Sulfonamides-induced oxidative stress in freshwater microalga Chlorella vulgaris: Evaluation of growth, photosynthesis, antioxidants, ultrastructure, and nucleic acids

Sulfadiazine (SD), sulfamerazine (SM1), and sulfamethazine (SM2) are widely used and disorderly discharged into surface water, causing contamination of lakes and rivers. However, microalgae are regard as a potential resource to alleviate and degrade antibiotic pollution. The physiological changes of Chlorella vulgaris in the presence of three sulfonamides (SAs) with varying numbers of –CH3 groups and its SA-removal efficiency were investigated following a 7-day exposure experiment. Our results showed that the growth inhibitory effect of SD (7.9–22.6%), SM1 (7.2–45.9%), and SM2 (10.3–44%) resulted in increased proteins and decreased soluble sugars. Oxidative stress caused an increase in superoxide dismutase and glutathione reductase levels but decreased catalase level. The antioxidant responses were insufficient to cope-up with reactive oxygen species (hydrogen peroxide and superoxide anion) levels and prevent oxidative damage (malondialdehyde level). The ultrastructure and DNA of SA-treated algal cells were affected, as evident from the considerable changes in the cell wall, chloroplast, and mitochondrion, and DNA migration. C. vulgaris-mediated was able to remove up to 29% of SD, 16% of SM1, and 15% of SM2. Our results suggest that certain concentrations of specific antibiotics may induce algal growth, and algal-mediated biodegradation process can accelerate the removal of antibiotic contamination.

Effects on chlorophyll fluorescence. The contents of chlorophyll a were significantly decreased for C.
vulgaris treated with different concentrations of SAs except 30 and 90 mg/L SD than for the control cells at day 7 (Fig. 2). The highest chlorophyll a content in L/D groups was observed for C. vulgaris treated with 90, 30, and 10 mg/L of SD, SM1, and SM2, respectively, while the chlorophyll a content in the algae treated with 270 mg/L SD was higher than that detected in algae treated with 270 mg/mL of SM1 and SM2. The value of ΦPSII was found to significant decrease with an increase in SA-concentrations (90-270 mg/L). However, treatment with low concentrations (10-30 mg/L) of three SAs had no significant change.
Effects on biochemical parameters. The protein content significantly increased following treatment with the four concentrations of SAs as compared with BC group (Fig. 3a). The highest protein contents reported after treatment with 270 mg/L of SD, SM1, and SM2 were 168%, 219% and 204% of the control, respectively. The soluble sugar content of all concentrations of SAs remained basically unchanged compared with BC group and significantly decreased in some SA-treatments, such as 30 mg/L SD and 90 mg/L SM2 (Fig. 3b).
With increasing SA-concentrations, the SOD activities increased, reaching the highest values at 10 mg/L SOD for SD, but at 90 mg/L SOD for SM1 and SM2 (Fig. 3c). The maximum activities of SOD were 181%, 160%, and 193% of the control, respectively. SOD activity reflected the toxic effects of 270 mg/L SAs in the order of SD < SM1 < SM2. However, CAT activity decreased following antibiotic treatment (Fig. 3d) and GR activity was most susceptible to SM2 of the three antibiotics (Fig. 3e).
MDA did not change significantly under the four different SD exposure levels and from 0 to 30 mg/L of SM1 and SM2 exposure levels, but with higher concentration of SM1 and SM2 it decreased (Fig. 3f). The content of H 2 O 2 in C. vulgaris increased upon treatment with all concentrations of SAs (Fig. 3g). The anti-·O 2 − activity was significantly increased in the algae treated with 10 and 30 mg/L of SD as compared with those treated with the same concentrations of SM1 and SM2 (Fig. 3h).
