Introduction

Chromium pollution is a serious problem in many countries with industrial activities related to leather tanning, cement production, stainless steel manufacturing, dye pigments, chrome plating, and chromite mining (Coetzee et al. 2020). Chromium is released to environment mainly as hexavalent [Cr(VI)] and trivalent [Cr(III)] forms, being the first one more toxic to animals and plants (Shahid et al. 2017). The solubility and mobility of chromium in the environment depend on its concentration and oxidation state. In addition, Cr(VI) is more soluble than Cr(III) and can remain for long time periods without reduction to less toxic Cr(III) form (Sharma et al. 2020a). Mostly, Cr(VI) is discharged as wastewater into aquatic systems producing a large contamination in the world, but especially in so-called third-world countries (Sharma et al. 2012). Thus, many attempts have been made to remove Cr(VI) from polluted waters, which are of great importance in terms of environmental protection and human health (Nur-E-Alam et al. 2020). Although different procedures have been developed to remove Cr(VI) from polluted waters, many of them are expensive, difficult to applied, and not environmentally friendly (Malaviya et al. 2020). However, hydroponically grown plants can directly remove toxic metals from polluted waters in a process termed phytoremediation (Ali et al. 2020). Plants used in phytoremediation must have a considerable capacity to absorb and accumulate metals and a high metal tolerance to decrease the treatment time of polluted water (Stephenson and Black 2014). Among aquatic plants used in heavy metal phytoremediation, Salvinia species have been recognized as useful tool to remove Cr(VI) from polluted waters in many tropical and subtropical regions of the world (Dhir 2009). They are free-floating aquatic ferns, with small, spongy, floating green leaves (fronds) positioned in pairs along a common stem and brown root-like submerged leaves (lacinias) (Kumari et al. 2016).

Aquatic plants can remove Cr(VI) from polluted waters by both external (surface adsorption and/or absorption) and internal (accumulation inside tissues) mechanisms (Augustynowicz et al. 2020). In this sense, chemical reactions such as Cr(VI) reduction and/or complexation with organic acids have also been reported (Espinoza-Quiñones et al. 2009; Gomes et al. 2017). Particularly, the ability to reduce Cr(VI) to Cr(III) has been demonstrated in several aquatic and terrestrial plant species (Lytle et al. 1998; Espinoza-Quiñones et al. 2009; Duarte et al. 2012; Augustynowicz et al. 2013a, b). In microorganisms, reduction mechanisms of Cr(VI) include enzymatic and/or non-enzymatic pathways (Cheung and Gu 2007). Enzymes catalyzing Cr(VI) reduction have been isolated and characterized from different bacteria genus (Ahemad 2014; Baldiris et al. 2018). By contrast, specific enzymes that catalyze Cr(VI) reduction have not yet been detected in plants, but due to the wide variety of oxido-reductase-type enzymes that plants possess, it could be thought that one or more of these enzymes could carry out the reduction of Cr(VI) (Pradedova et al. 2017). Also, Cr(VI) reduction can occur by a non-enzymatic pathway through reactions involving different plant metabolites such as soluble thiols, phenolic compounds, and organic acids (Sharma et al. 2021). However, it is still unclear what is the functional mechanism to achieve the reduction in plants. Thus, the aim of the present study was to investigate in S. minima plants exposed to Cr (VI) the presence of Cr (III) by means of XANES analysis and the possible implication of an enzymatic reduction and a non-enzymatic pathway, which involves thiols, phenolic compounds, and the expression of a thioredoxin.

