Impact of different enzymes on biofilm formation and mussel settlement

Enzymes have been known to impact the biofilm forming capacity. However, how the enzymes mediate the biofilm formation and macrofouling remains little known. Here, we investigated the effects of the three kinds of proteases, four kinds of glycosidases and one kind of lipase on the detachment of biofilms of Shewanella marisflavi ECSMB14101, identified biofilm total proteins response to enzyme treatments, and then tested the effects of biofilms treated with enzymes on the settlement of the mussel Mytilus coruscus plantigrades. The results showed that the cell density of bacteria in biofilms formed at different initial bacterial density were noticeably reduced after treating with all tested enzymes, and Neutrase and α-Amylase exhibited best removing efficiency of > 90%. Bacterial total proteins in S. marisflavi biofilm noticeably reduced or disappeared after treated by Alcalase. For the settlements of the mussel M. coruscus plantigrades, inducing capacities of S. marisflavi biofilm were noticeably suppressed and downregulation was > 75% at the initial density of 5 × 106 cells/cm2. Thus, the tested enzymes could effectively remove the adhered bacterial cell, inhibit the biofilm formation and finally suppress the mussel settlement. Our findings extend novel knowledge to developing eco-friendly approach to control micro- and macro-fouling.

Biofilms are mainly composed of surface-associated microbial communities and their secreted extracellular polymers, as well as some organic or inorganic particles [1][2][3] . As the initial microorganism colonies on the surface, bacteria have been proved that they could induce recruitment of many macrofouling invertebrates including mussels [4][5][6][7] . Previous studies have found the biofilm of Shewanella sp. has an obviously inducing activity for marine invertebrates 8,9 , thereby both of them can cause the harm of microfouling and macrofouling 10 .
To reduce or eliminate the harm of fouling organisms, one useful way is to inhibit or reduce the initial settlement organisms and then affect the settlement of macrofouling organisms subsequently 11 . The techniques of surface antifouling mainly using paints containing copper, tin, insecticide, oxide or chloride compounds, and some new antifouling coatings were also used currently, such as paints containing nano materials 12,13 . However, most of them could exist in the marine environment for a long time and toxicity to many marine organisms 14 . Therefore, it is essential to search environment-friendly antifouling materials. Hydrolases have many characters, such as the mass production easily, the biological degradation efficiently, the species diversity and the weak toxicity as well as they could degrade the matrix molecules. However, whether these enzymes could be used in degrading extracellular polymeric substances (EPS) on biofilms or cell wells and destroying the structure of the compounds, such as the Quorum sensing signal molecules, remains little known. This may be able to control or reducing the biofouling.
The mussel Mytilus coruscus is a typical macrofouling and commercial species in China [15][16][17][18] . It could settle on the ship, the water pipes, cooling facilities, port construction or other marine facilities to cause some fouling damages. So, it can be used as the modes of fouling organisms 7,9,[15][16][17][18] . Therefore, the present study will investigate the effect of the three kinds of proteases, four kinds of glycosidases and one kind of lipase on the detachment of biofilms of Shewanella marisflavi ECSMB14101 9,19 and then test the effects of biofilms after enzymes treated on the mussel settlement. Meanwhile, the change of biofilm proteins before and after enzymes treatment was also examined.

Results
Enzyme activities. The activities of three kinds of protease were shown in Fig. 1A-C. The activity of Alcalase showed a rising trend in the concentrations ranging from 0-6 g/L, and the higher activity about 0.44 at the concentration of 6 g/L in which the reaction was reached equilibrium (Fig. 1A). No variance of the activity was observed among various concentrations (P > 0.05, Fig. 1A). The activities of Neutrase and Papain had similar tendency of Alcalase and the suitable concentration of the two enzymes were 8 g/L and 7 g/L, respectively ( Fig. 1B and C). Similarly, no variance of the two enzyme activities was observed among various concentrations (P > 0.05; Fig. 1B and C). The activities of four glycisides were shown in Fig. 1D-G. There was no significant difference between the activities of the four enzymes at the different concentrations (P > 0.05). The activity of α-Amylase was reached equilibrium state at the concentration of 6 g/L. Meanwhile, the activities of Pectinex and Cellulase were increased gradually with the increasing of enzymes concentration, but their period of stable were not obviously (Fig. 1E, F). So, the activities of them were acquired at the concentration of 10 g/L and 8 g/L, respectively. The activities of Lysing enzyme obviously increased, and then showed stable change with the increasing concentration of enzymes (Fig. 1G). The reaction was kept in balance at 12 g/L with enzyme activity of 4.07. The activity of Lipolase was shown in Fig. 1H. No variance of the activity was observed among various concentrations (P > 0.05). Meanwhile, the activity of lipolase was changed slowly at the concentration of 10 g/L and the activity was about 3.78 at this time.

