The effects of aqueous extract from watermelon (Citrullus lanatus) peel on the growth and physiological characteristics of Dolichospermum flos-aquae

Nowadays, the increasing Dolichospermum (Anabaena) blooms pose a major threat to the aquatic environment and public health worldwide. The use of naturally derived chemicals from plants to control cyanobacteria blooms has recently received a tremendous amount of attention. This study investigates the possibility of transforming watermelon peel (WMP) into a biological resource to allelopathically inhibit Dolichospermum flos-aquae blooms. The results demonstrated that the growth of D. flos-aquae was efficiently restricted by the aqueous extract of watermelon peel (WMPAE) in a concentration-dependent manner. Cell viability decreased quickly, intracellular structural damage occurred, chlorophyll a in algal cells degraded, and photosynthesis was clearly inhibited. At the same time, the levels of reactive oxygen species in viable cells increased significantly, as did malondialdehyde levels, indicating that WMPAE elucidated strong oxidative stress and corresponding damage to D. flos-aquae. Capsular polysaccharide (CPS) levels increased in all treatment groups, which represents an adaptive response indicative of the development of resistance to WMPAE stress and oxidative damage. Despite this, WMPAE had clear inhibitory effects on D. flos-aquae. These findings provide fundamental information on an allelopathic system that could be a novel and attractive approach for suppressing D. flos-aquae blooms in small aquatic environments, especially aquaculture ponds.

Toxic cyanobacterial blooms have adverse influences on aquatic ecosystems and water quality 1,2 . Specifically, the toxins released by cyanobacteria can harm other aquatic creatures and endanger human health [3][4][5] . Dolichospermum (Anabaena), Microcystis, and Aphanizomenon are the most common species predominating cyanobacterial blooms 6 . Dolichospermum flos-aquae is one of the primary cyanobacterial species in Chaohu Lake in China 7 , with its blooms usually occurring during spring, autumn, and winter 8 . The D. flos-aquae cells can produce various toxins, such as microcystin, anatoxin-a (s), and anatoxin-a [9][10][11][12] . Recently, the frequency and persistence of these blooms have increased, posing a great threat to the water environmental quality and public health. Therefore, there is an exigent necessity to develop practical, economical, and environmentally friendly approaches for the control and elimination of cyanobacteria blooms, including those dominated by D. flos-aquae.
In recent years, methods to use plant's naturally produced chemicals to control cyanobacteria blooms have been suggested due to an increasing in-depth cognition of the allelopathic effects of plants [13][14][15][16] . The advantage of using chemicals derived from plants is that they are easy to degrade (making them environmentally friendly) and are very effective at suppressing algal growth 13,17 . The elimination and control of cyanobacteria bloom by plant-derived algaecides has been explored rather extensively for freshwater environments. Notably, barley straw has been applied in field and laboratory trials to suppress Microcystis blooms 13,18 . Aqueous extracts from Spartina alterniflora has also been utilized to control Microcystis aeruginosa blooms 16 . Furthermore, eucalyptus extract, an algicide, has the potential to control M. aeruginosa 19 . However, in comparison with Microcystis, the inhibitory effects of plant-derived chemicals on Dolichospermum spp. have yet to be extensively researched 20 . Bioassays of the effects of WMPAE on D. flos-aquae. D. flos-aquae solution (80 mL; cell density of 7 × 10 7 cells mL −1 prepared from cells in exponential growth) was transferred into axenic 250-mL Erlenmeyer flasks. Then, 24, 48, or 72 mL of WMPAE (0.012 g mL −1 ) was added with sterile BG-11 medium to reach a total volume of 160 mL. The corresponding concentrations of WMPAE were designated the 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 treatment groups. The initial algal density for each experiment was approximately 3.5 × 10 7 cells mL −1 . Cultures without WMPAE served as a control group (0 g L −1 group). Each group had three replicates. All flasks were maintained in a light incubator (64 μmol photons m −2 s −1 ) at 25 ± 1 °C under the light/dark cycle of 14 h:10 h. All culture flasks were shaken 3 times daily, and their position was varied randomly to decrease any potential light influence on the results.

