The Enzymatic and Non-Enzymatic Antioxidant System Response of the Seagrass Cymodocea nodosa to Bisphenol-A Toxicity

The effects of environmentally relevant bisphenol A (BPA) concentrations (0.3, 1 and 3 μg L−1) were tested at 2, 4, 6 and 8 days, on intermediate leaves, of the seagrass Cymodocea nodosa. Hydrogen peroxide (H2O2) production, lipid peroxidation, protein, phenolic content and antioxidant enzyme activities were investigated. Increased H2O2 formation was detected even at the lowest BPA treatments from the beginning of the experiment and both the enzymatic and non-enzymatic antioxidant defense mechanisms were activated upon application of BPA. Elevated H2O2 levels that were detected as a response to increasing BPA concentrations and incubation time, led to the decrease of protein content on the 4th day even at the two lower BPA concentrations, and to the increase of the lipid peroxidation at the highest concentration. However, on the 6th day of BPA exposure, protein content did not differ from the control, indicating the ability of both the enzymatic and non-enzymatic mechanisms (such as superoxide dismutase (SOD) and phenolics) to counteract the BPA-derived oxidative stress. The early response of the protein content determined that the Low Effect Concentration (LOEC) of BPA is 0.3 μg L−1 and that the protein content meets the requirements to be considered as a possible early warning “biomarker” for C. nodosa against BPA toxicity.


Introduction
The ever-increasing leaches from plastic pollution, have unavoidably led to some endocrine disrupting chemicals (EDCs) to be ubiquitous in nearly all the aquatic environments [1,2]. EDC compounds have detrimental effects on several organisms, since they disrupt the hormonal equilibrium. The most common EDC is bisphenol A (2,2-bis [4 -hydroxyphenyl]propane; BPA). This plasticizer is an organic chemical used as an intermediate in the production of numerous plastic utensils [3]. Due to the increasing demand for the aformentioned products, BPA production has constantly grown in the last years. For instance, the global demand was 5.5 million tons in 2011 [4] and it was expected to have grown at a rate of 4.6% per year the period from 2013 to 2019 [4]. BPA discharges which enter into coastal systems via rivers or rain [5], have steadily increased the BPA concentrations found in river waters (0.014-21.0 µg L −1 ), in seawater (<0.0005-2.470 µg L −1 ), in wastewater plant effluents (0.087-5.625 µg L −1 ) and in industrial wastewater treatment wastes (1-150 µg L −1 ) [6,7]. Thus, BPA leaching into the aquatic environment represents a potential risk [2]. Therefore, extensive toxicity studies concerning its impacts on aquatic organisms have been conducted mainly focusing on animals (i.e., fish, amphibians, crustaceans and tion time with the respective values in the control is also given. N = 54 values per concentration an time.   Superoxide dismutase activity (SOD) at the two lower BPA solutions (0.3 and L −1 ), on the 4th day, was significantly (p < 0.05) increased compared to the co (1.6-2.0-fold) and the higher solution (3 μg L −1 ) and continued to be slightly elevated 0.05) on the 6th day. At the highest treatment (3 μg L −1 ), SOD activity was equal to 0.05) or lower (p < 0.05) than the control and the two lower treatments in the experi (Figure 3a).
Protein content at 1 μg L −1 BPA was lower (p < 0.05) than the 3 μg L −1 on the 2nd and at the 0.3 and 1 μg L −1 BPA treatments, it was lower (p < 0.05) than the contro 0.05) and the 3 μg L −1 BPA treatment on the 4th day ( Figure 3c).
Lipid peroxidation, as expressed by MDA (Malondialdehyde) content, at the 3 μ was higher (p < 0.05) than the control and the two lower BPA treatments, at the start o experiment, whereas it was lower (p < 0.05) than the control and the lowest BPA sol on the 4th and 6th day (Figure 3d).

