Antimony induced structural and ultrastructural changes in Trapa natans

Antimony (Sb) is considered as a priority toxic metalloid in the earth crust having no known biological function. The current study was carried out in a hydroponic experiment to study the accumulation of ecotoxic Sb in subcellular level, and to find out the ultrastructural damage caused by Sb in different vegetative parts of Trapa natans. Sb-induced structural and ultrastructural changes of T. natans were investigated using scanning electron microscope (SEM) and transmission electron microscope (TEM). Experimental plants were exposed to different Sb(III) treatments: SbT1 (1.5 μmol/L), SbT2 (40 μmol/L) and SbT3 (60 μmol/L). Calculated bioconcentration factor (BCF) and translocation factor (TF) showed that at higher concentration (SbT2, SbT3), T. natans is a potent phytoexcluder whereas it can translocate a substantial amount of Sb to the aerial parts at lower concentration (SbT1). SEM analysis revealed Sb-mediated structural changes in the size of stomatal aperture, intercellular spaces and vascular bundles of different vegetative tissues of T. natans. TEM results showed subcellular compartmentalization of Sb in vacuole and cell wall as electron dense deposition. This is considered as a part of strategy of T. natans to detoxify the deleterious effects under Sb stress conditions. Fourier transform infrared spectroscopy (FTIR) study of plant biomass revealed possible metabolites of T. natans which can bind Sb.


Results
Plant responses to Sb toxicity. No visual toxicity symptoms were observed in SbT1 and SbT2 groups of T. natans. On contrary to this, chlorosis and fragility in leaf were observed in SbT3 group of plants to a certain extent. In addition, other toxicity symptoms such as browning and twisting of roots were also visible in SbT3 group of T. natans.
Effect of Sb on total chlorophyll content. Chlorophyll content is often accepted as an indicator of environmental stress. Significant reduction (P < 0.05) in chlorophyll content associated with application of elevated concentration of Sb was observed in the present study. Decreased level of total chlorophyll content with increasing Sb concentration and exposure time was found in the experimental plant species (Fig. 2). This difference in chlorophyll content was significant (P < 0.05) according to ANOVA and Tukey HSD test, in T. natans grown in different Sb treatments (Supplementary Tables 1, 2).