Materials and methods

Plant material and metal treatment

Healthy S. minima plants with uniform size were collected from an unpolluted 50-year-old man-made pond located in Tucuman province (26° 50′ S, 65° 12′ W). Plants were thoroughly washed with running tap water to remove plant debris and sediment particles. After washing, plants were transferred to plastic containers (30 L capacity) containing 1:10 Hoagland solution for 3 days under outdoor conditions to generate new and rapidly growing leaves (Prado et al. 2010). Next, plants with uniform size of lacinias and fully expanded fronds were washed with distilled water, blotted on filter paper, and transferred to plastic trays (10 × 8 × 4 cm) containing 150 mL of 20 mg L−1 Cr(VI) solution. Twenty plants (∼35 g, FW) per tray were used. Cr(VI) solution was prepared in distilled water. It is noteworthy that Cr(VI) solution was not prepared in Hoagland’s solution to avoid chelation and/or ion competition for cell wall binding sites between Hoagland ions and Cr(VI) ionic species. The pH of freshly prepared Cr(VI) solution was 6.7, ranging between 6.6 and 6.8 during the cultivation period. After planting, trays were transferred to a growth chamber for 7 days under controlled conditions: 200 μmol m−2 s−1 photosynthetic active radiation, 16 h:8-h day/night light regime, 27 °C:22 °C day/night temperature, and 80 % relative humidity. Cultivation period was chosen based on preliminary assays carried out in our laboratory indicating that S. minima plants can well grow for at least 9 days in distilled water without nutrient supply (Prado et al. 2010). Water loss by evapotranspiration was compensated daily adding distilled water up to initial volume. To avoid an excessive change in Cr(VI) concentration, the treatment solution was renewed totally 3 days after the cultivation started. After metal treatment, plants were harvested, rinsed in distilled water, and cut to obtain fronds and lacinias. Fresh weight (FW) was determined by weighing fronds and lacinias immediately after cutting, whereas the dry weight (DW) was obtained after drying at 80 °C in a hot air oven for 2 days.

Metal accumulation

Powdered dried fronds and lacinias weighing approximately 0.5 g were ashed in a muffle furnace at 450 °C for 5 h. Ashed samples were digested in a mixture of HNO3/HClO4 (3:1, v/v) at 115 °C for 15 min following the USEPA 3051 protocol (USEPA 1994). Chromium concentration in the digested samples was analyzed for chromium by flame atomic absorption spectrometry (FAAS) using a Perkin–Elmer 373 spectrophotometer. The conditions for the determination were 357.9-nm wavelength with a slit of 0.7nm and a relative noise of 1.0 with acetylene/air flame. The calibration curve was constructed from 1000 ppm AccuStandard in distilled water with a linearity range up to 5 mg/L. Every ten samples, the 4 mg/L standard that should produce a signal of 0.2 absorbance units was read, under the measurement conditions. This value was used as the “Characteristic Concentration Check Value” and allowed to monitor the operation of the equipment, according to the Manual of AA Atomic Spectrometry -Perkin Elmer. Metal content was expressed as mg g−1 DW.

XANES analysis

X-ray absorption near edge structure spectra (XANES) around the Cr K-edge (5989 eV) of freeze-dried samples of lacinias and fronds as well as Cr reference compounds (containing Cr(III) in the form of Cr2O3 and Cr(NO3)3, Cr(VI) in the form of K2Cr2O7, and Cr(0) (elemental chromium) as a Cr metal foil) were acquired at the XAFS2 beamline in the Brazilian Synchrotron Light Laboratory (LNLS, Campinas Brazil) (Figueroa et al. 2016). Measurements were performed in fluorescence mode using a Si(111) crystal monochromator in the energy range from 5950 to 6200 eV with a ion chamber as Io detector and a Germanium 15 elements detector to collect the fluorescence signal (Mirion Technologies [Canberra], Inc., USA). Plant samples were placed in sample holders with Kapton® windows. The absorption of Cr metal foil and reference compounds were measured in transmission mode using two ion chambers as detectors X-ray absorption spectra were normalized by standard methods using the ATHENA software (version 2.1.1) which is part of the IFEFFIT package to obtained normalized XANES spectra (Ravela and Newville 2005). Three scans of each studied sample were measured to improve the quality of the data.