Impacts of enzymes on S. marisflavi biofilm formation. Cell densities of treated and untreated S.
marisflavi biofilms were shown in Fig. 2. The initial bacteria density of the four groups were 1 × 10 6 , 3 × 10 6 , 5 × 10 6 and 10 × 10 6 cells/cm 2 , respectively. As the results shown, in the initial concentration of 1 × 10 6 cells/cm 2 , the density of bacteria in the biofilm treated with enzymes changed significantly (P < 0.05). The treatment with Pectinex was the most obvious result with bacterial density of 3 × 10 6 cells/cm 2 and the removal rate of 78.34% (Fig. 3). However, there was no significant difference on the bacterial densities in the biofilms treated with four glycosidases (P > 0.05). The death rate was highest in the group treated with Alcalase and the rate was 50.14% (Fig. 4).
In the initial concentration of 3 × 10 6 cells/cm 2 , the bacterial densities in S. marisflavi biofilms treated and untreated with enzymes had significant differences between each other (P < 0.05, Fig. 2). The treatment effects were best by Neutrase and α-Amylase with minimum bacterial density. The corresponding removal rates were 90.9% and 90.28%, respectively (Fig. 3). In addition, the death rates were relatively high in the groups treated with Lipolase, Lysing enzymes and Alcalase. The highest rate reached 48.8% by Alcalase (Fig. 4). Significant differences of the bacterial densities between the treated and untreated groups at the initial concentration of 5 × 10 6 and 10 × 10 6 cells/cm 2 were found (P < 0.05; Fig. 2). The density was lowest in the group treated by Neutrase and α-Amylase, respectively (Fig. 2). Compared with another seven kinds of enzymes, the Alcalase had the best treatment effect and the death rate reached 45.22% and 43.36%, respectively (Fig. 4). www.nature.com/scientificreports/ Images and total proteins of bacteria on the biofilms treated by Alcalase. The images of S. marisflavi biofilm were shown in Fig. 5A. As the results shown, in all groups with initial bacterial density of 1 × 10 6 , 3 × 10 6 , 5 × 10 6 and 10 × 10 6 cells/cm 2 , after treating with Alcalase, the bacterial density was reduced obviously and the ratio of living and dead bacteria was also changed significantly. The analysis of the total bacterial proteins extracted from the biofilms treated by Alacalse were shown in Fig. 5B and Fig. S1. The image of the SDS-page shown that the changes of the total proteins were obviously and  www.nature.com/scientificreports/ some protein bands reduced or even disappeared. The gray values and the molecular weights were calculated by the software of CS Analyzer 3.0 (ATTO, Japan) ( Table 1). Analysis of the gray values indicated that three bands (No. 1,4,18) were almost disappeared with the gray value decreased more than 80% in the detected 28 bands.
The gray values decreased less than 20% were deemed as no change and the bands number of 7, 22, 23, 27, 28 were liked this. Other bands were decreased in different degree.
