Superparamagnetic iron oxide nanoparticles (SPION) are widely used for biomedical application in cancer therapy due to their unique physicochemical nanoscale properties, such as enhanced reactivity, heat, and magnetic-field generation (see review Singh et al. 2010). Although the accumulation of SPION in the environment come from natural sources, such as fire and volcanic eruptions, the extensive use of engineered SPION may contaminate the aquatic environment and interact with biotic components of the ecosystem. Indeed, engineered metallic nanoparticles (NPs) are known to be potentially cytotoxic for many aquatic organisms because they alter many cellular processes and generate reactive oxygen species (ROS) (Klaine et al. 2008; Bhatt & Tripathi 2011; Peralta-Videa et al. 2011). Therefore, the accumulation of hazardous NPs, such as SPION, in aquatic ecosystems may represent a biological risk of toxicity, which must be evaluated. However, many factors must be considered in a toxicological risk assessment such as bioavailability of NPs in aqueous solution, uptake by aquatic organisms, and potential toxicity at a cellular level. Previous toxicological studies have identified that metallic NPs possess many physicochemical characteristics—such as their size, shape, chemical composition, aggregation degree, and their solubility, which causes the release of toxic metal ions—that are directly related to their toxicity potential (Griffitt et al. 2008; Jiang et al. 2009).

In aquatic ecosystems, photosynthetic organisms represent the main source of biomass production essential for the viability of animals of higher ecological trophic levels. Furthermore, the bioaccumulation of metallic contaminants in aquatic plants and algae can reach animal organisms, thus altering both biodiversity and the food chain of the ecosystem (Flinders 2006). However, the bioaccumulation effect of SPION causing toxic action on the plant cellular system has not been extensively examined, and the few studies on the toxicity of Fe3O4 NPs that were performed focused on terrestrial plant species grown hydroponically. For instance, Mushtaq (2011) reported inhibitory effects of Fe3O4 NPs after 6 days on cucumber (Cucumis sativus) seed germination and root elongation. They showed that the germination index, considering both relative root elongation and seed germination, decreased when cucumber seeds were exposed to 500, 2,500, and 5,000 µg mL−1 of Fe3O4 NPs. In another study, the effect of magnetite NP uptake and accumulation in pumpkin (Cucurbita maxima) plants grown hydroponically was investigated (Zhu et al. 2008). In this work, the investigators showed that at different aggregations of Fe3O4, NPs were taken up and translocated throughout various plant tissues (root, stem, and leaves), thus indicating NP transport pathways and bioaccumulation into the plant system. Recently, it was also shown that Fe3O4 NPs cause a decrease in net photosynthetic rate (based on absorbed carbon dioxide for fixation) and Chl a content when alga Chlorella vulgaris was exposed for 72 h to 100 and 200 103 µg L−1 of Fe3O4 NPs (Chen et al. 2012). Therefore, more studies must be performed to better understand the bioaccumulation effect of SPION on major physiological processes, such as photosynthesis and growth, of aquatic plants and algae.

Indeed, the entire physiological status of plant cells is highly dependent on photosynthetic activity, which it responsible for the supply of adenosine triphosphate and nicotinamide adenine dinucleotide phosphate involved in cellular metabolism. Therefore, any inhibition of photosynthetic processes may result in the alteration of the whole physiology of the plant cell (Seaton & Walker 1990; Gerst et al. 1994). For practical application, photosynthetic electron transport and carbon assimilation as the main indication of energy storage can be evaluated by PSII-dependent chlorophyll (Chl) fluorescence emission (Baker 2008). Particularly, the change in rapid Chl a fluorescence induction has been shown to be dependent on water-splitting system functions at the PSII RC and to electron transport activity related to oxido-reduction states of respective electron acceptors quinone A (QA), quinone B (QB), and plastoquinone pool (Strasser and Strasser 1995; Dewez et al. 2007). Ecotoxicological studies have shown that many photosynthetic-based fluorescence parameters can be sensitive biomarkers to both metallic and organic pollutants because these have different mechanisms of toxicity (Popovic et al. 2003; Ralph et al. 2007).