Sb concentration in vegetative tissues. Result of Sb concentrations in different plant organs in all
three Sb treatments showed that Sb concentrations in roots were somewhat higher (P < 0.05) than those in leaves and stems (Table 1; Supplementary Tables 3, 4, 5, 6) except in treatment SbT1 where Sb concentration in stems was slightly higher than roots. Among the three treatments, highest Sb concentrations were observed in leaves (58.60 mg/kg), stems (50.24 mg/kg) and roots (73.66 mg/kg) of treatment SbT2. Plant species exhibiting BCF  www.nature.com/scientificreports/ and TF > 1 is considered as suitable plant for phytoextraction of metals/metalloids 38,39 . In the present study, calculated BCF values for all three Sb treatments were greater than 1 (Table 1). On contrary to this, TF values for all the Sb treatments were found < 1.
SEM X-ray microanalysis of leaf, stem and root tissues. Leaf, stem and root of Sb treated plant were examined by scanning electron microscope (SEM) along with energy dispersive X-ray (EDX) unit. This electron microscopy study revealed different Sb induced structural alterations of leaves, stems and roots of T. natans treated with SbT3. Decreased size of stomatal aperture was evident from the SEM micrographs of the abaxial side of Sb treated leaf (Fig. 3b, Supplementary Fig. 1b) with respect to the control leaf (Fig. 3a, Supplementary Fig. 1a). There was a significant difference of 3.73 µm in the stomatal diameter of control (8.80 µm) (Fig. 3a) and Sb treated leaves of T. natans (5.07 µm) (Fig. 3b). In EDX spectra of the control plant leaf ( Supplementary Fig. 1c) Sb was absent whereas presence of Sb was confirmed in the EDX spectra of Sb treated plants ( Supplementary Fig. 1d).
The SEM micrograph of stem of control plant group did not show any abnormality of the vascular bundles (Fig. 4a). Loss of shape and partial collapse of the xylem vessels in the SEM micrograph of treated stem confirmed Sb toxicity in T. natans (Fig. 4b). However, cell walls of xylem vessels were not collapsing completely. The EDX spectra clearly showed absence of Sb in transverse section of control stem ( Supplementary Fig. 2), whereas Sb weight percentage of 1.20, 1.97 and 0.45 were found in epidermis, cortex and metaxylem vessels, respectively ( Supplementary Fig. 3).
In comparison to the control plants (Fig. 5a), the roots of Sb treated T. natans plants showed damages of parenchyma cells (Fig. 5b) which lead to the alteration of cell shape and reduction of intercellular spaces. Metaxylem vessels were uniform in control roots; however, in treated roots metaxylems were partially collapsed resulting in loss of cell shape ( Supplementary Fig. 4). Absence of Sb in the control root was confirmed by EDX spectra (Supplementary Fig. 5). However, the EDX spectra of treated root revealed highest enrichment of Sb in the epidermis (1.55 wt%) followed by cortex (0.52 wt%) and metaxylem vessels (0.12 wt%) ( Supplementary Fig. 6).  www.nature.com/scientificreports/ TEM analysis of leaf and root tissues. The changes in the chloroplast structure of leaf cells of Sb treated T. natans were examined by transmission electron microscopy (TEM). The TEM micrographs of control leaf cells of the plant showed no abnormality or disorganization in the ultrastructural view of the chloroplast (Fig. 6a). But, the chloroplasts of the Sb treated plant cells showed disintegration of the inner membrane, disorganization of the structures of grana and stroma, and accumulation of starch (Fig. 6b,c). Ellipsoidal shapes of chloroplasts were observed in the treated plant cells (Fig. 6c). Several small vacuoles (without metal/metalloid deposition) were found in mature root cells of control plant ( Fig. 7a) whereas, the TEM micrograph of Sb treated root cell showed presence of vacuoles with electron dense precipitation (Fig. 7b). The cell wall of root cell did not exhibit any ultrastructural alteration in control T. natans (Fig. 8a). On contrary to this, some electron dense precipitates were observed in the intercellular spaces of the root cells which are generally adsorbed on the cell walls (Fig. 8b).
Fourier transform infrared spectroscopy (FTIR) analysis. FTIR spectra of the control and Sb treated plant biomass of T. natans are presented in Fig. 9 which displayed a number of absorption peaks in the region of 400-4000 cm −1 . Shifting of the peak position in the FTIR spectra of the Sb-loaded biomass with reference to that of the control plant biomass indicates the binding of Sb with different functional groups present in the biomass ( Table 2).

Discussion
Plant responses to Sb toxicity. Chlorosis of the leaves was observed in Sb treated T. natans plants (SbT3), which is considered as one of the most predominant symptoms of toxic metal/metalloid phytotoxicity 15 . The chlorophyll content of T. natans was found to be decreased with the increase in Sb concentration. Pan et al. 48 reported the impact of Sb on chlorophyll content and the photosynthetic efficiency of maize treated with Sb concentration of 10, 50, 100, 500 and 1000 mg/kg in the soil, and found that both parameters were negatively influenced only at higher concentrations. The cause of decrease in chlorophyll level could be due to the inhibition of two enzymes of chlorophyll biosynthesis, namely ferredoxin NADP + reductase and δ-aminolevulinic acid dehydratase (δ-ALAD) 49 . Liu et al. 50 documented that enhanced chlorophyllase action can inhibit photosynthesis in plants under stress conditions.

Sb concentration in vegetative tissues. Calculated BCF (255.22) and TF (0.96) values for T. natans
grown in treatment SbT1 evinced that although it falls in the category of phytostabilizer, a significant amount of Sb is translocated to shoots. However, as the initial concentration increased, TF values for treatments SbT2 (0.74) and SbT3 (0.50) decreased where the plant acts as a potent phytostabilizer, and exclusion of Sb from aerial vegetative tissues can be considered as Sb tolerant strategy of the plant species. Besides translocation of Sb from root to shoot, certain amount of Sb may have directly absorbed by the leaves of T. natans as the leaves are in contact with the solution. For example, Maine et al. 51 also reported that the direct contact of leaves with metal solution is the main cause of increase of chromium in the aerial parts of some floating macrophytes, although they are poor translocator of the metal from roots to the aerial parts. Variation of translocation factor (TF) and enrichment factor (EF) with metal concentration has been reported by Liu et al. 52 Literature studies showed that quite a few plants such as Hygrophila auriculata and Cyperus exaltatus (Pb), Sphaeranthus gomphrenoides, Pluchea dioscoridis and Cyperus articulates (Cd), Sphaeranthus gomphrenoides, Typha capensis, Pluchea dioscoridis (Ni) were identified as potential excluders of metals/metalloids from their substrate 53 .