Mussel settlement on S. marisflavi biofilms treated with enzymes. The plantigrade settlement rate was only 18% in Glass without biofilms. In Fig. 6, in the initial concentration of 1 × 10 6 cells/cm 2 , the settlement of plantigrades in positive control group (untreated biofilms) was highest, with a rate of 38%. There was no variance between the groups of positive control and three protease treatments (P > 0.05), but the settlement of plantigrades in other groups with five enzymes had significant differences with the positive control (P < 0.05). What's more, the activities of S. marisflavi biofilms treated with α-Amylase and Pectinex decreased obviously (P < 0.05) and the settlement were 22% and 20%, respectively. In the initial concentration of 3 × 10 6 cells/cm 2 , the settlement of plantigrades in all enzyme-treated groups were noticeably reduced in comparison to untreated groups (P < 0.05), and the group treated with α-Amylase had the best inhibitory activity with a rate of 22%. When the initial concentration was 5 × 10 6 cells/cm 2 , the untreated biofilm had the highest inducing activity of 78%, which was noticeably greater than enzyme treated biofilms (P < 0.05). In the initial concentration of 10 × 10 6 cells/cm 2 , there was no variance among the biofilm groups treated with Neutrase, Pectinex, Lysing enzymes, Lipolase and untreated groups (P > 0.05). Moreover, the group treated with Cellulase had a high inducing activity of 55%.

Discussion
In the present study, the eight enzymes can reduce bacteria in the S. marisflavi biofilms formed from the different initial bacterial densities. The effects of Alcalase, Neutrase and α-Amylase were better than other enzymes on reducing bacteria. Meanwhile, the residual bacteria density and the ratio of living bacteria and death bacteria in the biofilms treated with Alcalase was lower than those treated with Neutrase and α-Amylase. These results indicated that Alcalase had an obviously effect on the death of bacteria in biofilms.  20,21 . There are two ways for enzymes acting on bacterial biofilms, one of them is to get into cells to make bacteria death directly, other way is the enzymes acting on the EPS, and all of them are aim to degrade the biofilms 22 . As the results shown, the mechanism of Alcalase maybe relate to the enzymes directly acting on the cells of bacteria which had a high death rate. Meanwhile, the results of SDS-page also proved that the enzymes could degrade some proteins in the cells which were similar to Pettitt's research 23 . The lower bacterial density and death ratio of biofilms treated by other seven kinds of enzymes maybe caused by the second mechanism as well as by combing with the two kinds of mechanisms. Further researches should be conducted by analyzing the change of EPS.
In many previous studies, enzymes have inhibitory effects on the formation of biofilms 24 , such as pathogenic bacteria 25,26 , oral microbial biofilms 27,28 and industry biofilms 14 . However, the potential of these enzymes in the marine environment is still relatively less known. The results shown that the hydrolases could inhibit the formation of S. marisflavi in marine environment and this may be due to the proteins on the biofilms, but what those proteins in present study were still unknown. Fortunately, the commonly used methods of two-dimensional electrophoresis and mass spectrometry will be applied to analysis the proteins.
Enzymes had obviously inhibition effects on the settlement of some invertebrates 22,23,29 , such as barnacles. That maybe due to the QS was hydrolyzed by enzymes or produced some chemicals that could suppress recruitment of benthic animals 29 . The present study had demonstrated that enzymes could not only affect the formation of bacterial biofilms but also further inhibit the mussel recruitment. Based on the mechanism of release coating 12,[30][31][32][33][34] , the enzymes with inhibitory effect, could be mixed in various coatings and then applied for marine antifouling. But the complicated marine environment with the change of temperature 35 , salinity or pH 36 may affect the activity of enzymes or cause the inactivation of some enzymes 37 . The retention time of the enzymes in the coating determined by the above conditions. In addition, the enzymes could also affect other creatures in short time in the marine environment which may affect the balance of the whole ecosystem 38 . Thus, it is very essential to develop more eco-effective materials, such as enzymes.
In conclusion, the results confirmed that the effect of detaching biofilms of S. marisflavi were related to the kinds of enzymes and the density of bacteria on biofilm. Simultaneously, the change of bacterial biofilm may lead to the variance in the settlement of plantigrades. Therefore, the study of enzymes was important in the fields of microfouling and macrofouling.

Materials and methods
Enzymes and enzymatic activities. In the present study, the enzymes were bought from Sigma or Novozymes, three kinds of proteases, four kinds of glycosidases and one kind of lipase were examined and some characteristics of those enzymes were shown in Table 2.