The freshwater plant Lemna gibba is a convenient model organism used in toxicological investigations due to its rapid reproduction rate (1–4 days), small size, and floating leaves, which permit rapid uptake of soluble compounds (Lewis 1995). Duckweeds (Lemna sp.) are frequently used for testing the effects of different substances on growth inhibition over a period of 7 days (OECD 2006). In this study, L. gibba plants were exposed to NP suspensions of Fe3O4 (SPION-1), Co0.2Zn0.8Fe2O4 (SPION-2), or Co0.5Zn0.5Fe2O4 (SPION-3) for 7 days to characterize their toxicity on the entire plant system by evaluating both photosynthetic processes and growth rate. These forms of SPION were selected to determine if their physicochemical properties can induce different degrees of toxicity in this plant model. Therefore, we were able to detect and characterize the toxic effects of these SPION by monitoring changes in different cellular and biochemical biomarkers.

Materials and Methods

Biological Material

The aquatic macrophyte L. gibba (CPCC no. 310) was obtained from the Canadian Phycological Culture Centre (Waterloo, Ontario, Canada). Plants were grown in an inorganic culture medium (pH 6.5) prepared as described in Frankart et al. (2002): KNO3 202 µg mL−1, KH2PO4 50.3 µg mL−1, K2HPO4 27.8 µg mL−1, K2SO4 17.4 µg mL−1, MgSO4·7H2O 49.6 µg mL−1, CaCl2 11.1 µg mL−1, FeSO4·7H2O 6 µg mL−1, H3BO3 5.72 µg mL−1, MnCl2·4H2O 2.82 µg mL−1, ZnSO4 0.6 µg mL−1, (NH4)Mo7O24 0.043 µg mL−1, CuCl2·2H2O 0.078 µg mL−1, and CoCl2·6H2O 0.054 µg mL−1. Cultures were maintained in a growing chamber (Conviron Controlled Environments, Winnipeg, Manitoba, Canada) under an illumination of 100 μE m−2 s−1 provided by cool white fluorescent lamps (Sylvania GRO-LUX F40/GS/WS, Drummondville, Canada) with a light-to-dark cycle of 16:8 h and a constant temperature of 24± 1 °C.

Synthesis of NPs

SPION as Fe3O4 (SPION-1), Co0.2Zn0.8Fe2O4 (SPION-2), and Co0.5Zn0.5Fe2O4 (SPION-3) were produced by the polyol method (Basti et al. 2010; Ben Tahar et al. 2012). The precursor salts Co(CH3CO2)2·4H2O (Acros, 97 %), Zn(CH3CO2)2·2H2O (Fluka, > 99 %), and Fe(CH3CO2)2 (Acros, 95 %) were accurately weighed in the appropriate stoichiometric ratios and then dissolved in 125 mL of diethyleneglycol to reach a nominal Fe3+ concentration of 0.2 M. The hydrolysis ratio, h, as defined by the water to metal molar ratio, was adjusted to ≈ 1 for SPION-1 and to ≈ 10 for the two others SPION by adding a small amount of distilled water. The resulting mixture was then heated to reflux under mechanical stirring for 3 h. After slow cooling, the resulting precipitate was washed with ethanol under sonication with intermittent centrifugation and dried in air at 60 °C.

Physicochemical Characterization

Powders were characterized by X-ray diffraction (XRD) on a PANalytical X’pert Pro in the 2Ө (o) range 10–90 with a scan step of 0.05 using the CoK α radiation (λ = 0.1789/0.1792 nm). The unit cell parameter and peak indexing were performed with the program McMaille (Le Bail 2004), and the refinement was performed by the least squares method. The average crystal size was inferred from the broadening of the (311) diffraction line (2θ (o) ≈ 41) by applying the well-known Scherrer equation and considering a pseudo–Voigt line profile shape with a polycrystalline silicon as standard.