Total thiols (TT), non-protein thiols (NPT), and protein-bound thiols (PBT)

Samples of fronds and lacinias (1.0 g FW) were homogenized with 3 mL of 20 mM Tris-HCl buffer (pH 8.6) containing 1 mM dithiothreitol (DDT) in a cooled mortar and pestle (Linde and Garcia-Vazquez 2006). Homogenized extracts were centrifuged at 14000 g for 10 min at 4 °C and supernatants used for determinations of TT and NPT. To determine TT, aliquots of supernatants (0.2 mL) were mixed with 0.6 mL of Ellman’s reagent (5 mM DTNB, [5,5′-dithiobis-(2-nitrobenzoic acid)], in 100 mM Tris-HCl buffer, pH 8.0), 0.8 mL of 100 mM Tris-HCl buffer (pH 8.0), and 0.5 mL of distilled water. After incubation at 37 °C for 60 min, the absorbance was read at 412 nm against a blank without supernatant. To determine NPT, aliquots of supernatants (0.5 mL) were mixed with 0.1 mL of trichloroacetic acid (TCA) 50% (w:v) and 0.4 mL of distilled water and maintained in ice for 20 min. After centrifugation at 14000 g for 10 min at 4 °C, supernatants were used to determine NPT as described above. PBT was calculated by subtracting NPT from total thiols (TT). Thiol concentrations were calculated from a calibration curve prepared with pure reduced glutathione (GSH) and expressed as μmol GSH equ. g−1 FW (equ. = equivalent).

Thioredoxin h (Trx h) expression

Total mRNA of fronds and lacinias from both control and Cr-exposed plants were obtained from 0.5 g FW of plant samples following the magnetic streptavidin particles method (mRNA Isolation Kit - Roche Molecular Biochemicals, Gmbh, Mannheim, Germany) (Pagano et al. 2000). The synthesis of the corresponding cDNA was performed by retrotranscription using a viral retrotranscriptase that synthesizes DNA from an RNA template (RevertAid M-MuLV reverse transcriptase system - Fermentas International Inc., Burlington, Ontario, Canada). Amplification of cDNA fragments was carried out by the polymerase chain reaction (PCR) in a thermocycler (MyCycler, BIO-RAD, Hercules, CA, USA). The separation and identification of amplified cDNA fragments was performed by horizontal electrophoresis on 1.5 % agarose gel (McDonell et al. 1977). Obtained gel was stained with the SYBR Green reagent (Invitrogen, Carlsbad, CA, USA), visualized by transillumination and analyzed using the UVP Doc-It LS Image Acquisition Software system (Ultra-Violet Products Ltd. Cambridge, England). To sequence cDNA fragments obtained by PCR, they were cloned into pGEM-T vector (Promega Corp., Madison, WI, USA).

Cr(VI) reduction by fronds and lacinias extracts

Fronds and lacinias extracts were prepared from healthy plants of S. minima maintained in distilled water without Cr(VI) for 2 days. Briefly, 1 g FW plant samples were homogenized with 3 mL of 0.2 M sodium phosphate buffer pH 7.0 using a pre-chilled mortar and pestle. Resulting homogenates were centrifuged at 14000 g for 15 min, at 4 °C, and supernatants were collected. The capacity of fronds and lacinias extracts to reduce Cr(VI) was established by measured residual Cr(VI) in the reaction mixture after incubation a 37 °C. Briefly, assay mixture (1 mL) containing K2Cr2O7 to final concentration of 60 μM, 0.6 mL of 0.2 M sodium phosphate buffer, pH 7.0, and 0.4 mL of supernatant aliquot was incubated at 37 °C. After incubation, tubes were boiled in a water bath for 2 min. Remaining Cr(VI) in the reaction mixture was determined with 1.5-diphenylcarbazide reagent (Memon et al. 2005). Cr(VI) concentration was calculated from a 10–100 μM K2Cr2O7 standard curve. Reaction mixtures without supernatant and without Cr(VI) were also incubated. To ensure that there was no spontaneous Cr(VI) reduction and/or Cr(VI) adsorption to tube walls, tubes containing Cr(VI) solution only were also incubated. After each incubation time, the Cr(VI) was determined with 1.5-diphenylcarbazide. Results did not show differences between 0- and 120-min incubation periods. By assuming that 100 % Cr(VI) concentration corresponds to 0 min, the reduction percentage of Cr(VI) was determined using the following equation:

$$ \boldsymbol{R}\%=\mathbf{1}-\left(\frac{\boldsymbol{Y}}{\boldsymbol{X}}\mathbf{1}\mathbf{00}\right) $$

where R% indicated the percentage of reduced Cr(VI), Y corresponds to Cr(VI) concentration after the incubation (t min), and X represents the initial Cr(VI) concentration (0 min).