Effects on ultrastructure. The damage caused to the ultrastructure could be visualised with TEM for the cells from BC group and L/D groups cultivated in the presence of 270 mg/L SAs (Fig. 4). We observed more single cells in BC group than in L/D groups (see Supplementary Fig. S1). A typical cell from BC group showed a rigid cell wall, while the cell walls showed wrinkles and separated from cell membranes in the cells from L/D group. The lobed chloroplast showed a neat distribution and was located at the periphery. Thylakoids appeared mostly in three-or two-layered stacks throughout the chloroplast. In comparison, the chloroplasts appeared irregular and the thylakoids appeared disordered following L/D groups, suggestive of the SA-mediated inhibition of photosynthetic in the algae. We also observed significant changes in mitochondria. The mitochondrial size was larger in L/D groups than in BC group.
Effects on nucleic acids. DNA damage observed in the cells from BC group and L/D groups containing 270 mg/L SAs are shown in Fig. 5. The cells from BC group had a mean tail DNA value of 2.1%. The mean values of tail DNA were 28.6%, 30.1%, and 32.5% in algae treated with SD, SM1, and SM2, respectively.

Discussion
Cell biomass and μ value are the commonly used parameters to quantify algal growth 14,15 . These antibiotics at some concentrations range (i.e., 30 and 90 mg/L SD and SM1) may serve as carbon sources to promote algal growth, which is consistent with the previous reports 16,17 , wherein 5 and 1 mg/L levofloxacin improved the dry cell weight of C. vulgaris and Scenedesmus obliquus, respectively. Wang et al. 18 found that the recovery capacity of algal species was consistent with the decrease in the growth inhibition effects during incubation, as observed following SD treatment in the present study. EC 10 is a commonly used measurement endpoint to evaluate toxicity in ecological risk assessment and ecotoxicology 19 . These results are consistent with the reported for C. vulgaris,    www.nature.com/scientificreports www.nature.com/scientificreports/ which showed low sensitivity to SAs 20 , although SD (72 h EC 50 : 2.19 mg/L) and SM1 (72 h EC 50 : 11.9 mg/L) for Scenedesmus vacuolatus were slightly more potent 21 . Taken together, the results of algal biomass, μ, and 96 h EC 10 value used as the comprehensive factors to determine the phytotoxicity of pollutions to algae 22,23 indicated that C. vulgaris was more sensitive to SM2, followed by SM1 and SD.
Chlorophyll a fluorescence is a sensitive parameter under environmental stress conditions, acting as an indicator of energy conversion in photosynthetic organisms, and it provides a relatively strong signal for Chlorophyll a concentration 24 , which has been used as a potential biomarker for the evaluation of antibiotic stress 25 . Chlorophyll biosynthesis is altered upon exposure to pharmaceuticals, resulting in the inhibition of algal growth 26 . In the present study, the decrease in chlorophyll content was consistent with algal growth inhibition (Figs. 1 and 2), which may be attributed to the reactive oxygen species (ROS)-mediated damage to the photosystem and chlorophyll biosynthesis in algae 27 . The relatively high chlorophyll content of cells may serve as a protective mechanism to scavenge the accumulated ROS in chloroplasts 28 . Thus, the scavenging efficiencies of C. vulgaris for ROS induced by SAs were in the order of SD > SM1 > SM2. The ΦPSII is considered as a main indicator for assessment of the photosystem II efficiency and for the photosynthetic capacity of algae 18 . In the present study, the value of ΦPSII was significantly inhibited at high doses (90-270 mg/L) at day 7, which may be ascribed to the disturbance of chlorophyll synthesis under stress of high levels of SAs as chlorophyll a is particularly vital for the functioning of PS II 29 .