Sb-induced structural and ultrastructural alterations.
Stomatal closure is one of the key molecular and physiological responses of plants to restrict water loss when plants are in stress. Sb toxicity causes stomatal closure in the leaves of the experimental plant species. Inadequate concentration of carbon dioxide via stomatal closure leads to photosynthesis inhibition 54 . An intricate system of signaling pathways regulates stomatal closure, where abscisic acid (ABA) plays the major role in association with jasmonic acid (JA), cytokinins, ethylene, and auxins 55 . Daszkowska-Golec and Szarejko 56 also reported that shrinkage of guard cells under stress conditions leads to the stomatal closure. In the present study, a significant damage to metaxylem vessels was observed.  www.nature.com/scientificreports/ Damage of vascular bundle, especially the xylem vessels, is of great concern as it forms an integrated network that connects all parts of the plant and is a principal water conducting tissue. Kasim 57,58 also reported decrease in diameter of metaxylem vessel in vascular bundle of leaf midrib which led to reduction in shoot growth in Zn treated Phaseolus vulgaris and, Cu and Cd treated Sorghum bicolor. However, there is contradicting report in this regard where increase in the diameter of metaxylem vessel was also observed in roots of Matricaria chamomilla during early exposure days of Pb treatment 59 . The most obvious consequence of the decrease in size of the metaxylem is the reduction of upward movement of water and mineral from root to shoot. The ultrastructural changes caused due to Sb in the vascular bundle may also modify the water status of leaves 60 , indicating declining of water level in leaves which in turn leads to stomatal closure. Moreover, researchers have reported toxic effects of metal in the epidermis and parenchyma cells in the cortex of root and stem in the form of disintegration of cells, loss of shape and size 57,58 . In our study, the highest weight percentage of Sb was observed in the epidermis region of stem. Similar result of very high concentration of Ni in epidermis area has also been reported in shoots of Alyssum inflatum 61 . Detection of Sb in the xylem vessels of root and shoot suggests transportation of Sb through xylem vessels from root to aerial parts. Although the shapes of the metaxylem vessels changed under Sb stress in the present study, xylem cell walls were remained intact which can be attributed to the heavily lignified composition of cell walls for providing mechanical strength as well as preventing collapse of vessels. Integrity of chloroplast ultrastructure is necessary for normal performing of photosynthesis in the plant cell as the whole process of photosynthesis is completed in the chloroplasts. Our experimental results revealed disorganization of chloroplast ultrastructure in Sb treated plants. Researchers have shown that toxic metals/ metalloids affect the chloroplast structure by degradation of grana and stroma lamellae along with enhancement of the quantity and dimension of plastoglobuli 62 . Starch accumulation was also evident within the chloroplast  However, on contrary to the higher plants, 50-60% Cd was reported in the chloroplast of cell-wall deficient Chlamydomonas reinhardtii and plant vacuole deficient Euglena gracilis 69,70 . Although, above results indicate capability of chloroplasts to accumulate toxic metals, but generally this is not the target organelle of plant cell having cell-wall and vacuole for accumulation of metal/metalloid. The results of our study also indicate that Sb accumulation in vacuoles is the most effective system for maintaining a very low cytoplasmic Sb concentration in T. natans. Cell wall also played an important role in the storage of metals/metalloids as it acts as the first barrier for entry of metal/metalloid into symplastic compartments of the cell and one of our recent publications also showed that Cd was mainly accumulated in the cell wall 71 . The binding of Sb in the cell wall and Sb compartmentalization in the cytosol have been proposed as possible mechanisms for detoxification in Sb hyperaccumulator ferns like Pteris cretica and Pteris fauriei 13 . This is consistent with our results as TEM micrographs of the Sb treated root    www.nature.com/scientificreports/ cells of T. natans confirmed electron dense deposition of Sb in the vacuoles. As stated earlier, the storage and Sb deposition as fine precipitates in the vacuoles of root cells indicate Sb detoxification mechanism of cell in this experimental plant species preventing high concentration of the metalloid in the cytosol 72 . It is also reported that intracellular ligands such as phytochelatins (PCs) and metallothioneins (MTs) play significant role in the metal detoxification mechanism 73 . Various metal transporter protein families such as ATP-binding cassette (ABC), cation diffusion facilitators (CDF), heavy-metal P-type ATPases (HMA), and natural resistance-associated macrophage proteins (NRAMP) are involved in this detoxification process. The ABC transporters are reported to be involved in toxic metal/metalloid transport into the vacuole and among these transporters, two subfamilies, viz. the multidrug resistance-associated proteins (MRP) and pleiotropic drug resistance (PDR), are mainly active in this process 74 . The observed Sb depositions along the cell wall of the root cells of Sb treated T. natans could be considered as one of Sb tolerance mechanisms. The importance of cell wall in binding of toxic metals in plant cell has already been documented 15,75 . The binding of cationic elements to the negative-charged pectic compounds (e.g. galacturonic acid) in cell wall takes place through passive ion exchange process 76 . Under stress conditions breakage of membrane allows binding of more cations in the newly exposed protein sites resulting in higher cation exchange capacity (CEC) in the plant cell 77 .  www.nature.com/scientificreports/ analysis different metabolites such as proteins, lipids, carbohydrates and secondary metabolites, especially flavonoids, tannins, saponins are found as possible plant constituents responsible for binding of Sb in the biomass of T. natans.

