadmium effects on mineral nutritional II: micronutrients

The concentrations of Fe and Mn were not affected by Cd in shoots, but the concentrations of Cu and Zn decreased as the concentration of Cd increased in the treatments (Table 2). In roots, the concentrations of Cu, Fe, Zn, and Mn increased with the increase in Cd concentration in the soil, with a slight reduction in the concentrations of these elements being observed only at the highest concentration of Cd used (Table 2).

Table 2 Micronutrients concentration (mg kg−1) in shoots and roots of Dolichos lablab L. under increasing concentrations of cadmium
Full size table

3.3 Cadmium effects on growth and phytoremediation potential of D. lablab L

The growth and biomass accumulation of D. lablab L. decreased linearly according to the increase of Cd concentration in the soil (Table 3). Based on the root/shoot ratio, shoots were more sensitive to Cd than roots (Table 3). The highest Cd concentration used drastically decreased biomass yield (around 75 %), while the lowest Cd concentration used led to a smaller biomass reduction (around 40 %) (Table 3). Cd was mainly accumulated in the roots and the accumulation increased linearly in both, shoots (Fig. 1a) and roots (Fig. 1b), according to the increment of Cd available in the soil. The soil-plant transfer factor (TF), translocation index (TI %), and tolerance index (TI) were higher for the lowest Cd concentration used (Table 4).

Table 3 Growth measurements of Dolichos lablab L. under increasing concentrations of cadmium
Full size table
Fig. 1
figure 1

Cadmium (Cd) concentration in shoots (a) and in roots (b) of Dolichos lablab L. under increasing concentrations of cadmium. Means (n = 4) with the same letter are not significantly different by Tukey’s test at 5 %

Full size image
Table 4 Phytoremediation indicator parameters for Dolichos lablab L. under increasing concentrations of cadmium
Full size table

3.4 Oxidative stress indicators

There was no increase in hydrogen peroxide content in both leaves and roots in all treatments (Fig. 2a). Lipid peroxidation was higher in the 200 μM Cd treatment in both leaves and roots (Fig. 2b).

Fig. 2
figure 2

Hydrogen peroxide concentration (a) and malondialdehyde concentration (b) in leaves and roots of Dolichos lablab L. under increasing concentrations of cadmium (Cd). Upper cases compare the means of leaves, and lower cases compare the means of roots. Means (n = 4) with the same letter are not significantly different by Tukey’s test at 5 %

Full size image

3.5 SOD isoenzymes and activity

Five SOD isoenzymes were observed in leaves (Fig. 3a) and six in roots (Fig. 3b). In leaves, total SOD activity or differential SOD isoenzyme activity patterns showed higher activity under 100 μM Cd, while in roots Cd did not affect SOD activity. In roots, the most active SOD isoenzyme was Cu/Zn-SOD.

Fig. 3
figure 3

Superoxide dismutase activity and biochemical characterization in leaves (a) and roots (b) under increasing cadmium (Cd) concentrations. a I, Mn-SOD; II and III, Fe-SOD; IV and V, Cu/Zn-SOD; b I, Mn-SOD; II and III, Fe-SOD; IV–VI, Cu/Zn-SOD. S standard bovine SOD

Full size image

3.6 Cadmium effects on CAT, GPOX, APX, and GR

In leaves, CAT, APX, and GPOX activities increased in response to increased Cd concentrations (Fig. 4). CAT and APX exhibited maximum activity when exposed to 100 μM Cd (Fig. 4a, c). In a similar manner to leaves, root APX activity increased under Cd stress, while GPOX activity decreased (Fig. 4c, b). APX and GPOX activities were higher in roots than in leaves, whereas CAT activity was higher in leaves than in roots. When GR activity is concerned, no differences in activity between both tissues and Cd treatments were observed (Fig. 4d).