The size and shape of the prepared particles were analyzed on a JEOL-100-CX II transmission electron microscope (TEM) operating at 100 kV equipped with an energy-dispersive spectrometer with a probe diameter of 200 nm, which also provides the different metallic atomic ratios. The powders were ultrasonically dispersed in a solution of oleic acid in heptane (5 % weight). One drop of the suspension was deposited on a carbon-coated copper grid, and the solvent was evaporated at room temperature. TEM image treatment (mean particle size and shape) was made from at least 250 particles collected from 10 micrographs using the software package SAISAM (Microvision Instruments S.A.S., Evry cedex, France).

In the preparation of stock solutions, SPION were suspended in culture medium and homogenized by vortexing for 1 min, and then the solution was sonicated in ultrasonication bath for 30 min at 4 °C to break up aggregates. After sonication, particles in suspension were vortexed (1 min), and all of the concentrations of SPION used for treatment were prepared in culture medium of L. gibba. Furthermore, the zeta potential of SPION in culture medium was determined by the electrophoretic mobility method using a ZetaPlus particle sizer (Brookhaven Instrument, USA) with 90Plus Particles Sizing software (Ver. 4.20).

Experimental Treatments

Three triple-fronded plants were used for experiments and were placed in Petri dishes (with open lid) containing 6 mL of growth medium and initial concentrations of 0, 12.5, 25, 50, 100, 200, or 400 µg mL−1 of SPION. Plants were monitored for 7 days. For each experiment, all treated and control media were renewed every 24 h to maintain constantly the exposure of L. gibba plants to initial concentration of SPION and their soluble fraction.

Growth-Inhibition Test

Plant growth was determined by the change in fresh weight or frond number every day for 7 days. The relative growth rate (RGR) was evaluated according to Hoffmann and Poorter (2002) as RGR = (ln Wd7− ln Wd0)/7 days where Wd7 represents the fresh weight or frond number of plants at day 7, and Wd0 represents the initial fresh weight or frond number. The inhibitory response for specific growth rate in percentage (Ir) was determined by frond number or fresh weight according to Organisation for economic Co-operation and Development (OECD) guidelines for the testing of chemicals (OECD 2006) as follows: Ir = ((μC– μT)/μC) × 100 where Ir represents the inhibitory response (in  %) for the average specific growth rate; μC represents the mean value of frond number or fresh weight in the control group; and μT represents the mean value in the treatment group.

Chl Content

At 7 days of treatment, entire plants were homogenized in 100 % methanol in darkness. Then total Chl was extracted at 65 °C for 10 min and centrifuged for 5 min at 10,000 g. The content was determined spectrophotometrically (spectrophotometer Perkin-Elmer Lambda 40) according to Wellburn (1994) as follows: total Chl (μg mL−1) = (24.93 × A652.4 + 1.44 × A665.2).

Fe, Co, and Zn Content in Plant Biomass

After treatment, plants were dried at 105 °C for 24 h and weighed to determine dry weight. Then plants were placed in an acid-washed glass tube containing 4 mL of HNO3 and 500 μL of H2O2. The digestion process was performed during 48 h at 129 °C before dilution in MilliQ purified water for atomic absorption spectrometry (Varian SpectrAA 220 FS) quantification of iron (Fe), cobalt (Co), and zinc (Zn). Obtained Fe, Co, and Zn concentrations were normalized to dry weight.

Soluble Fraction of Fe, Co, and Zn

Solubility of free Fe, Co, and Zn released from SPION suspensions were determined in L. gibba medium. SPION suspensions were incubated during 24 h under the same conditions as described previously for L. gibba. After incubation, NPs were removed by centrifugation at 12,000g for 30 min. The supernatant was removed with care and diluted in 10 % HNO3. Quantification of free Fe, Co, and Zn in solution was determined by atomic absorption spectrometry (Varian SpectrAA 220 FS).

