Balanus amphitrite cyprids produce complex adhesive substances that enable their attachment to surfaces and impart a strong detachment resistance from most immersed substrata. The colonization of man‐made structures by barnacle cyprids and other marine organisms is a troublesome and costly phenomenon, for which controlling strategies are actively sought. In this work, we expand previous investigations about the susceptibility of cyprid adhesives to unpurified proteases in solution by evaluating the interplay between these secreted biomolecules and a surface‐confined purified protease. The strategy involved the covalent immobilization of the enzyme Subtilisin A to maleic anhydride copolymer thin films through the spontaneous reaction of anhydride moieties with lysine side chains. This enabled the production of bioactive layers of tunable enzyme surface concentration and activity, which were utilized to systematically evaluate the effect of the immobilized enzyme on cyprid settlement and exploratory behavior. Surfaces of increasing enzyme activity displayed a gradual decrease in cyprid settlement levels (approaching inhibition) as well as an increase in post‐settlement adhesion failure (evidenced by significant numbers of detached metamorphosed individuals). High activities of the bound enzyme also affected pre‐settlement behavior of cyprids, reducing the velocity and total distance moved while increasing the amount and speed of meander compared to the controls. The here‐reported low enzyme surface concentrations found to be remarkably effective at reducing cyprid settlement hold promise for the use of immobilized enzymes in the control of marine biofouling.

1. Introduction Biofouling describes the undesirable accumulation of organic material and organisms, from unicellular to invertebrate species, on man‐made surfaces.1 Biofouling is a worldwide problem affecting a multitude of industrial processes and products, including food processing, medical devices, membrane separation, cooling systems, and ships' hulls. In the marine environment alone, more than 4000 species of marine organisms are recognized as responsible for biofouling.2 Marine biofouling has been attacked from different fronts, from deterring organisms as they settle using biocides1, 3 to more elaborate and environmentally‐friendly options based on the principle of ‘non‐stick’ or ‘fouling‐release’ surfaces, which do not jeopardize marine life.4-7 Several marine organisms use proteinaceous adhesives to attach to surfaces.8 Proteolytic enzymes are effective agents against settlement of marine bacteria, algae and invertebrates, their proposed mode‐of‐action being the enzymatic degradation of the proteinaceous components of the adhesives.9-14 To date, research has focused on either the use of commercial preparations of proteases, which include up to 80% additives (such as sorbitol and calcium formate)14 or the incorporation of pure or commercial preparations into potential antifouling paints and coatings.15, 16 Where commercial preparations are employed, biological results are compromised by the spurious effects of non‐enzymatic components and, in general, by the limited knowledge of the actual surface concentration and activity of the enzyme. Hence, it is difficult to draw conclusions about the impact of the soluble or matrix‐incorporated enzyme on the observed biological response. Balanus amphitrite is a sub‐tropical sessile crustacean considered to be a serious pest due to its rapid colonization of immersed man‐made objects and its widespread geographical distribution throughout the sub‐tropics.17, 18 Barnacle cypris larvae explore a surface by ‘walking’ using their paired antennules bearing the attachment discs. In this exploratory phase, cyprids are capable of detaching, leaving behind deposits of temporary adhesive ‘footprints’17 that subsequently act as signaling molecules to induce the settlement of additional cyprids.19 On finding a suitable surface, the commitment to settlement is accompanied by the release of proteinaceous cement (cyprid permanent cement) that embeds the paired antennules and cures within one to three hours to form a discrete matrix.17 The attached individual subsequently metamorphoses and develops into the calcified adult barnacle.17 As an adult, a third, discrete adhesive is produced, which is renewable and has 90% protein content.18 The adult cement forms a thin disc between the basis of the adult barnacle and the surface to which it is attached.18 Cyprid temporary adhesive, required for exploration and settlement by the larvae, is composed primarily of protein19, 20 and is sensitive to hydrolysis by proteolytic serine proteases.9, 14 Aldred et al.9 demonstrated that the mode‐of‐action of Alcalase® (a commercial preparation containing the protease Subtilisin A) in preventing cyprid settlement was to degrade the proteinaceous temporary adhesive rather than deterring them from settling. The degradation of footprints was observed by atomic force microscopy: footprints disappeared entirely within 30 min of exposure to the enzyme. Conversely (as also observed by Pettitt et al.14), cyprid permanent cement, while initially susceptible, became resistant to attack by Alcalase® within 15 h of release onto the substratum. In this work, we aim to determine the susceptibility of the adhesives produced by Balanus amphitrite cyprids to the purified enzyme Subtilisin A bound to polymer nanocoatings. Subtilisin A (or Subtilisin Carlsberg; EC 3.4.21.62) is a well‐studied serine protease with high specificity for the hydrolysis of proteins in aqueous media.21 The extracellular alkaline protease is produced by Bacillus licheniformis, has an average molecular weight of 27,280 Da, and comprises 274 amino acids.22 The protease was covalently‐bound to poly(ethylene‐alt‐maleic anhydride) copolymer films via the spontaneous reaction of anhydride moieties with the lysine amine groups of the protein.23 Previous work showed that these bioactive layers were effective at reducing the adhesion strength of spores of the green alga Ulva linza and cells of the diatom Navicula perminuta in a manner that correlated with the enzyme surface concentration and activity.24 The present study differs from the work carried out on algae in that the functional surfaces presented to barnacle, Balanus amphitrite, cypris larvae were of reduced enzyme content, activity and thickness. Nevertheless, the immobilized enzyme was able to interfere with settlement of cyprids and their permanent fixation to the surface, whereas algal attachment was not prevented. As for algae, a correlation between the protein content of the cyprid adhesive(s) and the efficacy of the immobilized enzyme is indicated.