Fig. 4
figure 4

Activity of the antioxidant enzymes CAT (a), GPOX (b), APX (c), and GR (d) of Dolichos lablab L. under increasing concentrations of cadmium (Cd). Asterisk indicates significant difference between means of leaf and root activity under the same Cd concentration by Tukey’s test at 5 %. n = 4

Full size image

3.7 Differentially expressed proteins under Cd stress identified by MALDI TOF-TOF MS/MS

Identified proteins, in both roots and shoot, were categorized in nine different classes: cellular signaling, carbohydrate metabolism, amino acids metabolism, secondary metabolism, energetic metabolism, abiotic and biotic stress response, protein metabolism, antioxidant metabolism, and photosynthetic metabolism, according to the function retrieved from uniprot (Table 5). Differentially expressed proteins under Cd stress (100 μM) in both, root and shoot, were distributed in almost all protein classes, except for secondary metabolism and protein metabolism classes which presented only differentially expressed protein from roots and photosynthetic metabolism class which presented only differentially expressed protein in leaves (Table 5).

Table 5 Identification and analysis of differentially expressed proteins in roots and leaves of Dolichos lablab under Cd stress
Full size table

The majority differentially expressed proteins from roots belonged to cellular signaling, carbohydrate metabolism, amino acids metabolism, secondary metabolism and protein metabolism classes, while those from shoot belonged to energetic metabolism, abiotic and biotic stress response, antioxidative metabolism, and photosynthetic metabolism classes (Table 5).

The same proteins were not found to be differentially expressed in both, leaves and roots, therefore each organ presented unique regulation of their proteins. In leaves, photosynthetic metabolism proteins were down regulated as well as proteins from amino acids metabolism (Table 5). However, the opposite was observed in proteins from carbohydrate metabolism which were upregulated in the presence of Cd. Moreover, proteins belonging to classes directly related to stress such as antioxidant metabolism and abiotic and biotic stress response were upregulated in the presence of Cd in leaves (Table 5). In roots, proteins belonging to abiotic and biotic stress response were shown to be upregulated in presence of Cd (Table 5).

4 Discussion

4.1 Potential of cadmium phytoremediation of D. lablab L

The increase in industrialization and excessive use of fertilizers contributes to increased concentration and availability of several heavy metals in the environment (Monteiro et al. 2011). Cadmium, as a highly toxic heavy metal, impairs plant growth, which may ultimately lead to plant death (Gallego et al. 2012; Azevedo et al. 2012; Gratão et al. 2015).

In the present study, the observation that macronutrients were not affected in D. lablab L. under increasing concentrations of Cd (Table 1) is a characteristic that is not normally observed in other plant species. Most authors have reported contrasting effects caused by Cd on the nutritional status. Tezotto et al. (2012) found that Cd decreased P content in leaves of coffee plants, although they did not observe growth reduction, while Vernay et al. (2007) reported that excess of chromium (Cr) affected Ca and Mg content in Lolium perene L., which was accompanied by growth reduction. It is therefore acceptable that excess of Cd may lead to impairments in macro- and micronutrient minerals and any situation disturbing their absorption reflects negatively on plant growth capability (Maathuis 2009). The results obtained in this study do not only support that the response is plant species specific, showing the importance to study each metal-plant interaction, but also suggesting that D. lablab L. is able to keep normal concentrations of most macronutrients under Cd-stressful situation, at least under the conditions tested in this study, in order to support its growth.

Additionally, elements that have specific roles in decontamination and energetic metabolism, such as S and Mg (Arruda et al. 2015), whose concentrations increased under Cd exposure may indicate a strategy to keep the metabolism active in a growing attempt, which may cause a dilution effect by biomass increase as observed in Calopogonium mucunoides (de Souza et al. 2012a). Elements whose concentration decreased under Cd exposure may be explained by inhibition or competition between Cd and other metallic nutrients for carrier proteins as observed by Bertoli et al. (2012), who observed that K levels decreased in tomato plants under increasing concentrations of Cd. As Cd slightly affected only the content of K and Mg in roots of D. lablab L. (Table 1), we propose that the homeostasis of these elements in this plant species was not kept only in the roots, since the translocation of these elements to the above-ground parts of the plants was not affected.