Chl Fluorescence Measurement

Photosynthetic electron transport was monitored from the change in the rapid increase of Chl a fluorescence emission using a Plant Efficiency Analyser fluorometer (Hansatech, King’s Lynn, Norfolk, UK). Before measurements, plants were placed in darkness for 30 min. The fluorescence induction was triggered using a 6-second saturating flash at 3,500 µE m−2 s−1. Fluorescence intensities at O (F O), K (F K), J (F J), and I (F I) transients were determined at 20, 300 µs, 2, and 30 ms, respectively. Fluorescence intensity at P level reaches maximal value (F M) under saturating illumination. Different photosynthetic-based fluorescence parameters related to the functional state of PSII were calculated according to Strasser et al. (2004) as follows: The maximum quantum yield of primary photochemistry of PSII—i.e., F V/F M = (F M−F O)/F M—was used to measure the maximum efficiency of PSII electron transport. The absorption of photons by light-harvesting antenna (ABS) per active reaction center (RC), estimating the relative size of PSII light-harvesting antenna complexes, was determined by the ratio ABS/RC = ((F K−F O)/0.250) × (1/(F J−F O)) × (F M/(F M−F O)). The relative variable fluorescence yield at J transient was calculated according to V J = (F J−F O) (F M−F O), thus estimating the fraction of QA in its reduced state.

ROS Formation Per Viable Cells

The formation of ROS in L. gibba fronds was determined according to the procedure explained in Babu et al. (2003). After being washed in culture medium, plants were exposed to 5 μM of 2′,7′dichlorodihydrofluorescein diacetate in 1 mL of solution medium for 30 min at 25 °C. The formation of ROS was determined using a fluorescence plate reader (SpectraMax M2e MultiMode Microplate Reader). The fluorescence emission was measured at 530 nm by using a light excitation wavelength at 485 nm. Fluorescence data were normalized by fresh weight.

Viability of L. gibba plant cells was estimated using the fluorescein diacetate (FDA) method according to Mayer et al. (1997). Four intact plants from each treatment and control were exposed to 5 μM of FDA in 1 mL of solution medium. A light excitation at 485 nm was used to measure the fluorescence emission at 530 nm. Fluorescence data were normalized by fresh weight. The ratio of ROS level per cellular viability was evaluated to normalize the formation of ROS in L. gibba fronds.

Statistical Analysis

All treatments were performed in two exposure test series with three replicates. Means and SDs were calculated for each treatment. Significant differences between control and treated plants were determined by one way analysis of variance and Dunnett’s multiple comparison test and were determined for p < 0.05. The half maximal effective concentration (EC50) was obtained by the inhibitory concentration versus the response fitting to the toxicity data.

Results

Physicochemical Characteristics of SPION

XRD is a rapid analytical technique used for phase identification of a crystalline material providing information on unit cell dimensions. The XRD patterns of all prepared powders are shown in Fig. 1a. All diffraction lines are consistent with those of typical spinel-like ferrites without evidence of impurities. The calculated cell parameters (Table 1) are close to those reported for ferrites with similar chemical composition (Vaidyanathan et al. 2007). In addition, the diffraction peaks are clearly broadened as the result of decreased particle size. The estimated crystallite average size (sizexrd) is approximately 10 nm for SPION-1 and 5 nm for SPION-2 and SPION-3. TEM images of the prepared powders are also given in Fig. 1b. They consist of approximately spherical NPs, almost nonagglomerated, with an average diameter (sizetem) of 9 nm for SPION-1 and 4 nm for others NPs. We noticed that the concordance of particle size inferred from XRD and TEM indicated that the produced NPs could be considered as single crystals. Measurement of zeta potential indicated that SPION are negatively charged in the media, and values were −26.37, −26.05, and −79.15 for SPION-1, SPION-2, and SPION-3, respectively (Table 1). These values were found to be stable in the culture medium during the experimental exposure.