BPA Effects on Antioxidant Enzyme Activity, Total Protein Content and Lipid Peroxidation
Superoxide dismutase activity (SOD) at the two lower BPA solutions (0.3 and 1 µg L −1 ), on the 4th day, was significantly (p < 0.05) increased compared to the control (1.6-2.0-fold) and the higher solution (3 µg L −1 ) and continued to be slightly elevated (p > 0.05) on the 6th day. At the highest treatment (3 µg L −1 ), SOD activity was equal to (p > 0.05) or lower (p < 0.05) than the control and the two lower treatments in the experiment (Figure 3a).
Protein content at 1 µg L −1 BPA was lower (p < 0.05) than the 3 µg L −1 on the 2nd day, and at the 0.3 and 1 µg L −1 BPA treatments, it was lower (p < 0.05) than the control (p < 0.05) and the 3 µg L −1 BPA treatment on the 4th day ( Figure 3c).
Lipid peroxidation, as expressed by MDA (Malondialdehyde) content, at the 3 µg L −1 was higher (p < 0.05) than the control and the two lower BPA treatments, at the start of the experiment, whereas it was lower (p < 0.05) than the control and the lowest BPA solution on the 4th and 6th day (Figure 3d).

Total Phenolic Content
Leaf epidermal cells under BPA treatment are filled with dark (probably phenolic material) (Figure 4a-d). Mann-Whitney U-test showed that the total phenolic content at the lowest treatment (0.3 µg L −1 ) was elevated (p < 0.05) compared to the control and the highest solution from the beginning (1st and 2nd day) and at the end of the experiment (8th day) (Figure 4e). At 1 µg L −1 the phenolic compound production also started from the beginning of the test (1st and 2nd day), but this increase was significant (p < 0.05) on the 4th day (  material) (Figure 4a-d). Mann-Whitney U-test showed that the total phenolic content at the lowest treatment (0.3 μg L −1 ) was elevated (p < 0.05) compared to the control and the highest solution from the beginning (1st and 2nd day) and at the end of the experiment (8th day) (Figure 4e). At 1 μg L −1 the phenolic compound production also started from the beginning of the test (1st and 2nd day), but this increase was significant (p < 0.05) on the 4th day ( Figure 4e). Finally, at the 3 μg L −1 BPA, their values increased relative to the control, mainly on the 2nd day (p < 0.05), and on the 4th day (p > 0.05) as well ( Figure 4e). Changes in total phenolic content in Cymodocea nodosa leaves, at 0.3, 1 and 3 µg L −1 BPA treatments and in the control, exposed for 1, 2, 4, 6 and 8 days; mean ± SE from three subsamples (e). Different letters express significantly different values between treatments and the control and between treatments (Mann-Whitney U-test; p < 0.05).