Determination of algal cell density and pigments.
A microscope (Olympus CX31, Japan) and hemocytometer were used to count the cell number every 24 h, each vegetative cell was defined as a one-unit cell. Pigments were also measured every 24 h over the experimental period. Algal cells in 5 mL culture solution were condensed through centrifugation (8000× g for 10 min). The collected cells were resuspended in 5 mL cold 80% acetone, and then placed in the dark at 4 °C for 24 h to extract chlorophyll a (Chl a) and carotenoid. Finally, the contents of Chl a and carotenoid were measured and calculated using the formula described by Lichtenthaler and Buschmann 22 .
In vivo single-cell Raman spectra were also used to describe the variation in carotenoid content. Culture solutions (1 mL) were centrifuged for 5 min at 1500× g, and the cell pellets were resuspended in 0.3 mL PBS solution after washing with 0.01 M PBS solution (pH 7.2-7.4) twice. Following this, the micro-Raman spectra of D. flosaquae cells were measured on a DXR Raman Microscope (ThermoFisher, America) excited by a 532 nm laser with 7 mW at 25 ± 1 °C. The laser was focused on D. flos-aquae cells using a × 10 objective. All spectra were collected twice with an acquisition time of 5 s, and wavenumber range of 600-1800 cm −1 .
Transmission electron microscopy (TEM) assays. Culture solution (35 mL) was centrifuged for 15 min at 1500× g to collect D. flos-aquae cell samples. These were fixed and dehydrated, and ultrathin sections were prepared using the procedures depicted by Ozaki et al. 23 . Examination of the stained samples was performed on a transmission electron microscope (Hitachi 7700, Japan). Representative images of D. flos-aquae cells treated with 5.4 g L −1 WMPAE at 24 h and 48 h were retained for presentation.

Observations and measurements of cell viability.
A quantitative measurement of cell viability was performed on a 96-well microplate using CCK-8 assay (Cell Counting Kit-8, Beyotime Biotechnology, China). Firstly, the culture solutions were condensed through centrifugation (1500× g for 15 min). The collected cell pellets were further washed using 0.01 M PBS solution (pH 7.2-7.4), and then resuspended in PBS solution to obtain an algal solution with a density of 100,000 cells mL −1 . Following this, 10 μL of CCK-8 solution and 100 μL algal solution (containing 10,000 cells) were combined in a microplate well. After shaking for 5 min in the dark on a micro-vibrator, the 96-well microplate was put into an incubator for 2 h under darkness at 35 °C. Finally, a microplate spectrophotometer (Spark 10 M, Tecan, Switzerland) was used to measure the absorbance of each microplate well at 450 nm. The cell viability of D. flos-aquae in the treatment groups was quantified according to the following equation: Cell viability (%) Data analysis. All control and treatment groups were performed in triplicate, data are presented as the mean ± standard deviation (SD). Independent-sample t-tests and Shapiro-Wilk tests were implemented in SPSS software (13.0, USA) to test the normality of the data and to detect differences between the treatment groups and control group, respectively. A result with p ≤ 0.05 was considered statistically significant. The units examined for cell viability, CPS, MDA, ROS, and pigment levels were defined using the cell-counting method. The growth inhibition efficiency of the WMPAE on D. flos-aquae was estimated by:

Results and discussion