2. Results 2.1. Characterization of the Bioactive Nanocoatings The maleic anhydride (MA) copolymer thin films that provide a platform for the immobilization of the protease Subtilisin A have been extensively characterized regarding their physicochemical properties, both in dry and swollen states.25-28 The structure and properties of enzyme‐containing MA copolymer films were recently investigated as a function of the polymer carrier features and of the enzyme concentration used during the immobilization process.23, 24 In this work, enzyme‐containing coatings were developed that permitted a reduction in enzyme surface concentration and activity (compared to those described previously23, 24) by using enzyme concentrations in solution from 0.01 to 3 mg ml−1 for enzyme immobilization to poly(ethylene‐alt‐maleic anhydride) films. Scheme 1 depicts the layered structure of the bioactive surfaces, shows the enzyme layer thickness as a function of the enzyme concentration in solution ([Es]) used for immobilization, and includes two AFM amplitude images of the coatings of lowest and highest enzyme layer thickness. Figure 1 shows the activity and surface concentration of these bioactive layers in relation to [Es] and to the incubation time in artificial seawater (ASW*). The activity of the immobilized catalyst increased with the protein content on the surface and appeared more depleted than the latter after 48 h incubation in ASW* (e.g., for [Es] = 3 mg ml−1, the residual activity after incubation is 14% whereas the residual protein content reaches 46% ‐if the means are considered‐; similarly for [Es] = 0.25 mg ml−1). Absorbance spectroscopy measurements of the incubation solutions after 48 h revealed the absence of active enzyme in the media (lack of absorbance at 405 nm when in the presence of 50 vol% 0.2 mM Suc‐AAPF‐pNA in PBS). Furthermore, the determination of protein concentration via absorbance spectroscopy at 280 and 555 nm (back‐calculation of the protein content from the fluorophore concentration) yielded values below 200 ng ml−1, i.e. below the detection limit of the equipment (NanoDrop ND‐1000, Wilmington, Delaware, USA), suggesting the decrease in activity is likely due to denaturation processes occurring during incubation in salty media, and not to leaching. Scheme 1 Open in figure viewerPowerPoint Right panel: Schematic representation of the bioactive surfaces utilized in this work showing the enzyme surface immobilization strategy. Left panel: Thickness of Subtilisin A bound to PEMA films as a function of the enzyme concentration in solution used during immobilization. Values are mean ± SD. The two AFM amplitude images correspond to the coatings of lowest (left) and highest (right) enzyme layer thickness Figure 1 Open in figure viewerPowerPoint Activity (filled symbol) and surface concentration (unfilled symbol) of Subtilisin A bound to PEMA films as a function of the enzyme concentration in solution used during immobilization. Values (mean ± SD) with and without incubation in artificial seawater for 48 h are presented The thickness of the enzyme layer bound to PEMA films (Scheme 1) displayed an increase as the [Es] increased, which is consistent with the protein content data. Thickness was in the range 1–4 nm, indicating the presence of a monolayer (mean diameter of Subtilisin A is 4.5 nm).29 Static water contact angle measurements of non‐aged samples, confirmed the hydrophilicity of the enzyme‐containing nanolayers, with water contact angles of 30° ± 5° for the active coatings and of 45° ± 8° for the denatured ones (PEMA control = 27° ± 5°). AFM measurements of non‐aged samples yielded RMS roughness values of 0.55 ± 0.2 nm without statistical differences between active and denatured coatings (PEMA control = 0.61 ± 0.16 nm). 2.2. Cyprid Settlement Assay 2.2.1. Inhibitory and Antifouling Effect of Subtilisin A in Solution The effect of 0, 0.5, 1, or 1.5 μg ml−1 Subtilisin A in ‘Tropic Marin’ artificial seawater (ASW) on cyprid settlement onto MA surfaces after 24 and 48 h is presented in Figure 2. PEMA and poly(octadecene‐alt‐maleic anhydride) (POMA) films without bound enzyme were considered for these assays. Settlement was inhibited on POMA coatings at all [Es] and incubation times tested. Settlement levels on PEMA coatings were particularly high in the absence of Subtilisin A in solution. When exposed to 1 and 1.5 μg ml−1 enzyme in solution, settlement on PEMA coatings was significantly reduced at both 24 and 48 h compared to the control (p ≤ 0.001), whereas for 0.5 μg ml−1 enzyme, settlement was only significantly reduced at 24 h (p ≤ 0.01). Based on the observed inhibitory character of the POMA films (p = 0.82), the further evaluation of the bioactive coatings with barnacle cyprids was restricted to the use of PEMA as platform for enzyme immobilization. Figure 2 Open in figure viewerPowerPoint Mean settlement of Balanus amphitrite cyprids to POMA and PEMA films in the presence of 0 (control), 0.5, 1, and 1.5 μg ml−1 Subtilisin A in ‘Tropic Marin’ artificial seawater. Settlement is expressed as a percentage of the total number of cyprids dispensed on each sample. White bars show settlement after 24 h, dark bars after 48 h. N = 6; error bars = + 95% confidence error intervals. 