Micronutrients are also of extreme importance for cellular metabolic maintenance, mainly because they are involved in redox reactions (Hänsch and Mendel 2009). Plants have different ways to respond in relation to absorption, transport and accumulation of micronutrients under heavy metal exposure. As demonstrated by Safarzadeh et al. (2013), rice plants subjected to Cd exposure resulted in diminished concentrations of Zn, Cu, Mn and Fe, and due to the essentiality of these elements, quantities below the optimum might have impact on plant development. For instance, similar results were found by Rezvani et al. (2012) in Aeluropus littoralis, a graminae plant species. Nevertheless, that is not the case for D. lablab L. whose data obtained were different revealing increment of micronutrients in the roots, whereas in the leaves no deficiency was observed (Table 2), suggesting that the nutritional status was not affected (Tables 1 and 2). Therefore, it appears that in D. lablab L. Cd toxicity does not have an apparent effect on macro or micronutrient acquisition or translocation, possibly pointing out to a kind of tolerance mechanism. Cadmium is generally related to growth decrease in plants (Konotop et al. 2012) especially at high concentrations (Azevedo et al. 2012), and a certain degree of growth reduction was observed in D. lablab L. (Table 3), nevertheless, such an effect cannot be explained or related to an impairment in mineral nutrient uptake or translocation.

Some authors reported growth reduction of at least 90 % with Cd concentrations similar to those used in this study (López-Millán et al. 2009; Hediji et al. 2010). Although we have observed some growth reduction at the lowest Cd concentration (5 mg kg−1) (Table 3), D. lablab L. was able to maintain 60 % of the growth when compared with the control (Table 4). Moreover, this concentration has been considered toxic for several plants species, for instance, for tomato plants (Dourado et al. 2014). An important aspect to be taken into account is a possible root tolerance, which can be observed by the obtained Cd root:shoot ratio (Table 3), which was similar to the control and also the amount of Cd accumulated in the roots. The higher Cd tolerance exhibited by D. lablab L. is important to support the use of this plant species for phytostabilization of areas with forthcoming risk of contamination, as suggested by Göhre and Paszkowski (2006), and this root tolerance may be related to a root barrier, which prevents the accumulation of excessive metals in edible aerial parts (Disla et al. 2014). As phytostabilization prevents heavy metal spreading in soil as well as leakage, it is very plausible this use for diminishing risk of contamination of uncontaminated surrounding areas.

Based on the results obtained and on the concept of Reeves and Baker (2000) in which a Cd hyperaccumulator species should accumulate in its aerial parts at least 0.01 % of its dry mass as Cd, D. lablab L. cannot be classified as hyperaccumulator. However, we propose the use of this plant species for phytostabilization, based on the amount of Cd accumulated in roots (Fig. 1), the high TF observed for the lowest Cd concentration (Table 4) as suggested by Kabata-Pendias and Pendias (2001), and also on the high root/shoot ratio observed at the lowest Cd concentration used (Table 3). Therefore, due to the high biomass produced by D. lablab L. we reinforce the idea of using non-hyperccumulator plants as phytoremediators (Souza et al. 2013b) once all parameters used to support our idea allowed us to conclude that root growth was less affected by Cd, which can be an important aspect to overcome Cd stress and successfully be established as a phytoremediator.

4.2 Antioxidative response of D. lablab L. to cadmium

The ability of D. lablab L. to grow in the presence of different Cd concentrations suggested some degree of Cd tolerance, which did not influence nutrient balance (Tables 1 and 2) despite of plant growth reduction at the highest Cd concentration (Table 3). A reported mechanism involved in Cd tolerance is the maintenance of the redox balance in the cell, mediated by the enzymes of the antioxidant pathway or non-enzymatic antioxidant compounds (Gratão et al. 2005; Fidalgo et al. 2011). The SOD enzyme is involved in superoxide dismutation and the different SOD isoenzymes are located in different cell compartments (Azevedo et al. 1998), therefore responding differently depending on the plant species, genotype, organ, and cell location (Gonçalves et al. 2009).

