Anti-algal activity of the 12-5-12 gemini surfactant results from its impact on the photosynthetic apparatus

A rapid amplification of algal population has a negative impact on the environment and the global economy. Thus, control of algal proliferation is an important issue and effective procedures which reduce algal blooms and control algal fouling are highly desired. Gemini surfactants are considered to have a low environmental impact, therefore they seem to be a promising group of detergents which could reduce algal blooms in water systems. Furthermore, due to their emulsifying properties they could replace algaecides added to antifouling paints and decrease algae adhesion to various surfaces. In this study the toxic effect of the 12-5-12 gemini surfactant was investigated on Chlorella cells and close attention was paid to a potential mechanism of its action. At the high cell density (10.05 × 107 cells/mL) a dose-dependent cell death was found and the IC50 value was reached at the concentration of 19.6 µmol/L after 72-h exposure to the surfactant. The decrease in chlorophyll autofluorescence shows that the photosynthetic apparatus seems to be the target of the tested compound. The presented studies indicate that gemini surfactants could effectively reduce algal blooms in water systems, and if added to paints, they could decrease algal growth on external building walls or other water immersed surfaces.

www.nature.com/scientificreports/ detergents, their toxicity seems to depend on the alkyl chain length. Hexadecyltrimethylammonium bromide (CTAB) and hexadecyltrimethyl-ammonium chloride (CTAC), whose alkyl tails consist of 16 carbon atoms, show stronger algaecidal activity, in comparison to dodecyltrimethylammonium bromide (DTAB) and dodecyltrimethylammonium chloride (DTAC), whose alkyl tails consist of 12 carbon atoms only 49 . Liang and coworkers indicated the strong toxic effect of CTAB on C. vulgaris cells, and 50% reduction in cell viability was shown at the concentration of 1.64 µmol/L 50 . Similar effect was obtained for Nannochloropsis gaditana cells, and the IC 50 value was reached at the concentration of 2.14 µmol/L. In the same study, IC 50 value for DTAC was reached at higher compound concentrations (34.86 µmol/L) 22 . Although in our studies IC 50 value for the 12-5-12 surfactant was reached at the concentration of approximately 20 µmol/L, initial cell density was 100-1000 times greater than that used in the studies described above. Thus, the 12-5-15 surfactant seems to show much higher algaecidal activity in comparison to DTAC and at least similar or higher, in comparison to CTAB. Thus, the obtained data indicate algaecidal activity of the 12-5-12 gemini surfactant on Chlorella cells, however the compound toxicity strongly depended on its concentration and the initial cell density. All further experiments were performed with the cell density ranging from 7.5 × 10 7 to 10 × 10 7 cells/mL.
The 12-5-12 surfactant does not seem to damage plasma membrane directly. It is accepted that surfactants destabilize plasma membrane integrity thus inducing cell death. To estimate dynamics of the 12-5-12 activity, algaecidal effects were studied after exposing Chlorella cells to its concentration of 20 µmol/L at different time points. After 2-h treatment the fraction of dead cells enlarged slightly up to 2.9%, however in comparison to the control no statistical significance was found. The number of non-viable cells increased with the prolonged incubation time to 32.4%, 49.3% and 47.7% after 24-h, 48-h and 72-h exposure, respectively (Fig. 4). The high percentage of viable cells after 2-h treatment might indicate that the 12-5-12 surfactant does not destabilize  The columns show mean values and the error bars represent a standard deviation from three independent treatments. Statistical significance was assessed with the two-way ANOVA followed by the post-hoc Tukey's honestly significant difference (HSD) test. An asterisk indicates statistically significant differences between the compound concentrations and a number sign points to statistically significant differences between the various cell densities (p < 0.05). www.nature.com/scientificreports/ plasma membrane directly and toxic effects may result from the secondary damage of plasma membrane. To validate this hypothesis the plasma membrane potential was monitored with DiOC6(3) according to the previously described procedure 51 . After 2-h exposure to the 12-5-12 surfactant most of the living cells (99.7%) exhibited green fluorescence which indicates plasma membrane hyperpolarization. However, during following hours of incubation the number of DiOC6(3) positive cells decreased in the fraction of viable cells to the level of 51.3% after 72-h exposure (Fig. 5A). At the same time a reduction of fluorescence intensity below the level of control cells was also observed (Fig. 5B). The low number of dead cells after 2-h exposure and reversible changes in the membrane potential during the following hours of treatment may indicate that the integrity of plasma membrane is not rapidly lost upon the 12-5-12 application. For example, lipoplexes of the zwitterionic phospholipid 1,2-dimyristoyl-snglycero-3-phosphatidylcholine (DMPC) and a gemini surfactants were found to have little effect on plasma membrane 52 . Furthermore, the extent of plasma membrane disorganization was significantly reduced during a treatment with gemini surfactants consisting of shorter spacers and resembled the result obtained for DTAB 35 . Thus, algaecidal activity of the tested compound may result from changes in overall cellular metabolism which then affects the stability of lipid bilayer.