Discussion
Laboratory toxicity data concerning BPA effects on aquatic macrophytes are scarce and have been conducted on a limited number of species. For instance, BPA effects were addressed on the freshwater angiosperms Lemna gibba and Ceratopyllum demersum and the seagrass Cymodocea nodosa. In L. gibba, at higher than the environmentally relevant concentrations, BPA affected its growth and survival, photosynthetic activity (chlorophyll fluorescence, photosynthetic pigments), and sugar and starch contents [16,38] and caused a severe cellular damage as indicated by the increase MDA levels [15]. Similarly, in C. demersum, BPA caused protein and photosynthetic pigment reductions [39]. BPA concentrations applied in our experiments (0.3, 1 and 3 µg L −1 ) were within the environmentally relevant concentrations (0.08-12.5 µg L −1 ; [7,8]). Specifically, 0.3 and 1 µg L −1 were among those found in seawater and 3 µg L −1 was in the range of those in surface waters and emitted from wastewater treatment plants [6][7][8]. Therefore, our study further consolidates the notion that BPA, even at low concentrations, is detrimental to aquatic organisms [3], including C. nodosa [20,23,25], as discussed below.
Limited ecotoxicological studies exist addressing the effects of BPA on seagrasses, and these are limited on C. nodosa [20,23,25]. However, in these studies the antioxidant mechanisms, the alterations in the cell membrane permeability and the protein content, have not been investigated. In C. nodosa, BPA (i) disrupted the cell cytoskeleton (actin filament, AF and microtubules, MT) at 0.03 and 0.1 µg L −1 BPA, respectively [20], (ii) retarded growth of juvenile leaf blades, adult leaves and plagiotropic rhizome internodes (10-day EC 50 values: 0.137, 0.029 and 0.025 µg L −1 , respectively [25]), and (iii) affected PSII function probably via the H 2 O 2 increase (positive H 2 DCF-DA staining) firstly at 0.5 µg L −1 BPA exposure [23]. Under BPA exposure the disruption of the electron transfer between the PSII and PSI reaction centers on aquatic microalgae and terrestrial plants [22,[40][41][42] resulted in ROS (superoxide, O 2− , hydrogen peroxide, H 2 O 2 , etc.) generation [22,26,43]. In a previous study, in C. nodosa, under BPA exposure, a decrease of the photosynthetic parameters ΦPSII and ΦNPQ, and an increase of ΦNO were observed, which supported the role of H 2 O 2 in the C. nodosa response against BPA toxicity [23]. In this study, in C. nodosa intermediate leaf blades, BPA induced H 2 O 2 overproduction even at the lowest concentrations used (0.3 µg L −1 ), and the shorter duration of exposure (Figures 1 and 2; Table 1) (see also [23]). At 3 µg L −1 , H 2 O 2 generation was elevated compared to the control and the other treatments, mainly the first 2-4 days of BPA treatment (Figures 1 and 2; Table 1). It has been stated that, up-regulated H 2 O 2 production could cause both oxidative damages, but can also trigger retrograde signaling, conferring BPA-tolerance [23], possibly by activating the antioxidant machinery. Oxidative-stress related parameters (both enzymatic and non-enzymatic antioxidant parameters) in marine flora species have been found to change as a response against various stressors, [34,35,[44][45][46][47][48]. Likewise, in C. nodosa BPA is a stressor that induces ROS production, especially H 2 O 2 (this study; [23]). H 2 O 2 is a parameter considered as an early marker of various stressors (i.e., [49]). H 2 O 2 is relatively stable and a compound with high longevity [50]. Therefore, when H 2 O 2 overproduction exceeds the antioxidant defense capacity can create oxidative stress in mitochondria, chloroplasts and peroxisomes [51], and can activate protective mechanisms, including the enzymatic antioxidant system (e.g., SOD, APX, CAT) [52]. In C. nodosa leaf blades the activity of SOD, an enzyme which catalyzes the dismutation of O 2− radical to H 2 O 2 and O 2 (e.g., [53]), was increased compared to the control at the two lowest treatments (0.3 and 1 µg L −1 ) on the 4th and up to 6th day, following an increase of H 2 O 2 production pattern (see CTCF ratio). In contrast, at the highest treatment concentration (3 µg L −1 ), SOD activity was lower or equal compared to the control and the other BPA solutions used (Figure 3a). The observed inhibition of SOD activity at high BPA concentrations, despite the H 2 O 2 generation, was also confirmed by other studies (e.g., [26], in Oryza sativa leaves), and could be attributed to the fact that high BPA doses are able to overwhelm the self-defense system, resulting in functional and structural cell damages, which seems to be also the case in C. nodosa leaves, since BPA damaged the structural integrity of chloroplasts, dictyosomes and endoplasmic reticulum [23]. APX activity, which has the highest affinity to convert H 2 O 2 to H 2 O [54], was activated by BPA, primary, at 1 and 3 µg L −1 on the beginning of the experiment and at 0.3 µg L −1 on the 6th day (Figure 3b), following the increase of H 2 O 2 production ( Figure 1). The possible dissimilarity among variations of ROS production and APX activity (Figures 1 and 3b) could be attributed to the fact that, other enzymes (i.e., catalase, CAT), which share the same substrate (H 2 O 2 ) could be possibly activated. For instance, in a previous study CAT and APX activities showed dissimilar trends, due to the unavailability of their common substrate [55].
Lipid peroxidation of polyunsaturated fatty acids is revealed by MDA formation [56]. The endogenous H 2 O 2 BPA-derived increase could trigger MDA creation, a phenomenon recorded in both aquatic and terrestrial flora under BPA toxicity [13,43,46,57]. The overproduction of H 2 O 2 in C. nodosa leaf blades at 3 µg L −1 on the 2nd day (maximum value of CTCF ratio) (Figure 1), indicated the inability of APX, which also increased on this day compared to the control, to successfully scavenge H 2 O 2 production, as the membrane lipid peroxidation increased (maximum MDA values) (Figure 3d). The decrease of MDA on the other incubation days, at this treatment could be attributed to the decrease of H 2 O 2 (Figures 1, 2 and 3d) or/and to a disruption of enzymatic defense mechanism due to structural and functional cell injuries [23].
Protein contents decreased in marine diatoms with the application of BPA [58], in the model plant Arabidopsis thaliana [59] and in the soybean plants [60], in a time and concentration related manner, with the BPA-derived oxidative stress being the reason of the observed protein content decrease. Likewise, the reduction in C. nodosa leaf protein content, mainly at the two lowest treatments on the 4th day and at 1 µg L −1 on the 2nd day (Figure 3c), may indicate an inhibition of protein biosynthesis or acceleration of protein decomposition [60], due to oxidation, even from the beginning of the experiments (Figures 1 and 2). This early response of the protein content (even from the 4th day), at the two lowest environmentally relevant BPA concentrations, determines that the LOEC value (Low Effect Concentration) is 0.3 µg L −1 , a LOEC value consistent with what is already reported, based on other markers [25].
Polyphenols are an important class of specialized metabolites that play crucial physiological roles throughout the plant life cycle, including responses to stress. It is well known that phenolic biosynthetic pathways are usually activated under harmful environmental conditions, such as drought, extreme temperatures, salinity, metal pollution, and ultraviolet radiations [61]. Secondary metabolites, such as phenolic compounds, were considered suitable "biomarkers" against abiotic and biotic stress in marine angiosperms, which contributed to their antioxidant activities [44,62,63]. BPA seems to be a factor inducing phenolic compound increase (Figure 3). The rise in the phenolic compounds at 0.3 µg L −1 BPA-treated C. nodosa leaves, compared to the control, even from the beginning of the experiment (1-2 days), which was then overstimulated at the end of the experiment (8th day), may indicate an activation of an antioxidant protective mechanism (see elevated CTCF ratio; Figure 2); this is also obvious for 3 µg L −1 BPA, on the 2nd experimental day.
Biochemical parameters (e.g., phenolics and proteins) are highly dynamic and variable among seagrasses, as they are modulated by external environmental or internal cellular factors [64]. However, biochemical "biomarkers" especially proteome-related ones recently have been considered as a promising tool in ecotoxicological research [65]. In our experiments on the 4th day, the stimulation of the enzymatic antioxidant mechanism (see high SOD activity in 0.3 and 1 µg L −1 BPA treatments) and the non-enzymatic mechanism (see high phenolic content at 1 µg L −1 on 4th day) may mitigate the overproduction of H 2 O 2 and lead to the recovery of SOD activity, and of protein and phenolic contents, on the 6th day, at the control levels (Figures 3c and 4, respectively). The early change in the protein levels was detected, and since most "biomarkers" are proteins (biotransformation enzymes, oxidative stress related proteins, cell death related proteins etc.) [66], it could help to conclude that protein levels could meet the requirements to be an early warning "biomarker" of BPA toxicity on C. nodosa.