The result of the contents of protein increased in response to the treatment of microalgae with antibiotics as compared with control microalgae was also observed by Aderemi et al. 30 . This observation may be attributed to the increase in enzyme synthesis or other energy-producing fractions and a corresponding increase in energy expenditure 30,31 . Soluble sugar plays an important role in carbon partitioning, photosynthesis, and osmotic homeostasis 32 . Any decrease in the soluble sugar content may be associated with reduced photosynthesis, inhibition of cell division, and osmotic imbalance 33 . SOD neutralises ·O 2 − as a primary scavenger and catalyzes its conversion to H 2 O 2 , which is eventually scavenged by CAT to H 2 O 34, 35 . The fact that SM2 induced SOD activity suggests that the rates of ROS scavenging and H 2 O 2 regeneration were increased 36 . The effect was similar to that in other algae exposed to many other environmental contaminants such as herbicide 37 and dibutyl phthalate 38 , and may be interpreted as a consequence of the oxidative stress associated with the overproduction of ROS 39 . On the other hand, SD induced no obvious changes in the enzyme activity. CAT activity is indicative of the ability of the alga to ameliorate toxicity under stress condition 40 . GR maintains the redox balance in the cellular environment by converting glutathione disulphide (GSSG) to glutathione (GSH) pool in the cytoplasm under stress conditions 41 . The decline in CAT and GR activities could be explained by they were activated unsuccessfully and changed digressively under various concentrations of SAs exposure, so that the alga is not fully competent to remove H 2 O 2 42 , which is a non-radical ROS involved in signalling in normal cells and causes toxicity at high concentrations 43 . The increase in GR activity may result in rise in the capacity of quenching ROS 36 , as well as the increased anti-·O 2 − activity presented in the present study, which mainly affected by non-enzymatic antioxidants which act to extinguish ROS and induce antioxidant enzyme activity 44 . As an ROS, ·O 2 − induces significant oxidative damage through LPO and is harmful to organisms at high concentrations 45 . The increase of MDA indicates oxidative stress. The activity of MDA was negatively affected by SM1 and SM2. This may be related to the rise of SOD, attributed to its strongly effects on ROS scavenging.
Through TEM analysis, we found that SAs mainly affected cellular morphology, including cell wall, chloroplast, and mitochondrion. Mitochondrial amplification may be associated with the increase in energy requirement in response to adverse environmental conditions such as nutrient or toxicant stress 46 . Eguchi et al. 20 reported that the inhibition of C. vulgaris growth in response to SAs was ameliorated following folic acid supplementation, indicating that the antibiotics inhibited folate synthesis in green algae. Thus, SAs may interfere with the normal cellular development and folic acid synthesis in the cytoplasm, mitochondrion, and chloroplast 47 .
Modifications in DNA integrity following exposure to organic pollutants may serve as early signals of potential degradation of ecosystems 48 . In the present study, the visible tail of comet and shrunken cell nucleus could be observed as per the study of Zhang et al. 49 , and indicated the moderate damage caused to algal cells. The increased nuclear fragmentation suggested that the toxic effects of SAs were in the order of SM2 > SM1 > SD. Białk-Bielińska et al. 21 showed that the higher the number of -CH 3 groups in the side R chain, the lower was the www.nature.com/scientificreports www.nature.com/scientificreports/ toxicity. This observation may be related to the difference in the pH of the medium environment and different pKa values, resulting in differences in bioavailability 21 .
Green algae may serve as mediators of antibiotic degradation in light conditions 50,51 . These results are in line with those previously reported, showing negligible contribution of abiotic factors in the removal of SAs 1 . The differences in the metabolism efficiencies of these three SAs are explained by Batista et al. 52 . The degradation of SM2 was slower than that of SM1 and SD because of the increase in the number of -CH 3 and the corresponding increase in the steric hindrance to radical addition. The biodegradation, bioaccumulation, and/or bio-adsorption may play crucial roles in the removal of SAs.

Materials and methods
Microalgal culture. Axenic strain of C. vulgaris (Code: FACHB-8) originally purchased from the Freshwater Algae Culture Collection of the Institute of Hydrobiology (FACHB-Collection), Wuhan City, China, were cultivated in 500 mL Erlenmeyer flasks containing BG-11 media 53 . Cultures were maintained in a homoeothermic incubator at 25 °C ± 1 °C under 3000 lux illumination with a light-dark period of 12:12 h. To retain the exponential growth phase, algae were aseptically transferred to fresh media every 3-4 days 30 . Deionised water at 18.25 MΩ purity was obtained using Milli-Q Gradient-A 10 Millipore water deioniser.  Table 2) purchased from Sigma-Aldrich (Steinheim, Germany) at> 98% purity were freshly prepared in BG-11 media before each toxicity test. Four concentrations (10,30,90, and 270 mg/L) of each antibiotic were used against C. vulgaris. The four test concentrations were attained by diluting a small amount of the stock solution in a geometric ratio (common ratio = 3) with the solutions of culture media for the microalgae. We used 0.1 mol/L sodium hydroxide (NaOH) as solvent to maintain pH 10 to fully dissolve SAs. Then, we added 0.1 mol/L hydrochloric acid (HCl) to adjust the initial pH 7 of the test medium 54 .