2.2.2. Antifouling Potential of Immobilized Subtilisin A Barnacle cyprids of Balanus amphitrite were exposed to bioactive PEMA coatings of increasing activity and to the respective denatured controls for 24 and 48 h to determine settlement levels of these organisms on the test surfaces. The mean settlement of cyprids displayed a significant (p = 0.036) trend with respect to increasing the enzyme surface activity (Figure 3A). At both incubation times, the number of settled cyprids decreased with increasing activity; the cut‐off activity for action being of the order of 50 [pNa] min−1 103. The denatured coatings were extensively fouled at both incubation times, with no significant differences observed (p ≥ 0.121). Excluding the case of the coating of lowest activity, a significant difference in settlement between active and denatured coatings of the same enzyme surface concentration was observed after both incubation times (p < 0.05), clearly highlighting the inhibitory effect displayed by the immobilized active enzyme. Figure 3 Open in figure viewerPowerPoint A) Mean settlement of barnacle cyprids onto PEMA active coatings of increasing activity and the corresponding denatured controls. Settlement is expressed as a percentage of the total number of cyprids dispensed on each sample. Settlement on the control PEMA coating was 43 ± 9.5% after 24 h and 63.6 ± 7.5% after 48 h. B) Percentage of cyprids metamorphosed and thereafter detached for PEMA active coatings of increasing activity and the corresponding denatured controls. The percentage of detached cyprids is calculated as the relative fraction of detached individuals to the total number of settled on each surface. The percentage of detached cyprids onto the PEMA control was 0% after 24 and 48 h. White bars show settlement after 24 h, dark bars after 48 h. N = 12 (active coatings) or 6 (denatured coatings); error bars = + 95% confidence error intervals. The active immobilized enzyme not only inhibited settlement, but also caused a fraction of settled (and metamorphosed) cyprids to detach from the surface (Figure 3B). The percentage of detached cyprids (relative to the total number of settled individuals) on the active coatings was significantly higher than on the denatured controls for the two highest activities (p ≤ 0.043). Attachment of cyprids to the denatured coatings was strong: less than 1% of the attached and metamorphosed cyprids subsequently detached from these surfaces in clear opposition to the pattern observed for the active surfaces. The number of detached individuals from the active slides displayed an apparent increase with increasing enzyme surface activity, but the differences were not significant (p ≥ 0.05). However, even at the lowest activity tested (i.e. 1.3 [pNa] min−1 103), the relative number of detached individuals was higher than from the PEMA control (0% at both times). It was noteworthy that detached juveniles had a normal basis and were found to behave as normal without any sign of compromise in their health and viability (95% surviving after 14 days incubation with some being able to re‐attach to glass surfaces during the course of their maintenance). 2.2.3. Comparison Between Immobilized Enzyme and Equivalent Amount of Enzyme in Solution The amount of enzyme immobilized on the coating having the highest activity (i.e., 63 [pNa] min−1 103) compares to 1 μg ml−1 of enzyme in solution. Although the amount of enzyme is the same in both cases, the activity of the soluble enzyme is higher (750 [pNa] min−1 103) than the activity of the immobilized enzyme (essentially due to the random immobilization strategy employed). Unlike previous observations with spores of the green alga Ulva and the diatom Navicula,24 immobilizing the enzyme did not enhance the antisettlement effect compared to the equivalent amount of enzyme in solution at any of the considered incubation times (p ≥ 0.05) (Figure 4). Figure 4 Open in figure viewerPowerPoint Mean settlement of barnacle cyprids onto PEMA coatings exposed to 0 and 1 μg ml−1 Subtilisin A in solution and onto PEMA coatings with bound active and denatured (‐D) enzyme. The amount of bound enzyme is equivalent to 1 μg ml−1. Settlement is expressed as a percentage of the total number of cyprids dispensed onto each sample. White bars show settlement after 24 h, dark bars after 48 h. N = 6; error bars = +95% confidence error intervals 2.2.4. Tracking As shown in Figure 5A, surface‐bound enzyme reduced the total distance moved by cyprids, although only the sample of highest activity was significantly different from the blank PEMA control (p = 0.005). Similar effects were observed for the mean velocity (Figure 5B, p < 0.001). The mean angular velocity of cyprids increased significantly at the highest activity (Figure 5D, p = 0.000), as did their meander (Figure 5C, p = 0.000). The total turn angle, however, was not significantly different between treatments (data not shown, p = 0.326). Figure 5 Open in figure viewerPowerPoint A) Total distance moved, B) velocity, C) meander, and D) angular velocity for barnacle cyprids exposed to active and denatured enzyme coatings of different activities. Bars show means + 95% confidence error. The line connects medians.