Plant Collection and Experimental Conditions
The seagrass Cymodocea nodosa (Ucria) Ascherson 1870 collected, during June from the eastern coast of the Gulf of Thessaloniki, Northern Aegean Sea (40 • 24 24.03 N, 22 • 53 43.21 E), were kept for 24 h in seawater under laboratory conditions (16/8 h day/night regime, ambient temperature 21 ± 1 • C, irradiance of 120 µmol m −2 s −1 ) to acclimatize. Seawater used in the experiment was also collected from the sampling area and had a BPA concentration below detection limit (DL = 10 ng L −1 ) [25].
Seagrass consisting of horizontal rhizome, vertical rhizomes, roots and leaf shoots were incubated in non-BPA-based polypropylene (PP) copolymer (PPCO) constantly aerated aquaria, containing 20 L of 0.3, 1 and 3 µg L −1 bisphenol-A (BPA; Sigma, Taufkirchen, Germany) solutions in filtered seawater (0.45 µm Whatman GF/C) and filter seawater (used as a control). It should be noted that attention was paid to the selection of leaf shoots without epiphytes, while only the first intermediate leaves were used for our experiments, as the adult leaves are usually covered with epiphytes. BPA solutions were changed every two days, to adjust to the original levels out of an abundance of caution, although BPA in seawater remains stable for 35 days [67]. The experiments were run in triplicates. From each BPA solution and the control after 1, 2, 4, 6 and 8 days, blades of the first intermediate leaf [31] with nearly the same size (mean leaf length ± SE: 86 ± 12.3 mm) from at least six shoots per dose and day (two per aquarium) were randomly removed for H 2 O 2 production measurements. Additionally, each day from each treatment, intermediate leaves from another nine shoots (three per aquarium) were selected for the estimation of antioxidant enzyme activities and protein content, and another nine shoots for the estimation of lipid peroxidation, and other nine shoots for the estimation of the total phenolic compound contents.