Tests were carried out according to the OECD Test Guideline 201 55 with minor modifications. To obtain sufficient biomass for the analysis of various parameters, the algal growth was performed in 250 mL Erlenmeyer flasks each containing 150 mL of test solution 3 . A specific volume of C. vulgaris culture in the exponential growth phase was diluted with a known volume of BG-11 medium autoclaved at 121 °C for 30 min to achieve a constant cell density   www.nature.com/scientificreports www.nature.com/scientificreports/ 679 nm wavelength using an ultraviolet-visible spectrophotometer (Unico UV-2800, Shanghai, China). The algal biomass was measured by algal volume based on a previously established methodology 58 . The value was converted to algal biomass (mg/L) based on the linear relationship between OD 679 and biomass calculated as follows: Cell biomass of C. vulgaris (mg/L) = 4021.9 × OD 679 − 8.6817 (R² = 0.9995) The specific growth rate (μ) was calculated by fitting the algal biomass to an exponential function using the following equation: where N 2 is the algal biomass at time t 2 and N 0 is the algal biomass at time t 0 . The chlorophyll a fluorescence and effective quantum efficiency (ΦPSII) were measured with a pulse amplitude modulated fluorometer (Phyto-PAM Walz, Effeltrich, Germany) equipped with an emitter-detector-fibreoptic unit using an irradiance of 16 μmol photons·m −2 ·s −1 PAR and by an actinic light illuminating dark-acclimated cells at intensity equivalent to the incubation light (264 μmol photons·m −2 ·s −1 PAR), respectively 59 . cell harvesting and enzyme extraction. Following incubation, algal cultures (10 mL) from BC group and L/D groups were harvested in 15 mL sterile tubes by centrifugation at 3,500 rpm for 10 min at 4 °C. The resultant pellet was resuspended in 0.01 mol/L sodium phosphate buffer (pH 7.4) and centrifuged. Homogenisation was carried out at 4 °C using a manual homogeniser and the enzymes obtained from the disrupted cells were extracted in 500 μL sodium phosphate buffer. The supernatant was collected for biochemical analysis.
Analysis of energy substances. Assays were performed using traditional methods. Optical density values were measured with a spectrophotometer. The protein content in the crude extract was determined with the Coomassie Brilliant Blue (G-250) method, as reported by Bradford 60 at 562 nm wavelength using bovine serum albumin as standard and expressed as milligram per milligram algal fresh weight (FW). Soluble sugar content was estimated as per the method described by Mccready et al. 61 at 620 nm and expressed as nanogram per milligram FW.
Analysis of antioxidant ability. Superoxide dismutase (SOD) activity was assayed by the nitro blue tetrazolium (NBT) method described by Beauchamp and Fridovich 62 . One unit (U) of SOD activity was defined as the amount of enzyme required to inhibit the rate of NBT reduction by 50% at 550 nm per milligram tissue protein. Catalase (CAT) activity was determined by measuring the initial rate of the decrease in the absorbance at 405 nm over 1 min at 30 °C in a spectrophotometric assay using hydrogen peroxide (H 2 O 2 ) 63 . One U of CAT activity was defined as the degradation of 1 μmol H 2 O 2 per second per milligram tissue protein. Glutathione reductase (GR) activity was determined at 340 nm with some modification in the microtiter plate assay described by Gutterer et al. 64 . One U of GR activity was defined as the degradation of 1 mmol nicotinamide adenine dinucleotide phosphate (NADPH) per minute per gram tissue protein.