3. Discussion Our reported study aimed at exploring the antifouling potential of a surface‐confined purified protease against Balanus amphitrite barnacle cyprids. The bioactive coatings tested had a similar structure but lower enzyme surface concentration, thickness, and activity than those employed in previous investigations with marine algae.24 Preliminary assays with barnacle cyprids revealed that changing the surface activity from 63 to 489 [pNa] min−1 103 (same activity range as tested for algae) had no influence on cyprid settlement after 48 h (see supporting information, Figure S1). This result prompted us to lower the surface activity of the immobilized enzyme to the range 1.3–63 [pNa] min−1 103, hypothesizing a dependence between surface activity and cyprid settlement could thus be made evident. The resulting bioactive coatings were essentially monolayers (thickness of the immobilized enzyme layer between 1 and 4 nm; values in the range of the average diameter of the enzyme molecule (4.5 nm)29). The layers exhibited an increase in enzyme surface concentration and activity for increasing concentrations of the enzyme in solution used during immobilization. Exposure of the bioactive layers to artificial seawater (ASW*) for 48 h decreased both activity and protein concentration on the surfaces. If any enzyme was released from the surface into the supernatant, it was inactive and hence did not contribute to the digestion of the adhesive substances secreted by cyprids during exploration and settlement. Denaturation and autolysis processes might explain these findings, which stress the need to further improve the stability of the layers, e.g. by oriented immobilization strategies. In the whole range of enzyme surface concentrations tested, nanorough, hydrophilic bioactive layers were obtained whose surface roughness and wettability were essentially independent of the protein content on the surface. This invariance in surface roughness and wettability over a range of enzyme surface activity is crucial to narrowing the spectrum of variables that might affect cyprid settlement and adhesion strength17, 30-34 and to analyze the cyprid‐surface interaction mainly in terms of the activity of the immobilized enzyme. The evaluation of the effect of increasing concentrations of enzyme in solution ([Es]) on cyprid settlement using MA coatings revealed the highly‐inhibitory character of POMA films in contrast to the highly‐inductive character of its counterpart PEMA. The inhibition of settlement observed for POMA films in the absence of enzyme seemed to have pre‐determined the results found at other [Es]. As these coatings were not leached prior to the assays, the release of inhibitory compounds from POMA coatings cannot be excluded. However, the leaching of inhibitory compounds appears unlikely because the chemicals used for the preparation of POMA surfaces were the same as for PEMA surfaces (for which no sign of inhibition was observed). The settlement inhibition found on POMA films may instead be associated with other surface‐related inhibitory features, such as their more hydrophobic character and/or different surface charge density.31-33 In a comparative assay using immobilized Subtilisin A on POMA films, in both active and denatured forms, plus POMA and acid‐washed glass controls (Figure 6), all tested POMA‐based surfaces were inhibitory to settlement regardless of whether the enzyme was immobilized to the POMA surface or not, or whether the bound‐enzyme was active or not. As with the enzyme in solution, the settlement‐inhibitory properties of the POMA layer may still have been ‘recognizable’ by cyprids, even when an enzyme monolayer (thickness = 2.7 ± 0.4 nm) was immobilized onto it in either active or denatured form. These observations illustrate the ‘determining’ effect of the polymer carrier used for immobilization, not only on the properties of the bound catalyst but also on the resulting biological response. Figure 6 Open in figure viewerPowerPoint Mean settlement of barnacle cyprids onto conditioned POMA, POMA + bound Subtilisin A with activity 59.2 ± 11 [pNa] min−1 103 (Active), its denatured control (Denatured), and acid‐washed glass. Settlement is expressed as a percentage of the total number of cyprids dispensed on each sample. White bars show settlement after 24 h, dark bars after 48 h. N = 6, error bars = + 95% confidence error intervals. For PEMA films, barnacle settlement was found to be dependent on [Es]. The strong inductive character of the PEMA control ([Es] = 0) vanished at [Es] ≈ 1 μg ml−1, revealing that relatively low [Es] are effective at deterring settlement of barnacle cyprids onto PEMA even after 48 h incubation. These findings are in agreement with previous studies by Pettitt et al.14 showing that Alcalase® (a commercial