Antialgal activity of WMPAE. The growth curves of D. flos-aquae and the algal inhibition efficiency of WMPAE are presented in Fig. 1. The D. flos-aquae in the 0 g L −1 group (control) grew quickly during the cultural period, with cell density increasing from 3.5 × 10 7 at 0 h to 7.93 × 10 7 cells mL −1 at 96 h (Fig. 1A). The growth of D. flos-aquae was inhibited in all three WMPAE treatment groups, with both time-and concentration-dependent trends observed during the 96-h exposure period. Growth inhibition was obvious in all three treatment groups from 24 h onward, with inhibition efficiencies of 13%, 24%, and 54% for the 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE treatments, respectively (Fig. 1B). Pronounced algicidal effects were noted at 96 h, with inhibi-Inhibition efficiency (%) = (Mean Cell density control − Cell density treatment ) Mean Cell density control × 100. www.nature.com/scientificreports/ tions of 73%, 79%, and 95% for cells exposed to 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE, respectively (p ≤ 0.05, Fig. 1B). It has been reported that extracts from Hydrilla verticillate can suppress the growth of D. flos-aquae 27 , as can compounds from the Bacillus cereus strain 28 . Moreover, Kaminski et al. 29 suggested that Lemna trisulca could naturally eliminate D. flos-aquae. Similarly, the present results showed that the growth of D. flos-aquae could be inhibited efficiently by WMPAE. WMP is a common agricultural waste product. However, there were different watermelon varieties. So, in the pre-experiments, all the WMPs of various types available in the fruit market during winter to summer were tested and the results showed reproducible the inhibition effects (data not shown). Furthermore, our previous investigations revealed that A. flos-aquae could be suppressed by WMPAE 21 . The ability to also inhibit the growth of D. flos-aquae shows it has promise as a D. flos-aquae bloom control method.
Cell viability of D. flos-aquae. The cell viability in D. flos-aquae treated with WMPAE was determined quantitatively by a CCK-8 assay ( Fig. 2A) and observed qualitatively using fluorescence microscopy (Fig. 2B). In the 1.8 g L −1 and 3.6 g L −1 WMPAE treatment groups, cell viability showed an obvious increase over the first 2 h of exposure before subsequently decreasing. The viability in treated D. flos-aquae cells decreased to 43.7%, 13.6%, and 10.5% of the control level after 24 h of exposure to 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE, respectively. Fluorescence images were obtained after staining algal cells with FDA. In the control group (0 g L −1 WMPAE), the filaments were very long and emitted strong green fluorescence (Fig. 2B1), indicating healthy algal cells. After exposure to 5.4 g L −1 WMPAE, some cells in the filaments had lost their viability (Fig. 2B2) and the filaments became shorter as exposure prolonged (Fig. 2B3). Decreased cell viability can serve as a basic indicator of dying algal cells 30 . Some chemicals, such as gallic acid and H 2 O 2 , can lead to the loss of cell viability  Intracellular structure of D. flos-aquae. The intracellular structure of cyanobacteria could be destroyed by many allelochemicals 33,34 . Therefore, we examined the ultrastructure of D. flos-aquae cells in 0 g L −1 and 5.4 g L −1 WMPAE groups after 24 h and 48 h of exposure. Representative images taken by TEM are presented in Fig. 3. It is shown in Fig. 3A,B that the D. flos-aquae cells in the 0 g L −1 WMPAE group retained their structural integrity. The thylakoids (Th), a uniform cell wall (CW), plasma membrane (PM), photosynthetic lamellae (PL), and a normal nucleoid zone (NZ) were observed. However, the intracellular structures of D. flos-aquae cells treated with WMPAE were distinctly damaged (Fig. 3C-F). After 24 h of exposure, the intracellular structure became uneven, the NZ disappeared, and vacuolation was observed (see arrow 1 in Fig. 3C). We also observed clear disconnections between cells in the filaments (see arrow 2 in Fig. 3D), which contributed to the gradual shortening of the long D. flos-aquae filaments. Intracellular structure damage of D. flos-aquae was even prominent after 48 h of exposure, in which vacuolation (arrow 1) and pyknosis (arrow 3) were both observed (Fig. 3E,F). Moreover, the plasma membranes and cell walls were separated (arrow 4 in Fig. 3F), and the photosynthetic lamellae became blurred (arrow 5 in Fig. 3F). These phenomena demonstrate that the cell structures of D. flos-aquae were damaged by the