Analysis of oxidative damage. The level of malondialdehyde (MDA), a product of liquid peroxidation, was assessed with the evaluation of thiobarbituric acid reactive substance (TBARS). TBARS formed was measured with a spectrophotometer at 532 nm wavelength and quantified as MDA equivalent using 1,1,3,3-tetramethoxypropane as standard 65 . MDA content was expressed as nmol TBARS per milligram tissue protein. H 2 O 2 level was estimated following the method described by Liu et al. 66 . H 2 O 2 binds to molybdic acid to form a complex and the absorbance of H 2 O 2 content was recorded at 405 nm wavelength and expressed as mmol per gram tissue protein. The ability of C. vulgaris to eliminate superoxide anions (·O 2 − ), which are biologically produced by NADPH oxidase in metabolic processes to protect against invading pathogens, was determined using a commercially available inhibition and generation of ·O 2 − kit (Nanjing Jiancheng Bioengineering Institute) 67 . One U of anti-·O 2 − was defined as the decrease in the value relative to inhibition observed with 1 milligram vitamin C per gram tissue protein at 550 nm.
Analysis of ultrastructure and comet assay. On day 7 of growth, two samples (10 mL) containing the highest concentrations of each antibiotic were centrifuged at 3,500 rpm for 10 min. The algal cells obtained from one sample were fixed with 2.5% glutaraldehyde and treated as per the modified method described by Lapaille et al. 68 for microstructure analysis using transmission electron microscopy (TEM, HT7700, Hitachi, Japan).
The other sample was used for the detection of DNA damage with single cell gel electrophoresis or comet assay. The protocol used was as per the method described by Tice et al. 69 with some modifications. Images were digitised using light and fluorescence microscopy with an Eclipse Ni epifluorescence microscope (Nikon) equipped with an inbuilt white and external mercury lamp light source (Model C-SHG1; Nikon). For documentation, a DS-Ri2 digital sight camera (Nikon) was used. Fifty nucleoids were evaluated from each group.

Determination of antibiotic concentration.
To determine the concentration of antibiotics in the test systems, the supernatants from samples were immediately decanted for further analysis during 7 days. Samples were filtered through 0.22 μm polytetrafluoroethylene (PTFE) filters prior to use for ultra-high performance liquid chromatography (UPLC) analysis. The instrument used was a ACQUITY UPLC H-Class system composed of a quaternary solvent manager, a sample manager FTN (Flow-Through Needle), a column heater and a photo-diode array (PDA). Empower 3 Pro software (version 2010) was used 70 . An ACQUITY UPLC BEH C18 chromatography column with dimensions 50 × 2.1 mm and particle size 1.7 µm was used for LC at a constant flow rate of 0.5 mL/min. A total of 2 μL of each sample was injected using an auto-sampler. The mobile phase comprised methanol and 1‰ formic acid in water (LC-MS grade; 10:90 v/v). The column oven temperature was Scientific RepoRtS | (2020) 10:8243 | https://doi.org/10.1038/s41598-020-65219-2 www.nature.com/scientificreports www.nature.com/scientificreports/ controlled at 40 °C. Detection was performed at 256, 258 and 262 nm, respectively. Quantitative determination was performed using external standards. Limits of quantification (LOQ) and determination (LOD) were determined by using the respective calibration solutions, the LOQ of SD, SM1, and SM2 were 0.03, 0.05, and 0.05 mg/L, respectively, the LOD were 0.01, 0.02, and 0.02 mg/L, respectively. Statistical analysis. Percent growth inhibition was calculated for the response variable μ. The effective concentrations of SAs that inhibited algal growth by 10% (EC 10 ) were calculated. Statistical differences in parameters between the BC group and L/D groups were analysed after normalisation with Shapiro-Wilk test, and homogeneity of variance was detected using Levene's test by one-way analysis of variance (ANOVA) of SPSS statistics software (version 17, Chicago, IL, USA). A value of p < 0.05 was considered significant as per Fisher's Least Significant Difference (LSD) and Tamhane post hoc tests. Comet analysis was carried out using CASP program (Casplab.com) to measure DNA migration parameters (i.e., % Tail DNA).