Soluble phenolics (SP) and insoluble phenolics (IP)

SP extraction was made with ethanol 96 % according to Swain and Hillis (1959) with minor modifications. Briefly, samples of fronds and lacinias (1.0 g FW) were extracted with 3 mL 96 % ethanol, incubated in darkness at room temperature for 48 h, and centrifuged a 3000 g for 5 min. Supernatants were recovered and used to SP determination. Aliquots of supernatants (0.1 mL) were added with 0.2 mL (1:1 v/v) of Folin–Ciocalteu reagent and 1.8 mL of distilled water. After standing at room temperature for 2 min, 0.8 mL of 7.5 % Na2CO3 was added and left standing again at room temperature for 5 min. Next the absorbance was read at 760 nm. Precipitates from SP extraction were washed twice with 2 mL ethanol 96% and centrifugation at 3000 g for 5 min. Washed precipitates were dried a 37 °C for 48 h and used to obtain IP (cell wall-bound phenolics). IP extraction was adapted from Assabgui et al. (1993). Dried samples (0.5 g) were hydrolyzed with 2 mL of 2 N NaOH in a water bath at 60 °C for 60 min. After cooling, solutions were slowly acidified up to pH 2.0 with 5 N HCl and extracted with ethyl acetate. Following ethyl acetate fractions were taken near dryness under a stream of N2 gas and dissolved in 0.5 mL of 96 % ethanol. Solubilized phenolics were determined using the Folin–Ciocalteu reagent as described above. Concentrations of SP and IP were determined using a standard curve made with pure phenol and expressed as μmol phenol equ. g−1 FW.

Statistical analysis

For all determinations, at least three replicates were analyzed, and two independent experiments were performed. Results are presented as the mean of all replicates, and bars represent standard error (SE). Significant differences in numerical results were established by using one-way analysis of variance (ANOVA) and treatment means compared by the Tukey’s multiple comparison test at p < 0.05.

Results

Metal accumulation

Chromium accumulation in lacinias and fronds of S. minima plants after 7 days of Cr(VI) exposure is shown in Fig. 1. The highest accumulation of metal was observed in lacinias reaching 4.43 mg Cr g−1 DW, while in fronds accumulation metal was very low (0.98 mg Cr g−1 DW). To confirm that Cr(VI) was the unique oxidation state present in the treatment solution, it was determined at the beginning and ending of the experimental period in presence and absence of KMnO4 (Memon et al. 2005). No differences were observed between two determinations (not shown).

Fig. 1
figure 1

Chromium accumulation in fronds and lacinias of S. minima plants exposed for 7 days to 20 mg L−1 Cr(VI) concentration. Data are mean ± SE of six replications from two independent experiments. Different lowercase letters on bars indicate significant differences between lacinias and fronds (p < 0.05)

Full size image

XANES analysis

Figure 2 shows XANES spectra of Cr accumulated in fronds and lacinias as well as those corresponding to Cr(0), Cr(III), and Cr(VI) standard compounds. The spectra of the reference samples [Cr(0), Cr(III) and Cr(IV)] present very particular characteristics in each case that makes it very easy to distinguish. In particular, XANES spectrum of Cr(VI) standard exhibits a characteristic pre-edge peak from 5980 to 6000 eV, while in Cr(III) and Cr(0) standards this feature is missing. After a 7-d incubation period, XANES spectra of plant samples do not show the characteristic Cr(VI) pre-edge peak. On the other hand, plant samples show the energy of their absorption edge (first inflection point of the curve) in accordance with the presence of Cr(III), indicating that is the unique oxidation state of chromium accumulated in Salvinia plants.

Fig. 2
figure 2

XANES spectra of chromium accumulated in lacinias and fronds of S. minima plants exposed for 7 days to 20 mg L−1 Cr(VI) concentration, as well as Cr(III) and Cr(VI) standard compounds

Full size image

Total thiols (TT), non-protein thiols (NPT), and protein-bound thiols (PBT)

Concentrations of TT, PBT, and NPT were significantly higher in lacinias than in fronds of Cr-untreated and Cr-exposed plants (Fig. 3). However, Cr(VI) exposure resulted in significant increases of all thiol species in both tissues. In Cr-exposed plants, thiol values for fronds and lacinias ranged between 0.93 and 3.69 μmol GSH equ. g−1 FW and 1.73 and 5.27 μmol GSH equ. g−1 FW, respectively. Maximum increases of thiols were 210.5 % (TT), 223.5 % (PBT), and 86.0 % (NPT) and were observed in fronds.