WMPAE. Through TEM analysis, it was demonstrated that M. aeruginosa cells in pyrogallic acid experience nucleoid disintegration, pyknosis, vacuolation, and photosynthetic lamellae rupture 33 . Moreover, Chen et al. 34 reported that vitamin C could induce extensive damage of the cell walls, membranes, and thylakoid structure in M. aeruginosa. Additionally, in our previous study, WMPAE could induce intracellular structural damage in A. flos-aquae 21 . Similarly, D. flos-aquae cells also displayed structural damage after exposure to WMPAE. The intracellular structure became inhomogeneous firstly over time, along with the vacuolization and the disappearance of NZ. Pyknosis and the separation of the plasma membrane and cell wall occurred subsequently. In addition, the photosynthetic lamellae became blurred. On the one hand, these phenomena indicated that the cell structure of D. flos-aquae was damaged by WMPAE. On the other hand, the gradual changes of algal cells were considered as a process of D. flos-aquae being damaged by WMPAE, and therefore, the specific broken inclusion structures apparent in the TEM images of D. flos-aquae cells provide insight into the basic inhibitory mechanisms of WMPAE.

ROS levels and lipid peroxidation of D. flos-aquae.
Oxidative stress can lead to cellular damage through undermining the basic functioning of some molecules and intracellular balance 35 . MDA could act as a good indicator of lipid peroxidation and oxidative stress in cyanobacteria 36,37 . In the present study, MDA and intracellular ROS levels were measured to evaluate the oxidative stress in D. flos-aquae under WMPAE (Fig. 4). Treatment with WMPAE significantly accelerated MDA accumulation in cyanobacteria cells in a dosedependent and time-dependent manner (Fig. 4A). The MDA concentrations after 48 h of exposure to 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE were 3.0 times, 5.1 times, and 7.2 times, respectively, than that of the control group. Abnormal lipid peroxidation in algal cells caused by bioactive compounds have been observed by many researchers 18,38 , including an MDA level increase in M. aeruginosa after luteolin exposure 38 . Furthermore, it was also found that the MDA levels of A. flos-aquae cells 39 and M. aeruginosa cells 40 increase greatly after treatment with Sagittaria trifolia extract. In this study, the significant increase in the MDA level of D. flos-aquae was also observed during exposure to WMPAE. This result clearly demonstrates heavy oxidative stress and severe damage to cell membranes by WMPAE 37 .
In contrast to the results for MDA, the ROS levels in all treatment groups decreased gradually during the first 24 h of exposure (Fig. 4B). The enhanced MDA level (Fig. 4A) of D. flos-aquae with increasing WMPAE was a distinct indication of intracellular oxidative stress, which is commonly caused by increased ROS production 41 . Considering the large decreases observed in the cell viability of D. flos-aquae, the lower ROS levels might be due to the dramatic loss of viable cells after treatment with WMPAE 42,43 . Therefore, the ROS levels in the viable cells treated with different concentrations of WMPAE were recalculated based on the cell viability (Fig. 4C). Compared to the viable cells in the control group, ROS levels in viable D. flos-aquae cells of the treatment groups increased significantly. This result was demonstrated time-dependent effects, increasing to 172%, 386%, and 401% of the control level after 24 h of exposure to 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE, respectively. An enhanced ROS level can destroy normal metabolic function and lead to cell death 35,42 . Our quantitative results showed that in viable cells, both ROS and MDA levels markedly increased, indicating that the WMPAE can induce oxidative stress in D. flos-aquae cells. These findings are similar to the observations of Mecina and coworkers 18 , who reported that Hordeum vulgare extract could induce a rise in ROS levels in M. aeruginosa.