All the experiments for determining the enzyme activity were conducted in autoclaved filtered sea water (AFSW) with pH = 8.15 at 18 °C. The concentration of each enzyme with the maximum activity was used to treat the S. marisflavi biofilms. Protease activity was measured by the method with a minor modification 39 . In details, as substrate, the 2% (w/v) of azocasein was placed in AFSW at 18 °C, and then protease was applied to this substrate at concentrations (0.5-10 g/L) for different incubation times (3-12 min). The reaction ended through addition of 10% trichloroacetic acid (TCA), and then took a suitable amount of supernatant liquid after centrifuging for 30 min at 3500 rpm to react with Folin-Ciocalteu's phenol Reagent for 20 min under the condition of water bath www.nature.com/scientificreports/ at 18 °C. The casein concentration was determined spectrophotometrically (UNIC 2100spectrophotometer) at 450 nm. The blank control group changed the order of adding protease and TCA solution. Protease activities were evaluated as the amount of enzyme required to generated one μmol casein per minute. Glycosidase activity was determined via the amount of released sugars during the hydrolysis of the appropriate substrate by enzymes 11,40,41 . The reaction was started with the mixture enzyme liquid at 0-18 g/L in phosphatecitrate buffer (pH: 8.15), and was then added 12.5 g/L substrate in AFSW at 18 °C for incubation. The soluble starch, pectin, carboxymethy cellulose and β-glucan were as substrates of α-Amylase, Pectinase, Cellulose and Glucanase respectively. The blank control group changed the order of adding substrate and DNS solution. Enzymatic activity was determined via enzyme content required to produce 1 μmol sugar min −1 .
Lipase activity was determined by using Lipase Activity Assay Kit. The concentrate of Lipase was diluted with AFSW to tested concentrations (1-12 g/L), and then acted on the substrate with the concentration of 465.35 μmol/L in AFSW at 18 °C (pH: 8.15) for incubation. Finally, the absorbance was detected at 420 nm. Enzymatic activity was defined as the amount of enzyme required to produce one μmol sugar per minute. The bacteria for the formation of biofilm. Stock cultures of S. marisflavi ECSMB14101 9,19 , were obtained from the storage bacteria from our laboratory. This species was isolated and purified by the ZoBell 2216E agar plate, which came from the natural biofilms in the sea, Zhoushan Zhejiang province. Previous studies have found that this species has an obviously inducing activity of M. coruscus recruitment 9,42,43 .
Formation of S. marisflavi biofilms and treated with enzymes. The formation of monospecific bacterial biofilm according to our methods 9,17,44 . S. marisflavi was cultured in ZoBell broth. The bacterium S. marisflavi were resuspended in AFSW to form the bacteria concentrated solution. The bacteria solution and AFSW into sterile glass Petri dishes with a half of a sterile glass microscope slide (38 mm × 26 mm) to form the bacteria biofilm with varying initial bacterial density at 18 °C for two days. Six replicates of each density were conducted.
The S. marisflavi biofilm on the glass slide was washed to clean unattached cells up and then immersed into Petri dishes with enzymatic preparation (20 mL) at the desired concentration for 1 h. Then, they were rinsed three times with AFSW gently for the later experiments.
Mussel settlement bioassays. Plantigrades of 10 individuals were kept on glass Petri dishes containing an untreated or treated S. marisflavi biofilm as well as 20 mL of AFSW. The blank controls used the sterile glass slides. The settlement inducing activity of biofilm was determined by settlement rate of plantigrades on S. marisflavi biofilms after 6 h, 12 h, 24 h, 48 h.
The detachment of adhered bacteria and determination of bacterial cell survival and density. The detachment of adhered bacteria by enzyme treatments was according to a previous method 11 and bacterial survival and density were described by Yang et al. 9 . The treated and untreated S. marisflavi biofilms were all rinsed three times as before and then stained. Amounts of viable and total bacteria on S. marisflavi biofilms were counted. Three biological replicates were set up.
Analysis of the total bacterial proteins. The treated and untreated bacterial biofilms were collected, centrifuged and washed twice with 1% PBS. After S. marisflavi cells were disrupted by sonication for 20 min at 4 °C, the mixtures were rotated by WH-986 Rotation Mixer for 2 h at 4 °C and then the total proteins in the supernatant were collected by removing the precipitation after centrifugation. The proteins were quantified using the RCDC method. The resulting protein patterns were captured and evaluated through Bio-rad Image lab software (Hercules, CA, USA).