Fig. 1
figure 1

XRD patterns (a) and TEM images (b) of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4)

Full size image
Table 1 Microstructural characteristics of prepared SPION NPs
Full size table

Solubility and Bioaccumulation of Fe, Co, and Zn

Because all treated and control media were renewed every 24 h, the soluble fraction of free Fe, Co, and Zn released from all SPION suspensions was determined in culture medium for that time period (Fig. 2). The quantity of free Fe, Co, and Zn in aqueous solution was dependent on metal species, SPION physicochemical characteristics, and concentration. According to our results, the quantity of free Fe was dependent on SPION-1 (Fe3O4) concentration and increased significantly (p < 0.05) by 2-fold for 400 µg mL−1 of SPION-1 suspensions compared with exposure concentration of 12.5 µg mL−1. However, free Fe per total Fe was equal to 0.6 % for 400 µg mL−1 of SPION-1 compared with 10.6 % for an exposure concentration of 12.5 µg mL−1. Such solubility was different for SPION-2 (Co0.2Zn0.8Fe2O4), for which the quantity of free Fe, Co, and Zn significantly (p < 0.05) increased respectively by 1.5-, 5.5-, and 22.5-fold at 400 µg mL−1 compared with 12.5 µg mL−1. However, the quantity of free metals in the soluble fraction compared with exposure concentration of SPION-2 (400 µg mL−1) decreased significantly (p < 0.05) by 13.4, 7.2, and 4 % for Fe, Co, and Zn, respectively. Furthermore, the concentration of Fe, Co, and Zn in the soluble fraction significantly (p < 0.05) increased, respectively, by 1.4-, 2.4-, and 1.5-fold for SPION-3 (Co0.5Zn0.5Fe2O4) treatment of 400 µg mL−1 compared with 12.5 µg mL−1. However, the quantity of free Fe, Co, and Zn, compared with the treatment (400 µg mL−1), decreased significantly (p < 0.05) by 26, 18, and 32 %, respectively. In addition, the quantity of free Zn in the soluble fraction was greater than that of Co and Fe for SPION-2 and SPION-3, respectively, and this was dependent on the composition and amount of total Zn made in the SPION (Fig. 2). Therefore, these results indicated that the concentration of Fe, Co, and Zn in the soluble fraction was dependent on the quantity of SPION suspensions in the media. Moreover, the contribution of free Fe, Co, and Zn provided by the culture medium was negligible when compared with their total amount in the soluble fraction (see Biological Material).

Fig. 2
figure 2

Soluble fraction of Fe, Co, and Zn released from different concentrations of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions in L. gibba medium after 24 h of exposure. Samples exposed to SPION were significantly different from the control at p < 0.05 (*)

Full size image

The bioaccumulation of total Fe, Co, and Zn was also quantified in plant biomass for L. gibba exposed 7 days to SPION (Fig. 3). The accumulated Fe, Co, and Zn concentrations in plants were highly dependent on the exposure concentration of SPION. The bioaccumulation of Fe increased significantly (p < 0.05) by almost 10-fold when L. gibba was exposed for 7 days to 400 µg mL−1 of SPION-1 compared with 12.5 µg mL−1. However, the bioaccumulation of Zn was greater when L. gibba was exposed to SPION-2 compared with SPION-3 under similar exposure concentrations (Fig. 3).