Pigments of D. flos-aquae. The WMPAE significantly affected the Chl a and carotenoid concentrations
in D. flos-aquae cells (Fig. 5). Chl a is one of the major light-harvesting pigments in cyanobacterial cells 44 . Chl a reduction in cyanobacteria induced by allelochemicals or other environmental stressors have previously been reported 45,46 . We observed an evident Chl a reduction in treated D. flos-aquae in a dose-dependent and timedependent manner (Fig. 5A). In the 0 g L −1 WPMAE group (control), the Chl a concentration was maintained steadily at 3.1 mg 10 -10 cells during the 96-h exposure period. After 72 h of exposure, the Chl a concentrations of D. flos-aquae in the 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE group reduced to 76%, 53%, and 26% of the control group level, respectively. The maximum decrease in Chl a (to 14% of the control level) was observed in the 5.4 g L −1 WMPAE treatment group after 96 h of exposure. These results imply that WMPAE affects the photosynthetic pigment of D. flos-aquae significantly, and that Chl a is susceptible to degradation under WMPAEinduced stress 47  www.nature.com/scientificreports/ www.nature.com/scientificreports/ Unlike Chl a, carotenoid concentrations in treated D. flos-aquae varied throughout the exposure period (Fig. 5B). In the 1.8 g L −1 group, the carotenoid concentration decreased slightly over 24 h; however, it exceeded that of the control group from 48 h onward, reaching a maximum value of 6.3 mg 10 -10 cells at 96 h. In the 3.6 g L −1 group, the carotenoid concentration remained stable, similar to the control group. The carotenoid concentration in the 5.4 g L −1 group decreased significantly, reaching 0.20 mg 10 -10 cells at 96 h, which represented 8.8% of the control group level. Raman spectra of D. flos-aquae cells produced by in vivo single-cell Raman spectroscopy confirmed the variations in carotenoids (Fig. 6). In the Raman spectra, 1513 cm −1 was assigned to conjugated C = C stretching vibrations, 1155 cm −1 was allocated to C-C vibrations coupled to C-CH 3 stretches or C-H in-plane bending, and 1000 cm −1 was allotted to CH 3 stretching modes, these data were considered characteristics of carotenoids 48,49 . The peak intensities were inhibited significantly in D. flos-aquae cells treated www.nature.com/scientificreports/ with 3.6 g L −1 and 5.4 g L −1 WMPAE, while those of D. flos-aquae cells treated with 1.8 g L −1 WMPAE were markedly enhanced. As carotenoids have outstanding antioxidant activities, their increase is indicative of the cells' remarkable capacity to resist oxidative stress and dissipate excited energy in non-photochemical ways 50,51 . In this study, the obvious carotenoid rises in D. flos-aquae cells treated with 1.8 g L −1 WMPAE hinted that WMPAE may bring oxidative stress and strengthen the pressure on energy dissipation in D. flos-aquae 52 . However, the carotenoid concentrations of D. flos-aquae treated by 45% WMPAE were obviously decreased, that may invalidate such a boycott mechanism at high exposure levels 52 .