Fig. 3
figure 3

Accumulation of TT, NPT, and PBT in fronds and lacinias of S. minima plants exposed for 7 days to 20 mg L−1 Cr(VI) concentration. Data are mean ± SE of six replications from two independent experiments. For each evaluated thiols, different lowercase letters on bars indicate significant differences between lacinias and fronds (p < 0.05)

Full size image

Thioredoxin h (Trx h) expression

The expression of an h-type thioredoxin in lacinias of S. minima plants exposed to Cr(VI) is shown in Fig. 4. Using specific primers PsH1F (GTGCATATGGCAGGTTCATCAGAAGAG) and PsH1R (GTGGATCCCTAAGCATTAGATGAAGCC), designed from the sequence of pea thioredoxin Trx h1 (PsTRXh1, AJ310990) and cDNA of S. minima, a complete sequence of a 377 bp fragment of Trx h1 and two partial sequences of 266 and 288 bp fragments of Trx h2 type were obtained from floating and submerged leaves. According to the BLASTN nucleotide sequence program, a program of the BLAST family (Basic Local Alignment Search Tool), the 377 bp fragment exhibited high homology (98 %) with Trx h mRNA of Pisum sativum (Trxh gene, AJ31099.1), 91 % homology with Trx h1-1 mRNA of Galega orientalis (Trxh1-1 gene, HQ446218.1), and 87 % homology with Trx h1 mRNA of Glycine max (Trxh1 gene, EU144127.1). Likewise, the 266 bp Trx h1 partial fragment showed high homology (99 %) with Trx h isoform 1 mRNA of Hordeum vulgare subsp. vulgare, clone NIASHv3019I20 (Trxh1-1 gene, AK373063.1), 99 % with Trx h isoform 2 mRNA of Hordeum vulgare subsp. vulgare (Trxh1-2 gene, AY245455.1), and 94 % with Trx H mRNA of Triticum aestivum cv. Michael (TrXH gene, EU584496.1). The 288 bp Trx h2 partial fragment showed high homology (99 %) with Trx h isoform 2 mRNA of Hordeum vulgare subsp. vulgare, clone NIASHv1045G18 (Trx1-2 gene, AK357069), 96 % with Trx H mRNA of Triticum aestivum (TrxH gene, AF420472.2), 88 % with Trx h mRNA of Oryza sativa Japonica Group RTRXH2 (Trxh gene, AB053294.1), and 87 % with Trx H-type, clone 325551, mRNA of Zea mays (TrxH gene, EU969009.1) (data not shown).

Fig. 4
figure 4

Expression of h-type Trx gene in fronds and lacinias of S. minima plants exposed to 20 mg L−1 Cr(VI) during 7 days. F, fronds; L, lacinias; M, DNA marker. The gel is representative of three independent replicates

Full size image

Cr(VI) reduction by fronds and lacinias extracts

The reduction of Cr(VI) by extracts of fronds and lacinias is showed in Fig. 5. Cr(VI) reduction percentage was higher with fronds extract than with those of lacinias. Maximum reduction percentages 16.5 % (fronds) and 12.5 % (lacinias) were observed after a 120-min incubation period. Longer incubation times did not show increases in Cr (VI) reduction (data not shown).

Fig. 5
figure 5

Time dependence of Cr(VI) reduction by extracts obtained from fronds and lacinias of Cr-untreated S. minima plants. Data are mean of six replicates from two independent experiments. Bars indicate standard error

Full size image

Soluble phenolics (SP) and insoluble phenolics (IP)

Concentrations of SP and IP in fronds and lacinias increased under Cr(VI) exposure (Fig. 6). In fronds and lacinias of Cr-treated plants, SP contents were strongly increases. Maximum increases were 93.1 % (fronds) and 199.9 % (lacinias), respectively. The IP concentration in Cr-untreated plants was slightly higher in fronds than in lacinias, whereas in plants exposed to Cr(VI) was higher in lacinias. Maximum increases occurring in Cr-treated plants were 121.1 % (fronds) and 183.7 % (lacinias).