Our results revealed the presence of up to six distinct SOD isoenzymes in the leaves and roots classified as Mn-SOD, Fe-SOD, and Cu/Zn-SOD (Fig. 3). Total SOD activity, based on the sum of SOD band intensities, was clearly higher in roots than in leaves, nonetheless, Cd exposure did not result in any major or significant change in band intensity or band distribution/presence pattern, except in leaves at 100 μM of Cd, which exhibited a slight increase (Fig. 3a, b). Yet, it is clear that the majority of the SOD activity observed in this plant species is due to Cu/Zn-SOD and also Mn-SOD activities when roots are concerned. The potential higher SOD activity already encountered in D. lablab L. independent of stress still have to be challenged in future studies possibly with higher Cd concentrations and also by testing other metals so that a better understanding of SOD activity thresholds and responses are determined for this plant species. Moreover, adaptive responses that can take place (Gratão et al. 2008a) cannot be ruled out and should be taken into account for D. lablab L.

Hydrogen peroxide is produced by the cell metabolism, including as a result of SOD activity, which can then interact with metallic ions to produce hydroxyl radicals, which is highly active and responsible for most of the cellular damage (Apel and Hirt 2004; Kim et al. 2008). Therefore, an efficient hydrogen peroxide scavenging system is required by the cells and the enzymes responsible for these processes are peroxidases (PODs) (Jouili et al. 2011). The results obtained for D. lablab L. suggest that the PODs tested strongly act on hydrogen peroxide degradation, besides presenting differential response between plant tissues (Fig. 4a–c). Several stressing factors such UV-radiation, ozone exposure, drought, senescence, and heavy metals lead to an increase in enzymatic activity of PODs (Gratão et al. 2008b; Cruz et al. 2013). The heavy metal-induced oxidative stress on POD activities have already been reported by some authors (Andrade et al. 2010; Dominguez et al. 2010; Pereira et al. 2011; Remans et al. 2012), which clearly indicated that the enzymes responses are tissue and/or organism specific. Yu et al. (2013) reported significant increase in several antioxidant enzymes, including CAT, corroborating with our results that the excess of hydrogen peroxide is scavenged by different PODs from the antioxidant pathway. Hydrogen peroxide scavenging can also occur by the activity of APX and GPOX, as observed in this study in the root tissue (Fig. 4b, c), frequently being much more effective than CAT (Caverzan et al. 2012). APX appears to be very effective in eliminating hydrogen peroxide in roots, which can be explained by its Km (Mittler 2002), as well as the amount of isoenzymes and substrate available (Mittler and Poulos 2007).

Nadgorska-Socha et al. (2013) observed an increase in GPOX activity in leaves under high concentrations of Cd, a response also observed in the present study, suggesting that this pattern of response in leaves may be normal once it was also observed in other reports for other plant species (Monteiro et al. 2011; Roychoudhury et al. 2012). On the other hand, the response of this enzyme was the opposite in roots, in which we observed a decrease in its activity (Fig. 4b), indicating that Cd can influence the response of the same enzyme differentially in each tissue. Another fact that could explain this decrease in GPOX activity in roots could be the high activity of APX in this tissue (Fig. 4c). Wang et al. (2008) working with Thlaspi caerulescens and Brassica juncea subjected to Cd stress observed a decreased CAT activity with simultaneous increase in the activity of generic PODs.

As already mentioned, the APX enzyme is very effective in hydrogen peroxide elimination, but APX is also important for the integration of the ascorbate-glutathione cycle (Suzuki et al. 2012; Baxter et al. 2014). Thus, ascorbate needs to be constantly regenerated and it mainly depends on GR activity to regenerates reduced glutathione (GSH), the electron donor for dehydroascorbate reduction (Mittler 2002; Zhao et al. 2009). GR is strongly induced under several environmental stresses (Opdenakker et al. 2012) due to its key role in keeping cellular redox homeostasis, which corroborates the suggestion that D. lablab L. can be considered a Cd-tolerant species in low to moderate Cd-contamination environment due to the relative good stability of GR activity (Fig. 4d).