The 12-5-12 surfactant reduces chlorophyll autofluorescence. Not only the integrity of plasma membrane but the physiological condition of a photosynthetic apparatus was also investigated during the compound treatment. It is assumed that the reduction in chlorophyll a autofluorescence indicates damage of the photosynthetic apparatus 53 , thus it was monitored in a population of viable cells during exposure to the 12-5-12 surfactant. The obtained data indicate that in the fraction of viable cells (Fig. 6A) chlorophyll a autofluorescence declined with the increasing compound concentrations (Fig. 6A′). Furthermore, the highest decrease in this parameter (over 60%) was obtained already after 2-h exposure to the concentration of 20 µmol/L and it persisted during the following hours of treatment ( Fig. 6B-B′). The obtained data seem to be in agreement with the previous research which indicated that detergents affect integrity of chloroplast membranes and reduce chlorophyll fluorescence in isolated photosystems [54][55][56][57] . Thus, the decrease in chlorophyll autofluorescence observed in our study right after the exposure to the 12-5-12 surfactant may indicate a dysfunction of the photosynthetic apparatus. However, one might ask why the 12-5-12 surfactant does not distinctly disrupt the plasma membrane but it affects mostly the functioning of chloroplasts?
Differences in lipid and fatty acid compositions of plasma membrane as well as of mitochondrial and chloroplast envelope could underlie various resistance reactions to the tested compound. Generally, detergent resistant membranes are rich in saturated fatty acids with a lower level of unsaturated ones 58,59 . Chloroplast lipids contain a relatively high percentage of polyunsaturated fatty acids in comparison to the other cell membranes. Furthermore, thylakoid membranes are rich in galactolipids, e.g. monogalactosyldiacyl-glycerol (MGDG) and digalactosyldiacylglycerol (DGDG). The former, due to its structure, does not favor the formation of a stable bilayer [60][61][62] . Thus, lipid composition could make chloroplast membranes more prone to detergent treatment.
It is possible that the 12-5-12 penetrates the plasma membrane without causing a serious damage of a lipid bilayer and the further loss of the plasma membrane properties results mostly from changes in cellular biochemistry triggered in response to reduced chloroplast functionality.
The 12-5-12 surfactant affects the photosynthetic apparatus. To confirm that the 12-5-12 compound affects functioning of the photosynthetic apparatus we applied an indirect approach based on the nature www.nature.com/scientificreports/ of photochemical processes in chloroplasts. Energy harvested by photosystems may be emitted as chlorophyll fluorescence, however light energy exceeding photosynthetic capacity is dissipated into heat due to activation of fotoprotective mechanisms, e.g. the non-photochemical quenching (NPQ). Both the NPQ and the photoinhibition, which can be induced in extreme cases, were found to decrease efficiency of photoautotrophy [63][64][65][66][67] . Thus, if the exposure to the 12-5-12 surfactant triggers photosystems overexcitation and photoinhibition which are followed by decrease in cell viability, lower light intensity could diminish this effect and reduce the number of dead cells.
The obtained data indicate that the percentage of dead cells decreased by about 70% at reduced light conditions during exposure to the concentration of 20 or 30 µmol/L (Fig. 7A). Interestingly, at the same time the intensity of chlorophyll a autofluorescence was slightly increased (Fig. 7B). Thus, the decrease in the number of dead cells at reduced light intensity may indicate that during exposure to the 12-5-12 surfactant, normal light conditions becomes the source of the high light stress which affects the functioning of photosynthetic apparatus and contributes to the following extensive cell death. Next, the algaecidal activity of the 12-5-12 surfactant was investigated towards cells which were grown in darkness. Unexpectedly, although dark conditions decreased the number of dead cells, the reduction was not as high as in the case of reduced light conditions. The fraction of non-viable cells declined from 50.6 to 36.1% and from 58.7 to 43.6% at the concentration of 20 and 30 µmol/L, respectively (Fig. 8). At the same time, the numbers of dead cells were also monitored at normal light conditions in the media supplemented simultaneously with the 12-5-12 surfactant and glucose to see whether it can reverse the negative impact of the reduced efficiency of photoautotrophy. The obtained data indicate that an addition of 1% glucose decreased the number of non-viable cells to the values similar to those observed at dark conditions. The fraction of dead cells declined to 32.3% and to 42.4% at the concentrations of 20 and 30 µmol/L, respectively. Interestingly, the cells treated with 20 μmol/L of the 12-5-12 surfactant showed a statistically significant difference between normal and dark conditions (Fig. 8).