Fig. 6
figure 6

Accumulation of soluble (SP) and insoluble (IP) phenolics in fronds and lacinias of S. minima plants exposed for 7 days to 20 mg L−1 Cr(VI) concentration. Data are mean ± SE of six replications from two independent experiments. For each evaluated phenolics, different lowercase letters on bars indicate significant differences between lacinias and fronds (p < 0.05)

Full size image

Discussion

It is generally accepted that excessive uptake of Cr(VI) affects the plant development and provokes several morpho-physiological and biochemical disorders (Sharma et al. 2021). Thus, plants able to accumulate high amounts of Cr(VI) must have efficient mechanisms to counteract the metal toxicity and also limit its transport to leaves. In this context, the root-shoot mobility of Cr(VI) in Cr-accumulating plants is very low and sometimes its concentration in roots being more 100 times higher than in shoots (Caldelas et al. 2012). In agreement with these findings, our results demonstrated that Cr(VI) is mostly accumulated in lacinias (submerged roots) (Fig. 1). Similar results were also communicated for other Salvinia species such as S. rotundifolia, S. auriculata, S. natans, and S. herzogii (Espinoza-Quiñones et al. 2009; Thilakar et al. 2012; Maine et al. 2016; Prado et al. 2021)

It is well known that the distribution and translocation of Cr(VI) within plants depend upon the plant species and the metal concentration in the growth medium (Shahid et al. 2017). However, the oxidation state of Cr also plays an important role in the translocation process (Sharma et al. 2021). The Cr(III) has a tendency to bind to functional groups of cell walls (e.g., –COOH, –OH, and –SH) of root cells, which interfere with its mobilization and transport within the plant (Augustynowicz et al. 2013a, b). In this way, it is expected that the reduction of Cr(VI) to Cr(III) occurs in plants able to absorb high levels of Cr(VI), without implying marked changes in metabolic processes. It has been suggested that while the reduction of Cr(VI) to Cr(III) mainly occurs in root tissues, this can also take place in leaf tissues (Aldrich et al. 2003). In agreement with this assumption, we demonstrated by XANES analysis the ability of S. minima plants to reduce Cr(VI). XANES spectra of both lacinias and fronds showed the lack of the pre-peak and positive shift along the energy axis when compared with Cr(VI) reference spectrum (Fig. 2). This clearly indicated that Cr(III) was the unique chemical form of Cr occurring in fronds and lacinias of Salvinia plants exposed to Cr(VI). The reduction of Cr(VI) to Cr(III) could also occur outside of roots in the rhizosphere (Pradas del Real et al. 2014); however, in our study, the Cr(III) was not found in the growth solution. This indicates that the reduction of Cr(VI) takes place inside S. minima tissues. A similar result was communicated for the aquatic plant Callitriche cophocarpa (Augustynowicz et al. 2013a, b). Reduction of Cr(VI) to Cr(III) in lacinias of S. auriculata has also been communicated (Espinoza-Quiñones et al. 2009); however, until now, the biochemical mechanisms underlying the reduction of Cr(VI) have not been thoroughly analyzed for S. minima.

Thiols, ascorbic acid, phenolics, lipoic acid, diol-containing molecules and carbohydrates (e.g. ribose, fructose and arabinose), as well as enzymes (e.g. Fe(III)-reductase and quinone reductase) have been shown to reduce Cr(VI) (Saha et al. 2011; Santana et al. 2012). This suggest that there could be different mechanisms operating in Cr(VI) reduction al. 2012; Kaszycki et al. 2018). In this context, our data show significant accumulations of thiols as well as soluble and insoluble phenolics in Cr-exposed fronds and lacinias; however, accumulation patterns were different indicating probably different roles in Cr(VI) detoxification.