Conclusions
The present study is an appraisal of the structural as well as ultrastructural alterations in the different organelles of T. natans due to subcellular accumulation of Sb. This study will also help to shed some light on the accumulation pattern and ultrastructural modifications of T. natans under Sb toxicity. In the electron microscopy studies, Sb was found as electron dense precipitate mainly in cell wall and vacuoles which can be considered as Sb tolerant mechanism of the studied plant species. Besides, accumulation of starch in the chloroplast, disorganization of chloroplast ultrastructure, change of shape of chloroplasts were some of the noticeable changes that were evident in T. natans due to Sb stress. FTIR analysis confirmed the possible functional groups of various metabolites present in the biomass of the plant species for binding of Sb ions. It is evident from the present study that in general T. natans is a suitable plant for phytostabilization of Sb by storing proportionately higher amount of the element in the rhizosphere restricting their mobilization and rendering them harmless. But at low concentration substantial amount of Sb was translocated to the stem and leaf. Although the plant cannot be clearly classified as phytoextractor of Sb at lower Sb concentration (as the TF slightly less than 1), but can be considered as a borderline case. Thus, T. natans which is abundantly available in many natural wetlands of Assam can be regarded as one potential candidate for phytoremediation of Sb.

Materials and methods
Description of macrophyte. T. natans (Water chestnut) is an herbaceous aquatic floating-leaved macrophyte with a rosette of floating leaf belonging to the monogeneric family Trapaceae that grows plentifully in the freshwater lakes of Assam (Fig. 1). The plant is also grown commercially in many parts of India for its nutrient rich edible seeds. T. natans of similar size and weight were collected from Joysagar Pond of Sibasagar District, Assam for the experimental purpose and washed thoroughly in running tap water and deionized water to avoid any surface contamination. Preparation of antimony stock solution and experimental design. Antimony (SbIII) stock solution of 100 μmol/L was prepared by dissolving SbCl 3 (Merck analytical grade) in Milli-Q (18.2 MΩ cm conductivity) water. An outline of the experimental design is presented in Supplementary Fig. 7. The acclimatized plants were transferred to 4 sets of 2 L opaque non-reactive round plastic containers (depth 10 cm and diameter 28 cm) having three tubs for each set. Set I was considered as control group (SbT0) containing 2 L of 0.2X HS without addition of Sb(III) solution. Plants with uniform size were placed in set II, III and IV filled with 0.2 X HS, but supplemented with desired amount of Sb(III) stock solution for making final Sb(III) concentrations of 1.5 μmol/L (SbT1), 40 μmol/L (SbT2) and 60 μmol/L (SbT3), respectively. Plants were allowed to remain in contact with the treatment solutions for 10 days under natural photoperiod and temperature. The pH of the treatment solutions was adjusted to 6.0 with dilute NaOH or HCl. All the treatment groups were arranged in a completely randomized design and three replicates of experimental plant species were set up for each treatment group. After 10 days of experimental period, plants were harvested, washed with Milli-Q water and separated into leaves, stems and roots. The loss of water volume in the containers due to evapo-transpiration was maintained by adding deionized water. A pilot study was conducted in order to design the concentration of the Sb in the experiment. Three different concentrations of Sb(III) were chosen considering the fact of pilot study that the experimental plant species was found to be tolerant to these concentrations to a great extent, although plants exposed to SbT3 showed some visual toxicity symptoms in the study.