Fig. 3
figure 3

Content of Fe, Co, and Zn in L. gibba plant exposed for 7 days to different concentrations of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions. For SPION-1 and -3, samples exposed to 25 µg mL−1 were significantly different from the control at p < 0.05 (*). For SPION-2, samples exposed to 50 µg mL−1 were significantly different from the control at p < 0.05 (*)

Full size image

Inhibition of Plant Growth

When L. gibba plant was exposed for 7 days to SPION at concentrations varying from 12.5 to 400 µg mL−1, a decrease in relative growth rate based on fronds number and fresh weight was noticed (Table 2). The relative growth rate based on fresh weight was more affected by SPION toxicity compared with frond number. Under a concentration exposure of 400 µg mL−1, the relative growth rate based on fresh weight decreased significantly (p < 0.05) compared with the control by 37, 43.5, and 36.5 % for SPION-1, SPION-2, and SPION-3, respectively (Table 2). In addition, the inhibitory responses for specific growth rate based on fresh weight were 56, 59, and 52 % for SPION-1, SPION-2, and SPION-3, respectively (Fig. 4). However, the inhibitory responses for specific growth rate based on frond number were, respectively, 39, 51, and 49 % for SPION-1, SPION-2, and SPION-3. Moreover, when L. gibba plants were exposed for 7 days to different concentrations of SPION suspensions, the EC50 value of plant growth based on fresh weight was 244, 183, and 237 µg mL−1, respectively, for SPION-1, SPION-2, and SPION-3.

Table 2 Change in RGR based on fresh weight or frond number of L. gibba plants exposed for 7 days to different SPION suspensions
Full size table
Fig. 4
figure 4

Inhibitory response of L. gibba plant growth (% Ir of total frond number or biomass as fresh weight) after exposure for 7 days to different concentrations (12.5–400 µg mL−1) of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions. Samples exposed to SPION were significantly different from the control for p < 0.05 (*)

Full size image

Inhibition of Photosynthesis

When L. gibba plant was exposed 7 days to SPION concentrations varying from 12.5 to 400 µg mL−1, the inhibitory effect of SPION was showed by a significant (p < 0.05) decrease of Chl content compared with the control by 51, 60, and 58 % for SPION-1, SPION-2, and SPION-3, respectively (Fig. 5). The Chl a/b ratio increased significantly (p < 0.05) compared with the control by 115, 156, and 132 %, respectively, for SPION-1, SPION-2, and SPION-3, thus indicating a stronger inhibition of Chl b synthesis. The change of photosynthetic electron transport was also investigated to determine SPION toxic effects on processes related to PSII activity. The effect of SPION on primary photochemical reactions of PSII was monitored using Chl a fluorescence emission. We found a significant decrease of fluorescence yields at J, I, and P levels, which was related to SPION concentration without any changing in their time of appearance on the kinetics (data not shown). Furthermore, we were able to characterize in more detail the SPION effect on PSII photochemistry by using the change of photosynthetic-fluorescence parameters estimated from the fluorescence kinetic (Fig. 6). A significant effect on photosynthetic electron transport was observed at < 400 µg mL−1 of SPION treatment. However, there was not a significant decrease in the value of the maximal quantum yield of PSII (FV/FM) for SPION-1, SPION-2, and SPION-3, thus indicating that a large proportion of PSII was still able to perform primary photochemical reactions. The change of VJ values (Fig. 6) was indicative of a significant (p < 0.05) increase of Q −A /QA(total) due to inhibition of PSII electron transport flow toward the plastoquinone pool. The value of ABS/RC also increased significantly (p < 0.05) compared with the control by 42, 70, and 39 %, respectively, for SPION-1, SPION-2, and SPION-3, thus indicating a decrease of functional PSII RCs. Moreover, the change of the PSII performance index was used as a global parameter integrating all PSII photochemical reactions. When L. gibba plants were exposed during 7 days to 400 µg mL−1 of SPION, the PI value showed a significant decrease (p < 0.05) compared with the control by 83, 86, and 79 % for SPION-1, SPION-2, and SPION-3, respectively (Fig. 6). Under the same experimental conditions, the EC50 on the change of parameter PI was 57, 17, and 200 µg mL−1 for SPION-1, SPION-2, and SPION-3, respectively.