Thiols play an important role in the protection of plants against Cr(VI) toxicity by protecting labile macromolecules against attack of free radicals generated during Cr-induced oxidative stress (Zagorchev et al. 2013). Glutathione (GSH) is a low-molecular-weight thiol that plays a central role in the scavenging of metal-induced ROS (reactive oxygen species) through the GSH-ascorbate cycle (Hernández et al. 2015; Sharma et al. 2020b; Sharma et al. 2020c). Furthermore, Levina and Lay (2004) have proposed that GSH is directly involved in the cellular metabolism of Cr(VI) compounds through the reduction to less toxic Cr(III) form. On the other hand, protein thiols such as thioredoxins (Trxs) are considered as protective metabolites and participate in regulatory mechanisms in plants exposed to heavy metals (Zagorchev et al. 2013). Trxs are protein-disulfide reductases that perform multiple functions related to cellular redox homoeostasis. The modulation of thiol-dependent redox system by metal ions via Trx and Grx (glutaredoxin) systems has been demonstrated in mammalian cells and bacteria (Ouyang et al. 2018). In this regard, the involvement of Trxs in the reduction of Cr(VI) has been demonstrated in a strain of Streptomyces violaceoruber isolated from the Yellow River at the People’s Republic of China (Chen et al. 2014). Consistent with this finding, we demonstrated the expression of an h-type thioredoxin in lacinias of S. minima plants exposed to Cr(VI) (Fig. 4). This could indicate that Trxs are also involved in Cr(VI) tolerance mechanism operating in plants able to accumulate high amounts of this metal. However, the h-type thioredoxins constitute the most complex cluster of Trxs, due to the identity and functions of Trx h members that are still largely unknown (Pivato et al. 2014; Boubakri et al. 2019). Thus, complementary studies are needed to get better knowledge on thiol accumulation, expression of h-type Trxs, and metal accumulation in Cr-exposed plants.

The reduction capacity of fronds and lacinias extracts of Salvinia species for the conversion of Cr(VI) to Cr(III) has been largely unexplored. Our data showed that S. minima crude extracts were able to reduce Cr(VI). Plant extracts usually contain sugars, glycosides, phenolic compounds, polyphenols, and even vitamins. Hydroxyl groups from sugars, phenolics, and polyphenols can affect the reduction state of heavy metals (Dubey et al. 2017). In this sense, several authors have demonstrated that extracts of both lacinias and fronds of Salvinia species are rich in phenolics and polyphenols (Devi et al. 2015; Gini and Jothi 2018). Consistent with these findings, our data showed high contents of SP and IP in fronds of Cr-untreated S. minima plants, which was coincident with their higher capacity to reduce Cr(VI) in vitro. In addition, SP and IP increased significantly in Cr-treated plants, suggesting a role in Cr(VI) reduction (Fig. 6). On the other hand, IP accumulation in lacinias agrees with previous reports indicating that increases of cell wall structural phenolics (lignin, suberin, cell wall-bound polyphenols) are related to a greater heavy metal accumulation in roots (Ederli et al. 2004; Kováčik and Klejdus 2008). Although data of this study have helped define the potential redox implications of thiols in Cr(VI) reduction, more thorough studies are needed to better understand the reduction process from beginning to end.

Conclusions

The present study demonstrated by XANES analysis that Cr(III) is the only oxidative state of Cr occurring in S. minima plants grown in presence of Cr(VI). In this context, our data showed significant accumulations of thiols as well as soluble and insoluble phenolics in Cr-exposed fronds and lacinias. The accumulation patterns observed were different indicating probably different roles in Cr(VI) detoxification. We suggest that operating mechanisms would be related to the amount of Cr absorbed and the organ considered. Thus, in lacinias, where higher Cr accumulation occurs, the most predominant mechanisms could be the metal complexation with IF and TBP, with a lower participation of soluble compounds (e.g., SF, NPT). While in fronds, with low Cr accumulation, the reduction mechanism would be more related to SF and NPT, which could also include an unknown enzyme able to reduce Cr(VI).

The expression of an h-type thioredoxin in Cr-exposed lacinias also suggest a possible role in the process of Cr(VI) reduction. Despite recent progress, our knowledge of the biochemical interactions between thiols and other molecules in Cr(VI) reduction is still incomplete.