Hoagland's nutrient solution.
Chlorophyll content measurement. Leaf chlorophyll content of T. natans was estimated according to Arnon 82 on 3rd, 5th and 10th day of the experiment. Fresh plant leaf (0.5 g) was homogenized with mortar and pestle, and extracted in 10 mL of 80% chilled acetone. The absorbance of the extract was recorded in UV-visible spectrophotometer at wavelengths of 645 and 663 nm.

Analysis for Sb bioaccumulation in plant vegetative tissues.
To determine the Sb accumulation efficiency in different parts of T. natans, leaves, stems and roots were separated. Drying of harvested plant materials was carried out using standard protocol 21 . The plant materials were digested with a mixture of HNO 3 (69%) and H 2 O 2 (30%) in microwave digestion system (Anton Paar Multiwave GO) at 105 °C for 30 min and then cooled for 3 min. Samples were diluted to a final volume of 12.5 mL with deionized water 83 . The blank and certified reference material (SRM-1573a tomato leaves, NIST standard reference material, USA) were used for quality control. Sb concentrations in the digested samples were measured by ICP-OES (Optima 2100, Perkin Elmer) with detection limit 3 µg/L. www.nature.com/scientificreports/ Procedure for microscopic study. For the microscopic study leaf samples from control group (SbT0) and Sb treated group (SbT3) were cut out from the middle portion of the leaf whereas root and stem samples were excised from 2 cm below and above, respectively, the shoot-root intersection. For scanning electron microscopy analysis (SEM) small pieces of leaves, stems and roots (3-4 mm) were instantly fixed in 3% glutaraldehyde prepared in 0.05 M phosphate buffer for duration of 90 min, followed by secondary fixation in 2% osmium tetroxide prepared in 0.01 M sodium cacodylate buffer for 30 min 84 . Samples were dehydrated in a graded acetone series (30-100%, v/v). SEM photographs were obtained using SEM model JEOL-JSM-6390 LV along with EDX unit, with an accelerating voltage of 15 kV and 20 kV.
The sample preparation protocol for transmission electron microscopy analysis (TEM) involves fixation, sectioning and staining of the sample 85 . Leaf and root samples were fixed in 2.5% glutaraldehyde in 0.05 M potassium phosphate buffer (pH 7.1) for 3 h. Osmium tetroxide was used as a stain and fixative for studying morphology in biological electron microscopy. Then the samples were dehydrated in a graded ethanol series (30-100%, v/v) and embedded in Spurrs epoxy resin. Ultrathin sections of the blocks were obtained using an ultramicrotome. Sections were post-stained with basic lead citrate and uranyl acetate for microscopic examination using JEOL TEM.
Fourier transform infrared spectroscopy (FTIR) analysis. FTIR spectroscopy was used to elucidate different functional groups responsible for binding of Sb ions in leaves, stems and roots of SbT0 and SbT3 group of T. natans. For this purpose, different parts of T. natans obtained after experimentation were freeze-dried for 24 h using a laboratory freeze-dryer (LSI LyoLab) to preserve its bioactive components. The leaf, stem and root powder (0.0035 g) of the freeze dried biomass were mixed with KBr (0.5 g) as the base material to form pellets 80,86 and FTIR spectra (400-4000 cm −1 ) were obtained using FTIR spectrometer (Perkin Elmer Spectrum100). FTIR spectra of plant samples before and after absorption were compared.

Data analysis. Bioconcentration factor (BCF). Bioconcentration factor (BCF) is a ratio which indicates
the capacity of the plant to bioaccumulate a specific metal in roots in regards to metal concentration in the medium 87 , and was calculated using the following formula: Translocation factor (TF). Translocation factor (TF) was calculated to assess the ability of the experimental plant species for translocation of the metal from the roots to the aerial parts 35 . The following formula was used to calculate TF 88 : Statistical analysis. All the results of the experiment were presented as mean ± standard deviation (SD) of three (n = 3) replicates. One-way analysis of variance (one-way ANOVA) followed by post hoc test (Tukey's Honestly Significant Difference test) was performed for all the measured variables using SPSS 23.0 to check the significant difference in the evaluated parameters of the Sb treated plants with respect to the control plants. A probability of 0.05 was considered as significant for evaluation of critical values differences.