Fig. 5
figure 5

Change in total Chl and ratio of Chl a/b when L. gibba plant was exposed for 7 days to different concentrations of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions. For total Chl, samples exposed to 25 µg mL−1 SPION were significantly different from the control (p < 0.05 (*). For the ratio of Chl a/b, samples exposed to SPION-2 and -3 at 50 µg mL−1 and to SPION-1 at 100 µg mL−1 were significantly different from the control at p < 0.05 (*)

Full size image
Fig. 6
figure 6

Change of different photosynthetic-based fluorescence parameters when L. gibba plant was exposed for 7 days to different concentrations of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions. FV/FM = maximal PSII quantum yield; ABS/RC = ratio between the number of active PSII RCs and light-harvesting Chl antenna size; relative variable fluorescence yield at J transient VJ; PI = performance index of PSII photochemical activity. For ABS/RC and VJ, samples exposed to 400 µg mL−1 of SPION were significantly different from the control for p < 0.05 (*). For PI, samples exposed to SPION-1 and -3 from 25 µg mL−1 and to SPION-2 from 12.5 µg mL−1 were significantly different from the control at p < 0.05 (*)

Full size image

Production of ROS per Viable Cells

The formation of ROS per viable cells was determined for L. gibba plants exposed 7 days to SPION (Fig. 7). When L. gibba plants were exposed 7 days to 200 and 400 µg mL−1 of SPION, the production of ROS increased significantly (p < 0.05) compared with control, thus indicating induce cellular oxidative stress under these conditions. Moreover, in L. gibba plants exposed for 7 days to 400 µg mL−1 of SPION, the production of ROS per viable cells increased by 2-fold for SPION-1 and SPION-2 compared with SPION-3.

Fig. 7
figure 7

Change in the production of ROS per viable cells when L. gibba plant was exposed for 7 days to different concentrations of SPION-1 (Fe3O4), SPION-2 (Co0.2Zn0.8Fe2O4), and SPION-3 (Co0.5Zn0.5Fe2O4) suspensions. Samples exposed to 100 µg mL−1 of SPION increased significantly in ROS compared with the control at p < 0.05 (*)

Full size image

Discussion

Alteration of Plant Physiological Status

This study showed original results concerning the toxic effects of SPION on L. gibba fronds, suggesting a complex mode of action that was dependent on SPION chemical composition, concentration, and dissolved metallic ions. On one hand, the bioaccumulation of free Fe, Co, and Zn from the soluble fraction contributes to the toxicity in the plant system. Solubilisation of metallic NPs into free metal ions is usually considered one of the most common mechanisms causing toxic effects in aquatic organisms (Franklin et al. 2007; Aruoja et al. 2009). However, the soluble fraction per total SPION was not proportional when comparing low (12.5 µg mL−1) with high (400 µg mL−1) initial concentrations of SPION. Indeed, the ratio of the concentration of free Fe, Co, and Zn contents per total concentration of SPION decreased with increasing concentration of SPION, thus indicating a change in their physicochemical properties related to their solubility. In contrast, the bioaccumulation of Fe, Co, and Zn, as well as their effects in L. gibba plant, increased with the increasing exposure concentration of SPION, suggesting that SPION suspensions also contribute to the toxicity impact. Recently, it was shown for agglomerates of Fe3O4 that NPs accumulate into the entire plant system and thus represent a source of toxicity (Zhu et al. 2008; Mushtaq 2011). Moreover, the ROS level per cell viability was greatest at an exposure SPION concentration of 400 µg mL−1 when the bioaccumulation of Fe, Co, Zn reached the greatest levels. Previous studies showed that many metallic NPs induce production of ROS by way of Fenton reaction or the disruption of physiological processes (Klaine et al. 2008; Hu et al. 2009). Therefore, the toxicity caused by bioaccumulated Fe, Co, and Zn in plant biomass may be due to both free ionic metals released from SPION in the soluble fraction as well as bioaccumulated SPION. However, no current analytical approach permits discrimination of the toxicity contribution between SPION and its free ionic metals.