Hydrogen Peroxide Production
BPA treated and untreated intermediate leaf blades at each exposure time were incubated with 2 , 7 -dichlorofluoresce in diacetate (H 2 DCF-DA, Sigma) in dimethyl sulfoxide (DMSO) [34]. The observation was made under a Zeis AxioImager Z.2 fluorescence microscope equipped with an MRc5 Axiocam. The corrected total cell fluorescence (CTCF) was measured, with the Image J software using the equation: CTCF = integrated density-(area of selected cell * mean fluorescence of background readings); CTCF values were attained for the tip, middle and basal leaf part, totally 54 CTCF values (3 areas per leaf part × 3 parts × 6 leaves) per concentration and day and the mean CTCF ratio, as the mean ratio of CTCF values in treated to that in the un-treated leaves was also calculated [49].

Estimation of Antioxidant Enzyme Activities, Total Protein and MDA Content
At each BPA concentration at the 2nd, 4th and 6th day, three subsamples were implemented. Each subsample (3 leaves per subsample, 100 mg wet weight) were grounded in liquid nitrogen and afterwards homogenized in 3 mL of 50 mM sodium phosphate buffer (pH 7.8), containing 0.1 mM EDTA and 2% (w/v) polyvinyl pyrrolidone (PVPP). The homogenates were centrifuged at 16.500× g for 30 min at 4 • C. Total protein content (mg g −1 wet weight) was determined in the supernatants, based on the principle of protein-dye binding (Coomassie Brilliant Blue G-250), using bovine serum albumin (BSA) as standard [68].
For the quantification of SOD activity, a reaction mixture (5 mL) of 50 mM potassium phosphate, 0.1 mM EDTA, TritonX-100 0.025% (pH 7.8), 13 mM methionine and 0.075 mM nitro blue tetrazolium (NBT) were used; in aliquots (50 µL) of the extracts 0.002 mM riboflavin was added. SOD activity (Units mg −1 protein) was determined under UV light at 560 nm, based on the inhibition rate of nitro blue tetrazolium (NBT) to photochemical decline [69].
APX activity was estimated according to a modified method [69]. Three subsamples (3 × 100 mg wet weight) were incubated in the reaction mixture (5 mL) consisting of 50 mM potassium phosphate, 0.1 mM EDTA (pH 7.8), 0.5 mM ascorbic acid, an aliquot (50 µL) of the supernatant and 0.1 mM H 2 O 2 . The alteration in absorbance was monitored at 290 nm for 1 min. APX activity (Units mg −1 protein) was measured using the extinction coefficient 2.8 mM −1 cm −1 .
MDA content was determined according to the method described in [70], in triplicate. Three subsamples (3 × 100 mg wet weight) were grounded in liquid nitrogen and homogenized after the addition of 0.1% trichloroacetic acid (TCA) and centrifuged at 15.000× g for 10 min, at 4 • C. The supernatants (1.25 mL) were combined with 3.75 mL of 0.5% thiobarbituric acid (TBA), diluted in 20% TCA, heated at 95 • C for 30 min, and centrifuged for 3 min. MDA concentration (nmol g −1 wet weight) was measured spectrophotometrically (Camspec M501, Single Beam Scanning, UV/Visible, Shimadzu, Tokyo, Japan) at 532 and 600 nm, using the extinction coefficient of 155 m −1 M −1 cm −1 .