preparation whose active component is Subtilisin A) had a minimum inhibitory concentration of 1.1 μg ml−1 for barnacle cyprid settlement onto polystyrene well plates. The enzyme in solution might degrade the temporary adhesive used by cyprids during exploration, hence making it difficult for them to commit to settlement. The enzyme is also known to target the cyprid permanent cement (prior to curing), as demonstrated by Aldred et al.9 for Alcalase. Compared to previous work carried out with marine algae using immobilized Subtilisin A,24 enzyme surface concentrations and activities that proved effective against barnacle cyprid settlement were much lower, suggesting that cyprid adhesive(s) is more sensitive to the immobilized protease. Subtilisin A bound to PEMA films steadily decreased the number of settled cyprids on the surfaces with increasing surface concentration and activity. The denatured controls were all highly‐fouled, settlement effects therefore being dependent on the presence of active enzyme. An enzyme surface activity of ca. 50 [pNa] min−1 103 reduced settlement levels to ca. 5%; any further increase in surface activity had no further influence (see supporting information). The activity levels of the bound enzyme that significantly reduced cyprid settlement were substantially lower than the levels needed for Ulva linza zoospores (ca. 300 [pNa] min−1 103)24 or for Navicula perminuta diatom cells. No clear explanation for this difference in sensitivity can be offered based on available knowledge of cyprid adhesives. The cyprid ‘temporary adhesive’, involved in exploratory behavior,9, 17, 19, 20 and the cyprid cement, which fixes the larva permanently to the substratum, are proteinaceous.17 The temporary adhesive is thought to comprise a glycoprotein–the settlement‐inducing protein complex–which has 15% N‐linked glycans, very similar to the 17% N‐linked glycan35 composition reported for Ulva sp.. On the other hand, the adhesive secreted by Navicula is largely polysaccharide with minor fractions of protein.36, 37 So while a higher protein content could explain a greater sensitivity of both cyprid adhesives to Subtilisin A compared to Navicula adhesive, and of cyprid cement compared to Ulva adhesive, this explanation is unlikely to hold for the comparison of cyprid temporary adhesive to Ulva adhesive. Furthermore, recent work from Barlow et al.38 on adult cement has shown around 50% of the adhesive to be made up of amyloid‐like β‐sheet proteins. Subtilisin A is highly effective in breaking down amyloid39, 40 and while the available evidence suggests that the cyprid permanent cement and adult cement differ,9, 18, 41 the pronounced effect of Subtilisin A on cyprid permanent adhesion may point to amyloid content in cyprid cement. Together with the reduction in cyprid settlement at increasing enzyme activity, an increasing number of individuals failed to permanently attach to the bioactive layers. Individuals that committed to settlement, secreted adhesives, and initiated metamorphosis most likely had their adhesion strength weakened by the surface‐bound enzyme, which ultimately resulted in their detachment. This observation was supported by the near absence (<3%) of detached cyprids from the denatured controls. Detached individuals had a shape indicative of late stage metamorphosis, suggesting that adhesive failure may have occurred during the movements associated with molting the cyprid exuvium. The lack of macroroughness or topographic features that could have strengthened the attachment of cyprids may have also facilitated adhesive failure.30 These individuals were found to have mostly (ca. 80%) flat bases, suggesting contact with the surface during final stage metamorphosis (individuals that molt in the water column show abnormal rounded bases (SLC personal observation)). Detached individuals behaved as normal without any sign of compromise in their health or viability. The enzyme‐containing PEMA surfaces tested appear to affect the consolidation of adhesion of barnacle cyprids to the substrate without permanently interfering with earlier behaviors (i.e. commitment to settlement and metamorphosis) nor with the viability after detachment. These facts point to the non‐biocidal character of the bioactive PEMA layers, a highly‐appreciated feature in environmentally‐friendly strategies for fouling control. When comparing immobilized with soluble enzyme at equal amount, a similar antifouling response was obtained although the activity of the soluble enzyme is higher (ca. one order of magnitude) than that of the immobilized enzyme. The localization of the immobilized enzyme at the interface between adhesive and surface might have counterbalanced its lower activity levels when compared to equal amount of enzyme in solution. Possibly, the time‐scale