The obtained data seems to be in agreement with the previous research which indicated that biosurfactants decrease the photosynthesis efficiency 68 . However, it is worth asking why darkness and glucose addition did not spectacularly reduce the number of dead cells as it was found at reduced light conditions. This observation could be explained by the fact that glucose added to media at light conditions was found to increase the level of MGDGs 69 , these of galactolipids which possibly make chloroplast more sensitive to detergents. Thus, although glucose addition diminishes the negative effect of sugar starvation caused by the 12-5-12 treatment, at the same  Supplementary Fig. 1S), which may result from intensive cell divisions 70 and the starch accumulation, respectively 71 . Interestingly, some contradictory studies indicate that the glucose addition to media may increase cell volume 71,72 . FSC and SSC signals must be interpreted with a caution, however it is still possible that autospores after hatching are less resistant to detergents due to a reduced thickness of cell wall or its specific chemical composition. Although cell wall thickness was slightly lowered in hatching autospores of Chlorella in comparison to mother cells 73 , more pronounced differences between mother and daughter cells were found in the case of Parachlorella kessleri 74 . Thus, intensive divisions in the presence of glucose might lead to a lower resistance of cell population to the 12-5-12 compound, which manifests itself in a higher than expected number of dead cells. Another possible explanation may come from mitochondria which could be key players in this case. Although mitochondria contain more phospholipids and a lower percentage of highly unsaturated galactolipids 60,75 , which makes them more resistant to detergents than chloroplast, plant mitochondria were found to fuse and elongate in darkness 76 . In mammals and yeast cells mitochondria fusion seems to be crucial for maintaining their functionality in the case of high energy demand or under stress conditions 77,78 . The 12-5-12 surfactant might have a limited effect on respiratory capacity under light conditions, on the other hand it could impact mitochondrial envelope to the extent that may reduce their capability of fusion and functionality in darkness. However, this hypothesis is not directly supported by our results and could be investigated in the future.
Nevertheless, all of this indicates that the 12-5-12 surfactant affects functioning of the photosynthetic apparatus which seems to become sensitive to normal light conditions. Cell death observed in our study may result from sugar starvation, however changes in the overall cellular biochemistry, which could be the source of a secondary damage to plasma membrane, should be taken into consideration as well.

The 12-5-12 induces reactive oxygen species (ROS).
Since damage of the photosynthetic apparatus may generate reactive oxygen species (ROS), their level was estimated during the exposure to the 12-5-12 surfactant at the concentration of 20 μmol/L. The obtained results indicate that an oxidative burst started already 2 h after the compound application and 76% of viable cells were gated as ROS positive (Fig. 9A,B). This effect was diminished when 50 µmol/L of the N-acetyl cysteine (NAC), an antioxidant, were added 1 h before ROS detection. Although NAC reduced ROS level, it simultaneously induced a considerable increase in the number of dead cells (Fig. 9B,F). During the following hours of exposure to the 12-5-12 surfactant, the percentage of viable   www.nature.com/scientificreports/ The observed alterations in the ROS level prompted us to investigate whether antioxidants present during the exposure to the 12-5-12 surfactant could reduce its algaecidal activity. Interestingly, 50 µmol/L NAC did not markedly decrease the number of dead cell, while 200 µmol/L ascorbic acid (AA) was found to reduce the fraction of non-viable cells to the levels similar to those observed after glucose addition at normal light conditions or in darkness. The impact of AA was the more visible the higher 12-5-12 concentration was used (Fig. 10). Thus, the obtained data clearly indicate that ROS participate in the 12-5-12-induced cell death. This finding is consistent with previous literature indicating that high light stress and following burst of ROS in chloroplasts may damage thylakoid lamellae and plasma membrane 79 .