Significance of Photosynthetic-Based Fluorescence Parameters

It was reported earlier that metals such as cadmium, copper, and chromium ions directly inhibit PSII electron transport at the water-splitting system of the RC (Küpper et al. 2002; Popovic et al. 2003; Dewez et al. 2005; Ait Ali et al. 2006). Here we showed evidence that changes of selected photosynthetic-based fluorescence parameters were indicative of the alteration of PSII functional integrity caused by SPION effects, especially under high exposure concentration (400 µg mL−1). By using the yields of fluorescence transients, we were able to evaluate the performance index of PSII activity, which represents an integrative indicator of all PSII photochemical reactions (Strasser et al. 2004). Compared with others fluorescence parameters, change in PI value was the most sensitive biomarker of the deteriorated PSII functions caused by all SPION toxicity. In a previous study, change of PI value was indicative of the decreased number of active PSII RCs as well as damage to the oxygen-evolving complex when Spirodela polyrhiza was exposed to chromate toxic effects (Appenroth et al. 2001). In our study, a high concentration of SPION (400 µg mL−1) showed evidence of strong inhibitory effects on energy transfer, from light-harvesting complex to photosynthetic RCs, on the deterioration of the PSII water-splitting system and the inactivation of PSII RCs.

Comparison Between SPION-1, SPION-2, and SPION-3

Currently, there are almost no data available concerning the toxicity of SPION on aquatic organisms. Only one study concerned the toxic effects of Fe3O4 NPs (35 nm) on the green alga Chlorella vulgaris exposed during 72 h and showed induction of oxidative stress and alteration of photosynthesis under treatment from 200 to 1,600 µg mL−1 (Chen et al. 2012). In our study, according to the change of used cellular and biochemical biomarkers, the exposure of SPION-1, SPION-2, and SPION-3 suspensions to L. gibba caused several alterations to the entire plant cellular system. However, based on our results, we were not able to compare the complex toxicity mechanisms between SPION-1, SPION-2, and SPION-3 suspensions, which may come from both the uptake of NPs and the presence of metal ions in the soluble fraction. Indeed, no current protocols and analytical methods permit the precise quantification of the proportion of bioaccumulated NPs from metal ions in plant cells without altering their physicochemical states or the determination of their contributions to the impact of cellular toxicity. Therefore, this study permitted only the detection and characterization of the toxic effects of these SPION by monitoring changes in different cellular and biochemical biomarkers.

Ecotoxicological Significance

The toxic potential of SPION suspension as hazardous contaminants in the aquatic environment must be investigated to develop specific bioassays for toxicity assessment. Our results showed that the toxicity risk of SPION suspension in aqueous solution can be estimated by the physiological alteration of the aquatic plant L. gibba. In this regard, the L. gibba plant was shown to be a sensitive bioindicator of SPION toxicity in laboratory bioassay testing. However, in a natural freshwater environment, more factors may alter the physicochemical properties of NPs in aqueous solution, which must be investigated. Indeed, the agglomeration and sedimentation of SPION can be affected by natural organic colloids, which may modify their bioavailability by changing their solubility. It has been shown that the presence of humic and fulvic acids in the natural environment may cause sedimentation of several NPs (Domingos et al. 2009; Scown et al. 2010). Therefore, further studies should be aimed at determining the behaviour of SPION in the aquatic environment and how natural conditions can alter their potential source of toxicity.

Conclusion

In this study, we characterized the toxicity of three SPION suspensions having different chemical compositions on the aquatic plant model L. gibba, which was shown to be a valuable bioindicator of toxicity. The toxicity impact was indicated by physiological alterations such as decreased Chl content, deterioration of photosynthetic activity, induction of oxidative stress, and inhibition of growth. Our results, based on the change of several biomarkers, showed that these SPION have a complex toxic mode of action on the plant system, thus affecting its viability. Therefore, the plant model L. gibba was shown to be a sensitive bioindicator of SPION cellular toxicity and thus can be used in the development of laboratory bioassay toxicity testing.