of the settlement process in barnacle cyprids is long enough to allow the kinetically less‐favored immobilized enzyme to act effectively on the secreted adhesives and/or to hinder the mechanisms of recognition of substrate ligands required for successful settlement to occur. Finally, when evaluating the behavioral pattern of cyprids exposed to the coatings of highest and lowest surface activity (and the corresponding controls), the observed reduction in total distance moved and velocity in all tested coatings compared to the blank control (PEMA) suggests that this effect was, at least in part, due to the presence of protein on the surface (i.e. enzyme in active or denatured form). Mean angular velocity and meander were significantly increased in the highest activity coating compared to all other treatments, clearly showing that cyprid behavior was affected by the bound enzyme. This differs from the effect of soluble Alcalase reported by Aldred et al.,9 where relatively high concentrations of the commercial enzyme preparation had no effect on behavior. The increase in meander is usually due to increased searching behavior, but observations by eye suggested that some cyprids altered their swimming behavior adopting a corkscrew‐like pattern, while others were relatively inactive, which on balance may explain the reduced ‘total distance moved’. In either case, cyprids experiencing natural hydrodynamic conditions rather than the static conditions of the settlement assays, would likely return to the plankton and, as future adhesion is not compromised, could attach to a more favorable surface.

4. Conclusions The covalent immobilization of the protease Subtilisin A to PEMA copolymer films provided a suitable platform for the evaluation of the interplay between the bound enzyme and the adhesive substances secreted by Balanus amphitrite barnacle cyprids. The bound enzyme appeared to affect the consolidation of adhesion of cyprids to the bioactive layers in a manner that depends upon the surface activity. The postulated mode‐of‐action is based on the proteolytic degradation of the adhesive substances secreted upon attachment that eventually results in the detachment of already metamorphosed individuals without further effects on their health or viability. Changes in the swimming behavioral pattern of cyprids exposed to high activities of the bound enzyme were also demonstrated. Since the strong effects observed on settlement and adhesion strength of barnacle cyprids were gained by using low amounts of immobilized enzyme, the present study may prove valuable in the design of antifouling surfaces that seek to incorporate active, non‐biocidal agents.42

5. Experimental Section Reagents: Poly(octadecene‐alt‐maleic anhydride) (POMA) (MW = 30,000–50,000) was purchased from Polysciences Inc. (Warrington, USA). Poly(ethylene‐alt‐maleic anhydride) (PEMA) (MW = 125,000) was purchased from Aldrich (Munich, Germany). Subtilisin Carlsberg, or Subtilisin A (alkaline protease from Bacillus licheniformis; type VIII; E.C. no.: 3.4.21.62), N‐succinyl‐Ala‐Ala‐Pro‐Phe‐pNa (Suc‐AAPF‐pNA), citric acid, 5 M sodium hydroxide solution (for molecular biology), phosphate buffered saline (PBS) tablets, and sodium bicarbonate were purchased from Sigma‐Aldrich (St. Louis, USA). Tetrahydrofuran, sodium sulphate, and dimethylsulfoxide (DMSO) were obtained from Fluka (Steinheim, Germany). Hydrogen peroxide solution (35 vol%, not stabilized), calcium chloride, and magnesium chloride hexahydrate were purchased from Merck (Darmstadt, Germany). Ammonium hydroxide solution (28–30 wt%), toluene and acetone were obtained from Acros Organics (Geel, Belgium). Potassium chloride and sodium chloride were purchased from Riedel‐de Haen (Seelze, Germany). Ethanol and paraffin wax were obtained from VWR International (Fontenay sous Bois, France). 3‐aminopropyldimethylethoxysilane (97 vol%) was obtained from ABCR GmbH (Karlsruhe, Germany). 5‐(and‐6)‐carboxytetramethylrhodamine, succinimidyl ester (5(6)‐TAMRA, SE) mixed isomers was acquired from Invitrogen (Eugene, Oregon, USA). PD‐10 desalting columns packed with Sephadex G‐25 medium were obtained from GE Healthcare (Buckinghamshire, UK). Glass coverslips (hydrolytic class 1) were purchased from Menzel‐Gläzer (Braunschweig, Germany). Ultrasonically cleaned microscope Nexterion Glass B slides were obtained from SCHOTT (Jena, Germany). The Petri dishes (Steriplan) employed in the tracking experiments were from Duran Group (Mainz, Germany). P 96‐well plates (Rotilabo, F‐profile) were obtained from Carl Roth GmbH (Karlsruhe, Germany). Quadriperm dishes were from Greiner Bio‐One Ltd. (Stonehouse, UK). De‐ionized double distilled water was obtained from a Milli‐Q Gradient A10 water purification system (Millipore (Molsheim, France)). ‘Tropic Marin’ artificial seawater (ASW) from Tropic Marine Centre (Chorleywood, UK) was used at pH 8.0 and 3.3 w/v% salinity in all biological assays. Unless otherwise stated, all chemicals were of the highest available grades. Preparation of Bioactive Nanocoatings: Poly(ethylene‐alt‐maleic anhydride) (PEMA) and poly(octadecene‐alt‐maleic anhydride) (POMA) copolymer thin films were prepared as previously described.23, 26 Briefly, glass and silicon coverslips (for the characterization assays) or ultrasonically cleaned microscope glass slides and glass dishes (for the biological assays) were sonicated in water and ethanol for 30 min, then oxidized in a mixture of water:hydrogen peroxide:ammonia (volume ratio 5:1:1) at 70 °C for 10 min, dried at 120 °C for 1 h and subsequently modified with 3‐aminopropyldimethylethoxysilane by exposure to the compound in vapor phase overnight. The aminosilane‐modified samples were thereafter rinsed in toluene, dried at 120 °C for 1 h and spin‐coated with PEMA (0.15 wt% in THF:acetone (weight ratio 2:1)) or POMA (0.08 wt% in THF) solutions. Stable covalent binding of the maleic anhydride copolymer film was achieved by annealing at 120 °C for 2 h to generate imide bonds with the underlying aminosilane layer. Covalent immobilization of Subtilisin A onto freshly prepared PEMA copolymer films was carried out by exposing the polymer layers to varying concentrations of the enzyme in solution ([Es]). Enzyme immobilization proceeded as disclosed in previous work23, 24 although departing from much lower enzyme concentrations (Subtilisin A was dissolved in PBS, pH = 8.6 (adjusted with 1 and 5 M sodium hydroxide solutions) at concentrations between 0.01 and 3 mg ml−1). The reduction in the enzyme concentrations used for immobilization was motivated by the lack of a settlement trend with regard to enzyme surface activity in the same activity range considered for marine algae (Figure S1, Supporting Information). After overnight exposure to the enzyme solution, samples were rinsed 10 times with distilled water and were thereafter either used for assays or denatured. Denaturation of the active enzyme‐containing coatings was achieved by heating the samples at 120 °C for 45 min. Characterization of the Bioactive Nanocoatings: The characterization of the bioactive coatings utilized in the biological assays with barnacle cyprids followed Tasso et al.23, 24 and involved the determination of immobilized protein amount, enzyme layer thickness, catalytic activity, stability upon incubation in aqueous media (aging), and surface roughness and wettability. Briefly, the thickness of the immobilized enzyme layer was determined by single‐wavelength ellipsometry (SE400, Sentech Instruments GmbH, Berlin, Germany) of freshly‐prepared coatings (to avoid dehydration) using a five‐layer model approximation (silicon/silicon dioxide/maleic anhydride bound to aminosilane/enzyme/air) and a refractive index for the enzyme layer of 1.37543 (N = samples per condition = 4; m = measurements per sample = 6). The surface concentration (i.e. amount of immobilized enzyme per unit area) of aged and non‐aged coatings was evaluated by confocal Laser Scanning Microscopy (cLSM) (TCS SP5, Leica Microsystems, Wetzlar, Germany) using immobilized TAMRA‐labeled Subtilisin A and the previously‐determined surface concentration of the highest surface concentration coating (0.24 ± 0.03 μg.cm−223) as scaling factor (N = 4; m = 3). The activity of the enzyme‐containing coatings was determined by following the conversion of the substrate Suc‐AAPF‐pNA into peptides and phenylnitroaniline (pNa) through absorbance spectroscopy (TECAN Magellan GENios, Tecan, Austria) at 405 nm (N = 5; m = 1). The stability of the enzyme‐containing coatings was assessed by evaluating the residual activity and surface concentration of the bioactive layers after incubation in artificial sea water (ASW*) for 48 h at room temperature (see a previous report24 for the composition of ASW*). Atomic Force Microscopy (AFM) (MFP‐3D, Asylum Research, Santa Barbara, USA; scan size = 4 μm) in tapping mode was employed to determine the surface root mean square (RMS) roughness of the coatings as freshly‐prepared (N = 2; m = 2). Static water contact angles (OCA30, Dataphysics Instruments GmbH, Filderstadt, Germany) of active and control coatings were measured in air using de‐ionized water and utilized as an indicator of the surface wettability (N = 4; m = 3). All characterization experiments (except for AFM) were run twice on independent sample batches. Balanus amphitrite (syn. Amphibalanus amphitrite44) cyprids: Adult, brood stock barnacles, supplied by Duke University Marine Lab, were maintained as per Hellio et al.45 Nauplii were cultured and grown to the cyprid stage as per Hellio et al.45 Cyprids were stored for 3 days at 6 ± 1 °C