The inter-relationship between the activities of APX and GR is clearly noticeable once an increase in GR activity was observed in roots (Fig. 4d), which probably is related to the high demand for ascorbate. In accordance with our results, Smeets et al. (2005) observed that the ascorbate-glutathione cycle efficiency is the main mechanism against Cd-induced oxidative stress in beans. Additionally, Li et al. (2013) confirmed in Pistia estratiotes L. that POD and SOD activities were increased under Cd stress. So, these results are in agreement with our findings and suggest that once the balance of hydrogen peroxide production in the ascorbate-glutathione cycle is altered, other PODs rapidly respond to keep the hydrogen peroxide content lower enough to avoid this molecule to induce free radicals formation that ultimately cause oxidative stress.

It is clear that enzymes from the antioxidant pathways respond differently and the decrease in activity of one type of POD may lead to an increase in the activity of another related enzyme, acting as a compensatory mechanism as reviewed by Gratão et al. (2005). Such responses may obviously differ in different organs as observed in D. lablab L. for CAT and APX activities between roots and leaves. Since high activities of enzymes of this pathway can be related to a better growth performance under stressful situations (Siddiqui et al. 2013), it is possible to infer that the observed behavior serves as an indicative of D. lablab L. versatility to cope with Cd exposure.

The pattern of response observed in roots (Fig. 4), clearly differed from that observed for leaves, which might be explained by the fact that this organ is the first to get in contact with the metal, therefore requiring a fast and effective response. In this way, we can assume that the high efficiency of the root tissue in keeping low the level of ROS such as hydrogen peroxide and superoxide is intimately related to D. lablab L. tolerance under Cd stress. Although we did not measure all ROS, the effect of Cd on growth, MDA, hydrogen peroxide, and enzyme activities, do allow us to argue that D. lablab L. root system appears to be able to face and deal with any excess of ROS eventually produced by the Cd treatment, at least under the conditions tested in this study.

4.3 Proteomic changes of D. lablab L. under Cd exposure

The two-dimensional proteomic-based approach was applied to elucidate which proteins were regulated by 100 μM Cd exposure in D. lablab L. Only recently proteomic studies have been carried out for plants under heavy metal stress and significantly contributed for the understanding of metabolic pathways that are affected, as well as identification of proteins related to tolerance (Farinati et al. 2009; Sharmin et al. 2012; Wang et al. 2012). This approach has been used mostly for hyperaccumulator plant species (Ingle et al. 2005; Tuomainen et al. 2006; Farinati et al. 2009; Walliwalagedara et al. 2010; Zhao et al. 2011), however, heavy metal-tolerant non-hyperaccumulator plant species do need to be characterized with the goal of understanding their potential as phytoremediators.

An important shift in metabolic pathways has been observed in D. lablab L. In roots, we could detect the upregulation and downregulation of four and two proteins, respectively, related to carbohydrate metabolism. The upregulation of enzymes involved in carbohydrate metabolism seems to be a common feature as indicated by Zhao et al. (2011) and Wang et al. (2012) in studies with Cd and Ni hyperaccumulator species. The isoenzymes of triose phosphate isomerase and glyceraldehyde 3-phosphate dehydrogenase from roots (Table 5) exhibited more than 50 % increase in their abundance. As these two enzymes are part of the glycolytic pathway, it seems reasonable to conclude that the root is using carbon to generate energy to defend itself from Cd stress. As our work was not with a hyperaccumulator species, it may be also possible to suggest that the metabolic shift may be shared between hyperaccumulators and non-hyperaccumulators, and it is interesting because it indicates that metabolic versatility under stressful situations may integrate different metabolic pathways, such amino acid metabolism, once methionine synthase was also upregulated.

In the amino acid metabolism class, the enzyme glutamine synthetase was downregulated in both roots and leaves (Table 5) indicating that Cd directly affects N assimilation, resulting in growth disturbances in plant in contaminated soils, especially because the product of this enzyme serve as substrate for synthesis of tolerance compound, such as glutathione. However, Hradilova et al. (2010) reported an increase in glutamine synthetase, which the authors correlated with increased tolerance of flax plants to Cd.