Total Phenolic Compound Content
Phenolic compounds were analyzed in 3 freeze dried subsamples [71] s (100 mg dry weight per subsample) per concentration and day. Total phenolic content (mg g −1 dry weight) was determined in the extracts, with a modified Folin-Ciocalteu method [72,73]. An aliquot of the extract (250 µL) was combined with 2.5 mL Folin-Ciocalteu reagent (previously diluted with water 1:10 v/v), 250 µL distilled water and 2 mL sodium carbonate 0.7 M. The mixture was incubated in a water-bath for 30 min at 37 • C and the absorbance was measured at 760 nm with a spectrophotometer (PharmaSpec UV-1700, Shimadzu, Tokyo, Japan). Gallic acid was used as a standard. The Folin-Ciocaltefreu formulation, is known to be less affected by interfering compounds, and to avoid precipitation, give greater color, less variation and better recovery [71].

Microscopy
Leaf specimens or leaf cross sections were placed on microscope slides in sea water and observed either under a Zeiss AxioImager Z.2 or a Zeiss Axioplan microscope, both equipped with a MRc5 Axiocam (Zeiss, Berlin, Germany). Differential interference contrast (DIC) microscopy was applied, and imaging acquisition was performed with the AxioVision SE64 4.8.3 or the ZEN. Blue 2.3. software following the manufacturer's instructions as already reported [23].

Data Analysis
Analyses on raw and ln-transformed data indicated unequal variances and severe violation of the normality assumption. Mann-Whitney U-test analysis was carried out to determine differences in the response parameters among BPA concentrations and between BPA concentrations and the control at each incubation time; these approaches consider average values, 54 values for CTCF, and for enzyme activity, MDA, and protein contents, the three subsamples at each BPA concentration and time. The above statistical analysis was carried out using IBM SPSS ® 25.0.

Conclusions
In summary, our results demonstrated that H 2 O 2 formation was detected in intermediate C. nodosa leaves even at the lowest BPA treatment, from the beginning of the experiment, causing the antioxidant defense mechanisms to be activated, even at environmentally relevant concentrations. ROS overproduction resulted in protein content decline at the two lower BPA exposures, on the 4th day and lipid peroxidation, at 3 µg L −1 of BPA, at the beginning of the test. The enzymatic and non-enzymatic antioxidant defense mechanism (i.e., SOD and phenolic compounds) effectively scavenges excess ROS, on the 4th day, re-establishing protein content to control levels on the 6th day. Therefore, this early warning response of the protein content determines the LOEC of BPA to be 0.3 µg L −1 , and additionally, it meets the requirements for protein content to be considered as a possible early warning 'biomarker' for C. nodosa against BPA toxicity. However, it should be noted that biochemical parameters in ecotoxicological studies should be studied in combination as they are interrelated.
Author Contributions: P.M.: conceptualization and design, methodology, resources, acquisition of data, analysis and interpretation of data, formal analysis, writing original draft, revising it critically for important intellectual content, visualization, supervision. D.K.: acquisition of data, visualization, analysis of data. A.K.: acquisition of data, visualization, analysis of data. I.-D.S.A.: acquisition of data, interpretation of data, revising it critically for important intellectual content. All authors contributed extensively to its finalization. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.