Photosynthetic activities of D. flos-aquae. Changes in algal photosynthesis under stress can be reflected in chlorophyll fluorescence parameters 53,54 . The parameters Fv/Fm , F ′ v /F ′ m , and NPQ of PSII are widely used as perceptive indicators of intracellular damage in cyanobacteria [55][56][57] . Figure 7 presented the changes in PSII Fv/Fm , F ′ v /F ′ m , and NPQ of D. flos-aquae induced by WMPAE. Significant inhibition of both Fv/Fm and F ′ v /F ′ m was observed in a dose-dependent and time-dependent manner. As shown in Fig. 7A,B, the Fv/Fm and F ′ v /F ′ m of D. flos-aquae declined promptly after 4 h of exposure to 3.6 g L −1 and 5.4 g L −1 WMPAE. After 48 h of exposure, the photosynthetic yield of D. flos-aquae in the 1.8 g L −1 , 3.6 g L −1 , and 5.4 g L −1 WMPAE treatment groups was significantly inhibited by 39%, 64%, 72% (for Fv/Fm ) and 39%, 63%, 70% (for F ′ v /F ′ m ), respectively. Significant decreases in photosynthetic yield of D. flos-aquae were observed after exposure to WMPAE, this result has also been reported in previous studies 21,58 . The photosynthetic yield of PSII in M. aeruginosa could be reduced significantly by metribuzin 56 . Dramatic decreases in the PSII photosynthetic yield of Microcystis flos-aquae were also observed under the stress of erythromycin 58 . Moreover, the Fv/Fm and F ′ v /F ′ m values in M. aeruginosa were decreased by ferulic acid 59 . Additionally, the filtrate from Alexandrium minutum could intensely  www.nature.com/scientificreports/ inhibit the quantum yield of PSII in Chaetoceros muelleri 60 . Fv/Fm and F ′ v /F ′ m reflect the photosystem capacity of light conversion into chemical energy, and their reduction in D. flos-aquae cells treated by WMPAE indicates an inhibited conversion capacity 41,59 .
Non-photochemical quenching (NPQ) is a safely and non-radiatively deactivation process of excess excitation energy in PSII, which may aid in avoiding photo-damage in the photosynthetic system 44,54,61 . NPQ could be used as one of the most apposite signals of inhibitory effects due to its high susceptibility and response speed 57,62 . Compared with photosynthetic yield, the NPQ of D. flos-aquae changed in a slightly different manner after exposure to WMPAE. Specifically, a palpable reduction in the NPQ value of D. flos-aquae was only demonstrated in the 5.4 g L −1 WMPAE treatment group (Fig. 7C). Furthermore, the NPQ of D. flos-aquae in the 5.    65 . Algae adapt to stress and defend themselves by increasing their EPS content 43 . It has been reported that EPS can scavenge H 2 O 2 to further avoid oxidative damage to the algal cells 66 . Moreover, Zhang et al. 67 verified the ability of CPS in Thalassiosira pseudonana on weakening toxicity of CdSe quantum dots. We also observed a CPS rise in D. flos-aquae after WMPAE treatment (Fig. 8). The CPS contents in all treatment groups increased markedly after 24 h. The relative CPS contents of the treatment groups when compared to the control group after 72 h of exposure were 4.98 times (1.8 g L −1 WMPAE), 7.82 times (3.6 g L −1 3 WMPAE), and 11.48 times (5.4 g L −1 WMPAE) higher. These large increases in CPS contents are consistent with findings in Microcystis cells, in which two morphological Microcystis strains showed obvious increases in CPS in response to exposure to Acorus calamus root hexane extract 43 . The gradual but pronounced increase in CPS content of D. flos-aquae hints that the initial response of algal cells to WMPAE stress is to synthesize more CPS, allowing stronger interactions between its hydrophobic components and the toxic substance 66,68,69 .

Conclusions
The algacidal efficiency of WMPAE on D. flos-aquae blooms were investigated in the present study. The analyses explored several physiological and photosynthetic indexes, as well as the inhibitory mechanisms involved. Our results indicate that WMPAE can inhibit growth, decrease cell viability, induce intracellular oxidative stress, and cause intracellular damage in D. flos-aquae. Furthermore, WMPAE also degrade Chl a in algal cells and intensively impair the photosynthesis in D. flos-aquae. Carotenoid (in the low exposure level only) and CPS contents in all treatment groups were stimulated. These responses represent attempts to increase the resistance of the algal cells to WMPAE-induced stress. Overall, our results suggest that WMP, an agricultural waste product, is a promising candidate for the control of D. flos-aquae blooms, which pose environmental and human health threats. However, for successful administration more detailed research is required in the future e.g., allelochemical analyses.

Data availability
The data are available from the corresponding author on reasonable request.