In conclusion, the presented studies show the strong algaecidal activity of the 12-5-12 surfactant. The toxic effects were estimated on high cell density, indicating that the tested detergent could be a promising compound which reduces algal blooms at lower concentrations than other available agents. Our indirect approach, based on monitoring the number of dead cells at reduced light intensity, in darkness and after addition of glucose, may indicate that the 12-5-12 compound negatively affects functioning of chloroplast causing the reduction in efficiency of photoautotrophy and the increase in ROS level. Thus, the algaecidal activity of the 12-5-12 surfactant may result from two coexisting events: the reduction of photosynthesis intensity followed by a gradual deprivation  www.nature.com/scientificreports/ of available sugar and ROS-induced secondary damages to plasma membrane. The former assumption seems to be supported by the fact, that sugar addition to the media decreased the number of dead cells. The latter may be underpinned by the fact that the burst of ROS appeared right after the compound application. However, further detailed analyses are needed to estimate which processes of photosynthesis are affected by the tested compound and close attention should be paid to the environmental impact of the 12-5-12 surfactant.

Material and methods
Chemical agents.  compound was synthesized and prepared for biological investigations according to the previously described procedures 80 . Propidium iodide (PI), 3,3′-dihexyloxacarbo-cyanine iodide (DiOC6 (3) Cells treated with a very high concentration of surfactant were found to lose green pigment and these cells were used as a control to determine whether a laser 405 induces chlorophyll a autofluorescence. As indicated in supplementary materials, cells without chlorophyll which were excited with a laser 405 or 488 showed a low degree of the fluorescence intensity. In turn, cells containing chlorophyll displayed shift into a higher level of the fluorescence intensity after being excited with a laser 405 and detected with 655/8 emission filter. At the same time no increase in the fluorescence intensity was recorded after excitation with a laser 488 and detection with 610/20 emission filter. It may indicate that a detector with 610/20 emission filter does not collect a high percentage of the chlorophyll b fluorescence, which could be also excited with a laser 488. Only after the addition of PI, cells containing chlorophyll show a high level of the fluorescence intensity while excited with a laser 488 ( Supplementary Fig. 2S). Furthermore, to check that cells counted as PI positive are not false positive due to chlorophyll b excitation, cells were also investigated under a fluorescence microscope. In these cells strong red fluorescence was localized in nuclei after surfactant treatment ( Supplementary Fig. 3S).
The mean value of chlorophyll autofluorescence and the number of dead cells were estimated based on measurements of 100,000 cells. The data was exported and analyzed with FlowJo software.
Measurements of plasma membrane potential. Plasma membrane potential was monitored with DiOC6(3) according to the previously described method 32 . Aliquots of the cell suspension were stained with DiOC6(3) to a final concentration 0.25 μmol/L for 1 h. To discriminate viable and non-viable cells, 9 μmol/L PI was added to the samples 30 min prior the flow cytometry analyses. Staining was performed at room temperature in darkness. PI fluorescence was measured as described above. DiOC6(3) fluorescence was excited with a blue laser 488 nm and gathered with a 530/30 emission filter. To correct the overlap between DiOC6(3) and PI fluorescence, a fluorescence compensation was performed. For each sample 100,000 events were counted. The data was exported and analyzed with FlowJo software. The cells showing a high value of fluorescence were classified as DiOC positive ones.
Reactive oxygen species (ROS) detection. ROS were detected with the CellROX Green Reagent. The samples were incubated with the reagent at a final concentration of 5 μM for 1 h at 37 °C in darkness. Next, the cells were washed twice with PBS (pH 7.4) by centrifugation (1500g, 10 min, 4 °C). To discriminate viable and non-viable cells 9 μmol/L PI was added to the samples for 30 min prior the flow cytometry analyses at room temperature in darkness. PI fluorescence was measured as described above. CellROX Green Reagent fluorescence was excited with a blue laser 488 nm and gathered with the 530/30 emission filter. To correct the overlap between CellROX and PI fluorescence, a fluorescence compensation was performed. For each sample 100,000 events were counted. The data was exported and analyzed with FlowJo software. The cells showing a high value of fluorescence were gated as ROS positive ones. www.nature.com/scientificreports/ Statistical analyses. Statistical analyses were performed with Statistica 13 software. Normality was assessed with the Shapiro-Wilk test, and equality of variances was estimated with Brown-Forsythe test. Depending on a data set, the differences between mean values were assessed with the Student's t test, the one-way or two-way analysis of variance (ANOVA) and followed by post-hoc Tukey's test.

Data availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.