in 0.2 μm filtered natural seawater prior to all assays. Settlement Assay: The settlement assays with barnacle cyprids concentrated on the effects of both soluble and surface‐bound enzyme on cyprid settlement. The general methodology of the ‘barnacle drop settlement assay’ was as described elsewhere.46 Enzyme immobilization was performed in situ and immediately before the assays to ensure the properties of the biolayers were as characterized. Assays determining the effect of soluble enzyme used the base maleic anhydride (MA) coatings exposed to the enzyme buffer (PBS; pH = 8.6) overnight. Slides were then placed into Quadriperm dishes and exposed to a 1.5 ml drop of 0, 0.5, 1 or 1.5 μg ml−1 of active enzyme in ‘Tropic Marin’ artificial seawater (ASW) with 40 cyprids in each drop. Dishes were incubated in the dark at high humidity and at 28 °C. Settlement and mortality were enumerated at 24 and 48 h. The assays with immobilized enzyme used freshly‐prepared active and denatured samples (as previously described). These slides were carefully dipped into paraffin wax along each edge of the slide to confine cyprids to the central area of enzyme immobilization. This resulted in 3 distinct areas: the central area with immobilized enzyme, the interface wax‐immobilized enzyme, and the wax area itself. Twelve (active coatings) or six (denatured and PEMA control coatings) replicates of each concentration were thus produced. Surfaces were thereafter placed in Quadriperm dishes and exposed to a 1.5 ml drop of ASW containing 40 cyprids. Dishes were treated as in the previous experiment. At the end of each incubation period (24 and 48 h), the total number of settled cyprids was determined and their location noted (i.e. central, interfacial or wax area). Any individuals that had settled on the wax were excluded from data analysis. Larvae that did not settle after the 48 h experimental period were observed for signs of abnormal behavior. During the experiments, a number of settled individuals were found to detach from the bioactive surfaces after having partially or fully metamorphosed. These detached individuals were carefully removed after the 48 h enumeration and kept for 2 weeks in ASW with two weekly changes of water and feeding (Tetraselmis suecica), at which time any dead individuals were noted. Settlement results are presented as mean percent settlement (i.e. mean number of settled cyprids expressed as a percentage of the total number of dispensed cyprids) with 95% confidence intervals. Normality of the distribution was assessed by means of an Anderson Darling normality test (Minitab v15.1.0.0, Minitab Ltd., Coventry, UK). Settlement data were analyzed for statistical differences using the Kruskal‐Wallis method and a post‐hoc Dunn's test (GraphPad Prism v5.0, GraphPad Software, San Diego, USA). Tracking Assay: Tracking using EthoVision 3.1 software (www.noldus.com) was carried out as per Marechal et al.47 Thirty 3‐day‐old cyprids were tracked for 5 min each on active layers (activities 1.3 and 63 [pNa] min−1 103) and the respective denatured and PEMA controls. Total distance moved, linear velocity, meander, total turn angle and angular velocity were analyzed. Normality of the data was evaluated using Anderson Darling (Minitab). Normal data underwent a one‐way ANOVA followed by post‐hoc Tukey's test (Minitab). Otherwise, a Kruskall Wallis test was carried out with a post‐hoc Dunn's test (GraphPad Prism v5.0).

Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements M.T. and S.L.C. contributed equally to this work. This work was funded by the AMBIO project (NMP4‐CT‐2005‐011827) under the 6th Framework Program of Research and Technological Development of the European Community and in part by an U.S. Office of Naval Research award (N00014‐08‐1‐1240). MT wishes to express her gratitude to Mr. Andreas Janke and the technical staff of the Max Bergmann Center of Biomaterials Dresden for AFM imaging and lab support, respectively. The authors are grateful to Mr. Robert Mutton for his collaboration during the execution of the biological assays, Dr Nick Aldred for his help with the behavioral assays and wish to thank Prof. Dan Rittschof and Beatriz Orihuela Diaz (Duke University Marine Lab, North Carolina, USA) for the collection and supply of barnacle brood stock. The authors are equally thankful to Dr. Ana L. Cordeiro for her contribution in the initial design of the experiments and for reviewing this manuscript, as well as to Ms. Florencia Tasso for help with graphic design.

Supporting Information Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer‐reviewed, but not copy‐edited or typeset. They are made available as submitted by the authors. Filename Description adfm_201101173_sm_suppl.pdf106.3 KB suppl Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.