Tolerance of D. lablab L. to Cd appears to be closely related to carbohydrate and amino acids metabolism considering that molecules from primary metabolism are being converged to the energy produced by the Krebs’ cycle in an attempt to overcome the stressful situation.

The upregulation of isocitrate dehydrogenase in roots under Cd stress suggests that this enzyme increased in abundance as a plant response to supply substrate for the synthesis of compounds that may confer tolerance, such as organic acids that are used as organic chelators to attenuate metal toxicity. Yet, an opposite view was reported by Sánchez-Pardo et al. (2013) who observed a 22 % decrease in abundance of this enzyme, suggesting that it may lead to drastic effects on nitrogen assimilation, as this enzyme provides α-ketoglutarate for ammonium incorporation in the GS/GOGAT pathway. For D. lablab L., it may not favor nitrogen assimilation, since a decrease in abundance of both isoenzymes of glutamine synthetase was observed, which may explain the effects of Cd on growth disturbance.

What is particularly interesting is the link between primary metabolism with the secondary metabolism, as the latter is not involved with energy supply, but several metabolites from primary metabolism are substrate for the synthesis of secondary metabolites, and increases in two enzymes from the secondary metabolism such cynnamoil CoA reductase and isoflavone reductase were found (Table 5).

It is known that Cd also influences some proteins from the secondary metabolism, which responded with upregulation in roots in this study and was also reported by He et al. (2011). One of the upregulated enzymes was cynnamoil CoA reductase, an enzyme involved in lignin biosynthesis, which suggests possible Cd interference in auxin metabolism, leading to higher lignification and cell shortening (Elobeid et al. 2012). Lignification is a common response under biotic and abiotic stress. It the case of biotic stresses, it appears that lignification acts as a physical barrier to biological invasions, while when abiotic stresses are concerned, Podazza et al. (2012) studying Cd response in Citrus concluded that the increase in lignifications is responsible for increase in Cd immobilization in root cell walls. The fact that the responses to both biotic and abiotic stresses have a point of convergence, agrees with the data obtained, which showed the increase in proteins related to both types of stresses such as chitinase in roots and heat shock proteins (HSP) and lectin in the leaves (Table 5).

Some HSPs are known as important environmental stress biomarkers (Bierkens 2000). Such as observed in this work, Basile et al. (2015) also reported increased expression of HSP in Lemna minor under stress by Cd and other metals, which confirms the role of this chaperone protein as a biomarker. In relation to lectin and chitinase, it is difficult to determine their roles under Cd stress since these two proteins are involved in biotic stress responses, but we may suggest that the signaling pathway may be similar in response for both biotic and abiotic stress (Smékalová et al. 2014; Prasch and Sonnewald 2015).

In leaves, despite of the upregulation of carbohydrate and amino acid metabolic proteins, photosynthesis appears to be impaired since it was observed downregulation of proteins involved in the photosynthetic process, such as the small chain of RubisCO and oxygen-evolving enhancer protein (Table 5). It is expected the decrease in photosynthesis under Cd stress as observed by Zancheta et al. (2015) studying response of sorghum and jack-bean to Cd, and even in Cd-tolerant plants, such as sunflower, the effect is the same as observed by Lopes et al. (2015) that studied Cd effects on proteins using a proteomics approach. Additionally, Cd also impairs chlorophyll content, which is directed related to photosynthesis as observed in tomato plants (Dourado et al. 2013).

Cd appears to induce a shift in the metabolism and these changes are converged to upregulation of stress response proteins and at some extent to downregulation of proteins involved in growth. This behavior is well supported under stressful situation in which organisms tend to direct their resources to overcome the stress while growth remains stunted (Prasch and Sonnewald 2015).

5 Conclusions

The approaches used in this research allowed us to support the hypothesis that non-hyperaccumulator plant species can be used as phytoremediators. Based on the minor alterations in nutritional status, the high efficiency of the enzymatic antioxidant response and positive regulation of root proteins involved in carbohydrate, amino acids, and nitrogen metabolism, a positive scenario to overcome Cd-induced stress is established, making D. lablab L. a good candidate for phytostabilization purposes.